Reprinted from Ruopora, Vol. 42, July and August, 1940 CONTRIBUTIONS FROM THE GRAY HERBARIUM Se ; OF HARYARD UNIVERSITY ~ “ v Ceo CXXXI a Reprinted from Ruopora, Vol. 42, July and August, 1940 CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY CXXXI I. Studies in the Genus Hedysarum in North America. BeMent) Co OLIN oe gc ee ee i a een eS 217 II. Some Spermatophytes of Eastern North America. Oe Ai PRENAUI so ek ee 239 III. On two weedy Crucifers. Rerep C. Rouuins........ 302 Daves or Issuxr Pages 217-276 and Plates 507-605. ..........0...0..00 00.005. 15 July, 1940 Pages 281-306 and Plates 606-625.....................05. 13 August, 1940 CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY—NO. CXXXI I. STUDIES IN THE GENUS HEDYSARUM IN NORTH AMERICA REED C. Rouuins* (PLATE 597) SETACEOUS processes upon the reticulate lines or rib-like mark- ings of the loments of Hedysarwm are found in species which occur in northern Africa, certain parts of Europe and in Asia, but no species with this unusual characteristic has been recorded from America. It was, therefore, quite surprising when in 1937 plants of this genus with fruits bearing numerous setae were discovered growing in a decidedly undisturbed native habitat in the Uinta Basin of northeastern Utah. This discovery imme- diately raises the question as to whether these outgrowths upon the fruits are merely parallel developments in unrelated sections of Hedysarum or whether they indicate a close fundamental relationship between all species that possess them, despite wide geographic separation. If the latter is true, it will be agreed that a new interpretation of the geographical distribution of Hedy- sarum must be made, properly relating our new plant to those of the Old World. Since its discovery, the new plant has received intermittent attention with the view to determining its relation- ships and probable position in Hedysarum as a whole. To elucidate facts of relationship, distribution and speciation, and 1 Society of Fellows of Harvard University. 218 Rhodora [JULY in order to present a codrdinated picture of the genus as it occurs in North America south of the Yukon Territory and Alaska, a rather detailed consideration of Hedysarum has been necessary. The Yukon Territory and Alaska have been excluded from the present treatment because much of the material which ought to be studied is in the hands of Dr. Eric Hultén in Sweden and its return to America must await the termination of the present war. There are no questions involving name-priority which would be affected by specific names based on plants from this ea. De Candolle! in his monumental work upon the Leguminosae used, in part, muricate or glochidiate processes on the fruits to characterize section Echinolobium of Hedysarum and to separate it from the only other section recognized, Leiolobium. The two sections were clarified in his Prodromus? by the actual listing of the species belonging to each. That this classification was drawn along artificial lines apparently was first recognized by Basiner,’ who reclassified the genus by using an entirely new set of char- acters. The essential features of Basiner’s treatment have been accepted and used by Boissiert and more recently by Fedt- schenko® in his world-wide monograph of the genus. The latter work® has been the most valuable single reference during the course of the present study, but the treatment of American plants is not satisfactory and reflects, perhaps, an inadequacy of material upon which the monographer based his decisions regarding our species. All the American plants were placed by Fedtschenko in “subtribe” Gamotion, which supposedly contained only those species with at least the lower stipules united. Actually, of the two separable groups of species found in America, one has the lower stipules frequently partially free and those above often en- tirely free, the other has the lower stipules always united and the upper somewhat united or in rare instances nearly free. On the basis of united or free stipules alone, some of our plants would fall into “subtribe” Eleutherotion where they obviously do not 1Mem. Legumin. 345 (1825) * Prod. Syst. Nat. 2: 340-44 (1825). ; Acad. P. 6: 45-97 (1846) 4 Fl. Orientalis 2: 511— 5 Acta Hort. Petrop. 19: 1 85-325 (1902). ®I am indebted to Mr. F. J. Whitefield, one of my colleagues in the Society of Fellows, for translating several passages from the Russian 1940] —- Rollins, ~The Genus Hedysarum in North America 219 belong. In view of this fact, it appears that the use of this stipule-character, without others to support it, leads to a some- what artificial division of Hedysarum, at least in so far as the American species are concerned. Plants of Hedysarum in North America are divisable into two natural groups. One, which fits into section Obscura of Fedt- schenko, has prominently veined leaflets; articles of the loment wing-margined, surface reticulations or areolae of the articles nearly as broad as long; unequal calyx-teeth which are much shorter than the tube, and linear wing-auricles which are united under the standard and equal or exceed the claw of the wings in length. In this group belong H. alpinum, H. occidentale and H. sulphurescens. In the other group, which does not fit clearly into any of the published subdivisions of the genus, the leaflet- veins are hidden; the articles of the loments are wingless, surface reticulations or areolae are transversely oblong to rectangular, hence much longer than broad; the nearly equal calyx-teeth are linear-subulate and longer than the tube; and the wing-auricles are free, short, broad and less than a third the length of the wing- claw. Here I place H. boreale, H. Mackenzii and H. gremiale. The characters which have been used to separate these two groups are surprisingly definite and have stood the test of dozens of flower-dissections in American material. It doubtless is true that this subdivision can be employed in classifying certain Asiatic species of Hedysarum, as a cursory examination of some of them has indicated, but it is not my purpose to so extend the present investigation. As indicated above, H. boreale, H. Mackenzii and H. gremiale are not well-placed in any of the established subdivisions of Hedysarum, but according to the treatment of Fedtschenko they must be referred to section Multicaulia. Into this section both spiny-fruited and spineless-fruited species are admitted. That species with both types of fruit are sufficiently related to be Placed in the same section of H edysarum appears to be in accord with the facts, for my own studies indicate that H. gremiale is more nearly related to H. boreale var. cinerascens than to any spiny- or setose-fruited species from Africa, Europe or Asia. It must be concluded, therefore, that the discovery of a setose- fruited species of Hedysarum in America does not mean that 220 Rhodora [JULY there has been a recent genetical connection between this species and those of the Old World, but that this striking feature of the loment has probably arisen independently in America from forms without spiny fruits. Such a supposition is supported to some extent by the fact that occasional plants of H. boreale var. cinerascens tend to have muricate reticulations on the loment- segment over the seed, and, in at least one case (Tweedy no. 132 from the Tongue River in Wyoming), short but definite nubbin- like spines have been developed. I am indebted to the curators of herbaria in the following institutions who have loaned material or made facilities available for my use: Gray Herbarium of Harvard University (G) ; Forest Service, U. S. Department of Agriculture (FS); North Dakota Agricultural College (NDA) ; New York Botanical Garden (NY) ; U.S. National Museum (US). Dr. Theodor Just of the Univer- sity of Notre Dame has supplied pertinent information concerning some of Greene’s types. Cited collections followed by the symbol (R) are in my own herbarium. SYNOPSIS OF THE GENUS HEpysaruM L. 1In NortH AMERICA, EXCEPTING ALASKA AND THE YUKON TERRITORY Herbaceous perennials; stems several to numerous from & ligneous root, decumbent to erect, terete, more or less longitudi- nally grooved, usually appressed-pubescent at least above, often densely so; leaves odd-pinnate, petiolate, leaflets nearly sessile, often apiculate, mostly puncticulate above; stipules united or sometimes free, chartaceous; inflorescence racemose, axillary, peduncled; flowers erect to reflexed, pink to purple, yellowish or white; calyx bracteolate, campanulate, five-toothed, pubscent; corolla glabrous, wings and standard shorter than the keel; stamens diadelphous (9 & 1), included; fruit a loment with elliptical to suborbicular articles; single-seeded articles pubes- cent or glabrous, areolate, wing-margined or the wings absent; loments usually stipitate. Key To THE SPECIES ra the articles of rranony wing-margin as broad as long; leaflets conspicuously veined. 1940] — Rollins, —The Genus Hedysarum in North America 221 b. eels of the loment 3.5-5 n broad, narrowly wing- margined, nearly pala "e salightly oblong; flowers less than 18 mm. ong; leaflets mostly oblong to nar- rowly oblong, 4-7 (-10) mm. wide . American varieties of H. alpinum. b. — of the loment 6-13 mm. broad, conspicuously lets mostly ovate to ovate-lanceolate, 6-14 e. C. eect rae ea ipl. 15-18 mm. vatets oe glabrous ove; loments glabrous; keel truncate ..... 2. H. sulphurescens. c. Vowels ScadiaoniT Ale, 16-20 m a ook ake usually ove; loments Habieeat or very rarely Wialrons:* heel POUNCE . J ooo. acct te cee eee 3. H. occidentale. a. Auricles of the wings not united, blunt, much shorter "ibis the claw; calyx-teeth nearly equal, linear-su bulate; arti- cles of the loment wingless, areolae transversely elongated ; leaflet-veins hidden. d. Articles of the loment covered with setae; loments divari- cate; leaflets _ emia on rare OES cs Ln oo 4. H. gremiale. ’ d. Articles of the loment setae; loments oe endent; leaflets a Sher above or pubescent on es. e. Inflorescence elongated; stems branched above, 2-6 dm igh; ogee dark bro own; leaflets pubescent on both surfaces or aga peorhig usually dull slate-colored ove; Gone rs 12-19 mm. lo ong; nodes of the loment narrow, aa poe half as ; basal as the articles; articles Bg eR rei deh a ae ga Wiel esas mie 6 5. H. boreale. ee congested ; stems usually unbranched above, -3.5 dm. high; stipules whitish-translucent, brown- ercakea . teatlets green and glabrous above, rarely aS pubescent; flowers 18-21 mm. long; nodes of the loment more than half as broad as the articles; : BOCiGiee GB es el 5 Nn Cees eo eae ee 6. H. Mackenziv. 1. AMERICAN VarreTIEs oF H. atpinum. Herbaceous perennial, stems numerous, terete, longitudinally grooved, branched above, sparsely temas marspctnn 2-7 dm. high; stipules connate, rown, 5-15 mm. long, lower large and obtuse, upper acute; leaves petiolate, 618 6 m. long; leaflets 15-21, glabrous and obscurely puncticulate ahaa: sparsely pubescent (particularly along mid-vein and margins) below, prominently veined, broadly lanceolate to oblong, usually obtuse, rarely nearly acute, apicu- late, 10-25 mm. long, 5-10 _ : wide; inflorescence eo elongated, often somewhat secund; flowers deflexed, 12-18 m long, reddish-purple ; calyx oe van ar teeth unequal, 1-2 m long, shorter than the tube, upper short and eiansslie: sivas narrower and longer; standard broadly spatulate to obovate emarginate, 11-14 mm. long; wings narrowly oblong to linear, blunt, 10-13 mm. long, about 2 mm. wide, claw 2-3 mm. long, wing-auricles Gnenas united beneath standard, psa the claw; 222 Rhodora [JuLy loments mostly stipitate, glabrous to appressed- peers’, pend- ent; articles 2-5, rarely one, 5-7 mm. long, 3.5-5 mm wide, suborbicular to slightly longer than broad, wing-margined; are- olae mostly polygonal, nearly as broad as long. Key To THE AMERICAN VARIETIES OF H. ALPINUM L. Loments glabrous to very sparingly pubescent along the margins. Flowers 12-15 mm. “iad inflorescence elongated; ste ms nearly erect, 3-5 dm. high ................+-- la. var. americanum. Flowers 15-18 mm. m. long; i ree subcapitate ie some- what elongated; stems decumbent, 2-4 (—5) dm. high ib var. grandiflorum. Loments pubescent on both surfaces, rarely glabrate .. "Ie. var, philoscia. la. Var. AMERICANUM Michx. ex Pursh, Fl. Am. Sept. 2: Sr graye in RHoporA 28: 216 (1926) ; Raup in Contrib. Arn. Arb : 181 (1934) ; Bull. Nat. Mus. Can. 74: 8 (1935). H. alpinum americanum Michx., Fl. Bor.-Am. 2: “a (1803). H. alpinum sensu Marie-Victorin, Fl. Laurent. 352 (1935), non L. JH. al- pinum subsp. americanum Fedtsch. in Acta Hort. Petrop. 19; 255 (1902) in part. H. alpinum var. americanum f. albiflorum Fern- ald in Ruopora 35: 275 (1933). H. americanum Britt. in Mem. Torr. Bot. Club 5: 201 (1894) ; Britt. and Brown, Ill. Fl. 2: 311 ( Meld Rydberg, Fl. Rky. Mts. adj. Plains 524 (1917) ; Fl. Pr. Pl. Cent. N. Am. 487 (1932). H. boreale sensu De Candolle, Prod. Syst. Nat. Veg. 2: 343 (1825) in part; Hooker, Fl. Bor.-Am. 1: 155 (1834) ; cee and Gray, .N. Am. 1: 356 ( ie : Wood, British Columbia and probably Alaska. NerwrounpLanp: Bard Harbor Hill, Highlands of St. John, Aug., 1925, Fernald & Long 28627 (G); ‘Bishop Falls, valley of Exploits River, July, 1911, Fernald, Wiega nd & Darlingt on 5800 (G); Grand Falls, valley of Exploits River, July, 1911, Fernald, Wiegand & Darlington 5798 (G). QuEBEc: between Mont Louis and Riviére & Pierre, Aug., 1923, Fernald & Smith 25875 (G); Gulf of St. Lawrence, east of Marten River, Gaspé Co., July, 1922, Fernald & Pease 25171 (G); St. John River , Gaspé Co. , Aug., 1904, Collins, Fernald & Pease s. n. (G); Tourelle, Gaspé Co., July, 1924, Pierce & Hodge 7a (G); Bic, Rimouski Co., July, 1904, Collins "& Fernald s.n. (G); Aug, 1927, Rousseau 26823 (G); Renard River, Anti- Pep Island, Aug., 1927, Marie-Victorin & Rolland- Germain 354 (G) ; Natiskotek River, Anticosti Island, Aug., 1927, Marie- Lara & Rolland-Germain 27356 (G); between Baldé and Baie des Chaleurs, Bonaventure Co., Aug., ‘1904, Collins, Fernald 1940] — Rollins, —The Genus Hedysarum in North America 223 & Pease s.n. (G); Lake St. Jean, July, 1921, Marie-Victorin s. n. (G). New Brunswick: Gorge of the Aroostook River, Victoria Co., Aug., 1909, Fernald 1962 (G); July, 1902, Williams, Collins & Fernald s. n. (G); Connors, St. John River, July, 1903, Pease 2262 (G). Marne: St. John River, St. Francis, July, 1932, Pease & Goodale 67773 (G); Aug., 1893, Fernald 26 (G); Fort Fairfield, July, 1902, Williams, Collins & Fernald s. n. (G); Fort Kent, July, 1908, Mackenzie 3536 (NY). Vermont: Wil- loughby, July, 1887, EH. & C. E. Faxon s. n. (G); Aug., 1874, Congdon s. n. (G) ; Smuggler’s Notch, Mt. Mansfield, July, 1894, Eggleston s. n. (G, NY); Aug., 1877, E. & C. E. Faxon s.n. (G); July & Aug., 1877, Pringle s.n. (G). MAanrrosa: 6 miles east of Forest, June, 1906, Macoun & Heriot 70783 (G). SASKATCHEWAN: 30 miles east of Touchwood, July, 1906, Macoun & Heriot 70784 (G); Duck Lake, July, 1913, Johnson 1375 (NY). ALBERTA: Kootenai Plains, North Branch Saskatchewan River, June, 1908, G). British Cotumsta: vicinity of Hudson Hope, June, 1932, Raup & Abbe 3626 (G). 1b. Var. grandiflorum, var. nov. Herba perennis; caulibus decumbentibus, 2-4 (—5) dm. longis; floribus purpurascentibus, 15-18 (-19) mm. longis.—H. alpinum sensu Fernald in RHopora 13: 119 & 129 (1911); ibid. 28: 216 (1926); ibid. 35: 275 (1933) ; Raup in Contrib. Arn. Arb. 6: 181 (1934); non L. 4. alpinum var. americanum sensu Ostenfeld in Vidensk. Selsk. Skrift. I Klasse, no. 8, 55 (1909), non Michx. ex Pursh—Lab- rador, Newfoundland, northern Alberta and British Columbia. Laprapor: Forteau, 1870, S. R. Butter s. n. (G). Newrounp- LAND: Port & Port, July, 1921, Mackenzie & Griscom 10332 (G, US) ; Table Mountain, region of Port & Port Bay, July, 1914, Fernald & St. John 10849 (G); Cook Point, Pistolet Bay, July, 1925, Fernald & Gilbert 28622 (G); Anse aux Sauvages, Pistolet Bay, Aug. 11, 1925, M. L. Fernald, K. M. Wiegand & Bayard Long 28625 (G, rypz); west of Big Brook, Straits of Belle Isle, Long & Gilbert 28620 (G); Sandy Cove, Straits of Belle Isle, Aug., 1924, Fernald, Long & Dunbar 26810 (G) ; Killdevil, Bonne Bay, Aug., 1929, Fernald, Long & Fogg 1835 (G) ; Eastern Point, region of St. John Bay, July, 1929, Fernald, Long & Fogg 1833 ). Aperta: head of Malique Lake, July, 1908, Brown 1218 {G, NY); Cataract Cr., headwaters of the Saskatchewan and Athabasca Rivers, Aug., 1908, Brown 1452 (G); head of Smoky 224 Rhodora (JULY River, Aug., 1911, Riley 36 (G). BrivtsH ConumBia: Mt. Selwyn, July, 1932, Raup & Abbe 3967 & 4091 (G). 46. Var philoscia (A. Nels.) comb. nov. H. philoscia A. Nelson in Proce. Biol. Soc. Wash. 15: 185 (1902) ; Caalier and Nelson, Man. Bot. Cent. Rky. Mts. 300 (190 9). H. boreale sensu Rydb., Fl. a Mts. er Plains 524 (1917) ; Fl. Pr. Plains Cent. N. Am. 487 (1932); non Nuttall —Saskatchewan and Alberta to South Dakota and Wyoming. SasKATCHEWAN: with- out locality, 1858, EZ. Bourgeau s. n. Pas NY). AL perta: Fort Saskatchewan, July, 1 938, Turner 58 & 59 (G). SourH Dakota: Rochford, Black Hills, July, 1892, Rudberg 640 (G) ; Black Hills, July, 1872, Greene 13 (NY); near Custer Peak, Lawrence Co., June, 1929, Palmer 37554 (G); ; hing Pennington Co., June, 1929, Palmer 37509 (G). Wyo : Boyd, Weston Co., July, 1910, A. ie tages (G); Willow Ae Albany Co., July, 1897, A. Nelso 7 (G); Crow Cree ; any Co., ’Aug., 1903, A, “peas pref (G); Laramie Hills, ee Co., July, 1901, E. Nelson 622 (G, NY). In eastern Asia, as in America, there are several phases and varieties of H. alpinum. The exact application of Linnaeus’ name to Siberian material, except in the broad sense, has not been attempted in the present study. It is evident from a careful examination of Siberian specimens of H. alpinum in the Gray Herbarium and in the United States National Herbarium, that the plants heretofore passing as H. boreale in such works as Torrey and Gray’s Flora! and Gray’s Manual,? and as H. ameri- canum in Britton and Brown’s Flora? and Rydberg’s Flora* are not specifically distinct from those of eastern Asia. However, the North American plants do differ in certain minor ways and should be considered as separate varieties of a wide-ranging species, H. alpinum, which extends from Asia across the north to New- foundland, Gaspé and Maine and southward along the moun- tains of western America. Such a treatment indicates clearly the relationships of our plants with those of Asia, and at the same time shows that the American plants have certain special char- acteristics which are not possessed by those of the Old World. The differences separating the Old World plants from those of the New, particularly those emphasized by Hooker,® are at best 1 Fl. N. Am. 1: 356 (1838). ? Robinson and Fernald in Gray’s Manual 7th. ed. 518 (1908). 2 Til. Fl. 2: 311 (1897). ‘Fl. Rky. Mts. adj. Plains 524 (1917). 5 Fl. Bor.-Am. 1: 155 (1834). 1940] —_ Rollins, —The Genus Hedysarum in North America 225 only trivial. Certainly the actual differences are not of sufficient importance to justify specific segregation and, in my opinion, it is a mistake to obscure the natural relationships of our plants by giving them a separate specific epithet. The misapplication of the name H. boreale is dealt with under that species. H. alpinum in America has three geographic varieties which are very closely related, but which have certain characteristics peculiar to themselves. Variety philoscia is very similar to var. americanum except for its densely pubescent instead of glabrous fruits. This difference is not absolute, for there are specimens with fruits pubescent along the margins or even with a very few trichomes along the edges of the flat surfaces of the loments which I have referred to var. americanum. Plants of the latter type are apparently of rare occurrence, but they indicate that intermediates between the two varieties actually exist and that attempts to establish either as a distinct species should be re- garded with suspicion. Variety grandiflorum is a more dwarfed, larger-flowered plant than its nearly related var. americanum, and the two are usually quite easily distinguished, but here again, as far as herbarium material shows, there is a gradual transition from one to the other. Variety grandiflorum inhabits the head- lands of Newfoundland and barrens of northern Canada, while var. americanum ranges southward in more favorable habitats. Their most distinctive characters have been set forth in the key above. . H. sutpHurEscens Rydberg. Herbaceous perennial, stems several to numerous from a ligneous root, shallowly furrowed longitudinally, branched above, appressed-pubescent, dm. igh; stipules united, brown, chartaceous, lower obtuse, 1—-1.5 cm. ong, upper acute to acuminate, reduced; leaves petiolate, 8-12 em. long; leaflets 9-17, elliptical to ovate-oblong, usually apicu- late, conspicuously veined, sparsely pubescent below, glabrous and puncticulate above, 15-30 (-40) mm. long, 5-10 (—15) mm. wide; inflorescence axillary, racemose, elongated; flowers pendent, ochroleucous to yellow, 15-18 mm. long; calyx pubescent, teeth unequal, shorter than tube, 1-3 mm. long, upper shorter and broader than the lower ; standard obovate-cuneate, emarginate, 12-14 mm. long, 6.5-8 mm. wide; wings obtuse, 12-14 mm. long, : mm. wide, wing-auricles linear, united under standard, equaling the claw, 3-3.5 mm. long; keel sharply truncate; loments Pendent, stipitate, articles 1-4, conspicuously wing-margined, 226 Rhodora [JULY glabrous, asymetrically elliptical, 8-12 (15) mm. long, 6-8 (9) mm. wide; reticulations not raised, polygonal.—Bull. Torr. Bot. Club 24: 251 (1897) ; Mem. New York Bot. Gard. 1: 257 (1900) ; Piper in Contrib. U. S. Nat. Herb. 11: 367 (1906) ; Coulter and Nelson, Man. Bot. Cent. Rky. Mts. 300 (1909) ; Rydberg, Fl. Rky. Mts. adj. Plains 523 (1917). H. flavescens Coult. and Fisch., Bot. Gaz. 18: 300 ( 1893), non H. flavescens Regel and Schmalh. ex Regel in Bull. Soc. Sci. Moscow 34: 21 (1882). H. boreale Nutt. var. flavescens (Coult. and Fisch.) Fedtsch. in Bull. Herb. Boiss. 7: 256 ( 1899). H. boreale Nutt. var. albiflorum Macoun, Cat. Canad. Pl. 1: 510 (1884). H. albiflorum (Macoun) Fedtsch. in Acta Hort. Petrop. 19: 252 (1902). H. boreale Nutt. var. leucanthum sensu M. E. Jones in Proc. Calif. Acad. Sci. 5! 677 (1895), non H. Mackenzii Richards. var. leucanthum Greene, Pitt. 2: 294 (1892) —Alberta and British Columbia to Washing- ton, Idaho and Wyoming. Atperta: Pipestone Valley, July, 1906, Brown 425 (G) ; Bow River Valley, June, 1906, Brown 127 var. albiflorum). Montana: upper Marias Pass, Aug., 1883, Canby 93 (G); McDonald’s Peak, Mission Range, July, 1883, Canby 90 (G); Cutbank Creek, Glacier Nat. Park, July, 1934, G. N. Jones 5425, 5438 and 5513 (G); Mt. Haggin, near Ana- conda, July, 1915, M. E. Jones s. n. (G); Bozeman, July, 1895, Shear 5269 (US) : Bridger Mts., Gallatin Co., Aug., 1902, W. W. Jones (G); Baldy Mountain, Park Co., June, 1912, Eggleston 8079 (G) ; Pioneer, July, 1898, J. K. Uhl s. n. (G, NDA); Bear- tooth Mts., 17 miles southwest of Red Lodge, Carbon Co., July, 1939, Rollins & Munoz 2828 (G); West Fork of Sun River, Lewis and Clark Nat. For., Aug., 1912, Saunders 174 (FS) ; north slope of Pryor Mt., Beartooth Nat. For., June, 1926, Williamson 2 (FS). Wyomrne: Little Tongue River Canyon, Big Horn Mts., Sheridan Co., June, 1936, L. & R. Williams 3112 (G) ; 20 mi. west of Dayton, Sheridan Co., July, 1935, L. Williams 2364 (G); Beartooth Butte, Park Co., Aug., 1937, L. & R. Williams 3767 (G); near Cody, Yellowstone Nat. Park, July, 1930, Churchill s. n. (G); Wraith Falls, Yellowstone Nat. Park, July, 1899, A. & E. Nelson 5706 (G). Ivano: south end of Lake Pend d’Oreille, July, 1892, Sandberg, MacDougal & Heller 748 (G); Targhee Nat. For., Aug., 1911, Willey 161 (FS). WasHINGTON: * near Winthrop, Okanogan Co., July, 1934, Thompson 10913 (G); 1940] Rollins,—The Genus Hedysarum in North America 227 road to Salmon Meadows, Okanogan Co., June, 1931, Thompson 7024 (G). Fedtschenko! expressed doubt concerning the ultimate validity of H. sulphurescens (H. albiflorum) as a specific entity, stating that it was very close to H. alpinum var. japonicum and an un- designated variety of H. obscurum. Recently, Hara? named var. japonicum, in the sense of Fedtschenko, as a species, which seems to indicate that the relationship is not as close as originally supposed. A careful study of the Asiatic plants referred to, shows that they are in the same species-group, but that they are not conspecific with the American plants. Our plants belong to the “alpinum” group, but are actually most closely related to H. occidentale on account of the large, widely wing-margined loments. Besides having yellowish instead of reddish-purple flowers, H. sulphurescens differs from H. occidentale in having glabrous instead of pubescent fruits, smaller flowers and a more sharply angled truncate keel. These two species probably had a common origin, but in my opinion, they have now developed differentiating characters which are constant enough to merit for each the rank of a species. H. sulphurescens occupies a unified phytogeographical area in the northern Rocky Mountains and adjacent ranges. 3. H. occipeNTALE Greene. Herbaceous perennial, stems sev- unequal, 1—-3.5 mm. long, upper short, triangular, lower nearly subulate, shorter than the tube; standard obovate-spatulate, emarginate, 13-15 mm. long, 6.5-7.5 mm. wide at widest point; Wings 13-15 mm. long, 2-3 mm. wide, linear-oblong, wing- auricles united under the standard, linear, equaling the claw, mm. long; loments pendent, stipitate; articles 1-4, elliptical, conspicuously wing-margined, pubescent or rarely glabrous, 9-14 mm. long, 7-13 mm. wide, reticulations polygonal.—Pitt. 3: 19 ? Fedtschenko, op. cit. p. 253. * Journ. Jap. Bot. 15: 52 (1939). 228 Rhodora [JULY (1896); Piper in Contrib. U. 8S. Nat. Herb. 11: 366 (1906) ; Piper and Beattie, Fl. Northw. Coast 225 (1915); G. N. Jones in Univ. Wash. Pub. Biol. 5: 188 (1936). H. marginatum Greene, Pitt. 4: 138 (1900); Rydberg, Fl. Colo. 215 (1906) ; Fl. Rky. Mts. adj. Plains 524 (1917); Coulter and Nelson, Man. Bot. Cent. Rky. Mts. 300 (1909). H. wintahense A. Nelson in Proc. Biol. Soc. Wash. 15: 186 (1902); Coulter and Nelson, op. cit. p. 300. H. lancifolium Rydb. in Mem. New York Bot. Gard. 1: 256 (1900); Fl. Rky. Mts. adj. Plains 524 (1917) —Wash- ington to Montana, Colorado and Utah. Montana: Moser Mt., Flathead Nat. For., Aug., 1925, Kirkwood 2187 (G, NY) ; Thomp- son Falls, Aug., 1909, Butler 5058 (NY); Jocko Range, Aug. 1880, S. Watson 95 (G); near Gunsight Lookout Station, Flat- head Nat. For. July, 1928, Liebig 303 (FS); West Fork Teton River, Lewis and Clark Nat. For., Aug., 1921, Lane D2-3 (FS). Wyomrnc: headwaters of Clear Creek and Crazy Woman Creek, Big Horn Mts., July-Aug., 1900, Tweedy 3193 (NY); Soldier’s Park, Big Horn Mts., Aug., 1898, T. A. Williams s. n. (US); on the Red Grade near the top, eastern slope of the Big Horn Mts., June, 1934, Rollins 503 (G, NY); Teton Pass, July, 1920, E. B. & L. B. Payson 2096 (G, NY); July, 1901, Merrill & Wilcox 977 (G, NY); Two-gwo-tee Pass, July, 1932, L. Williams 955 (G); Mt. Wagner, southeast of Smoot, Aug., 1923, Payson & Armstrong 3749 (G); Evanston, Uinta Co., June, 1900, A. Nelson 7198 (G, NY, tsorypes of H. wintahense); near Big Muddy Creek, between Fort Bridger and Evanston, June, 1938, Rollins 2323 (G); Ashley Nat. For., Uinta Co., June, 1924, Kane 6 (FS). Uran: near Mill Creek, Summit Co., July, 1926, E. B. & L. B. Payson 4881 (G, NY). Coxtorapo: White River Nat. For., June— July, 1910, Reynoldson 81 (FS); 6 miles northwest of the Rio Grande Reservoir, Hinsdale Co., Aug., 1936, Rollins 1503 (G, NY); Pagosa Springs, July, 1899, Baker 429 (G, NY); near La Plata, July, 1898, Baker, Earle & Tracy 464 (G, NY); Silverton, Aug., 1897, Shear 5227 (NY); foot of Mt. Hesperus, La Plata Mts., Aug., 1892, Eastwood s. n. (NY); Rio Grande Nat. For., July, 1924, Lister 75 (FS). Ipano: hills southeast of Victor, Teton Co., July, 1920, Z. B. & L. B. Payson 2167 (G, NY); Caribou Mt., Bonneville Co., July, 1923, Payson & Armstrong 3538 (G); ridges south of Wiesner’s Peak, July, 1895, Leiberg 1366 (NY); divide between St. Joe and Clearwater River’s, July, 1895, Leiberg 1213 (G, NY); Waterfall Canyon, Targhee Nat. For., July, 1929, Richwine 4 (FS); head of Georgetown Canyon, Caribou Nat. For., June, 1926, Phinney 89 (FS). Wasuincton: Olympic Mts.: without definite locality, July, 1890, Henderson 1850 (G); 1889, Grant 156 (G); Aug., 1895, Piper 2227 (US); June, 1900, Elmer 2529 (US); Mount Angeles, 1940] Rollins—The Genus Hedysarum in North America 229 Aug., 1931, Thompson 7831 (G); July, 1933, Thompson 9471 (G, US) ; Hurricane Ridge, Sept., 1937, Thompson 14176 (G, US); Blue Mountain, Deer Park Recreational Area, Aug., 1938, Rollins & Chambers 2693 (G) ; Bogachiel Ridge, headwaters of the Hoh River, Aug., 1938, Rollins & Chambers 2704 (G); Mt. Colonial Bob, July, 1931, Thompson 9968 (G, US); Aug., 1933, Thompson 9968 (G); Mt. Baldy, July, 1897, Lamb 1318 (US). H. occidentale is most nearly related to H. sulphurescens, from which it differs in having reddish-purple instead of light yellow flowers, usually pubescent instead of glabrous fruits and more leaflets on each compound leaf. There are 13 to 19 leaflets with pubescent upper surfaces in H. occidentale, whereas in H. sul- phurescens the 9-15 leaflets are glabrous above. Also, the flowers and fruits of the former are uniformly larger than those of the latter. Both these species are related in a general way to the American varieties of H. alpinum, but the very much larger and more broadly winged loment-articles of H. occidentale and H. sulphurescens are not matched by any other American species of the genus. The known geographical distribution of H. occidentale is of interest because of the total absence of this plant from the inter- vening area between the Olympic Mountains of western Wash- ington and the mountains of northern Idaho. Many other plants of boreal dispersion have a similar distribution. Doubtless a continuous range once existed to the northward, but such a con- tinuity could hardly have survived the glacial activity which is known to have taken place in the area. This explanation pre- supposes a preglacial migration of H. occidentale from the north to the Olympic Mountains on the one hand and to the Rocky Mountains on the other. Plants from the two areas are alike in all details, hence there is no question about their belonging to the Same species. H. marginatum, described from Colorado, and H. unitahense, described from Wyoming, do not differ in any signifi- cant way from H. occidentale. Indeed, Nelson,' in his citation of specimens accompanying the original description of H. uinta- hense, mentioned a Henderson specimen from the Olympic Moun- tains of Washington as probably belonging to the species he was describing. H. lancifolium Rydberg appears to be a leaf-form of H. occidentale. The actual type was not found at the New York * Proc. Biol. Soc. Wash. 15: 156 (1902). 230 Rhodora [JULY Botanical Garden, but specimens annotated by Rydberg and plants coming from the type-locality have proved to be narrow- leaved forms of the latter species. There are several minor variations in H. occidentale, but they are mostly quantitative and are not correlated with each other or with any phytogeographical area. For example, the length of the calyx-teeth varies a millimeter or more, the size and to some extent the shape of the leaflets vary, and the total height of the plants varies with habitat and altitude. The loments are nearly always pubescent, but an occasional collection from Montana or Washington may have glabrous fruits. There is 4 slight difference, in some cases, as to the way in which the trichomes are disposed upon the loment. Often they are ap- pressed, but in a number of collections the hairs are spreading and may even be slightly crooked. These variations, in so far as I am able to discern, are not of any real significance as far as classification is concerned, but might easily have resulted from the differences in habitat and elimatie conditions under which the plants grew. 4. H. gremiale, sp. nov. (PLATE 597). Deep-rooted, perennial herb; stems numerous from a ligneous caudex, greenish, ascend- ing, branched, densely pubescent with small simple appressed richomes, terete, 3-6 dm. high; leaves odd-pinnate, densely ap- pressed-pubescent; leaflets 5-13, oblong to elliptical, 1-2 cm. long, 5-10 mm. wide, often apiculate; stipules brownish, pubes- cent, chartaceous, very fragile, lower united, upper nearly free; inflorescence racemose, in fruit 1—-1.5 dm. long; flowers numer- ous, erect, 1-1.5 em. long; pedicels pubescent, 2-4 mm. long; calyx furnished with two small bracteoles, densely pubescent, calyx-teeth about equal, narrowly subulate, tipped with red, 4-5 mm. long; corolla pink to reddish-purple, drying purplish-pink ; standard obovate, emarginate, 12-14 mm. long, about 1 em. wide; wings about 1 em. long, 3-3.5 mm. wide, auricles blunt, broad, not united, about 1 mm. long, claw broad, about 2 mm. long; keel blunt, 13-15 mm. long; loments stipitate, articles 1-5, either closely joined or with a connective of variable length, flattened, suborbicular to slightly longer than broad, prominently and loosely reticulate-nerved, rather densely appressed-pubescent, wingless, 6-8 mm. broad; nerves or costae of the articles sup- porting numerous spine-like or setaceous processes, these purple- tipped, 3-5 mm. long and usually with a few scattered trichomes upon them; articles one-seeded, mature seeds not seen. eee RED eee ee ae ee 1940] —_ Rollins, —The Genus Hedysarum in North America 231 Herba perennis multicaulis; caulibus erectis vel adscendentibus pubescentibus, 3-6 dm. altis; foliis imparipinnatis petiolatis; foliolis 5-13, oblongis vel ellipticis undique pubescentibus; stipulis fuscis connatis pubescentibus; inflorescentiis axillaribus racemosis; floribus erectis; calycis subcylindricis, lobis subulatis 4-5 mm. longis; corollis siccatis roseo-purpurascentibus 13-15 mm. longis; leguminibus articulatis stipitatis compressis; arti- ulis suborbicularibus pubescentibus ciliatis reticulato-rugosis atis immarginatis in costis setosis vel subspinulosis.— i, cinerascens sensu Graham in Ann. Carn. Mus. 26: 251 (1937) in part, non Rydberg. H. utahense sensu Graham, ibid. p. 252 in part, non Rydberg—Known only is Uran: Uintah County: heavy adobe soil in a narrow ravine, 14 miles west of Vernal, Uinta Basin, roel 16, 1937, Reed C. "Rollins 1733 (G, Typx, R, ISOTYPE) ; 18 miles north of Vernal, Uinta Basin, June, 1937, Rollins 1757a (G, R); bench west of the Green River, north of the mouth of Sand Wash, Uinta Basin, May, 1933, Graham 7912 (G, US); Vernal- Manilla road north of Vernal, June 19, 1933, Graham 8156 (US); Uinta Basin, June, 1912, Peterson s. n. (US). Flowering plants of H. gremiale are difficult to distinguish from varieties of H. boreale which have both leaflet-surfaces pubes- cent, but with developing or mature fruits, there is no need for question as to which species one is observing. In the very young stage, fruits of H. gremiale do not show any signs of the very marked setae which later appear upon the flattened surfaces. As the fruit enlarges, small tubercles appear at various points along the surface reticulations. Soon these tubercles glongate into peculiar setae or spine-like processess which are sparsely covered with simple trichomes. H. gremiale is actually most closely related to H. boreale var. cinerascens, which it resembles in general habit. Both are pubescent throughout, though H. gremiale is much less densely covered with trichomes and less silvery in appearance than H. boreale var. cinerascens. A further clue to the relationship between these species is the occasional occurrence of very abbreviated tubercles on the loments of H. boreale var. cinerascens. This suggests a comparatively recent genetical connection between the two. H. gremiale apparently occurs only locally in the Uinta Basin of northeastern Utah; however, the plants were very abundant in the two places where I observed them. The habitat is in the geologically young foothills of the Uinta Mountains, near the 232 Rhodora [JULY bottom of the Basin. This limited distribution in a geologically young area points to a recent origin for H. gremiale. I should suggest H. boreale var. cinerascens or some other phase or variety of H. boreale as the probable ancestor. 5. H. BoreaLe Nuttall. Herbaceous perennial, stems several pubescent, 2.5-6 dm. high, branched above; stipules brown, chartaceous, triangular with a subulate tip, lower united, upper nearly free; leaves short-petioled, 4-8 cm. long; leaflets 9-13, linear- oblong to nearly elliptical or those of the lower leaves obovate, densely pubescent on both surfaces to glabrous above, puncticulate above, 3-8 mm. wide, 1-2.5 cm. long, obtuse; in- florescence racemose, elongated; bracts brown, subulate, equaling or exceeding the pedicels: flowers erect, numerous, carmine, 12-19 mm. long; calyx pubescent, teeth nearly equal, subulate, 3-5 mm. long, longer than the tube; standard obovate to broadly cuneate, reals Seas 12-17 mm. lon ng, 7-12 mm. wide; wings 10-14 mm. ong, 2.5-4 mm. wide; claw hig 2-3 mm. long, wing-auricle blunt, free, about 1 mm. lon ; loments pendent to somewhat divaricate, betes epiate artile 2-5, orbicular to subor- bicular, 5-7 mm ong, appressed-pubescent, not wing-margined, Aattened, rugose when mature, reticulations alievenily: elongat ted. Key To THe Varieties oF H. BoREALE Flowers 12-16 mm. long, erect or the lower tardily reflexed; oo not interrupted; leaflets 10-15 (-20) mm. Leaflets glabrous to sparsely pubescent above, articles of the loment rugose, but not deeply wrinkled, short —— a iecews Su vee Ree Eel sek a) vas cid te ae 5a. var. typicum. silvery-canescent throughout, articles of jo lian — deeply wrinkled over the seed, short tubercles often — thine to broadly linear; pubescence smooth, ry hairs appressed, less than 1 mm. lo 5b. var. cinerascens. Leaflets obovate: pubescence shaggy, silvery hairs spread- t15m ONE OTS ee a 5c. var. obovatum. Flowers es i ee ‘lower ag ty inflorescence inter- rupted, leaflets 15-25 CeO) Wii! WE coke esa ree ee . var. ulahense. a. H. BOREALE Nuttall, var. typicum. H. boreale Nutt., Gen. N. oe Pl. 2: 110 (1818) ; Journ. Acad. Sci. Philad. 7: 19 (1834). H. Roezlianum Prantl, Ind. Sem. Hort. Wirceb. 8 (1873) ? H. ae Greene, Pitt. 3: 212 (18 97); Rydb., Fl. Colo. 216 (1906) ; Fl. Rky. Mts. ee! Sg 524 (1917). H. Mackenzii sensu Rydb. in Mem. N. Y. Bot. Gard. 1: 257 (1900); Fedtsch. in hte Hort. Petrop. 19: oe (1902) in part, non Richardson. H. ae necogmte 1940] Rollins,—The Genus Hedysarum in North America 233 pabulare A. Nels. in Proc. Biol. Soc. Wash. 15: 185 (1902) ; Rydb., Fl. Colo. 215 (1906); Fl. Rky. Mts. adj. Plains 524 (1917); Coulter and Nelson, Man. Bot. Cent. Rky. Mts. 300 (1909) ; Wooton and Standley i in Contrib. U. S. Nat. Herb. 19: 373 (1915) ; Tidestrom in Contrib. U. S. Nat. Herb. 25: 333 1925). H. pabulare, var. rivulare L. O. Williams in Ann. Mo. Bot. Gard. 21: 344 (1934). H. Mackenzii Richards., var. pab- ulare Kearney and Peebles in Journ. Wash. Acad. Sci. 29: 485 (1939). H. cinerascens sensu Tidestrom in Contrib. U. 8. Nat. Herb. 25: 333 (1925), non in owe H. utahense sensu Graham in Ann. Carneg. Mus. 26: 252 (1937) in part, non Rydberg. — Alberta to Oklahoma, pesiinss and Idaho. ALBERTA: Rosedale Coulee, near Rosedale, July, 1915, Moodie 1078 (G, NY); Rose- dale Trail, near Rosedale, June, 1915, Moodie 1020 (G); Red Deer Lakes, July, 1879, Macoun 105 (G). NorrH Daxora: Donnybrook, July, 1935, Stevens & Kluender 132 (G, US); Range 92, Township 149, Dunn Co., June, 1936, Heidenreich s. n. (NDA); Sanish, July, 1923, Stevens s. n. (NDA); Fort Buford, 1890, Havar 2&3 3 (NY); Medora, Aug., 1923, Stevens s. n. (NDA) ; fabs 1938, Stevens & Brenkle 38-011 (G). OKLAHOMA: near Shattuck, Ellis Co., June, 1914, Clifton 3200 (G). Mon- TANA: North Fork of Bear e,, Gallatin Nat. For., Gallatin Co., Whitham 1811 (FS); 1 mile west of Teal Lake, July, 1901, Spragg 8326 (G); Ear Mountain, Lewis and Clark Nat. For., July, 1921, Butter D3-12 (FS); Jefferson Nat. For., Aug., 1927, Park 65 FS), WYoMING: c: Gilbert Creek, Park Co., July, 1937, L&R. Williams 3539 (G, NY, R); Undine Falls, Yellowstone Nat. Park, July, 1899, A. & BE. N elson 5679 (G); along Snake River, Teton Co., July, 1932, L. Williams 975 (G, 1soTyPE of a ste var. rivulare) : bars of Gros Ventre River, Teton Nelson 1087 (G) ; July, 1901, Merrill & Wilcox 998 (G, Ne, US): : Bates Creek, Natrona Co. , Ju ly, 1901, Goodding 201 (G, US); 20 miles west of Big Piney, ‘Sublette Co. , July, 1922, H. B. & L. B. Payson 2617 (G); 14 miles east of Evanston, Uinta Co., July, 1939, Rollins & Munoz 2875 (G, R). Cotorapo: Canon City, Aug., 1896, Shear 3768 (NY); June, 1917, E. L. Johnston & Hedgecock 638 (G, NY); Trinidad, Aug., 1912, Beckwith 91 (NY); June, 1917, "B. Ls Johnston 617 (G); Trinchera Creek, about 20 miles northwest of Branson, July, 1937, Rollins 1864 (G, R); Cimarron, Gunnison Co., June, 1901, Baker 274 (G); mouth of Wolf Creek, White River, Rio Blanco Co., May, 1935, Graham 9044 (G, US); Paradox, Montrose Co., June, 1912, Walker 151 (G); Naturita, Montrose Co., May, 1914, Payson 322 (G); Mancos, June, 1898, Baker, Earle & Tracy 83 (G); Du- Tango, May, 1916, Eastwood 5311 (G). New Mexico: near Cimarron, June, 1929, Mathias 556 (G); Algodones, June, 1887, 234 Rhodora [JULY Tracy & Evans 139 (NY); Canoncito, Santa Fe Co., June, 1897, A. & E. Heller 3732 (G) ; between Gallup and Albuquerque, May, 1931, McKelvey 2338 (G). Ipano: Clyde, Blaine Co., July, 1916, Book Cliffs, Uinta Baan, July, 1935, Graham 9842 (US) ; ‘Soldier’s Summit, 1894, M. E. Jones 5592 (NY); Bryce Canyon, Garfield Co., July, 1938, Rollins & Chambers 2453 (G, R); Pleasant Cr., Powell Nat. For., Garfield Co. , July, 1915, Hanks 6 (FS) ; Millard Co., June, 1938, Jensen s. n. (G); Ju ab, June, 1902, Goodding 1076 (G); mesa east of Monticello, July, 1911, Rydberg & Gar- rett 9203 (NY, US) ; Convulsion Canyon, Sevier Co., July, 1930, Albertson 41 (FS); south of Mexican Hat, June, 1930, Goodman & Hitchcock 1345 (G). Arizona: vicinity of Flagstaff, July, 1898, MacDougal 214 (G, NDA, NY). Orecon: oe Creek, Wallowa Co., July, 1897, Sheldon 8628 (G, NY); Aug., 1898, Cusick 2104 (G); head of North Fork of Imnaha River, Wallowa Co., July, 1928, Reid 738 (FS Bb. Var. cinerascens (Rydb.), 1 comb. nov. HA. cinerascens Rydb. in Mem. N. Y. Bot. ore 1: 257 (1900) ; rope 7 and Nelson, Man. Bot. Cent. Rky. Mts. 299 (1909) ; Rydb., 1. Rky. Mts, adj. Plains 524 (1917); Fl. Pr. and Plains Cent. N Am. 487 (1932). H. canescens Nuttall nT. &°G., ¥1. Am. |: 357 (1838), non H. canescens L., Sp. Pl. 2: 748 ae H. M acquenzii f. canescens Fedtsch. in Acie Hort. Petrop. 19: 274 (1902). . Macquenzii v. canescens Fedtsch., ibid. in index p. 362.—Saskatchewan and Alberta to Wyoming. SasKATCHEWAN: Quappelle, June, 1879, Macoun s. n. (NY); Whiteshore Lake, Aug., 1906, Macoun & Heriot 70786 (NY); Bare Hills, Aug., 1906, Macoun & Heriot 70785 (G, NY); Moose Jaw, July, 1880, Macoun s. n. (G); without locality, 1858, E. Bourgeau s. ni. , NY). Atperta: Cypress Hills, June, 1894, Macoun 4532 (G) ; Medicine Hat, May, 1894, Macoun 4531 (NY) ; Milk River Ridge, June, 1883, Dawson s.n. (G). WuiTHOUT DEFINITE LOCAL- ity: Lewis River, Rocky Mts., Nuttall (NY, 1sorypr ?); Rocky Mts., Nuttall (G, isoryPE? possibly same as previous collection). Montana: Sec. 28, T. 10 N., R. 10 E., Jefferson Nat. For., June, 1925, Bouham 25 (FS); Yellowstone ‘River, 1878, Havard s. 1 ( G); Midvale, July, 1903, Umbach 372 (NY, US); Lima, June, 1895, Shear 3363 (NY, US); June 29, 1895, Rydberg 2721 (NY); There is some question as to whether var. canescens Fedtschenko, though based on the Pasty eye H. canescens Nuttall, should not take precedence over the combination e made here, because of its possible priority in the varietal category. The fact that the steation H. Macquenzii, v. canescens was dubiously made only in the index of Fedtschenko’s monogtaph, l. ¢., makes it almost imperative that the legitimate valid H. cinerascens be taken up in order to make the nomenclature of this variety definite and clear 1940] Rollins,—The Genus Hedysarum in North America 235 Pitta hs 1901, Scheuber 222 (NY, US); Sixteen Mile Creek, July, 1883, Scribner 32 (G); Spanish Creek, Gallatin Co., June, 1901, Vogel s s. (G); Pa rk County, June, 1889, Tweedy s. 2: (NY). Weoiine: Shirley Basin, Aug., 1908, A. Nelson 9179 (G, NY); Red Bank, Big Horn Co., July, 1901, Goodding 332 (G, NY, US) ; Headwaters of Tongue River, Big Horn Mts., July, 1898, Tweedy 1382 (NY); Dayton-Kane Road, Sheridan Co., June, 1932, Dickson 250 (FS) ; Lower Blackrock, Teton Nat. For., Rosencrans 39 (FS) 5c. Var. obovatum, var. nov. Herba perennis argentea pubes- centia; foliolis obovatis; pilis ca. 1.5 mm. longis—Northern Nevapa: Thorpe Creek, east of Lamoille, Elko Co., Humbolt National Totes. July 25, 1928, Harold H. Price 168 (FS, TYPE). 5d. Var. utahense (Rydber comb. nov. H. utahense Ryd- adj. Plains 524 (1917) ; Tidestrom in Contrib. U. 8. Nat. Herb. 25: 333 (1925). H. boreale? sensu Durand. in Journ. Acad. Nat Sci. Philad. 11: 162 (1859), non Nuttall. H. Mackenzii sensu Watson, Bot. Fortieth Parallel 78 (1871), non Richardson — Northern UranH: Wasatch Mts., July, 1869, Watson 294 (G, NY); ak Douglas, June, 1906, Garrett 1798 (G); May, 1908, Clemens (G) ; vicinity of Salt Lake City, May, 1883, Leonard 55 (NY, TYPE); Mt. Nebo, Aug., 1922, Harris 622402 (G); Ogden Canyon, July, 1902, Pammel & Blackwood 3705 (G): Rock Canyon, near Provo, June, 1925, Garrett 3324 (G) Lehi, May, 1916, W. W. Jones 170 (G); Salina mers te Jun 94, M. E. Jones 53199 (NY); Brigham, May, 1 Zundel 193 (NY); between Linder and Pleasant Grove, Utah Co., June, 1917, Eggleston 13870 (US); South Sink, Garden City, Cache Nat. For., July, 1927, Craddock 20 (FS). H. . boreale has been widely misunderstood, possibly because of Nuttall’s' own suggestion that his plant was “H. alpinum ? Mich. Fl. Am. 2. p. 74.” Apparently realizing his error, perhaps be- cause he became familiar with the species of Michaux, Nuttall clarified his position by listing? one of Wyeth’s specimens from the “sources of the Missouri” as “Hedysarum boreale, H. Macken- zu. of Richardson, not H. alpinum of Michaux.” Whatever led many authors, including Torrey and Gray,® Gray,* Britton and Brown® and Rydberg,® to apply the name H. boreale to one or 1 Gen. N, Am. Pl. 2: 110 (1818). * Journ. Acad. Sci. Philad. 7: 19 (1834). ). SFI. N. Am. 1 (1838 *Man. Bot., ed. 2, 98 (1856) STll. Fl. 2: 392 3). * FI. Rky. Mts. adj. Plains 524 (1923). 236 Rhodora [JuLy another of the varieties of H. alpinum is not at present entirely clear. Several points brought out by Nuttall in his original description of H. boreale such as “stipules . . . subulate,” “ar- ticulations of the loment . . . rugose,” and “calix subulate” could hardly be applied to any of the American varieties of H. alpinum. Most suspicious of all, when one attempts to utilize the usual interpretation given in most floras and manuals, is Nuttall’s statement of habitat “in arid and denudated soils around Fort Mandan, on the banks of the Missouri.” Those familiar with the Fort Mandan region of North Dakota and the usual habitat for any of the varieties of H. alpinum are aware that no single species of Hedysarum is apt to be found in both habitats. Five collections of Hedysarum from North Dakota have been supplied by Dr. O. A. Stevens of the North Dakota Agricultural College for my study. All of these collections, one of which is from Dunn County in the Fort Mandan area, are the same species, H. boreale. In recent manuals plants comparable to these have been passing as H. pabulare and H. cinerascens, or in some instances as H. Mackenzii. Durand! long ago seems to have been on the right track as to the true identity of H. boreale when he noted that, “I cannot but consider H. boreale & H. canescens of Nuttall, and H. Mackenzii of Richardson, as forms of the same species, which it is impossible to separate.” H. boreale is one of those species with several recognizable varieties and numerous variants of less stable character. In several areas in its wide geographic range from Saskatchewan to Oklahoma and Arizona, trends of development are observable. Most prominent, perhaps, are those which have given rise to the large-flowered type found chiefly in northern Utah which Rydberg named H. utahense and the silvery-canescent type from northern Wyoming, Montana and adjacent southern Canada named H. canescens by Nuttall. Unfortunately, var. obovatum is not well enough known for its relationships with other varieties of H. boreale to be at all clear. Greene named another variant H. carnosulum, but the only points which are at all even measurably different from H. boreale var. typicum are the shorter stem- internodes and smaller leaves. These, it seems to me, are hardly sufficient to warrant keeping it up even in varietal rank. * Journ. Acad. Sci. Philad. 12: 162 (1859). = 7 NA ee rte te nT RO ne a ao a ly lal ei ia Si it pee cece ae Be a SW eA ea emmy Sr a nh eh ee IR Redd be eas AU cde dn dnt pL EN Eye ce PP aE SP 1940} — Rollins,—The Genus Hedysarum in North America 237 Neither the original publication nor the type, if there be one, of H. Roezlianum have been available to me, but Fedtschenko! has reproduced Prantl’s original diagnosis and notes in their entirety. The description is certainly that of a Hedysarum and, in my opinion, H. boreale or possibly one of its varieties. H. Roezlianum is based upon plants grown in the botanical garden of the University of Wiirzburg from seeds supplied by Rozl, who supposedly collected them in California. Since the genus Hedy- sarum is not known to occur in California and since Ré6zl is known to have traveled rather widely? in the Great Basin and Rocky Mountain regions, I think it is safe to assume that the seeds came from some locality in these areas. 6. H. Macxenzu Richardson. Perennial; stems several to numerous from a ligneous caudex, mostly simple above, terete, shallowly furrowed above, sparsely appressed-pubescent, 1-4 dm high; stipules white-translucent with brown streaks, united, tips of the upper free; leaves petiolate, 4-8 cm. long; leaflets 4-6, oblong to broadly ‘lanceolate or nearly elliptical, thick, glabrous to very sparsely pubescent and puncticulate above, appressed- pubescent below, 15-25 mm. long, 4-10 mm. wide, o e acute, not apiculate; inflorescence racemose, subecapitate; flowers 5-12, purple, erect but soon pendent, 18-21 mm. long; bracts subulate, pubescent, broadly scarious-margined; bracteoles linear, mm. long; calyx villose, teeth linear-subulate, nearly equal, longer than tube, 3.5-6 mm. long; standard cuneate to broadly spatulate or nearly obovate, emarginate, 15-17 mm. long, 9-11 mm. wide; wings oblong, 14-16 mm. long, 3-4 mm. wide, claw broad, about 3 mm. long, wing-auricles blunt, rou ced, free about 1 mm. long ; loments pendent to divaricately ascending, stipitate, 3-8-articled; articles nearly orbicular, not a mg margined, pl abenewrt Aewtetona 4-6 mm. tobe 5-8 mm and Gray, Fl. N. Am. 1: 357 ees Basiner in Mem. ad Ba 181 = Bull, Nat. Mus Can, 74: 148 (19 35). H. ameri- canum Mackenzii Britt. in Mem. Torr. Bot. Club. 5: 202 (1894). Newfoundland and Quebec; Manitoba and northwestward; also * Acta Hort. Petrop. 19: 324 (1902). * Garten- und Blumenzeit. Hamb. 422 (1874). 238 Rhodora [JULY eastern Siberia. Without definite locality, presumably the Cana- dian Arctic, Richardson s. n. » ISOTYPE?). NEWFOUNDLAND: Green Gardens, Cape St. George, July, 1922, Mackenzie & Gris- com 11005 and 11053 (G). Quesrc: Vaureal River, Anticosti Island, July, 1925, Marie-Victorin et al, 20871 (G). Manrrosa: Churchill, Hudson Bay, July—Aug., 1910, J. M. Macoun (G). Norruwest Trrrirory: Chesterfield Aug., 1933, Gardner 434 (G) ; Bernard Harbour, Aug., 1915, Johansen 331 (G). ALBERTA: Calgary, July, 1903, Barber 211 (G); near Fortress Pass, July, 1927, Ostheimer 82 (GQ); Ptarmigan Lakes and vicinity, he 0 (G); between Salt Mt. and Junction Lake, June, 1928, Raup 2802 (G). British ConumsBta: Mt. Selwyn, July, 1932, Raup & Abbe 3757 (G); vicinity of Hudson Hope, June, 1932, Raup & 235 (G). H. Mackenzii is a close relative of H. boreale Nutt. and might with some propriety be considered a variety of it, but there are some good reasons for keeping the two as separate species. H. Mackenzii is an arctic-alpine species which extends, interruptedly, from eastern Siberia to Newfoundland, and southward along the Rocky Mountains in western Canada. On the other hand, H. boreale is not a high-mountain species at all, but rather in- habits the low hills of the plains region of southern Canada and the western plains states together with the intermontane basins and lower mountain slopes of the Rocky Mountain Region. H. Mackenzii has fewer, larger, more brilliantly colored flowers and & more globular inflorescence than H. boreale, in which the flowers are a dull reddish-purple and disposed in an elongated raceme. On the whole, the two species differ considerably in habit and general aspect as well as in a number of technical characters which have been emphasized in the key. Considering the fact that H. Mackenzii and H . alpinum var. americanum (H. americanum Britt.) belong to two entirely different sections of the genus, it is almost incredible that Britton should have con- sidered the two to belong to the same species. His combination TA A ET AEM BLO ey es ete eT De here Sees a mc ik am a i any iad a al | | 1940} Fernald,—Spermatophytes of Eastern North America 239 H. americanum Mackenzii must have been the result of a very hasty judgment. Puate 597. HepysaruM GREMIALE Rollins, sp. nov.: FIG. 1, ISOTYPE, - 1/3, from 14 miles west of Vernal, Utah, Rollins 17 33; Fic, 2, lomen x 3; FIG. 3, upper portion of inflorescence, “2 Il. SOME SPERMATOPHYTES OF EASTERN NORTH AMERICA M. L. FERNALD (PLATEs 598-625) In an attempt to place the flora of the area covered by Gray’s Manual upon a basis of greater precision it is found necessary to check the treatment of every genus and species. In the present paper notes which have accumulated during the past year are presented. In several cases nomenclatural transfers are neces- sitated by study of the actual types or of photographs of them purchased with aid of appropriations for research from the De- partment of Biology of Harvard University. The photographs of types of Linnean species have been received through the cooperation of Mr. Spencer Savage, Assistant Secretary of the Linnean Society of London, and of Dr. John Ramsbottom, Keeper of Botany at the British Museum of Natural History. For those of Lamarck, Desrousseaux and Michaux I am indebted to the always helpful Professor H. Humbert and M. R. Metman of the Muséum National d’Histoire Naturelle of Paris. To all these gentlemen I extend my grateful appreciation. Some transfers are made from indefinite trinomials (published without clear statement of rank) or of plants originally described as subspecies. Most unfortunately, the term subspecies, clearly understood and correctly used by some of the most accurate of Old World systematists, has become debased and confused by & group of relatively inexperienced taxonomists (chiefly in this country) and its incorrect use is being urged by them, in the sense of the long-established term varietas. As correctly used the two are by no means of the same rank. The subspecies of the best taxonomists is a subdivision of an aggregate-species, Gesam- tart or species collectivus, the subspecies often consisting of 240 Rhodora [JULY geographic varieties (“races”) and by many of us often treated as true species. A clear and concise statement of the true rela- tion of these categories is contained in the Preface (p. vi.) to Hayek’s splendidly clear Prodromus Florae Peninsulae Balean- icae, 1. in Fedde, Repert. Sp. Nov. Reg. Veg. xxx!. (1924). This I reproduce: I. Subspecies . . . , i. e. formae valore systematico magno, quae plerum- que facile distingui possunt, quarum origo communis autem formis intermediis vel alio modo patet arietates..., i.e. variationes valde distinctae distributione geographica propria. ey oe j a) Subvarietates... , Variationes similibus characteris a¢ priores sed aut constantia minore aut distributione geogra- phica minus distincta a) formae -.. +, Variationes valoris systematici parvi, plerumque sine distribu- tione geographica propria. As an illustration we may look at Hayek’s treatment of the cosmopolitan weed, Stellaria media (as it occurs in the Balkans). It is broken by him into three parallel subspecies: subsp. ewmedia, pallida and neglecta. These differ in relatively fundamental characters, in which many systematists see true species: subsp. eumedia with outer sepals obtuse, petals about equaling calyx, stamens 3-5, styles recurved at apex, with forma apetala like it but apetalous; subsp. pallida with outer sepals lance-acuminate, petals minute or wanting, stamens 1-3, styles divaricate at base; and subsp. neglecta, a robust plant, with petals equaling or eX- ceeding calyx, stamens 10. Subsp. neglecta, treated by many students as a good species, has two varieties: var. macropetala With stem pilose in lines, the petals longer than calyx; and var. pubescens with stem and peduncles densely viscid-puberulent, the petals equaling the densely pubescent sepals. Hayek cor- rectly differentiated between subspecies and varieties. Another illustration of correct use of the terms is in Hackel’s monograph of the Andropogoneae in DC. Mon. Phan. vi. (1889). Hackel (PP. 383-385) broke the familiar Andropogon scoparius Michx., which he considered an aggregate-species, into two subspecies: subsp- a. genuinus and subsp. b. maritimus. The former, including ght eral subvarieties and forms, is the wide-ranging North American cespitose plant, with each fertile and sessile spikelet accompanied by a pair of plumose pedicels, the latter either with rudiments eee —— a a a a Oi os ie a ef ek STS et aaa eo 1940] Fernald,—Spermatophytes of Eastern North America = 241 or without; the second subspecies, subsp. maritimus, has each sessile spikelet accompanied by a single hairy pedicel which terminates in a well developed staminate spikelet. Hackel’s A. scoparius, subsp. maritimus consisted of var. a. genuinus, which is A. maritimus Chapm., a low and stout stoloniferous plant with solitary culms and reflexed or divergent short leaves, found on the sandy coast of the Gulf of Mexico; and var. 6. divergens (Anderss.) Hackel, based on a previously unpublished species, A. divergens Anderss. in herb., the latter a very tall plant forming dense stools, with prolonged slender ascending leaves, known only in pinelands of Texas. Nash, Hitchcock and all American students of grasses who correctly know these two plants treat them as perfectly distinct species, A. maritimus Chapm. and A. divergens Anderss. In other words, the term subspecies, correctly used by Hayek and by Hackel, covers a taxonomic concept of far greater value than the term varietas. A quite similar interpretation will be found to underlie the subspecies and varieties of those masters who have long cor- rectly used these categories in botany; the term subspecies, correctly used, is of higher rank than the term variety (varietas) as used by the overwhelming majority of outstanding leaders in taxonomy from Linnaeus down. The substitution of the term subspecies for varietas has no justification in sound taxonomy of plants; and those who so substitute it not only cheapen and obscure the categories but give concrete evidence of a lack of familiarity with the best work in taxonomy. I am quite aware that, taking their cue from the statement in Philosophia Botanica, some like to argue that Linnaeus used the term varietas, the subdivisions of species designated in his Species Plantarum by greek letters, only for garden “creations” and abnormalities. A little experience with Species Plantarum shows, however, the error of such a claim; an extended experience with the works of Linnaeus and his illustrious and wise associates and later editors shows that in actual practice he generally designated as varieties indigenous plants which he considered to be natural (often geographic) variations within the broad limits of his specific concept. The misconception of those who interpret the Linnean use of the term varietas by stressing the statement in Philosophia Botanica, rather than the actual prac- 242 Rhodora [JULY tice of Linnaeus, has been well discussed by Ramsbottom in his Presidential Address before the Linnean Society of London, delivered at the 150th Anniversary Celebration in May, 1938. No more wholesome document, in these days of many confused and superficial ideas, can be recommended for the careful con- sideration of all taxonomists. I venture to quote briefly from this masterly address, Linnaeus and the Species Concept: To return to ‘Philosophia Botanica’ we find that varieties are defined as in the ‘Fundamenta’ [1736], — Linnaeus had in mind chiefly what own as variations. There are as many varieties as there are different plants, produced from the nits of the same species. (Varietates tot sunt, quot differentes plantae ex ejusdem speciei semine sunt pro- ductae.) He adds that a variety is a plant changed by an accidental cause due to the climate, soil, cig bheage ete. It is consequently re- duced to its original form by chan oil. i ; mutata a causa Rockton: Climate, Solo, Salons Ventis , &e., reducitur Ppsirag in Solo mutato.) Further, the kinds of varieties are size, abund- ance, crispation, colour, taste, smell. ipso Nesey sunt Magni- win : : saepius Varietas.) For that reason he would have ignored them mn the general discussion of his ideas “published for the sake of my pupils,” but that they often have economic, artistic, and medicinal uses. ~ But it is not reasonable to judge Linnaeus’s ideas solely by his didactic ‘Philosophia Bota His greatest botanical work, his Golden Book, Perhaps equally striking is the treatment of varieties in ‘Species a h e bear in mind the definitions repeated two years iations in ly a a a. taney is now one of the characteristics: thus var. B. ( lvus) - po proper Lilio-Asphodelus is described as “hybrida & ee = ts. a. and B. of Adonis annua sae distinet. 8 so is shown by the fact that his earlier writings contain similar notes. ‘Hortus Cliffortianus,’ 1737, is remarkably fully annotated and contains much of interest in showing the development of Linnaeus’s ideas. Even here varieties are treated for the most part as permanent forms and ALL PELE OP Ligeanenemee i ce ama eT 1940} Fernald,—Spermatophytes of Eastern North America 243 ahienne to their i pasa pate species epebaen ad species proprias F (of Suecica’ was published in 1745. Here again several _species are regarded as related to others . . . or perhaps only varieties .. . ; americana planta abe yore est.) Varieties may be inconstant . or constant as var. 8. of Alnu Tho ug h it is obvious, therefore, that there was no change of view, we are still” left with the problem of explaining the categorical statements in ‘Philosophia Bitaiien, Is not the explanation the simple one that this was a text-book, and that too much stress has been placed upon it them a general account of certain aspects of botany at a stage where it was probably considered better to receive categorical statements than 0 wander into the realms of real philosophy.* The modern fad of certain botanists, to substitute the hereto- fore clear term subspecies, erroneously used and often misunder- stood by them, for the long established varietas, as used correctly for more than two centuries, is, as stated, a practice which cheap- ens the status of true subspecies and makes for inaccuracy and misunderstanding. As shown so clearly by Ramsbottom’s analy- sis of Linnaeus’s more scientific writings, it glorifies a text-book precept of 1736 and completely ignores the actual practice in Hortus Cliffortianus, Flora Suecica and “his greatest botanical work, his Golden Book, maximum opus et aeternum . . . ‘Species Plantarum’.” Naturally, in a work dealing with all known plants Linnaeus had to recognize the innumerable garden products, such as the great series in Brassica or Lactuca, but he did so under strong protest: “Botanists differ from florists in their conception of varieties in this respect: that the former bestow varietal names y way of defining and expressing in words some unique char- acteristic in the variety: but the latter do reverence to the objects of their worship with names showing their devotion. . . . The objects of Botanists and Florists in regard to varieties are the same, but with this difference, that the Florist begins to play his part where the Botanist leaves off... . Hence the species of * Ramsbottom in Journ. Linn. Soc. Lond. 150th Sess., oe pp. 199-205 (1938). 244 Rhodora [JULY Botanists become the Classes of Florists, the Varieties of Botan- ists the Orders of Florists.”! In spite of his inclusion of some horticultural products as varieties, a large proportion of the varieties actually recognized by Linnaeus in Species Plantarum are, as Ramsbottom points out, wild plants in nature with strong morphological or geographic differences, the varieties as main- tained by the great majority of taxonomists for two centuries. A few from very many illustrations from ed. 1 (1753) are here given, selected as plants familiar to those who are intimate with the North American flora. Circaea lutetiana (European) had @ var. B. canadensis (p. 9), which is the eastern North American C. quadrisulcata (Maxim.) Franch. & Sav., var. canadensis (L.) Hara in Ruopora, xli. 287 (1939). Alopecurus geniculatus (Pp. 60) had a var. B., which is universally recognized as a distinct species, A. aequalis Sobol. Juncus effusus (p. 326) had a var. 4, “floribus sessilibus,” which is the very distinct J. arcticus Willd. Phytolacca americana (p. 441) from Virginia had a Mexican variety B.; but in the 2d edition Linnaeus himself separated them as two species, P. decandra (Virginian) and P. octandra (Mexi- can) and everyone who understands them keeps them apart. Magnolia virginiana had 5 varieties: a. glauca, B. foetida, Y- grisea, 3. tripetala, ¢. acuminata; but in ed. 2 (1762) Linnaeus treated them, correctly, as species, M. acuminata, tripetala, ete. Eupatorium purpureum (p. 838) had a var. B., which in ed. 2 became the species E. maculatum L. These and scores of other cases (in Myosotis, Erythronium, Arenaria, Prunella, Thlaspi, Sisymbrium, Arum, etc.) sufficiently indicate that in actual scientific practice Linnaeus by no means confined the term varietas to garden products and sports, but used it for native plants with such strong morphological and geographic isolation that, in many cases, his own later studies showed them to be true species. I have never found myself able to join this modern group who base their thesis upon lack of clear understanding. I prefer to be a follower of the practice (not merely the text-book precept) of Linnaeus and of the tremendous army who by their clear understanding and painstaking studies have won lasting honor in the field of taxonomy: such outstanding scholars, who have * Linnaeus, Critica Botanica (1737), transl. of Sir Arthur Hort, 190, 191 (1938). ee ee ee ee ee " — ore 1940] Fernald,—Spermatophytes of Eastern North America 245 correctly used the honorable old term varietas, to enumerate a few, as N. J. Andersson, Arnott, Ascherson, Babington, Bentham, Bertoloni, Blume, Blytt, Boissier, Boott, Alexander Braun, Robert Brown, Briquet, Buchenau, Bunge, Chamisso, Cogniaux, Cosson, Crépin, Decaisne, A. P. DeCandolle, Alphonse DeCan- dolle, Desfontaines, Desvaux, Dunal, Eichler, Endlicher, Engel- mann, Engler, Fée, Fenzl, Fischer, Franchet, Fries, Gaertner, Garcke, Gaudin, Gay, C. C. Gmelin, C. F. Gmelin, 8. C. Gmelin, Godet, Godron, Gray, Greene, Grenier, Grisebach, Hackel, Hemsley, Hillebrand, Hoffmann, J. D. Hooker, W. J. Hooker, Hornemann, Jacquin, Koch, Kunth, Lamarck, Lange, Ledebour, Lehmann, Lindley, Link, Loudon, Martius, Maximowicz, Meisner, Mertens, C. A. Meyer, Michaux, Miller, Moquin- Tandon, Johann Mueller, Nees, Nuttall, Palisot de Beauvois, Persoon, Poiret, Presl, Regel, Reichenbach, Roemer, Ruprecht, Sargent, Schimper, Schlechtendal, Schrader, Schultes (several), Sprengel, Swartz, Torrey, Trattinick, Trinius, Turezaninow, Underwood, Urban, Ventenat, Warming, Watson, Wettstein and Willdenow. The easy-going but often quite misleading practice of using trinomials without designation of rank, whether subspecies, vari- ety, form or some other status, is an Americanism which does not reflect sound scholarship nor a desire to be quite clear to others. Those who thus avoid making their categories perfectly definite leave the burden of clarification to the unfortunate users of their work. It is not enough to say, like one author cited on a succeeding page, that “The trinomial in botany is usually referred to as a variety, although the designation subspecies would ap- pear to be more reasonable.” By thus “passing the buck,” such authors automatically invite the outlawing of their names, for, as shown in the discussion of Descurainia, these names are not validly published by the International Rules of Botanical Nomen- clature. Far from being definite the trinomial may be hopelessly indefinite. For instance, in his Working List of North American Pteridophyta (1901), the late B. D. Gilbert enumerated the variations of Athyrium Filix-femina, mostly on p. 15. He there had as no. 221 “f. rubellum f. nov. Gilbert,” but in the Appendix where he described it (p. 35) it changed to Athyrium filix-femina rubellum Gilbert, with the statement three times made that it is 246 Rhodora [JULY a variety, and the objectionable “var.” definitely finding place in the discussion: “In var. rubellum” ete. Gilbert, evidently, we not sure whether he wanted to call the plant a forma or a variety, though in the second effort he did settle on “var.” put into the discussion. The trinomial alone told nothing. How much better to be absolutely definite. The time taken and the clarity of mind necessary correctly to write the abbreviation are slight; the con- fusion created by its omission or by an explanation of the cate- gory hidden somewhere in explanatory text, which no author with generous attitude toward readers of various linguistic stocks should expect them to dig out, is enormous and enduring. As stated, authors who do not clearly designate the category to which their names belong are likely to have their combinations ignored. POTAMOGETON BERCHTOLDI Fieber, var. tenuissimus (Mert. & Koch), comb. nov. P. pusillus L., var. tenuissimus Mert. & Koch in Rohling, Deutschl, Fl. i. 857 (1823). - BERcHTotp1, var. lacunatus (Hagstrém), comb. nov. P. lacunatus Hagstrom, Crit, Researches Pot. 120, fig. 53 (1916). » PERCHTOLDI, var. polyphyllus (Morong), comb. nov. d pusillus, var. polyphyllus Morong in Bot. Gaz. v. 51 (1880) an ee Torr. Bot. Cl. iii2. 46, excluding plant of Fresh Pond 3). P. BERCHTOLDI, var. colpophilus (Fernald), comb. ergs pusillus, var. colpophilus Fernald in Mem. Am. Acad. xvii. 90, t- 20, figs. d and e, and t. 35, fig. 5 (1932). Ever since it was shown by Dandy & Taylor in Journ. Bot. Ixxvi. 90-92 (1938), that the type of Potamogeton pusillus L. (1753) is really P. panormitanus Biv. (1838) and that the plant passing as P. pusillus must be called by the earliest available specific name, P. Berchtoldi Fieber (1838), American botanists, Wishing to use the varietal names correctly, have awaited their transfer by the two English botanists. A letter to them sent long before the present war, inquiring if they proposed to make the transfers, having brought no reply and many students inquiring what they shall call the different varieties, there seems no dis- courtesy in getting them properly issued, that they may be available for use. FIMBRISTYLIS caroliniana (Lam.) , comb. nov. Scirpus carolin- vanus Lam. Illustr. i. 142 (1791). S. puberulus Michx. FI1., Bor. cecal 1940] Fernald,—Spermatophytes of Eastern North America 247 Am. i. 31 (1803). Fimbristylis puberula (Michx.) Vahl, Enum. 11. 289 (1806). Scirpus carolinianus Lam. was collected in Carolina by Fraser. By Index Kewensis it is referred to the synonymy of Fimbristylis castanea (Michx.) Vahl, but the type, of which Professor Hum- bert and M. Metman send me a very clear photograph, shows that the Lamarck species was the usually smaller F. puberula, with relatively small ellipsoid-ovoid spikelets. It is very well matched by the photograph published by me of typical F. pu- berula in RuoporA, xxxvii. t. 388, fig. 4; also by the type of S. puberulus Michx. (photograph before me) and by material from Petty’s Island, Delaware River, New Jersey, October 27, 1866, C. F. Parker, by the Virginian specimens of Fernald & Long, nos. 3751 and 3746, by Wiegand & Manning, no. 553 from Robeson Co., North Carolina and Wiegand & Manning, no. 554 from Tallahassee, Florida. The following transfers are neces- sitated. F. CaROLINIANA, forma pyenostachya (Fernald), comb. nov. F. puberula, forma pycnostachya Fernald in Ruopora, xxxvii. 396, t. 388, fig. 3 (1935). . CAROLINIANA, forma eueyela (Fernald), comb. nov. F. pu- berula, forma eucycla Fernald, |. c. figs. 1 and 2 (1935). Wuar Is ARISAEMA TRIPHYLLUM? Puates 598-600. Linnaeus in 1753 (Sp. Pl. 965) recognized a single American Arwm with ternate leaves. This species, Arum triphyllum, was, however, a composite one, with two designated but unnamed varieties based upon earlier descriptions and figures, while the only specimen seen by Linnaeus himself was supposed to be the same as Dracunculus sive Serpentaria triphylla Brassiliana, in Prod. descripta of Bauhin, Pinax, 195 (1671). What the latter may have been is a problem, for in his Prodromus Bauhin gave a full description of Serpentaria triphylla Brassiliana with a round- Ish and 3-lobed (subrotundum, trifidum) leaf; spathe 5 inches long, dark purple, with white stripes, the spadix blackish, oblong and 2-cleft at tip (flos . . . longitudine unciarum quingq; atroru- bens, striatus, venis albicantibus per medium, . . . pistillo nigri- cente, oblongo, in summo bifido). Furthermore, it was stated unequivocally to come from Brazil: “Ex Tououpinambault Bras- iliae anno 1614. allata.”—Bauhin, Prodr. 101 (1671). In view 248 Rhodora [JULY of the subrotund leaf and the bifid spadix, in addition to the source of the plant in Brazil, the single synonym given by Lin- naeus for the Virginian Arum acaule, foliis ternatis of Gronovius, must have been erroneously associated with the Gronovian species. Linnaeus cited under the Brazilian reference the treatment of Dodart, Memoires, 81 (1676), and others have cited Dodart’s remarkably clear plate. The latter, said to be Dracunculus sive Serpentaria triphylla Brasiliana, was drawn from a specimen secured from Canada! It is an excellent representation of the commonest form of the wide-ranging species which has been erroneously passing as A. triphyllum (var. zebrinum Sims) ; and it represents nothing Brazilian. Blume treated the reputed plant of Brazil as a distinct species, Arisaema brasilianum Blume, Rumphia, i. 96 (1835), but he added nothing definitely clarifying its identity. In Flora Brasiliensis, iii?, 47 (1878) Engler explie- itly states that “Tribus V. AREAE Engl. cum subtribubus 1) ARISARINAE Schott (gen. Arisarum . . . , Arisaema Mart., Pinella Ten.) .. . omnino extra-brasiliensis est”; and in 1879 (Araceae in DC. Mon. i. 534) Engler definitely reduced Arisaema brasilianum to the aggregate eastern North American species. In view of the confusion, however, as to A. brasilianum, based upon a Te- puted Brasilian plant, it is wisest not to consider Dracunculus sive Serpentaria triphylla Brassiliana, at least of Bauhin, as 4 primary basis of Arum triphyllum L. In order that the Linnean treatment may be clearly understood it is here reproduced. 4. ARUM acaule, foliis ternatis. Gron. virg. 113. _ triphyllum. Dracunculus s. Serpentaria triphylla brasiliana. Bauh. pin. 195. prodr. 101. Dod. mem. 81. B. Arum minus triphyllum s, Arisarum, pene viridi, virginianum. Moris. hist. 3. p. 547. s. 13. t. 5. f. 43. y- Arum s. Arisarum triphyllum minus, pene atroruben- c. 77. 7, & It will be seen that the single unequivocal basis of true Arum triphyllum L. was the Gronovian Arum acaule, foliis ternatis. It is important, therefore, that a specimen of this plant, which Linnaeus had in his own herbarium prior to 1753 and which he ne Oe SC _ Se ee ne a Se aN eR aM Ree Ee LE Saal Oe Re A ae Ses Py PUT a TNA, Cia rae ee TN tn EN, Re rere A a PERE Se ei le 1940] Fernald,—Spermatophytes of Eastern North America 249 marked “4. triphyllum,” consequently the Typs of the species, is the green-spathed form of Arisaema pusillum (Peck) Nash in Britton, Man. 229 (1901), based on A. triphyllum, var. pusillum Peck, N. Y. State Mus. Rep. li. 297 (1898). Pate 598, ric. 1, shows the leaf and one of the spathes of the TtypE of Arum triphyllum, < 1, from. the sheet in the Lin- nean Herbarium, the original photograph sent by Mr. Savage. An exactly similar specimen is in the me Gronovian Herbarium at the British Museum,a photograph of it sup- plied through Dr. Ramsbottom. Lin- naeus gave no original description, merely | citing the Gronovian account. Gronovious, however, described i the spadix from the | notes of Banister and | of Clayton as green $3Arum triphyllum fiue Dracuncutus triphyllus Virginianus aa Kai ch q q (pene viridi) and he cited as identical Arum minus triphyl- lum s. Arisarum pene viridi Virginianum of : Morison’s Plantarum Historiae Oxonensis, iii. sect. xiii, 547, no. 44, t. 5. fig. 43 (1715). This is the reference given by Linnaeus as the sole basis of his Arum triphyllum, 6., and Morison’s plate 5, fig. 43 shows a plant (our text-figure 1) readily referable to Arisaema pusil- i3 Trext-Fig. 1. Morrison’s figure of Arum triphyllum. 250 Rhodora [Jour lum. It should be noted, however, that in the text (p. 547) Morison cites no figure under his no. 44, Arum minus triphyllum sive Arisarum pene viridi Virginianum, but the plate was cited by him under his no. 43, Arum triphyllum Virginianum flore pallido, pistillo atrorubente, although on the plate he called it Arum triphyllum sive Dracunculus triphyllus Virginianus, thus introducing a confusion which I do not attempt to clear. The main point is, that, regardless of early literary confusions, the actual Typ of Arum triphyllum must be taken as the plant which Linnaeus had before him (our PLATE 598, FIG. 1). Until 1901, when both Arisaema pusillum (Peck) Nash (our PLATE 598, Fic. 3) and A. Stewardsonii Britton (our PLATE 600, FIG. 4) were put forward as species, botanists generally supposed that we had in temperate eastern North America a single poly- morphous species, to which several specific names had been need- lessly given. So far as I can determine A. Stewardsoni has 8 clear title; the identity of A. pusillum as true A. triphyllum has been discussed. There remains the hope that one of the earlier names may clearly belong to the species (PLATE 599) which has generally passed as A. triphyllum, the plant with leaves glaucous and dull beneath (when fresh), the lateral leaflets strongly rounded on the lower side, the horizontally rounded summit of the spathe-tube with a flange 2-8 mm. broad, the broadly oblong- ovate abruptly acuminate hood 3-6 em. broad, the fruiting head 3-6 cm. long, the depressed-globose or reniform seeds deeply invaginated at base. This is the common and often rather coarse species of rich woods and thickets from New Brunswick and adjacent Quebec to southeastern Manitoba, south to South Carolina, Tennessee, Missouri and eastern Texas. Its hood varies from purple to green, often with pale longitudinal stripes: Contrasting with the wide-ranging and inland plant which erroneously passes as Arisaema triphyllum, true A. triphyllum (A. pusillum) is usually smaller (pLaTE 598) ; its leaves are green and lustrous beneath, the lateral leaflets acute at base; the flange at the summit of the spathe-tube is narrow (0.5-2 mm. wide); the narrowly oblong to lance-ovate attenuate hood is only 2-3 cm. broad and solidly green or solidly purple on the inner face; without pale stripes above the base; the fruiting heads are only 1.5-2 em. long; the obovoid seeds stipitate and not clearly in- 1940} Fernald,—Spermatophytes of Eastern North America 251 vaginated at base. This is a plant of the southern Coastal Plain and Piedmont, occurring in wet woods and about spring-heads from Florida to Kentucky, northward to southeastern New York, Connecticut and southeastern Massachusetts. Arisaema Stewardsonii (PLATE 600, Fic. 4) in the fresh ma- terial is strikingly different. Its tuber frequently sets off stolons or stalked plantlets; its foliage is similar to that of the preceding species, but its spathe has the inrolled tube sharply and deeply corrugated with white ridges (corrugations disappearing in dry- ing), green or purple, the summit tapering gradually into the narrowly ovate (2-6 cm. broad) attenuate green to purple hood; fruiting head 2.5-3.5 em. long; seeds as in A. triphyllum (pusil- lum). A. Stewardsonii is a plant of wet or swampy woods and thickets from Nova Scotia and Prince Edward Island to Minne- sota, south to northern New Jersey and Pennsylvania. Returning to the species with fresh foliage glaucous beneath, the lateral leaflets rounded at base, the spathe with smooth tube and horizontally rounded broad-flanged summit, several early names seem to belong to it. It was, probably, the Arum s. Arisarum triphyllum minus, pene atrorubente, virginianum of Plukenet, Almagestum, 52, t. 77, fig. 5; therefore, as the only cited plant, Arum triphyllum, y. of L. Sp. Pl. 965 (1753). Pluke- net’s description is very brief and quite inconclusive and his figure is very far from satisfactory. It might have been con- ventionalized from a specimen of the glaucous-leaved plant. The first reasonably clear name for this species is Arum atroru- bens Ait. Hort. Kew. iii. 315 (1789). Aiton recognized two North American species: A. triphyllum, described quite correctly (in the sense of the Linnean type) “foliis ternatis, lamina lanceolata acuminata”; and the new A. atrorubens “lamina ovata,” culti- vated from Virginia by Philip Miller in 1758. Aiton, to be sure, took his specifie name from the not wholly definite Arum s. Arisarum triphyllum minus, pene atrorubente virginianum of Plukenet (see above); but since Aiton had fresh material with ovate (as contrasted with lance-acuminate) leaflets, his name may be accepted. Unfortunately, Mr. Weatherby, who sought, When in England, for Aiton’s type, could find nothing to stand for his species. I am, however, adopting for the plant which has recently passed as Arisaema triphyllum the name A. atrorubens 252 Rhodora [JULY (Ait.) Blume, Rumphia, i. 97 (1835). Its identity is reasonably certain. Both Arisaema triphyllum and A. atrorubens vary greatly in the color of the spathes. These color-forms have mostly been named as varieties. In so far as I can identify the old names I have taken them up. The bibliography of the two long confused species follows. Forma pusillum (Peck), comb. nov. A. triphyllum, vat. pusillum Peck in N. Y. State Mus. Rep. li. 297 (1898). A. pusil- lum (Peck) Nash in Britton, Man. 229 (1901) —Hood full purple- brown on inner face. PLatE 598, FIG. 3 stripes. PraTE 599, Fic. | orma zebrinum (Sims), comb. nov. Arum triphyllum, 4. purple or bronze within, purple to greenish without, with long pale stripes on inner face. PLate 599, Fia. 2. orma viride (Engler), comb. nov. Arum triphyllum, var. B. virens Michx. Fl. Bor.-Am. ii. 188 ( 1803), photograph of type ™ Gray Herb. Arisaema atrorubens, B. viride Engler in DC. Monog?. 1. 5386 (1879). Aris. triphyllum, var. viride (Engler) Engler, flanzenr. iv**, 200 (1920). Aris, triph., forma viride (Engler) Farwell (as vivide) in Am. Midl. Nat. xi. 50 (1928), xii. 53 (1930) —Spathe green, without or with only faint stripes. PLATE 599, Fic. 3. ci In order to clarify the identities I am showing characteristi¢ spathes of the different species and forms laid out to show the summit of the sheath and the flange at the junction of sheath and hood, in each case, X 1. ‘ ‘ i Cla A. triphyllum, forma pusillum from Chestnut Neck. Mullica River, ye | Republic, Atlantic County, New Jersey, B. Long, no. 12,064. PLaTe 4 1940] Fernald,—Spermatophytes of Eastern North America 253 Little Mud Pond, 4 miles northeast of Dorter’a Lake, Pike County, Penn- sylvania, Fogg, no. 10,732. The seeds of A. triphyllum, x 5, are shown in PLATE 598, Fic 5, these from Little Neck, Princess Anne County, Virginia, Fernald & Long, no. 22; characteristic seeds, X 5, of A. atrorubens from south of Milltown, Chester County, Pennsylvania, Adams & Tash, no. 558, are shown in PLATE 599, Fic. 4. Two southern plants, Arisaema quinatum (Nutt.) Schott, with mostly quinate leaves and A. acuminatum Small, should be noted. The former is apparently only a frequent phase of A. atrorubens which often has the lateral leaflets lobed; while A. acuminatum (PLATE 598, Fic. 4) shows nothing but great size to separate it from typical A. triphyllum. The stalked obovoid seeds are too similar. I am leaving it as A. triphyllum, var. acuminatum (Small) Engler but gigantic plants of more northern A. triphyl- lum (rarely collected because of their great size) suggest that var. acuminatum is likely to prove of no taxonomic importance. The characteristic spathe, ric. 4, is from Duval County, Florida, A. H. Curtiss, no. 2681. Typically Arisaema triphyllum has the leaflets quite green on both sides, the relatively narrow leaflets attenuate or acuminate at both ends, the hood solidly colored (greenish or dark purple) and without clear striping. On the Cumberland Plateau of Tennessee and in western North Carolina there is an anomalous plant, with relatively broad and strongly rounded leaflets glau- cous beneath, the spathe as in A. triphyllum, forma pusillum, except that it is clearly striped. In its extreme it is very definite, but one collection from Biltmore, North Carolina, with broadly rounded leaflets, has the solidly purple hood of forma pusillum. I am, therefore, treating the extreme plant as a variety because of its three characters. Further material may necessitate another disposition of it. I call it ARISAEMA TRIPHYLLUM (L.) Schott, var. montanum, var. nov. (TAB. 600, Fic. 1-3) , foliolis late obovatis vel ovatis apice subulato- aristatis subtus glaucis; spathe laminis intus purpurascentibus valde pallide striato-vittatis TENNESSEE: rocky slope at Caney Creek Falls, alt. 1200 feet, May 13, 1934, E. B. Harger, no. 7745 (TyPE in Herb. Gray.). NortH CarouiNa: deep and rich woods, Biltmore, May 15, 1897, Biltmore Herb., no. 1288°. 254 Rhodora [JULY In PLATE 600, Fic. 1 shows a characteristic leaf and Fic. 2, a spathe, X 1, of the TypE; Fic. 3 a spathe opened out, x 1, of Biltmore no. 1288», Desr. in Lam. Encyel. iv. 26 (1796). Helonias augustifolius Michx. FI. Bor.-Am. i. 212 (1803). Amianthium angustifolvum (Michx.) Gray in Ann. Lye. N. Y. iv. 121 (1848). Z. angusti- ichx, ats. in c. Am. Acad. xiv. 280 (1879). ae angustifolius (Michx.) Small, Fl. Se. U. S. 251, Desrousseaux thought that his new Melanthiwm densum might perhaps be the same as Veratrum luteum L. and gave the latter as a doubtful synonym. Consequently, later authors, including Index Kewensis, have assumed without examining Desrousseaux’s material or detailed description that his M. densum is Chamaeli- rium. Desrousseaux’s material, in Lamarck’s herbarium, Was collected in Carolina by Fraser. Its inflorescence, X 1/4, is shown In PLATE 601, Fic. 1. Fic. 2 is Michaux’s type, also X 1%, of Helonias angustifolius. That they are the same there can be no question. Differing from the other eastern American variations of Stel- laria calycantha (discussed below) in the great size of all its parts; our other eastern plants, true S. calycantha (S. borealis Bigel.) and its vars. isophylla and floribunda, having usually smaller leaves, mature calyx 2-3.5 (rarely -4) mm. long, mature capsules 3-5 mm. long, shorter styles and smaller seeds. Var: laurentiana was mistaken by me in Ruopora, xvi. 151 (1914) for the Pacific American 8. borealis (or calycantha), var. Bon- gardiana Fernald, which has similarly long calyx, capsule, styles ears ate = SNe I eer at 1940] Fernald,—Spermatophytes of Eastern North America 255 and seeds, but in maturity has the fruiting peduncles abruptly reflexed from the base. The probable necessity to take up for the familiar Stellaria borealis the less known name S. calycantha makes it unfortu- nately necessary to transfer several varietal names and also to evaluate the characters recently emphasized in the group. In order that the situation may be clear it is important to illustrate the various points by photographs taken by Mr. Hodge. The transfers follow. . CALYCANTHA, var Simecoei (Howell), comb. nov. Alsine Simcoei Howell, Fl. Nw. Am. i. 83 (1897). 8S. borealis, var. Simcoei (Howell) Fernald in Ruopora, xvi. 150 (1914). . CALYCANTHA, var. isophylla (Fernald), comb. nov. S. borealis, var. isophylla Fernald, 1. c. (1914). Pate 602, Fic. 2; PL. 603, Fic. 4. S. CALYCANTHA, var. floribunda (Fernald), comb. nov. S. ships var. floribunda Fernald, |. ¢. 151 (1914). Prats 603, FIG, 2. S. CALYCANTHA, var. Bongardiana (Fernald), comb. nov. S. borealis, var. Bongardiana Fernald, |. c. (1914). S. sitchana Steud., var. Bongardiana (as Bongardia) (Fernald) Hultén, FI. Aleut. Isl. 165 (1937). PLATE 602, Fic. 3. S. CALYCANTHA, var. sitehana (Steud.), comb. nov. S. sitchana Steud. Nom. ed. 2, ii. 637 (1841). S. borealis, var. sitchana Prey Fernald, |. c. (1914). Pare 602, Fic. 5; PL. 603, Fics. and 3. Fenzl,1 to whose judgment in the Caryophyllaceae everyone defers, had reduced Stellaria borealis Bigel. to S. calycantha (Ledeb.) Bongard, based upon Arenaria calycantha Ledeb. (1812) and, therefore, with a specific name older than S. borealis Bigel. ( 1824) ; and there was special significance in the fact that Ledebour’s species was thus disposed of by Fenzl in Ledebour’s Flora Rossica. When, however, in 1914 I doubted this identity and stood by the clearly described S. borealis Bigel., I based my skepticism upon the original diagnosis of Arenaria calycantha Ledeb. in Mém. de l’Acad, Sci. St. Pétersb. v. 534 (1812). Lede- bour, along with Boissier, Schlechtendal, Bigelow and Torrey, Stands out as one of the most accurate phytographers of his time. I, therefore, disliked to take up for S. borealis, with erect flowers on naked peduncles, the name given by Ledebour, for his 1 * Fenal in Ledebour, Fl. Ross. i. 382 (1841). 256 Rhodora [JULY Arenaria calycantha from Siberia was described with “peduncu- lis .. . diphyllis” or again ‘“Pedunculi supra medium diphylli... Flores nutantes, interdum bractea ovata.’”’ Dr. Eric Hultén, how- ever, in his Flora of Kamtchatka, ii. 64 (1928) unequivocally states that “the type-specimen in Hb. Hort. Petrop. of Arenaria calycantha Leprs. which I have seen is identical with our plant [S. borealis], and it must therefore bear the name Stellaria calycantha.” There is, consequently, no course, assuming that the plant Hultén examined was the actual type and that he cor- rectly understood the characters of S. borealis, but to take up the very imperfectly described S. calycantha. To this extent I am following him (and before him, Fenzl and Ledebour), with the reservation that a good photograph of the type, when it can be secured, may change the interpretation; but I cannot follow him when he says: “Fernald . . . includes as varieties under S. borealis BicEu. several forms which in my opinion represent distinct species” (p. 61). In his Flora of the Aleutian Islands, 164, 165 (1937), Hultén treats S. sitchana (S. calycantha, var. sitchana), with S. sitchana, var. Bongardiana (Fernald) Hultén, but misspelled Bongardia (S. calycantha, vat. Bongardiana), as a species quite distinct from S. calycantha (S. borealis) , quoting the characters I had used for them as varieties (merely the overlapping differences in length of calyx and cap- sule), to which he added “other characteristics” —‘“the styles are about twice as long (often broken on herbarium specimens), the stem is distinctly quadrangular and scabrous, especially on the edges, and, in many cases at least, short petals are developed. S. calycantha has a smooth, not quadrangular stem and no petals. As, furthermore, Fernald’s varieties have a geographical area of their own and in general appearance differ considerably from typical S. calycantha, I think they together form a separate species.” As to the general appearance which Hultén finds to “differ considerably,” PLATE 602, FG. 2, shows the upper part of one of the type specimens, < 1, of S. calycantha, var. isophylla (from New Hampshire), with sepals 2-3.5 mm. long; ric. 3 is from the upper half of a specimen, x 1, of var. Bongardiana from Atka Island (Dall), cited by Hultén and labelel by him as S. sitchana, var. Bongardiana. The difference in appearance does not see™ 1940] Fernald,—Spermatophytes of Eastern North America —.257 to me specific. Pate 603, ric. 1, is a specimen, X 1, of the re- puted species, S. sitchana, from Oregon (Sheldon, no. 8241) ; ria. 2 a specimen, X 1, of S. calycantha, var. floribunda from near Lake Superior (John Macoun). I fail to get the specific dissimilarity in appearance, except that the mature fruiting peduncles of the former are more generally (but not always) reflexed than in the latter. Piatre 602, ric. 4, shows a mature calyx and capsule, x 5, from an Anticasti (Victorin) specimen of the eastern S. calycan- tha, var. laurentiana; PLATE 602, Fic. 5, is a mature calyx and capsule, also x 5, from Alaskan var. sitchana (Wrangel, July 22, 1891, Cooley). The differences are not striking. Nor do the calyx and capsule, x 5 (Fic. 6), of S. calycantha from the White Mountains (type area of S. borealis) show any appreciable dif- ference except in size, FIG. 6 from Ice Gulch, Randolph, New Hampshire, September 2, 1890 (Churchill). If, as Hultén asserts, the type of Stellaria calycantha “is identi- cal with our plant [S. borealis]” and if, as he also states, the wide-ranging S. calycantha, with relatively short calyx, capsule and styles, differs from the more localized and larger-flowered S. sitchana by having a “smooth, not quadrangular stem and no petals,” it is most remarkable that that master of accurate obser- vation and description, Jacob Bigelow, should, in his original description of S. borealis from the White Mts. of New Hampshire, have defined his new species as having “Stem... angular. . . Petals white, deeply cloven” (Bigelow, Fl. Bost. ed. 2; 182); that, still earlier, André Michaux, getting one of the narrow- leaved varieties (S. calycantha, var. floribunda) of the small- flowered series “in borealibus Americae septentrionalis” (which, as shown in his herbarium at Paris, meant Lake Mistassini and the Saguenay River) , should have described it, as his Spergulas- trum lanceolatum, “floribus petaliferis . . . Petala brevissima, ovalia” (Michx. Fl. Bor.-Am. i. 275); that Hooker, with ma- terial from “Throughout Canada” which, as shown by his detailed and critical discussion (his p. 99), he clearly understood, should have described S. borealis unequivocally “petalis bipartitis calyce aucto trinervia vix longioribus” (Hook. Fl. Bor.-Am. i. 94) ; that Torrey & Gray, both familiar with the common eastern plant, should have characterized S. borealis with “petals (sometimes none) 2-parted, nearly the length of the... sepals” (T. & G, Fl. 258 Rhodora [JULY N. Am. i. 185). Michaux, Bigelow, Hooker, Torrey and Gray all knew their plants; and their uniform characterization of S. borealis (or calycantha) as having petals (sometimes not) is supported by the experience of others who have long known (some of us for more than half-a-century) the wide-ranging and highly variable plant with relatively short sepals, capsules, styles and seeds. As to Hultén’s other point, that the larger-flowered western series which he calls a distinct species, Stellaria sitchana, has “the stem ... scabrous,” while the small-flowered S. calycantha (S. borealis) “has a smooth . . . stem,” I can only regret that he did not study a very full series of the North American plants. Had he examined with a medium-power binocular a good series of such plants as S. calycantha and vars. isophylla and floribunda, he would have found the stems of some plants glabrous, of others fully as scabrous (and that only remotely) as in much of the material of S. sitchana. PuatE 603, FIG. 3, shows a portion of stem, x 10, of S. calycantha, var. sitechana from Wrangel, Alaska, July 22, 1891 (Cooley) ; while ric. 4 is a similar piece, also < 10, of stem of var. isophylla from Madison County, New York (House, no. 17,651). The reputed “specific” difference, that the stem of the former is scabrous, that of the latter smooth, obviously is an unstable one. The one usually decisive character which distinguishes vars. sitchana and Bongardiana, with mature calyx 4-5.5 mm. long and mature capsule 5-8 mm. long, from the series of Stellar calycantha, with mature calyx 2-3.5 (-4) mm. long and capsules 3-5 mm. long, is the strong reflexing of the fruiting peduncles 12 the former series (this one often apparently constant, though only tardily developed, distinctive character not mentioned by Hultén), for the local var. laurentiana, of shores of the Gulf of St. Lawrence in Quebec, with ascending fruiting peduncles, otherwise inseparable from the western var. Bongardiana. This character, however, loses its virtue when the fruiting peduncles of the small-flowered eastern plant become reflexed (see PL. 603, FIG, 2). Typical Stellaria calycantha has glabrous or only remotely scabrous stems. On the mountains from Washington to Montan@ and northern California var. Simcoei closely resembles it in its ph 1s i pp ea alli 1940} Fernald,—Spermatophytes of Eastern North America — 259 short and broad leaves, but the upper branches are definitely pilose. It is, therefore, noteworthy that Dr. G. N. Jones, un- questioningly accepting (FI. Pl. and Ferns of Mt. Rainier, 76) Hultén’s verdict, that S. sitchana is a distinct species, reduces without comment the local and uniquely pilose-stemmed var. Simcoei to the synonymy of the wide-ranging glabrous S. caly- cantha. Right here is a key-note to the whole situation. The series is one of the hundreds of plastic boreal species, growing extensively across or near the areas in which Pleistocene activity was extensive. Such species appear in different areas in some- what different phases. To me these are geographic varieties, so confluent in their different characters that the sorting of specimens into piles with more than a single character semi-constant is seemingly impossible. Until some stable character besides the tardy reflexing of the peduncles is demonstrated for the western Stellaria sitchana I must continue to retain it within the plastic and wide-ranging S. calycantha. Lycunis aupina L., var. americana, var. nov., omnibus partibus majoribus quam in planta Europaea; foliis radicalibus subcoriaceis 1.5-6.5 em. longis 2-8 mm. latis; foliis caulinis 3-7-jugis, imis 1.5-5.5 em. longis 2.5-10 mm. latis; calycibus mm. longis; petalis roseis 8.5-14.5 mm. longis 3-6 mm. latis—Greenland, Labrador, Newfoundland and eastern Quebec, southward confined to magnesian soils. Type: serpentine and Magnesian limestone barrens, northeastern bases and slopes of Blomidon (“Blow-me-down”) Mountains, Bay of Islands, New- bovine July 24, 1910, Fernald & Wiegand, no. 3395 (Herb. ray). Forma albiflora (Lange), comb. nov. Viscaria alpina (L.) Fenzl, forma 6. albiflora Lange, Consp. Fl. Groenl. 19 (1880). The albino form of var. americana is rare in comparison with the ordinary rosy-flowered plant. The North American plants are usually coarser throughout than Lychnis alpina, var. typica! of Europe. Occasional collec- tions from Iceland and Italy are as stout as ours and some from Iceland and Norway have flowers approaching ours in size. Furthermore, dwarfed American plants (especially from Green- land) may be as small as in some European specimens; but the average and ranges of size of the different organs shows a marked 1 LYcHNts atprys L., var. typiea. L. alpina Sp. Pl. i. 436 (1753). 260 Rhodora [JULY geographic segregation. 36 series of the European plant (typical L. alpina), including 130 plants, and 64 series, including 175 plants, of var. americana give the following results. Var. TYPICA Var. AMERICANA Radical leaves 1-3.5 cm. long, Radical leaves thicker and 2-5 mm. broad; cauline leaves firmer, 1.5-6.5 em. long, 2-8 mm. 2-4 pairs, the larger leaves 1-3 broad; cauline leaves 3-7 pairs, cm. long and 2-5 mm. broad; the larger leaves 1.5-5.5 cm. long calyx during anthesis 3-5 mm. and mm. broad; calyx long; petals (including claw) 6-8 during anthesis 5-7 mm. long; (rarely -9) mm. long, 3-3.5 (-4) petals (including claw) 8.5-14.5 mm. broad. mm. long, 3-6 mm. broad. The stature is variable but the following comparison is illu- minating. Of var. typica 38% of the plants are less than 1 dm. high, of var. americana only 13%. Of var. typica 39% are from 1-1.5 dm. high, of var. americana 29%. Of var. typica only 14% are from 1.5-2 dm. high, of var. americana 29%. Of var. typoca only 7% of the plants are 2-3 dm. high, of var. americana 27%; while no plants seen by me of var. typica are more than 3 dm. high, but 2% of var. americana are 3-4 dm. high. Similarly, all but a few of the plants of var. typica have the stem (dried) at most 2 mm. in diameter at the first cauline node; but most plants of var. americana have the stems 2-4 mm. in diameter. SILENE CAROLINIANA Walt., var. pensylvanica (Michx.), comb. nov. S. pensylvanica Michx. Fl. Bor.-Am. i. 272 (1803). S. caro- liniana, subsp. pensylvanica (Michx.) Clausen in RHopor, X#)- 580 (1939). S. CAROLINIANA Walt., var. Wherryi (Small) comb. nov. %: Wherryi Small in Torreya, xxvi. 66 (1926). S. caroliniana, subsp. Wherryi (Small) Clausen, |. c. 582 (1939). Wuat 1s AcTAEA ALBA? (PLATE 604). Two clearly defined species of Actaea occur in the northeastern states and southern Canada. They are distinguished as follows. A. aupa Bigel. in Eaton, Man. ed. 2: 123 (1818) and Fi. 29) ; Tort. inson 8, no. 2 (1768) nor A. spicata, B. alba L. Sp. Pl. 504 (1753) nor nomenclaturally at least, the following, resting upon Oe : names: A. americana Pursh, a. alba Pursh, Fl. Am. Sept. ii- . 1940} Fernald,—Spermatophytes of Eastern North America 261 (1814) and A. brachypetala DC., a. alba DC. Syst. 1, 385 (1817). A. pachypoda Ell. Sk. ii. 15 (1821) : Mackenzie in Torreya, xxvili. 53 (1928). A. brachypoda Rydb. Fl. Pr. Pl. 345 (1932), ascribed to Elliott and obviously an error for A. pachypoda Ell. —Pedicels in fruit becoming relatively thick, red or reddish; petals slender, mostly truncate; stigma during anthesis broadly sessile; fruits white, capped by the broad sessile purple stigma (whence the colloquial name Douu’s Eyss) ; seeds (3—) 5-10, 4-5 mm. long. Puiate 604, Fic. 1 A form of A. “alba,” as above defined, has dark-red fruits. This is A. ALBA, forma rUBROCARPA Killip in House, N. Y. State Mus. Bull, 243-244: 40 (1923). It seems to be an extreme with darker fruits than in A. brachy- petala DC., 8 microcarpa DC. Syst. i. 385 (1817), described from near Boston with “baccis parvis albis subrubellis, pediculis incrassatis.”’ 2. Fale te ek Teresi dace Be Ral rae A oe yy escheat ge. en Saal bam it Bey p Meet cutee oe Ee EM i . ee cee = . eee Se ac Dalton oes gulch 6 Siege Loeo a ey Caste HS gia elas cle ee Ear eae aaa os selina ee 1940] Fernald,—Spermatophytes of Eastern North America 295 Rusus laevior (Bailey), stat. nov. R. permixtus, var. laevior Bailey, Gentes Herbarium, ii. 348 (1932). Rubus laevior, occurring at the base of Cape Cod, differs in so many characters from R. permixtus Blanchard, found from southwestern Maine across New Hampshire and Vermont to central New York and originally described from the Connecticut Valley, that I find them quite as different as the other local species of the genus. The contrasts follow: R. perMiIxtus. Primocanes with 1000-5000 soft purple hairs and gland-tipped bristles with few prickles intermixed on a dm. of young growth; leaflets of primocane-foliage rhombic-ovate to obovate, abruptly short-pointed, the terminal ones 2.5-6.5 em. long, and on heavily gland- ular petiolules only 0.7-2 cm. long; inflorescence an elongate raceme, the rachis, pedicels and calices villous-tomentose; bladeless bracts mostly lobed or cleft, lanceolate, elliptic or oval. R. Larvior. Primocanes with fewer and paler prickles and glands; leaflets of primocanes more gradually acuminate, the terminal ones 6-9 cm. long and on glandless or very sparsely glandular petiolules 2.3-3.5 cm. long; inflorescence a short corymbiform raceme or open cyme, the rachis, pedicels and calices minutely pilose to glabrescent; bladeless bracts linear-lanceolate, entire. : CENTELLA erecta (L. f.), comb. nov. Hydrocotyle erecta oe Ee (1788). H. ficarioides Michx. Fl. Bor.-Am. i. 161 (1803), not Lam. (1789). H. repanda Pers. Syn. i. 202 (1805). Glyceria repanda (Pers.) Nutt. Gen. i. 177 (1818). Chondrocarpus re- pandus (Pers.) by implication only, though ascribed by later authors as a binomial to Nutt. Gen. ii. Errata (1818). Chondro- carpus erectus (L. f.) Nutt. ex Wats. Bibl. Index, 425 (1878), as error for Nuttall’s merely implied C. repandus. H. asiatica sensu Coult. & Rose, Revis. N. Am. Umb. 136 (1888), not L. (1753). H. asiatica, var. Floridana Coult. & note a c. (1888). Centella erecta, recently generally known as C. repanda, was clearly described by Linnaeus filius from Jamaica. It is the only member of the genus known from that and adjacent West Indian Islands, Fawcett & Rendle citing only the one species, as Hydrocotyle asiatica. The younger Linnaeus had dwarf plants 296 Rhodora [AuGust “foliis cordatis crenatis, scapis paucifloris longitudine petio- lorum”’, but plenty of West Indian specimens also show pro- longed petioles, Jamaican material, coll. by March in 1858, show- ing them 2 dm. long and many times overtopping the scapes. Walter’s Hydrocotyle reniformis (1788) was beautifully de- scribed, with the generic and specific characters emphasized: “foliis reniformibus, dentatis, erectis. Scapis, 2, 3’¢ 1, 2, 3” floris, involucro diphyllo”, and by Watson’s Bibliographic Index and by Index Kewensis it is clearly made synonymous with the later H. repanda Pers. (1805). Walter’s trivial name was quite available, though seven years later than that of Linnaeus filius, when the Persoon name was perpetuated and finally placed under Centella. The plant varies greatly in stature in response to the simplest of edaphic factors. When it grows on open mud, without shade, its creeping stems hug the ground and the clusters of leaves will be crowded, with petioles down to 2.5-0.5 em. long, with blades, naturally, reduced in area (down to 2.5-1.5 cm. long) and increased in thickness. When it grows among tall vegeta- tion the loosely ascending petioles extend up to 1-3 dm. in length and the blades increase in size (up to 4-8 em. long) and become relatively thin (as in most herbs under parallel condi- tions). Upon this difference and an implied difference in the size of fruit Coulter & Rose attempted to differentiate C. “asiatica,” including Hydrocotyle repanda Pers., with petioles “7.5-10 em. or even 30 cm. long” and growing from “Maryland to Florida and west to Texas”; and C. asiatica, var. floridana, “Petioles 2.5 em. long or less, making the leaves appear in rather close clusters along the rootstock, more or less pubescent; fruit somewhat larger and more or less pubscent,” this short-petioled series occurring in “Florida; also in the West Indies and Central America”; the type being A. H. Curtiss, no. 988, from near Jacksonville, Florida (in herb. Coulter). Fortunately or U2 fortunately, the sheet of Curtiss, no. 988 which came to the Gray Herbarium contains both developments of the plant; a tuft with petioles up to 2 dm. high and blades up to 5.5 em. long, and 8 — repent strand with short petioles and small blades. Similarly in the West Indies both extremes and all kinds of transition . | | 1 - i 1940] Fernald,—Spermatophytes of Eastern North America 297 occur, a Jamaican sheet showing, as already noted, petioles 2 dm. long. In our experiences in Virginia the largest and smallest extremes prove to be mere responses to light and moisture. Small, who was not averse to habit-species, went no farther than Coulter & Rose, but Nannfeldt, finding what seemed to him differences in size and proportions of fruits, treated these plants as two species, C. repanda and C. floridana. Unfortunately he rather reversed the ranges given by Coulter & Rose, citing C. repanda only from the West Indies to Texas, and north to South Carolina; while “C. floridana (C. et R.) Nannf. n. sp.”, with “Spec. orig.: Purpus Nr. 5237 [from San Luis Potosi, Mexico]” was described “Planta robusta ... Petioli.. . usque ad 15 cm. longa” and specimens were cited from north through eastern Virginia into Maryland. In other words, Nannfeldt, although deriving his specific epithet from Coulter & Rose, applied it largely to the plant which they excluded from their C. asiatica var. floridana. This difference arose through Nann- feldt’s feeling that the Maryland and Virginia material had fruits somewhat broader, “4.5-5.0 mm.” broad, while his C. re- panda should have them “3.5-4.0 mm.” broad. I have tried this character and get no satisfaction from it. Nannfeldt states that his description of the fruit of C. repanda was made from a Lindheimer specimen from Texas, from at least 800 miles away from the TypE region in eastern South Carolina. His de- scription of the fruit of his C. floridana was made, not from the Jacksonville typ, but from a specimen from San Luis Potosi, more than 1100 miles from Jacksonville and in a very different vegetational area. The Purpus plant, as represented in the Gray Herbarium, does have some fruits nearly 5 mm. broad; the type number of C. floridana (Curtiss, no. 988) shows, in the Gray Herbarium sheet, fruits down to 3.5 mm. broad. The futility of struggling to recognize two species here is evident. A word should be said regarding Hydrocotyle cordata Walt. Fl. Carol. 113 (1788). This name occasionally appears, with indication of doubt, in synonymy of Centella erecta. Walter had four species under Hydrocotyle and I am under obligation to my friend Ramsbottom for a very clear photograph of the four as represented in the Walter herbarium at the British 298 Rhodora [ AuGuSsT Museum. H. uwmbellata L. and H. americana L. were correctly described and are represented by characteristic specimens. 4. reniformis Walt. was clearly described, with the generic char- acters of Centella and the tuft of long-petioled leaves was al- notated (apparently by the late Keeper, James Britten) “ZH. asiatica (repanda)”. Returning to H. cordata, described merely “foliis cordatis integris erectis”, the specimen is a single charac- teristic leaf, correctly identified in the same hand as the last, “Limnanthemum trachyspermum.” Walter’s Hydrocotyle cor- data long antedated Villarsia cordata Ell., basis of Nymphoides cordatum (Ell.) Fernald in Ruopora, xl. 338 (1938). The latter name, under Nymphoides, cannot now be changed. Hydrocotyle cordata Walt., being of the same date and published in the identical volume with Anonymus aquatica Walt., cannot disturb the name Nymphoides aquatica (Walt.) Ktze. for the coarse southern species. Ziz1a_trifoliata (Michx.), comb. nov. Sison_ trifoliatum Michx. Fl. Bor.-Am. i. 168 (1803). Z. aurea, var. Bebbii Coult. & Rose in Bot. Gaz. xii. 138 (1887). Z. Bebbii (Coult. & Rose) Britton in Mem. Torr. Bot. Cl. ii. 35 (1890). In view of the similarity of various species of Zizia and a: Thaspium in the flowering condition it is most regrettable that the rule of priority of specific epithets forces upon a species ° Zizia the specific epithet trifoliata when there is also a Thaspwm trifoliatum (L.) Gray. Fortunately, however, the two species are quite dissimilar in appearance in the flowering condition. They will not be confused in fruit and foliage. Sison trifoliatwm Michx. was from the high mountains of Carolina. That it is the plant generally known as Z. Bebbii will be clear from the : photograph of the Typ, X 14 (our PLATE 623, Fic. 1). FXG. a is a portion, & 14, of characteristic Z. Bebbii from Warm Spring* Georgia, Tracy, no. 8917. BreWeERIA patens (Desr.), comb. nov. Convolvulus patens Desr. in Lam. Encyel. iii. 547 (1789). Breweria patens has been generally confused with B. tricho- : santhes (Michx.) Small. The latter species, resting nomencla- — turally on Convolvulus trichosanthes Michx. Fl. Bor.-A™- * : 137 (1803), was described as having fascicles of 1-5 flowers, but : 1940] Fernald,—Spermatophytes of Eastern North America 299 the type shows all the peduncles with 3-5 flowers and is a good match for extremely large specimens of B. aquatica (Walt.) Gray. Desrousseaux beautifully described the very slender plant of sandy pine barrens and openings from North Carolina to northwestern Florida (our PLATE 624), with widely divergent 1-flowered peduncles and linear leaves. His material, sent by Fraser from Carolina, was characterized: “Convolvulus filiformis suberectis foliis linearibus, mucronatis; pedunculis patentissimis folio longioribus” and his fuller account emphasized the obtuse leaves about 1 inch long and 2 lines broad, solitary flowers on horizontally divergent peduncles with 2 oval to lanceolate bracts at summit, the sepals oval and pointed. That it is not the Con- volvulus trichosanthes Michx. (= Breweria aquatica) is clear from PLATE 624, Fic. 1, showing a portion, X %, of Desrous- seaux’s type. Fic. 2 is from a plant, < 1, from Liberty County, Georgia, Wiegand & Manning, no. 2632. The following specimens are referred to B. patens. NorrH Carotina: dry open pine woods, 2 miles south of Pinebluff, Moore Co., Wiegand & Manning, no. 2631; sandhill, 12 miles north of Laurinburg, Scotland Co. , Godfrey, bi ‘5047: sand-ridge at Carolina Beach, New Hanover Co., Godfrey, no. 4705; sand- ridge, 5 miles west of Clin n, Sampson Co., Godfrey, no. 4495. CAROLINA: without ae of tocaltt ty, FRAZER, TYPE (see PLATE 624). Gerorcia: sandy roadside, 4 miles southwest of Hinesville, Liberty Co., Wiegand & Manning, no. 2632; sandy soil near hapinte ) Biltmore Herb. no. 14953. Forma: dry pine barrens, De Funiak Springs, A. H. Curtiss, no. 5903 (distrib. as Evolvulus sericeus Sw.) ; De Funiak, Tracy, no. 9198. Hovustonta adarorys (Lam.), com mb. nov. Gentia na nigricans Lam. Encyel. ii. 645 (1786). Houstonia angustifolia Michx. Fl. Bor.-Am. i 3 (1803) . The heretofore problematic Gentiana nigricans was described with “fleurs . . petites, purpurines ou bleuatre, nombreuses, & disposée au sonnet de la plante en cime corymbiforme. Leur limbe est quadrifide & un peu pubescent 4 l’interieur.” Lamarck had his doubts about it and really suspected the true generic affinity, commenting, “Il faudra peut-étre la ranger parmi les Houstonia, avec lesquelles il paroit qu’elle a des rapports.” TE 625, Fics. 1 and 2 are the two plants, < 1, of the TYPE of Gentiana nigricans; FIG. 3 is a similar portion of Houstonia 300 Rhodora_ [Avcust angustifolia from Georgia, Boykin, near the type region of Michaux’s species. That they are the same no one is likely to question. described but excluding synonyms S. spicatus Lam. and Prenan- thes autumnalis Walt. Mulgedium leucophaeum Willd. DC. Prodr. vii?. 250 (1838). L. leucophaea (Willd.) Gray in Proe. Am. Acad. xix. 73 (1883), not Sibth. (1833). L. spicata sensu Hitche. ex Britt. & Brown, Il. Fl. iii, 276 (1898), not Sonchus spicatus Lam. Encyel. iii. 401 (1789), basinym. It is most difficult to understand how Lactuca biennis, the erect large-leaved plant with thyrsoid-paniculate inflorescence of erect heads with pale- or dirty-bluish to whitish ligules should have been confused with the beautifully described Sonchus spicatus Lam., with the diagnosis “Sonchus pedunculatis squamatis nut- antibus, spica longissima virgata, foliis runcinatis” and wi the full account of this “Nouvelle espéce trés-distinguée de toutes les autres par la disposition de ses fleurs. Sa tige est haute de deux A trois pieds, trés-simple . . . feuilles . . roncinées - - les supérieures sont fort étroites, linéaires-lancéolées, entiéres; plus petites que les autres. Les fleurs sont purpurines, pencheées ou méme pendantes . . & disposées en un épi effilé, trés-long, & terminal . . . Cette espéce a été trouvée dans la Caroline mérid- ionale par M. Fraser, qui nous l’a communiquée”. Lamare clearly described Prenanthes autumnalis Walt.; in fact he cited Walter’s species as a probable synonym! Incidentally, Fras could not have missed Prenanthes autumnalis in South Carolin’, he could hardly have found Lactuca biennis there. Small giv the southern limit as in North Carolina, and the only specimens from that state are from the Blue Ridge counties. Lamare ve familiar with our plant. He clearly described it, with py? idal panicle and with the sessile pappus russet, as Sonchus racemosus. The photograph of his type is perfectly character tic; but unfortunately there is already a Lactuca racemos The identity of Sonchus spicatus is so apparent from Lamarek® lucid account that, since noting the error of identifying sr coarse Lactuca with it, I have merely waited until I could 8° 1940] Fernald,—Spermatophytes of Eastern North America 301 a photograph of Lamarck’s type before attempting to clarify the situation. The photograph of the type of Sonchus spicatus, received through the codperation of Professor Humbert and Messrs. Metman and Cintract,! is conclusive; the clarification is only tentative. The identity of the many North American plants described in the 18th and early years of the 19th centuries by European authors under Lactuca, Sonchus and Mulgedium are quite im- possible to decipher from the meager descriptions alone. Yet, until they can be painstakingly studied, with fully representa- tive American specimens for comparison, we can only take the statements of others regarding them. In modern would-be exact taxonomy no saying is more true than “you can trust nobody”. In the present instance the untrustworthiness of past identifi- cations is patent. The confusion seems to have started with the great and almost unexcelled scholar, A. P. DeCandolle, who, correctly describing Mulgedium leucophaeum (Willd.) DC. Prodr. vii?. 250 (1838), “caule erecto glabro apice paniculato, - . . paniculae ramis racemosis . . . Capitula . . juniores albi demtim coerulescentes”, cited Sonchus spicatus as a synonym but with the comment: “Nomen Lamarckii etiamsi vetustius sed falsum omisi”. Starting with DeCandolle, the misidentification of S. spicatus has stood uncorrected for more than a century. Asa Gray, calling the coarse species of Lactuca, L. leucophaea, wrote in 1883, “The large synonymy of this species may be still further extended; for it must be the S. racemosus as well as S. spicatus of Lamarck, S. biennis of Moench, and it may also be S. multiflorus Desf.”—Gray, Proc. Am. Acad. xix. 73 (1883). These with other synonyms and “sensu” names appeared under L. leucophaea (Willd.) Gray in the Synoptical Flora. As indi- cated in the opening paragraph of this note, the latter name is a later homonym. Some other names are similarly disbarred. Other names treated by Gray as synonyms, such as S. pallidus ~ Torr., seem very doubtfully identified. Torrey’s brief account called for a yellow-flowered plant of fields, with lanceolate- ensiform leaves and flowering in July. That sounds like some *A package of photographs of types, arranged for two years ago, was mailed in Paris on September 1 and was delivered in Cambridge on December 6, 1939, more than 3 months en route! 302 Rhodora [ Auaust form of Lactuca canadensis. The identity of Sonchus biennis Moench is not absolutely established but Gray had no doubt about it; and Moench’s comparison of it with S. alpinus L., a European species with which our plant was much confused by early European authors (including Linnaeus) makes the identi- fication reasonable. I, therefore, take up L. biennis, at least until Moench’s type at Marburg can be examined. At any rate, we can use for our plant neither of the familiar names, L. leucophaea nor L. spicata. The following forms should be noted. L. BIENNIS, forma integrifolia (T. & G.), comb. nov. Mul- gedium leucophaeum, 8. integrifolia T. & G. Fl. ii. 499 (1843). L. spicata, var. integrifolia (T. & G.) Britton in Mem. Torr. Bot. Cl. v. 350 (1894). L. spicata, var. aurea, forma integrifolia Jennings in Ann. Carnegie Mus. xiii. 443, pl. 33 (1922). L. BIENNIS, forma aurea Vana comb. nov. L. spicata, var. aurea Jennings, |. c. 440 (1922). Ill. ON TWO WEEDY CRUCIFERS Reep C. Rouurns! Durine the last few years, weed specialists and agronomic botanists in America have become aware that two species of pernicious crucifers were passing in weed-surveys and bulletins as the same plant. Weed-manuals have usually given the com mon name of these plants as “hoary cress” or in some cases “white-top”. From the striking similarity of the two species, which have frequently been found growing in the same field, it is no wonder that they have been confused. Yet, taxonomically, the two have usually been known in separate genera under te names of Lepidium Draba L. and Hymenophysa pubescens C. A. Meyer. Both are introductions from the Old World and are apparently spreading rapidly, particularly in western Nort America. Repeated queries regarding the systematic position of these species have prompted a detailed examination of each with a view to determining their generic relationships. Historically, Lepidium Draba, so-called, has often bee? thought of as an aberrant species in the genus Lepidiuwm. Lin- 1 Society of Fellows of Harvard University. ate ne er Se ne i 1940] Rollins —On two weedy Crucifers 303 naeus himself in the tenth edition of his Systema and second edition of Species Plantarum shifted it from Lepidiwm to Coch- learia. Since that time, “Z. Draba” or one of its numerous subspecies, varieties or forms has been placed, by different authors, in no less than five genera! other than Lepidium. Al- most without exception, treatments of Lepidiwm have either excluded “ZL. Draba” or placed it by itself in a separate section or subsection of the genus. Thus, nearly everyone who has dealt with the plant has been impressed by its singular peculiari- ties and was not satisfied to give it equality with the general run of species in Lepidium. Some of the salient points of differ- ence between “ZL. Draba” and the other species of Lepidium may be summarized as follows: 1, the fruits of “L. Draba” are indehiscent, those of Lepidium are dehiscent; 2, the siliques of “L. Draba” are neither strongly flattened nor carinate-margined, while in the rest of Lepidium the siliques are strongly flattened contrary to the narrow septum and the margins are either car- inate or at least strongly compressed; 3, the siliques of “L. Draba” are somewhat inflated (markedly so in var. repens), whereas the siliques of Lepidium proper are uninflated; 4, the nectar-glands in “L. Draba” are comparatively large and well developed, completely surrounding the base of the single stamens and subtending the paired stamens but in the rest of Lepidiwm the nectar-glands are small, poorly developed, merely subtend- ing the single stamens and only weakly developed below the paired stamens, or are absent entirely. According to Schulz,? myrosin-cells are found in the vascular bundles of “L. Draba”, but none have been found in the vascular bundles of those species of Lepidium proper which have been investigated. Although the latter point does not have any practical taxonomic importance, it adds weight to the evidence against a presumed close direct relationship between “L. Draba” and other species of Lepidium. Taking all available evidence into consideration, it appears to be a mistake to continue “L. Draba” as a species in the genus Lepidium. The earliest generic name to which this species may * Nasturtium, Crantz, Crucif. 91 (1769); Cardaria, Desvaux, Journ. de = 3: 163 — raba, Baumg., Enum. Stirp. Transilv. 2: 232 (1916); Corton Wallr., Sched 340 (1822); Fhysolepidion, ichcaik, Enum. Pl. Nov. 97 (184 Engler’s Pflanzenf. 17b: 416 and 476 (1936). 304 Rhodora [Aucust be referred is Cardaria of Desvaux 1. c., hence the plant in question should be known as Cardaria Draba (L.) Desv. The striking habital resemblance between Cardaria Draba and Hymenophysa pubescens suggests a closer relationship between the two than has usually been admitted. A detailed study of H. pubescens has not revealed a single valid reason for its not being considered congeneric with C. Draba. The siliques of H. pubescens are inflated and indehiscent; the seeds are large and have incumbent cotyledons; there is only one seed in each loculus of the ovary; the petals are broad-limbed and narrow- clawed as in C. Draba; and the style is of the same type as that found in the latter species. Most authors have separated Hymenophysa from Lepidium on the basis of its subglobose inflated silique and broad septum, but these characteristics are shared also by C. Draba var. repens, and to a lesser extent by typical C. Draba. The most important character which Car- daria Draba and Hymenophysa pubescens have in common 18 an indehiscent silique. In the species of Lepidium with which I am familiar, the siliques are definitely dehiscent. The extent of development and disposition of the nectaries of C. Draba and H. pubescens are similar, and both species possess myrosin-cells in their vascular bundles which seemingly further indicates close relationship. The available evidence indicates that EE pubescens should be placed in the genus Cardaria, which I should constitute as follows: 1. Carparta Drasa (L.) Desv.; based on Lepidium Draba L., Sp. Pl. 2: 645 (1753). Type species of the genus. The oldest American specimen of C. Draba in the Gray Her barium, is E. L. Greene no. 783, collected near Yreka, Californi@, in 1876. The species is now widely distributed in North Amer ica as a noxious weed. It is particularly troublesome in the slightly alkaline soils of many irrigated districts in western America, though it is by no means restricted to this type of habitat. Specimens of C. Draba have been seen from Nov Scotia, Massachusetts, Connecticut, New York, New Jersey; District of Columbia, Illinois, Nebraska, Idaho, Wyoming: Colorado, New Mexico, Arizona, Utah, California, Oregon and Washington. 1940] . Rollins,—On two weedy Crucifers 305 la. C. Drapa (L.) Desv., var. REPENS (Schrenk) O. E. Schulz; based on Physolepidion repens Schrenk, Enum. Pl. Nov. 97 (1841). Specimens of the variety have been seen from Alberta, South Dakota, California, Oregon and Washington. O. E. Schulz, l. ¢., has named these plants forma macrocarpa, but the additional epithet is not necessary for a clear understanding of variety repens. 2. C. pubescens (Meyer), comb. nov.; based on Hymeno- physa pubescens C. A. Meyer in Ledeb. Ic. Pl. 2: 20 (1830) tab. 165, and Fl. Alt. 3: 181 (1831). The plants of this species which have been introduced into North America are not typical of the species as originally described or of specimens coming from Central Asia. Our weeds have a much more elongated fruiting raceme and smaller siliques than those described and figured by Meyer, I. c., and by Busch. There are four specimens of C. pubescens from the Altai region in the Gray Herbarium, all of which have a short dense raceme, compact subcorymbose inflorescence and larger siliques than specimens from North America. On the other hand, a speci- men collected in 1939 in the territory of Neuquen, Argentina by A. Chicchi, is typical of the Altai plants. The Argentina collection, so far as I am aware, is the first collection of this weedy crucifer from South America. It should be pointed out that Schulz’s illustration in Die Pflanzenfamilien? is of the plant Which we now have as a weed in the United States and Canada. His figures so nearly match specimens collected in Idaho by Mrs. Soth, which were known to have been sent to him at the Berlin Herbarium,® that it is not improbable that these draw- ings were actually taken from Idaho-grown material. As near as I have been able to learn, the weedy Cardaria pubescens as found in North America is an undescribed variety which perhaps originally came from some area south and west of the Altai dis- trict of central Asia. Import-records indicate that the seeds of these plants were brought to America as impurities in Alfalfa seed* The variety which is extant in North America is named as follows: * Fl. Sib. et Orient. Extr. 107 (1913). *Engler's Pflanzenf. 2 Aufl. 17b: 477 (1996). * Science 62: 509 (1995 ( : *Mr. Herbert Groh of Ottawa is actively interested, and has a forthcoming paper on this subject. 306 Rhodora [ Aucust | 2b. C. puBESCENS (Meyer) — var. elongata var. nov. Herba perennis; inflorescentiis 4-10 cm. longis; siliquis 2.5-3.6 mm. latis. Muicuican: Ypsilanti, hac. 1919, C. Billington s. n. (G); Aug., 1919, B. A. Walpole s. n. (G). saa Pocatello, July & Aug., 1925, Mrs. M. E. Soth s.n. (G, NY, US). Wyom-a ING: near Powell, Park Co., shrak 1933, Rollins 324 (R). CoLo-— RADO: Fort Collins, B. Thornton 1 (US) : near La Jara, Aug. 1926, M. W. Talbot s. n. (Us) "Caciwenieia: edge of an alfalf | field, near Sacramento, June, 1932, Bellue s. n. (US). OREGON: near Redmond, Sept., 1922, Whited 499 (G); near Burns, Harney _ Co., July 9, 1933, ¥ chores 11960 (G, Type; NY, US, © isotypes); Klamath Falls, Jun 1 Applegate 3603 (aa ; son 9047 (G, US) ; May, 1935, Phomusn 11539 (G, P near Tonasket, June, 1931, Thompson 7107 (G, US) ; wheat held Pullman, July, 1925, R. F. Hazton s. n. (G). Presumably the | same plant has been ‘reported from Pennsylvania by J. M. Fogg, Jr.,! but I have not seen specimens of the collections cited. Out plants are neither of the following species which have not turne up as weeds in North America. @ Although I have not seen specimens of Hymenophysa fen E ) strata and H. macrocarpa, judging by their descriptions and note 7 concerning them, they are also to be included in Cardaria. 3. C. fenestrata (Boiss.), comb. nov.; based on Hymenophys fenestrata Boiss. in Ann. Sci. Natur. Bot. 177: 197 (1842) Turkestan -e. macrocarpa (Franch.), comb. nov.; based on Hyment physa macrocarpa Franch. in Ann. Sci. Natur. Bot. 15%: 23 (1883). Persia. 1 RHopora 39: 190 (1937). Rhodors Plate 598 Photo. H. G. Fernald. ? F iM ARISAEMA TRIE ls L v M type (leaf and smaller spathe), x 1.0! Ant . ke oe LU ere "Vi 2 _spathe, i aid open, < 1, from type-region, Virginia; FM" seeds {rom rg = A. TRIP HYLLUM, forma PUSILLUM: FIG. 3, spathe, laid open, « 1, from he: Jers ‘ A. TRIPHYLLUM, Var. ACUMINATUM: FIG. 4, spathe, laid open, x 1, fre Flor 2 » FOc Rhodora Plate 599 Photo. H. ( t. Fernald. _ ARIsamma ATRORUBENS : 2, from Pennsylvania. \TRORUBENS, form: ATRORUBENS. FIG. 1, spathe, laid open, X 1, from Maine; Fic. 4, se¢ ds. A. A. ZEBRINUM: FI forma VIRIDE: FIG. 3 2. G “ Sp t € { Ope x ( Virginia } i h : lal | I n, \ L from 1 Spa he ( opel x | om South ( aroina | ; laid i 5 . i; l Rhodora : Plate 600 Photo. H. G. Fernald. 1 ARISAEMA tipple M, var. NTANUM: nigral nd 2, leaf am spathe, * from the TYPE; FIG. 3, spathe, laid open, X Getty North Carolin: \. Stew ARDSONII: FIG. 4, spathe, laid ope x 1, from Pennsylv: ania. seagoing Rhodora Plate 601 ey 7 ae JARAOWMLA a Angus fc a aus Se thes Z 'd af carr) act ) ‘ ; het de Ait thihy A of rye gus. felaca 4 7 4 ite Om Arn ut, OPE, Fat te . Lat Poa post flor N tifianr. Arasas : C ohm coe ce Chat aus jg7 & tt / VA (eur Ju Sat f Ee, ALM gf of hors 7 bumusey boa The Caro hrs acd a early Fic. 1, Ty a Type of MELANTHIUM DENSUM Desv.. X 1 Cype of HEtontas ANGUSTIFOLIUS Michx.. x Lf Rhodora Plate 602 | | > | Photo. W. H. Hodge. ARIA CALYCANTHA (S. borealis): 1a. 6. fruiting calyx, x 5, from White Mou tain, New Hampshire (type-region of S. borealis ba ‘ar. LAURENTIANA: FIG. 1, portion of TYPE, X 1; Fic. 4, fruiting calyx, Anticos tl : ar. ISOPHYLLA: FIG. 2, portion of Typ i. Var. Bona ARDIANA: FIG. 3, portion of pls int, x 1, from Atka Island, Alask@- Var. SITCHANA: FIG. 5 , fruiting calyx, , from Wrangel, Alaska . -& froll x 5,3 Rhodora Plate 603 Photo. } V. H, Hodge. STELLARIA CALYCANTHA, Var. SITCH: Fic. 1, portions of plant, < 1, from Oregon ; 1G. 3, portion of stem, x 10, from Wr san “ab iska. ,At. FLORIBUNDA: FIG. 2, portion of plant, x 1, from Ont: uP. Var. ISoPHYLLA: FIG. 4, portion of stem, X 10, from New York. Rhodore Plate 604 Photo. W. H. Hodge. ACTAEA PACHYPODA: FIG. 1, fruiting raceme (dried), from Nova Se A. RUBRA, forma NEGLECTA: FIG. 2, fruiting raceme ‘(ered x 1, from V sami Rhodora Plate 605 Photo. W. H Hodge 2 . » RorIPPA ISLANDICA, Var, MICROCARPA: FIG. 1, isotype of Nasturtium palustre, var. P >» microcarpum, > l, from Amur (2 plants) ; FIG. 2, fruiting raceme of isotype of R. islandica, var. Fernalk liana, i. from Maine. ar. HISPIDA: FIG. 3, siliques, < 5, from Idaho, of Nelson & Macbr-de, no. 1318, distributed as FR. terrestris globosa Nelson and cited by Butters & Abbe as R. LS andica, vai glabrata; Fic. 4, siliques, x 5, from New York; Fia. 5, silique, x 5, irom New Mexico. of Heller & Heller, no. 3743. cited by Butters & Abbe as R. slandica var glabrata; FIG ) tetra Lrpe ll ite silic ue, xX 5 { om Sague ni Co ia. 6 traci Li 1q 0, Vy “9 Quebec; Fig. 7, silique, < 5, from Lake St. John, Caine: FIG. 8, young tri- or tetracarpellate silique, x 5, from Rhode Island. R. BARBAREAEFOLIA: FIG. 9, raceme, X 1, of authentic specimen from Amur. Rhodor: Plate 606 Photo. H. G. Fernald. FIG. 2, summut ol {UBUS THOLIFORMIS: FIG. 1, characteristic primocane-leaf, x 1; ; FIG. 3, lower surface of leaflet, x 10. petiole and bases of leaflets, x 5 , Rhodors Plate 607 Photo. H. G. Fe rnald, RUBUS THOLIFORMIS: FIG. 1, portions of floricane and inflorescences, < 1; 9m, 2, flowers and pedicels, X 5; Fic. 3, portion of primocane, Plate 608 Rhodora Photo. H. G. Fe rnald. | of RuBUS SPICULOSUS: FIG 1, characteristic primocane-leaflets, < 1; FIG. 2, summit tials and bases of leaflets, x 5; FIG. 3, portion of primocane, x 3. thodora Plate 609 H. G. Fi rnald. I )» I 3 ap : : . . . ‘ , Tas 5 siete PICULOSUS: FIG. 1, tip of flowering shoot, x 1; FIG. 2, calyx and pedicel, X 4. Rhodora Plate 610 Mi E > 5 7 i SERENE mh nes temiireatatsenny Photo. H. G. Fernald. ‘ _ : re . ie er sul- XUBUS ACULIFERUS: FIG. 1, characteristic primocane-leaflets, x 1; FIG. 2, lowe 10. i face of leaflet, « Rhodora Plate 611 Ruy ‘UBUS ACULIFERUS: FI 9 o G FIG . 1, portion of raceme, * 1: FIG. 2, calyx and pedicel, vee sp portion of primocane, « 3 Rhodars Plate 612 ; { ; 7 { Photo. H. G. Fernald. 9 » ie ew is ae RvuBUS ADENOCAULIS: FIG. 1, portion of primocane and oo leaf, X 1; ™ lower surface of leaflet, x 10; Fic. 3, portion of primocane, Rhodora Plate 613 Photo. H. G. Fernald. RUBUS ADENOCAULIS: FIG. 1. medium-sized inflorescence, * 1; FIG. 2, flower-buds and pedicels, « 5. Rhodora Plate 614 Photo. HG: r Fernald. errr "ATACIATY - - - ° \UBUS ADENOCAULIS: tip of vigorous flowering sprout, x 1. Rhodors Plate 615 Photo. H. G. Fernald. RUBUS ADENOCAULIS: fruiting racemes, < 1. Plate 616 Rhodora bos Ln > > @ af? =~ -~ t _ i ee -_ t = ae lata Plate 617 a Rhodor lowe ring branches, l,i ADJACENS: FIG. : LUBUS Rhodora Plate 618 Photo. H. G. Fernald. {UBUS ADJACENS: fruiting racemes. 1. Rhodor: Plate 619 Photo. Ht. GQ. rnald. {UBUS BRACTEOLIFERUS: FIG. 1, portion of primocane and a primocane-leaf, x 14; FIG. 2, portion of primocane, < 3. Rhodore Plate 620 Photo. H. G. Fe rnald. RUBUS BRACTEOLIFERUS: FIGS. 1 and 2, fruiting racemes, X 1. Plate 621 Rhodora noe ah yor eX eh ba iad SS 2" i . gee +s a3 Bi]: A, ; vt, 4h; A Fernald. G. H. Photo. primocane- L: characteristic fruiting raceme, \ — = = = = ‘ Ce A & Oo = = — es ~~ — Sf an ZS a — — ~ 5 r surface of leaflet, FIG. » no 5 2 -— i << om ae AN _ _ — = sa Ry no mm & mA Ae tH er Rhodora RUBUS PALI typical floric DIVAGUS: FI cane . Fic. 1, br ape oo. Py | nches alyx id { ) Plate 622 t vigorous f pedicel, Ic yricane, 0. 9 portion O} Rhodora Plate 623 ? Ao yy PLANTS € Pie j ¢ 4 are Photo. H, G. Fernald. ZIZIA TRIFOLIATA: FIG. 1, TYPE, X 14, of Sison 2 noeealaga Michx.; Fic. 2, characteristic plant, x 1%, of Zizia Bebbii (Coult. & Rose) Britt Rhodora Plate 624 Photo. H. G. plant Mic ee Fernald. / : , ring Bre prireees PATENS: FIG. 1, TYPE, K 1%, of Convolvulus patens Desr.; FIG 2, flowe [| oF ie weria trichosanthes sensu Small, hes the not Convolv nulus trie “thosanth Rhodora Plate 625 Pi. nha aT aC, oy ange Me. : g we mH Mi, ‘@ Kit, > fae. ay 1a em he Calor a geo tinns " por We flows Gam. die Gtntiann gta onank dar Photo. H. G. Fernald. HovusToNIA NIGRICANS: 1 and 2, type of Gentiana — Lam., X 14; Fi. 3, characteristic branch of Diigo Susuatilelin Michx, x 4, from Goenka INDEX New scientific names are printed in full-face type Fess nigrum, 292 an ae um baccis niveis, [et rubris], 3, 265 Adee cites: alba, 260-265, pl. a f. rubrocarpa, ag 264; ame cana, 261, x a , 260 0, 3. pa vet 261; sbrankiv nace a. alba, 261, 8. microcarpa, 261, rubra, 261; brachypoda, 261; eburnea, 261; longipes, 261; neglecta, 261; pach- ypoda, 261, 263, 264, f. rubro- carne 264; rubra, 261, 262, f. neglecta, 261-264, pl. 604; spicata, , 265, B. alba, 260, 362, 263, 265, y: rubr ra, 261 Adonis annua, var. a., 242, var. B., pecs s, 24 Alopecurus —— 244; genicu- latus, 244, var. B., 244 a. genuinus, 240, 8. maritimus, 240, 241; subsp. maritimus, var. divergens, 241, a. genuinus, 241 An moe es 240 pl. 599, f. Seieeront: 262; 253, pl. 599; eae 248; hasta- ae 249252, :. 508. var. acuminatum, 253, pl. 598, var. mon , , pl. 600, f. p 3 , 262, 253, l. 598, var. genet ae ty e 248; atrorubens, 251-253; minus triphyllum, ete., 249, 250; tri- 307 phyllum, Zao pl. 598, £B., 248, 249, 248, 251, £. atro- parca, 252, sive Dracuncu- us, 250, 6. viride, 252, a. zebrina, a pod B. virens, 252, Vir rginia- nu oe Aranda rio Astragalus loneifolius, 267 filix-femina, 245, rubel- 245, f. rubellum, 245, var. fabian 246 Blackberry, 286 Br Breweria aquatica, 299; patens, 298, 299; trichosanthes, 298 Bugula odorata lusitanica, 263 Bossekia parviflora, 275 assica, 243 Camelina, 274; barbareaefolia, 270, Carat “ 304, 306; Draba, 304, repens, » f. macro- sain ; 305; fenestra arpa, . ; pubescens, 305, var. elongata, ardiolepis, 303 arya, 292; spp., 292 ryophyllaceae, 255 soteosict dentata, 292 us americanus, 292 Sicia 296, 298, asiatica, 295, 296, floridana, 2 295, var. floridana, reece floridana, D095, 5 200; repanda, 295, 9 = rocarpus erectus, 295; repan- Christophoriana Americana, 265 Cimicifuga racemosa, f. dissecta, issecta, Cirecaea lutetiana, "244, va B. canadensis, 244; peeing lr aa var. re 344 ee 03 volvulus patens, 298; tricho- miethen: 298, 299 , 302 rs, 302 rears orientale, 263 308 Descurainia, 245; intermedia, 266; pinnata brachycar Pe nig be ubsp. brachyearpa 266; achy- a, 266; Siehandictdena: 266; Tishardeont, 266 Draba, 303 Dracunculus sive Serpentaria tri- phylla, ete., 247, 248 Erythronium, 244 eying maculatum, 244; pur- pur 244 eeaivules sericeus, 299 — lis caroliniana, eucycla, 247, 247; hewtrl 247; puberula, 247, f. eucycla, 247, f: pycnostachya, Fraxinus pennsylvanica campestris, 266 ; Gaura coccinea parviflora, 267, var. parviflora, 267 Gentiana nigricans, 299 Glyceria repanda, 295 Hedysarum, 217-220, 236, 237; rag a 226, 227; alpinum, 219-225, 229 235, num, 222, subsp. ciecleaniin, 222, var. americanum, 222, 223, 225, 238, f. albiflorum, 222, var. difio: orum, 222, 223, 22. var. japonicum, 227, var. philos- cia, 222, 224, 225; americanum, 229, 224, 238, Mackenzii, 237, 239; boreale, 219, 221, 222, 224, 225, » 232, 235-238, var. albi- var. : marginatum, 228, 229; § Multicaulia, 219; Obse cura, 219; obscurum, INDEX occidentale, 219, 221, 227, * 233 6, 9; uintahense, 228, 229; utahense, 231, 233, 235, 236 Helonias angustifolius , 254 Hemerocallis ilio-Asphodelus, var. Hoary Cress, 3 Houstonia, wa tere ya oe 299, , , 299 Hydrcodtyie. 207, pears nee asiatica, 295, 298, v r. Floridana Hymenophysa, 304; fenestrata, 306; macrocarpa, 306; pubescens, 302, 304, 305 Juncus arcticus, 244; effusus, 244, var. a., 244 Lactuca, 243, 300, 301; biennis, 300, 3 : , var f. integrifolia, 302, var. integrifolia, 302 Lamium m= um, 268 Leguminosae, 218 Levee "302-3045, Draba, 302- 04, var. repens, Limnanthemum cracihioesiciiall spice dioica, 292 ge bois npn, 259, rota 259, var. typica, 259, 260 Magnolia acuminata, 244; tripeselss 4; virginiana, , €& acum nata, 244, 6. foetida, 244, % glauca, 244, y. grisea, 244, 6. tri- petala, Melanthium densum, 254 osses, Muleedia um, 301; leucophaeum, 300, 301, 8. integrifolia, 302 Myosotis, 244 270, var. majus, 272, var. mie carpum, 270, 271, pl. 605, ¥ INDEX 309 minus, 272; palustris, 267; ter- restre, var. hispidum 273 Nymphoides, 298; aquatica, 298; cordatum, 298 Physolepidion, 303; repens, eo Phytolacca americana, 24 ro Bs 244; decandra, 244; emeldne 244 Pinella, Potamogeton Berchtoldi, 246, var. colpophilus, , var. lacuna- tus, 246, var. polyphyllus, 246, ; tenuissim tus, pusillus, 246, var. ealnoutilen: 6, var. tenuissimus, oe rv autumnalis, 300 Prunella, 24 Pyrola, 33 Quercus ilicifolia, 292; prinoides, 292 Radicula palustris, 267, var. micro- carpa, 271 Ranunculus aquatilis, 268 Rhododendron roseum, Rhus Toxicodendron Negundo, 267 Rorippa barbareaefolia, 270, 273, 274, pl. 605; hispida, 272, 274, var. Fernaldiana, 268, 270-272, 274, pl. 605, var. glabrata, 270, 272-274, pl. 605, 268 270, 272- 274, 75; tomen- Rubus abbrevians, 287-289; (subg. Eubatus) § Abbrevi iantes, 289; (subg. Eubatus, § a rmes) erus, 2 > pis. 610, 611; (subg. "Eubati § ; 8. 288-290, pls. 612-615; (subg. Eubatus, § Tholiformes) adja- 292 2, 290— 616 618; _ (subg. Eubatus), § Alle- gheniens 293; ns, 289, 294; argu re 294; biformis 294; a 289; Boyntoni, 293; (subg. Eu- batus, o7 Tholiformets) bracteo- liferus, 293, pls. 619, es caesius, 276; canadensis, 236, 86; cubitans, ba: frondisentis, 283; x setosus, 291, 292; jacens, 282, 92, (subg. Eubatus), aevior, 2002 ok ul- 85, 286; novanglicus, 282; nutkanus, 278, ari- us, 275, var. iateradenten, hor 293, 294, 622; florus, 275, 281, var. bifarius, 275, var. genuinus, 275, r. grandi- oe 275, var. 5 ianlaaiia, 5; malac us, 275, var. ut- tli, oblige var. parvifo lius, ee, 276, var. scopulorum, 275, Selasiiea, “O75; parvitolion. 281; rmixtus, 282, 292, 295, var. aevior, 295; (subgen. ee Persis tentes 6, § Triviales, 276, § Setosi, 350, setosus 287; severus, 282; (subg. Euba- tus, 3 ive mes) 8} spiculosus 2 ls. 608, 609 * tardatus, liformis, 286, p 607; — 292; velutinus, 275: vigil, 2 Sassafras, 292; albidum, 292 Scirpus carolinianus, 246, 247; pu- berulus, 246, 247 Silene carolinianay subsp. pensyl- vanica, msylvanica, ~ 260, s ubsp. Wherr rryi, i, 260, var. Wherryi, 60; pensylvanica, 260; erryi, 260 Sison rifolintuty, 298 Sisymbrium, 244; brachycarpum, 266; canescens, var. apo arell pum, sir Bag eat door 267; natum, brachycarpum, 566: Richardson, 266; Rich: ardsoni- Seseah cis “BOL: alpinus, 302; biennis, 300-302; 300; m mul- tiflorus, 301; pallidus, 301; race- 310 INDEX mosus, 300, 301; spicatus, 300, 240, subsp. pallida, 240; sitchana, 3 256-2 59, var. Bongardiana, 255, Spergulastrum lanceolatum, 257 na, 254, 255, var. Biri Tetrapoma pyriforme, 273 bunda, 255, var. isophylla, 255, Thaspium, 298; trifoliatum, 298 var. Simcoei, 255, var. sitchana, Thlaspi, 255; calycantha, 254-259, var. Tracyanthus angustifolius, 254 , 255, 256, 258, pl. go ar orfbunds, od 'age, _Yeretrum luteum, 24 : Bhs , var. isophylla, » f 254. 206, 256, 258, pls. 602, 603, Viscaria alpina, f. 8. albiflora, 259 var. laurentiana, 254, 257, 258, 292 . 602, var. Si “ i 256, 258° Xanthoxylum americanum, var. sitchana, 255-258, pls. 602, es angustifolius, 254; den- 603; media, 240, subsp. eumedia, 8S, 254 240, subsp. neglecta, 240, var. "098; aurea, var. Bebbii, 298; macropetala, 240, var. pubescens, aapebbit 298; trifoliata, 298 g CONTRIBUTIONS Reprinted from Ruopora, Vol. 42, No. 501, September, 1940. CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY CXXXII A LIST OF PLANTS FROM INTERIOR ALASKA EpiraH SCAMMAN CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY—NO. CXXXII A LIST OF PLANTS FROM INTERIOR ALASKA! Epita SCAMMAN Or the great territory of Alaska—in extent equal to one-fifth the size of the United States—the Interior has received little attention from botanists. In Southeastern Alaska, the Aleutian Islands, and the Bering Sea region, many collections have been made and reports published, since the earlier years of Russian occupancy. Ledebour’s “Flora Rossica’’ remains today a most valuable work for all students of Alaskan plant life. But the vast Interior, cut through by the Yukon River—a natural highway for the gold seekers and adventurers of early days—and its drainage system, including the valleys of the Tanana and the Koyukuk Rivers, with Mt. McKinley Park as the southern boundary, contains & great wealth and variety of plants. Much of the region is un- glaciated, except locally, and has, therefore, afforded a safe and secure home in which plant species have lived undisturbed for countless ages. On the mountains of this old region may be found growing in close proximity arctic-alpine plants, many circumpolar, cordilleran species of the Rocky Mts. and British Columbia, and Asiatic types, “transgressing” seven or eight hundred miles inland from the coast of Bering Sea. The distribution of plants in the Interior is, there- fore, of considerable interest, and this list includes new localities and extensions of range for many species. ' Cost of publication defrayed by the author. 310 Rhodora [SEPTEMBER The summers of 1936 and 1937 were spent by the writer in Alaska for the purpose of collecting specimens for the Gray Herbarium. Small collections were made in Juneau and Seward, but are excluded here, as this list covers only vascular plants of the Interior, obtained from a number of different localities: Mt. McKinley National Park; Black Rapids, on the Richardson Highway, extending from the coast to Fairbanks; Fairbanks; Miller House, Eagle Summit, and Circle Hot Springs, all three near the Steese Highway, the only road from Fairbanks, the metropolis of the Interior, to Circle City on the Yukon; and Wiseman, the “ Arctic Village” on the Middle Fork of the Koyukuk. The stations in this area extend from McKinley Park, 63° 43’ no. lat., to Wiseman, about 67° 30’ no. lat. Mr. McKIntry Park Two visits were made to the Park, one of three days only, July 5-7, 1936, and a longer period, June 13-22, in 1937. Spring was late in Alaska in 1937, and when I reached Savage River Camp, twelve miles from the railroad entrance to the Park, on June 13th, practically nothing was in bloom. From a collector’s point of view it looke quite hopeless. But several days of warm weather with sunshine much of the night, causing underground ice to thaw and surface snow to melt, resulted in an almost unbelievable transformation. The stony banks of the Savage River, with swamp sand beyond, and open, parklike woods behind the Camp, were masses of brilliant color— Silene, Anemones (four species), Delphinium, Aconite, Papaver, Parrya, Drabas, Saxifragas, Dryases, Potentillas, Rhododendron, Lupine, Dodecatheon, Mertensia, and Polemonium. As I was eaget to get farther north, I remained in the Park only long enough to collect 115 species. The majority of these were obtained within two or three miles of the Camp. A day’s trip to the Rangers’ cabin at Igloo, Sable Pass and Polychrome Pass yielded a few less common finds—V iola biflora, Draba alpina, Potentilla nivea and, on a rock pile near Sable Pass, Astragalus falciferus and several critical species of Oxytropts. A clump of pinkish-purple Douglasia in bloom at Polychrome Pass ws the chief reward for the days spent at the Park. Rapips An impressive demonstration of the forces of nature was staged at Rapids on the Richardson Highway between my two visits there, August 7-10, 1936, and August 25-28, 1937. The Hunting Lodge 1940] Scamman,—List of Plants from Interior Alaska 311 situated on the Big Delta River, 138 miles south of Fairbanks, on the theastern slopes of the Alaska Range, elevation 2130 feet. In 1936 a small glacier could be seen from the Lodge, merely as a glittering spot of ice high up in the mountains. A year later it had become a “runaway” glacier, had moved downward five miles, taking shrubs and all other vegetation with it, and stopped, a great wall of ice 300 feet high, on the opposite bank of the Big Delta. On a dry, rocky bluff near the river a new variety of Lesquerella arctica grew in abundance. Tiny, low-growing Sazxifraga Eschscholtzit, and Woodsia alpina and Crepis nana were also collected at Rapids, with other plants one would expect to find in the two habitats—spruce and birch woods bordering the swift-flowing glacial streams, and the mountain-sheep “pastures” at higher altitudes above the tree line. MILLER House A brief acquaintance with the plants of Eagle Summit, seven miles from Miller House, in 1936, made me realize the fine possibilities for collecting arctic-alpine plants there. So I returned the following year, happy to be back with my kind, hospitable friends, Mr. and Mrs. Frank Miller. There in a tiny log cabin next door to the road- house, post-office, and general store—all three in one—I remained from July 2 to 26. In the center of the “ Circle diggings,”’ 116 miles north of Fairbanks, this roadhouse is located in a valley, through which flow several mining creeks, Mammoth, Mastodon, and Miller. Plants typical of the Interior valleys of an elevation of about 2000 feet were found here. Large clumps of the rusty-backed swords of Dryopteris fragrans shared the dry hillsides above the creeks with Arctostaphylos Uva-ursi and A. rubra, Vaccinium Vitis-Idaea, var. ee (Mountain Cranberry), Empetrum nigrum, Pedicularis labra- dorica, Geocaulon lividum, Silene repens, Arnica attenuata, Saxifraga ricuspidata and Arenaria obtusiloba. In the open tundra below were white pools of Eriophorum with Carices, Andromeda polifolia, Vaccin- tum uliginosum, Rubus Chamaemorus, Polygonum viriparum, and the Straight pink plumes of P. Bistorta. Parnassia Kotzebuei and P. palus- iris, var. neogaea, Linnaea borealis, and tangled Hedysarum alpinum, var. americanum were sheltered by willows and alders in moist shady FR Flat patches of Arenaria physodes were abundant, and delicate Cr eprs elegans, with long tap-root, penetrated the spaces between the Piles of round stones thrown up by mining operations. Along the 312 Rhodora [SEPTEMBER lower part of the trail from Miller House to Eagle Summit shrubs and tall, weed-like plants of the familiar roadside association flourished luxuriantly—Rubus idaeus, var. canadensis, Potentilla fruticosa, Rosa acicularis, Senecio lugens, Aster sibiricus, Solidago multiradiata, 8. de- cumbens, Polygonum alpinum, var. lapathifolium and Epilobium angus- tifolium. The latter is the most striking and conspicuous plant in the Interior, covering the low hills around Fairbanks with masses of brilliant color, visible for miles. EAGLE SUMMIT An Alaskan “travelogue” describes Eagle Summit, elevation 3880 feet, as a barren summit, the highest point on the road between Valdez on the southwestern coast to Circle City on the Yukon. The extreme top is level, with disintegrated rock fragments, dry and bar- ren, indeed. In winter it is bleak, wind-swept, and bitterly cold, and is considered one of the most difficult places to cross by dog team. But in July all the cave-like hollows and wet slopes form an alpine rock garden of great beauty. Large snow patches remain nearly all summer between the slopes—a favorite place for caribou to seek refuge from mosquitos and flies, as well as from the heat. The borders of wet, springy ground where water trickles down from the snow afford a perfect home for Ranunculus nivalis, also Claytonia sarmentosa and C. tuberosa, Senecio Kjellmanii, Sazxafraga Hirculus, S. foliolosa, Dodecatheon frigidum, Parrya nudicaulis, Cardamine purpurea, and numerous others. Slightly higher in the rocky crannies grow different species—tw° dwarf Salices, Oxyria digyna, Arenaria arctica and Arenaria macrocarpt, Silene acaulis, var. exscapa, Ranunculus pygmaeus, Anemone narciss'- flora, Cardamine bellidifolia, Saxifraga rivularis, S. punctata, S. bronch- ialis, subsp. Funstonii, Dryas octopetala, Cassiope tetragona, Gentiana glauca, Arnica Lessingii, Saussurea densa, and Antennaria alaskana. Five species of Pedicularis were close neighbors in spongy ground near the base. ; It was a delight to discover several rarer plants—Campanula wn flora, inconspicuous and easily missed; fragrant and charming Erir- chium aretioides; a pink and white Papaver, Oxytropis Mertensiana, the first record for the Interior and third for Alaska; and Eutrema Ed- wardsii, known usually from the northern coast. Here on the slopes of Eagle Summit circumpolar, Asiatic or North Pacific, and cordillera® species lived side by side. 1940] Scamman,—List of Plants from Interior Alaska 313 PorcuPiInE DoME The red letter day of my two Alaskan summers came on July 12, 1937, when Mr. Miller and I climbed Porcupine Dome, the highest peak, elevation 4810 feet, of the whole region between Fairbanks and the Yukon. Wearing high rubber boots, our faces hidden by mosquito netting, and armed with a gun, for protection against unfriendly bears, we made slow progress across the “niggerheads.” Our hike ended with a climb almost on hands and knees over the massive rock-pile to the flat plateau which formed the top of the Dome. There growing beside Salix phlebophylla and Antennaria alaskana, was a delicate little Potentilla, glabrous, with three-foliolate leaves. It proved to be otentilla elegans, a rare plant of the mountains of Siberia, reported but once before on the North American continent, on the coast. Here it was more than 600 miles inland. It bears a close resemblance to a rare endemic of the White Mountains. Potentilla biflora, found occasionally on the northwest coast, but rare in the Interior, had been collected on the way up. Ata lower altitude in wet, springy ground below a snow patch were a few plants only of a single-flowered, purplish-magenta Claytonia—very character- istic and unlike any Claytonia I had seen either in the field or in the herbarium. This new species—recently described by Dr. Eric Hul- tén—and Potentilla elegans were enough thrills for one day, even though we missed meeting a grizzly. CrrcLte Hot Sprines This is the chief resort of Interior Alaska, 9 miles southeast of the Steese Highway and about 138 miles north of Fairbanks. The hot baths of healthful mineral water are very popular, as are also the fine vegetables grown in the neighborhood of the springs in the hotel garden. Plants with a more southern range are found here, growing luxuriantly around the springs, in spite of the nearness to the Arctic Circle. I spent only several days at the Hot Springs in 1936 (July 17-22), but was repaid by finding Juncus filiformis, not previously reported from Alaska, and northern extensions of range for several Plants. Caltha natans, of rare and local distribution, grew in ditches with Ranunculus Purshii, subsp. yukonensis. WISEMAN Wiseman is a small village of Eskimos and gold miners, on the Middle Fork of the Koyukuk, about 75 miles north of the Arctic 314 Rhodora [SEPTEMBER Circle. It is reached by a plane trip of 200 miles from Fairbanks. One night in 1936, July 12, I flew there and back to view the midnight sun, picking up a few common plants in the town. But most of my specimens, the first collected or recorded from Wiseman, were obtained from August 2-12, 1937. It was late in the season for the best col- lecting, however. The river valley and ravines are well wooded with Picea, Populus tacamahacca and Betula papyrifera, with Populus tremuloides on higher altitudes in the hills. Juniperus communis, var. montana is occasional and Betula glandulosa very common, as it is everywhere throughout the Interior. Three Orchids grew abundantly in the woods about the town—Cypripedium passerinum, Habenaria obtusata and Habenaria hyperborea. Spiranthes Romanzoffiana was occasionally found, prob- ably the most northern station on record. Boschniakia rossica, called “corn pipes” by the Eskimo children, is very plentiful. Masses of Dryas Drummondii, with beautiful feathery styles, lined “Main Street,” and Epilobium latifolium gave color to the river bars. unusual “finds” were true Oxytropis deflexa, the second station in Alaska; the rare Antennaria pulcherrima, and Artemisia alaskana. This is the only report of the latter, with the exception of the type specimen from the lower Yukon. Most of the Wiseman plants are noticeable for their unusually tall growth, due probably to the rapid forcing by so many hours of sunlight. A few specimens, labelled “ Along the Yukon River,” were collected when the steamer stopped to load at various wood-piles between Tanana and Eagle, August 20-27, 1936. Astragalus yukonis from Tanana is the most noteworthy. The material from Fairbanks is scanty and not representative, as I gathered plants there only while waiting to go on to other places, 5° missed some, well-known and common. I have also included several from Curry, on the Alaska R. R. between Anchorage and the Park, too far south to be classed as truly in the Interior. While in Alaska I appreciated the courtesy shown me by Professor Gasser of the University at Fairbanks, Mrs. Ada Sharples of Juneau, author of a popular Flora of Alaska, and Mr. J. B. Anderson, who kindly offered me while in Juneau, the use of his private herbarium, containing the largest and finest collection of Alaskan plants in the Territory. 1940] Scamman,—List of Plants from Interior Alaska 315 My deepest appreciation is expressed to Professor Fernald for his generous encouragement, kindly interest and helpful guidance, with- out which this study of flora of the Alaskan Interior could not have been undertaken. On a visit to the Gray Herbarium in 1938, Dr. Eric Hultén of Lund looked over many of my specimens, giving me valuable aid in their identification and record of distribution. He also took several critical species back to Sweden for later study, and has since kindly described and named the new Claytonia. “Contributions to the Flora of Alaska,” by A. E. Porsild, of Ottawa, in Rxopora, 1939,! has been a constant guide and inspiration. To him I am greatly indebted for identifying and annotating a number of my puzzling specimens, and for his generous and helpful advice in regard to this list. It is a pleasure to express my gratitude to Mr. C. A. Weatherby for his assistance in the classification of the Pteridophyta; to Dr. Hugh M Raup, the Salices; to Mr. Reed Rollins, the Cruciferae; and to Dr. G. Haglund, the Tarazaca. List or VascuLarR PLANTs COLLECTED IN INTERIOR ALASKA (407 species and varieties, 847 numbers) I. PTERIDOPHYTA Borrycatum LuNnarIA (L.) Swartz. Not common and seen only once. It may easily have been overlooked, however. Rapids, no. 28. Woopsta ttvensis (L.) R. Br. One xposed dry cliffs above the Susitna River. Curry, nos. 2 and 565 W. atpra (Bolton) S. F. ied Crevices in rocks overhanging Gunnysack Creek. Rapids, n Plants in this collection ae piensa stipes, as is often the case in typical W. alpina of the Old World, which in all other charac- ters they resemble. W. ctapetia R. Br. Frequent around Wiseman on moist rocks and in damp, mossy hollows shaded by overturned tree roots, no. 871. Cys STOPTERIS FRAGILIS (L.) Bernh. Abundant in shaded ravines. ge oe mo pontan fern grows luxuriantly in ripe places. Park, he Rapids, no. 4; Circle Springs, nos. 3 a TERETIS NODULOSA (Michx.) Nieuwl. Pisialen ‘nodulosa Michx. In alder thickets along the Susitna River, oe reaching the eight of = feet. Curry, no. 564; Fairbanks, n Drropreris rracrans (L.) Schott. Thelypieri fragrans “-) Niet. Apparently a most common fern of the north-central In- ' Porsild, A. E., Riopona 41: 141-183, 199-254, 262-301 (1939). 316 Rhodora [SEPTEMBER terior. At Miller House hundreds of plants grow on dry, stony hill- Sena nos. 15 and 700; Park, no. 574; Circle Springs, no. 14; Wise- man, no. 872. D. pas eile (O. F. Mill.) Watt, var. pruatata (Hoffm.) Watt. See Fernald, Contrib. Gray Herb. 76: 147 (1926), for discussion of D. spin- os d D. austriaca. Common in southern iene ras the coast, but rare in the central Interior. Circle Springs, D. sprnutosa (O. F. Miill.) Watt, var. AMERICANA Ooch, ) Fernald, in Ruopora 17: 48 (1915). Rare. Circle Springs, no. 16. Scales on stipe and a few on rachis, pale brown, concolorous. D. Puecorteris (L.) C. Chr. Phegopteris wid acanensy Fée; The- lypteris Phegopterts (L.) Slosson. Common i oe Alaska but rare in the Interior. Rich woods by the river. Cine D. Linnazana C. Chr. Phegopteris Dryopteris (L.) Wee; rile teris Dryopteris (L.) Slosson. Rare in cool, moist woods. Circle Springs, no. 10; Wiseman, no. 913. These specimens have a few glands on the rachis, and belong with forma GLANDULOSA Tryon in Fern Jour. 29: 4 (1939). ATHYRIUM FILrx-FEMINA (L.) Roth ex Mertens, var. SITCHENSE Rupr. ex wom Abundant and of tall, luxuriant growth near the hot springs. Circle Springs, no. 19. Forma stricruM (Gilbert) But- ie o. 20. SETUM ARVENSE L. Common and widely distributed. Fair- bank no. 34; Miller House, no. 30; along Yukon River, no. 31; Wise- man, E. eet Ehrh. In ie and poplar thickets along banks of ger a Fairbanks, no. i E. sytvaticum L. See F yeah in Ruopora 20: 129 (1918). Often in woodland. Park, no. 576; Fairbanks, no. 33; Circle Springs, no. 32. . PA eee a Abundant on mud bars and in shallow water. Fairbanks, no. 860; Wiseman, no. 873. FLUV say = L. In sicsigh and ponds. Fairbanks, nos. 34-2, 860-2; ee House, no. 702. E. varrecatum Schleich. Occasional along borders of creeks. Miller ‘fon no. 701; Wiseman, no. 874. Lycopoprum Sezaco L. Frequent on mossy ledges in alpine situa- tions. ~ no. 577; Rapids, no. 1044; Eagle gana: nos. 35, 7 L. AN sit L. er rich woods. Rapids, L. comptanatum L. Circle Springs, no. 41; in open ai? higher slopes, Wiseman, no. 875. SELAGINELLA SIBIRICA (Milde) Hieron. See — FI. Aleut. Isl. 62 (1937). On dry, bare cliffs. Wiseman, no. 1940] Scamman,—List of Plants from Interior Alaska 317 II. SPERMATOPHYTA GYMNOSPERMAE Picea GLauca (Moench) Voss, and PICEA MARIANA (Mill.) B.S.P. constitute the coniferous forests of the Interio ARIX cA, (DuRoi) Koch. Cocoa! in swamps. Fair- banks, no. 43. Fonvaite COMMUNIS L., var. MONTANA Ait. J. sibirica Burgsd. On es , gravelly slopes, not common. Park, no. 579; Wiseman, no. 879, ANGIOSPERMAE MOoNOCOTYLEDONAE TRIGLOCHIN PaLUsTRIS L. Common in marshes and along the rivers. Fairbanks, no. 46; Wiseman, n HALARIS CANARIENSIS £. Introdu ced. Fairbanks, no. 1087. HIEROCHLOE oporaTa (L.) Wahlenb. Park, no ; H. auprna (Sw.) Roem. & Schult. On alpine summits. Park, no. 581; Eagle oe no. 706. ARCTAGROSTIS ARUNDINACEA (Trin.) Beal. In the hills above Wise- man, no. 881. A. Latrrota (R. Br.) Griseb. Miller oo no. 709. Agrostis scaBra Willd. Circle Springs, no. 48. ALAMAGROSTIS CANADENSIS (Michx.) Nuit. ., var. LANGSDORFI (Link) Inman. GC. Lan gsdorffii Trin. See Ruopora 24: 143 (1922), and 32: 43-44 (1930). Very common. Rapids, nos. “es 1049; Miller House, no. 710; Wiseman, no ESCHAMPSIA CAESPITOSA (L.) Beauv. Fairbanks, no. 50. TRISETUM SPICATUM (L.) Richter. Miller Hous se, no. 53. BEcKMANNIA SyziGAcHNE (Steud.) Fernald in yaat ah 30: 27 (1928). Widespread and common. Miller House, no. 707; Circle 883. oO. Poa apina L, Stony soil. Park, no. 604. P. arctica R. Br. P. rigens Hartm. In alpine situations. Eagle , 71 o. 708. 1a CEFCERIA GRANDIS Wats. In sein abe: the Chena. Fair- Deora ruBRa L. Varies greatly. Rapids, no. 1050. F. auraica Trin, Miller House, no. 705 ited dy d cabins. cbr hee no. ca Scribn. In dry ground around cabins 318 Rhodora [SEPTEMBER ERIOPHORUM OPACUM (Bjérnstr.) Fernald in Ruopora 7: 85 (1905); 27: 203-10 (1925). Very common in swampy ground. This and the following species form large “niggerheads” in the tundra. Park, no. 587; Circle Springs, no. 65; Wiseman, no. 888. E. VAGINATUM L. iller House, n: 13. E. catuitrrx Cham. Occasional in bogs and — the edges of small ponds. Miller House, no. 712; Circle Springs, n E. Scneucnzeri Hoppe. Miller House, no. 711; Wiseshaie no. 887. E. mepium Anders. E. Chamissonis C. A. Meyer, var. albidum sensu Fernald. Rapids, no. 1048. E. ancustiroLium Roth. Widespread and abundant in wet places. Miller House, no. 714; a larger form, no. 715; Circle Springs, no. 64; Wiseman, no. 889. ScrrPus AMERICANUS Pers. Circle Springs, no. 61. CAREX CAPITATA L. re. In swamp near Fairbanks, no C. pISPERMA Dewey. C. tenella Schkuhr. Beside creek at Miller House, no. 718. C. BRUNNESCENS Poir. Fairbanks, no. 693. C. suptnaA Wahlenb. Rare. Rapids, no. 1045. C. scrrpoipEA Michx. Common. Park, no. 585; Rapids, no. — 1046; Circle Springs, no. 69; Wiseman, n C. scrrPOIDEA Michx., var. CONVOLUTA Kikenth. Miller House, no. 719. C. concinna R. Br. Park, n C. caPrLuaRis L. Cicenaanak. on bah banks. Rapids, no. 72; Wiseman, no. 886. C. ANGARAE Steud. Synops. Cyper. 190 (1855). C. og Schkuhr, var. inferalpina sensu Fernald in Ruopora 35: 220-223 (1933), non Wahlenb. Very common. Rapids, no. 1047; Miller House, no. 717; Circle Springs, no. 75. C. srytosa C. A. Meyer. Wiseman, no. 992. Probably first record for the Interior. C. popocarra R. Br. Park, no. 582; Eagle Summit, no. 720-A. C. Totmter Boott. Park, no. 586; Miller oe a 722. C. ATROSQUAMA Mackenzie. Miller House, no. C. aquatitis Wahlenb. Common in swamps a gree creeks and re Miller House, no. 716; Circle Springs, no. 80; Wiseman, N0- 3 C. pHysocaRPA Pres]. Miller House, no. 724. C. MEMBRANACEA Hook. C. membranopacta Bailey. Comm Park, nos. 84, 584; Rapids, no. 85; Miller House, no. 723; Circle Springs, no. 83; Wiseman, no. a RoTUNDATA Wahlenb. In low, marshy ground. Wiseman, 10- C. ROSTRATA Stokes. In a ditch in Fairbanks, no Ncus BuFoNIus L. Frequent along paths or penree Fair- banks, nos. 89, 1085; Circle Springs, no. 88. 1940] Scamman,—List of Plants from Interior Alaska 319 J. BaLticus Willd., var. Hazenxu (E. Mey.) Buch. Common in wet places. Park, no. 591; Fairbanks, no. 90; Wiseman, no. 998. J. ritirormis L. Rare. The first station for this slender Juncus reported in Alaska. Circle Springs, no. 93. J. aupinus Vill. Fairbanks, no. 94. : : - CASTANEUS Smith. Abundant and widespread in the Interior. The most common Juncus in the places I visited. In marshy ground. Fairbanks, no. 96; Miller House, no. 726; Circle Springs, no. 95; Wiseman, no. 999. LuzuLa PARVIFLORA (Ehrh.) Desv. Alpine slopes. Park, no. 589. L. conrusa Lindeb. Also an alpine species. Miller House, no. 729 i MULTIFLORA (Retz.) Lej. L. campestris (L.) DC., var. multiflora (Ehrh.) Celak. Park, no. 588. Var. FRIGIDA (Buch.) G. Sam. ris, var. frigida Buch. See Hultén, Fl. Aleut. Isl. 125 T. coccinga Richards. Found occasionally in higher, more alpine situations, Rapids, no. 103; Eagle Summit, no. 731. : YGADENUS ELEGANS Pursh. Abundant in grassy plains and thin, pen woods. Rapids, no. 106 (in fruit), no. 1005 (in flower); Wise- man, nos. 105, 891. ey LLIUM SCHOENopRASUM L., var. stprricuM (L.) Hartm. A. sibir- tcum L. Occasional on sandy shores. On banks of the Yukon, Rampart, no. 107. ss _ Luoypta serorina (L.) Reichenb. This delicate little “ Alp Lily 1s Tare in the mountains. A good-sized colony along the rocky shore of Savage River, Park, no. 592. A single plant was seen in a crevice at Eagle Summit. Tris serosa Pall. A tall, beautiful purplish-blue Iris, abundant along roadside ditches and on marshy ground in Fairbanks, no. 110 (in fruit), no. 688 (in flower). YPRIPEDIUM PASSERINUM Richards. Abundant in spruce woods along the Koyukuk River, Wiseman, no. 892. As I was there in August my specimens were all in fruit, but the following year Tom Brady, a miner of Wiseman, sent me several in flower, no, 892-A. TTATUM Swartz grows in Fairbanks. I saw mounted speci- ba mens there, but did not collect it myself. ? HABENARIA sp. These plants, growing in marshland in Wiseman, no. 894, seem to 320 Rhodora [SupTeMBER belong to the Habenaria hyperborea or H. dilatata group, as the lip is entire, about 5 mm. long, broader at the base and slightly dilated. Basal leaves are rather wide and obtuse, and the flowers greenish- white. H. OBTUSATA (Pursh) Richards. Occasional in rich woods. Rapids, nos. 112, 1006; Wiseman, no. 893. SPIRANTHES ee Cham. & Schlecht. Rather rare and local. Circle Springs, no. 113; Wiseman, no. 895 Several plants were found in moist ground on hills above the town. This is probably the most northerly station ever recorded for S. Romanzoffiana. Catypso BULBosA (L.) Oakes. A mounted specimen, collected along the Steese es Os gee was seen at the Tanana Valley Fair in Fairbanks. DIcOTYLEDONEAE PoruLus TREMULOIDES Michx. The aspen is common in dry soils bare asa the Interior. Here and there in the hills above Wiseman, 897. a ee Felarr an Oe, Miller. P. balsamifera DuRoi, not L. See Jour. Arn. Arb. 10: 55 (1929). Common on flood plains along creeks and rivers. Pes House, no. 114; on the banks of the Koyukuk, Wise- 896. mick RETICULATA L. One of the most common prostrate willows in the hills of the Interior. Park, nos. 115, 595; Rapids, no. 116; Wiseman, no. 903. The two latter belong to var. ORBICULARIS _S. arctica Pallas. S. crassijulis Trautv. Common in alpine re- pas Park, nos. 118, 594 (Savage River Camp), 600; Paxson, no. 1 S. ROTUNDIFOLIA Trautv. Eagle Summit, no. 735. This small willow with bright green, round leaves and short pistillate catkins is rare in the mountains. S. PHLEBOPHYLLA Anders. Growing in mats on bleak, bare moun- tain summits. Porcupine Dome, no. 736. The stems of this species are crowded with skeletonized leaves. S. PSEUDOPOLARIS Floderus. Rare. Eagle Summit, no. 117. 5. STOLONIFERA Coville. Eagle Summit, nos. 120, 73 A low-growing willow of alpine regions, characterized by many stolons. S. NiPHOCLADA Rydb. Frequent along the Richardson Highway 0? the banks of dace streams. Paxson, no. 121; Rapids, no. 1039; Fairbanks, no. 1084; Wiseman, no. 904. 1940] Scamman,—List of Plants from Interior Alaska 321 These specimens seem to belong to the general complex of S. brachy- carpa Nutt. and S. niphoclada Rydb. The leaves tend to become gray or spotted when dry, are rounded at the base, on short petioles. S. etauca L. A circumpolar willow which varies widely. The bisias Noles Valley representative was named S. Seemanti by Ryd- ee ony House, nos. 123, 733. cA L., var. ACUTIFOLIA Schn. See Bot. Gaz. 66: 327 (1918), and 67: "60 (1919); Ruopora 33: 241-4 (1931). Very common along streams. Paxson, no. 124; Rapids, no. fhetee Circle Springs, no. 125; Miller House, no. 126; Ft. Yukon, no ; Wiseman, no. 899. S. Barcuayr Anders. Paxson, n oie S. ALAXENSIS (Anders.) Cov., var. Lona (Rydb.) Schn. The felt-leaf willow is a characteristic common tree of valleys of the In- terior. Park, no. 597; Miller House, nos. 130, 732; Wiseman, no. 901 S. BEBBIANA Sarg. S. rostrata Richards. A shrub or small tree common in woods and along the banks of Shen Pr sloughs. Park, no. 593; Circle Springs, no. 131; Fairbanks, no. 1083. S. ARBUSCULOIDES Anders. Occasional. Pati no. 599; Miller House, no. . PULCHRA Cham. Apparently rather rare in the Interior. Park, 598. Myrica Gate L. Rare in this region. Along the Yukon River, no. 133. BrETULA PAPYRIFERA Marsh. The common white birch of the In- terior forests. Abundant and widespread. iller House, no. 738; along va Besse River, no. 134; Wiseman, no. B. GLANDULosa Michx. A vast portion of the tundra is oe by this diva ue Miller House, nos. 135, 739; Wiseman, no. uLosa Michx., var. srprrica (Ledeb.) Blake. ig no. 136; Wikemas, no. 907. . ERMANI X GLANDULOSA. Park, no. 601. ALNUS crispA (Ait.) Pursh. A. fruticosa Rupr. Abundant along the edges of creeks and streams. Rapids, no. 1037; Wiseman, no. 908. A. stnuata Rydb. Along river banks, near Fort Yukon, no. 137. GrOcAULON Livipum (Richards.) Fernald in Ruopora 30: 21-24 (1928). Comandra livida Richards. Common in sterile soil on hill- sides, often associated with Pedicularis labradorica. Rapids, nos. 141, 1015; Circle Springs, no. 140; Wiseman, no. 909. The leaves are frequently variegated. Rumex mexicanus Meisn. Along Yukon River, no. 142; Miller House, no. 740. R. arcticus Trautv. In bogs, Wiseman, no. 910. R. Acerosa L. On stony slopes, near Savage Camp, Park, no. 602. Oxyrta pieyNa (L.) Hill. Common in damp rocky crevices in the 322 Rhodora [SEPTEMBER mountains. Rapids, nos. 144, 1042; Park, no. 603; Eagle Summit, nos. 143, 741. PoLyGonuM AVICULARE L. Wiseman, no. 911. P. viviparuM L. Abundant throughout the alpine ett Park, no. 149; Eagle | Summit, nos. 150, 744; Wiseman, no. Pp; BrsToRTA L. P. plumosum Small. Scattered over a tundra, this plant from Asia with plumy rose-pink spikes, occurs less frequently than the preceding circumpolar, bulb-bearing ng Park, no. 147; Miller House, no. 743; Eagle Summit, no. 148, 745 P. auprinum All., var. LAPATHIFOLIUM Cham. & Schlecht. A tall, conspicuous plant, very common in thickets along roadsides and in gravelly places. Miller House, no. 746; Circle Springs, no. 151. It is called “ Wild Rhubarb,” and, when young and tender, is some- times used for greens. P. HyDROPIPEROIDES Michx. Circle Springs, no. 152. P. Convotvutus L. Occasional. Fairbanks, no. 1074; Circle Springs, no. 153. Cuenoraunte capiratuM (L.) Asch. Blitum capitatum L. Com- mon around cabins and in vacant lots. Fairbanks, nos. 154, 1076 . ALBUM L. A common weed. Fairbanks, no. 1077. CLAYTONIA Ener oe a C. A. Meyer. Montia sarmentosa Robin- son. rge clumps of this delicate, pink Claytonia grew in wet, spri ingy spots ie the base of Eagle Summit, nos. 156, 747. A small specimen (in bloom) was brought to me by an Eskimo in Wiseman, no. 914. C. rusERosa Pall. Occasional with the preceding at the base of the Summit, no. 748 The ie bile of this white-flowered plant are eaten by Eskimos and Indians. Both these Claytonias have been considered endemic to the North Pacific or to the Bering Sea region. C. ScaMMANIANA Hultén in Botaniska Notiser 4: 826-827, fig. 1 (1939). Porcupine Dome, July 12, 1937, no. 749 A few plants of this brilliant purplish-magenta Claytonia made 4 vivid patch of color at the base of a melting snowbank on the lower slopes of the Dome. I saw it only in this one spot throughout the region. It is very characteristic in appearance, low in growth, single- flowered, with long, narrow basal leaves. STELLARIA CALYCANTHA (Ledeb.) Bong., var. ISOPHYLLA Fernald. Not common. iller House, no. 750-A. S. CALYCANTHA (Ledeb.) Bong., var. FLORIBUNDA Fernald. Miller House, no. 752. S. cRAssiFoLIA Ehrh. Alsine anaes (Ehrh.) Britton. Occa- sional in sloughs. Fairbanks, no. 1940] Scamman,—List of Plants from Interior Alaska 323 S. Lonerpes Goldie. Alsine longipes (Goldie) Cov. Common throughout the region. Very variable. Park, no. 611; Eagle Sum- mit, no. 750; Circle Springs, no. 159. glaucous form with stiffer leaves described as var. LAETA (Richards.) Wats. Rapids, nos. 160, 1008; Miller House, no. 751. S. a Muhl. Alsine longifolia (Muhl.) Britton. Miller House, no. 752 Shes Sag (L.) "Cyril. Introduced. | Fairbanks, no. 158; Wiseman, 920. ; hoe viscosuM L. Introduced. Wiseman, no. 919. C. BeertnciANum Cham. & Schlecht. Very common throughout the region. Park, no. 607; Fairbanks, no. 1055; Miller House, no. 753; Eagle Summit, no. 163 C. BrErtnc1anum Cham. & Schlecht., var. GRANDIFLORUM (Fenzl) Hult. Fl. Aleut. Isl. 165 (1937). This larger variety, with petals much longer than the green sepals and acute leaves, is not as Sona. In wet places. Rapids, nos. 162, 1009; Eagle Camp, no. 698. SPER He aan RuBRA (L.) J. & C. Presl. Probably introduced. Fairbanks, no. 1067. Apiecite: LATERIFLORA L. Moehringia lateriflora (L.) Fenzl. Not very common, but widely distributed in woods and thickets. Curry, no. 571; Miller House, no. 760; Circle Springs, no. 164. A. pHysopes DC. Merckia physodes Fisch. This North Pacific species seems to be very plentiful wherever it is found in this region. In gravel and mining debris at Miller House, no. 755; Circle Springs, no. 166; Wiseman, no. 915. : DAWSONENSIS Britton. Rare. Along road at the base of Eagle Summit, no. 167; Wiseman, no. 918. amet _(Wablenb,) Sm. On stony ground. Park, no. 610; wae gee nos. yee A. Rossi “4 pie On damp slopes. Conspicuous for its reddish- purple sepals. Miller House, no. 168; Eagle Summit, no. A. OBTUSILOBA Se Fernald in Bacio 21: 12-15 (1919). Park, no. 609; apids, no. 170; in great clumps on dry cliffs and ledges back. of Miller Hgiek a 59. A. ARCTICA Steven. Minuartia Biri mca Aschers. & Graebn. ai no. 173; Eagle Summit, n 4, A. MA ACROCARPA Pursh. jase gs macrocarpa (Pursh) Ostenf. Park, no. 171; Eagle Camp, no. 697; Eagle Summit, nos. 172, 757. Both this species and the preceding were very abundant on the mountains in the Park and the Eagle Summit region. They are some- times confused, if not found in fruit. See Hultén, Fl. Kamtch. 2: 79-81 (1928) for descriptions of the two species. The stems of A. macrocarpa are matted-cespitose instead of tufted; the lower leaves are denticulate and faleate; the stem-leaves are connate-clasping and 324 Rhodora [SEPTEMBER very glandular; the sepals and petals are longer than and of different shape from those of A. arctica; the flat filaments widen more gradually to the dilated base. SILENE AcAULIS L., var. ExscaPa (All.) DC. This arctic species forms large cushions in the mountains of the Interior. Abundant in gravelly places along the Savage River. Park, no. 605; Eagle Sum- mit, nos. 175, 76 S. acauLIs L., var. SUBACAULESCENS (F. N. Williams) Fern. & St. John in Ruopora 23: 119-120 (1921). This western cordilleran variety is rare in the Interior of Alaska. It was found only in the Park, nos. 176, 606. : S. REPENS Patrin. Occasional on exposed slopes at altitudes of about 2000 feet in places where one finds Pedicularis labradorica, Arnica attenuata and often Dryopteris fragrans. Miller House, nos. 177, 762; on road between Miller House and Eagle Summit, no. 178. S. Wituiamsir Britt. in Bull. N. Y. Bot. Gard. 2: 168 (1901); see Porsild in RHopora 40: 212 (1938). In dry places on the banks of the Chena River, Fairbanks, nos. 179, 1056. Lycunis TayLortaz Robinson in Proc. Amer. Acad. 28: 150 (1893). See Ruopora 34: 22-25 (1932). Occasional in alpine regions. Eagle Camp, nos. 699, 764. L. apeTALA L. Melandriwm apetalum (L.) Fenzl. Seen only at the base of the higher summits. Eagle Summit, no. 180; Porcupine Dome, no. 763. Nupnar variecatum Engelm. Nymphozanthus variegatus (Engelm.) Fernald. Observed commonly in ponds and small lakes in the lower regions of the Interior. CaLrHa NaTANS Pall. This small white-flowered Caltha is probably rare in the Interior. Collected only once in a wet, boggy area with Ranunculus Purshii, subsp. yukonensis, Circle Springs, no. 181. It has an interesting distribution from northern Mongolia and Kamtchatka, in the Bering Sea region, with isolated stations 1 Alberta, Athabasca, and northern Minnesota. C. pavustris L., var. asartrotia (DC.) Huth. Very common I southwestern Alaska and observed in ditches along the R. R. tracks from McKinley Park to Fairbanks. Collected it at Seward, but neglected to get specimens from the Interior. Probably common 12 the lowlands. ACTAEA RUBRA (Ait.) Willd. A. spicata L., var. rubra Ait. _Appar- ently very rare in the central and northern Interior. Circle Springs» no. 183. Plants with white berries (forma NEGLEcTA (Gillman) Rob- inson—A. eburnea Rydb.) as well as red grew tall and luxuriantly “ thickets and thin woods in the neighborhood of the Hot Springs. Bo specimens in fruit. Circle Springs, no. 184. 1940] Scamman,—List of Plants from Interior Alaska 325 De.rnintum Menzies DC. Frequent in the mountains of the Interior. Eagle Summit, nos. 186, 765; on the way to Porcupine Dome, no. 766. 5 D. scopuLoruM Gray, var. GLAucuM Gray. D. Brownti Rydb. A tall plant often four or five feet high. Collected in two places only, but it appeared to be common in thin woods in lower regions. Rapids, no. 1014; Fairbanks, no. 185. ACONITUM DELPHINIFOLIUM DC. Widespread and common throughout the Interior, in thickets on hills and slopes of mountains. Rapids, nos. 188, 1013; Circle Springs, no. 187; Eagle Summit, no. 771; Wiseman, no. 922. ANEMONE PATENS L., var. WOLFGANGIANA (Bess.) Koch. Pulsatilla ludoviciana (Nutt.) Heller. My specimen from the Park, no. 613, was in fruit, but a beautifully pressed flower was given me by a young girl in Fairbanks. A well-known and beloved wild flower of the Interior, called gen- erally “Wild Crocus.” The very large purple blossom comes very” early in the spring as soon as the snow melts. A. PARVIFLORA Michx. One of the earliest flowers to bloom at Savage River Camp, in the Park, about the middle of June, no. 616. In Wiseman, no. 923, I collected both flowers and fruit the first week in August. Other collections were made at Rapids, no. 191, and on the road below Eagle Summit, no. 190. A very common Anemone of wide-spread distribution. The backs of the sepals are usually tinged pink or blue. A. PARVIFLORA, var. GRANDIFLORA Ulbr. in Engl. Bot. Jahrb. 37: 251 (1905). Park, nos. 189, 617. This variety, with the large flowers nearly two inches in diameter, was the most abundant at Savage River Camp. It seemed quite distinct, the sepals in my specimens lacking the bluish tinge. A. MULTIFIDA Poir., var. HUDSONIANA DC. A. globosa Nutt. See Rwopora 19: 14] (1917). Occasional in gravelly places. Rapids, no. 192, A. NARCISsIFLoRA L. A. zephyra A. Nels. One of the most beauti- ful of the Alaskan wild flowers, growing sometimes in large masses, in dry, rocky soil in the mountains of the Interior. General throughout the territ ory. Park, nos. 614, 615; Eagle Summit, nos. 195, 767. Often the plants of higher altitudes where there is little soil are single- owered, described by Eastwood as var. UNIFLORA. - Ricuarpsonn Hook. This low-growing bright yellow Anemone prefers moist, springy ground. Common along the southwestern Vr nv and also in the Bering Sea region, but occasional in the Yukon Valley. Park, no. 612 326 Rhodora [SEPTEMBER Ranuncuus Pursa Richards., ssp. YUKONENSIS (Britt.) Porsild. R. yukonensis Britt. in Bull. N. Y. Bot. Gard. 2: 168 (1901). Occa- sional along edges of ponds. On the mud of a dried-out swamp. Both flowers and leaves very small. Circle Springs, no. 198. R. nrvauis L. Park, no. 618; Eagle Summit, nos. 200, 769. True to its name this bright yellow buttercup is found often abundantly just below melting ice or snow patches in the mountains. A characteristic field mark is the brown- or black-hairy calyx. R. pyamarus Wahlenb. A small arctic-alpine plant, very rare in the mountains of the central Interior. Seen only on Eagle Summit, nos. 201, 770. R. sceLteratus L. In ditches near Ester Creek, Fairbanks, no. 202. THALICTRUM SPARSIFLORUM Turcz. In ditches along wooden side- walks in Fairbanks, no. 210. .. PapaveR ALASKANUM Hultén, Fl. Aleut. Isl. 190, tab. 10 (1937). Park, no. 619. In bloom the middle of June on a sunny exposed spot where the snow had just melted. These were low-growing plants with many old petioles, deeply dissected grayish-green leaves, well covered with stiff hairs, and pale yellow flowers with the central projection of the stigma absent, P. microcarpum DC. A tall, slender, large-flowered, deep yellow pepe frequent in the hills. Eagle Summit, no. 772; Wiseman, 20- One plant only, very unusual and distinctive, with white petals, bordered by a wide pink band, and with dark green leaves, was col- lected at Eagle Summit, no. 212. CorypaLis PauctFLora (Steph.) Pers. A low plant with several large pinkish-lavender spurred flowers, apparently common in the alpine meadows of the Park, no. 620, but rather rare elsewhere in the Interior. It grew also at Eagle Summit (field notes). C. SEMPERVIRENS (L.) Pers. Apparently rare. On a rocky slope behind the R. R. station. Park, no. 213. uTREMA Epwarpstt R. Br. Collected only near top of Eagle Summit, no. 774. This cireumpolar, arctic species, known usually from the norther? Coast and the Bering Sea region is rare in the higher mountains © the Interior. See Fernald in Mem. Gray Herb. 2: 337 (1925), ~ map of general distribution. Brassica campestris L. Introduced. Miller House, no. 218. 1940] Scamman,—List of Plants from Interior Alaska 327 RorIPPA BARBAREAEFOLIA (DC.) Kitagawa in Journ. Jap. Bot. 13: 137 (1937). In damp places near the mines. Miller House, no. 777. R. curvisttiqua (Hook.) Bess. Occasional. Park, no. 223; Rap- ids, no. 224. ARBAREA ? PLANISILIQUA C. A. Meyer. Not well developed. Curry, no. 569. CARDAMINE PRATENSIS L. Not common in the Interior. In damp thicket beside the Savage River, Park, no. 629; Wiseman, no. 1188. C. PRATENSIS, var. ANGUSTIFOLIA Hook. Park, no. 628. C. puRPUREA Cham. & Schlecht. This beautiful little arctic plant mountains of the Interior. Park, no. 627, (white-flowered form) no. 626; Eagle Summit, nos. 226, 779. é C. BELLIDIFoLIA L. Rare in the mountains. Eagle Summit, nos. 227, 780. oe C. BELLIDIFOLIA L., var. BERINGENSIS A. E. Porsild in Trans. Royal Soc. of Can. ser. 3, sect. 5, 32: 31 (1938). This sturdy, vigorous vari- ety with broader and shorter siliques was collected in damp ground on lower slopes of Eagle Summit, no. 781 (listed by Porsild in Ruopora 41: 234 (1939). All previous collections are from the islands and nearby shores of Bering Sea. New to the Interior of Alaska. LESQUERELLA ARCTICA (Wormskj.) Wats., var. SCAMMANAE Rol- lins in Am. Journ. Bot. 26: 421 (1939). A new variety of L. arctica, taller in growth, with long slender pedicels and leaves, grew abundantly on a dry, gravelly bank near Gunnysack Creek. It was collected both years in the same place, but not seen elsewhere. L. arctica is rare in Alaska, as, with the exception of specimens from the Bering Sea region, it has been reported only from the head of the Chitina River. The type of the new variety is no. 216, Rapids, Aug. 7-10, 1936. The second collection, Rapids, no. 1000. LASPI ARVENSE L. In waste grounds around roadhouses and cabins. Introduced. Rapids, nos. 217, 1002; Fairbanks, no. 863. EPIDIUM APETALUM Willd. Fairbanks, no. 864. i NESLIA PANICULATA (L.) Desv. Introduced. Fairbanks, no. 1069. CaPSELLA BURSA-PASTORIS (L.) Medic. Observed in several places about settlements, but failed to collect it. RABA ALPINA L. Seen only once. Park, no. 625. re. DR. NIvALIs Lilj. Delicate, tufted, in dry gravels near Savage River. Park, no. 621. : 91 : PRAEALTA Greene. A Rocky Mountain species. Park, nos. 4-B and 622, R. GLABELLA Pursh. See Fernald in Rnopora 36: 333 ae oo in alpine situations in Interior. Park, no. 214-A; Rapids, no. 215 Dr. BorEALIs DC. An Asiatic-Northern Pacific species. Park, 0. 623-A. n 328 Rhodora [SepTeMBer Descuratnia sopntomes (Fisch.) O. E. Schulz in Engler, Pflanzenr. 4: 105: 316 (1924). Common and abundant in settlements and mining districts. Park, no. 632; Miller House, no. 776; Wiseman, no. 926. ARabis LyraTa L,., var. KAMCHATICA Fisch. Very common at Sew- ard and along the coast where it was one of the first plants to bloom in the spring, but not seen as often in the Interior. Rapids, nos. 229, 1001 A. ryrata L., var. GLaBra (DC.) Hopkins in Ruopora 39: 93-4 (1937). “ie ere stony places. Park, no. 630; Rapids, no. ee A. Sie 8 A. Nelson. Probably rare in Alaska in dry sandy places. Along R. R. track from Anchorage to Curry, no. 568; P no. 633. ERyYSIMUM CHEIRANTHOIDES L. Very rem and of widespread distribution in the central Interior. Probably introduced. Fait- banks, no. 221; Miller House, no. 775; Circle Springs, no. 220; along Yukon River, no. 222; Wiseman, no. 925. Parrya Nupicauuis (L.) Regel. P. macrocarpa R. Br. In oper woods in the Park, no. 631; Eagle Summit, nos. 225, 778 A showy plant with large flowers varying from white and pink to lavender and purple. It is popularly called “Wild Stock” in Alaska and is fairly common in Alpine regions in moist tundra. There is considerable variation in the leaves and in the degree of glandular pubescence. SepuM Roseum (L.) Scop., var. INTEGRIFOLIUM (Raf.) Hultén, Fl. Aleut. Isl. 205 (1937). Rhodiola integrifolia Raf. See also Porsild in Ruopora 41: 239-240 (1939). Common throughout the Tecritay in moist mossy crevices. Park, no. 230; Circle Springs, no. 231. 1. Also seen at Eagle Summit, no. 783. BoykInia RicHaRpsontt (Hook.) Gray. This tall conspicuous plant is seen occasionally in damp saat aa in subalpine regions- Park, no. 233; Wiseman eo fruit), n SAXIFRAGA RIVULARIS L. In wet cals crevices in | hr mountains. — no. 235 (in fruit); Eagle Summit, nos. 234-A - ADSCENDENS L. One specimen resembles closely "this cordilleran wale Very rare in Alaska. Eagle Summit, no. 234-B. S. FOLIOLOsA R. Br. S. stellaris L., var. comosa Retz. ‘Most of the flowers on the delicate stem are replaced by tiny tufts of green leave In wet ground beside a small ditch. Eagle Summit, no. 791. . new to the flora of the Interior. S. HIERACIFOLIA Waldst. & Kit. Occasional in alpine tundra. Park, no. 638; Circle Springs, no. 236; Eagle Summit, no. S. REFLEXA Hook. Fl. Bor. -Am. 1: 249, tab. 85 (1840). 8. yukon- ensis Small. On a gravelly bank, Park, no. 637. Apparently rare, as it has been recorded in two other stations only in the Interior. ane ee oa Raa NN Alaa COIR Prt: pre net 1940] Scamman,—List of Plants from Interior Alaska 329 S. punctaTa L. sens. lat. Common on moist ledges in the Eagle Summit region, nos. 237, 238, 786; almost at the very top of Porcupine Dome, no. 787. This group is very variable. See Hultén, Fl. Aleut. Isl. 213-214 (1937) and Sv. Bot. Tidskr. 30, 3: 324, fig. 5 (1936). Most of my specimens have the inflorescence more open, dark purple capsules and purplish-tinged leaves and scape, as in ssp. insularis Hultén. S. TRICUSPIDATA Rottb. Common on dry ledges in all the subalpine regions where I collected. Sometimes forming large mats over rocks, and readily separated from the next species by the rigid leaves with three sharp teeth. Park, nos. 239, 636; Rapids, no. 1003; Miller House, no. 788; Wiseman, no. 929. - BRONCHIALIS L. ssp. Funston (Small) Hult. Fl. Kamtch. 3: 12-17 (1929). Fairly common also in stony places. Park, no. 636-A; agle Summit, nos. 240, 789; Wiseman, no. 930. S. Escuscuotrz Sternb. Rapids, no. 1041. This dwarf cespitose Saxifraga of Asiatic-Western American distri- bution is rarely found. Probably the second record for the Interior. S. FLAGELLARIS Willd. Occasional in the mountains of the Interior. In a swampy meadow in Park, nos. 241, 635; Wiseman, no. 931. The specimens in the two collections differ considerably. No. 931 has a much larger single flower with very long flagella. S. Hrrcutus L. Park, no. 242; Eagle Summit, nos. 243, 790. A common and beautiful Saxifrage, with large bright yellow flowers, the petals usually red-spotted, growing in wet boggy places in the mountains. S. oppositirot1a L. This well-known circumpolar species seemed to be surprisingly rare in the central Interior. Park, no. 634. ARNASSIA PALUSTRIS L., var. NEOGAEA Fernald in Ruopora 39: 310-312 (1937). Common and abundant throughout the region, eXx- cept in the higher altitudes. Park, no. 247; Rapids, no. 1004; Big Delta, no. 250; Fairbanks, no. 1070 (a freak form); Miller House, no. 793; Circle Springs, no. 249; Wiseman, nos. 248, 932. Prof. Fernald has described in detail, with plates, the differences between the American plant and the head form. Several of my speci- mens, especially those from Wiseman, seem to show transitional forms between the variety and the typical form of the species. P. Korzesver Cham. & Schlecht. Found often in damp thickets in sub-alpine regions. Miller House, no. 792; Eagle Mining Camp, no. 695; Circle Springs, no. 251; Wiseman, no. 1189. IBES TRISTE Pall. Occasional. Wiseman, no. 927. 330 Rhodora [SEPTEMBER PIRAEA BEAUVERDIANA Schneid. See Hultén, Fl. Kamtch. 3: 38- 41 10n0). S. betulifolia Am. auth., non Pall.; S. Stevenza Schat Rydb. Very common along roadsides, i in thickets, and dry tundra throughout. Miller House, no. 794; Circle Springs, no. 253; Wise. man, no. 933 RUBUS IDAEUS L., var. CANADENSIS Richardson. A delicious wild raspberry common in thickets of the Interior. Miller House, no. 796; Circle Springs, no. 254. R. CuamaEmorus L. The Cloudberry or Baked-apple eg is abundant in moist tundra and peat bogs throughout. Park, Miller House, no. 795; Circle Springs, no. 255; Wiseman, no R. acautis Michx. In moist ground near Savage River. Park, no. 639. R. arcticus L. Wiseman, nos. 935, 935-A. The specimens from Wiseman seem to belong here, although 2. arcticus and R. acaulis are confusing and critical in Alaska. Stems leafy, calyx-lobes densely pubescent, petals obovate and less clawed. FRAGARIA YUKONENSIS Rydb. A small-fruited strawberry with long runners, common on banks of sloughs. Fairbanks, no. 1072. OTENTILLA NoRvEGIcCA L. Common in low ground. Miller House, no. 798; Fairbanks, no. 260; Wiseman, no. P. ntveEaA L. See Hultén, Fl. Kamtch. 3: 68-69 (1929) for discussion of this and P. uniflora. Occasional. Park, no. 642; Rapids, no. 261. P. UNIFLORA ae On bare rocky places in the mountains. Not common. Par 641; Wiseman, no. 936. Pi? She eee ‘L. San dy banks of rivers. Rapids, nos. 262, 1010; Eagle on the Yukon, no. 263. P. Nurr Eagle, no. 264; Along Yukon River, no. 265. P. GRACILIS Dea New to interior of Alaska. Wiseman, no. 938. P. srrtora Willd. On the slopes of Porcupine Dome, no. 300. A tufted alpine Potentilla, very characteristic, with thick woody caudex, leaves with linear divisions, calyx-lobes and bractlets of a reddish tinge, as is also the stem, and pale yellow petals. According to Wolf it has a wide range from the Himalayas and mountains of Central Asia to arctic regions of eastern Asia and Western N. A., but is nowhere common. It has been collected at Cape Thompson, and several places on the Seward Peninsula, but this, a far inland station, is the first record for the Interior. P. ELEGANS Cham. & Schlecht. in Linnaea 2: 22 (1827); Ledeb. be Ross. 2: 56; Lehmann, Rev. Potentill. 53, fig. 1 (1856); Wolf, American continent. 1940] Scamman,—List of Plants from Interior Alaska 331 The finding of this dainty, low, tufted Asiatic Potentilla, with caudex crowded with persisting brown stipules, three-foliolate leaves, hairy calyx, and pale yellow petals, a little longer than the sepals, was a happy experience of the summer. At home in the mountains of Siberia and north-western Mongolia, it was collected once before by Thornton, in the Sawtooth Range on Seward Peninsula, and reported by Porsild, Ruopora 41: 246 (1939). Its closest relative is P. Robbin- siana Oakes, a rare endemic of the White Mountains, N. H. - PALUSTRIS (L.) Scop. Comarum palustre L. In wet places, , ho. 267. . FRUTICOS sa L. This shrubby Potentilla, known in the Interior as Tundr. ra Rose,” is one of the most abundant and characteristic plants in the region. In lowlands, beside roads, in dry tundra, and in sub- alpine locations. Fairbanks, no. 1071; Miller House, no. 799; Wise- man, no. 268. so common in Park (field notes). GeUM MACROPHYLLUM Willd., var. PERINCcIsUM (Rydb.) Raup in Ruopora 33: 172-176 (1931). Common around settlements. Fair- banks, no. 275; Circle Springs, no. 274. G. Rosext (R. Br.) Sér. Sieversia Rossii R. Br. A beautiful plant, abundant in wet, springy ground, in the higher mountains; often below snow pa atches, Eastern Asiatic-Western American. Park, is Rydb. in N. Am. Fl. 224: 387 (1908). Along the banks of rivers. Tanana, no. 285. Both Hultén and Porsild think the American plant should not be separated. S. sitcHENsIs C. A. Meyer. Not seen north of the Alaska Range. Beside a brook in the woods above Rapids, nos. 284, 1011 - ? Menziesu Rydb. Form with oblong spikes and purple calyx. In same places as above. Rapids, nos. 286 and 1011-A. Dryas Drummonptt Richards. Occasional in gravel bars along rivers. Park, no. 277. Abundant in large mats along the paths of the village, Spat with the striking feathery akenes, as it was in August, but several with the low-stemmed nodding yellow flowers. iseman, nos, 278, 941 R. OCTOPETALA L. Very common in dry, stony places in the mountains. Growing so clowdly § in thick mats that the ground looks as though covered by snow patches. Park, no. 645; Rapids, nos. 282, 1012; Eagle Summit, nos. 281, 803. ; Dr. INTEGRIFOLIA M. Vahl. Not as frequent as the A vipers Spe- cies. In gravels in the mountains. Park, nos. 279, 644 332 Rhodora [SEPTEMBER Dr. INTEGRIFOLIA M. Vahl, var. sytvatica Hultén in Sv. Bot. Tidskr. 30, 3: 527, fig. 2a (1936). Below Ft. Yukon on wooded bank, no. 280; Wiseman, in woods, no. , Rosa acicuLaris Lindl. Very common everywhere except in the high mountains. Fairbanks, nos. 287, 1057; along Yukon River, no. 288; Wiseman, no. 942. Also at Miller House, but did not collect it there. The beautiful wild rose of Alaska, with very prickly stems, pyriform fruit, and large fragrant flowers. TRIFOLIUM PRATENSE L. Introduced. Fairbanks, no. 1079. T. nypripum L. Introduced. Fairbanks, no. 1 Mepicaco Luputina L. Miller House, no. 811. Lupinus arcticus Wats. Common on dry sandy slopes throughout the region. In open woods in the Park, nos. 289, 646; Circle Springs, no. 290 AsTRAGALUS FRIGIDUS (L.) Gray, var. LITTORALIS (Hook.) Wats. Occasional in subalpine locations. Park, no. 649; Miller House, n0- 3, ; : A. atpinus L. Generally common. Park, no. 648; Rapids, no 292; Fairbanks, no. 690; Wiseman, no. 943. . YUKONIS Jones, Revision of N. A. Astragalus, p. 89 (1928). parse: along the Yukon River, no. 294. Possibly the first record for aska. This weak-stemmed, much branched Astragalus, with flowers cap tate in bloom and pods ascending, is very rare in herbaria. It was described from specimens found by Gorman near Ft. Selkirk, Yukon Territory. Later collected by Eastwood at Whitehorse, and recently by Raup on the Athabaska. A. raLctFERUS Hult. Gynophoraria falcata Rydb. See Hultén in Sv. Bot. _Tidske. 80, 3:526. Rare. On rock pile near Sable Pass, Park, no. 647. Second report for Alaska. Oxytropis MaypELuiana Trautv. 0. campestris DC., var. melan- ocephala Hook. Rare in the high mountains of the Interior. Eagle Summit, no. 805; on way down from Porcupine Dome, no. 5-A. Characterized by chestnut-colored stipules. O. MerTenstana Turez. Porcupine Dome, no. 809. A distinctive eastern Asiatic species, very rare in the Interior, prob- ably the first station, but reported previously twice from Peninsula. The third record for Alaska. < arctica R. Br. Eagle Summit, no. 806; Porcupine Dome, 9° 1940] Scamman,—List of Plants from Interior Alaska 333 O. PYGMAEA (Pall.) Fernald in Ruopora 30: 153 oe. Occa- sional in the mountains. Park, no. 650; Eagle Summit, no. 808. O. HUDSONICA (Greene) rahanee es Near Palyohrecia .. Park, no. 651. A plant with a strong tap-root, leaflets 18 or more, usually attenuate, and very glandular calyx. O. cracittis (A. Nels.) K. Schum. sng common. Park, no. 295; Rapids, no. 296; Wiseman (in fruit), n O. DEFLEXA (Pall. ) DC. Wiseman, no. This specimen has truncate calyx-tubes, satth broad, rectangular sinuses between short teeth. It is, therefore, the true Siberian 0. deflexa, not O. retrorsa Fernald, the American plant which formerly passed in America as O. deflera. See Ruopora 30: 140 (1928). 0. retrorsa has lanceolate, approximate calyx-lobes and acute sinuses. Second record for Alaska HepYsARUM ALPINUM L., var. AMERICANUM Michx. A variable terior. Park, nos. 304, 653; Rapids, nos. 302, 1018; Fairbanks, nos. 301, 689; Miller House, no. ee Circle Springs, no. 300; along Yukon River, no. 303; Wiseman, no. 94 . Mackenen Richards. Occasional i in thickets along rivers and streams. Rapids, no. 1017; Wiseman, nos. 306, 946. ERANIUM ERIANTHUM DC. Not seen north of the Alaska Range. Curry, no. 566. Linum Lewisn Pursh. Fairbanks, no. 865. Empetrum nicruM L. The Crowberry i is aes in heaths and bogs, especially in subalpine regions. Park, no. 308; Rapids, no. ara Miller House, no. 812; Circle Springs, no. 309; Wiseman, nos. , 948. VioLa BIFLORA L. The Violas are among the rarest plants of the Interior. This yellow-flowered species was collected but once, in a — thicket on a slope back of the ranger cabin at Igloo, Park, no. v7? sees L. In swamps, Circle Springs, no. 312; Fox near Fairbanks, no. 313. Not xe material to be sure of this. Collected V. epipsila Ledeb. in Seward on the southwestern coast, but did not see it in the Interior. _ SHEPHERDIA CANADENSIS (L.) Nutt. The “Soap berry” is common 'n open woods and thickets along streams. Park (in regia “e 655; Fairbanks, no. 315; Circle Springs, no. 314; Wi iseman, no. CIRcAEA ALPINA Li Circle Hot Springs, no. 327 Apparently the second record from the erick: Porsild reports it from Tanana Hot Springs, in Raopora 41: 263 (1939) 334 Rhodora [SEPTEMBER Epitopium aNGustirotium L. Chamaenerion angustifolium (L.) Scop. The Fire-weed is, doubtless, the most spectacular plant in Alaska, covering entire hills and meadows with a rose-magenta blanket. It also fringes both the Richardson and Steese Highways. Paxson, no. 316; Gahinks. no. 1060; Wiseman, no. 950. E. ANGUSTIFOLIUM, forma ALBIFLORUM (Dumort.) Haussk. Occa- sional. Wiseman, no. 318. E. ANGUSTIFOLIUM, fours SPECTABILE (Simmons) Fern. Miller House, no. 813. E. uatirouium L. Chamaenerion latifolium (L.) Sweet. The large-flowered “ River Beauty,” as it is sometimes called in Alsakis | Is often found on gravel bars in rivers and creeks in subalpine ee For map of general distribution see Fernald in Mem. Gray 2: 337 (1925). Park, no. 320; eg House, no. 814; Wists? nos. 321, 951, (a pale pink form) no E. LATIFOLIUM L., var. ceric Haussk. Park, no. 323; Wiseman, no. 953. E. paLustrE L. In wet ground by a creek. Miller House, no. 815. E. GLanpULosuM Lehm. A northern station for this species. Cir- cle Hot Springs, no. 324. UPLEURUM AMERICANUM Coult. & Rose. Common in ett gravels. Park, no. 329; along Yukon River, no. 330; Wiseman, n Conr0sELINUM CNIDIFOLIUM (Turcz.) Porsild in fhe pee 41:2 267- 268 (1939). C. dawsoni Coult. & Rose. On the bank of the Chena Slough, Fairbanks, no. 1081; along Yukon River, no. 333. Heracteum Lanatum Michx. This very tall, large co grew in a woody thicket back of Gunnysack Creek, Rapids, no. 332 Did not see it farther north. Cornus CANADENSIS L. Common in wooded areas, also in d open places. Rapids, n o. 1021; Fairbanks, nos. 337 and 691; Miller C. canadensi sis X suecica Hult. Fi. Keak 253-254 (193 “See as Porsild in Ruopora, 41: 270 (1939). Occasional. Park, no. 656; near pea Dome on Richardson Highway, no. 339; Circle Springs, pepe C. sroLonirera Michx., var. BarLeyt (Coulter & Evans) Drescher. A common shrub in the Yukon Valley and central interior region. Fairbanks, no. 342; Circle tek o. 341. Moneses UNIFLORA (L.) Gray. mip moist woods. Park, no. 344- Pics UNIFLORA, Var. RETICULATA (Nutt.) Blake. Rapids, nos. 345, iis SECUNDA L. A wooded ravine, Wiseman, no. 956 P. npa L., var. opTusata Turcz. Rich woods, Rapids, sas 1019; Panes no. 349; Circle Springs, no. 348. P. Granpirtora Radius. Circle Springs, no. 351. 1940] Scamman,—List of Plants from Interior Alaska 335 P. GRANDIFLORA, var. CANADENSIS (Andres) A. E. Porsild. Wiseman, no. 955-B. P. GRANDIFLORA, var. GorMANII (Rydb.) A. E. Porsild. Common in open spruce woods near Savage River, Park, no. 657; Eagle Summit, no. 817. See Porsild, Ruopora, 41: 271-273 (1939), for discussion of this species and varieties. : - ASARIFOLIA Michx., var. INCARNATA (DC.) Fern. in Ruopora 6: 178 (1904). Wiseman, no. 955—-A. : LEDUM GROENLANDICUM Oeder. The Labrador Tea is very common in muskegs in the central Interior. Fairbanks, no. 355; Circle Springs, no. 354; Wiseman, no. 957. L. paLustRE L., var. pecuMBENS Ait. L. decumbens (Ait.) Small. A dwarf, more northern species, with narrowly linear leaves, growing in the mountains. Park, no. 658; Eagle Summit, nos. 356, 820. RHODODENDRON LaPPpontcuM (L.) Wahlenb. Occasional on alpine slopes and also on the banks of the Savage River in the Park, no. 664. LoIsELEURIA PROCUMBENS (L.) Desv. The little “ Alpine Azalea is found occasionally on bare mountain slopes. Park, no. 659; Eagle Summit, no. 818; Circle Springs, no. 357. ; ASSIOPE TETRAGONA (L.) D. Don. Very common and widespread on mossy ledges and alpine summits. Park, nos. 361, 660; Eagle Summit, no. 362; Wiseman, no. 958. NDROMEDA Po.rrot1a L. Common in peat bogs and tundra throughout. Fairbanks, no. 363; Eagle Summit, no. 819; Wiseman, no. 959. CHAMAEDAPHNE CALYCULATA (L.) Moench. In low bogs. Fair- banks, no. 365; Circle Springs, no. 364; Wiseman, no. 960. ao _ ARCTOSTAPHYLOS Uva-Urst (L.) Spreng. Bearberry or Kinni- kinick is found now and then on dry bare slopes. Park, no. 661; Circle Springs, no. 366; Wiseman, no. 961. ; A. RuBRA (Rehder & Wilson) Fernald. A. alpina (red-fruited form) of Richards. Very common in this region. Circle Springs, no. 367; Wiseman, no. 962. The thin, deciduous leaves turn scarlet in the fall, causing the hills and lower mountain slopes of central Alaska to become masses of brilliant color. VaccInIuM uLIcINosum L. The highly-prized blueberry of the Interior, Common in heaths and tundra. Fairbanks, no. 368; Wiseman, no. 963. L., var. minus Lodd. Common on dry slopes. Park, oe Se (in flower); Circle Springs, no. 369; Wiseman, no. 964 (in ruit). DIAPENSIA OBOVATA (Fr. Schmidt) Nakai in Nakai and Koidzumi, Trees and Shrubs of Japan proper, 194 (1922). D. lapponica L., var. 336 Rhodora [SEPTEMBER obovata Fr. Schmidt. Rather rare on mountain ledges. Top of Poly- chrome Pass, in Park, no. 665; rocky cliff above the Lodge, Rapids, no. 371; Eagle Summit, no. 821. See Porsild in Trans. Royal Soc. of Can., ser. 3, sect. 5, 32: 35 (1938), for discussion of D. obovata and D. lapponica. D. obovata, loosely caespitose with trailing branches, and of a reddish-green tinge, with short, thick styles is of Eastern Asiatic-Western American distribu- tion, and all records west of the Mackenzie should be referred to it according to Porsild. Doverasta GorMANt Constance in Am. Midland Naturalist, 19: 257 (1938). A tiny caespitose plant with several rose-purplish flowers was found on a bare summit near Polychrome Pass in the Park, on June 20th, no. 675. This very rare little plant I put with some hesitation here. The backs of the leaves are pubescent with forked hairs and the margins not definitely ciliolate, which corresponds to Constance’s description. But specimens of this and D. arctica Hook. are so rare in herbaria that it is difficult to make determinations without more material. D. arctica, a rare arctic plant of the region west of the Mackenzie, is known in Alaska only by a sheet from Bering Strait. The type of D. Gormanii, and one other collection mentioned by Constance, are from Yukon River Valley and Lake Kluane in Yukon Territory respectively, so this is the first record for Alaska. AnpRosace CHAMAEJASME Host, var. arctica R. Knuth. ?4. Lehmanniana Spreng. Occasional in sandy places. Near Savage Rive prin mp, Park, no. 667 , A. SEPTENTRIONALIS L. A. Gormanii Greene. Rare. Dry cliff back a R.R. station in Park, no. 687. DoprEcaTHEON FRricipuM Cham. & Schlecht. A beautiful western- arctic “Shooting Star,” common in wet, springy ground in high alpine regions of the Interior. Park, nos. 372, 666; Eagle Summit, n0s- 373, 822. TRIENTALIS EUROPAEA L., var. arctica (Fisch.) Ledeb. Along the wooded bank of the river at ‘Curry, nos. 572, 573. No. 573 has narrow, acute leaves instead of broad and obtuse, and lanceolate petals. Also seen in a willow thicket in Fairbanks. GENTIANA PRoPINQUA Richards. See Hooker, Fl. Bor.-Am. 2: 62, a rad isa Twelve Mile, on Steese Highway, no. 377; Wiseman, Os. These two collections are true G. propinqua, tall, but delicate stem® purplish, many branched, with a slender corolla, 10-15 mm. long, 4? generally short, very unequal calyx-lobes. 1940] Scamman,—List of Plants from Interior Alaska 337 Other collections—Park, no. 375; Rapids, no. 1024; Miller House, no. 379; Circle Springs, no. 378—show much stouter plants with stem and leaves yellowish-green, cauline leaves more connate-clasping and calyx-lobes longer and united in a tube. Although they are dif- ferent from typical G. propinqua, they can not be satisfactorily placed in any other species. G. prostrata Haenke. This low-growing, often almost prostrate Gentian with small blue flowers is found only occasionally in damp ground in the mountains. Park, no. 380; Rapids, no. 381. G. ciauca Pall. A rare alpine plant with corolla of a strange P. putcHerrimum Hook. Circle Springs, no. 388. PLaciopoturys Cusick (Greene) Johnston. Introduced, Fair- banks, no. 1080-a. - COGNATUS Johnston. Dry ground, Fairbanks, no. 1080. Amsincx1a Menzies (Lehm.) Nels. & Macbride. Probably intro- duced. Rapids, nos. 395, 1023. ; ErrrricuruM aretromEs (Cham. & Schlecht.) DC. Eagle Summit, no. 826-A; Porcupine Dome, no. 826. A charming little fuzzy plant of the high mountains, with a rosette of hairy leaves at the base, rising from a woody root covered with brown leaves; the flowers in a capitate head, blue with a yellow eye, and very fragrant. Myosoris apestris Schmidt, ssp. asratica Vestergr. Twelve Mile Summit on the Steese Highway, no. 392; Wiseman, no. 967. The “Forget-me-not” is the Territorial flower, found in alpine meadows, and moist ledges of the mountains. MERTENSIA PANICULATA (Ait.) G. Don. One of the most “oe Ge Ous and graceful plants of open woods, common throughout A e Interior. Curry, no. 567; Park, no. 669; Paxson, no. 397; Miller House, no. 827; Eagle, on the Yukon, no. 399. : ity) TACHYS SCOPULORUM Greene. S. palustris L., subsp. pilosa (Nutt. Epling. Fairbanks, no. 866. 338 Rhodora [SEPTEMBER SCUTELLARIA ale ar aie ra pisbes See Fernald in Ruopora 23: 86 aay). am denne MENTHA CANADENSIS t. var. Pleat Benth. In swampy land, beside the] hot facies Circle Springs, no. 404. PRUNELLA VuLGaRIs L., var. + aniienie (Barton) Fernald, forma IODOCALYX Fern. in Ruopora 15: 179-186 (1913). Fairbanks, no. Linaria vuuearis Hill. In waste land, probably a garden escape. Fairbanks, nos. 405, 1058. VERONICA ALPINA L., var. UNALASCHCENSIS Cham. & Schlecht. V. Wormskjoldit Roem. & Schult, Rare. A mossy slope at Twelve ile Summit, on Steese Highway, no. 408. See Fernald in Ruopora 41: 450 (1939). . PEREGRINA L. Waste land. Fairbanks, no. 1068. Lacotis GLauca Ga ertn., var. LANCEOLATA Hult. FI. yee 4: 105 (1930). Rare. Near snow patch on Eagle Summit, n CASTILLEJA HYPERBOREA Pennell in Proc. Acad. Nat. Sei. Phil. 86: 532 noe Srey Summit, nos. 411 and 828 C. pati (L.) Spreng., subsp. CAUDATA Pennell. Rapids, n 1025. Dee pean of this collection is yellow-villous and ee glandular. Circle Springs, no. 413; Wiseman, no. 412. C. sp. Other col.ections of Castilleja await further study. EvuPHRASIA MOLLIS (Ledeb.) Wettst. In a low thicket on a hill behind the Lodge , Rapids, nos. 416, 1026. E. SUBARCTICA Raup i in Ruopora 36: 87-88 (1934). Fairbanks, no. 868. PEDICULARIS VERTICILLATA L. Occasional in subalpine regions. Park, no. 417; Rapids, no. 418; Wiseman, no P. LABRADORICA Wirsing. Common on dry hillsides, not usually in high altitudes. Park, nos. 419, 672; Miller House, no. $31; Cir cle Springs, no. 420; Wiseman, no. 971 P. supetica Willd. Fairly common in tundra. Miller House, 20. 829; along road to Eagle Summit, no. 422. P. Orpen Vahl. Abundant in wet ground below snow banks at base of Eagle Summit, but not seen elsewhere, nos. 421, 833. ‘ P. LANGSDORFFII Fisch. Park, no. 670; Eagle Summit, nos. 423, 34, A striking alpine Pedicularis with tall thick pink spikes. Rarely reported from the Interior. P. ranata Cham. & Schlecht. Rare. Eagle Summit, nos. 424, 832. Reported by Porsild in Ruopora 41: 287 (1939), from the Alaska Range. Probably second record for the Interior. P. caprrata Adams. On dry stony slopes, occasional in alpine regions. Park, no. 671; Eagle Summit, nos. 425, 830. BoscHNIAKIA Rossica (Cham. & Schl echt.) B. Fedtsch. Rapids, no. 427; Wiseman, no. 973, where it was very abundant. 1940] Scamman,—List of Plants from Interior Alaska 339 A strange looking plant, parasitic on roots of alder and spruce, growing in woods and thickets. It has a wide distribution in Alaska, as I saw it also in Seward. The Eskimo children called the plants “corn pipes.” PrincuricuLa vuLGaris L. Many of the vom rosettes were seen in a marsh in Wiseman, but I failed to collect t PLANTAGO MAJOR 1. var. ASIATICA (L.) Detain Circle Springs, no. 428. GaLIuM BOREALE L. Very common in dry soil on banks of streams and rivers in low areas. Fairbanks, no. 867; Circle Springs, no. 431. A tall, showy Galium with many white flowers in compact panicles. G. TriripuM L. In moist places near Circle Hot Springs, no. 434. /IBURNUM PAUCIFLORUM Raf. Very common in thickets along rivers. Rapids, no. 1016; Circle Springs, no. 439. The red fruit of this tall straggling shrub, often called “High-bush Cranberry,” is much prized for jellies and pies. INNAEA BOREALIS L. Specimens from Wiseman, nos. 437 and Cae a to the typical form. See Fernald in RHopora 24: 210 L. Borzauis L., var, AMERICANA (Forbes) Rehder. Plants from Miller Siiiees no. 837, have the funnel-shaped corollas. Common in rich woods and shaded thickets. VALERIANA CAPITATA Pall. A common characteristic plant of moist 5 and subalpine tundra. Park, nos. 441, 674; Wiseman, no. ommon in be Interior. Park, no. 442; Rards, no. 444; Beale Summit, nos. 443, 839; Wiseman, no. 976. - UNIFLORA L. Rare in the mountains, but no doubt easily over- My only collection was from Eagle Summit, no. 840. SoLipaGo MuLTIRADIATA Ait. The most common and abundant Solidago with wide-spread distribution in dry soil. Park, no. 445° Rapids, no. 448, 1033; Circle Springs, no. 447; Wiseman, nos. 446, 977. . DECUMBENS Greene, var. OREOPHILA (Rydb. ) Fernald in Ruopora 38: 201-204 (1936). Occasional on sandy banks. Miller House, no. 841; Eagle, on Yukon River, no. 449. S. LEpIDA DC., var. ELONGATA (Nutt.) Fernald in Raopora 17: 8-10 (1915). Along the Chena River, Fairbanks, nos. 451, 870. Apparently the second record for the Interior. STER JUNCEUS Ait. Circle Springs, no. 455. Reported but twice previously from Alaska. A. re L. Very common throughout the region, especially On sandy banks of streams. Park, no. 453; Rapids, no. 1027; Fair- banks, no. 1059; along Yukon River, no. 454; Wiseman, nos. 456, 978. 340 Rhodora [SEPTEMBER ERIGERON ELATUS Greene. E. acris L., var. arcuans Fernald. See Ruopora 40: 347 (1938). Wiseman, no. 980. E. ancuiosus Gaudin, var. cree eect (DC.) Hara in Ruopora 41: 389. E. elongatus Ledeb. E. acris L., var. asteroides of Am. Auth. Very common in subalpine catas Rapids, nos. 458, 1028; verte House, no. 843; Wisecinns: nos. 457, 979. E. LoncHopnytius Hook. In a sandy location not far from the Cheha Rivet n . 1064. Apparently rare, as only other reports are from Fairbanks. E. compositus Pursh, var. TrrrFIpUS (Hook.) Gray. Probably oc- casional in high mountains. On a rock pile near Sable Pass in Park, no. 676-a. Tufted with finely cut leaves and white flowers. E. srg mind err Gray. Along the Yukon River in thickets, Tanan E. CAESPI riba Nutt. Along the Yukon River, no. 461. E. rapicatus Hook. Park, no. 676; Eagle Camp, nos. 459, 696. Low, caespitose, with strong tap-root, growing in rocks. Similar in appearance, but in the Park specimen the leaves are less linear and perma to be slightly lobed and base of heads more villous. RIA PHILoNIPHA A, E. Porsild in Ruopora 41: 294, pl. 554 (idee) eg no. 678. This alpine species, recently named and described, ‘differs from A. monocephala, its nearest relative, according to Porsild 1. ¢. “by the thinner tomentum of leaves, the elongated offsets, tall and slender stems and by the larger pistillate heads.” It has a wide range from the Bering Sea, the mountains of interior Alaska and Yukon Territory to the Arctic coast east of Mackenzie. A. ALASKANA Malte in Ruopora 36: 107 (1934). Rare in bleak, exposed gravelly places on high mountains. Eagle Summit, no. 464; on the very top of Porcupine Dome, no. 845. A. Larner A. E. Porsild in ese 41: 293 with pl. 554 (1939). Dry gravelly location near the Savage River, Park, no. 677 (mentioned by Porsild). “ By its densely matted growth, the very leafy stems and compact glomerulate ivory-white heads, it differs strikingly from all other boreal Antennarias” Porsild 1. c. The TYPE was collected by H. © Laing at the head of the Chitina River. My specimens constitute the second pases: for this species. ULCHERRIMA (Hook.) Greene. This attractive Antennaria, often 15 inches high, grew abundantly in low open woods near the Koyukuk River, Wiseman, nos. 463, 982. 1940] Scamman,—List of Plants from Interior Alaska 341 Rarely reported from Alaska. This is probably the second record. A. OCCIDENTALIS Raf. Fairbanks, no. 1061. : } A. phen Ledeb. A. multiflora Hook. Occasional on river banks. Fairbanks, no. 1062. Leaves only pinnately divided. MATRICARIA MATRICARIOIDES (Less.) Porter. M. capris sey Buch. Common around settlements. Fairbanks, no. 475; Wiseman, no. 987. i i teep, ARTEMISIA BOREALIS Pall. Found growing plentifully on a s rocky hill near Gunnysack Creek, Rapids, nos. 447, 1034. A rare spe- cies in interior Alaska. A. ALASKANA Rydb. in N. A. Flora 34: 281 (1916). Gravel bars of a creek in Wiseman, no. 984. These match the photograph of Rydberg’s TYPE SPECIMEN in Gray Herbarium, collected by I. C. Russell on Yukon River between Nulato and Nowikakat, July 23-27, 1889. The second report for this species. A. Tries Ledeb. sens. lat. Very variable. Several specimens re- semble var. UNALASCHCENSIS Besser (var. elatior Torr. & Gray). eg mon around roadhouses and towns. Park, no. 480; Rapids, nos. 482, 1035; Miller House, no. 848; Wiseman, nos. 481, 983. P PETASITES FRIGIDUS (L.) Fries. Common in alpine main So os a Park (in flower), no. 685; Wiseman (basal leaves only), no. ARNICA ATTENUATA Greene. A tall plant with strongly attenuated leaves, common in thickets and on dry hillsides in subalpine oo ark, no. 681; Miller House, no. 849; along Yukon River, = ae A. Loutseana Farr. Park, no. 679; on et ee ee no. 850, : : A Cordilleran species found at Lake Louise and other apr ie the Canadian Rockies, in the mountains of interior Alaska, ‘ee v2 Gaspé Peninsula and western Newfoundland. See Fernald in Ruo- DORA 35: 368, pl. 270 (1933). A. Lessinei Greene. See Hultén, Fl. Kamtch. eerie ager Close beside the preceding species on Eagle Summit, nos. 486, Sol. Also in the Park, no. 485. 342 Rhodora [SEPTEMBER An Asiatic-Western American Arnica, with pale yellow ligules, brownish pappus and nodding heads. A. sp.? Resembles A. et but the rays unusually long for that species. Miller House, no. 850-a. SENECIO VULGARIS L. Probably introduced. Rapids, no. 1030. S. paLustris (L.) Hook. In wet places in the mining dumps after the ground has aa pag by water pipes. Fox and Ester Creeks near Fairban Oo. i TRE aa er che 3 to 4 feet high. S. ENS Richards. Abundant in sings along roadsides in ais and subalpine regions. Par . 492; Rapids, nos. 493, 1032; Fairbanks, no. 494; Miller House, no. 853; Wiseman, no. 985. Tall and weedy, characterized by the black-tipped bracts of the involucre. S. PAUPERCULUS Michx. Fairbanks, no. S. Fricipus (Richards.) Less. Cineraria “Vrigida Richards. Occa- sional in the mountains. Park, nos. 682, S. Kyettmanr A. E. Porsild in Roopora 41: 299 (1939). Cineraria frigida Richards., f. tomentosa Kjellm. Growing in moist places be- neath snow patches on Eagle Summit, nos. 854, 855. The heads are larger in this species, rootstock stouter, and whole plant woolly and tomentose. Rare in high mountains. S. aTRopURPUREUS (Ledeb.) Fedtsch. in Fedtsch. & Fler. Fl. Eur. Russ. 992 (1910). Cineraria atropurpurea Ledeb.; S. integrifolvus Kjellm. Eagle Summit, nos. 496, 856; Porcupine Dome, 856-a. A striking plant with orange rays and stem and leaves gray with lanate wool, found commonly on the mountain slopes and dry alpine tundra in the Eagle Summit region. Usually a number of heads in & cluster, but several plants were collected at higher altitudes with a larger single head. There is a resemblance between this species and specimens in the Gray Herbarium of S. pyroglossus Kar. & Kir. from the mountains of the northwestern part of Mongolia. S. RESEDIFOLIUS Less. See Fernald in Ruopora 26: 113-116 (1924), and Mem. Gray Herb. 2: 259 (1925), for map of general distribution. Rare in the mountains. Eagle Summit, no. 852. S. conrerMinus Greenm. On stony ground near the Big Delta River at Rapids, nos. 490, 1031. l SAUSSUREA ANGUSTIFOLIA DC. Fairly common in dry gravelly places and subalpine tundra of interior on Park, no. 498; Miller House, no. 858; Circle Springs, no. 499. S. ? monticota Richards. Wadena no. 986. 1940] Scamman,—List of Plants from Interior Alaska 343 A tall plant, with the heads longer-peduncled. Similar to S. remoti- flora Rydb. These three have been grouped under S. alpina, and more recently under S. angustifolia sens. lat., but at least S. densa seems distinct in the field. S. pensa (Hook.) Rydb. Eagle Summit, nos. 497, 857. Low alpine plant with heads in a close cluster. CrEPIS NANA Richards. Youngia nana oe Rydb. Occa- sional in dry soil on oe slopes. Park, nos. 505, 684; Rapids, nos. 506, 1029; Wiseman, no. 989. Low and tufted, Spies stemless. C. ELEGANS Hook. Apparently rare, but abundant at Miller House, in the piles of stones thrown out by placer mining, nos. 507, 85 A taller plant, many-stemmed with a long tap-root. TARAXACUM MUTILUM bao (Group a Dahlst.) (de- terminavit G. Haglund 1938). Wiseman, 90. Aachen: ed $93 verosimiliter”) ean Vulgaria Dahlst.). Fairbanks, n iY noeendienndin tc Dahlst. (Group Glabra Dahlst.). Park, no. 686. Lacruca scartota L. Introduced. Miller House, no. 508. L. PULCHELLA (Pursh) DC. Along Yukon River, no. 509. INDEX New scientific names are printed in full-face type Achillea borealis, 341; multiflora, os ; occidentalis, 341; sibirica, Aconite, 310 Aconitum ae en gh 325 aea eburnea, 324; rubra, 324, f. ~~ ge 324; ; spicata, var. rubra, Agropyron latiglume, 317 ge scabra, 317 proses Allium Sebeseasenen var. sibiri- cum, 319; sibiricum, 319 Alnus crispa, 321; fruticosa, 321; sinuata, B31 manii, 336; Lehmanniana, 336; septentriona lis, 336 Anemone, 325; globosa, 325; mul- Angiosperma e, 317 oe 34; alaskana, 312, 313, monoce shala, 30; pilonipha, 946; paleharritia” Antennari rias, 34 Arabis divaricarpa, 328; lyrata, var. glabra, 328, var. kamchatica, 328 : tagrostis arundinacea, 317; lati- Arctostaphylos al ina, 33 311, 335; Uva-ursi, 31 1 ia Arent arctica Bi 24 Rossii, 3 oe f a ‘3,4 a vee I Lessing 312, “341; Artemisia gee hc da 314, ; 341; ar tica, 341; borealis, 341; frigida, 341; Tilesii, 341, var. om 341, var. unalaschcensis, 341 Aspen, 3) ‘ae "junceus, 339; sibiricus, 312, Astragalus, 332; alpinus. 332; falci- ferus, 310, 332; fri sh pee var. lit- toralis, 332; yukonis , 33 Athyrium Filix-femina, a sitch- ense, 316, f. strictum, Azalea, Alpine, 335 Baked-apple Berry, 330 Barbarea planisiliqua, 327 Bearbe Reckitaciik Syzigachne, 317 Betula Ermani X glandulosa, 321; a, 31 14, 321, ere sibirica, 321; papyrifera, 314, 321 Birch, White, 321 Blitum capitatum, 322 Bluebells, 339 Blueberry, 335 Boschniakia rossica, 314, 338 aria, Botrychium Lun Boykinia Richardsonii, 328 Bupleurum americanum, 334 Haleioun; 326 Calamagrostis Pig pees nsis, Langsdorfi, 31 Langsdorfhi 317 Caltha, 324; si eo 324; palustris, var rifolia, Calypso bul , 320 Campanula lasiocarpa, 339; uni- a coe oe ee apsella Cardamine "belo aka B12, 327, var. berin 827; pratensis, 327, var. cnaanuite lia, 327; pur- purea, Carex angarae, ae Tg sewartom 318; atrosquama, ns, 318; capillaris, Ole: , capitata, 318; cinna, 318; disperma, 318; miei, 318; , var. 31 346 Pate 311 tetragona, 312, 335 iope Castilleja, 338; sp., 338; hyperborea, 338; allida, subsp caudata, 33 Cerasti Bee m, 323, var andi gr. florum Chamaedaphne calyculata, hamaenerion angustifolium, 334; latifolium, Chenopodium album, 322; capita- Ohenes ported 333 peers atropurpurea, 342; fri- a, 342, f. rodeathek: 342 Claytonia, 313, 322; a oe 312, 322; Scammanian . Comarum palustre, 331 ea rane cnidiifolium, 334; Dawso 34 Corn P Pause 339 — Sense ress: ue en var. inter- 334 suecica, 334; stolonifera, Pa." Baileyi, 334: s, 33 pipet done spect a 326; semper- rens, anberry, Mountain, 311 ‘iene cick: 311, "343: nana, 311, 343 Crocus, Wild, 325 333 Cystopteris fragilis, 315 Delphinium, 310; Brownii, — enziesii, 325; scopulorum, glaucum Descha mpsia caespitosa, 317 id vata, 335; obovata, 335, 336 Dicotyledoneae Dodecatheon, “310; frigidum, 312, Douglasia, ane arctica, 336; Gor- 327; borealis, 327; nivalis, 327; manii, Draba alpina, 310, 327; glabella, ta, 3 8 ee 314, 331; in- tegrifoli 1, wer. 'sylvatica, 332; cotowetali. 312, 331 INDEX Dryases, 310 Dryopteris austriaca, 316; fragrans, 1 - Linnaeana, 316, i eatin 316; Phegopteris, 316; spinulosa, 316, var. ameri- cana, 316, var. dilatata, 316 See innovatus, 317 Em rum nigrum, 311, 333 Epilobium angustifolium, 312, 334, f. albiflorum, 334, f. spec ctabile 334; paninbns: 3 lat ga 314, 334, var. camel m, 334; palustre, 334 Bauien um arvense, 316; fluviatile, oru : 318; callitrix, 318; Chamissonis, var. albidum, 318; medium, 318; cheuchzeri, 318; vaginatum, 318 Beiohan aretioides, 312, 337 Erysimum cheiranthoides, 328 Euphrasia mollis, 338; subarctica, Eutrema Edwardsii, 312, 326 Fern, 315 Festuca altaica, 317; rubra, 317 Fire- , 334 Forget-me-not, 337 Fragaria yukonensis, 330 Galium, 339; boreale, 339; trifidum, 9 Gentian Danieaien ei 312, 337; propin- qua, 336, 337; prostrata, "33 Geocaulon lividum, 1, 321 ranium erianthum . Geum macrophyllum, var. perinel ymnosperm Gracpharate pet 332 Habenaria sp., 319; dilatata, & = hyperborea, 314, "320; obt ’ ’ ameri- ysarum alpinum, var canum, 311, — O ackensi 333 INDEX 347 Heracleum lanatum, 334 Hierochloé alpina, 317; odorata, 317 igh-bush Cranberry, 339 Hosdans jubatum, 317 Iris, 319; setosa, 319 chris 319; alpinus, 319; meng er ar. Haenkii , 319; bufonius, 318; eastaneus 319; Sitjormia, 313, 319 Juniperus c mmunis, var. montana, 314, 317; pr rms 317 Kinnikinick, 335 Labrador Tea, 335 Lactuca clans. 343; scariola, 343 Lagotis glauca, var. lanceolat ata, 338 Larix laricina m decum ns, 335; groenlandi- cae 335; palustre, var. decum- ns Lepidium apetalum, 327 Lesquerella arctica, 311, 327, var. Seamman ste 327 . Linaria vulga s, 338 Linnaea wt 311, 339, var. americana, , 339 Linum Lewi wisi, Lloydia serotina, 319 Loiseleuria , 335 Lomatogonium rotatum, f. tenui- ae 310 Lupinus arcticus, — Luzula campestris, var. frigida, 319, var. multiflora, 310: confusa, 319; japonica, 319; multiflora, 319, L var. frigida, 19; 1 Lychnis apetala, 324; ; Tayloriag, 324 ycopodium ann um, 316, var. pungens, 316; se et 316, var. =e on, 316; complanat- 316; Sek. 316 es Matric matricarioides, 341; Me eolens, 34 Medicago i lupulina, Saag elan Menthe | onakeues ensis, Deg Mihai Merckia physodes M , 323 Meee, 310; paniculata, 337 — arctica, 323; macr rocarpa, Mocbringa lateriflora, 323 “sp rie Pen iflora. , 334, var. reticu- vp a 317 Montia sarmentosa, 322 Myosotis alpestris, 337 yrica Gale, 1 Neslia paniculata, 327 Nuphar variegatum, 324 Nymphozanthus variegatus, 324 Onoclea —— 315 Orchids, 314 nica, 333; 332; Mertensiana, 312, 332; pyg- ea, 333; retrorsa 333 Papaver, 310, 312; alaskanum, 326; pala um, Parnassia Kotzebuei, 311, 329; pal- erie: var. neogaea, ,3 a, 310; macrocarpa, 328; nudi- s, 312 , 338; capitata, 338; brads. 311, 321, 324, 338; anata, 338; Lan gsdorffii, 338; eri, 338; sudetica, 338; verti- cilla Petasites frigidus S canariens is, a317 Phegopteis Dryopteris 316; poly- podioides Picea, 314; oe 317; mariana, 317 nguicula vulgaris, 339 Plagiobothrys genet iak 337; Cu- sickii, 337 Plantago major, var. asiatica, 339 Pleurogyne rotata, 6. tenuifolia, 337 Poa bite 317; alpina, 317; retica, 317; rigens Polemonium, 310; acutiflorum, 337; cherrimum Polygonum alpinum var. lapathi- folium, 312, 322; aviculare, 322; Bistorta, 311, 322; Convolvulus, 322;h ydropiperoides, 322; plumo- sum, 322; viviparum, 311, 322 Poppy, 326 Populus balsamifera, 320; hacea, 314, 320; eden on 314, os a 313, 330, 331; biflora, ar. groe 348 INDEX Nuttallii, 330; pacifica, 331; sewn 331; pensylvanica, 330; obbinsiana, 331; uniflora, 330 Potentillas, 310 Prunella vulgaris, var. lanceolata, f. 8 aly X, Pteretis sag 315 Ranunculus hyperboreus, 326; ni- valis, 312, 326; Purshii, subsp. yukonensis, 313, 324, 326 pyg- maeus, 312, 326; dent nei Bb: yukonensis, Raspberry Raw deursn, 310; lapponicum, SS mages Borger Saponm 328 Ribes e, 329 River uty, Rorippa barbareaefolia, 327; curvi- siliqua, Rosa acicularis, 312, 332 332 Rose, Rubus acaulis, cot —— 330; aemorus, , 330; idaeus, var. canadensis, 312, Rumex Acet tosa, 321; arcticus, 321; mexicanus, 321 Salices, 312, 315 Salix alaxensis, var. longistylis, 321; oides, 321; arctica, 320; ae phylla, 313, 320; pseu udo polari 20; pulehra, 321; vetioelata: 320, var. orbic 20; rostrata, 321; baleatierrs: § 320; Seemanii 321; stolonifera, 320 es rba Menziesii, 331; micro- phala, ; offici nalis, 331; fitshentla 331 Saussurea alpina, 343; angustifolia, 342, 343; ensa, 312, 343; monti- cola, 342; remotiflora, 343 29; adscendens, Ao pba ere subsp. Funstonii, 312, 329; Bechecheltaii, 311, 329; agellaris, 329; foliolosa, 312, 328; hieracifolia, ; Hirculus, 312 329; oppositifolia, 329; punctata, 312, 329, ssp. insularis, 329; pera were 328; rivularis, 312, 328; tellaris, var. comosa, 3 : tri- coeradehd, 311, 329; yukonensis, Saxifragas, 310 Saxifrage, 329 cirpus at iieieiah) 318 Scutellaria epilobiifolia, 338 um roseum, var. integrifolium, Selaginella sibirica, 316 Silene, — acaulis, var. exscapa, 312, var. subacaulescens, 324; esciny 311, 324; Williamsii, 324 Soap-berry, 333 Solidago 3: 339; — ey sons oreophila, lepi cen 339; ceusttieadintts 312, aria rubra, 323 3 unda . : . ’ . . ifolia, 323; crassifolia, ye —— ta, 323; , 330 Swertia perennis, var, obtusa, 337 Taraxaca, 315 343: Taraxacum kamtchaticum, 343 Kjellmanii, 343; mutilum, Tha bepos f ge Dryopteri is, 316; ire s, 315; Pheropteris, 31 Thiaspi arvense, ‘ig Tofieldia coccinea, 319; minima, 319; palustris, 319 sien Trient talis europaea, var. archi” 336 INDEX 349 Trifolium hybridum, 332; pratense, Triglochin palustris, 317 Trisetum ea 317 Tundra Rose, 331 Vaccinium uliginosum 311, 335, var. alpinum, 335; Vitis-Idaea, var. minus, 35 Valeriana capitata, 339 Vero ‘ ‘gee var. unalasch- egrina, 338; censis, $88 oreg Viburnum pauciflorum, 339 Viola biflora, 310, 333; epipsila, 333; palustris, 333 Violas, 333 Wild > 325; Rhubarb, 322; Stock, 328 Willow, 320, 321; Felt-leaf, 321 Willows, 31 1i, 320 Woodsia alpina, 311, 315; glabella, 315; ilvensis, 315 Youngia nana, 343 Zigadenus elegans, 319 CONTRIBUTIONS FROM THE GRAY HERBARIU OF HARVARD UNIVERSITY ie a i ae Reprinted from Ruopora, Vol. 42, Nos, 502-504, October-December, 1940 CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY CXXXIII A CENTURY OF ADDITIONS TO THE FLORA F VIRGINIA By M. L. Fernautp Dates or IssvuE Pages 355-416 and Plate @96... 0... 2 1 October, 1940 Pages 419-498 and Plates 627-649..............--..--- 9 November, 1940 Pages H0S-E3l 17 December, 1940 CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY—NO. CXXXIII A CENTURY OF ADDITIONS TO THE FLORA OF VIRGINIA! M. L. FerRNALp (Plates 626-649) Part I. I rinerary oF Five FIELD-TRIPS Aided by a generous and greatly appreciated grant from the Penrose Fund of the American Philosophical Society, I was able to continue botanical exploration on the Coastal Plain of south- eastern Virginia at intervals from June to October, 1939. Half the grant was used to cover the expenses of an all-summer bo- tanical reconnaissance of the lower Santee valley in South Caro- lina (the home of Thomas Walter and a collecting ground of Stephen Elliott) by my students, Messrs. Robert K. Godfrey and Rolla M. Tryon, Jr., their results to be published in a sepa- rate paper. With the portion of the grant retained for the Vir- ginla work I employed our already proved and efficient helper, Mr. Leonard Birdsall, maintained the established headquarters at Mrs. Bowman’s at Century House, south of Petersburg, and met the expenses of equipment and a portion of those of travel. Mr. Bayard Long, most happily, was able to join me for all five trips, most happily because the eastern Virginia work has PG Exploration done with the support of a grant from the AMERICAN PHILOSOPHICAL ETY. 356 Rhodora [OcroBer now become intensive and wholly eclectic, with little or no atten- tion devoted to the ubiquitous and well known species, our eyes being always on the alert for something which “does not register.” Occasionally the failure of a plant to “register” proves humiliat- ing but in two out of three cases the plant we do not at once recognize from the car or in walking through the unspoiled areas justifies a special stop; most often it is quite new to the known flora of the region and it frequently leads to the discovery of a new and productive habitat. Without Long’s keen knowledge of Coastal Plain plants and his persistence in following clues to successful discoveries too many rare and excessively localized species would be overlooked. As it is, we are often amazed to find in supposedly familiar areas plants of great significance which, diverted by some other species, we must previously have overlooked. For instance, in June, 1939, attracted by an un- usual composite, we slowed down near Applewhite’s Church—to collect for the first time the remarkably distinct Tetragonotheca helianthoides,? a southern monotype (Map 1), based by Linnaeus partly upon a Virginian specimen of Clayton’s, but certainly rare in eastern Virginia. That discovery soon led us to a rich strip of undisturbed woodland along the Nottoway near Carey Bridge, on the plantation of Mr. S$. R. Westbrook, who readily granted us permission to botanize over the area. Our interest absorbe by many unusual species, we perhaps had blind spots for others. At any rate, returning in June, 1940, we found on the springy wooded slopes which we had canvassed during three days at the same season the year before such extraordinarily conspicuous species as an Aconitum, leaning and twining on the neighboring shrubs, a plant not yet in flower but presumably A. uncinatum of the interior; the true and famously rare Sphenopholis pallens 2 In this, as in preceding vest of this series, the authors of species are omitted in the narrative if they are in Gray’s Manual. The preceding papers on the work in Virginia are as follows: Fernald & an ays of Botanizing in tern Virginia, wW ald eniona ree D outheas DORA, xxxvii. 129-157 and 167-189, 20 plates (1935)—Contrib. Gray Herb. cvil; nald, Midsummer Vascular Plants of paper abil Virginia, Ruopora, xxxvii. 378-413 and 423-554, 22 plates (1935)—Cont a gaint Bi o. CIX; Fernald, Plants from the Outer — bore! oF Virginia, Ruopvo pid and 414-452, 13 plates (1936)—Co . ray Herb. CXV; Local Mids ur of eee Inner Coastal Plain of Southeastern Virginia, prance XXxix - 321-366, 379-415, 433-459 and 465-491, 4 plates (1937)—Contrib. Gray Herb. no. CXX; Noteuerthe Plants Southeastern ide ia, Ruopora, xl. , 434-459 467-485, 27 plates (1938)—Contrib. Gray Herb. CXXIIL ; Last Survivors in the Flora of Tidewater Virginia, Rwopora, xli. 465-502, $0058 an d 564-577, with 14 plates (1939)— Contrib. Gray. Herb. no. CXXVIII 1940] Fernald,—Additions to Flora of Virginia 357 (Spreng.) Scribn.,? of which, as stated by Hitchcock (Man.), “The only other specimen [besides the original of more than a century ago, sent by Muhlenberg to Sprengel, supposedly from Pennsylvania] known is the type of Eatonia aristata collected [somewhere] in South Carolina by Curtiss’; and, beneath them, carpets of a reclining rhizomatous Cardamine, with showy petals deeply cleft like those of some species of Silene. A few miles away, in another area north of Applewhite’s Church, where we had extensively botanized one year earlier, there was a very definite and apparently heretofore unrecognized Carex, although it exists, misidentified, in some of our herbaria from Georgia and Florida. In these cases, each species is so highly localized that, passing a rod or two on either side of their small colonies, we might easily have missed them all the second year. They again illustrate the truism, that no area is ever fully worked out. Needless to state, it was Long who first detected three of the four. Our first field-work of the summer of 1939 was from June 11— 24. We had some problems left over from previous years which demanded immediate attention. On our first trip on the inner Coastal Plain, in June, 1936, we were amazed to find in open woods and clearings a species of Andropogon in the prime of flowering and early fruiting. Dreading the complexity of the genus, we tentatively set it aside as an aberrant colony of A. scoparius, a variable species normally flowering in September and October. But the plant could not be thus disposed of. Every June it abounds in open woods, clearings and fields as a regular element in the late-spring and earliest-summer vegetation, from Dinwiddie County at the west, York County at the east, thence south nearly to the North Carolina line; and Godfrey & Tryon got it, over-ripe and shattered in July, in southeastern South Carolina. Best of all, it has strong morphological characters Separating it from A. scoparius. It is a vernal species, nearest related to an autumn-flowering endemic of Texas. It will be described and illustrated (pLaTE 626) in Part II. Another characteristic plant of the region, also found by us in June, 1936, and subsequently much collected, from May into July, on the rich wooded bottomlands, is a gigantic quillwort, Sir pallens true, not that of most treatments, which is the transcontinental S. inter- media Rydb. 358 Rhodora [OcrToBER Isoétes. Flaccid and sprawling in the richest woods, with bulbs up to 3 cm. in diameter and loose rosettes up to 9 dm. across, it always fascinates us, for it is decidedly not an aquatic; and on many bottomlands, from Dinwiddie County to Greensville and western Nansemond, it is impossible to walk in June without trampling upon it. It proves to be the little known J. Engel- manni var. caroliniana A. A. Eaton, described from along creeks tributary to the Tennessee River in westernmost North Carolina but subsequently found in Georgia (Map 2), another of the up- land and inland types reappearing on the Coastal Plain of Virginia. These were a good start and on our first full day in the field (the 12th) we began brand new discoveries. On one of the earlier trips we had hoped for good things about Taylor’s Mill- pond in southeastern Greensville County but, alas, it is a con- ventional millpond, formed by damming a stream, with water lapping the drowned marginal forest and without a natural beach, such as we are constantly seeking. Taylor’s Millpond was promptly given a black mark. On June 12th, however, the region came into botanical favor. After making brief visits to old areas near Skipper’s and Dahlia, we took the dirt road east from Skipper’s (that from Dahlia has become impassable), plan- ning to explore via any side road which looked tempting. Sud- dently, beyond an open pine wood, we saw a swale. That meant moisture and perhaps a sphagnous bog. Sphagnum was there in abundance but we were on a ruined remnant of several acres of what was once a good habitat. The plow had been there, al- though, as is so often the case, the heavy and soggy clay land is unfit for agriculture. Only a few of the original bog plants re- mained, Panicum strigosum and some Droseras; but here was the greatest assemblage imaginable of the bog species of Lycopodium, which seemed to have profited by cultivation of the land. alopecuroides and L. inundatum vars. adpressum and Bigelovit formed a nearly continuous carpet, and at one end of the swale there was the rare L. carolinianum. It is sometimes listed as a Virginian, but in their monograph of the genus (1900) Lloyd & Underwood were unable to cite any specimens from between southeastern North Carolina and southern New Jersey, although it has long been known at a single station in Prince George 1940] Fernald,—Additions to Flora of Virginia 359 County, Maryland. In June L. carolinianum was very imma- ture but we made regular monthly calls, to watch its develop- ment. On our September visit the swale was under two feet of rain-water. We then wallowed and tumbled about the sub- merged furrows, well above our knees in muddy water, and, reaching arm’s-length below, pulled up strobiles of L. carolini- anum 74% em. long, of L. alopecuroides 1 dm. long and of L. inun- datum var. Bigelovii 1.1 dm. long, of var. adpressum 1.2 dm. long. After this hydropodic and most novel adventure among the Lycopodia we gladly replaced Taylor’s Millpond upon the botanical map. Taylor’s Millpond further justifies its existence as a focal point of botanical note because slightly to the southeast a dirt road leads across and along the wooded bottomland of Fontaine Creek into North Carolina. The bottomland, where not too heavily wooded, is a tangle of southern sedges, Carex louisianica Bailey up to 7.5 dm. high (Mackenzie says 2-6 dm.), C. crus-corvi var. . virginiana Fernald, Scirpus atrovirens var. flaccidifolius Fer- nald, and several others. At the upper border of the bottom- land there is a fine clump of Amsonia Tabernaemontana, the typical variety, not recorded by Woodson in his monograph of the genus as native north of South Carolina. Farther out, at the margin of the Creek, Sagittaria Weatherbiana Fernald min- gles with an orchid with oblanceolate leaves 3-3.5 em. broad. In June the latter was not yet flowering, and the colony was deep under water until the end of September, when, from among the clay-crusted leaves, the characteristic long raceme showed it to be Spiranthes cernua var. odorata (Nutt.) Correll (S. odorata Nutt.), a giant plant for its group, though in September and October we found it on tidal marshes farther north reaching a greater stature (up to 7.5 dm. high). On the 14th, Godfrey, on his way to South Carolina, joined us for half-a-day. After showing him some of the specialties between Petersburg and Emporia, we drove east to look at the bottomland of Three Creek, between James River Junction and Grizzard. In what was once a sphagnous bog along the South- ern Railway Sarracenia flava and a few other bog species still held their own (though apparently completely obliterated by ditching operations later in the summer). Here was a new in- 360 Rhodora [OcroBER land limit for Zrigeron vernus, and the second station in Virginia and east of the Mississippi valley of Eleocharia tenws var. verrucosa Svenson. On the bottomland, where we went to show Godfrey Glyceria arkansana Fernald and other species not re- corded from neighboring North Carolina, we were greatly im- pressed by Peltandra virginica. In New England, New Jersey and Pennsylvania the plant we know by that name has a tightly rolled green spathe with closely appressed pale margins which loosen without spreading more than just enough to allow the entrance of insects and snails to the closely surrounded whitish spadix. Here on Three Creek, however, the spathe loosely opens, its creamy-white margin spreading and fully ex- posing the deep orange-yellow spadix. At the summit of the enclosed ovary the spathe is circumscissile, a deliquescent band soon developing at this point, so that the limb of the spathe very soon drops off, leaving a truncated young fruit. This prompt circumscission of the limb of the spathe I have not seen in Massachusetts. Furthermore, the pale spadix of the northern plant is covered with flowers essentially to its tip; the orange- yellow spadix of the eastern Virginian has the upper inch or s0 often sterile. On all the streams we have yet visited in south- eastern Virginia Peltandra is uniform; and Dr. Lily M. Perry in- forms me that this plant with open spathe and yellow spadix is what she knew in Georgia. P. virginica was based on the brief description by Gronovius of the Virginian plant of Clayton. It is probable that the northern plant is not P. virginica. The spathe and spadix of each will be shown in pLaTEs 627 and 628. Attending the meeting of the Botanical Society of America at Mountain Lake and Blacksburg, where we were the guests of Professor and Mrs. Massey, we got back into our special field- work on the 18th. Wishing to check on Peltandra, we took 4 side road, leading from near the Prince George-Sussex County line from the Jerusalem Plank Road across to Stony Creek. On the first bottomland, that of Jones Hole Swamp, we were col- lecting the calla-like spathes when a passing fisherman, looking down from the bridge, remarked: “Oh! getting water lilies?” That seems to be a general name for Peltandra virginica in the region, more appropriate to it than to the northern plant. Oe the wooded bottomland of Jones Hole Swamp Rhododendron 1940] Fernald,—Additions to Flora of Virginia 361 canescens was frequent, the first we had met in Virginia, though we afterward got it in woods along other portions of the Notto- way system. This is true R. canescens, as restricted by Rehder, who, in his monograph of the group, gives its range as extending “from southwestern Tennessee and southern central North Caro- lina” southward and southwestward. Somewhat northeast of Stony Creek and again both north and south of that botanically productive center swampy woods were characterized by pin oak, Quercus palustris, which we also found along Three Creek, far- ther south. We had never before met it on the Coastal Plain of the state and Sargent (Man.) cites it as occurring in Virginia south of the Potomac only in the mountains (Wythe County). It has, however, been reported from several counties, including Amelia, which approaches the Coastal Plain, and Charles City County which is upon it. As it turned out, the day was pri- marily devoted to trees and shrubs. Driving eastward to Sussex Courthouse, then to Homeville and Waverly, we stopped on the west side of the Nottoway, where the ash-keys attracted us by their extreme smallness. They were on Fraxinus americana, var. microcarpa Gray, a southern extreme which we had not previously met. On the east side of the river, nearer Homeville, the forest was so rich that we returned next morning for further collecting. Carya ovata var. pubescens Sargent, recorded from Alabama and Mississippi northward into South Carolina and Tennessee, abounded, and with it the southern C. pallida Ashe. t one point in the drier forest above the bottomland Arabis canadensis abounds, our first Coastal Plain station; and near it was a carpet of Paronychia fastigiata, var. paleacea Fernald, heretofore known only from Delaware and Pennsylvania to Il- lmois and Tennessee. Another extension southward to Virginia was noted when we stopped, as we usually do, at the border of Assamoosick Swamp, northeast of Homeville. A bramble there was one of the arching or doming but tip-rooting series midway between dewberries and high blackberries. I cannot separate the material from that of Rubus Janssonii Bailey, of southern New England. Exploring, still unsuccessfully because of the clearing of land and the turning in of hogs on most boggy areas, with the hope of rediscovering Pursh’s stations of 137 years ago, some of them 362 Rhodora [OcrToBER on “boggy meadows near Dr. Gray’s,t Southampton”, others simply from “swamps, Southampton”, we covered much of the county. Pursh’s material, at the Philadelphia Academy, shows familiar Southampton County plants, such as Stillingia sylvatica, Asimina parviflora, Quercus laevis, Sarracenia flava, Amianthium Muscaetoxicum, Lobelia glandulifera, Carphephorus bellidifolius and tomentosus, and others which are now very local. But we wanted to find in the county Gentiana Stoneana Fernald (G. linearis in Pursh’s herbarium), from “swamps, Southampton and Asclepias paupercula, “wet swampy woods, Southampton : We already have them from farther east, so that their rediscov- ery is of secondary importance. But of the very first importance is the fact that Pursh got in “swamp, Southampton” the very rare Litsea geniculata, probably not subsequently collected in the state; and “in rich hedge-rows, Dr. Gray’s”, Baptisia villosa, a species also collected by Canby near Franklin in 1867. As yet these two are unknown as living species in the state. During these searches we found on the old road from Sebrell to Court- land an area of white sand, with the remnant of a forest of Catesby’s oak, Quercus laevis, and with Q. cinerea Michx. at a new northern limit. Stillingia is abundant, perhaps where Pursh got it, but at its northernmost known station; Bulbostylis cilia- tifolius (Ell.) Fern. also is there at its northern limit, and at the margin of Assamoosick Swamp Nyssa sylvatica var. — Fernald abounds. The area was so promising that we registere it for later visits. ‘ I have referred in the opening paragraphs to the region os Applewhite’s Church and of Carey Bridge. Here is an area 0 alternating rich loamy (often caleareous) woodland, more ster- ile sandy woods and wooded bottomland. The unique Tetrago- natheca (Map 1), with four big foliaceous bracts subtending the sunflowerlike heads, the distinctive achenes becoming consp!¢U- ous, was found in two spots. In the woods were Onoemnoe virginianum and a complex series of Houstonia, from aber H. lanceolata to extreme H. tenuifolia. It begins to look as! Gray’s old treatment, of combining several of these as variations of one species, were sound. Polygala polygama abounds and 1s : h and *Edwin Gray, proprietor of much of Southampton County in the late 18t early 19th centuries, buried between Sebrell and Wakefield. 1940] Fernald,—Additions to Flora of Virginia 363 strikingly unlike the more northern plant in its very lax raceme with distant and relatively large flowers. It is true P. polygama of Walter at its northern limit, the plant of the North, with more crowded and mostly smaller flowers being var. obtusata Chodat. Viola triloba var. dilatata (Ell.) Brainerd (“southwestern Lou- isiana northward to northern Oklahoma, southern Missouri, and southern Illinois; thence eastward to northern Georgia and western North Carolina”) was there, with scattered Chamae- lirtum luteum and Carex striatula. Viburnum rufidulum, rarely flowering, was here heavily in bloom, some of the trees 20 feet high, with trunks 4 inches in diameter, the blackish bark broken into rectangular blocks suggestive of the bark of Diospyros. Small trees of Vaccinium arboreum (up to 20 feet high, with trunks a foot in diameter, the main branches 6 inches through) were also very beautiful, with their loads of milk-white drooping flowers. At the point where Mr. Westbrook’s bridge crosses Three Creek an old woodroad leads up the latter stream to pits from which the shell-marl was long ago dug out. Here, as exclusively in lime as elsewhere in eastern Virginia, Equisetum arvense and E. hyemale var. affine abound; and the thousands of plants of Ponthieva racemosa (Walt.) Mohr give encouragement that this southern orchid may hold its own in Virginia. In deep woods along the bottomland Ampelopsis arborea occurs and near it sterile plants which could be only the rare Chelone obliqua; and later in the year we here established a new southern limit on the Coastal Plain for the Canadian and Alleghenian Prenanthes altissima. At a crossing of Three Creek slightly farther up-stream, on the road north from Applewhite’s Church, Rhododendron canes- cens abounds; and on the steep slopes with Antennaria solitaria we got our first Coastal Plain material of the upland Cunila origanoides, The last full day in the field in June (23rd) we went to some of the old ferry-landings on the Nottoway and the Blackwater. The tidal shore at Knight’s Seine Beach (Battle Beach of the con- tour-sheet) on the lower Nottoway was showy with masses of flowering Zizaniopsis, with great swales of Rhynchospora caduca Ell. and other nice species farther back, but our great prize here 364 Rhodora [OcToBEerR was the first collection of the season (by no means the last) of the estuarine Isoétes saccharata, not seen by Pfeiffer from south of the Potomac at Alexandria. Later in the summer and autumn it proved to be a regular inhabitant of tidal shores. Since Battle Beach is almost in North Carolina, just above the junction of the Nottoway and the Blackwater which form the tidal Chowan River, the Isoétes will certainly be found in North Carolina. At Cobb’s Wharf,5 on the west bank of the lower Blackwater, there are good swales. Panicum albomarginatum, not too common in Virginia, was in fine condition and Lysimachia producta, which we had not previously collected in the state, abounded, quite by itself, with no L. quadrifolia nor L. stricta (sometimes thought to be its parents) seen. Best of all, there were splendid great clumps in perfect vernal flowering of one of the very rarest of grasses, Panicum cryptanthum Ashe, which had been unknown between southeastern North Carolina and its local stations in New Jersey. These old landings, long ago cleared of heavy timber, thus offering sunshine, as compared to the darkness of the neighboring cypress and gum forests, were so full of local species that we at once conceived the idea that the old landings on the east side of the Blackwater would yield their quota of novelties. The contour-sheet showed below South Quay bridge a tempting series, George’s Bend, Milk Landing, Cox Landing, Sandy Landing and Cherry Grove; and Wyanoke at a point which, if it were west of the Blackwater, would be in Virginia. At this area along the Blackwater, the boundary between Virginia and North Caro- lina suffers what the geologist would call a nonconformity oF perhaps a fault. From below False Cape, on the Atlantic coast, westward to the Blackwater the boundary line is at latitude ap- proximately 36° 48’, but from the Alleghenies eastward to the Blackwater it is nearer 36° 46’. We were told that when the early boundary surveyors came from the west and from the east in two parties they failed by more than half-a-mile to meet at the Blackwater and their miscalculation is still sacred. When, in 1936, I first approached western Nansemond I had left Long at home to make up sleep and with Carroll Williams 5On some of our earlier labels we mistook Smith’s Ferry on the Nottoway for Cobbs ce 1940] Fernald,—Additions to Flora of Virginia 365 crossed the bridge at South Quay from Southampton County. I hoped, somewhat blindly, to find a continuation southward of the white-sandy pine barrens which occur south of Zuni (farther up the Blackwater). As I then wrote, “Expressing my hopes to the drawtender, I received the reassuring reply, accompanied by a general sweep of his arm toward Nansemond County: ‘Thar’s a powerful lot of right smart sand over thar’.”® We con- sequently took the firm road southeastward to Factory Hill, thence to Somerton, and on many subsequent days this route and its arteries have led us through southwestern Nansemond. It would not serve, however, if we were to get to the old landings on the Blackwater. We had formerly let the wheels of the car make our decision for us and they had automatically followed the hard road. Now, putting our own minds to the task, it was clear from the contour-sheet that between the Factory Hill road and the river there are no good roads; the area has a loose mesh of dotted lines indicating only “poor roads” and paths, and much of the region is uncleared. Following the first “poor road” southward from east of the bridge, we soon found ourselves in loose white sand, the road crossing deeply drowned and loosened or floating corduroys over the rain-swollen branches, so that Long and I had to get out and, standing in deep water, hold floating logs at each end of the submerged bridge while Leonard “bucked” the bridge, the water often flooding the body of the car. Only a driver with Leonard’s skill would attempt such a road, but through two years now he has not balked, except when, trying a swollen branch and finding the water nearly up to our hips, we have concurred in his decision not to cross. We did not reach any of the landings on the Blackwater that afternoon. The familiar black clouds were assembling in the north and west, ominous thunder was increasing and we, ob- Viously, must get out of this “poor road” before the next cloud- burst. But the continuous white sand, with forest dominated by Catesby’s oak, Quercus laevis, and by Q. cinerea, with long- leaf pine (Pinus palustris) quite evident, with Carphephorus bellidifolius, Scleria nitida Willd. and other pine-barren special- ties everywhere, it was obvious that at last and at the last Moment of our last day of the June trip we had unwittingly *RHopors, xxxix, 360 (1937). 366 Rhodora [OcToBER stumbled upon the “powerful lot of right smart sand” which we had been missing for four years. The pine barrens of south- western Nansemond occupy more than 12 square miles and they could not be investigated until the July trip. On our July trip (12-28) old problems also had to be looked into. In 1936, on the bottomland of the Nottoway at Cypress Bridge, we got two small individuals which were not quite typical of Lysimachia radicans Hook., primarily of the Mississippi basin. We needed material collected through the season and showing all variations. So, on July 17 we went to Cypress Bridge to start a series of monthly inspections of the plant. At the original station freshets had apparently wiped it out, but at the western end of the bridge the plant, in flower, was abundant about cold spring-heads, thence extending far out into the submerged borders of the bottomland. Collections then made, with others up to late autumn, show it to be good L. radicans (MaP 3), with considerable leaf-variation, but the plant, which fruits in the Mississippi basin, was here all sterile, whether because too thoroughly drowned or because of its efficient vegetative repro- duction we cannot say. Another year, if flowering freely above high-water mark, it may produce fruit. Although L. radicans (Steironema radicans) occurs in Virginia, I am unable to locate any specimens from other Atlantic states. Its chief development is along the lower Mississippi Valley. This characteristic isola- tion in Virginia is one which no longer surprises us. With it was a Galium which puzzled us. Its fruit and foliage-form were those of G. obtusum Bigel., a species ordinarily with whorls of 4 leaves. The Cypress Bridge plant was producing 5 and 6 leaves at the primary and some secondary nodes. This was hardly regular, but flowering material, collected in June, 1940, shows the corollas of G. obtuswm; we will leave it at that. For some years we had looked suspiciously at the group of Rhynchospora cephalantha Gray and R. microcephala Britton. The latter is common in peaty soils of southeastern Virginia, the former rare and confined to sphagnous bogs. But that 1s not all. A third plant, also rare, somewhat like R. microcephala but clearly different, had been found; and in drowned peaty hollows and margins of pools there is a fourth, very much coarser than any of the others. It was now blooming and we made a point aa EXPLANATION ol Exssting faposed affer Lag Pagina! Of PLELSTOCENE sce. Pe ineullean? cere. Orifttess a nunatak areas or portly exposed PLEISTOCENE, & Areas mountain avellay Oring the FULTSTOCENE. . . ty YF x LA tH ite RE ey ee ane a eh P, AG it mt ER goats bales Ui: ey 2: iss li eS OR Fra aN } eo SY i A . =2=— = eo : o mn =o, VF —— = Map 1, Range of TeTrRAGONOTHECA HELIANTHOIDES; 2, of IsofreEs ENGEL- MANNI var. CAROLINIANA; 3, of LyYSIMACHIA RADICANS; 4, of AMMANNIA Korner; 5, of Juncus MEGACEPHALUS; 6, bicentric Range of Litazopsis CAROLINENSIS. of securing it, with the hope that its anthers might give good characters. Since my student, Miss Shirley Gale, had for two years been closely studying and monographing Rhynchospora 368 Rhodora [Ocroser in North America, I asked her help on this problem. Our results will be included in Part II. One rainy and dreary morning (the 19th), disliking to spend the day, when identities of plants are so disguised by mist and loads of rain-drops, upon an open habitat, we reasoned: since we are bound to get soaking wet and it is a dismal day, why not tackle the Great Dismal Swamp? Accordingly, in the after- noon, at the lumber camp of the Camp Lumber Co., southeast of Suffolk, we followed the lumber-road into the Swamp. As we already knew, the forest and its flora are essentially those of many large flooded swamps and pocosins or dismals in southern Virginia. However, Ilex coriacea (Pursh) Chapm., which we had met nowhere else, is there, and we were particularly pleased to make its acquaintance and to see how really different it is from I. glabra. In October the contrast became emphasized, for, whereas J. glabra retains its hard drupes through the winter and is used as a Christmas green, the drupes of J. coriacea in October were soft and pulpy and most of them had dropped. Only by long search could we find any branchlets with drupes still intact. A smooth-leaved Persea interested us and I had a vain hope that we had at last found P. Borbonia, but it is not that; in outline the leaf is similar to that of the pubescent P. palustris (Raf.) Sarg. Later in the season we secured it also in wet depressions of the pine barrens of western Nansemond. Its identification must await further study, for the material of the group from most American herbaria is now buried in Hol- land, where, long before the war, it was sent on loan. On our second trip to the region of Taylor’s Millpond we re turned to Fontaine Creek, where the bottomland woods are neat the northern limits of some species. We were quite unprepared to find them the extreme southern limit of any range but in one recently cut area there was a good colony of the northern Lactuca canadensis var. longifolia (Michx.) Farwell,’ heretofore know? only farther north or inland, from Quebec to Saskatchewan, south to New England, New York, the upland of North Carolina, Ohio, Indiana, Missouri, etc. We had found, when we had only a fragment of the day free for botanizing, that the waste places, dumps, neglected road- 7See Fernald, Ruopora, xl. 481 (1938). 1940] Fernald,—Additions to Flora of Virginia 369 sides and the railroad yards (especially that of the Norfolk and Western at the eastern margin of Petersburg) supplied some novel weeds; furthermore some weeds, rare or scanty in 1938 or 1939, like Potentilla recta and Chondrilla juncea, have in 1940 become very abundant and aggressive. So, on the 21st, having only the forenoon clear, we set out to weed the town. Botani- cally, if not economically, the operation was successful and we brought back a large series of weedy Euphorbias, including the southern Euphorbia prostrata Ait., which had been known in the state only through collections of Grimes’s. Acalypha ostryae- folia, very distinct and almost handsome, abounds on steep weedy slopes near the Appomattox but along the Norfolk and Western, where it might have been indigenous, but too doubtfully so, since, except for a somewhat shaky record for New Jersey, it is other- wise known in the Atlantic States, from South Carolina south- ward, primarily as a weed. On one open and weedy patch there was a tall amaranth, with prolonged but much interrupted spici- form inflorescences. This proves to be Amaranthus Torreyi (Gray) Benth., a western prairie type, a species which may soon abound about Petersburg. Our earliest Virginia work, from 1933 to 1935, centered on Virginia Beach. We then became very cognizant of the strong contrast between the flora of the open shore and marshes from Cape Henry to Sand Bridge and the western side of Back Bay and that of the eastern shore of Back Bay and the islands (Knott's Island and Cedar Island, the only ones formerly Visited). Along the mainland shore or on the west side of Back Bay several species occur which we did not see on the islands and the eastern shore of the Bay. Conversely, there are many Species on the islands or at False Cape which we have never ound on the west side of the Bay. Thus, along the shore or on the sands between Cape Henry and the inner shore near Munden the following, among others, are characteristic, yet we have never got them from the outer shore of the Bay or from the islands: Aristida lanosa, Cyperus retrorsus var. Nashii (Britton) Fern. «& Grise., C, erythrorhizos, Fuirena pumila Torr., Xyris difformis, Tillandsia usneoides (on trees), Sesuvium maritimum, Nelumbo and Heliotropium curassavicum. On False Cape or Cedar Island the following, unknown on the mainland side of the Bay, we knew 370 Rhodora [Ocroser to be characteristic: the austral Phalaris caroliniana Walt. and Eleocharis albida; the boreal E. halophila Fern. & Brackett; and E. Lindheimeri (Clarke) Svenson at the only known coastal area east of Texas; the southern Dichromena colorata and Juncus megacephalus; and the remarkable Iresine rhizomatosa of the Mississippi basin and southwestward but with another isolated area on the Potomac; and Ampelopsis arborea and Lippia nodi- flora. These contrasts indicated that the islands of Back Bay need much more exploration by competent botanists. Conse- quently, when that energetic and successful champion of con- servation of the flora of Cape Henry, Miss Sally Ryan, invited us to her home at Virginia Beach and asked me to suggest some area where we could do worth-while botanizing, I suggested se- curing a motor-boat and getting out to the islands. On the 22nd, therefore, with Miss Ryan’s co-worker on con- servation, Miss Mary Leigh, my correspondent for some yeals, and with Mr. George Mason, Director of the Mariners’ Museum at. Newport News, we drove to the headquarters (C. C. C. camp) of the National Wildlife Refuge at Pungo. The friendly Man- ager of the Refuge, Mr. Harry A. Bailey, with his associate, Mr. Ewell, a native of the region, drove us down the beach to the developing headquarters of the Refuge, back of and below Little Island Life Saving Station. Thence we were taken in one of the splendid launches of the Biological Survey by Mr. Ewell to the southeastern corner of Long Island, and he waited pa- tiently among the mosquitoes and flies for some hours while we waded along the marshy margin of the Island. Before we had left the plank-walk leading from the landing at the outer border of the marsh we wanted to jump in, to get the rare plants which were in sight. Najas guadalupensis and Anacharis occidentalis, flecked with Lemna perpusilla, filled the water; Ammannia Koeh- nei, MAP 4 (represented in our larger herbaria from only ® im stations between Florida and the Hackensack Marshes), in solid phalanx, and the white and lilac-tinged flowers of the sub-tropical Bacopa Monnieria and the panicles of the local marsh grass, Diplachne maritima Bicknell, were conspicuous We waded at the inner margin of the marsh (in fresh water, oat were assured, but with a remarkable number of halophytes; Scirpus robustus, Spartina cynosuroides and Lythrum linear’ 1940] Fernald,—Additions to Flora of Virginia 371 through shoulder-high and taller vegetation. Progress was slow and difficult but we wanted to go slowly in order to see the plants. Amidst the commoner marsh species there were limited colonies of Dichromena colorata and of Juncus megacephalus (map 5) which we had known in Virginia only at False Cape, the latter following the coastal marshes northward, but in the south- ern half of its range venturing inland, and unmistakable Verbena scabra Vahl, a tropical species which we already knew as an ex- cessively local plant in eastern Virginia. At the border of the marsh, beneath the bushes, Carex Frankii, an inland species, had reached the coast, although we had never before met it in Princess Anne County; and farther back, just as on Cedar Island, the Mis- sissippi basin Jresine rhizomatosa abounded, and Physalis angu- lata, the first we have had in the Gray Herbarium from north of North Carolina, was occasional. We were approaching the small pond back of the marsh and south of the landing, and Mr. Ewell Was in sight, coming to say that it was time to leave, for it was important to get back to Little Island in order to take advantage of low tide in driving up the beach. It was necessary, however, to see what was in the pond. It is fortunate that we did so, for this pool contains a continuous carpet of closely interlocked plants of Lilaeopsis carolinensis Coult. & Rose (Map 6). Li- laeopsis, a very primitive genus of the Umbelliferae, without true leaves but with clavate or broadened and jointed stems function- ing as leaves, has one species on tidal shores of Atlantic North America from Florida to Nova Scotia, another on the Pacific coast from California to British Columbia, others (very local) In Arizona and Mexico and the others (except L. carolinensis) local species of various parts of South America (tidal shores to more than 4000 m. in the Andes), Tasmania, New Zealand and Australia. The antiquity of the genus (Map 7) is evident; and at the last moment on Long Island we were pulling in mats of L. carolinensis. Described in 1897 from a single collection from eastern North Carolina (presumably near Wilmington), it is now known from three other stations in North America: Myrtle Beach, South Carolina; New Orleans; and our new sta- tion on Long Island. But along La Plata River in Argentina, Uruguay and Paraguay it is evidently frequent. L. carolinensis 1s clearly a bicentric species, with one area in warm-temperate 372 Rhodora [OcroBer eastern South America, there reaching its southern limit near lat. 35°, the other area in warm-temperate eastern North America, with the northern limit near lat. 36°. Such ranges will find amplification when we consider other plants of fresh tidal waters and marshes collected in the late summer and autumn of 1939. We were deeply grateful to Miss Ryan for having arranged so successful an expedition and to all our new friends who had helped her carry it through; but she and Miss Leigh, with the remarkable enthusiasm which they share, were not through for the day. When we reached headquarters near Pungo, although we had already had a long day and were soaking-wet, they offered to show us some of the specialties; but, since none of the party had ever seen Asarum arifolium, which Griscom and I had found in May, 1935, in woods near Creeds, we went in search of that. The station seems to have gone with its protecting for- est. After that the ladies introduced us to their station for Stewartia Malachodendron and took us to other interesting plants. When, after dark, we brought up at a fashionable restaurant for late dinner, the well-groomed official haughtily looked us over and shook his head. After demonstration that we were more respectable than our bedraggled clothes, we were ushered into a secluded room and allowed to eat where the sport- ing fraternity of Virginia Beach would not be shocked. Long and I were dead-tired when we reached the Ryan home; the ladies could have stayed up all night, going over the specimens! Next morning we drove, not too far away, in the vicinity of the Beach, for Long and I had to catch a noon train at Norfolk. Having heard that the area of sphagnous peat south of the Rifle Range was greatly altered and that deep ditching had obliterated the only known station in the world for the unique cleistogamous Hypoxis Longii Fernald in Ruopora, xxxvii. 410, pl. 393 (1935); both Long and I, as responsible for its original recognition, wer? anxious to know its fate. Since the discovery of the plant the Rifle Range has been greatly altered and, whereas we used to go and come at will in the region to the south, the area is now fenced off and with military guard, for not only rifle-practice but mortar-firing and other military training have greatly creased. The Commandant, as soon as I explained my errant 2 o — key qe) ler} S ll Se > [oh Qu —_s o =) 3 ™M oc ° es —— ° ba} 9 pe => < —e ng 03 es =| bo b=) ie) =] 1S) (GOODES POLAR, EQUAL AREA PROJECTION FOR THE CONTINENTS BY ) PAUL GOODE JULY i, 1928 prommctan, 8 ee harem ncaa Phe Ms ae The sone peter separ te premented Gan meet at the bef on cartier asumithe epost seme ie tt EEL A Veh YET BY KORO IC SAL SS LE Qe tenaene pike Bama cA f One PTF AK ALG. wy Zee A ROL RCCL ee ae) eee s is w cA ry OS TD a oss NGS We ee ae av at SS sa Fa : \K ye" aul Ms ie AI Sy ox 4h. eV Le PHY | LT * te 4 ¢ = ~ 40%, shod FT RIES - ae NS {? pi @ ey, LI Fh) Lie ‘ ‘ CS, P Ya a ol She Oe a [BES és be y Rs oe ee oe SMS CF SST xs j \) \ WEE CTS Map 21, Range of CypeRUS BREVIFOLIUS. the Chickahominy into the James, below Barrat’s Bridge (for- merly Barrat’s Ferry, the bridge opened during the summer of 396 Rhodora [OcToBER 1939). The still pretty nice series of species abounds there; but under water, completely submersed when we found it, was the tidal-shore form of Xyris caroliniana with floating ribbon- like leaves, which Long had been sending me from the lower Delaware. Here, again, was another identity, though this only of minor taxonomic importance, between the flora of the lower Delaware and of these southeastern Virginian tidal shores. The ereat prize of these marshes, however, was the stoloniferous and extensively creeping pantropical Cyperus brevifolius (Rottb.) Haussk. (Kyllinga brevifolia Rottb.), widely dispersed in tropi- cal and subtropical regions (Map 21) but heretofore unknown in eastern North America between its stations in Florida and ad- jacent Georgia and the isolated colony on the lower Delaware. This was pretty fine but, after helping dig a good series of plants and leaving Long on dry land to lay them into paper, | pushed farther out into deep water and found myself in a colony of @ purple-rayed Boltonia. This was the third species of the genus we had found in eastern Virginia and quite like one which Gris- com and I had collected on our late-September trip in 1933 on the tidal marshes of North Landing River. By current treat- ments all three go into the too inclusive B. asteroides, but the plant of tidal marshes is freely stoloniferous, the other two species of southeastern Virginia not at all or but slightly 50, and one of them has tiny white heads. The discovery of the plant of North Landing River had led Griscom and me to at- tempt a revision of the group. Our tentative results were held back pending receipt of photographs of types from abroad; but this discovery of a new colony was the signal to revamp and strengthen the unpublished treatment of seven years ago. This, with illustrations, will be found in Part II. Forced out by high tide, we drove west, hoping to beat the incoming tide up-river. We took a chance on the shore near Tettington, but there we found sand-beach, with vigorous weedy colonies of soy bean and peanuts and other signs of man’s in- vasion. The native flora, however, was interesting, for here, far up the James and mingled with typical plants of fresh sands, there were colonies of Spartina patens var. juncea, Pamcum amarum and other maritime species. Best of all, the uppe border of the beach was covered with characteristic Apocyn’” 1940] Fernald,—Additions to Flora of Virginia 397 sibiricum Jacq., the northern white-flowered species which Wood- son, in monographing the genus, recorded southward in the At- lantic coastal region only to Delaware and the District of Columbia. Still having a remnant of daylight when we reached Charles ' City Courthouse, we drove northward for a ten-minute glimpse of the Chickahominy at Long Bridge. The ten minutes were well rewarded. On the wooded bottomland in New Kent County we might have been in Southampton County, 50 miles to the south. Here were Leersia lenticularis, which we had never seen north of Sussex County, Hypoxis leptocarpa Engelm. & Gray, a southern species which was new to Virginia when we found it in Southampton, and Lysimachia radicans, with which we had been keeping regular appointments at Cypress Bridge in southern Southampton. The bottomlands of the Chickahominy evidently need close study; but we had been out since an early breakfast, had explored six remote localities and, with darkness coming on, a willing to leave the Chickahominy bottoms and “call it a ay”, Our September work had led us to the region north of the James, but we had not wholly forgotten the allurements of the Pine-barren regions of western Nansemond and southeastern Southampton. The sandy pine barrens and pinelands several miles south of Franklin, where “Long’s Flannel-weed”, Chrysop- sis Longii Fernald in Ruopora, xl. 467, pl. 531 (1938) and Trade- scantia rosea Vent. var. graminea (Small) Anders. & Woodson abound, are always fascinating. On most of our visits heavy tains had made the wood-roads too full of water-holes for com- fortable driving. On September 20th, however, the roads being Well dried out, we made a circuit out to Point Beach on the Nottoway (where Styrax americana is very fine), thence north- Westward to Round Gut on the same river and eastward via Wiggins School to the automobile road. Fine material of many ‘are species was collected and we were delighted to find Zenobia pulverulenta in Southampton County, much taller than in Nan- “emond (even if Pursh, nearly 140 years ago, “beat us to it” by collecting in Southampton the very rare Litsea geniculata, Which has not subsequently been found, he did not discover Zenobia!). The local Trichostema setaceum Houtt. (T. lineare) 398 Rhodora [Ocroper abounded and was very large, and in a patch of hickory and oak woods (usually relatively rich) Kuhnia ewpatorioides was at its easternmost station. In the sandy woods at Round Gut the Tephrosia spicata looked unusual. This was because the plants are nearly glabrous, instead of densely pilose-villous. I have been tempted to glorify the type-locality of this plant by apply- ing to the latter the name splanchnodita, but the glabrous or glabrescent plant occurs at other stations; I am, consequently, giving it a less suggestive name in Part II. Chinquapin, Castanea pumila, was heavily fruiting in the bar- rens and very puzzling. Some shrubs had the burs densely covered with erect long-rayed scales, others had the rays or bristles short and erect, while in others the remote scales had depressed and horizontally divergent rays, their tips often not touch- ing and thus leaving broad naked areas. Such differences in the cups of acorns would be strongly specific, but in Castanea pumila the different variations all have essentially uniform foliage and nuts. We ate the nuts inordinately and filled all receptacles we could find with more nuts to take home. Tragically, however, when we opened the containers at home, we found them squirm- ing with the fattest of grubs. We could merely take comfort by remembering the ancient conundrum about the half-worm in the partly eaten apple! I have carefully compared our ma- terial with the extensive series at the Arnold Arboretum. The shrub or small tree with broad naked spaces on the involucre is C. pumila var. Ashei Sudworth, heretofore known from south- eastern Texas, Arkansas, and Louisiana to Georgia and north- eastern North Carolina. The one with very long erect bristles is a close match for C. pumila var. Margaretta Ashe, which Ashe (for years familiar with the Carolinas) knew chiefly from “the upper edge of the longleaf pine lands of Texas and Louls- ana”. Ashe did not know it from the Atlantic States (only from western Alabama, Mississippi, Louisiana, Arkansas, Okla- homa and Texas). This isolation in southeastern Virginia like that of scores of other plants; the next step is to find va! Margaretta and the other plants in the Carolinas and Georg! Returning on the 22nd to the pine barrens of western Nanse- mond County, I asked Leonard to stop the car at the wood- road leading to Milk Landing, where in July he and Long had 1940] Fernald,—Additions to Flora of Virginia 399 been unmercifully peppered with seed ticks. I knew that I must start alone, for the memories of the place were still too sad for me to hope for Long’s company. But he soon caught up with me, fearing that I might discover something interesting and correctly reasoning that the seed ticks of two months before were no longer waiting for him. And when we got to Milk Landing the Blackwater was at low tide. The little ribbon of tidal vegetation under the overhanging shrubs was tiny and interrupted but in five minutes we had Sagittaria subulata, the submersed form of Xyris caroliniana and, near the head of tide which runs up from Albemarle Sound in North Carolina, Aneilema Keisak. It surely must grow in North Carolina as well as in Virginia! Slightly farther down-river, at Cox Land- ing, we got Eriocaulon Parkeri only 314 miles north of North Carolina. The sphagnous savannah-like bog at Tom Hunter’s was gor- geous with the orange and brown heads of Coreopsis oniscicarpa Fernald in Ruopora, xl. 472, pl. 533 and 534 (1938). The plants were up to 9 dm. high, often with 30 to 60 heads. This is the greatest station we know, but the species continues abun- dant eastward to the region of Cleopus. South of this area, not far from Cathole Landing, the white sands support, among super- abundant Carphephorus bellidifolius, fine colonies of Phlox Hentzii Nutt. (discussed in Part II) at its easternmost known station, Eupatorium tortifolium Chapm., which we had already got near Sandy Landing, a northern extension; and, the great prize of the day, extensive carpets of a prostrate Desmodium, how over-ripe but quite strange to us. It proves to be true D. glabellum (Michx.) DC., quite unlike the rare plant which Passes in our manuals under that name, the latter being D. humifusum Beck. Michaux’s South Carolina type is represented in the Gray Herbarium by a fragment (leaf) and by a very clear photograph. Otherwise, so far as I can find, the species 18 unknown. In late August, 1940, we secured a representative Series of specimens. These will be discussed in a later report. The discovery of Desmodium glabellum at dusk closed our ex- Ploration of the native flora for the month, except that on the 23d we went to Carey Bridge to get flowering material of Chelone obliqua (found too young earlier in the season). 400 Rhodora [OcToBER The regular 2-hour weeding of Petersburg, after the presses were finally emptied and we awaited the afternoon train, brought us, among other unusual adventives, a labiate which greatly puzzled us. It proves to be Hyptis mutabilis (A. Richard) Briq. var. spicata (Poit.) Briq., the West Indian and Floridan repre- sentative of a tropical American species. At Petersburg, grow- ing on a weedy bank, it is far from home. The weed which makes the fortunes of many residents of Petersburg and of Rich- mond is Nicotiana Tabacum. We were more interested in Hyptis mutabilis! In October (12-17) our limited time was mostly devoted to the fresh tidal shores, although we started off by returning Me the Great Dismal Swamp, near the Camp Lumber Companys plant southeast of Whitemarsh School. We were primarily after mature fruit of Ilex coriacea, already noted; but we found the wonderfully developed phyllodia of Sagittaria Weatherbiana Fernald which will be specially noted in Part II, extended east- ward the ranges of Leersia lenticularis and Scirpus divaricatus, and, in the clearings, found unusually strong and heavily fruiting plants of Viola esculenta Ell. : On the 13th we returned to “Shady Rest” for mature fruit of Aneilema Keisak. On this trip we noticed that the deep chan- nel which runs through the marsh was full of Potamogeton epihydrus var. Nuttallii (C. & 8.) Fern., here found for the first time south of the Potomac, a fitting companion for the northern P. Spirillus which we got here in September. High tide forcing us from the Chickahominy, we then went north to the York, where we could still have four hours of low water. We spent some time on the sandy beach north of Holly Forks, this pe of the upper York being essentially sea-shore, with carpets © Euphorbia polygonifolia and other maritime plants. On a ieee sandy slope there was an abundance of Sporobolus asper, whie we had never met in Virginia. Upon referring to Hitchcock s Manual it becomes evident that the species is new to ms ae It was obvious that for fresh tidal marshes we must go UP ‘ ; Pamunkey and the Mattaponi which unite to form the — salty York. So we drove up the former river and tried : marshes east of White House. Aneilema, the new Cass4 od Aeschynomene were there and we at once set ourselves the tas 1940] Fernald,—Additions to Flora of Virginia 401 EXPLANATION a Existing glece/ telds. fiposed after ats Co) or Ateirocene re woosed Oritttess « nunatak areas ‘or portly ex ering PLES: MEL Areas ye : a . : US §* 9° > =4 Peay ae { C4 a9 a OBB é a Map 22, Range of Bacopa cyctopHyLia (derived largely from Pennell) ; 2 of PoLYGONELLA aRTICULATA; 24, of BACOPA oBovATA; 25, of CARDAMINE NGI; 26, of GenTIANA VICTORINII. of following these three index-species to the northern limits of their range (a task not yet finished). Next day, after some exploration in other areas, we reached the Mattaponi at Horse Landing, near King William Courthouse. The shores were so fine and the tide so unaccommodating that we Teturned for a second day, at low tide. All the standard species, Including the variety of Rhynchospora macrostachya, were there, &S were the less common Eriocaulon Parkeri and Elatine ameri- cana, Spiranthes cernua var. odorata was there producing young 402 Rhodora [OcToBER plants (rosettes) at the tips of the prolonged roots;!° and in the outer mud, exposed at extreme low tide, Hypericum mutilum var. latisepalum Fernald, heretofore known only from Florida to Texas, was mixed with scattered individuals of the excessively rare and little collected Bacopa cyclophylla Fernald (Herpestis rotundifolia Gaertn. f., not B. rotundifolia (Michx.) Wettst.), this being the eighth known station and the first between Wil- mington, North Carolina and the two colonies on the Eastern Shore of Maryland (map 22). In southeastern Virginia Bidens coronata (L.) Britton is chiefly represented by var. trichosperma (Michx.) Fernald in Ruopora, xl. 350, t. 506, figs. 8 and 9 (1938); but along the Mattaponi some plants had coarse and often simple leaves and very large broad-based awns. They can be referred only to typical B. coronata, which in the Atlantic States had been unknown south of the lower Delaware. Farther up river, opposite Walkerton, the marsh was bordered by a towering thicket of Aeschynomene, so dense and 80 high above our heads that we finally abandoned the last lingering pretense that it is rare! And slightly below Walkerton, on the King and Queen side of the river, the index-plants were all seen, and with them Bidens coronata. Having trailed the quarry t King and Queen, the obvious step was to hunt for the group on the next river to the north, the Rappahannock. We made a tactical error, however, in going toward Tappahannock, for there the marshes are salt and it was obvious that we must g° much farther up river for fresh tidal shores. Consequently, We hour being late, we went to the nearest open shore, at Richmon Beach, on the Rappahannock in Essex County. The a character of the area was evident from the abundance on . sands of Diodia teres var. hystricina Fern. & Griscom of ee sands of Cape Henry and of Yorktown. In the thicket pa of the beach the giant Arundo Donaz, up to 15 feet high, he thoroughly naturalized. We selected small panicles ere folding them back and forth and tying their tips to the et of the inflorescences secured specimens which can be kept W! the limits of the herbarium-sheet. Quay. One trip was made to the pine barrens south of South : habit”—D- © Correll says: ‘often in dense clumps because of its stoloniferous Correll, Bot. Mus. Lfts. Harvard Univ. viii. 81 (1940). 1940] Fernald,—Additions to Flora of Virginia 403 At Tom Hunter’s we were happy to find the new Liliwm with mature fruit and abundant seeds; and in one seeping or springy spot Utricularia fibrosa, rare in Virginia, was flowering. Quer- cus cinerea Michx., abundant in all the pine barrens of south- eastern Virginia, is sometimes well-behaved, but it often mixes with the other oaks. One of the most striking hybrids, of which we found a characteristic tree, is X Q. subintegra Trelease (Q. cinerea X falcata). In 1940, at the northern limit of Q. cinerea, in Sussex County, we felt that its hybrids with various species were almost as abundant as typical Q. cinerea. One of our chief errands at this time was to collect, at last, flowering material of the Polygonella, of which we had found seedlings in July. The plant was very abundant, and we traced it across the state-line, nearly to Wyanoke in Gates County, North Caro- lina. The larger plants, heavily loaded with flowers and fruit, were 8 dm. high, but, alas, they are only the northern P. articu- lata (Map 23). Alas!—nevertheless the species is really a most interesting one to find in southeastern Virginia and northeastern North Carolina. There is an old specimen of Thomas Nuttall’s at the Philadelphia Academy, marked simply as from “Georgia”. Otherwise, there are no specimens of this essentially northern species in the herbaria of the Philadelphia Academy and the New York Botanical Garden nor in the Gray Herbarium from south of the Eastern Shore of Maryland. Our disappointment n not having a typically more southern species was, conse- quently, tempered. us our season of botanical exploration came to a close. It had been one of unusual discoveries in view of all the previous Work in the same area and, even if the record shows less than one hundred maintained species and geographic varieties, which by Some would be treated as species (for instance Andropogon virginicus, var. glaucus, maintained by Nash, Small and Hitch- Cock as a species, A. capillipes Nash), the inclusion of several Well marked forms which deserve designation (some of them usually called species, as in case of Commelina crispa Wooton or the 8reen-leaved form of Zenobia, maintained by Small as a Species, Z. cassinefolia) fully justifies the title I have given to 18 Teport of progress. The records in abbreviated form con- stitute Part II; for the convenience of users some of the data 404 Rhodora [ OcroBER diffusely stated in the Narrative is repeated. In Part III brief consideration will be given the phytogeographic problems raised by some of the discoveries, especially those of the fresh tidal shores. Part II. ENUMERATION AND CONSIDERATION OF THE MoRrE NoTeworTHY PLANTS As in previous reports species, varieties and forms of signifi- cance in reaching an understanding of the flora of the Coastal Plain of southeastern Virginia are enumerated. Some records are from collections of earlier years; a few are of plants in the Gray Herbarium, collected by others; and in a few cases speci- mens collected in 1940 are included in order to complete the record to the date of going to print. In most cases, where the plants were collected by Long and me, it has seemed unnecessary to repeat the names of the collectors. In course of identifying the material many genera or groups of species have been criti- cally studied. In so far as these revisional studies have grown out of the Virginia work they are here included. The photo- graphs and material for the plates have been prepared by my assistant, WALTER H. Hopcr, or by my son, Henry G. FERNALD. The maps showing world-ranges are on base-maps of the Goode series, copyrighted by the University of Chicago. The initial cost of photography and preparation of blocks has been partly met through appropriations for personal research from the De- partment of Biology of Harvard University; the cost of their re- production through the generous support of Mr. Long. The names of plants thought to be reported for the first time from Virginia are preceded by an asterisk (*). CHEILANTHES LANOSA (Michx.) Watt. Sussex County: ledges in rich woods at the “fall-line” along Nottoway River, above Double Bridge, about 6 miles northwest of Jarratt, no. 10,862. Close to the inner border of the Coastal Plain. See p. 38%: EQUISETUM ARVENSE L. SourHampton County: in lime-marl, wooded bottomland of Three Creek, northwest of Carey Bridge, no. 10,071. Equisetum arvense, so common in all damp habitat North, is rare in southeastern Virginia, and always, so far as have observed, in caleareous pockets. See p. 363. s in the we ee ey ee ee 1940] Fernald,—Additions to Flora of Virginia 405 southwest of Grizzard, no. 10,866. In originally describing the variety, Chapman used the spell- ing Lycopodium inundatum, var. appressum Chapm. in Bot. Gaz. iii, 20 (1878). Subsequently, however, he took up the alterna- tive spelling, var. adpressum Chapman, FI. So. U. 8S. ed. 2: 671 (1883), holding to this altered spelling in the 3rd edition (1897). Most authors have used the second (and by Chapman obviously preferred) spelling. If it be maintained by some that an author has a right to correct his own error (assuming that Chapman so considered his first spelling), then, immediately, an equally strong group will argue that the original spelling must stand. As one who has been forced through typographic, orthographic or stenographic errors into misspellings, I have claimed the right to correct them. This is quite different from alterations made by others. In the latter cases, unless an evident error is cor- rected they are not justified. In the case of Chapman’s name, of Course, the original spelling was etymologically as correct as the substitute. “L. CAROLINIANUM L. GREENSVILLE CouNTY: argillaceous and Fi meadow northwest of Taylor’s Millpond, nos. 10,075, Certainly a very rare plant in Virginia. We have met it only at this station, a much burned peaty meadow or swale, where the Lycopodia of bogs have a remarkable and perplexing develop- ment. I have been unable to trace the source of the record for Virginia, sometimes given. Lloyd & Underwood in Bull. Torr. Bot. Cl. xxvii. 158, 159 (1900), cited no material from between southeastern North Carolina and southern New Jersey, although there is a single station in Maryland. In August, 1934 (Clay- tonia, i. 3) Massey, in his account of the genus in the state, said: “No reports of its having been collected in Virginia are at hand.” See p. 35 IsolitEs sacCHARATA Engelm. Fresh tidal shores. Kine Wiutam County: Mattaponi River, at Horse Landing, near King William Courthouse, no. 11,508. New Kent County: 406 Rhodora [OcToBER Chickahomimy River at “Shady Rest”, southeast of Windsor Shades (Boulevard Postoffice), no. 11,507. CHARLES City County: James River at “Four Oaks”, below Harrison Point, no. 11,216. Nansemonp County: Blackwater River, Milk Landing, south of South Quay, no. 217. SouTHAMPTON County: Nottoway River, Knight’s Seine Beach (Battle Beach), no. 10,082. Pfeiffer, in her Monograph, cites no material from south of tributaries of the Potomac near Alexandria. Knight’s Seine Beach is barely not in North Carolina; I. saccharata will doubt- less be found in that state farther down the Nottoway or on the Chowan. See pp. 364, 391 and 393. . Encetmanni A. Br., var. carotiniana A. A. Eaton. GREENSVILLE, SOUTHAMPTON, NANSEMOND, Sussex and DinwiD- pie Countiss, frequent and often abundant on the wooded bot- tomlands of the larger rivers and creeks (Meherrin and_Not- toway Rivers, Fontaine Creek, Three Creek, Somerton Creek, Rowanta Creek) and even along small runs in the woods, fruiting from mid-May to August, usually in June, most colonies being eee and nearly unrecognizable by mid-summer (many nos.). Extension north from Georgia and mountains of North Caro- lina. See pp. 358, 367 and 375 and MAP 2. TaxopIUM pDisticHuM (L.) L. C. Richard, var. IMBRICARIUM (Nutt.) Sudw. (7. ascendens Brogn.). Common on wooded bottomlands and even on tidal shores of rivers, many collections from James City, Sussex, Southampton and Greensville Counties. Influenced by the persuasions ‘of those who see two species of Taxodium in the South, we called our collections made in June T. ascendens. Finally, however, noting that the characters sepa- rating the two were very fickle, we reached the conclusion of Sudworth and of the late Carl Mohr. The latter’s statement 18 one with which we have full sympathy: . . 544 ns orm passes freely into the species where the soil conditio are more favorable—Mohr, Pl. Life of Alab., Contrib. U. S. Nat. Herb. vi. 325 (1901). After a strong wind in early summer, many of the deciduous 1940] Fernald,—Additions to Flora of Virginia 407 branchlets bearing only the small leaves of the variety will be found on the floors of bottomlands. In Virginia, at least, this form is not confined to pond-margins, as it is sometimes said to be farther south. Probably var. imbricarium is not a true variety, but merely a state of development or a seasonal stage. TYPHA TRUXILLENSIS HBK. Princess ANNE County: brack- ish to fresh marsh along Back Bay, at eastern margin of Long Island, no. 10,476. Sussex County: argillaceous ditch by Nor- olk and Western Railroad, about 2 miles west of Waverly, no. 10,477. See pp. 374, 376 and map 8. Recorded in 1935 (Ruopora, xxxvii. 385) from shores of Back Bay, the first area known north of Florida. The station in Sus- sex County is 65 miles inland and extends the range northward to lat. 37° 2’, from the former northern limit at 36° 40’. *PoraMoceTon Sprrittus Tuckerm. New Kent County: Floating at outer border of fresh tidal marsh by Chickahominy River, at “Shady Rest”, southeast of Windsor Shades (Boule- vard Postoffice) , no. 11,219. Extension south from the lower Delaware River and adjacent waters of southwestern New Jersey and of Delaware. See p. 374. P. epinyprus Raf., var, Nurrauim (C. & §.) Fernald. New Kenr County: in open water, fresh tidal marsh by Chickahom- iny River, at “Shady Rest”, southeast of Windsor Shades (Boule- vard Postoffice), no. 11,541. Extension south from the Potomac. See p. 400. TRIGLOCHIN stRiATA R. & P. To the station on North Landing River, recorded in 1936, add the following, also in Princess Annr County: inner border of brackish to fresh marsh along Back Bay, at eastern margin of Long Island, no. 10,479; turfy knolls in shallow pools in brackish to fresh marsh along Back ay, east of Creeds, no. 10,876, very fine development, plants hearly 3 dm. high, with spikes more than 1 dm. long. See PP. 387, 388 and map 19. Sacirrarra supuLaTa (L.) Buchenau. To the few recorded stations add one in James Crry County: tidal mud along Pow- atan Creek, north of Jamestown Island, no. 10,877. Kine Wittam County: fresh tidal shore of Mattaponi River at Horse Landing, near King William Courthouse, no. 11,512, Nanse- MOND County: muddy tidal margin of Blackwater River, Milk Landing, south of South Quay, no. 11,220. See pp. 386 and 399. SUBULATA, var. NATANS (Michx.) J. G. CESS Anne County: shallow pools in fresh to brackish marshes along Back Bay, east of Munden, no. 10,878; east of Creeds, no. 10,879; 408 Rhodora [OcToBER along Nowney Creek, southeast of Back Bay Postoffice, no. 10,880. See p. 387. I am keeping as var. natans the narrower-leaved plant which was included by Chapman, along with the coarser Florida plant, under his Sagittaria natans, var. lorata. As I interpret S. subu- lata it consists of four fairly marked varieties as follows. S. supuLata (L.) Buchenau, var. typica. Alisma Pgh S. subul .) Buchenau in Abh. er. Bremen, ll. 4 (1871) —Dwarf; leaves linear, strap-shaped, obtuse to acutish, 2-12 cm. long, mm. broad, exceeding to shorter than scape, or.-Am. ii. 190 (1803)—photograph of type in Gray Herb. S. natans, var. lorata Chapm. Fl. So. U. 8. 449 (1860), in part coast, Florida to southeastern Virginia. Cl ar. GRACILLIMA (Wats.) J. G. Smith in Mem. Torr. Bot. : v. 26 (1894) and N. Am. Sp. Sagittaria and Lophotocarpus, 19, pl. 14 (1894). S. natans Michx., var. (?) gracillima Wats. in Gray, Man. ed. 6: 556 (1890) —Very elongate, up to 1 m. a more long, and submerged; leaves (phyllodia) prolonged, : ua mm. wide; scape prolonged; inflorescence 1-3 dm. long, with fe very remote whorls; bracts (at least of the upper whorl) = t herbaceous, elongate, mostly caudate-tipped, nearly poorer 6-10 mm. long; pedicels all elongate, the lower with pistil flowers, arched or spreading, 0.3-2 dm. long, not much thicke? , fruit unknown.—Deep water of streams, eastern Massachuse to southeastern Pennsylvania. 1940] Fernald,—Additions to Flora of Virginia 409 Var. lorata (Chapm.), comb. nov. S. natans Michx., var. lorata Chapman, FI. So. U. 8. 499 (1860) in great part. 8S. lorata (Chapm.) Small in No. Am. Fl. xviit, 52 (1909), as to type—The coarsest extreme; phyllodia 0.8-1.5 cm. broad, 2-9 dm. long; in- florescence with 3-6 whorls; sepals relatively large; fruiting head up to 1 em. in diameter; achenes 2-2.5 mm. long, with 5-7 crests.—Brackish waters, Florida. The somewhat mystifying bibliography in the treatment in the North American Flora gives the type locality of Sagittaria lorata (Chapm.) Small as “Carolina”. Chapman, in originally publishing S. natans, var, lorata, upon which S. lorata rests, said “Brackish water, along the west coast of Florida”. On the same page in the North American Flora, in the synonymy of typical S. subulata, we get the following entry: “Sagittaria natans lorata A. Gray, Man. ed. 5. 494. 1867”. Search for the latter com- bination fails to reveal it; but the combination, S. subulata natans (Michx.) J. G. Smith, properly published by Smith in his monograph of the genus (1894), failed to win citation in the North American Flora. S. WearHersrana Fernald in RuHoporA, XxxVii. 387, pl. 385 and 386, fig. 1 (1935). In summer the broad phyllodia are mostly shriveled, though conspicuous in early spring. On October 12, we found the plants of pools in the Great Dismal Swamp with completely shriveled mature foliage but with newly developing phyllodia of extra- ordinary beauty, with the whole breadth or a wide central band filled by large lacunae, in this suggesting the foliage of Potamo- geton epihydrus. It is evident that S. Weatherbiana makes its Principal growth from autumn to spring, as do Hottonia inflata and some other aquatics. See pp. 359 and 400 VALLISNERIA AMERICANA Michx. To the station recorded -in 1936 add anot er, also in Princess ANNE County: abundant 71 Sandy bottom of Back Bay (depth slightly more than 1 m.), Long Island, no. 10,881. See p. 387. ea Ps ANACHARIS DENSA (Planch.) Victorin. To the station in Din- Widdie County reported in 1938 add one in JaMEs Ciry County: Pool in cypress swamp back of Chickahominy River, below Bar- rat's Bridge (or Ferry), no. 11,221. . “JIPLACHNE MARITIMA Bicknell. Princess ANNE CouNnTY: nner border of brackish to fresh marsh along Back Bay, at eastern margin of Long Island, no. 10,522; shallow pools in 410 Rhodora [OcTOBER brackish to fresh marsh along Back Bay, east of Munden, no. 10,833; similar habitat east of Creeds, no. 10,884. See pp. 370 and 387. GLYCERIA ARKANSANA Fern. To the stations in Greensville and Surry Counties recorded in 1938 add the following. Sussex County: wooded bottomland, Three Creek, southwest of Griz- zard, no. 10,128; bottomland swamp, Nottoway River, southwest Homeville, no. 10,129. Greensvinte County: wooded bottomland of Fontaine Creek, southeast of Taylor’s Millpond, no. 10,127. See p. 360. RAGROSTIS HYPNOIDES (Lam.) BSP. Local range extended eastward to the Blackwater valley. NanseMoND County: sandy border of rill in swampy woods east of Milk Landing, south of South Quay, no. 11,223. E. PEREGRINA Wiegand. Range extended southward to Priv- cess ANNE County: roadside, open muddy and sandy shore of Back Bay, east of Creeds, no. 10,891. SourHaMPTON COUNTY: railroad yards and waste places, Franklin, no. 10,886. E. uirsuta (Michx.) Nees. To the station in Arlington County (see RHopora, xli. 500) add one in DinwIDDIE County: large clumps in cinders of freight-yard of Norfolk and Western Railroad, Petersburg, no. 10,892. The typical southern plant with strongly hirsute sheaths. See p. 390. pahannock, no. 11,527, plants 10-15 feet high. See p. 402. Triopra CHAPMANI (Small) Bush. SourHampTon CoUNTY dry white sand of clearings in oak and pine woods bordering ee samoosick Swamp, south of Sebrell, no. 10,894; border of sanay woods near Three Creek, northwest of Carey Bridge, no. il See pp. 378, 383 and map 15. : : Extension inland from Nansemond County. Experience 1? the field, now extending over several years, indicates that < Chapmani is abundantly distinct from T. flava. Hitchcoe! treated it as inseparable, but in 1934, Griscom and I, finding 7 in the sandy woods back of Cape Henry, recognized it as 4 least varietally separable, T. flava, var. Chapmani (Small) ak nald & Griscom, in Ruopora, xxxvii. 133 (1935). Subsequen observations show that, whereas the wide-ranging T. flava (L.) Hitche. prefers relatively good soil and is inclined to follow pee sides, T. Chapmani is restricted to sandy pinelands and 08% scrub. It not only has the spikelets long-stalked; its very 1940] Fernald,—Additions to Flora of Virginia 411 narrow leaves are bluish-green, 5-7 mm. broad, and strongly in- clined to become inrolled. Its open and skeleton-like panicle has the bases of the principal branches surrounded by ring-like pulvini with very long hairs, whereas the denser panicle of T. flava has the shorter hairs confined to the upper sides of the axils of the branches. 7. Chapmani reaches its northernmost limit, apparently, in the sands of southeastern Virginia. *SporopoLus asperR (Michx.) Kunth. New Kent County: steep sandy bank of York River, near mouth of Fillbate’s Creek, north of Holly Forks, no. 11,525. See p. 400. Hitchcock (Manual) indicates no station in the Atlantic States south of northern Maryland. SPARTINA cyNosuROIDES (L.) Roth. James Crry County: fresh tidal marsh of Chickahominy River, below Barrat’s Bridge (or Ferry), no. 11,236. Noteworthy as an extension inland beyond the saline shores. See the two following. S. ALTERNiFLoRA Loisel., var. PrLosa (Merr.) Fern. Inland to Cuartes Crry County: sandy tidal margin of Chickahominy River, Ferry Point, no. 10,911. S. parens (Ait.) Muhl., var. suncea (Michx.) Hitche. In- land to Cuartes Crry County: sandy beach of James River southeast of Tettington, no. 11,235. See p. 396. In dry fresh sand, associated with plants of strictly fresh habitats. ERSIA LENTICULARIS Michx. Cartes Crry County: bot- tomland woods by Chickahominy River, north of Roxbury, no. 11,238. New Kent County: similar habitat, north of Long Bridge, southeast of Quinton, no. 11,239. Nansemonp County: border of gum swamp at margin of Great Dismal Swamp, south- east of Whitemarsh School, no. 11,521. See pp. 397 and In view of Hitchcock’s map, indicating no station in the At- lantic States north of South Carolina, and our records of it from Greensville and Southampton Counties (RHopoRA, xxxix. 348, 353, 382 and map 22 (1937)), the stations in New Kent and in Nansemond Counties constitute notable local extensions of range. SEX County: roadside ditch southeast of Wakefield, no. 10,918. It is evident that this species is rapidly spreading northward. 412 Rhodora [OcroErR PANICUM CAERULESCENS Hackel. Range extended inland to SourHAMPTON County: sphagnous swampy woods southwest of Applewhite Church, no. 10,104. ” CRYPTANTHUM Ashe. SourHampton County: bushy swales and borders of swampy woods near Blackwater River, Cobb’s Wharf, no. 10,931. See pp. 364 and 383. One of the rarest members of the genus, heretofore unknown between the region of Wilmington, North Carolina, and southern New Jersey. Near Cobb’s Wharf (just north of the North Carolina line, on the west bank of the Blackwater River) there are several very large clumps. One-half of such a plant, broken into fragments, made 30 full sheets of representative material. The station, unfortunately, is within a rod or two of a plowed field. P. caprnuaRE L. New Kent County: ditch at border of damp woods, near Fillbate’s Creek, north of Holly Forks, no. 11,519. The first time we have met this elsewhere usually common species on the Coastal Plain of Virginia. P. AGRostoiwes Spreng., var. conpensum (Nash) Fernald. Tidal shores of James River, forming dense clumps up to 1.8 m. tall, with wide-branching panicles up to 3 dm. long and 1-15 dm. broad (several nos.). pp. and 391. Just such plants from tidal shores all the way to Texas are referred in the Gray Herbarium by Hitchcock & Chase to r. agrostoides. I can get no satisfaction in separating them from P. condensum Nash. They have the longer spikelet of that plant, also the firmer foliage with broader midrib. Along the James these very tall clumps with panicles much exceeding the “rarely more than 5 em. wide” of Hitchcock’s Manual form thickets at high-tide limit. Farther out, they pass into an outer band (as at Wilcox Wharf, for instance) with plants down to 6 dm. high and with dense lance-ellipsoid panicles only 3-4 em. thick. These are P. condensum as defined by the above authors, but they are clearly small individuals of the coarser plants which form part of the series of specimens listed by Hitchcock & Chase as “intermediate between P. agrostoides and P. condenswm - Unfortunately some of the specimens cited by them as inter mediate do have the coarse habit of the tall plants of the Jame and short spikelets which I cannot distinguish from those of Ee agrostoides. If P. agrostoides and P. condensum are distinct 1940] Fernald,—Additions to Flora of Virginia 413 species, I shall weleome having their morphological distinctions pointed out. EcHINOCHLOA cotonuM (L.) Link. James Crry County: tidal ore of Back River, opposite Jamestown Island, no. 10,939. The only time we have met this species, although Hitchcock (Man.) gives Virginia as its northeastern limit. Wiegand in his study of The Genus Echinochloa in North America, RHopoRA, xxiii. 49-65 (1921), cited it from South Carolina southward and Westward. See p. 385. “E. puNGENS (Poir.) Rydb. (E. muricata Mieke) dees NaNsEMonD County: roadside bordering sandy sphagn margins of thickets in pineland southwest of March Hill School, south of South Quay, no. 11,242. Not seen from Virginia by Wiegand, l. c. “E. Watrertr (Pursh) Nutt., forma LarvicaTa Wiegand. (EZ. longearistata Nash). PRINCESS ANNE County: open muddy d sandy shore of Back Bay, east of Creeds, no. 10,940. SOUTHAMPTON County: swaley tidal shore ‘es Shaped River, Knight’s Seine Beach (Battle Beach), no. 10,938. See _ 383. Wiegand cites no material from Virginia; Nash gave ite range, “South Carolina to Louisiana.” *“SETARIA VIRIDIS (L.) Beauv., var. Wemnmanni (R. & S.) Beauv. Prince GrorGE County: ’ cinders of freight-yard of Nor- folk and Western Railroad, east of Petersburg, no. 11,240. a Hubbard, in his Taxonomic Study of Setaria italica and its immediate Allies, Am. Journ. Bot. ii. 169-198 (1915), cited no material from south of New England. “ANDROPOGON (§ ScHIZACHYRIUM) ie ematurus, (TAB. 626, FIG, 1-3), ab A. scopario recedit planta dense vice 3-9 dm. alta; racemis 1-3 em. longis 3—7-articulatis longe pedun- culatis pedunculis filiformibus valde exsertis; rhachi undulato internodiis 3-5 mm. longis superne barbatis; spiculis sessilibus 7 mm. longis; spiculis pedicellatis saepe bene evolutis solitariis plerumque masculis 3.5-7 mm. longis——Dry open woods, clear- Ings and fields, southeastern VirGINIA: dry gravelly soil north- west of Grove, xork County, June 21, 1922, L. F. & Fannie R. Randolph, no. 353, as A. scoparius (in ’anthesis) ; dry sandy pine woods about 3 miles southeast of Petersburg, on headwaters of 414 Rhodora [OcToBEeRr thesis) ; argillaceous field near Century House, northeast of Bur- gess, Dinwiddie County, September 13, 1937 (shattered), Fer- nald & Long, no. 7310; June 17, 1938 (in anthesis) Fernald & Long, nos. 8095 (foliage green) and 8096 (glaucous) ; border of dry woods near Assamoosick Swamp, about 2 miles northeast of Homeville, Sussex County, August 24, 1938, Fernald & Long, no. 8922 (fruit scattered) ; hickory and oak woods and clearings east of Skipper’s, Greensville County, June 12, 1939 (in anthe- sis), Fernald & Long, no. 10,092 (ryPE in Herb. Gray.). SouTH CaRoLINA: grass-sedge bog or savannah, 1 mile west of Chicora, dred 24, 1939 (shattered), R. K. Godfrey & R. M. Tryon, Jr., n0. Andropogon praematurus (see p. 357), when first found by us in June, 1936, at once challenged interest because of its extra- ordinary flowering season. Subsequently we have seen much of it, the plant consistently flowering from early June through the month (presumably into early July) but the inflorescences be- come quite shattered and the fruit scattered by August. A. scoparius and its varieties are autumn-flowering plants, in eastern Virginia very immature in September and in their prime through October. In A. scoparius the culms are mostly 0.5-1.5 m. high, the leaves 3-6 mm. wide, the racemes 3-7 cm. long, and the pedicels at each node are paired and truncate or terminated by a sterile rudiment. The early-summer A. praematurus is ated ally only 3-6 dm. high, though exceptional woodland colonies may reach a height of 9 dm.; its leaves are mostly 24 mm. broad; the racemes only 1-3 cm. long, with internodes only 3-5 mm. long. Most remarkable, the sessile perfect spikelets, 1 stead of being accompanied by a pair of truncate bearded pedi- cels or these terminated by mere rudiments, usually are & companied by a single well developed staminate pedicelled spike- let (rics. 2 and 3), though the terminal group may have 2 pedi- celled spikelets. In well developed plants these characters are conspicuous; in some, however, only a few well formed staminate spikelets develop. All specimens display some of them and no second pedicel (except the terminal group). Andropogon praematurus thus belongs in the group of species which retain well developed pedicelle nate spikelets, these species being presumable ancestral types which the pedicelled spikelets have not become reduced to meré remarkable d stami- 1940] Fernald,—Additions to Flora of Virginia 415 rudiments or to truncate bearded pedicels, but in the three Ameri- can species with this character the pedicelled spikelets are mostly solitary, instead of paired. The new species can not, however, be placed with either of the other North American species with the solitary pedicelled spikelets staminate. A. divergens Anderss." is a coarser endemic of Texas, 0.8-1.5 m. high, with scarcely or barely exserted racemes (our FIG. 4) 3-4 em. long, the internodes of the rachis 4.5-7 mm. long, the sessile spikelets and the pedicelled staminate ones 6-8 mm. long. The three sheets of it before me (Cory, nos. 25,510, 25,840 and 26,064) were all collected in October. A. praematurus is certainly very closely related to it, but it differs in its much lower stature, its long-exserted racemes with internodes only 3-5 mm. long, the spikelets mostly smaller. These differences, associated with the Pronounced difference in flowering season and the geographic isolation, sufficiently distinguish A. praematurus. tom Andropogon maritimus Chapm., of the coast of the Gulf of Mexico, A. praematurus is abundantly distinct. A. maritimus is a stout and stoloniferous non-cespitose species with conspicuously distichous divergent leaves, racemes (Fic. 5) partly included and 4-6 em. long, the spikelets 8-10 mm. long. Whether the break in the range of A. praematurus (South- eastern Virginia, reappearing in southeastern South Carolina) is an actual one may be doubted. It is probable that the species occurs also in southeastern North Carolina, a region sharing many of the peculiar plants of Southeastern Virginia. In PLate 626, Fic. 1 is the type of Andropogon praematurus, X ¥%4; FIG. 2, a raceme, x 1, from me TYPE; FIG. 3, upper half of a x 5, from TE; Fic. 4, raceme, X 1, of A. divergens Anderss., from near Alba, Texas, Cory, no. 25,510; ria. 5, raceme, x 1, of A. maritimus Shee from Horn Island, Mississippi, Tracy, no. 3786. _ A. vircinicus L,, var. — (Muhl.) Hack. (A. capillipes Nash), Nansemonp County: dry white sand of pine barrens east of Cox Landing, south of South Guay, no. 10, 43, See pp. “1 Andropogon a and A. divergens were treated by Hackel as A. scoparius Michx., subsp. aritimus, a. genuinus and f. divergens, Hackel, Androp. (in DC. Mon. s divergens Anderss., Published in syno: autho: or is An ‘ RHopora, xxxvii . : misinterpreted A. i The plant there called by us A. scoparius, var. divergens 'S var. polycladus Scrib. & Ball. 416 Rhodora [OcToBER First from north of southern North Carolina. A. VIRGINICUS, var. TETRASTACHYUS (Ell.) Hack. To the single recorded mii kes station (at Cape Henry; see Fernald & Griscom in RuHopora, xxxvii. 142) add two in NANSEMOND OUNTY: seeping bank of ditch at margin of woods, about 2 miles southeast of Cleopus, no. 9513; dry white sare of pine barrens, east of Cox Landing, south of South Quay, n ANDROPOGON VIRGINICUS L., Var. HIRSUTIOR (Hackel) Hitche., forma tenuispatheus (Nash), comb. nov. A. glomeratus tenut- 7 akeade Shag in Small, Fl. Se. U. S. 61 (1903). A. tenuispatheus (Nash) N nN. Am. FI. xvii. 113 (1912). A. virginicus var. one ah (Nash) Fernald & Griscom in RHoporaA, XxXVil. 142 (1935). Forma tenuispatheus seems to be only a glabrous form of the hirsute-sheathed var. hirsutior. In making the transfer of the name tenuispatheus as a variety to A. virginicus, with var. hir- sutior as a hirsute form, A. virginicus, var. tenuispatheus, forma hirsutior (Hackel) Fern. & Grisc. 1. c. (1935), Griscom and I overlooked the fact that as a varietal name A. macrourus, ¥- hirsutior Hackel (1889) has priority over A. glomeratus tenw- spatheus Nash (1903). (To be continued) A CENTURY OF ADDITIONS TO THE FLORA OF VIRGINIA M. L. FERNALD (Continued from page 416) *CYPERUS RIVULARIS Kunth, forma eLurus (C. B. Clarke) Kiikenth. James Crry County: tidal marsh along Powhatan Creek, north of Jamestown Island, no. 10,946; similar habitat, Gordon Creek, east of Barrat’s Bridge (or Ferry), no. 11,260. Apparently all the plants of tidal estuaries from southern Maine southward belong to forma elutus, with nearly colorless Scales. The type of var. elutus C. B. Clarke, was from Pennsyl- vania, without designation of locality. In the tidal marshes the form becomes very tall (up to 6 dm. high). See p. 394. C. HASPAN L., var. AMERICANUS Boeckeler. Tidal estuaries of the James and the Chickahominy and tributary creeks inland to Charles City and New Kent Counties, also tidal marshes of ack Bay (many nos.). See pp. 376 and 386 and map 9. American variety of a pantropical type, reaching essentially equivalent latitudes north and south of the Equator in North and South America. C. GLopuLosus Aublet, This adventive species, rapidly spreading as a weed, often forms dense carpets in dooryards, where, mowed at frequent in- tervals, it maintains a good, green carpet through the mid- summer heat. It thus becomes an uninvited “lawn grass” of some value. “C._ BREVIFOLIUS (Rottb.) Haussk. (Kyllinga brevifolia Rottb.). James Crry County: fresh tidal marsh of Chicka- 420 Rhodora [NoveMBER hominy River, below Barrat’s Bridge (or Ferry), no. 11,266. See p. 395 and Map 21. First known area north of Georgia, except that on the lower Delaware. Like Cyperus haspan (see above), C. brevifolius is a pantropical type, with dispersal which suggests great an- tiquity, although in some areas it is evidently a recent adventive; and, like C. haspan var. americanus, its eastern American limits north and south of the Equator are in equivalent latitudes. ELEOcHARIS TENUIS (Willd.) Schultes, var. verRUCOSA Svenson. To the station in Dinwiddie County, recorded in 1938, add one in GREENSVILLE County: peaty swale by Southern Railway, northeast of Emporia, no. 10,137. See p. 360. DIcHROMENA COLORATA (L.) Hitche. To the station recorded in Rwopora, xxxix. 396, add another, also in Princess ANNE County: inner border of brackish to fresh marsh along Back “ee eastern margin of Long Island, no. 10,539. See pp. 370 an PsILOCARYA SCIRPOIDES Torr., var. Grimes Fern. & Grise. 10 Ruopora, xxxvii. 154, pl. 344 (1935). To the two original sta- tions add a fine one in NaNseMonD County: seepy sandy and peaty open spots in sphagnous savannah-like swale east of Cherry Grove, south of South Quay, no. 11,530. See p. 384. ULBOSTYLIS CILIATIFOLIUS (Ell.) Fernald in Rxopora, xl. 391 (1938). Local range extended northward in Southampton County and eastward into Nansemond County (many nos.). See pp. 362 and 379. Scirpus DEBILIs Pursh. Dinwipre County: springy sphag- nous swale about 5 miles east of Burgess Station, no. 10,970. South of the Potomac chiefly a plant of the interior of the State. We have met it nowhere else on the Coastal Plain. See p. 390 S. arrovirens Muhl., var. ruaccipirotius Fern. in xl. 396 (1938). Range extended into GREENSVILLE CountTY: wooded bottomland of Fontaine Creek southeast of Taylors Millpond, nos. 10,140, 10,544. Sussex County: botto mln Mb Pherae River, southwest of Homeville, no. 10,14!- p. ? RHODORA, ~ By mid-July the over-lopping and drowned inflorescences freely proliferate in the manner of the two following. into NAN- SEMOND County: border of gum swamp at margin of Great ee mal Swamp, southeast of Whitemarsh School, no. 11,532, ° 1940] Fernald,—Additions to Flora of Virginia 421 S. FONTINALIS Harper, var. VIRGINIANA Fern. in Ruopora, xi. 532 (1939). Range extended northward across the James to JaMEs Ciry County: swampy woods along Powhatan Creek, north of Jamestown Island, no. 10,974, old culms arching and trailing, the inflorescences producing many rooting young plants. 386. EriopHoruM vircinicum L. Dinwippre County: springy sphagnous swale about 5 miles east of Burgess Station, no. 10,975. Princess ANNE County: swampy and inundated woods, north of Blackwater River, no. 3779. It is not known where Clayton got his material, cited by Gronovius and by Linnaeus, nor the exact geographic source of the plant said to be from Virginia, described and illustrated by Plukenet. The species is frequent along the Blue Ridge and the Appalachian Upland but we have only twice met it on the Coastal Plain, although Britton & Hollick collected it in late September, 1890, in the Great Dismal Swamp in Nansemond County, along “the old canal” from Suffolk, where “masses of Eriophorum Virginicum were frequent.”—Hollick in Mem. Torr. Bot. Cl. ii. 56 (1890). See p. 390. *FUIRENA BREVISETA Coville. Dinwippie County: springy sphagnous swale about 5 miles east of Burgess Station, no. 10,976. Extension north from North Carolina. See p. 390. “RHYNCHOSPORA MACROSTACHYA Torr., var. colpophila Fer- nald & Gale, var. nov., planta 0.8-1.75 m. alta; foliis valde elongatis subflaccidis; achaeniis 3-3.8 mm. latis 5—5.8 mm. longis; 2. tidal marsh along Gordon Creek, east of Barrat’s Bridge (or erry), September 19, 1939, Fernald & Long, no. 11,275; tidal 422 Rhodora [ NovEMBER marsh along Powhatan Creek, north of Jamestown Island, August 22, 1939, Fernald & Long, no. 10,977. Prince George County: muddy tidal shore of James River, Jordan Point, August 16, 1938, Fernald & Long, no. 8983 (ryPE in Herb. Gray). See pp. 386, 393 and 401. As noted by me in Ruopora, xli. 533 (1939), Rhynchospora macrostachya is known on the Coastal Plain of Virginia only from the acid Great Dismal Swamp and from the tidal estuaries entering Chesapeake Bay. The plant of the tidal shores 1s extraordinarily large, with prolonged and almost flaccid leaves. Miss Shirley Gale, engaged in an intensive study of the genus, has made measurements which give the following contrasts. Typical RuyNcHospora MACROSTACHYA: plant 4.5 (in dwart speci- mens down to 1)—11.3 dm. high, in the Southeast becoming taller; leaves firm, 0.35-1.2 em. broad; achene 5—5.4 mm. long, 63.1 mm. broad; tubercle 1.7—2 (rarely —2.2) em. long, 1-1.8 mm. broad at base. Var. coLpopHILA: plant 0.8-1.75 m. high; leaves very long, sub- flaccid, 0.9-1.5 em. broad; achene 5—5.8 mm. long, 3-3.8 mm. broad; tubercle 1.8—2.3 cm. long, 1.8—2.4 mm. broad at base. Var. colpophila is a member of the remarkable flora which characterizes the tidal reaches of the Chesapeake, a group ° endemic and isolated species specially discussed in Part IIL. Tue Varieties or RHYNCHOSPORA CEPHALANTHA.—In 1935 I pointed out'? that the type of Schoenus axillaris Lam. (1791), which is the nomenclatural basis of Rhynchospora axillaris (Lam.) Britton (1888), can have nothing to do with the plant with subspherical glomerules to which Britton had applied it, I then showed that the first clearly identifiable binomial in this series is R. cephalantha Gray in Ann. Lyc. N. Y. iii. 218 (1835); and I took up R. cephalantha in the inclusive sense of Gray and recognized R. microcephala Britton as a species on account of more numerous and more remote and smaller glomerules of smaller spikelets with smaller achenes with more slender tubercle. During the summer and autumn of 1938 and again in 1939 Mr. Long and I were much perplexed by the series, for it, we noted that in bogs where deep sphagnum prevails plants very close to typical New Jersey R. cephalantha occur, but that inundated pinelands and shallow pools there is 4 much coarse? plant, with achenes and spikelets of R. cephalantha, the glomer- 12 Fernald in Ruopora, xxxvii. 403-405, plate 391 (1935). 1940] Fernald,—Additions to Flora of Virginia 423 ules tending to be very numerous and remote and the broad leaves remaining flat, whereas typical R. cephalantha has the narrow leaves soon becoming involute and the few glomerules less scattered. Miss Shirley Gale has joined me in a detailed study of the series and we are recognizing two well defined varie- ties of R. cephalantha. See p. 366. *R. CEPHALANTHA Gray, var. typica Fernald & Gale. RB. ce- phalantha Gray in Ann. Lye. N. Y. iii. 218, as to plant of Torrey from New Jersey and Long Island (1835); Torrey, Fl. N. Y. ii. 365 (1843) ; Gray, Man. 533 (1848) ; Fernald in Ruopora, xxxvii. 404, pl. 391, figs. 2 and 3 (1935). R. azillaris Britton in Bull. Torr. Bot. Cl. xv. 104 (1888) as to plant, in part, not as to name- bringing synonym, Schoenus axillaris Lam.—Culms 5.7-9 dm. high; leaves 1.5-2.5 (-3) mm. broad, tending to become strongly Inrolled in drying; glomerules chiefly confined to the summit or the upper fifth of the culm, terminal or in 1 or 2 (rarely 3) upper axils, the lowest then 2-10 (-15) em. apart; mature glomerules 5-2 cm. in diameter; spikelets castaneous to blackish-fuscous, 5-6 mm. long; achenes 2-2.4 mm. long, 1.4-1.6 mm. broad; tu- bercle 1.8-2.4 mm. long; bristles 3.8-4.8 mm. long.—Wet pine barrens and bogs, Long Island, New Jersey and Delaware; south- eastern Virginia, eastern North Carolina and southeastern South Carolina. The following are the more southern specimens. Vir- SINIA: argillaceous and siliceous boggy depression southeast of Petersburg, at head of Poo Run, Prince George County, June 19, 1936, F ernald, Long & Smart, no. 5661, with scattered glomer- ules (as R. microcephala), July 18, 1936, Fernald & Long, no. 6090 (distributed as R. microcephala) ; sphagnous argillaceous bogey depression just northwest of Wakefield, Sussex County. September 11, 1937, Fernald & Long, no. 7352 (as R. micro- cephala); sphagnous bog about 1 mile northeast of Dahlia, Greensville County, July 15, 1938, Fernald & Long, no. 8610. August 20, 1938, Fernald & Long, no. 8993. NortH CaRouINa: sphagnous bog at Method, Wake County, July 13, 1938, R. K. odfrey, no. 4985 (transition to var. pleiocephala in its distant glomerules); drainage ditch at Carolina Beach, New Hanover County, June 24, 1938, Godfrey, no. 4719 (broad leaves of var. pleiocephala) ; savannah 12 miles north of Jacksonville, Onslow County, August 6, 1938, Godfrey, no. 5756: savannah 5 miles east of Jacksonville, Onslow County, August 6, 1938, Godfrey, _ 8. SourH Caronina: grass-sedge bog or savannah, 12 miles north of Georgetown, Georgetown County, August 2, 1939, iebey & Tryon, nos. 752a, 1061; drainage ditch, 3 miles north T McClellanville, Charleston County, July 19, 1939, Godfrey & ryon, no. 677. 424 Rhodora [ NovEMBER Much of the material from southeastern Virginia has 3 or 4 remote glomerules, whereas 2 glomerules are more general in New Jersey. Plants with only 2 glomerules occur, however, in Virginia and exceptional New Jersey specimens show 1, 3 or 4. *Var. pleiocephala Fernald & Gale, var. nov., culmis crassis 0.6-1.2 m. altis; foliis planis 2.5-4.5 mm. latis; inflorescentiis 1.4-5 dm. longis, glomerulis axillaribus 4-7 remotis 1.2-2 cm. diametro, imis 0.6—1.8 dm. distantibus; spiculis fulvis vel castaneis 5 mm. longis; achaeniis 2.1-2.5 mm. longis 1.4-1.6 mm. latis; tuberculis 1.4-2.2 mm. longis; setis 3-4.4 mm. longis——Swamps, pond-holes, wet pinelands and ditches, southeastern Virginia to Florida and Louisiana. Vircrnia: abundant and dominating an exsiccated argillaceous pond-hole in woods, about 1 mile south of Mercy Seat Church, Surry County, August 23, 1938, Fernald & Long, no. 8994 (rypr in Herb. Gray), October 15, 1938, Fer- nald & Long, no. 9549: dominant in flat sphagnous pineland, Collier’s Yard, 3-4 miles southwest of Petersburg, Dinwiddie County, July 16, 1939, Fernald & Long, no. 10,548; same station, August 17, 1939, Smith & Hodgdon in Pl. Exsic. Gray.; pond- hole in pine and oak woods near Three Creek, north of Emporia, Greensville County, September 9, 1938, Fernald & Long, DO. 9283. Norru Caroirna: pineland at Nakina, Columbus County, August 29, 1938, R. K. Godfrey, no. 6347; low pineland at Dunn, Harnett County, August 25, 1938, Godfrey, no. 6122; without stated locality (presumably near Wilmington), M. A. Curtis; ditches near Wilmington, July 2, 1897, Biltmore Herb., no. 279%. SourH Carona: Santee Canal, August, ——, Ravenel (as & cymosa) ; sandy drainage ditch, 2 miles west of Salters, Williams- burg County, July 14, 1939, Godfrey & Tryon, no. ; Swamp depression in savannah, 15 miles north of Georgetown, Georgetown County, June 24, 1939, Godfrey & Tryon, nO. 62; grass-sedge bog or savannah, 12 miles north of Georgetown, Au gust 2, 1939, Godfrey & Tryon, no. 752; shallow peaty pond in pine barren, 9 miles north of Georgetown, July 21, 1939, God- frey & Tryon, no. 759; boggy ditch in pine barrens, 2 miles east of Meggett, Charleston County, July 16, 1927, Wiegand & M eo ning, no. 582 (as R. azillaris) ; drainage ditch, 3 miles north 0 McClellanville, Charleston County, July 19, 1939, Godfrey & Tryon, no. 675. Gxorcia: Waycross, August 18, 1909, 4: m Lang (Phil. Acad.) ; wet meadow, Leslie, Sumter County, Augus 17, 1900, Harper, no. 413 (as R. azillaris). Fiona: low pepe places, near Jacksonville, August ——, Curtiss, er oo swampy places in pine barrens near Jacksonville, June 18, Curtiss, no. 5016 (as R. axillaris); cypress swamp, Viclm R Eustis, Lake County, May 16-31, 1894, Nash, no. 845 (as 1940] Fernald,—Additions to Flora of Virginia 425 axillaris) ; swamp, Okeechobee region, Brevard County, May 14, 1903, Fredholm, no. 5821 (as R. fascicularis) ; swamp on prairie, Okeechobee region, October 29, 1903, Fredholm, no. 6175 (as R. axillaris); bogs and shady swamps, Apalachicola, July— August, ——, Chapman in Biltmore Herb. no. 279». Mussis- SIPPI: tidal marsh on Bilox Bayou, Harrison County, September Springs, St. Tammany Parish, August 12, 1912, Pennell, no. 4137 (Phil. Acad.). A single collection from New Jersey (Quaker Bridge, October 12, 1859, Wm. Boott) is as coarse as R. cephalantha, var. pleio- cephala but its few glomerules place it nearer var. typica. God- frey, no. 4985, from Wake County, and his no. 4719 from New anover County, North Carolina (see var. typica) are also transitional. The coarse southeastern plant which we are calling Rhyncho- spora cephalantha, var. pleiocephala was recognized by Gray When he originally described R. cephalantha; and he gave the latter name rather than the original manuscript R. biceps of Torrey because of the southern specimens (from Wilmington, North Carolina, to Florida and Louisiana). The original ma- terial which Torrey had designated R. biceps and which formed the primary basis of R. cephalantha was the plant of the Pine Barrens of New Jersey (and of Babylon, Long Island); the coarser several-headed plant of the South was appended to it by Gray and caused his rejection of Torrey’s name. Gray’s comment was as follows: nwo of Prof. Torrey, one which is in a good degree characteristic of € species. Later, however, in 1848, Gray (Man. 533) restricted the range of R. cephalantha to “Long Island and New Jersey.” We are, therefore, maintaining this more slender extreme of the species 48 var. typica. 426 Rhodora [ NovEMBER *Ruyncuospora chalarocephala Fernald & Gale, sp. nov. (FIG. 1 et 2), planta cespitosa, culmis gracilibus 1.8-8.3 dm. altis; foliis lineari-involutis 1— rare 2 mm. latis imis confertis; glom- erulis 3-7 remotis turbinatis vel laxe subhemisphaericis plerum- que 2—5-lobatis 0.9-1.8 cm. diametro; spiculis laxe adscendentibus vel explanatis; floribus solitariis; achaenio lenticulari-obovoideo laevi lucido 1.4-1.7 mm. longo 0.9-1 mm. lato marginato basi angustissime attenuato, tuberculo angusto subulato basi quam achaenii apicem valde angus- tiore 1-1.6 mm. longo; seti retrorse barbellatis tuberculo aequalibus vel id leviter su- perantibus 2.8-3.8 mm. long- is—New Jersey to South Carolina. New Jersey: wet (Phil. Acad.) ; Pleasant Mills, Burlington County, Septem- ber 24, 1887, C. A. Gross (Phil. Acad.); peaty, sandy RYNCHOSPORA CHALAROCEPHALA: FIG. bank along West Branch of 1, achene, < 10; Fic. 2, summit of in- Wading R., Harrisville, Bur- deg epee aha a eek lington County, July 21, 1932, 10; Fic. 4, summit of inflorescence, Fogg, no. 4545 (Phil. Acad.); A. peaty bog,near Beaver Branc of Wading R., southwest of Calico, Burlington County, July 28, 1937, Long, no. 50,935 (Phil. Acad.) ; Parkdale, Camden County, August 17, 1905, S. Brown, no. 56 (Phil. Acad.), September 1, 1911, Pennell, no. 3570 (Phil. Acad.) ; peaty and sphagnous pond-hole depression near Hard- ingville, Gloucester County, August 8, 1935, Long, no. 47,134 (Phil. Acad.); Weymouth, Atlantic County, August 19, 1883, C. A. Gross (Phil. Acad.) ; Atlantic Cranberry Meadows, Wey- mouth, August 26, 1922, G. W. Bassett (Phil. Acad.) ; moist pine barrens, Egg Hatbor City, Atlantic County, August 10, 1913, Mackenzie, no. 5558 (herb. N. Y, Bot. Gard.) ; Maurice River flats, east of Vineland, Cumberland County, August 12, 1928; Bassett & Long (Phil. Acad.) ; wet peaty margin of East Creek Pond, East Creek, Cape May County, September 25, 1920, on! no. 23,461 (Phil. Acad.) ; wet peaty pond-hole about 1 mile wes 1940] Fernald,—Additions to Flora of Virginia 427 of Bennett, Cape May County, September 26, 1920, Long, no. 23,488 (Phil. Acad.); moist sandy and peaty margin of pond- hole northwest of Ocean View, Cape May County, Long, no. 25,007 (Phil. Acad.); swamp 1 mile west of Dennisville, Cape May County, August 16, 1925, R. R. Driesbach, no. 3869 (Phil. Acad.) ; swamp, “Head of River”, Tuckahoe River, Cape May County, August 23, 1925, Driesbach, no. 3982 (Phil. Acad.). DELAWARE: moist soil, Lewes, August 15, 1895, Commons (Phil. Acad.). Vireinta: without definite locality, John Ball (as R. glomerata, var.); sandy and peaty border of Cat Pond, south of Benns Church, Isle of Wight County, September 18, 1937, Fernald & Long, no. 7357 (distrib. as R. microcephala), TYPE In Herb. Gray. Norru Caroutna: sandy soil, Beaufort (Car- taret County), August, 1903, Ruth, no. 9 (with memorandum, “Seems to be R. glomerata but I am not sure of it; the leaves are too setaceous for that”); drainage ditch near Hubert, Cartaret County, August 6, 1938, Godfrey, no. 5844 (as R. glomerata, exceptionally large plant with several glomerules borne in branching axillary cymes); pineland at Atlantic, Cartaret County, September 1, 1938, Godfrey, no. 6427 (as R. glomerata, var. minor); savannah 8 miles southwest of Jacksonville, On- slow County, September 1, 1938, Godfrey, no. 6469 (as R. gtomerata, var. minor) ; savannah 12 miles north of J acksonville, August 6, 1938, Godfrey, no. 5753 (as R. glomerata, var. minor) ; Pender County, September, 1880, Hyams, no. 4979 (herb. N. Y. Bot. Gard.) ; dried-out road-making sand-pit, 4 miles east of Bolton, Columbus County, July 5, 1927, Wiegand & Manning, no. 581 (as R. axillaris) ; Southport, Brunswick County, August 13, 1930, Blomquist (as R. microcephala). SourH Caro.iNa: g, 3 miles east of Georgetown, Georgetown County, July 21, 1939, Godfrey & Tryon, no. 774; grass-sedge bog or Savannah, 12 miles north of Georgetown, August 2, 1939, Godfrey : Tryon, no. 1060; sandy drainage ditch, west of Salters, Wil- lamsburg County, July 14, 1939, Godfrey & Tryon, no. 514; marshy border of lake, 8 miles southeast of Columbia, Lexington County, August 8, 1939, Godfrey & Tryon, no. 1339; damp soil, iken, July 28, 1866, H. W. Ravenel (as R. cephalantha?). LORIDA : Swamps, Sanford, November 3, 1927, S. Rapp (herb. Y. Bot. Gard.). Rhynchospora chalarocephala (from the loose heads) belongs ies Eurhynchospora, series Glomeratae, having the character- istic “wire-edged” smooth and lustrous umbonate obovoid achene 4wn out at base into a pronounced slender stipe. As shown by the Original identifications, it has been (often with doubt) Placed in R, glomerata, R. capitellata (R. glomerata, var. minor), 428 Rhodora [ NoveMBER R. “axillaris”, R. cephalantha and R. microcephala. Its closest affinity is with R. microcephala, since it has the spikelets 1- flowered, R. glomerata and R. capitellata having two or more florets to each spikelet, or, if with only one true floret, then with one or more terminal rudiments. Its tightly inrolled scales and small achenes also place it with R. microcephala. From that species (Fics. 3 and 4) it is distinguished in gross aspect by the character of the inflorescence. This is composed of 3-7 remote glomerules which are turbinate or subhemispherical at full de- velopment. Never tightly compacted, they are often subdivided into a cluster of as many as 5 smaller secondary and mostly approximate fascicles, thus suggesting the inflorescences of RK. glomerata and R. capitellata. As may be deduced from the shape of the fascicles, the spikelets are mainly ascending to loosely spreading, more rarely divergent. R. microcephala, on the other hand, as shown by the type, kindly placed at our disposal by Dr. A. C. Smith and Mr. Witt- rock, and as illustrated in Ruopora, xxxvii. t. 391, figs. 1 and 5 (1935), has the inflorescence normally composed of 4-6 remote globose heads, the spikelets of which are closely compacted, ascending to reflexed, thus presenting a solid aspect. Specimens of R. microcephala collected in shaded or unfavorable habitats may simulate R. chalarocephala, but in no case do they have the lateral glomerules lobed or forked. The basal leaves also afford a diagnostic character of secondary importance. Those of R. microcephala are commonly 1.5-3 mm. broad, and flat. In B. chalarocephala, however, they are rarely 2 mm. wide, being usually narrower, even to subfiliform, and usually involute 1 age. In borderline cases, where the gross appearance of the two species is similar, the identity can be determined by study of the mature achenes. Those of R. chalarocephala (FIG. 1), although of the same length and breadth as in R. microcephala (ric. 3) usually have the longer and narrower stipe passing more abruptly to the main body of the achene, which is, therefore, proportionally shorter. The ranges of R. microcephala and R. chalarocephala overlap; but their relative distribution varies. The former species, T@9& ing from New Jersey southward along the Coastal Plain, 1s local in New Jersey but attains a maximum development in south- 1940] Fernald,—Additions to Flora of Virginia 429 eastern Virginia and eastern North Carolina, dwindling, though locally present, in South Carolina, Georgia, Florida, Alabama and Mississippi. R. chalarocephala is relatively common at its northern limit in New Jersey and, although we have seen only two collections (one without stated locality) from Virginia, it is apparently common in eastern North and South Carolina. We have found only one collection from Florida. "R. pauuipa M. A. Curtis. Nansemonp County: sphagnous Savannah-like swale east of Cherry Grove, south of South Quay, no. 10,550, forming large stools 1.1 m. high. The first station between southeastern North Carolina and southern Delaware. See p. 381 and map 14. R. PerrLexa Britton. Local range extended to Dinwivp1e County: flat pineland, Collier’s Yard, 3-4 miles southwest of Petersburg, no. 10,981. Surry County: exsiccated argillaceous t of Cox Landing, south of South Quay, nos. 10,551-10,554, rang- _& In height from 2 dm. to 1 m.; similar habitat, southeast of a oye Landing, south of South Quay, no. 10,980. See p. 379. 0,154. UTHAMPTON County: wooded bottomland of Meherrin River, South of Hugo, no. 10,155. Both stations within a mile of North Carolina! p C. Franxn Kunth. Local range extended to the coast. TaN CESs ANNE County: in low woods along Back Bay, Long sland, no. 10,557. See p. 371. f Hi. 8 TUMESCENS Rudge X LoursIanica Bailey. Plants clearly ° this origin, cespitose like C. intumescens but with elongate 430 Rhodora [NoveMBER pistillate spikes pean of those of C. louisianica, local in Sussex County: wooded bottomland, Jones Hole Swamp, west of Coddyshore, any 18, 1939, no. 10, 174, ARISAEMA. In view of the revision in Ruopora, xlii. 247-254, plates 598-600, the following are recognized on the Coastal Plain of Virginia. A. TRIPHYLLUM (L.) Schott. (A. pusillum, forma pallidum E. H. Eames). Princess ANNE County: rich woods, Grea Neck, Fernald & Griscom, no. 4351; Little Neck, no. 3822. RINCE GEORGE CoUNTY: rich deciduous woods, Coggins Point, no. 9703. Henrico County: dark shady places, University of Richmond, Westhampton, May, 1933, Harriet M. Walton. SourHampron County: sphagnous pocket in rich deciduous woods northeast of Statesville, no. 8176; rich wooded slopes and spring-heads along Nottoway River above Care ey Bridge, no. 11,799. Type a Clayton specimen from Virginia. Bridge, n A. ATRORUBENS . (Ait.) B lume. Fruiting pian’, not showing the fresh spathes, but referable to the species, com We oar seen no typical A. atrorubens, with letibed full-purple ] A. ATRORUBENS, forma ZEBRINUM (Sims) Fern. Prince GEORGE County: rich wooded ravine southeast of Hopewell, no. 9702; sandy alluvial woods, bottomland of Powell Creek, Garysville, no. 7785. *A. ATRORUBENS, forma VIRIDE (Engler) Fern. PRINCESS ANNE County: damp woods, Virginia Beach, Fernald & Griscom, N0- 4349. Prince Grorce County: rich h deciduous woods, Co oe" Point, no. 9704; sandy alluvial woods, bottomland of Powel Creek, Ga no. 7784. Sussex County: wooded bottom- lands and swampy woods near Nottoway River, east of Stony Creek, no. 8175. ARE TWO SPECIES PASSING AS PELTANDRA VIRGINICA? On Pp. 360 I emphasized the marked differences in the freshly flower- ing material of the more northern and more southern plants which pass as Peltandra virginica. The two extremes are striking in fresh condition but, as in most aroids, the herbarium specimens are almost hopeless to make out. I am, therefore, merely 1940] Fernald—Additions to Flora of Virginia 431 showing photographs of characteristic fresh and passing spathes and a flowering spadix of each, x 1. PLATE 627 is of the southern plant, common in southeastern Virginia, thence to Georgia. It is characterized by the loosely opening white-margined spathe (Fic. 1) which, at the base of the limb soon becomes deliquescent (Fic. 3), the limb coming squarely off by circumscission. The summit of the orange- yellow spadix (rics. 2 and 4) is sterile or only weakly floriferous. The photographs were made from fresh material collected by the Appomattox River at Petersburg, Virginia, in June, 1940. PiaTE 628 is of the northern plant, occurring at least from southern Maine to New Jersey and Pennsylvania. Its greener, though pale-margined spathe (rics. 1 and 2) barely opens or is tightly rolled. The limb decays away, leaving a stub at the summit of the fruit (ric. 3); and the whitish spadix (ric. 4) flowers to the tip. The photographs are from fresh material collected in June, 1939, by Dr. Lyman B. Smith in Aberjona River, Winchester, Massachusetts. As stated, I am unable from herbarium material to reach a decision as to the importance of these characters. I merely present the problem, with the hope that many others will care- fully watch and most carefully collect the plants from New England southward and westward, recording their observations on the degree of opening of spathe, circumscission or gradual toting off of the limb, color of spadix and whether it flowers to the tip. If someone will volunteer to collate the material and teach a decision I shall be greatly relieved. When it is clearly settled whether we have two species or two well defined geographic varieties, the names must be care- fully weighed. The Linnean type is now underground and not available and those of Rafinesque are, if preserved, presumably also in European herbaria. Whether they would show more than most existing material in modern herbaria is questionable; but, until they are all traced and carefully considered and it is determined whether the contrasts evident in Virginia and from *nnsylvania to Maine are constant, it is, I feel, premature to make a decision which, at best, would be only tentative and too liable to upset. In 1890 Sereno Watson,!* stating that 13 Watson, Proce. Amer. Acad. xxv. 187 (1890). 432 Rhodora [NoveEMBER “Extreme forms received from Mr. A. Commons of Wilmington, Del., seemed to indicate that two species might perhaps be dis- tinguished”, went extensively into the matter. He studied the flowers, staminodia, ovaries and their number, and other tech- nical characters from variable fresh material. Besides the Commons material (with both open and tightly rolled spathes) and the Cambridge material, Watson had several lots of speci- mens (now showing little) sent by J. N. Rose from near Wash- ington. His conclusion was that “a study of the forms growing near Cambridge shows that no division can be safely made’. With relatively little experience in the group and a profound ignorance of the more technical characters, I hesitate, as stated, to make a hasty decision. As to the names, the following must be carefully considered: Arum virginicum L. and at least 9 proposed for species by Rafinesque; it must also be determined to which species o variety the formal names proposed by Blake apply. The final solution, I feel, should be deferred until the applications of the 10 specific names and the several formal ones can be satisfacto- rily determined. EriocavLton Parxert Robinson. To Grimes’s station at a exa, add the following. Kina Witu1am County: fresh a shore of Mattaponi River, at Horse Landing, near King bi boca Courthouse, no. 11,539. New Kent County: fresh tidal ae by Chickahominy River, at “Shady Rest”, southeast of Win cad Shades (Boulevard Postoffice), no. 11,538. CHARLES 10,988. Nansemonp County: muddy tidal margin of B “ water River, Cox Landing, south of South Quay, no. 11,283. pp. 386, 391, 399 and 401 and map 17. The last station is near the head of tide on a river rine to the Chowan in North Carolina. The plant should confidently be sought in the latter state. X- LACHNOCAULON ANCEPS (Walt.) Morong. Local phen tended eastward into NANseMonpD County: sphagnous 8aV 10,363. like swale east of Cherry Grove, south of South Quay, no. * See p. 381. to Xyris ampicua Beyrich. Local range extended ens or NANSEMOND County: sphagnous savannah-like swale 1940] Fernald,—Additions to Flora of Virginia 433 Cherry Grove, south of South Quay, no. 10,567, gigantic stools up to 3 dm. in diameter at base. Long (Phil. Acad.); shore of Red Mill Pond, near Lewes, Sus- 21, 1907, S. Brown (Phil. Acad.). Vircinia: fresh tidal shore of Mattaponi River, at Horse Landing, near King William Court- b erry), James City County, September 19, 1939, Fernald & Long, no. 11,284; muddy tidal margin of Blackwater River, Milk panding, south of South Quay, ee September , 193 mond County, September 22, 1939, Fernald & Long, no. 11,286. See pp. 396 and 399. Forma flaccida, in its very thin, often translucent, and mostly broad leaves is readily confused with X. difformis Chapm. It has, however, the small heads and flowers of X. caroliniana and the relatively few nerves in the leaves. X. difformis is usually ~arser, with the very thin and translucent leaves (emersed) 4mm. broad and 20-40-nerved, often falcate, half as long as to longer than the scapes, the latter when fresh conspicuously Winged; the mature heads are 1-2 em. long, and the seeds about 25-ribbed. It is locally abundant in the extreme South, coming horth to eastern Maryland, with a station near Laurel, Delaware pr swamps, Pepper’s Mill, A. Commons in Herb. Phil. cad.). Typical x. caroliniana, occurring from Florida to Louisiana, 434 Rhodora [ NoveMBER north to Nova Scotia, southeastern and central Maine, north- central New Hampshire, southeastern Vermont, eastern New York, thence west to Parry Sound, Ontario, Michigan and north- ern Indiana, is smaller, its opaque or subopaque leaves 0.5-6 mm. wide, mostly 1%—% as long as the mature scapes, and only 3-18-nerved (nerves vaguely seen by transmitted light). The scapes are less winged than in X. difformis, the mature spike 0.5-1.4 em. long, the seeds about 13-ribbed. In its thinner and mostly broader (5-8 mm. wide) elongate leaves (14 as long to nearly as long as scapes) X. caroliniana forma flaccida might be mistaken for X. difformis, especially when the leaves are translucent or not too encrusted with silt, but they have only 11-19 (rarely —23) nerves as in X. carolini- ana, mature heads only 0.8-1.3 cm. long, and the seeds of A: caroliniana. X. FLExuosa Muhl. (X. sree fey Local range eX- tended into Nansemonp County: wet sandy and peaty depres- sions in Lager: southwest of Marsh Hill School, south of Sout th Quay 10,992; very abundant in sandy and peaty pine bar- rens, pon of Cox Lenat, south of South Quay, no. 10,5 this area covering many square miles. IsLE or WIGHT County: esagt and peaty pine ana. south of Lee’s Mill, no. 12, 283. ee oC aac ELINA DIFFUSA Burm. f. (C. nudiflora of Am. auth., not L.). Kine Wrira M County: fresh tidal shore of Mattapon! HARLES City County: sandy tidal shore of James “Four Oaks”, below Harrison Point, no. 11,298. Princess ANNE County: low woods and clearings along Back Bay, Long no. 10,995. Norrotk County: Dismal Swamp, Wailaceton, July 17, 1891, A. B. Seymour, no. 39. SouTHAMPTON CounTY: wooded alluvial bottomland of Meherrin River, near Haley's Bridge, no. 9290. GREENSVILLE County: weed in fencerow, Em- poria, nos. 9291 and 10,996. See pp. 388 and 301. In his enumeration of stations for C. diffusa (as C. longicaulis) Pennell, Bull. Torr. Bot. Cl. xliii, 100 (1916), cited no specimens from north of Georgia except on ballast at Philadelphia. Al- though sometimes a weed in southeastern Virginia, C. diffus¢ is ; constituent element of the flora of bottomlands and in tida marshes. It has this dual behavior in many regions. If an adventive, it has frequently sought out very natural habitats. 1940] Fernald,—Additions to Flora of Virginia 435 Its pantropical distribution indicates its plasticity at least in recent times. THE VARIETIES OF COMMELINA ERECTA (PLATES 629-631).— One of the most bafflingly intricate series in Commelina is the group of plants which broadly constitute C. erecta L. In south- eastern Virginia we get four variations of it, which by the treat- ment in Small’s Flora of the Southeastern United States could be sorted into seven so-called species. Varying from plants 1 dm. high, with linear leaves at most 3 mm. broad and with fruiting spathes barely 1 cm. long to colonies 1.2 m. high, with lance-ovate leaves 4 cm. broad and spathes 3.6 cm. long, rang- ing from southeastern New York to Nebraska, thence south to the West Indies, Gulf States and Mexico, sometimes in the driest of sands, again in rich alluvium, these plants have naturally offered a fertile field for the “splitter”. But in his treatment of the group in 1916 Pennell’ sounded the call for retreat from further specific segregation; for, although he then maintained three species in the series (C. erecta L., C. angustifolia Michx. and C. crispa Wooton), he hesitated about recognizing more than one, C’. erecta, saying (p. 105): “Commelina angustifolia Michx. and C. crispa Wooton, until more fully compared living, are con- tinued as of specific rank. Both are probably better considered 48 geographic varieties of C. erecta.” Pennell forthwith reduced to C. erecta the recently proposed C. saxicola Small, and to C. angustifolia both C. Swingleana Nash and C. Nashii Small. With these reductions I am heartily in accord. But, even the "ecognition of three varieties, typical C. erecta and vars. angusti- folia (Michx.) and crispa (Wooton) Palmer & Steyermark, does hot satisfactorily settle the problems in the species. In order to see how in the most conservative treatment we have, Pennell’s, the three elements are defined we may quote his key: Bei lanceolate. Posterior petals usually larger, 12-25 L ong. Seeds mostly more or less oblong .........+-- 6. C. erecta O15 ager lanceolate. Posterior petals usually smaller, 436 Rhodora [NovemBre In 1938 Pennell wrote “I would now consider that this species includes as narrow-leaved subspecies the southern C. angustifolia Michx. and the southwestern C. crispa Wooton.”!> With this decision I entirely agree, except that the free intergradation of the three compel me to treat them as two geographic varieties and a minor form, rather than to dignify them as subspecies. There is practically no doubt about the identity of Commelina erecta L., although there has long been needless confusion of it with C. virginica L. (the coarse species with extensively creep- ing rhizome and stolons, the sheath of the leaf with erect summit fringed with long erect reddish bristles, whereas C. erecta has & fascicle of fleshy root-fibers and the sheath has a spreading flange-like summit short-ciliate with white hairs). It was long cultivated in Europe as C. erecta and Linnaeus described it “foliis ovato-lanceolatis” (rendered by Pennell simply “lanceo- late”, and by Small, Fl. Se. U. S. 242, “linear to linear-lanceo- late”). The plant of European gardens, “Habitat in Virginia”, was well shown by Dillenius in Hortus Elthamensis, his plate cited by Linnaeus and, in view of the confusion evident in the Linnean Herbarium, selected by the late C. B. Clarke in DC. Mon. iii. 181, to stand for C. erecta. Dillenius showed proadly lanceolate leaves nearly 3 cm. broad, the spathes about 3 ¢m. long. Just such a plant, tall (up to 1.2 m. high), with broad leaves on the primary axis, and with large spathes, occurs 1? relatively rich soil from southeastern New York to Kansas, south to Florida, Louisiana and eastern Texas. Much of the material of this largest extreme of the species has the spathes (PL. 629, FIG. 1) subglabrous to only minutely hirtellous but, creatine confusion, a considerable series (pL. 629, FIG. 3), differing m 2° other notable character, has the base of the spathe as shaggily villous with long white trichomes as in the western plants called var. crispa (low and with linear or linear-lanceolate leaves and small spathes). This large plant, with white-villous spathe- bases (PL. 629, Fics. 2 and 3), is regularly dispersed through much of the range of typical C. erecta and, although it oe from it as var. crispa does from var. angustifolia, it can be calle only a form, not a geographic variety. The type of var. angustifolia, C. angustifolia Michx., a8 ate 15 Pennell in Bartonia, no. 19: 21 (1938). 1940] Fernald,—Additions to Flora of Virginia 437 by a photograph of Michaux’s plant (our PLATE 630, Fics. 1 and 2) has merely narrowly lanceolate leaves (6-8 cm. long, 4-9 mm. broad) and spathes 1.6-2 cm. long. Just such a plant, low, with narrowly linear to narrowly lanceolate leaves mostly less than 1 dm. long and 0.4-2 cm. broad and mature spathes mostly 1-2 (rarely —2.5) cm. long, occurs in dry sands or rocky soils from Florida to Texas and northern Mexico, north to Delaware, West Virginia, southern Illinois, Missouri and Kansas. In this typical var. angustifolia, from which the types of C. Swing- leana Nash (“leaves ... 4-6 cm. long, 5-7 mm. wide . . .; spathe -..2 cm. long”) and of C. Nashii Small (“leaf-blades 4-8 cm. long . . .; spathes 1 em. long”) are inseparable, the spathe (PL. 630, ric. 3) is rather densely short-hirtellous; but the reputed var. crispa (“leaves .. . 3-7 cm. long, 4-6 mm. wide . . .; spathe -.- 1.5-2 em. long, 1 em. broad”) seems to differ from it only in having the base of the spathe villous with long white trichomes. Var. crispa (pL. 629, FIG. 4) is supposed to occur from northern Mexico to Missouri and Nebraska and about the head of Lake Michigan. In the material from southern Michigan and northern Indiana I find no such long hairs, but specimens from Nebraska southward and southwestward often display them (PL. 629, Fic. 4). Except for this single character I find nothing by which var. crispa differs from var. angustifolia. The situation is nearly parallel with that in the broad-leaved typical C. erecta; in fact, It is quite parallel, for some of the low and narrow-leaved plants With relatively small spathes, in the sands of eastern Virginia, (PL. 629, Fic. 5) have the long trichomes which place them with C. crispa, supposedly restricted to the Southwest. I can, there- fore, see in the latter only a barely recognizable form of var. angustifolia. The fact that, upon geographic grounds alone, highly competent students have been placing the plants of north- em Mexico impartially in C. crispa, though most of them are typical var. angustifolia, is eloquent. The plant of northern Indiana, often referred to C. crispa, lacks the long white pubescence of the spathes of true but Telatively unimportant var. crispa. In the latter the longer leaves are mostly 4-10 em. long, the mature spathes usually 1-2 (rarely ~2.5) em. long. The plant of northern Indiana and adjacent Michigan (pu. 631), thence locally to Texas, has the 438 Rhodora [ NoveMBER longer (but narrow) leaves 7-15 cm. long, the mature spathes mostly 2.5-3 em. long. It seems to be a well defined geographic variety, with which it is a pleasure to associate the name of its chief collector, the always alert CHARLES C. Dram. In peninsular Florida, extending down to the Keys, there is another extreme (PL. 629, ric. 6), placed by Small and others in C. angustifolia. In the latter and in var. Deamiana, however, the mature spathes are more than half as high as long, with the lower margin straightish or only slightly curved, and the larger leaves are 4-20 mm. broad. The plant of the Florida Keys and southern and central Florida is very low and slender, the leaves only 3-5 mm. broad, the strongly faleate spathes long-beaked, 1.8-2.8 cm. long and less than half as high as long. It is a close match for the isotype in the Gray Herbarium of C. hamipila Wright, of Cuba. Clarke kept C. hamipila apart as a species, though stating that it was scarcely separate from C. erecta. The meagre material before him suggested that the spathes were solitary and axillary, never clustered. That they may be either solitary or grouped is now apparent from the several collections at hand. No difference in seed is evident and I am treating C. hamipila asa characteristic variety of southern Florida and Cuba. As I see Commelina erecta it is a polymorphous species with pronounced but freely confluent varieties, as follows: a. Stems (0.45-)0.6-1.2 m. high; larger leaves of prim axes oe af lance-ovate, (0.9-)1-1.5 dm. lone 15- 2.5-3.6 cm ‘ ee ais es eee ees C2 ee Var. typica. : Sten 1-4.5(-7) dm. high; larger leaves of primary axes narrowly linear e o linear-lanceolate, 3-12(-20) mm. broad; mature spathes . long. Mature oe sac than half as high as long, the cede shay only slightly curved; larger leaves 4-20 m 8 —~2(-2.5 oe ci hg 4-10 cm. long; mature spathes 1 pid angustifolia. OME OSES 6 6 OSH 6 OO O04 6H 6.8 Oo, 88 + 6.4 9 FS r. Deamiand. CCCCEHH HTH OOS ERE H EHO HHH OES s 3-5 mm yee 00 Var. ha C. erecta L., var. typica. C. erecta L. Sp. 1. 41 9) “Habitat in Virginia”: Clarke in DC. Mon. iii. 181 ( me Pennell in Bull. Torr. Bot. Cl. xliii. 104 (1916). —Loamy oF ‘ ra soil, more rarely in rocky places, southeastern New Yo a Kansas, south to Florida, Litictnas and Texas. PLaTeE 629, Fi 1940] Fernald,—Additions to Flora of Virginia 439 On the Coastal Plain of southeastern Virginia we have it only from SourHAMPTON County: dry woods, thickets and clearings along Three Creek, Drewryville, no. 5701. *Forma intereursa, f. nov. (TAB. 629, FIG. 2 et 3), caulibus 0.45-1.2 m. altis; foliis majoribus primariis late lanceolatis 1-1.5 dm. longis 1.5—4 em. latis; spathis maturis 2.3-3.6 cm. longis basi longe albido-villosis—DELAWaARE: Laurel, September 21, 1880, A. Commons. Vircinia: sandy thicket by James River, Jordan Point, Prince George County, July 23, 1938, Fernald & Long, nos. 2, 8643 (Type in Herb. Gray, 1isoryPE in Herb. Phil. Acad.) ; sandy pine and oak woods south of Zuni, Isle of Wight County, July 20, 1938, Fernald & Long, no. 8644 (transition to var. angustifolia) ; dry sandy open pine and oak woods, 6 to 7 miles south of Franklin, Southampton County, June 19, 1938, Fernald & Long, no. 8183. SourH Carotina: Congaree River, June 19, 1855, Hexamer & Maier; Santee Canal, June, —, H. W. Ravenel. Geroreia: dry rocky woods on southwest slope of Kennesaw Mountain, Cobb County, July 12, 1900, Harper, no. 219; north- west slope of Stone Mountain, DeKalb County, July 10, 1900, Harper, no. 11. Trnnessre: cult. Harvard Bot. Gard. 1847, from French Broad River, Gray. Mussourt: woods, Eagle Rock, August 12, 1905, Bush, no. 3222; dry soil, Prosperity, June 17, 1909, Bush, no. 2242. oo Var. angustifolia (Michx.), comb. nov. C. angustifolia Michx. Fl. Bor.-Am. i. 24 (1803); Pennell in Bull. Torr. Bot. Cl. xliii. 108 (1916). C. virginica, 8 angustifolia (Michx.) C. B. Clarke in DC. Mon. iii. 183 (1881). C. Swingleana Nash in Bull. Torr. Bot. Cl. xxii. 160 (1895) : Small, Fl. Se. U. S. 242 (1903). C. Nashii Small, 1. ¢. (1903).—Dry sandy soil, Florida to New Mexico and Mexico, north to Delaware, West Virginia, southern Illinois, Missouri and Nebraska. PLATE 630. “Var. ANGUSTIFOLIA is known in Virginia from the following Stations. York County: sandy clearings near Capital Landing, Queen’s Creek, Grimes, no. 4081. Warwick County: Old Point Comfort, September 15, 1885, N. L. Britton. Princess ANNE County: dry pine and oak woods, Cape Henry, Fernald & Griscom, no. 2805. Ise or WicHT County: dry sandy pine barrens, south of Zuni, Fernald, Griscom & Long, no. 6565. SOUTHAMPTON County: dry sand, pine barren about 7 miles South of Franklin, no. 7373. NaNnsEMonp County: dry sandy Woods along Pitch Kettle Creek, north of Lake Kilby, Fernald, ong & Fogg, no. 4840 (unusually tall, but with small spathes). “Forma albina, f. nov., petalis albis—-Dry sandy barrens, Cape Henry, September 24, 1933, Fernald & Griscom, no. 2804 (TYPE in Herb, Gray). . . 440 Rhodora [NoveMBER Forma eda (Wooton), comb. nov. C. crispa Wooton Bull. Torr. Bot. Cl. xxv. 451 (1898) Pennell, ibid, xliii. 107 agin C. erecta, var. crispa (Wooton) Palmer & Steyermark n Ann. Mo. Bot. Gard. xxli. 417 (1935), without bibliog reletanise. and in Ruopora, xl. 131 (1938), ibetrpete —Throug much of the range of the var. Puare 629, rics. 4 a *Forma CRIsPA is known in Virginia at the foots ‘stations. Fauquier County: by railroad, Beverly Mill, H. A. Allard, no. 1013. New Kent County: dry sandy field by Chickahominy River, Lanexa, Grimes, no. 4172. Sussex County: dry sandy hickory and oak woods, Burt, no. 6133. NaNnsEMoND COUNTY: white sand of pine and oak woods and clearings near Cathole Landing, west of Factory Hill, no. 12,285. Var. Deamiana, var. nov. (TAB. 631), caulibus gracilibus 2-6 dm. altis; foliis lineari-lanceolatis vel linearibus majoribus 7-15 cm. longis 5-15 (-20) mm. latis; spathis maturis plerumque 2.5-3 em. longis—Micuican: “State Collection”. Inp1ana: sand hills on west or southwest side of Lake Maxinkuckee, Marshall County, August 19, 1915 and July 15, 1920, Deam, nos. 17, and 31,900; sand hill north of Ora, Starke County, July i? 1920, Deam, no. 31,852; sandy cut along roadside 6 miles sout of Vincennes, Knox County, July 8, 1915, Deam, no. 17,087; very sandy soil along railroad at Duncan Siding, about 4 miles south: east of Vincennes, August 18, 1931, Deam, no. 50,986; $ ‘: dunes, Dune Park, Porter County, July 30, 1897, Umbae wooded sand hill, Dune Park, August 16, 1897, Agnes Chase; inner faces of frontal dunes, Mineral Spring, Porter County, August 13, 1920, D. C. Peattie; bog, Chesterton, Porter County, August 9, 1925, J. R. Churchi ill; very sandy roadside cut, i miles north and 114 miles west of Morocco, Newton Oi August 30, 1916, Deam, no. 21,511; roadside sand-hill, 4% mue> northwest of Morocco, July 13, 1920, Deam, no. 31,66 (xvee Herb. Gray). ILLINOIS: Oquawka, July 45 1908, Gleason. Kansas: shady places, Riley County, July 20, 4 a Norton, no. 524. OKLAHOMA: sandy hillside near Pas ‘Le ~ County, June 20, 1914, O. W. Blakley, no. 1456; Arbuckle eae tains, near Davis, June 25, 1917, W. H. Emig, no. 786. 0 EXAS- 3% pe north of Jasper, Jasper County, May 18, 1937, Cory, no. 2: Var. hamipila (Wright), comb. n C. hamipila wre 4 Sauvalle, Fl. Cubana, 157 (1873) ~Stems 1-4.5 Sra aera oathes m. with prolonged beak—Cuba and Florida. The following are Ae F 1940] Fernald,—Additions to Flora of Virginia 44] 19, 1910, S. C. Hood; high pineland, vicinity of Eustis, Lake County, Nash, no. 387; dry sandy soil, open woodland, Kelsey City, Palm Beach County, F. R. Randolph, no. 105; riverside, East Fort Myers, Lee County, A. A. Eaton, no. 1416; pineland, vicinity of Fort Myers, J. P. Standley, no. 137; pineland, Mul- lock Creek District, southeast of Fort Myers, J. P. Standley, no. 430; Pine Key, Blodgett: Key West, Blodgett. PuatE 629, Fic. 6. of Commelina erecta from Safe Harbor, Pennsylvania, Heller & H eller, no. 686; Fic. 2, a flowering tip, X 1 of forma intercursa from the type-station ; ; # A Tt in part; Fic. 5, spathes, x 2, of forma crispa from Burt, Sussex County, Virginia, Fernald & Long, no. 6133; Fic. 6, spathe, X 1, of var. hamipila PLATE 631, C. erecta var. Deamiana: rics. 1, 2 and 3, portions of TYPE, X 1; Fic. 4, inflorescence, X 1, from Mineral Springs, Porter County, Indiana, D.C. Peattie. ; *ANEILEMA Keisax Hassk, Thoroughly typical of fresh tidal marshes and shores from King and Queen to Nansemond County, and, doubtless, into North Carolina, flowering from September to frost. The following are the actual specimens collected. (Boulevard Postoffice) , nos. 11,290, 11,542; Chickahominy River, Lanexa, no. 11,543. CHaries Crry County: James River at Four Oaks”, below Harrison Point, no. 11,282; Kimage’s Creek, | ’ 5; James River, Wilcox Wharf, no. 10,994; Morris Creek ams Bri : : L and at Cox Landing (no. 11,295 , south of South Quay. See pp. 386, 391-394, 399 ene 400 a ie 20. A characteristic plant, here recorded for the first time outside eastern Asia. It is a regular constituent of the estuarine flora, along with Cyperus haspan L., var. americanus, Eryngium “quaticum, Rhynchospora macrostachya, var. colpophila (see above) and the several other species which inhabit these peculiar 442 Rhodora [ NovEMBER habitats. I have studied the material most carefully with Dr. Hiroshi Hara, who has been able to supply finely fruiting Asiatic specimens. In all characters, including the seeds, our plant seems quite inseparable from that of eastern Asia. TRADESCANTIA ROSEA Vent., Var. GRAMINEA {ene Anders. & Woodson. Local range extended eastward to NaNsEMOND County: white sands ha i barrens tea of South Quay (several nos.). See p. 3 JUNCUS CAESARIENSIS Coville. Range in the state extended southward to Dinwippie County: springy sphagnous swale about 5 miles east of Burgess Station, no. 11,000 At our Henrico County station J. caesariensis was associated with an extraordinary series of species with disrupted ranges (see Ruopora, xli. 470, 473). So at the Dinwiddie County station it shares the honors with Fuirena breviseta, range eX- tended north from North Carolina, Scirpus debilis, at our only station on the Virginian Coastal Plain, and other plants of great rarity in the region. See p. 389. J. ABORTIVUS Chapm. To the single station pape north of South Carolina Gin Isle of Wight County) add from NANSEMOND OUNTY: very abundant in wet or moist sandy ind peaty depres- sions of pinelands and pine barrens south of South Quay (many nos. Ga different stations). See p. 379. MEGACEPHALUS M. A. Curtis. To the previously recorded stations (False Cape and vicinity) add another, farther north in Princess ANNE County: inner border of brackish to fresh marsh along Back Bay, at eastern mare of Long Island, no. 10,574. See pp. 370 and 371 and map 5 J. DIFFUSISSIMUS Buckley. To the two stations previously recorded add the following, in SovrHampton County: wet clear- ing slightly west of Branchville, no. 10,190; along woodroad in wooded bottomland, Meherrin River, southeast of Branchville, no. 10,192; wet sandy roadside ditch south of Shiloh, no. 10, 191; sandy. roadside ditch southeast of Windman’s Mill, no _ 12,088 GREENSVILLE County: argillaceous clearing in swa mpy W ; near Readjuster Bridge over Nottoway River, ieee 0 Orion, no. 12,039. TOFIELDIA RACEMOSA (Walt.) BSP. Aste range extended ir ward to Nansemonp County: sphagnous savannah-like sW# . cd of Cherry Grove, south of South ae no. 10,577. See P- ZIGADENUS GLABERRIMUS Michx. Local range extended eas ni ward to NANSEMoND County: wet peaty pine barrens, eas 1940] Fernald,—Additions to Flora of Virginia 443 Cox Landing, south of South Quay, no. 10,579. Iste or WicHT County: sandy and peaty pine barrens, south of Lee’s Mill, no. 12,294. See pp. 374 and 381. MELANTHIUM virGINIcUM L. Local range extended eastward to NANSEMoNnD County: wet bushy swale east of Cherry Grove, south of South Quay, no. 11,003. See p. 381. *“Litium CatesBaEI Walt., var. Longii, var. nov. (TAB. 632, FIG. 1-4), a var. typico recedit bulbi squamis ovoideis efoliosis; foliis oblongis, imis mediisque obtusis 6-13 mm. latis; tepalis laxe adscendentibus apice vix recurvatis ad 3.5 em. latis; capsulis subtruncatis 2.5-4.5 em. longis; seminibus 5-6 mm. longis ala 1.6 mm. lata.—Local, from southeastern Virginia to Georgia and Alabama. The following are placed here. Vircrn1a: sphag- nous savannah-like swale east of Cherry Grove, south of South uay, Nansemond County, July 21, 1939, Fernald & Long, no. 10,582 (Type in Herb. Gray, 1sorypr in Herb. Phil. Acad.), October 15, 1939, Fernald & Long, no. 11,553. Nort Caroutna: Pineland at Delway, Sampson County, August 25, 1938, R. K. Godfrey, no. 6176; pineland, White Lake, Bladen County, Au- gust 14, 1938, Godfrey, no. 5988; peaty grass-sedge savannah at Carolina Beach, New Hanover County, August 7, 1938, God- frey, no. 5903. Grorcra: Columbus, Boykin. ALABAMA: low, grassy pine barrens near Bolling, Butler County, August 28, 1885, J. Donnell Smith. See pp. 381 and 403. Typical Lilium Catesbaei (our PLATE 632, Fics. 5-10) has the slender bulb-scales often, or usually, terminated by long erect linear leaf-blades; its cauline leaves are acutish to long-attenu- ate, the broader ones (in a large series of specimens) 2—12 mm. broad. Its tepals have typically long slender claws and pro- longed recurving tips, the claw being usually 0.25-0.36 as long as the blade, measurements of flowers of 32 collections showing the claw averaging 0.26 as long as the limb or, in other words, the limb about three times the length of the claw. Measure- ments of the limb of the broadest tepal show a range from 4.5—9.5 em. long, with an average length (from 60 specimens) of 7 cm., the breadth ranging from 1-2.6 cm. (average 1.8 em.). The capsule narrows gradually to a beak and the seeds are 4-5 mm. long, with wing (at broadest point) 0.6-1.2 mm. wide. Var. Longii (PLATE 632, rics. 1-4) on the other hand, has the plumper bulb-scales without the leafy tips; the middle and lower cauline leaves are blunt and oblong, 6-13 mm. broad; the tepals are loosely ascending, without prolonged and recurving tips, consequently the proportion of claw and blade is very differ- 444 Rhodora [NovemBER ent, measurements of all the collections at hand showing the claw to be 0.3-0.42 (average 0.37) as long as the blade, in other words the blade about twice the length of the claw. In actual size the largest blades show a range from 4.5-6.5 (av. 5.5) em. long, and 1.8-3.4 (av. 2.6) cm. broad. The capsule is subtruncate at summit, not narrowed to a beak, and the seeds are larger, 5-6 mm. long, with wing 1.4-1.6 mm. wide. In their extremes the two plants are quite unlike. Unfortu- nately, however, transitions occur and several collections are as near one as the other. I am, therefore, constrained to treat the blunt-leaved plant with broad tepals scarcely recurving as 4 marked variety. It is noteworthy that at the northern limit of the variety it was in full flower on July 21. At its next known area to the south it is in full flower from August 7-14 (the collection of August 25 showing the flowers fading), while the Alabama material was collected in prime flower on August 28. Lilium Catesbaei, according to English writers and also grow- ers of lilies about Boston, is almost impossible to cultivate, mn part because it stands no freezing. Now that we have a variety of it, sturdier and perhaps handsomer than typical L. Catesbaet, with a natural northern station in Virginia, where ice frequently forms and the children look forward to out-door skating, there is hope that var. Longii may prove to be a successful plant in northern gardens. Fully ripe seed collected in October has been supplied to the Harvard Botanic Garden and to several private growers of lilies. In this study I have had the great advantage of having before me the material from the herbarium of the New York Botanical Garden. With its special concentration upon the Southeastern States, it was expected that this collection would add materially to the known stations of var. Longii. Very strikingly, howev : e 5 and 6, base and flower, x %, from Baldwin, Duval County, 1871, Nash, no. 2321; ric. 7, capsule, X 1. from Bluffton, South Carolina, ston, Mellichamp ; ria. 8, seeds, X 5, from Eutawville, South Carolina. Egg'e no. 5018 (N. Y. Bot. Gard.); ria. 9, seed, x 5, 1 Ms _, ieee Fria. 10, seed, x 5, from Florida, Chapman ard.). 1940] Fernald,—Additions to Flora of Virginia 445 ALETRIS AUREA Walt. Local range extended eastward to NaNsrMonp County: sandy and sphagnous margins of thickets in pineland southwest of Marsh Hill School, south of South Quay, no. 11,006. *TRILLIUM LANCEOLATUM Boykin. ?NorrotK County: Great Dismal Swamp, west of Wallaceton, April 24, 1926, Paul A. Warren, no. 413. Specimen (unidentified) received in exchange from the College of William and Mary. Professor Warren tells me that it could have come only from west of Wallaceton, but whether in Norfolk or southeastern Nansemond County he does not know. First record from north of Georgia. See p. 375 and map 11. SMILAX PULVERULENTA Michx. To the very few recorded stations add the following. Sussex County: rich woods _and bushy clearing just east of the “fall-line” along Nottoway River, Double Bridge, about 6 miles northwest of Jarratt, no. 11,302. SouTHAMPTON County: rich sandy and loamy woods along Three Creek, northwest of Carey Bridge, no. 10,210 tay). NortH Caronrna: open woods, Winston-Salem, June 30, 1921, P. O. Schallert. Typical Hypozis hirsuta, with a very inappropriate name, as the scapes and pedicels sparsely pilose, the expanding perianth loosely short-pilose but soon glabrate and the fruit sparsely pilose. Forma villosissima is conspicuous on account of the permanent and very dense long white villosity of scape, pedicels, perianth and fruit. At the type-station it is very scarce but there associated with Seymeria cassioides and other species of dry pinelands. H. LEPTOCARPA Engelm. & Gray. Range extended northward to New Kent County: bottomland woods by Chickahominy River north of Long Bridge, southeast of Quinton, no. 11,304. A considerable northward extension, from Greensville, South- 4mpton and Nansemond Counties. See p. 397. SisyRINcHIUM MucRONATUM Michx. Local range extended Guchward into GrEENsviLLE County: rich woods near Three teek, north of Emporia, no. 10,217. 446 Rhodora [ NovEMBER *>< GLADIOLUS GANDAVENSIS Van Houtte. Dinwippie County: old field, south of Petersburg, no. 10,593. Seen persisting in other old fields NNA FLaccipA Salisb. Prince Grorce County: roadside fill bordering wooded swamp, northwest of Disputanta, no. 10,008. URMANNIA BIFLORA L. Local range extended eastward into NANSEMOND County: sphagnous obese: like swale east of Cherry cide south of South Quay, no. 11,009. *< HABENARIA CanBy1 Ames. A rinple plant flowering with abundant iL blephariglottis and H. cristata, sphagnous savannah- like swale east of Cherry Grove, south of South Quay, no. 10,601. See p. 381. CLEISTES DIVARICATA (L.) Ames. NansemMonp County: scat- tered and scarce at two stations ae of South Quay, nos. 11,010 and 11 pli See pp. 380 and 384. *CALOPOGON PALLIDUS Chapm. NaNnsEMOND COUNTY: sandy and Seats pine — east of Cox Landing, south of South Quay, no. 10,603; moist pe aaa in white sand of pine barrens 1— 1% ce piney of Cherry Grove, south of South Quay, no. 12,054. See p. 380 and ees 13, Extension north from North Carolina. Identified by Dr. D. S. Correll. SPIRANTHES CERNUA (L.) L. C. Richard, var. ODORATA (Nutt.) Correll in Bot. Mus. Lfts. Harvard Univ. viii. 79 oe (S. Courthouse, no. 11,557 ” (some root-tips budding into new rosettes). CHARLES Crry County: fresh tidal marsh along Kit- tewan Creek, Weyanoke, no. 11,307 (plants up to 7.5 dm . high, completely submerged at high tide). GREENSVILLE Count: wooded bottomland of Fontaine Creek, southeast of Taylor's Millpond, nos. 10,224 and 11,308 (completely drowned during 94 and 401 much of es summer). See pp. 359, d PONTHIEVA RACEMOSA (Walt,) Mohr. Local range extende into Savemsarton County: rich marly woods along Three Creek, northwest of Cary See eg no. 11,487. See p. 363. f HEXALECTRIS APHYLLA (Nutt.) Raf. Local new stations ° very limited extent in Dinwmpm County, no. 10,605, and in OUTHAMPTON County, no. 10,606. *Carya ovata K. Koch, var, PUBESCENS Sargent. County: rich deciduous woods along Nottaway River, west of Homeville, no. 10,226. Described by Sargent from South Carolina and Tennessee southward. See p. 361. *CASTANEA PUMILA (L.) Mill., var. Marcaretta Ashe. Ne MOND County: Dry white sand of pine barrens east of Sandy Sussex south- 1940] Fernald,—Additions to Flora of Virginia 447 Landing, south of South Quay, no. 11,313; similar habitat, south- west of Marsh Hill School, south of South Quay, no. 11,014. SourHAMPTON County: white sand of pine and oak woods north of Point Beach, south of Franklin, no. 11,315. First records from northeast of western Alabama. Our speci- mens, having a well developed pale tomentum on the lower leaf- surfaces, belong to the form later called C. Margaretta var. arcuata Ashe (from Texas). See p. 398. *CASTANEA PUMILA (L.) Mill., var. AsHEr Sudworth. NansE- MOND County: sandy and peaty pine barrens east of Cox Land- ing, south of South Quay, no. 11,310. SourHampron County: sandy woods, southeast of Round Gut, southwest of Franklin, no. 11,309 Extension north from eastern North Carolina. See p. 398. 4 . QuERCUS PALUSTRIS Muench. Sussex County: by woodland brook northeast of Stony Creek, no. 10,617; border of wooded swamp north of Stony Creek, no. 10,618; wooded bottomland, Three Creek, southwest of Grizzard, no. 10,237. Although Sargent (Man.) cites the Pin Oak as occurring in Virginia south of the Potomac only in the extreme west (Wythe County), it is characteristic of river-swamps over a considerable area of southern Sussex County. It was reported in Claytonia, 1. 44 (1934) from Amelia and Charles City Counties as well as from counties farther in the interior. See p. 361. Q. Lazvis Walt. (Q. Catesbaei Michx.). NansEMoNnD County: very abundant in the sandy area south of South Quay, many hos. See pp. 362 and 365. XQ. Busan Sargent (Q. marilandica < velutina). GREENS- Mpc County: large tree on wooded bottomland by Nottoway iver, northeast of Purdy, no. 11,015. Sargent (Man.) cites x Q. Bushii only from Georgia, Florida, Alabama, Mississippi and Oklahoma. The tree near Purdy iruits heavily. See p. 383 * me Counry: a single large shrub, sandy and ae east of and Dixwippprm County: becoming abundant by roadsides m waste places, Petersburg, no. 11,018. See p. 390. 448 Rhodora [ NoveMBER *Rumex AcETOSELLA L., var. INTEGRIFOLIA Wallr. Surry County: roadsides and waste places, Surry Courthouse, no. 10,626. See p. 377. Very striking, with most of its rosette-leaves slenderly taper- ing to base and exauriculate. Infrequent in North America. *POLYGONUM HYDROPIPEROIDES Michx., var. breviciliatum, var. nov., planta subrobusta; foliis primariis lanceolatis acumi- natis subtus scabris 9-13 cm. longis 1.5-3.2 cm. latis; ochreae ciliis perbrevibus 0.8-1.2 mm. longis; ochreolis eciliatis vel ciliis ad 0.4 mm. longis; floribus valde exsertis; achaenio trigono— Dinwiddie County, Vircrnia: deep ditch in swale south of Burgess Station, July 16, 1938, Fernald & Long, no. 8698, August 15, 1938, no. 9044 (rvpn in Herb. Gray; 1tsorypE in Herb. Phil. Acad.). When he defined the eight geographic varieties of Polygonum hydropiperoides in Ruopora, xxviii. 22-28 (1926), Stanford said: “It is highly improbable that varieties recognized in this paper exhaust the list of admissable geographic varieties. Among possible fields for future geographic research, the south- ern seaboard states below Virginia and those of the lower Missis- sippi basin, which are not particularly well represented in the material at hand, suggest attractive possibilities.” Var. brevi- ciliatum is presumably in North Carolina, though as yet we have only the material from Dinwiddie County. It is at once sepa- rated from the other eastern American varieties as follows. From typical P. hydropiperoides by its coarser habit, broader leaves (1.5-3.2 cm. wide as against 1-1.5 cm.), cilia of the ochreae only 0.8-1.2 (as against 2-4) mm. long, and of the ochreolae wanting or barely 0.4 (as against 0.5-1) mm. long. Var. psilostachyum St. John, type from Sable Island, Nové Scotia, has glabrous leaves barely acuminate and only 48 em. broad, eciliate and glabrous ochreae and ochreolae, the latter with broadly open summit; var. macerwm Stanford, a very slender plant of Florida (only 5-6 dm. high) has leaves barely 6 cm. long and 1 cm. broad, cilia of the ochreae 2-4 mm. long; of the ochreolae 2 mm.; var. sanibalense Stanford, also fro e Florida, is stout, but with elliptic obtuse short leaves, long cilia of ochreae, and ochreolae broadly turbinate. Superficially V" breviciliatum closely resembles var. Bushianum Stanford of Oklahoma, and ‘yar. aspersfoliwm Sanford of Caliorma: ame 1940] Fernald,—Additions to Flora of Virginia 449 those varieties, however, have the ochreae and ochreolae long- ciliate, the cilia of the ochreae in var. Bushianum about 1 cm. long, in var. asperifolium 5-6 mm. long. Var. breviciliatum can scarcely be forced into any of these recognized varieties. *POLYGONELLA ARTICULATA (L.) Meisn. NansEMOND County: dry white sand of pine barrens east and southeast of Cox Land- ing and east of Sandy Landing, south of South Quay, nos. 10,632, 11,025, 11,562 and 11,563. Ist or WicHt County: dry sandy pine barrens and open woods, south of Lee’s Mill, no. 12,068. See pp. 379 and 403 and map 23. A very extensive area, the plants, flowering in mid-October, often reaching a height of 8 dm. Although given a range south to Georgia (“or Fla.”) by Small, Polygonella articulata is repre- sented in the herbaria of the Academy of Natural Sciences of Philadelphia and the New York Botanical Garden and in the Gray Herbarium by no previous collections from south of northern Worcester County, Maryland, except for an old speci- men bearing in Nuttall’s hand the indefinite data: “Georgia. Dr. B[aldwin]”. Whether it was actually collected in Georgia or whether there was confusion of data is an open question. Baldwin was born in Chester County, Pennsylvania; studied at the University of Pennsylvania; practiced medicine at Wilming- ton, Delaware, where he married; and, after moving to Georgia, made visits at Wilmington. The specimen which Nuttall had might have been picked in the North, the label being Nuttall’s, not Baldwin’s. In a letter dated “Jefferson, November 27, 1816”, Baldwin wrote a friend: The distance from our lodgings on the Sand Hills to Fort Barrington, Where We crossed the Altamaha, is about seven miles; and from thence ‘o this Place, forty one. I have frequently mentioned Fort Barrington ; ie tio puzzle a stranger to find any Fort, here. The —S ue he Feiry le, in the midst of a luxuriant vegetation, a short way below Among the more rare productions of the Sand Hills, we find the ‘ w age Mernonum: Polygonum articulatum, and Stipulicida fe ichaux;—all of which are found on the middle districts of a, and not on the sea-coast. . . . I suspect, with Pursu, that the f PA et articulatum, and polygamum, are merely varieties. The one ®, however, has always flores albi,—and is frequently more and the? /¢et in height. That the northern plant should be annual, outhern one perennial, is by no means extraordinary. oo in white sand of pine barrens, Wyanoke, Gates County, North Carolina, no. 11,564. ‘am Baldwin as quoted by Darlington, Reliquiae Baldwinianae, 334 (1843). 450 Rhodora [NoveMBER The last sentence quoted clearly shows that Baldwin knew as Polygonum articulatum (=Polygonella) in the Sand Hills of Georgia a white-flowered perennial, not the usually pink-flowered annual, but that he knew the northern plant as an annual (ap- parently not white-flowered). The Georgian source of the Nut- tall specimen is not verified by its reputed collector. *AMARANTHUs Torrey! (Gray) Benth. Dinwippig County: roadsides and waste places, Petersburg, no. 10,633 Native from Iowa to Texas and westward. A colony of vigor- ous plants, likely to spread. See p. 369. IRESINE RHIZOMATOSA Standley. Range extended farther north in Princess ANNE County: low woods and clearings along Back sae Ore Island, nos. 10,636, 10,637 and 11,028. See pp. 370 For discussion see Ruopora, xxxviii. 379 and 416 (1936) ; and xxxix. 483, map 58 (1937). *FROELICHIA GRACILIS (Hook.) Moq. Dinwippre County cinders of freight-yard of Norfolk & Western Railroad, Peters- burg, no. 10,639. Sussex County: similar habitat about 2 miles west of Waverly, no. 10,638. Hrnrico County: waste places and railroad ballast, Richmond, no. 12,3 A western species (Iowa to Arkeansan. and southwestward), ikely to become more established. See p. 374. FASTIGIATA (Raf.) Fern., var. PALEACEA Fern. in Ruopona, Xxxviii. 421, pl. 447, figs. 6 and 7 (1936). Sussex County: border of rich deciduous woods along Nottoway River, southwest of Homeville, n ops ap white sand of dry woods to Owen’s Store, no. 12,341; moist sandy roadside ditch, ei Nottoway River, at Green Church Bridge, northwest of Owens Store, no. 12 342. First from south of Delaware and Pennsylvania. See P- 361. *P. rrpartA Chapm. Dry sands and sandy woods of ISLE : WIGHT, ouuas ere SouTHAM MPTON COUNTIES (many numbers This is the plant erroneously reported in 1937 as P. Baldwint. I am a to Dr. Core for clarifying the identification. TUM MARITIMUM (Walt.) BSP. To the station fos Piecisy formerly reported add the following, also in PRINCES Anne County: open muddy and sandy shores of Back SS east of Bingen no. 11 029: east of tecada, nos. 11,030 and 11, See pp. 369 and 387. 1940] Fernald,—Additions to Flora of Virginia 451 *STELLARIA MEDIA (L.) Cyrill., var. GLABERRIMA G. Beck. Lawns of GREENSVILLE and SouTHAMPTON This is the pernicious weed most erroneously reported in Ruopora, xli. 489 and 540 (1939) as the native southern Stellaria prostrata Baldwin. The latter is still unknown in Virginia. My apology for the erroneous identification is the lame one, that I accepted the verdict of a student who was making an intensive study of the genus. As I have elsewhere pointed out, the motto of the true scientist unfortunately has to be: “you can trust no one.” We sometimes weaken and do so! SILENE NocTirLora L. Prince Grorce County: cinders of freight-yard, Norfolk & Western Railroad, east of Petersburg, no. 11,3835. Nupwar FLUVIATILE (Harper) Standley. Range in the state extended eastward to NANSEMOND CoUNTY: pages 4 margin a Blackwater River, Milk ee south of South Quay, n 10,643; seen, forming a definite band of vegetation near the en irae margin of the river, tte to the draw-bridge at ou hiseconc FLABELLARIS Raf., forma RiPARIuS Fernald in Ruopora, xxxviii, 171 (1936). New Kent County: bottomland woods by ramsey | River north of Long Bridge, southeast of Quinton, no. 11,336. Sussex County: gum and cypress swamp bordering ile. east of Littleton, no. 12,348. SourH- AMPTON County: dried-out pools, wooded bottomland, Meherrin River, southeast of Branchville, no. 10,259. Certainly local in southeastern Virginia. Our Varieties or RANUNCULUS BULBosUS. In the Northeast- ern States Ranunculus bulbosus L. is a tolerably uniform plant, with the three leaflets of the radical leaves cleft into 2 or 3 divisions, these again with short segments, the petioles and stems silky-villous to glabrate. In southeastern Virginia, however, the species there abundantly naturalized, is much more variable, three fairly definite varieties being recognizable in the field. These were worked out with me several years ago by Mr. Lud- low Grisecom but our notes never published. They are here in- cluded. Besides the typical form, R. bulbosus, var. typicus Erdner, we there find var, VALDEPUBENS (Jordan) Briq., Fl. Corse, i. 619 (1910), with foliage as in var. typicus, but the whole plant strongly spreading-villous, often giving a hoary effect; and var. DissEctus Babey, Fl. Jurassienne, i. 39 (1845), with the leaves finely dissected into elongate linear segments. 452 Rhodora [ NoveMBER *Var. VALDEPUBENS in the United a is chiefly in the South- east and only occasional northward into New York and Rhode : ave examined the a peat specimens. RHODE IsLaNp: Cumberland, May 30, 1911, C. H. Knowlton. New York: bank of Mohawk River, east of inc Saratoga Co., May 30, 1906, S. H. Burnham. VIRGI - Buckro oe, May 18, 1912, Robinson, no. 302; near Norfolk, April 23, 1911, PMc no. 4495: roadside east of Little Creek, Princess Anne Co., Fernald & Griscom, no. 4402; cinders of freight- -yard of Atlantic Coast Line Railroad, Petersburg, no. 12,07 *Var. DISSECTUS is more generally dispersed, but apparently local: Marne: Somesville, Mt. Desert Isl., June 15, 1928, G. L. Stebbins, Jr. MassacHUsErts: near Spring Pond, Peabody, June 11, 1896, J. H. Sears; Beaver Hill, Middlesex Fells Reserv., June 9, 1 920, N. T. Kidder; Blue Hills Reserv., June 10, 1920, N.T. Kidder. RHODE ISLAND: Nayatt, Barrington, May 30, 1911, M. L. Fernald; gg oan May 31, 1908, E. F. Williams. New York: Inwood, New York City, May 14, 1887, Mrs. L. M. Parker; Ledyard, Cayuga Co. ., Wiegand, no 6449. VIRGINIA: Campbell, Bedford Co., May 14, 1871, A. H. Curtiss; near Black- water River, Princess ‘Anne Co., Fernald & Griscom, no. 1. MicHican: Agricultural College, May 23, 1894, C. F. Wheeler. SASSAFRAS ALBIDUM (Nutt.) Nees. GREENSVILLE, SouTHAMP- TON, Sussex and Minaaiess Counties (many nos. The typical form of the species, with glabrous ‘nanshlue and leaves; often more abundant than the pubescent extreme. *Rorippa sEssILirLora (Nutt.) Hitch. Henrico County: James River, Richmond, May 11, 1894, J. R. Churchill. A characteristic species of the Mississippi Valley. RABIS CANADENSIS L. Sussex County: rich deciduous woods along Nottoway River, southwest of Homeville, no. 10,272. Our first station on the Coastal Plain. See p. 361. DRoSERA ROTUNDIFOLIA L. NANSEMOND CouNTY: SW . pressions in pine barrens south of Cox Landing, south ee South Quay, no. 10,663; sphagnous savannah-like swale east of Che e Grove, south of South Quay, no. 10,662. Seen some years ago SLE OF WIGHT County: in sphagnum near Joyner’s Bridge. It is a striking fact that the only stations we know 0M ms Coastal Plain (south of the James) for Drosera rotundifolia at near the eastern border of the Blackwater River. D. tnrerMepIA Hayne. Nansemonp County: sphagn way a swale east of Cherry Grove, south of South Q no. 10, § Sa- 1940] Fernald,—Additions to Flora of Virginia 453 Very local in southeastern Virginia: frequent in Princess Anne County but rare on the Coastal Plain farther west. j *SEDUM ALBOROSEUM Baker. GREENSVILLE County: natural- ized in roadside thicket near Dahlia, no. 9575 peaty and sandy thicket in pine barrens south of Lee’s Mill, nos. 12,358 and 12,359. SurrotK County: dry woods of a “ham- tional). SourHampron County: sandy woods southeast of Round Gut, southwest of Franklin, no. 11,340; swampy woods west of Wiggins School, south of Franklin, no. 11,341. Much of the material has the lower surfaces of the leaves strongly whitened but I am unable to separate it on any character (including seeds) from the thick-leaved shrub of Nova Scotia and New England with densely pubescent lower leaf-surfaces. Torrey & Gray, taking up Nuttall’s variety, cited it from the Mountains of Pennsylvania and from Louisiana. There has been no material in the Gray Herbarium from south of Nova Scotia, Maine and Massachusetts, where the lower surfaces are often, but not always, rufescent; but Mr. Long sends me for study the Pennsylvania material at the Philadelphia Academy. This consists of Nuttall’s type from “mts., Penn”, and character- istie specimens from Pocono (Wolle) and from Sand Patch, alt. 2500 feet, “summit of Allegheny Mts.”, Somerset County (C. F. Saunders). See p. 380. GILLENIA TRIFOLIATA (L.) Moench. Sussex County: dry sandy hickory and oak woods, Burt, no. 11,042. Our only Coastal Plain station. “POTENTILLA INTERMEDIA L. Surry County: roadsides and waste places, Surry Courthouse, no. 10,676. See p. 377. ; BUS GENICULATUS Kalt. Princess ANNE County: exten- ‘ively trailing, waste ground at old railroad terminal, Munden, 10. 10,674. Dinwippme County: roadsides and waste places, Petersburg, no. 10,673. One of the sprawling and climbing European species. R. Grimesi Bailey. Local range extended southward. Svus- Sex County: rich deciduous woods along Nottoway River, south- West of Homeville, no. 10,287. GREENSVILLE CouNTy: open 454 Rhodora [ NoveMBER thickets, clearings and borders of woods east of Emporia, no. 10,291. *R. JANssoni Bailey. Sussex County: wet woods, Assa- moosick Swamp, about 2 miles northeast of Homeville, no. 10,286. I am unable to find any points to separate this from the trailer of southern New England. See p. 361. *Rusus (Eupatus, § THOLIFoRMES) pernagaeus, sp. NOY. (raB. 633 et 634), arcuans, cannis simplicibus 6-7 dm. longis apice radicantibus; primocannis 2-3.5 mm. diametro subteretibus viridibus glabris sparse setosis setis aculiformibus, deinde un- cinatis; primocannae foliis ternatis subquinatis, immaturis supra strigoso-pilosis subtus subvelutinis; foliolis terminalibus cuneato- obovatis 2.7-3.2 cm. longis argute duplicato-serratis sublobatis; floricannae foliis ternatis foliolis anguste cuneato-obovatis, ram- orum sterilium foliolis 1.5-3.5 em. longis; ramorum floriferum foliolis terminalibus vel foliis simplicibus 1.5-2.5 cm. longis; corymbis 1-3-floris; pedicellis laxe adscendentibus 1-2 em. longis valde glandulosis plerumque bracteolatis; calycibus glandulosis lobis late ovatis 4 mm. longis deinde reflexis; petalis roseotinetis 6-8 mm. longis 4 mm. latis; fructibus ignotis—Isle of Wight County, Vircrnra: roadside thicket near Smithfield, April 5, 1938, Fernald & Long, no. 7879, distributed erroneously as R. pauxillus Bailey. Rubus pernagaeus (of the land of ham, Smithfield hams, from peanut-fattened hogs, having a reputation which has extended far from Virginia) was placed under R. pauzillus as a simple but quite unsatisfactory means of getting the specimens roughly identified. R. paucillus, characterized by Bailey as “Littlest of the upright blackberries in the United States”, is stiff and erect (instead of arching and tip-rooting) and belongs in § a Rydb. It has the primocane-leaflets ovate (instead of narrow y cuneate-obovate) ; its floricane-leaflets, likewise, broadly (instead of narrowly cuneate-obovate) ; “pedicels pubescent an sometimes with a few glandular hairs” (in R. pernagaeus 7 ously glandular); calyx apparently glandless, and petals are longer. R. pernagaeus in some ways resembles members © Procumbentes, though the canes are more arching and less pros trate than in theoretically characteristic members of the fe cumbentes. Its tiny corymb of 1-3 flowers is suggestive of ie section, but I am unable to place it with any described ae i of that group. R. Enslenii Tratt. sometimes has glandu 1940] Fernald,—Additions to Flora of Virginia 455 pedicels and calyx, but the very short pedicels and small pink flowers and the cuneate primocane-leaflets of R. pernagaeus keep it apart from R. Enslenii. R. centralis Bailey (type from south- ern Indiana) has glandular pedicels, but the plant is trailing, the primocanes branching, the terminal primocane-leaflets “broadly subcordate-ovate”, the floricane-foliage similar, and the calyx- lobes ascend in fruit. *SANGUISORBA MINOR Scop. Dinwippie County: dry field about 5 miles east of Burgess Station, no. 11,043. See p. 390. *CASSIA FASCICULATA Michx., forma JENSENI Palmer & Steyer- mark. Sussex County: sandy pine and oak woods south of Stony Creek, no. 11,348. An albino. *CaSSIA FASCICULATA Michx., var. macrosperma, var. nov. (TaB. 635, FIG. 1 et 2), planta 1-1.7 m. alta; ceaule ramisque piloso-hirsutis vel glabratis; pedicellis hirsutis; leguminibus plerumque 4-8.5 em. longis 5-10 mm. latis strigosis vel glabratis, suturis hirsutis, segmentis 5.5-7 mm. latis; seminibus 4-10, oblique rhomboideo-ovatis valde rostratis plerumque 5.5—7. ‘ longis 4-6 mm. latis—Fresh tidal marshes and shores, south- eastern Vircinta: sandy tidal shore of James River, at “Four Oaks”, below Harrison Point, Charles City County, September 16, 1939, Fernald & Long, no. 11,349; fresh tidal marsh along Kittewan Creek, Weyanoke, Charles City County, September 18, 1939, Fernald & Long, no. 11,350 (Type in Gray Herb.; ISOTYPE in Herb, Phil. Acad.), October 14, 1939, no. 11 973 (fully "pe legumes) ; fresh tidal marsh of Chickahominy River, below Barrat’s Bridge (or Ferry), James City County, September 19, 1939, no. 1135 Pamunkey River, southeast of White House, New Kent County, October 14, 1939, no. 11,574 (fully ripe); fresh tidal shore of Mattaponi River at Horse Landing, near King William Court- ruse, King William County, October 14, 1939, no. 11,575 (legumes rather small); fresh tidal shore of Mattaponi River, Cassia fasciculata, var. macrosperma is remarkable for its steat stature and large legumes and seeds; it is also noteworthy lor its apparent restriction to the tidal reaches of the rivers and creeks, partially or wholly submersed at high tide, partially “nersed at low tide. It thus forms a characteristic element in the estuarine flora of the region. 456 Rhodora [NovEMBER Typical C. fasciculata (C. Chamaecrista of authors, not L.) is mostly much lower (1.5-9 dm. high), with minute appressed pubescence, the legumes (ric. 3) 2.5-5 em. long and 4-6.5 mm. broad, with segments 3-4.5 mm. broad, the 4-13 seeds (Fic. 4) 3.5-5 mm. long and 2.5-4 mm. broad. Var. robusta (Pollard) Macbride, chiefly of the Mississippi drainage, differs from typical C. fasciculata primarily in being more spreading-pubescent; its legumes (Fic. 5) and seeds are not conspicuously larger. Pate 635, Fic. 1, shows portions of the Typr, X 1, of ee fasciculata, var. macrosperma; FIG. 2, seeds, X 3, from the type. Fic. 3 is ripe fruits, x 1, of typical C. fasciculata from west of White Bluffs, Dickaod County, Tennessee, Svenson, no. nape, FIG. 4, vente Se 3, of C. fasciculata from New Castle, Delaware, Benner, n 5 fruits X< 1, of var. robusta from Auburn, Alabama, Earle & toe, no. 30. *TRIFOLIUM PRATENSE L., forma prLosum (Griseb.) ae GREENSVILLE CouNTY: open thickets, clearings and borders 0 woods east of Emporia, n no. 11,051. An extremely long- ie form. *Mepicaco LtupuLiIna L., var. GLANDULOSA Neilr. PRINCE GrorcE County: siincinss of freight-yard, Norfolk & Western Railroad, east of Pet epee no. 11,352. *TEPHROSIA SPICATA (Walt.) T. & & var. semitonsa, me caulibus sparse ailesechindutie: rhachibus sorrenyees “folio is supra glabris vel glabratis; 0 Saeed sparse strigoso-pilosis. Vircin1A: Southampton County: white sand of pine an 729, woods at Round ps sowihivekt of Franklin, September 20, 19 % Fern ald & Long 11,353, plants stiffly pe a yok in ‘Her any woods and adjacent clearings, Kilby, giernte We 1085 Fernald, Long & Fogg, no. 4892; sandy woods, Facto Wi ht August 26, 1936, Fernald & ae no. 6 sle of Wi8 south of Zuni, August 24, 1936, 'Fonold é Long vt sand of dry woods and clearings east of Joyner’s ‘ aealee, tly 1940, Fernald & Long, no. 12,379. See p. 398. There are two strongly marked ato of Tephrosia spicata in southeastern Virginia. Assuming that the identification bY Torrey & Gray of Walter’s Galega spicata was correct and that the subsequent identifications of Small, Rydberg and others . Tephrosia or as Cracca) are correct, typical T. spicata 38 bie densely villous extreme extending from Florida to Louisiana 42 north to Tennessee and Delaware. It is common in southeas 1940] Fernald,—Additions to Flora of Virginia 457 Virginia, extending northward at least to Princess Anne and York Counties. In view of the fact that many segregates have been made from the traditional T. spicata, it is highly important carefully to check the type of Galega spicata Walt. Fl. Carol. 188 (1788) which had the wholly inconclusive diagnosis ‘“Spicis longis terminalibus.” That simple character belongs to several southern species. From its sparse pubescence extreme plants of Tephrosia spicata, var. semitonsa might be mistaken for T. hispidula (Michx). Pers., which is known from Virginia only through very old specimens without definite data. Although Small (Man.) separates T. hispidula (as Cracca hispidula) by ‘“Pubescence of bp: stem of appressed hairs” and Rydberg (in N. Am. Fl.) says “stem ... sparingly strigose or glabrate”, the type of Galega hispidula Michx. (as shown by a photograph) has some spread- ing hairs; so has the specimen from Virginia (old specimen of Torrey & Gray; Michaux said “in Virginia, Carolina et Georgia”) and some from South Carolina, Georgia and Florida. The clear- est distinctions between 7. hispidula and T. spicata, var. semi- tonsa are as follows: T. HISPIDULA. Stem, rachis and eet strigose-hispid ; principal . lets lanceolate, lance-oblong or oblanceolate, stri a neath pedicels filiform; upper calyx-lobes Halenaddanoaclicea: 15-3 m - SPICATA, Var. SEMITONSA. Stem, rachis and petiolules spreading-pilose or -villous; principal leaflets oblong-obov ate: appressed-pilose eath; pedicels stout; upper calyx-lobes lance-subulate, 2.5-3.5 mm. laa AESCHYNOMENE ViRGINICA (L.) BSP. To the stations recorded in 1939 add others (often very extensive) on fresh tidal shores in Kine Wiuiiam, New Kent, Cuartes Crry and James Crry Counties (many nos.). See pp. 385, 391, 393, 394, 400 and 402. *DESMODIUM GLABELLUM (Michx.) DC. Nansemonp County: white sand of pine and oak woods and clearings near Cathole Landing, west of Factory Hill, no. 11,356. A close match for the type of Michaux from South Carolina, our material and a fragment and photograph of Michaux’s type being the only representatives of the species I have seen. The more northern and larger-leaved plant passing as Desmodium glabellum is D. humifusum Beck. See p. 399. erm ANGusTIFOLIA (Pursh) Ell., forma subvelutina, ., caulibus calycibusque densissime subvelutino- Orn pilis cbpateheibeas: foliolis subtus dense subadpresso- pi losis.— 458 Rhodora [ NoveMBER With typical L. angustifolia or by itself, southern New Jersey to North Carolina. New Jerspy: sand, Cold Spring, Cape May County, August 30, 1917, Gershoy, no. 386, distrib. as L. angusti- folia & capitata; dry soil, Bennett, Cape May County, August 8, 1925, H. E. Stone, distrib. as L. oblongifolia. Virainta: sphag- nous bog about 1 mile northwest of Dahlia, Greensville County, August 20, 1938, Fernald & Long, no. 9077 (type in Herb. Gray; ISOTYPE in Herb. Phil. Acad.), growing mixed with typical L. angustifolia; sphagnous swale at head. of north fork of Mill Swamp, south of Emporia, August 18, 1939, Fernald & Long, no. 11,060. Nort Carotina: Bladensboro, September 3, 1908, Eggleston, no. 4043, in part (mixed with typical L. angustifolia). *STROPHOSTYLES UMBELLATA (Muhl.) Britton, forma ochro- leuca, forma nov., floribus ochroleucis. Greensville County, Vircin1a: open thickets, clearings and borders of woods east of Emporia, August 18, 1939, Fernald & Long, no. 11,065 (TYPE In Herb. Gray, isoryPE in Herb. Phil. Acad.) *GrRANIUM DissectUM L. James Crry County: cultivated field, 1 mile south of Williamsburg, Grimes, no. 3469 (reported as G. carolinianum). SourHamMpton County: roadside, Boy- kins, no. 10,310. Decidedly unusual in the Atlantic States. The only other material from the East in the Gray Herbarium is from Athens,. Georgia, but it is included in the Flora of the District of Colum- bia from College Park, Maryland. The species is widely dis- persed on the Pacific slope from southern British Columbia to California. *PoOLYGALA POLYGAMA Walt. SouTHAMPTON COUNTY: dry sandy pine woods near Nottoway River, above Carey Bridge, n°. 10,311; border of dry mixed woods by Applewhite’s Church, no. 10,312; rich wooded slopes and clearings by Three Creek, north of Applewhite’s Church, no. 12,122. See p. 362. We have never before met Polygala polygama on the Coastal Plain of Virginia, nor is it represented in the Gray Herbarium from between Montgomery County, Maryland (Great Falls and vicinity) and South Carolina. Wheelock in his monograph (Mem. Tor. Bot. Cl. ii. 140) cited nothing from Virginia and North Carolina. Our material is the true southern Polygala polygama of Walter, with very loose racemes, the flowers 4-6 mm, apart, 5-6 mm. long, on pedicels 1.5-3.5 mm. long, with the obovate wings much exceeding the capsule. The plant of Great Falls belongs to the wider-ranging and mostly northern val. 1940] Fernald,—Additions to Flora of Virginia 459 obtusata Chodat, with closer-flowered raceme, the flowers 1-4 mm. apart, 3-5(—6) mm. long, on pedicels 0.5—-2 mm. apart, the wings strongly rounded above and shorter than to exceeding the capsule. The latter extends northward to Nova Scotia, cen- tral Maine, central New Hampshire, Vermont, New York, south- ern Ontario, Michigan, Wisconsin and Minnesota. I have seen typical P. polygama only from Florida, Georgia, South Carolina and southeastern Virginia. PoLYGALA veERTICILLATA L. To the station in Greensville County, noted in 1938, add one in SourHampTon County: argil- nemeee clearing near Blackwater River, southeast of Ivor, no. In Ruopora, xl. 337, pl. 501 (1938), I held as true Polygala verticillata the plant which Linnaeus actually had in his own herbarium prior to 1753 and which closely matches his diagnosis and notes; but Pennell still urges!® (if I understand his long argument) that the species could as well rest on earlier specimens and concepts not so clearly identifiable as familiar to Linnaeus. If our identifications of the vague and mostly too inclusive Amer- ican species of Linnaeus are to be determined by psychoanalysis of what one supposes Linnaeus to have thought, or by the speci- mens less known to him and least matching his mature notes, rather than by the specimens actually before him in preparing Species Plantarum and which better match his original diagnoses and notes, we might as well give up. Different psychoanalysts and interpreters of the vague past rarely draw identical con- clusions from identical data. Similarly, no two modern botanists are likely to agree as to just what passed in the brain of Lin- naeus a couple of centuries ago. Interpretation of his species is difficult enough even when the inadequate specimens before him are studied. Whenever possible we should rest our conclusions on the clearest, rather than the vaguest evidence. Fe VERTICILLATA, var. IsocycLA Fern. To the few recorded sta- tons add the following. Sussex County: rich woods and bushy rating just east of the “fall-line” along Nottoway River, r ouble Bridge, about 6 miles northwest of Jarratt, no. 11,070; ty old field and border of woods, near Nottoway River, south- ,ast of Stony Creek, no. 12,392. Sournampton County: wooded ttomland, Meherrin River, southeast of Branchville, no. 10,313. “Pennell in Ruzopora, xli, 378-384 (1999). 460 Rhodora [NovEMBER P. vERTICILLATA, var. AMBIGUA (Nutt.) Wood. To the stations recorded in 1938 add the following. Sussex County: border of wooded swamp north of Stony Creek, no. 10,701. In the article above cited I referred to Polygala verticillata as a polymorphous species. Experience with it in the field for more than 50 years makes this polymorphism quite evident to me; and Mr. Long’s and my experiences in the field in eastern Vir- ginia and elsewhere show that the separation of adjacent colonies into varieties (to say nothing of “species”) is somewhat arbitrary. Plants referable to var. ambigua can often be separated from others called true P. verticillata only by a careful consideration of the degree to which intergradation can be discounted. It is, therefore, surprising to find Pennell, in his discussion of 1939, saying “These three species in constancy of characters, lack of intergradation, and differing areas of occurrence seem to me amply distinct specifically. After a long probation Polygala ambigua is now generally so recognized” (p. 378). The surprise is not that Pennell considers them three species; it is at the statement that “Polygala ambigua is now generally so recog- nized” [as “amply distinct specifically”] for this implies 4 uni- versality of judgment which is hardly consistent with recent practice. It is true that authors immediately following Nuttall and with inadequate material so treated it. It is also, as would be expected, true that it is maintained by Britton and Small, as well as by Pennell. But it will be admitted that all of these authors have viewed the genus from a somewhat local stand- point. It may be equally said that, in his Manual, Asa Gray was viewing the genus locally, but by his 3d edition (1862) he had become skeptical, saying of P. ambigua “nearly as 2 No. 9 [P. verticillata] (of which it is probably a mere variety) and later on he flatly so treated it. Chodat’s M onographia Polygalacearum (1891) can hardly be called a provincial treat- ment. Chodat viewed all the known species of the world; yet he saw nothing specific about the characters of P. ambigua. He treated it as P. verticillata, var. ambigua. Similarly Blake s monograph of all the species of Polygala in North America 18 broad-gauge in its specific concept and area; yet here (N. Am. Fl. xxv. 348) P. ambigua is treated only as a variety of P. verticillata. The counting up of specialists to make 4 total by 1940] Fernald,—Additions to Flora of Virginia 461 whom P. ambigua is generally recognized as a species is sub- ject to the factors which affect many other statistics. Since my own experiences have abundantly shown me that P. ambigua and P. verticillata frequently intergrade, I do not find myself in the “general” group referred to by Pennell, which excluded Asa Gray (in his more mature years), Chodat and Blake. *ACALYPHA OSTRYAEFOLIA Riddell. Dinwippre County: steep weedy bank by freight-yard of Norfolk & Western Railroad, Petersburg, no. 10,703. Although stated to occur from New Jersey to Florida and westward, there is no other material in the Gray Herbarium from the Atlantic states north of South Carolina. Acalypha caroliniana Ell., not Walt., is referred to A. ostryaefolia. Elliott knew it only from Paris (or Parris) Island in Beaters County, South Carolina “in cultivated land .. . very rare”. Messrs. Godfrey and Tryon found it as a weed about Gecrastown, South Carolina, in 1939. See p. 369. EvpHorsia potycontroua L. New Kent County: sandy beach of York River, near mouth of Fillbate’s Creek, north of Holly Forks, no. 11 1,584. Inland extension from the sandy coast. See p. 400. E. prosrrata Ait. Dixwippre County: cinders of freight-yard of Norfolk & Western Railroad, Petersburg, no. 10,704; road- sides and waste places, Petersburg, no. 11,367. Previously found, as a street-weed in Williamsburg, by Grimes. €e p. 369. E. pentata Michx. Soni GrorcE County: cinders of freight- yard, Norfolk & Western R. R., east of Petersburg, no. 11,366. ENRICO County: ou places ‘and railroad ballast, Richmond, ~ 12,393. HETEROPHYLLA L. Dinwippie County: roadsides and waste places, Petersburg, no. 11,072. Not previously represented ai the Gray Herbarium from north of South Carolina. See p. Thex CORIACEA (Pursh) Ch Nikeaniond County: wet hig near lumber camp of Camp Lumber Co., Great Dismal wamp, southeast of Whiterabehc! School, nos. 10,718 and 11,587. Shrubs up to 3 m. high. The black drupes become pulpy and juicy when ripe and promptly drop. Branches heavily loaded 462 Rhodora [NoveMBER with young drupes in July had lost all but a few lingering ripe ones in October—an unusual character in the genus and one which has not been emphasized. See pp. 368 and 400. *CEANOTHUS AMERICANUS L., var. INTERMEDIUS (Pursh) Trel. SourHaMPpTon County: dry sand, pine barrens about 7 miles south of Franklin, no. 7527, NaNsemonp County: white sand of pine and oak woods and clearings near Cathole Landing, west of Factory Hill, nos. 11,370 and 12,130. Characteristic of these dry white sands. Trelease, in Gray's Synoptical Flora, gave the range: ‘Tennessee . . . and 8. Carolina . to Louisiana .. . and Florida”. Small (Man.) gives it, as C. intermedius, the range, “Fla. to La., Tenn. and Ga.” Al- though extending northward into the pinélands of southeastern Virginia and well known on the sands of Middle Cape Cod,” var. intermedius seems to be lacking between Virginia and southeastern Massachusetts. AMPELOPSIS ARBOREA (L.) Koehne. To the very few known stations in eastern Virginia add the following. SOUT County: rich sandy and loamy woods along Three Creek, north- west of Carey Bridge, no. 10,327. Istz or WicHT COUNTY: waste ground, near Lee’s Mill, no. 12,407. Princess ANNE ouNTY: damp sandy ocr Cedar Island, no. 12,406. 7 *Vitis Lasrusca L., subedentata, var. nov. (raB. 63 ), foliis subtus pene ag penta vix sublanatis, ae subedentatis humeris plerumque latis vix elongatis.—U0ast” Plain, southeastern New York to Virginia. New York: Fisher's Island, August 10-15, 1920, St. John, no. 2811. New ge roadside along creek, Tumersville, Gloucester Co., June 24, 1 “ge RK. ft. 5, Tiestrom, no. 2 a moist places, Pleasantville Octo 13: 23° — a) . 3 ey se2s ® iva) ~~ Le | S BS i) No er: © hot Ber a SUE Rie 2 Pg o~ 35'S 25 i & ine ao 2 Es gg oO bes) Se 5 4e ro. 28 > & ‘Aan , 1939, Fernald & Long, no. 11,074 (TYPE in Herb. Genet: sete of gum swamp, west of Pungo, Pri t County, May 6, 1935, Fernald & Griscom, no. 4454; wet pea y clearings in woods of Pinus serotina, south of Grassfield, N ye nic August 4 and 5, 1934, Fernald & Long, no. 4027; . 20, 1939, Fernald & Long, no. 11,073. See p. 390 19 Fernald, The Cape Cod Ceanothus, Ruopora, xxxii. 161 (1930). 1940] Fernald,—Additions to Flora of Virginia 463 Typical Vitis Labrusca, as shown by the Linnean type (PLATE 636, Fic. 1) is the wide-ranging vine with leaf-margin coarsely dentate and with more or less porrect lobes or shoulders on the fruiting portions of the branches. This is the common form from Maine to southern Michigan, south (southward chiefly in the Piedmont and mountains, though reaching the Coastal Plain in South Carolina) to Georgia and Tennessee. The dense blanket of pubescence is relatively loose, the tangled hairs often distinctly showing under slight magnification (PLATE 636, Fic. 5). The Coastal Plain var. swbedentata has the margins of leaves ac- companying inflorescences with only obsolescent teeth, the sub- uli at the ends of the stronger veins relatively short, the shoulders usually poorly developed and rounded or broad and subhorizon- tal, and the dense felt of the lower surface very close and fine, its component hairs scarcely discernible under slight magnifica- tion (PLATE 637, FIG. 3). In fact the lower surface glistens as if varnished and in pressing it leaves a heavy brown varnish-like stain on the specimen-sheets. In Virginia var. swhedentata ma- tures and drops its fruit by late August. 636 shows, as Fig. 1, a portion of the type, X 1, of Vitis ptr Soa from . photograph received from Mr. Savage. Fics. 2-4 are leaf-marg X 1, from different specimens: Fic. 2 from Bedford County, Virginia, ae 20, 1871, A. H. Curtiss; rig. 3 from Chilmark, Martha’s Vineyard, Massa- on. In PLATE 637, FIG. 1 is a portion of the type, X 1, of var. subedentata. ys 2 is a leaf, x 1, fro m Chesapeake City, Maryland, Pe allabie no. 1,548; ric. 3, pubescence of lower surface, X 10, from the T “Sip inflexa, sp. nov. (TAB. 638, et TAB. 639, F 1-3), per his; Ghele minute stellato-puberulo 0.6-1.2 m. ita, his laxe a scendentibus: foliis lanceolato- vel lineari- oblongis breviter petio- latis, primarii iis 2.5-6 em. longis, 0.4-2 em. latis pagina superiore viridi sparsissime stallatocetr ones vel glabrata pagina evict pallida remote stellato-puberula, margine adpresso-serratis v Porrecto-dentatis; floribus plerumque corymbosis Ceanaibus et ad ramorum apices vel rare axillaribus pedunculis ad 1.7 em. longis; calycibus 6-10 mm. longis basi plus minusve villoso-hirsu- tis, lobis .deltoideo-aeuminatis; petalis late obovatis inaequaliter obeo rdatis 1.5 . longis flavescentibus basin — valde Striatis ; carpellis ¢¢ ca. 10, apice valde inflexis, matur 5 mm. altis horizontaliter ¢ costato-reticulatis dorso viniditnis hispidisque apice valde incurvatis plus minusve bidentatis, dentibus brevibus adscendentibus .—Southeastern Virginia, west-central Tennessee, 464 Rhodora [ NoveMBER Alabama and southeastern Missouri. Vircrn1a: Petersburg, Din- widdie County, Twomey; dry pine woods just east of the “fall- line’, along Nottoway River, Double Bridge, about 6 miles north- west of Jarratt, Sussex County, August 18 and September 21, 1939, Fernald & Long, nos. 11,076 and 11,372; border of sandy woods near Three Creek, northwest of Carey Bridge, Southampton County, August 19, 1939, Fernald & Long, no. 11,077, September 23, 1939, Fernald & Long, no. 11,373 (rypr in Gray Herb.; ISOTYPE in Herb. Phil. Acad.). TENNESSEE: cedar barrens, Lavergne, Rutherford County, May and September, 1882, Gat- tinger. AwaBAMA: locality not stated, Buckley. Missouri: Stoddard County, September 12, 1893, Bush, no. 5; sands, Camp- bell (Stoddard County), September 9, 1910, Bush, nos. 6293 and 6293A. Sida inflexa has been passing as S. Elliottii Torr. & Gray (see p. 382). That species rests primarily upon S. gracilis Ell. Sk. i. 159 (1816), not Richard. It was beautifully described by Elliott, from “the Sea Islands. Common about Beaufort” with “Stem ... glabrous; leaves linear . . .; peduncles solitary, axillary . - -; capsules (10) two horned, glabrous.” It is the very narrow- leaved plant occurring from southeastern South Carolina to southern Florida and Alabama, with nearly glabrous (often bushy-branched) stems 1.5-8 dm. high; linear cauline leaves mostly 1.5-5 em. long and 1.5-7 mm. broad; flowers mostly solitary in the axils and on peduncles up to 2.5 cm. long; calyx at most strigose on the ribs at base; carpels (PLATE 639, ric. 4) with prominent erect teeth, glabrous or nearly so on the back and with relatively weak cross-ribs on the sides. Sida inflexa is also related to the nearly glabrous S. rubro- marginata Nash in Bull. Torr. Bot. Cl. xxiii. 102 (1896) of Florida. In foliage the two are similar, but with quite different toothing. S. inflexa has the flowers chiefly in terminal corymbs, S. rubro-marginata axillary. §. inflera has the calyx (PL. 638, ric. 4) villous-hirsute on the ribs at base, S. rubro-marginata (pL. 639, Fic. 5) not; and the long erect teeth and quite different reticulation of the carpels (pL. 639, ric. 6) of the latter species show that S. inflera can hardly be forced into it, even by those most modern of young taxonomists who are so frequently main- taining that characters of the carpels, achenes and spores are unimportant as compared with shape of the foliage. 1940] Fernald,—Additions to Flora of Virginia 465 Sida inflexa is also somewhat related to S. neo-mezxicana Gray and to S. Lindheimeri Engelm. & Gray, especially in its stellate- puberulent stems. S. neo-mexicana, however, is a lower plant, with cinereous lower leaf-surfaces and calyx, the latter much lower than in S. inflera, the petals short, and the muticous car- pels (PL. 639, Fic. 7) without the transverse ribs which are so prominent in S. inflera. S. Lindheimeri has the flowers chiefly on long axillary peduncles, the very large calyx cinereous, the carpels (PL. 639, Fic. 8) with erect cinereous teeth and with obliquely ascending elongate reticulation. S. inflera in its less cinereous pubescence, its terminal corymbs, villous-hirsute base of calyx, and horizontally costate carpels with short (or no) hispid teeth is quite distinct from either S. neo-mexicana or S. Lindheimeri, It is noteworthy that all these species have at one time or another been included under Sida Elliottii. The old specimen of S. inflera from Petersburg, Virginia, and Bush’s plants of it from southern Missouri are the bases for the inclusion of S. Elliottii in Gray’s Manual, ed. 7; the Missouri and Tennessee material of S. inflera was cited in the Synoptical Flora as S. Elliottii. 8. neo-mexicana was placed in S. Elliott in Gray’s Plantae Wrightianae and in Torrey’s Botany of the Mexican Boundary; S. Lindheimeri was first treated as S. Elliottii, B terana Torr, & Gray; and S. rubro-marginata of Florida was — by Robinson in the Synoptical Flora, with the com- ment: “S. rubro-marginata . . . appears to be merely a broad- ae form of S. Elliottii.” Abcadatt material now at hand indicates its specific distinction from the latter. See p. 382. s oar : rom ; all from ee: aca gi In PLA , Fics. 1-3 are details of rh : Fig. 1, a flower, 2 porters reinforced by es). ti Nottoway River, "Doubis Bride, oe. Fernald & Long, ; had Fics. 2 and 3, carpels, x 10, from the Type. In PLATE 639 Elio hing figures are details of related ines: Fic. 4, caspel, x 10, Aas s ottii Torr. & Gray from iami, Florida, H. Curtiss Ts and ring of carpels, x 4, from onde of S. rubro-marginatt fash, ™Mpa, Florida, N ash, no. 2472; c. 6, rip carpel, 4 a of S. rubro- : o, Wrg Tac ripe cape, we 10, of S. Touihemen Howat: & Gray, fro he TYPE, 1843, L heimer, no. 24 466 Rhodora [NoveMBER Hrpiscus MILiTARIs Cay. Local range extended to New Kent County: bottomland woods by Chickahominy River, north of Long Bridge, southeast of Quinton, no. 11, STewarTIA MaLacHopenpron L. To the few known Virginian stations add one in Norro.K County: dry woods of a “ham- mock”, Great Dismal Swamp, west of Yadkin, nos. 11,078 and 12,131, _ tree-like shrubs up to 6 m. high. *Hyp UM MUTILUM L., var. LATISEPALUM Fernald. K1nc Wittn ae fresh tidal shore of Mattaponi River, at Horse Landing, near King William Courthouse, no. 11, Heretofore known only from Florida to Texas. See p. “402. H. Drummonpu (Grev. & Hook.) T. & G. GREENSVILLE County: border of fs ephones argillaceous field, northwest of Taylor’s Millpond, no. 11,080; open argillaceous border of wood- road northeast of Caskiie no. 11,081. The plump capsules burst under slight pressure, pushing out their masses of ripe and unripe seeds. The colored children, knowing the plant as “Nits and Lice”, demonstrated this fea- ture to us. They repudiated the very common and slender- fruited H. gentianoides as a member of the same group because it had “no lice”, the capsules being soft and unresponsive. The fact that they so sharply differentiated H. Drummondu indicates that it is less uncommon than we had supposed. King William Courthouse, no. 11,588. James City © tidal mud along Powhatan Creek, north of Jamestown no. 11,083. See pp. 386 and 401 and map 18. Extension south from the estuary of the Delaware. *VIOLA TRILOBA Schwein., var. pi~aATATA (EIl.) SouTHAMPTON County: rich sandy loam of woods by a River, northeast of Ivor, no. 10,735; rich mixed and deciduo woods near Nottoway River, above Carey Bridge, no. 10,336. First northeast of western North Carolina. See p. 363. ne in V. EscuLENTA Ell. To the two stations recorded add 0 itch NANSEMOND County: abundant and very large, ay Camp bordering wet woods and clearings near ‘Tomber cam rsh Lumber Co., Great Dismal Swamp, southeast of Whitema School, no. i O91. See p. 400. erd. V. EMARGINATA (Nutt.) LeConte, var, ACUTILOBA Brain oe: To the few recorded stations add one in SovurHAMPTON C00 337. border of dry mixed woods by Applewnite: s Chattels no Island, Brainerd. r 1940] Fernald,—Additions to Flora of Virginia 467 V. striata Ait. Prince Grorce County: rich wooded slope by James River, Indian Point, nos. 11,087 and 11,874. Our first station on the Coastal Plain of a characteristically inland species. See p. 3 AmMaNNiIA Korunel Britton. To the single oe Virgin- ian station (on York River) add the following in Princess ANNE ouNTy: brackish to fresh marsh along Back Bay, at gio margin of Long Island, nos. 10 ,741 and 11,088; similar habitat, Cedar Island, no. 12 416. See p. 370 and Map 4. GAURA BIENNIS L. Prince Grorce County: sandy shore of James River, City Point, no. 11,094. Our first Coastal Plain station. Eryncium aquaticum L. Very characteristic of fresh tidal shores and marshes at pa from the James River to the Matta- Poni (many nos.). See a SANICULA CANADENSI oe ar. grandis, var. nov., a var. genuina recedit foliis amplioribus ve fructibus crassioribus ; folii petiolati imi foliolis longioribus 5.5-13 cm. longis, folii subsessilis, imi foliolis longioribus 4.5—12 em. longis; fructuum triadibus ar 5) tae latis—Rich woodlands, western Vermont to Nebrasks, is © North Carolina, Tennessee, Missouri, Oklahoma and Tex Toes fr from Bristol, Vermont, July 25, 1898, Ezra geri (in er ay.). Seni canadensis consists of three strongly marked varieties, as shaw S. CANADENSIS L., var. genuina. S. canadensis L. Sp. Pl. 235 (1753) in_ part, as interpreted by Gray, Bicknell and later ais __ Larger leaflets of the wail developed petioled leaves long, 1.5-4(-5) em. broad; larger leaflets of lower subsessile leaves (at lower fork of s tem) 3-7 cm. long; triads of fruits (including tips of bristles) 7-9 mm. broad—Open woods, Florida to Texas, north to Rockingham County, New Hampshire, P lymouth, Bristol and Dukes Cos., Massachusetts, central and Southern rm Connecticut, Long Island, New Jersey, Pennsylvania, West Virginia, Ohio, Kentucky, Missouri and Okla This is the common plant of eastern Virginia. i Ss Fernald (supra). Var. typica H. Wolff in Engler, Pianzenr iv?28_ 67 (1913), not S. canadensis L., as shown 7s ts known to and cited by Linnaeus. Larger leaflets of well devine petioled leaves 5.5-13 cm. long, 2.5-6(-8) cm. Toad; larger vr of lower subsessile leaves (at lower fork of stem) 4.5-12 em. long; triads of fruits 1-1.5 em. broad—Of road acer eae (see above). 8 468 Rhodora [ NovEMBER Our only Coastal Plain stations in Virginia are along the James. Prince Grorce County: wooded bank of James River, City Point, no. 10,343 (plants up to 1.7 m. high). Surry County: rich calcareous wooded slopes along James River, Claremont Wharf, no. 10,344. Var. rLormpANa (Bickn.) H. Wolff, 1. c. 67 (1913). S. floridana Bickn. in Bull. Torr. Bot. Cl. xxiv. 581 (1897). Smaller through- out; the abruptly cuneate small leaflets with firm spinulose teeth; larger leaflets of petioled leaves 2-5 (rarely —8) cm. long; triads f fruit 5-7 mm. broad.—Dry sandy woods, Florida to Missis- sippi, north to southeastern Virginia. It would be perfectly possible to make a rational argument that the species known as Sanicula canadensis L., a southern species unknown in Canada, should be called S. marilandica L.; and that we should further reverse the established usage and take up for the more boreal and transcontinental S. marilandica of all recent authors the name S. canadensis. Such a complete reversal has more than the geographic argument in its favor. On the other hand, it is so evident that Linnaeus was utterly at sea regarding the real characters of the two and so confused the two elements in his writings and memoranda that it cannot be said that his own herbarium and the specimens he saw give un- equivocal support to such a reversal. The situation, under the two names, is as follows. S. canapEnsis L. The pre-Linnean references are to a Clay- ton plant of eastern Virginia described by Gronovius and sup- posed by him to be Sanicula canadensis, amplissimo lacimiato folio of Tournefort. Tournefort gave no further information ; consequently the only pre-Linnean reference of value is that of Gronovius. This account says “Pedunculi infra bifurcaturam caulis longi”, thus emphasizing the usual 2 long rays or branches so characteristic of the plant with divaricate and bifurcate 1n- florescence, the S. canadensis of Bicknell, Britton and Robinson & Fernald. But in his own herbarium Linnaeus had a Clayton specimen from Virginia (part of Clayton’s much confused no. 28—see below) which Linnaeus definitely marked S. canadensts. This is the northern large species, the S. marilandica sensu Bick- nell et al., and it was presumably what Linnaeus had in mind when he gave the simple comparative phrase of original diag- eee eee ees loess 1940] Fernald,—Additions to Flora of Virginia 469 noses: “Structura ita praecedentis [S. europaea] .. . sed planta decuplo saepe omnibus partibus major.” Asa Gray, in his manuscript notes on the Linnean Herbarium, specifically stated that the Linnean specimen of S. canadensis has long-exserted styles; and this observation was verified by Mr. Long and me in 1930. At that time, however, another Gronovian (Clayton) plant in the Clayton (Gronovian) herbarium was found to be short-styled S. canadensis sensu Bicknell et al. On the portion of the long-styled Clayton material of no. 28 (S. marilandica of recent authors) retained in the Gronovian herbarium occurs the original label: Sanicle. D. Clayton An. 1734. Num. 28. Claython ex Virginia an 1734. Num. 28. Lappula fere umbellata Astrantiae foliis virginiana. Plukn. Mant. 114. This was subsequently marked in a second (perhaps clerical) and: “Sanicula flosculis masculinis pedunculatis, hermaphro- ditis sessilibus flor. Virg. p. 31”, this phrase being the diagnosis given by Gronovius, FI. Virg. 31, for no. 28, which Linnaeus cited as his basis of S. marilandica. This Clayton material with long styles was marked by Asa Gray: “The type of Marilandica. A. G.” “The greater part of this was given to Linnaeus & he has wrongly named it : Canadensis. A. G.” However, there is another sheet of Clayton’s no. 28 in the Gronovian herbarium, Mr. Long and I studied it in 1930 and Professor Alfred Rehder then kindly photographed it. This specimen is appropriately discussed under - MARILANDICA L. Linnaeus referred to 3 sources and gave no hew diagnoses: (1) to Gronovius, p. 31 (7. e. Clayton’s no. 28, mM part at least); (2) to Hortus Upsaliensis; (3) to Ray. Hortus Upsaliensis referred back to the same accounts by Gronovius and by Ray, with 2 other references which were not cited by Linnaeus in Species Plantarum. Gronovius gave the brief diagnosis, “Sanicula flosculis masculinis pedunculatis; hermaphroditis sessilibus” and the citation, “Sanicula sylvatica floribus albis, foliis triscuspidatis. Clayt. n. 28”; also the teference to Ray which was later cited by Linnaeus. _ Ray’s account was detailed. It emphasized the small (min- ‘mis) fruit, the divaricate and dichotomous branching, the short (Pollicares) , simple peduncle in the forks, and the pair of leaves 470 Rhodora [NovEMBER at the forks; all very striking characters of S. canadensis sensu Bicknell et al., not of S. marilandica of Bicknell et al. One sheet of Clayton’s no. 28 has already been discussed. The other bears in the hand of Solander, apparently, the full text from Gronovius, p. 31, which referred to no. 28. This second sheet is perfectly characteristic of the bifurcate, divaricate-branched, small-leaved and small-fruited, short-styled plant (S. canaden- sis sensu Bicknell) which abounds in eastern Virginia. It is, thus, quite unlike the other sheet of no. 28, which bears the original annotation by Gronovius. Since the names Sanicula canadensis and S. marilandica, as used by Linnaeus, were hopelessly confused, as were the two species as treated by all authors up to Bicknell (1895), we should not lay too much stress upon attempts to retrace the vague mental processes underlying the confusion. Gray de- cided that the long-styled specimen of no. 28 should stand as S. marilandica, the short-styled specimen as S. canadensis. Bick- nell, the first monographer to clarify the formerly tangled group, followed Gray; and Wolff has followed their interpretations. These decisions should stand. Nothing would be gained by re- typifying the species on the basis of early misconceptions. *TorILIS JAPONICUS (Houtt.) DC. Prince Grorce County: waste ground near wharf, City Point, no. 10,354. Cited by Coulter & Rose from Baltimore and joreipieg *LILAEOPSIS CAROLINENSIS Coult. & Rose. Princess ANNE County: forming an extensive mat in shallow pool in 5 breekieh to ta along Back Bay, at eastern margin of Long Island, no The fourth known station in North America and the first north of southeastern North Carolina; species bicentric, its second area on the drainage-system of La Plata River in eastern South America. See p. 371 and map 6. L. cHInEeNsIs (L.) Ktze. To the few recorded stations sen the following. CHartes Crry County: sandy tidal shore of James Rivers at “Four Oaks”, below Harrison Point, no. 11,388. New Kent County: san dy tidal shore of York River, near ay a Fillbate’s Creek, ie of Holly Forks, no. 11,597. > ioncebslltaa CANESCENS (Michx.) G. Don. Kine WILLIAM County: steep wooded bank of Mattaponi River, at Horse 1940] Fernald,—Additions to Flora of Virginia 471 Landing, near King William Courthouse, no. 11,602 (shrubs up to3m. high). Sussex County: wooded bottomland, J ones Hole Swamp, west of Coddyshore, no. 10,363 (shrubs 3 m. high) ; rich woods and bushy clearing just east of the “fall-line”, along Not- toway River, Double Bridge, about 6 miles northwest of Jarratt, no. 11,099. SourHampron County: steep wooded slopes by Three Creek, northwest of Applewhite’s Church, no. 10,364 (strag- gling, 0.5 m. high), 10,365 and 11,882 (erect, 1.5 m. high); on steep slope at Round Gut, southwest of Franklin, no. 11,391; rich woods near Raccoon Creek, north of Mill Neck Church, no. 12,427. Nansemonp County: damp thicket, steep bank o branch entering Blackwater River, northwest of Duck’s Store, no. 12,428. See pp. 361 and 363. Rehder in Wilson & Rehder, Mon. Azaleas, 144 (1921) gave the range: “from southwestern Tennessee and southern central North Carolina to eastern South Carolina and northeastern Flor- ida to extreme southeastern Texas”, etc. i LATIFOLIA L., var. LAEVIPES Fernald in Ruopora, 53 40). With or apart from typical K. latifolia in the southeastern counties, *ZENOBIA PULVERULENTA (Bartram) Pollard. NanseMonp County: Chamaecyparis swamp in sandy and peaty pine bar- rens northeast of Sandy Landing, south of South Quay, nos. 12,149 and 12,150. Passing through transitional shrubs (such 48 no. 12,151) into the green-leaved and commoner - PULVERULENTA, forma nitida (Michx.), comb. nov. Andro- meda speciosa, var. a. nitida Michx. Fl. Bor.-Am. i. 256 (1803) ; Rehder in Bailey Cycl. Am. Hort. iv. 2007 (1902).—Leaves of fertile branches oval or elliptic-oblong, obtuse, shallowly crenate, Sreen both sides.—In Virginia known from two southern counties. NaNsEemonp County: swampy depressions and Chamaecyparis Swamps in pine barrens, from northeast of Cox Landing to east of Sandy Landing, south of South Quay, nos. 11,102-11,104, dis- tributed as var. nuda, no. 12,152. SourHamptron County: Swampy woods southeast of Round Gut, southeast of Franklin, pe 11,395 (shrubs up to 3 m. high). Passing into yet PULVERULENTA, forma nuda (Ventenat), comb. nov. eromeda cassinefolia Ventenat, Descr. Jard. Cels, 60, t. 60 oes . - Cassinefolia, var. nuda Ventenat, Jard. Malmais. ii. i 804). Z. cassinefolia (Vent.) Pollard in Bull. Torr. Bot. Cl. Sch, 231 (1895). Andromeda pulverulenta, var. nuda (Vent.) venelder, Ill. Handb. Laubholzk. ii. 526 (1911). Z. pulverulenta, “t. nuda (Vent.) Rehder in Mitteil. Deutsch. Dendrol. Gesells. 472 Rhodora [ NoveMBER for 1915: 226 (1915) —Leaves lance-oblong to narrowly ovate, acute at both ends, prominently toothed, green both sides —In Virginia known only from Nansemonp County: with the other forms, Chamaecyparis swamp in sandy and peaty pine barrens northeast of Sandy Landing, south of South Quay, no. 12,154. In western Nansemond County the three forms of Zenobia are clearly confluent. Furthermore, among the very few sheets in the Gray Herbarium there are two sent from South Carolina by M. A. Curtis, one the green-leaved with elliptic-oval round- tipped blades, as Andromeda speciosa Michx., a. nitida, the other of the similar shrub with strongly whitened blunt leaves, as A. speciosa, B. pulverulenta and marked, “Mixed helter-skelter with the other”. Curtis’s experience in South Carolina was obvi- ously like ours in Virginia. As forms they are strikingly differ- ent, but they certainly are not geographic varieties, much less two species—the disposition of them by Small. See pp. 385 and 397 and map 16. The application of the names needs clarification. The first name in the species was apparently Andromeda pul- verulenta Bartram, Travels, pl. 8, opp. p. 476 (1791), accom- panying a crude but recognizable drawing of a flowering branch (crude as to outline and toothing of leaf), with an evident at- tempt to indicate pulverulence on foliage and branches. There was no word of diagnosis and the name cannot be taken up 45 adequately published by Bartram. Willdenow, however, Sp. Pi. ii’, 610 (1799) properly described the species under Bartram’s name, citing the plate, so that the species Andromeda pulveru- lenta properly dates from Willdenow’s adequate publication (Bartram ex Willd.) ; but, whereas Bartram’s original plate ac- companied the description of “The North West of Cape Fear; here at Ashwood [the old estate of Colonel William Bartram] ... near three hundred yards over... and... eighty or ninety miles above the capes”, therefore near the inner border of the Coastal Plain of southeastern North Carolina, Willdenow ren- dered its geographic source as “Habitat in Florida.” In 1800, Ventenat, Descr. Jard. Cels, 60, t. 60, described and illustrated Andromeda cassinefolia, “Feuilles . . . ovales, dentées et munie d’une glande au sommet de chaque dent, souvent aigués, quelquefois obtuses, glabres, . . . d’un verd foncé”, ete: 1940] Fernald,—Additions to Flora of Virginia 473 The plate shows narrow leaves mostly acute at each end and coarsely toothed, a relatively infrequent extreme, like our no. 12,154. But Ventenat, like Willdenow in case of Bartram’s shrub, took the easiest course and said “Arbrisseau découvert par Michaux dans la Florida”, although Michaux himself (FI. Bor.-Am.) cited his own collections as discovered “in Carolina septentrionali, circa Fayette-Ville et Wilmington.” Andromeda speciosa Michx. FI. Bor.-Am. i. 256, clearly de- scribed “A. foliis ovalibus, subrotundis, obtusis, crenatis ser- ratisve”, is the least rare form of the species, with green, obtuse or round-tipped relatively broad and low-crenate leaves (except on vigorous leaders). Michaux divided his A. speciosa into two varieties: var. a. nitida (the typical variety) from “circa Fayette-Ville et Wilmington” and var. “8. pulverulenta: Barr. Tamis, foliis floribusque pulvere albo inspersis; qui candor certo morbus est . . . in Carolinae utriusque stagnosis.” The name pulverulenta has been sufficiently considered, except that Ven- tenat, Jard. Malmais. ii. 79 (1804) treated it as A. cassinefolia Vent. (1800), var. pulverulenta. Michaux’s A. speciosa, a. nitida is the shrub with round-tipped or obtuse green leaves. Although the name var. nitida was with- cut diagnosis, Michaux obviously meant it for’his typical An- dromeda speciosa. It was taken in this sense by Rehder in 1902 and I am so considering it. Small and others, following Willdenow and Ventenat, who re- spectively ascribed Bartram’s locality on Cape Fear River, North Carolina, and Michaux’s stations, “in Carolina septentrionali, crea Fayette-Ville et Wilmington”, to “Florida”, state the range of Zenobia cassinefolia as “Pinelands, Coastal Plain, NE Fla. to N. ©.” and of typical Z. pulverulenta as “Ga. (or Fla.?) to MC” fa the herbarium of the New York Botanical Garden, where one would expect specimens from Florida (Small for many years collecting there), and in the Gray Herbarium the only material (until our Virginia collections) is from North and South Carolina, and at the former institution there is a letter to Dr. Small, stating that in the National Herbarium there is no ma- ‘nial from south of South Carolina. “Florida” may safely be dropped from the range. See map 16. 474 Rhodora [NoveMBER *VACCINIUM ore” Andr. (Herpothamnus Small). NANSEMOND _ nee Pg ie extensive carpets in the sandy or , 105; east of Cox erg nos. 10,774 and 12,160; south- ere of f Cox Landing, no. 11 foi Se ripe October 15, ‘purple- black, lustrous, soft, juicy, s rir bland) ; southeast of set Landing, no. 11,106; 1-1% Cae. south of Cherry Grove, no. 11,398. Istm or Wicut County: dry sandy pine barrens south of Lee’s Mill, no. 11,889. Extension north from southeastern North Carolina. See p. 79 and map 12 PYXIDANTHERA BARBULATA Michx. To the single station re- F pine barrens south of Lee’s Mill, no. 11,893. _NANSEMOND OUNTY: very abundant, with Vaccinium crassifolium, in pine barrens south of South Gung: nos. 10,775, 10,776, and 11,716; rare in pineland southwest of Marsh Hill School, no. 11, 107. See p. *LYSIMACHIA PRODUCTA (Gray) Fernald. SouTHAMPTON County: bushy swales and borders of swampy woods near Black- water River, Cobb’s Wharf, no. 10,382. Our first material from south of the District of Columbia, except from western North Carolina. See p. 364. L. (§ Steironema) rapicans Hook. New Kent County: bot- tomland woods by Chickahominy River north of Long Bridge, southeast of Quinton, no. 11,402. SourHampTon County: about spring-heads bordering alluvial wooded bottomland of peeeeed River, Cypress Bridge, nos. 10,778, 10,779, 11,109 and 11,40 In Ruopora, xxxix. 438 (1937), I recorded the Cypress Bridge station and expressed some doubt as to the exact identity. We watched the plant from anthesis through prolonged drowning to maturity, when it set no fruit; there is now no doubt of its identity with the plant of the AGtidiennio! and Gulf drainage. See pP- 366 and 397 and map 3. *SABATIA ANGULARIS (L.) Pursh, forma cleistantha, f. nOV-, eorollis minutis tubulosis clausis pallidis. —Greensville County, VirGINIA: argillaceous and sphagnous meadow northwest - Taylor’s Millpond, August 29, 1939, Fernald & gees no. 11,11 (TYPE in Herb. Gra ; d S. AMOENA (Raf. ) G. Don. To the few recorded stations “om another in Princess ANNE County: brackish to fresh mars 1940] Fernald,—Additions to Flora of Virginia 475 along Back Bay, at eastern margin of Long Island, no. 11,113. See p. 389. nee aquaTicum (Walt.) Ktze. Princess ANNE County: shallow water near margin of Salt Pond, no. 10,788, station shown us by the eye Sally Ryan and Mary L el igh. RA oe AMERICANA IL., var. MICROCARPA Gray. SUSSEX County: wooded bottomed, Baath River, southwest of Homovitie no. 10,389. See p. 3 Although Small speaks of the small-fruited extreme as having no seeds, our material seems normal. Its fruits are not so small as in the original material from Alabama nor as in Small’s col- lection from Smyth County, Virginia, but decidedly smaller than in the regular run of F. americana. F. PENNSYLVANICA Marsh., var. AUSTINI Fernald in Ruopora, xl. 452, pl. 529, figs. 1 and 2 (1939). CHarutes Ciry County: wooded sandy margin of James River, Wilcox Wharf, no. 11,110. Our first station on the Coastal Plain of Virginia. NSYLVANICA, var. LANCEOLATA (Borkh.) Sarg. GREENS- VILLE Count: alluvial woods along Meherrin River, Emporia, UO, 23,111. Daa first station on the Coastal Plain of Virginia. “APOCYNUM sIBIRICUM Jacq. A. hypericifolium Ait. See Fernald in RHODORA, XxXxvVili. 397, 328 (1935). CHaRLEs City ihe sandy beach of James River, southeast of Tettington, 0. 11,405. See p.3 Seem in his eeu of the genus, Ann. Mo. Bot. Gard. xvii. 137 (1980), cited nothing from south of Delaware and the District of Columbia. “AMSONIA TABERNAEMONTANA Walt. SouTHAMPTON CouNTY: rich mere and deciduous woods near Nottoway River, above Carey Bridge, nos. 10,790 and 11,895. Gimariiie County Wooded bottomland of Foutaius Creek southeast of Taylor's Millpond, no. 10,394. See p. 359. Woodson, in his monograph of Amsonia, Ann. Mo. Bot. Gard. XV. 405-407 (1928), cited A. Tabernaemontana as only escaped from cultivation northeast of South Carolina. Along the Notto- way and Fontaine Creek it is a part of the strictly indigenous flora. The only Virginian material seen by Woodson was from Petersburg “data lacking” and referred by him to var. salici- folia (Pursh) Woodson. The variety abounds in rich woods 476 Rhodora [NoveMBER and clearings along Appomattox River, slightly above the “fall- line” about 2 miles west of Petersburg, no. 11,896. ACERATES FLORIDANA (Lam.) Hitche. To the single known station in Sussex County (now under the plow) add one (now swale by Southern iad = eeace - Emporia, no. 11, LANCEOLATA Walt. cess ANNE County: brackish to fresh end along Back Bass Pellitory Point, north- east of Munden, no. 11,117. As pointed out by me in Ruopora, xxxvii. 438 (1935), the plant of Princess Anne and Norfolk Counties is mostly var. paupercula (Michx.) Fernald, with linear leaves. The Pellitory Point station (rather extensive) is the first known to us between northeastern North Carolina and Delaware. See p. 387. A. PurRPURASCENS L. To the single Station in Nansemond County add one in Dinwwwpre County: rich deciduous woods about old marl-pits east of Burgess Station, no. 10,3 BREWERIA HUMISTRATA (Walt.) Gray. Local range ‘extended northward and eastward. New Kent County: dry clearing south of Providence Forge, no. 11,407. Surry County: dry thicket north of Surry Courthouse, no. 10,794. NANSEMOND County: dry sandy woods at Cox Landing, south of South Quay, no. 11,408. Ipomora HEDERACEA Jacq., var. INTEGRIUSCULA Gray. To the station in Princess Anne County recorded in 1935 add one in ie County: roadside fencerow west of Franklin, no. Puiox Henran Nutt. in Journ. Acad. Nat. Sci. be ibe 110 (1834). P. nivalis sensu Wherry in Bartonia, no. 11: 8 (1929); . * (18 In my earlier papers, without looking carefully into the matter, I erroneously took up the name Phlox nivalis Loddiges, in def- erence to the usage of Dr. Wherry, 1. c., and in later papers- When the original source is consulted, however, no adequate description is found. Loddiges showed a branch with a mass of white corollas, the tubes barely exserted from the calyx, and his text was as follows: 1940] Fernald,—Additions to Flora of Virginia 477 No. 780 PHLOX NIVALIS. Class. Order. PENTANDRIA MONOGYNIA. This is a native of Carolina and Georgia. It was sent to us by our excellent friend Dr. Wray, of Augusta, and flowered beautifully the last spring. We think it so different from both subulata and setacea, that it may well form another species. It is probably not quite hardy: we preserved it very well in a cold frame, and have increased it by cuttings. The soil should be light loam, with a little peat mixed. Like the setacea, it appears to be partly shrubby. That, of course, is not a diagnosis and, unfortunately, the Loddiges plate (which shows no analyses of the flowers, which would admit it under the International Rules) shows corolla- tubes barely exserted from the calyx, one of the few flowers (at the left) which show the corolla-tube with it exceeding the calyx by only 2mm. The many sheets in the Gray Herbarium checked by Dr. Wherry as “Phlox nivalis” have the long-exserted corolla-tube exceeding the calyx by 7-10 mm. Nuttall’s ac- count of P. Hentzii had the required diagnosis and a clear dis- cussion of the plant. Its exact geographic source was not given, merely “Sent to the herbarium of the Academy by my friend Mr. Hentz”, the plant said to be a “common species in the southern Pine barrens.” One of Nuttall’s original specimens (part of the TYPE) in the Gray Herbarium bears in Nuttall’s hand the clari- fying data: “Chapel Hill, N. Carolina. N. M. Hentz”. Ma- terial of the magenta-colored form (the only one known in southeastern Virginia) was collected near Chapel Hill on April 4, 1939, by A. S. Pease (no. 27,007). HyproLga quaprivatvis Walt. Local range extended east- co Princess ANNE County: margin of Stubby Lake, no. HeEtiorroprum curassavicum L. To the two stations already Tecorded add from Princess ANNE County: open mu n along Back Bay, Pellitory Point, northeast of Munden, no. ant a Similar habitat east of Creeds, no. 11,125. See pp. 369 ONosmoprum vircintanum (L.) A. DC. To the few recorded Stations add one in SourHAaMPTON CouNTY: dry sandy pine woods near Nottoway River, above Carey Bridge, no. 10,404. Sussmx 478 Rhodora [NovEMBER County: dry open sandy woods and thickets between vetigorce and ein Bridge, no. 12,179; sandy woods near Chub, n 12,448. See p. 362. VERBENA Hie SOE L. Drnwippre County: roadsides and waste places, Petersburg, no. 10,798. Henrico County: waste places and railroad ballast, Richmond, no. 12,451. Princess ANNE County: sandy clearing, Ragged Island, no. 12,452. Our first stations on the Coastal Plain of Virginia; apparently very local. V. scapra Vahl. To the two stations already recorded add two in Princess ANNE merece inner bo rder of brackish to fresh (T. stations in Isle of Wight Cmny recorded in 1937 add from SourHampton County: white sand of pine and oak woods north of Point Beach, south of Prank. no. 11,416; similar habitat, southeast of Wiggins School, no. 11,417. See p. 397. PYCNANTHEMUM VERTICILLATUM (Michx.) Pers. DiINwIDDIE OUNTY: springy sphagnous swale about 5 miles east of Burgess Station, no. 11,131. Heretofore known in the state only from the western aigae Associated in the swale with other localized species. See p. 3 Cuniza oricaNnowes (L.) Britton. Entering the Coastal Plain in Sussex County: rich woods and bushy clearing just east of the “fall-line” along Nottoway River, Double Bridge, about 6 miles northwest of Ja hee nos. 11 132 a nd 11,420. SourHAMPTON County: steep ooded slopes by Three Creek, northwest of Applewhite’s Chieeh : *HyYPTIS MUTABILIS “A. Haat Briq., var. spicata (Poit.) Briq. Dowdon County: roadsides and waste places, Peters- burg, no. 11,418. A tropical American species (variety chiefly West Indian) not previously reported from north of Florida. See p. 4 *PHYSALIS ANGULATA L. Princess ANNE County: borders of low woods and clearings along Back Bay, Long Island, nos. 10,811 and 11,137. 1940] Fernald,—Additions to Flora of Virginia 479 Although given a broad range, “Pa. to Minn. and southw.” by Robinson & Fernald in Gray, Man. ed. 7, there is no other in- digenous material in the Gray Herbarium from north of North Carolina and southern Illinois. In his monograph of the group, Mem. Torr. Bot. Cl. iv. 334 (1896), Rydberg was unable to cite specimens from north of North Carolina. See p. 371. P. pupescens L. Range extended westward into NansEMOND County: wood-road in swampy woods east of Milk Landing, south of South Quay, no. 10,810. *LyctuM CHINENSE Mill. Dinwippre County: roadsides and waste places, abundant and rapidly spreading, Petersburg, no. *CYMBALARIA MURALIS Gaertn., Mey. & Scherb. Dinwippie County: abundant on bank by railroad, Petersburg, no. 11,610. Pennell (Scroph. E. Temp. N. Am. 317) cites no material seen from Virginia. Kicks1a ELATINE (L.) Dumort. New Kent County: ditch at border of damp woods, near Fillbate’s Creek, north of Holly Forks, no. 11,611. Henrico County: waste places and railroad ballast, Richmond, no. 12,470. SourHampton County: cinders of freight-siding, Branchville, no. 10,414. C RA L., var. ELATIOR Raf. SouTHAMPTON County: with C. Cuthbertii Small, border of low woods south- west of Cypress Bridge, no. 11,488. A plant chiefly of the upland, here definitely on the Coastal lain. C. OBLIQUA L. SourHampton County: seeping wooded slope bordering bottomland of Three Creek, northwest of Carey Bridge, hos. 11,419 and 11,424. The only Virginian station (probably the original of Clayton) Own to Pennell, 1. c., is in Gloucester County. See pp. 363 and 399 C. Curuperrn Small. Local range extended eastward into NaNnsemonp County: wet bushy swale east of Cherry Grove, south of South Quay, nos. 11,138 and 11,614; wet peaty thicket M pine barrens, east of Cox Landing, south of South Quay, no. 11,423. See p, 384 ACOPA CYCLOPHYLLA Fernald in Ruopora, xli. 446 (1939) (Herpestis rotundifolia Gaertn. fil.). Kina Wiiuiam County: very scarce, fresh tidal shore of Mattaponi River, at Horse Land- 'ng, near King William Courthouse, no. 11,613. € eighth known station, connecting that at Wilmington, 480 Rhodora [NoveMBER North Carolina, with the two in eastern Maryland. See p. 402. and MAP 22. *GERARDIA RACEMULOSA Pennell. NanseMoND County: sandy and peaty pine barrens, east of Cox Landing, south of South Quay, nos. 11,145 and 11,433. Pennell, Scroph. E. Temp. N. A., map 115 (p. 434), indicates no station between the Eastern Shore of Maryland and south- eastern North Carolina. G. racemulosa seems like a pine-barren extreme of G. purpurea L. *UTRICULARIA BIFLORA Lam. Kina Winuiam County: fresh tidal shore of Mattaponi River, at Horse Landing, near King William Courthouse, no. 11,620. Cuartes Crry County: tidal water of Kimage’s Creek, Kimage’s, no. 11,438. Princess ANNE County: shallow pools in brackish to fresh marsh along Back Bay, east of Munden, no. 11,147; similar habitat, east of Creeds, no. 11,148. Sussex Counry: small sandy pond in woods north of Double Bridge, about 6 miles northwest of Jarratt, no. 11,439. There is no previous material in the herbarium of the New York Botanical Garden and in the Gray Herbarium from be- tween South Carolina and southern New England. See p. 387. U. riprosa Walt. Nansemonp County: seepy sandy and peaty open spots in sphagnous savannah-like swale east 0 Cherry Grove, south of South Quay, nos. 11,618 and 12,186. See p. 403. There is no previous material in the herbaria of the New York Botanical Garden and of the Philadelphia Academy nor in the Gray Herbarium from between southeastern North Caro- lina and southern Delaware and New Jersey. Mr. Lloyd G. Carr has reported it (Claytonia, iv. 23) from Augusta County. U. suncea Vahl. Local range extended to NANSEMOND County: seepy sandy and peaty open spots in sphagnous savan- nah-like swale east of Cherry Grove, south of South Quay, n°. 11,149. See p. 384. . VIRGATULA Barnhart. Local range extended to NANSEMONP County: with the last, no. 11,150. See p. 384. *RUELLIA sTREPENS L., forma CLEISTANTHA (Gray) 8. McCoy. Prince GrorGE County: wooded swamp by James River south of Indian Point, no. 11,152. *DiopIA TERES Walt., var. HIRSUTIOR Fern. & Grisc. PRINCESS Anne County: sandy fields, Long Island, no. 11,153. NaNSE- MonD County: dry white sand of pine barrens, east of 00x Landing, south of South Quay, no. 11,156. SourHAMPTON 1940] Fernald,—Additions to Flora of Virginia 481 County: border of sandy woods southeast of Round Gut, south- west of Franklin, no. 11,442. See p. 389 D. TERES, var. HYSTRICINA Fern. & Grisc. Essex County: sandy beach of Rappahannock River at Richmond Beach, south- east of Tappahannock, no. 11,621. Extension inland from coastal sands. See p. 402. *RICHARDIA scaBRA L. Dinwippre County: railroad cinders, scarce, Collier’s Yard, 3-4 miles southwest of Petersburg, no. 11,159. Sournampton County: weed in sandy field near Black- water River, Cobb’s Wharf, no. 11,160. Nanst : roadside bordering swampy woods north of Whitemarsh School, no. 11,161. A tropical American species, formerly known northward into North Carolina. See pp. 382 and 383. “EUPATORIUM TORTIFOLIUM Chapm. NANSEMOND County: dry white sand of pine barrens northeast of Sandy Landing, south of South Quay, no. 11,108; similar habitat, near Cathole Land- ing, west of Factory Hill, no. 11,448. Extension north from South Carolina. See pp. 384 and 399. ‘UHNIA EUPATORIOIDES L. To the few recorded Coastal Plain stations add one in SovrHaMPTON CoUNTY: dry hickory and oak ig north of Point Beach, south of Franklin, no. 11,453. See p. 398. *CARPHEPHORUS TOMENTOSUS (Michx.) T. & G., var. Walteri (Ell.), comb. nov. Liatris Walteri Ell. Sk. ii. 285 (1822), at least as to plant described. Istz or WicHT County: dry sandy Pine barrens south of Lee’s Mill, no. 12,486. NANSEMOND County: sandy and peaty pine barrens northeast of Sandy Land- ng, south of South Quay, no. 11,173. With typical pilose-leaved C. tomentosus and very distinct from it in its glabrous rosettes and only sparsely pubescent stems. Our plant is definitely what Elliott described from eastern South Carolina as Iiatris Walteri with “leaves lanceolate, acute, gla- brous, dotted, attenuate at base”, etc., though Elliott made the error of including Anonymos uniflorus Walt. Elliott’s note that ‘This plant appears to form an intermediate species between L. Bellidifolia and Tomentosa” is significant, but I find little to Place it near Carphephorus bellidifolius; its characters, except for the glabrous lower leaves, place it with C. tomentosus. The late Henry W. Ravenel sent it to Gray as Liatris Walteri from ‘ntee Canal and the late M. A. Curtis thus correctly identified 482 Rhodora [NoveMBER his material from the region of Wilmington, North Carolina. See p. 384. j Souipaco FisTuLosA Mill. Our most inland station is in SourH- AMPTON County: low woods, very scarce, near the pond, Wind- man’s Mill, south of Sunbeam, no. 11,464. From here eastward it becomes progressively more abundant. S. Exuiorrm T. & G. (typical). To the single Virginian sta- tion, in Henrico County, recorded in 1939, add the following, in NanseMonpD County: clearings and borders of wet woods north of Whitemarsh School, nos. 10,831 and 11,625; swampy depres- sions in pine barrens east of Cox Landing, south of South Quay, no. 10,832; sphagnous savannah-like swale east of Cherry Grove, south of South Quay, nos. 11,463 and 11,626. . ULMIFOLIA Muhl. Extending into the Coastal Plain in CuarLes City County: dry wooded bank of James River at “Four Oaks”, below Harrison Point, no. 11,461. A Synopsis or Bouronta (Puates 640-646).—In September, 1933, on my first trip to Virginia, Mr. Ludlow Griscom and I collected on the tidal marshes of North Landing River a Boltonia which did not readily work out by existing treatments of the group. In studying it we found other difficulties in the genus and then prepared a tentative outline of the more significant characters. The completion and publication of this study was delayed until the identity of some types, including those of Matricaria asteroides L., basinym of B. asteroides (L.) L’Hér., and of Chrysanthemum carolinianum Walt., referred by Gray to the synonymy of B. asteroides, could be established. The Walter type has not been found; but a sheet compared by Mr. C. A. Weatherby in October, 1935, and a photograph (our PL. 640, rics. 1 and 2) received from Mr. Savage in November of that year clearly settle that Matricaria asteroides L. is, as Mr. Griscom and I inferred from the Linnean diagnosis and the source of the type, “Pensylvania”, not the wide-ranging series with broad leafy corymbs, extending westward to [Illinois and beyond, as Gray inferred, but a local plant chiefly of the Atlantic States, with its chief concentration along the Susquehanna River in Pennsylvania and Maryland, though perhaps extending to northern Ohio, and known from western North Carolina; also with a geographic variety, the true B. glastifolia (Hill) L’Her. (our pL. 641), extending from southern New Jersey along the 1940] Fernald,—Additions to Flora of Virginia 483 coastal areas to Louisiana. Some other conclusions reached by Grisecom and me in the winter of 1933-34 are supported by additional collections; other conclusions are altered through new evidence. For instance, the commoner species of south- eastern Virginia (pL. 642), tall (up to 2.3 m. high), with small mostly white-rayed heads on the loosely paniculate branches, then not known to us, closely matches Walter’s account of his Chrysanthemum carolinianum from the region of Charleston; and, fortunately, a collection made by Mr. Robert K. Godfrey in Berkeley County, South Carolina, in September, 1939, is quite like the plant of southeastern Virginia. We are, therefore, safe in considering it Walter’s species, the type of which is presum- ably lost. In many points the outline prepared seven years ago is here adopted, with real regret that his other duties prevent my associate in the original study from continuing it at this time. The two coastwise species (B. asteroides and an undescribed one of southeastern Virginia and South Carolina, py. 643), with broad disks and long lilac ligules, often produce, even at flowering time, well defined subterranean stolons; the tall southeastern Species with small usually white-rayed heads (Chrysanthemum carolinianum Walt.) has a mass of fibrous roots, with no elongate stolons, at most producing sessile or subsessile superficial basal offsets in late autumn. Similar differences of habit apparently exist in the species of the interior but, most unfortunately, only one or two out of many sheets of specimens of them exhibit care- fully dug and washed subterranean parts. Nine-tenths of all the specimens I have seen are hastily broken or snatched frag- nents without bases. Until properly collected and intelligently laid-out specimens of these plants are available their treatment must be necessarily tentative. In this study I have been greatly aided by the use of the local material of the Academy of Natural Sciences of Philadelphia, most kindly sent me for examination by Mr. Long.! ne The following key, like many others published in recent years in my revisionary papers, ** made primarily for use in a new edit ny Published in the past have subsequently appeared (without serious alteration of 2 minor details) in books copyRIGHTED by others. If compilers of such books like my 7 : uct of weeks or months of concentrated study and of tedious measure- ri it is hoped that they will not attempt to place me in the —— position of S Be Q g = S B my ° ® R : 484 Rhodora [NovEMBER a. Phyllaries (involucral bracts) linear to linear-attenuate or subulate, 0.2-0.75 (rarely -1) mm. broad; disk 3-8 mm. road; awns ge or up to 3% as long as achene, shorter than disk-coro b. Phyllaries 051. mm. “broad, not long-attenuate nor sub- ulate; ligules 0.8-1.5 cm. lo ong, lilac or purplish; heads few, on strongly ascending to erect naked or few-bracted peduncles; plant often spreading by elongate eae ber 1. B. asteroides. b. Phyllaries 0.2-0.4 (rarely -0.5) mm. broad, long-attenuate to linear-subulate ; ae 5-8 mm. long, white or lilac; ads numerous, more or less diffusely paniculate or C; m pass c. Involucre of 2-3 closely imbricated often subequal series of phyllaries, not often extending abe the peduncles b ; peduncles 0.5-5.5 cm. long; achenes wing- oth ends; ligules mm. long, white (rarely lilac) ; isks 3-5 mm. ued: anthers included; awns about 0.1 mm. long ..... 2. B. caroliniana. myers betray elongate stolons; leaves ietling ——s es and short sl eed (0.5-2.5 cm. long); a .7-1 mm. long, the roa chovate schoae ....5. B. latisquama, var. microcephala. é Involucre fe) unequal series, the lower common! m bracts; peduncles stiff and straight, many of * em. ports isks 3-5 mm. broad; achenes ‘mostly, B. diffusa. equaling y oF quite as I mostly 1-2 m ced coarse plant with leafy corymb; ligules lilae or 5 vist 1-18 ¢ nea B. latisquama. 1. B. astrromes (L.) L’Hér. (PLare 640). Slender, simple or with loosely ascending branches, 2-7 dm. high, with basal offsets and stolons developed in autumn; leaves submembranaceous linear to oblong-lanceolate or oblanceolate, broad-based, scarcely petioled, the principal ones 2.5-12 em. long and 3-13 mm. broad; eads 1-23, usually few, on loosely sity Fine to erect naked or few-bracted peduncles 2 25-11 ¢ m. long; involucre of 2-3 series of subequal linear phyllaries 05-1. mm. broad; ligules lilae or pur- 1940] Fernald,—Additions to Flora of Virginia 485 haps also northern Ohio. New Jersny: low woods, Brighton, Sussex County, September 4, 1910, Mackenzie, no. 4778; border of pond east of Swartswood Lake, Sussex County, August, 1911, Mackenzie, no. 4922; muddy calcareous shores and flats, Shyster Pond, Warren County, July 24, 1920, Mackenzie. PENNsyL- -4. & E. G. Heller, no. 595; McCall’s Ferry, September 5, 1892, 8. Brown, July 17, 1903, S. Brown & B. H. Smith. Maryan: of Susquehanna River, Conowingo, Cecil County, July 29-31, 1924, Jos. Crawford; Havre de Grace, September 1, » Long; 14 mile south of Havre de Grace, July 19, 1902, G. H. Shull, no. 75. Norra Carona: sandy roadside, near se of divide, south of Tuxedo, Henderson County, August » 1927, Wiegand & M. anning, no. 3232. Two collections from northern OH10 (bay shore, Sandusky, September 28, 1898, FE. L. Moseley, and Continental, Putnam poutity, October 1, 1883, H. A. Young) seem like the Susque- anna plant but the material is inadequate. That true Boltonia asteroides, as shown by the Type, the speci- men in the Linnean Herbarium described in Species Plantarum (our PLATE 640, rics. 1 and 2), is the slender species of the Sus- quehanna Valley, eastward into northwestern New: Jersey, is “Pparent, although the name has erroneously been made to cover an assemblage of broadly corymbose and very leafy plants ex- tending westward into the Prairie States. Details of the local “astern species are shown in PLATE 640. 486 Rhodora [ NoveMBER rt. S. SavacE; Fic. 3, two heads, , from Harrisburg, Pennsylvania, August, 1858, Porter; Fic. 4, partly denuded receptacle (split), with mature fruit, x 2, from same specimen as FIG. 3; FIG. 5, achene, < 10, from mouth of Tucquan, Lancaster County, Pennsylvania, Heller & Heller, no. 575. = gS oF 2. cr] =] 3. a hb i) ~ is) 3 j=) S. Q d a ie) a a is) Qa 3 ° B uo) = al 3S oO i) _ 3 Typical Boltonia asteroides was early in European botanic gardens, probably derived from the Pennsylvania material sent by Bartram to Linnaeus. A beautiful sheet of it (somewhat overgrown in cultivation) is in the Gray Herbarium, collected by Jacques Gay from plants cultivated in Paris in 1822; and old specimens from other European gardens are there represented. Near the coast, on muddy shores or in tidal marshes (see Pp. 396) from southern New Jersey to Louisiana, Boltonia asteroides becomes stiffer, with firmer leaves and taller stems, and with more promptly stoloniferous habit. In this plant the principal leaves are narrowed into petiolar bases and the stiffer and straighter peduncles are often more bracteolate. Some speci- mens of this coastwise plant are a good match (taking into ac- count the fact that it had been in cultivation) for the original plate of Matricaria glastifolia Hill, basinym of B. glastifolia (Hill) L’Hér. This wider-spread variety becomes ish, narrowed to subpetiolar bases; peduncles often more brac- teate; awns of achenes well develope —Matricaria glastifoha Hill, . Kew, 19, t. 3 (1769). B. glastifolia (Hill) L’Her. Sert. Angl. 27 (1788) —Southern New Louisiana, Ellendale, August 16, 1877, A. Commons; moist open sandy de- pression east of Ellendale, September 5, 1925, Pennell, no. 12,872. rat’s Bridge (or Ferry), James City County, September 19, 1939, Fernald & Long, no. 11,466; brackish marsh of North Landing 1940] Fernald,—Additions to Flora of Virginia 487 River, Pungo Ferry, Princess Anne County, September 22, 1933, Fernald & Griscom, no. 2914; Northwest, Norfolk County, Sep- tember 6, 1893, Heller, no. 1248. Norru Carouina: ditch near Old Dock, Columbus County, August 29, 1938, Godfrey, no. 6337. Sourn Carotina: Santee Canal, September, —, H. W. Ravenel. Grorcia: edge of nearly fresh marshes of Altamaha River, just below Darien, McIntosh County, September 17, 1903, Harper, no. 2003. Louisiana: without stated locality, EZ. Hall. In priate 641, ric. 1 is a plant, x 2/5 from below Barrat’s Bridge, James City County, Virginia, Fernald & Long, no. 11,466; Fics. 2 and 3, heads, 2, from no. 11,466; ric. 4, achene, x 10, from Pungo Ferry, Princess Anne 2914. . *B. earoliniana (Walt.), comb. nov. (PLATE 642). Very tall, up to 2.3 m. high, from fibrous roots, without elongate stolons, freely branched with wide-spreading to subascending long pa- niculate branches; early (soon dropping) lower leaves broadly oblanceolate, membranaceous and blunt; principal cauline leaves at flowering time submembranaceous, lanceolate to linear- oblanceolate, tapering to apex and petiolar base; paniculate branches up to 7 dm. long, leafy, diffusely forking; peduncles filiform, 1.5-4.5 em. long; involucre of 2-3 close series, the linear-subulate phyllaries only 0.2-0.3 mm. broad; ligules white (rarely lilac), 5-7 mm. long; disks 3-5 mm. broad; achenes harrowly cuneate-obovate, narrow-rimmed; awns minute, about Woods, upper terrace of Nottoway River, southwest of Burt, Sussex County, September 20, 1937, Fernald & Long, no. 7678; alluvial woods, Nottoway River, southwest of Lambs, Sussex County, September 20, 1937, Fernald & Long, no. 7679; bushy swale 114 i east of Stony Creek, Sussex County, August 24, June 23, 1936, Fernald, Long & Smart, no. 5936 (young foliage) ; must yey ooded bottomland of Nottoway River, Courtland, Au- . eherrin River, near Haley’s Bridge, Southampton County, ‘gust 19, 1938, Fernald & Long, no. 9186. So 488 Rhodora [NovEMBER Cooper River, Berkeley County, October, 1847, Cranmore Wal- lace; an pe g along logging railroad, floodplain forest, along ntee Riv a suilee northeast of Pineville, Berkeley County, September i 1939, Godfrey, no. 8155. This tall and paniculately branched species, with usually white ligules, the lanceolate leaves attenuate to base and commonly to apex, agrees far better than other Atlantic-slope species with Walter’s description of his Chrysanthemum carolinianum. Walter’s account was as follows: Carolinia- foliis lanceolatis integris utrinque acumi- num 1, natis laevibus, caule 5 ad 6-pedali ramosissimo, floribus radio albis disco luteo Walter lived on Santee River. We now have two collections of the tall much branched plant with leaves “utrinque acuml- natis” and with white rays from that region. Even if the Walter type is never found the identification seems quite safe. 3. *B. Ravenelii Fernald & Griscom, sp. nov. (TAB. 643), planta 2-8 dm. alta paniculato-ramosa basi aicstas: foliis primariis membranaceis anguste obovatis vel late oblanceolatis pedunculis sparse eaetelace tine 2-5 cm. eae p a laribus lineari-subulatis 0.2-0.4 mm. latis 2-3-seriatis; ligulis Wasi 7-8 mm. longis; disco 58 mm. lato; antheris deinde xsertis; achaeniis vix aristatis ——Rich bottomland woods, very i southeastern Virginia and eastern South Carolina. VIR- 1846, H. W. Ravenel; ak swamps, Santas Canal, October, — Ravenel (type in Herb. Gra ay). A very little known species. The two Ravenel specimens, sent to Asa Gray more than 90 years ago, and the single collec- tion from southeastern Virginia are consistent in their thin and broad leaves, stiffly ascending peduncles, very small involucres, 1940] Fernald,—Additions to Flora of Virginia 489 lilac ligules and awnless achenes. The Virginia material, col- lected very late in the season, has most of the anthers exserted, and this unusual character shows in the older heads of the Ravenel material. In pLate 643, Fic. 1 is the original Ravenel material, x 2/5, the TyPE of B. Ravenelii at the left; rias. 2 and 3, heads, x 2, from the Type; Fic. 4, achene, X 10, from the tyPE; Fic. 5, heads, showing exserted anthers, < , from Fontaine Creek, southwest of Haley’s Bridge, Greensville County, Virginia, Fernald & Long, no. 9642. 26, 1903, Fredholm, no. 6151; cypress-head, Kelsey City, Palm Beach County, December 16, 1920, Small, DeWinkeler & Rane, & Earle, no, 191; Deer Island, October 12, 1895, Tracy, no. 4778. ARKANSAS: roadside ditch near Hot Springs, August 3, 1935, | ea i Scully, no. 66 UL i ! 0. 67; open dry soil, south edge of Hammond, Tangipahoa Par- ish, June 30, 1938, D. S. & Wood, Caleasieu Parish, July 18, 1938, Correll & Correll, . OKLaHoma: along snd streams, San Bois Mts., 1891, C. 8. Sheldon, no. 280. ExaAs: Dallas, 1873, Reverchon, no. 233; Houston, September 14, 1913, G. L. Fisher; neglected field, Montgomery Co., July 18-21, 1909, R. A. Dixon, no. 492; ponds, Hempstead, June 6, 1872, E. Hall, no. 308. 490 Rhodora [NovEMBER Boltonia diffusa is here taken up in its traditional sense, al- though Elliott’s type has not been studied. The range is usu- ally stated as extending northward into South Carolina. I have seen no material, however, from north of northern Florida and southwestern Georgia, and I am inclined to the interpretation that authors have merely assumed that Elliott’s species was described from South Carolina. Elliott’s very detailed account states, fortunately, that his species “Grows in damp rich soils between the Chatahouchie and Alabama”, 7. e. somewhere be- tween southwestern Georgia and northwestern Florida and western Alabama. Characteristic material is known from Sumter and Lee Counties, Georgia, in the drainage area of Flint River, which, at the extreme southwest corner of the state, joins the Chattahoochie to form Appalachicola River, whence B. dif- fusa is also known. Until authentic material from South Caro- lina is known, that state may well be omitted from the range. Much of the material from the Mississippi valley differs from typical Boltonia diffusa in its coarser habit, much more branch- ing stems and rather shorter and stiffer peduncles. Its bases are rarely collected, but the few seen usually show no stolons. This is B. pirrusa Ell., var. interior Fernald & Griscom, var. NOV. (TAB. 644), plerumque estolonifera; caulibus 0.6-1.2 m. altis basi 4-7 mm. crassis paniculato-ramosis, ramibus plerumque valde ramuliferis ; ig se snag phyllaribus lineari-oblongis apice vix subulatis. KENTU eed in field, northeast of Murray, Calloway County, July: 19, ‘1937, Smith & Hodgdon, no. 4078. og dry oak barrens, alt. 1100 feet, Tullahoma, Coffee Blankinship. ARrKANsas: waste, ais cotton- Pees Brinkley Monroe County, August 28, 1934, Delzie Demaree, no. 10, bottoms, Arkansas River, Pine Bluff, October 24, 1931, Demaree, no. 8786; Texarkana, Miller County, September 8, 1898, A. A. & E. G. Heller, no. 4229 (transitional). OKLAHOMA: near edge of pond in woods, near Copan, vests County, August 15, 1913, G. W. Ste no. 2094, as B. latisqua Cae interior is a pcm plant, in its oe extreme develop- 1940] Fernald,—Additions to Flora of Virginia 491 ment clearly belonging near if not with Boltonia diffusa, but in extreme forms approaching small-headed extremes of B. lati- squama. Fuller material, especially with carefully collected root-systems, is important to clarify the relationship here. In PLATE toe FIG. 1 as the typz, < 2/5, of Boltonia diffusa, var. interior ; Fic. 2, heads, x 2, from Type; FIG. 3, mature head, X 2, from Tullahoma, ‘Terhnnes, Scne no. 4257; FIG. 4, echone. < 10, from no. 4257. 5. B. LarisquaAMA Gray (PLATE 645). Coarse, 0.7-1.5 m. or more high, the base apparently not stoloniferous; stem strongly corymbose-paniculate and leafy, conspi icuously corrugated ; leaves firm, the primary ones lanceolate, those of the corymbose branches linear-attenuate and much smaller; peduncles leafy- bracted, 1.5-5 em. long; involucre of abo ut 3 series of firm narrowly cuneate- to spatulate-obovate aiibieh oe 1-2 mm. broad; ligules Mire or lilac, 1-1.8 em. long; mm. prey achenes obovat 2.5-3 m m. long, broad-winged; awns 1.5—-2 mm. long. ee Journ. "Sei, ser, 2, xxxiii. 238 (1862) —Prairies and banks of streams, Missouri, Kansas and Oklahoma, probably Texas; escape from cultivation eastward to New England. In piate 645, rics. 1 and 2 are portions of the typz, X 1, from near the mouth of the Kansas River, Parry; Fic. 3 is a characteristic head, x from Baker, Lincoln County, Oklahoma, August 25, 1895, Blan kinship; Fic. 4, a fruiting head, x 2, from = agen Sorevetca by Gray in the Har- vir ‘Botanic Garden, October 1865; Fic. 5, an achene, x 10, from the Var. OCCIDENTALIS Gray. Clearly stoloniferous; leaves larger and blunter ; ee ate naked; phyllaries more pointed. —Syn. Fl. i?, 166 (1884). B. occidentalis (Gray) Howell, FI. Nw. Am. i. ‘ae (1897) —Bottomlands, Union County, Oregon. When better known, var. occidentalis may prove to be a dis- tinet species. Howell in elevating it to specific rank added no new oo and gave no reason for separating it. ecognita Fernald & Griscom, var. nov. (TAB. 646), foliis eaulinis valde evolutis oblanceolatis vel badkpoabates: corymbis i 4 dm. latis foliosis; phyllaribus anguste oblongis 0.5-1.3 mm. atis; disco 0.7-1.5 em. latis; ligulis 1-1.8 cm. longis.—B. aster- oides sensu Am. auth., non (L.) L’Heér. St Wiichiona to ts and North Dakota, south to Kentucky, Missouri a Kan naturalized in New England and New Jersey. Type: Fort Snell. ing, Minnesota, August 20, 1891, FE. A. Mearns (Herb. Gray). Boltonia latisquama, var. recognita is the tall plant of woods and prairies, with broad corymbs of large heads, which has been 492 Rhodora [NovEMBER passing erroneously as B. asteroides. It seems, phylogenetically, to be the main variety of B. latisquama, of which typical B. latisquama is an extreme development in the drier prairies and plains. Var. decurrens seems to differ from the wide-ranging var. recognita only in its decurrent leaves, while var. micro- cephala is a small-headed extreme somewhat approaching B. diffusa, var. interior and perhaps mixing with it. In piate 646, Fics. 1 and 2 are portions of the type of var. recognita, xX 1; Fig. 3, a head, x 2, from the rrpr. Var. decurrens (Torr. & Gray) Fernald & Griscom, comb. nov. B. glastifolia B. ? decurrens Torr. & Gray, Fl. N. Am. 1. 188 (1842). B. decurrens (Torr. & Gray) Wood, Bot. and Florist, 166 (1870). B. asteroides, var. decurrens (Torr. & Gray) ngelm. in Gray, Syn. Fl. i?. 166 (1884). —Bottomlands of Illi- nois and Missouri. Var. microcephala Fernald & Griscom, var. nov. (TAB. 646), var. recognitae simillima; foliis primariis oblongis vel oblanceo- latis 0.6-1.5 dm. longis; pedunculis 0.5-2.5 em. longis; phyllari- bus lineari-subulatis vel -lanceolatis 0.3-0.4 mm. latis; ligulis mm. longis albidis; disco 5-8 mm. lato; achaeniis 1.5-2 mm. longis; aristis ca. 1 mm. longis.—Wisconsin, Illinois, Minnesota owa. Wisconsin: St. Croix Falls, August 1, 1900, ©. “. Baker. Iturnots: along Illinois Central Railroad, south of Harvey, Cook County, August 30, 1893, S. H. Burnham (TYPE in Herb. Gray); wet bottoms, Urbana, Gleason, no. 1961; low wet soil near Creek, Vienna, August 9, 1902, Gleason, no. 2741; Fox Bottom, near Mud River, August 18, 1914, Robert Ridgway: Minnesota: Lake City, August 15, 1883, W. H. Manning; ora Shakopee Lake, Louriston Township, Chippewa County, Augus 5, 1930, Hotchkiss & Jones, no. 362. Iowa: wet places, Johnson County, August, 1895, T. J. Fitzpatrick. In Pua’ , Fig. 4 is a group of mature heads, x 2, from the sh of var. microcephala; Fic. 5, an expanded head, x 2, from TYPE; FIG. ©, an achene, x 10, from Typr. ASTER GRANDIFLORUS L. Our northernmost station is in Bue Witu1am County: argillaceous roadside thicket northwest 0 King William Courthouse, no. 11,630. *A. tavaricus L. f. CHartes Crry County: escaped to dry roadside thicket, Holdcroft, no. 11,631. *A. UMBELLATUS Mill., var. brevi 8, var. Nov. Sas 2 647, Fics. 1-4), caulibus ad 1.4 dm. altis; foliis laneeols’ ellipticis vel anguste ovatis glabris; involucris 2.54 mm, a'Ui® phyllaribus 4~5-seriatis valde inaequalibus mediis linear oblongis obtusis ca. 0.5 mm. latis; ligulis 4-7 mm. longis.—Bogs 1940] Fernald—Additions to Flora of Virginia 493 and swamps, District of Columbia and eastern Virginia. Dis- TRICT OF CoLuMBIA: swamp, Terra Cotta, September 29, 1897 and September 3, 1915, Th. Holm. Vircrnia: sphagnous springy swales bordering Whiteoak Swamp, west of Elko Station, Hen- rico County, August 17, 1938, Fernald & Long, no. 9192; argil- laceous and siliceous boggy depressions, about 3 miles south- east of Petersburg, at head of Poo Run, Prince George County, October 18, 1936, Fernald, Long & Smart, no. 6894 (TYPE in erb. Gray; isorype in Herb, Phil. Acad.). Typical Aster wmbellatus Mill., ranging from Newfoundland to northern Ontario, south to western North Carolina, Ohio, Indiana and Illinois, has the involucre (Fics. 5-7, all x 4) 4-5 mm. high, of gradually lengthening narrow linear-lanceolate acute or obtuse phyllaries, those of the median series nearly equaling the inner ones. Commonly with lanceolate or lance- oblong leaves, it may have them relatively short and broad (elliptic to lance-ovate). The latter individuals are A. humilis Willd., small plants sent to Berlin from Pennsylvania but (as shown by Muhlenberg’s material and other eastern Pennsylvania broad-leaved plants) scarcely worth even formal designation, although some authors maintain them as a species or as a variety. The most extreme departure from typical A. wmbellatus in the Series with essentially glabrous but often scabrous leaves and gradually lengthening series of narrow phyllaries is the plant of wet cliffs described as Doellingeria umbellata, var. flexicaulis Ouse. As an extreme departure from the usually stiffly erect plant it is noteworthy but its involucre is not strikingly different and it seems to me a forma rather than a true geographic Variety 20 Var. pubens Gray, chiefly of the region from Upper Michigan to Saskatchewan, south to Iowa and Nebraska, is harsh with scabrous puberulence but otherwise resembles typical Aster umbellatus. Its involucres (rics. 8 and 9, both X 4) are in- clined (in dried material) to be more turbinate and the rela- tively few phyllaries are very narrow. On the Coastal Plain, from Florida to Texas, north to North Carolina and Arkansas, occurs A. wmbellatus, var. latifolius Gray, Syn. Fl. N. Am. i2. 197 (1884). Gray, and most later oa UMBELLATUS Mill., forma fllexicaulis (House), comb. nov. Doellingeria umbellata ill.) Nees, var. flexicaulis House in N. Y. State Mus. Bull. no. 254: 712 (1924). 494 Rhodora [NoveMBER authors, confused the really definite plant of the South by citing a synonymy of sensu names. Gray’s differentiation was based solely on breadth of leaf (“leaves from ovate-lanceolate to ovate, comparatively short, less narrowed or sometimes even rounded at base”; with the significant statement: “Extreme forms seem very Setechal from A. umbellatus, having leaves even 2 inches wide by 3 in length”). It is thus clear that Gray had a broad-leaved southern plant primarily in mind, but he failed to note the distinctive involucre and consequently included any plants with “ovate-lanceolate leaves’ and extended the range north to Pennsylvania and New Jersey. The first synonym was A. humilis Willd. Sp. iii. 2038, which, as already noted, is only a low and broad-leaved phase of typical A. umbellatus, although Willdenow so little understood his own A. humilis as to publish in 1806 in Hort. Berol. t. 67, a beautiful plate of A. infirmus Michx. as A. humilis! Gray’s next synonyms under A. umbellatus, var. latifolius were A. amygdalinus sensu Bertoloni (1847) and Doellingeria amygdalina Nees (1833), “chiefly, excl. syn.”, which is a misleading reference since the basinym, Aster amygdalinus Lam. Encyel. i. 305, no. 24 (1782) was founded (as shown by a photograph sent from Paris) on very character- istic A. umbellatus cultivated at Paris. The other synonym given by Gray under A. umbellatus, var. latifolius was Diplopap- pus cornifolius sensu Lindl. (1835), not A. cornifolius Muhl. ex. Willd. (1804). The synonymy given by Gray, consequently, is too vague to stand as the basis of his new var. latifolius, and only A. humilis could have been the nomenclatural basis for anything, the others being misapplications of older names by later authors. I am, therefore, taking Gray’s name for the plant of the southern Coastal Plain which, alone, was so marked by him in preparing the Synoptical Flora; and, since he specially commented on “Extreme forms ... very different from A. umbel- latus, having leaves even 2 inches wide by 3 in length”, I am designating as Type of var. latifolius the specimen he thus signi- fied. This was collected by Hale in Louisiana in 1844. Photo- graphs of portions of it, & 1, are shown in Piate 648, with the characteristic long involucres (5-7 mm. high), with phyllaries 0.8-1.5 mm. broad of it and other specimens shown, X 4. That 1940] Fernald,—Additions to Flora of Virginia 495 A. umbellatus, var. latifolius, as thus cleared of its encumber- ances, is a very distinct variety of the South will be apparent. In pLate 647, Fic. 1 is a portion of the TYPE, X 1, of Aster umbellatus, var. brevisquamus; Fic. 2, a head, to show involucre, x 4, from ; ric. 3, head, x 4, from Elko Station, Virginia, Fernald & Long, no. 9192; Fic. 4, involucre, x erra Cotta, District of Columbia, Holm. Fig. 5 is a head, X 4, of typical A. umbellatus from Ithaca, New Yor : e pubens, from the typn, Saskatchewan, Bourgeau; Fig. 9, involucre, X 4, from Madison, Wisconsin, August 30, 1893, Churchill. N PLATE 648, Fics. 1 and 2 are portions of the typr, xX 1, of A, umbel- : s mM TYPE; Fic. 4, involucre, X 4, from Alabama, Gates; ria. 5, involucre, X 4, from Gateswood, Ala- bama, Tracy, no. 8587; Fic. 6, involucre, x 4, from Georgia, September 30, 1903, Mrs. Taylor. ERIGERON veRNuS (L.) T. & G. Local range extended inland to GREENSvVILLE County: peaty swale by Southern Railway northeast of Emporia, no. 10,441. See p. 360. - BONARIENSIS L. To the single station in Norfolk County already reported add from Princess ANNE County: borders 0 k 11,188; © Island Life Saving Station, no. 11,189; clearing near old house, Cedar Island, no. 12,489. See p. 389. *PLUCHEA PURPURASCENS (Sw.) DC. Princess ANNE County: brackish to fresh marsh along Back Bay, at eastern margin of Long Island, no. 11,190; moist sandy shore, Ragged Island, no. 12,490. See p. 389. A tropical American species heretofore recorded northward only to the coast of Georgia. The Long Island plant is ex- treme, with lanceolate leaves cinereous-pilose beneath, the in- volucre heavily pubescent, its outer phyllaries ovate. Some other collections, from near Hampton, and from tidal shores of the Chickahominy, have narrow leaves but the larger heads and oblong or elliptical outer phyllaries of the common northern P. marilandica ( Michx.) Cass. Further observation may show that the latter usually broad-leaved and smoother plant may have to be treated as an essentially northern variety of the former. Jntil I have more field-experience with them in the transition- belt I am not ready so to treat them. TETRAGONOTHECA HELIANTHOIDES L. To the station in Glouces- ter County add two in SourHampton County: border of sandy Woods southwest of Applewhite’s Church, no. 10,447; rich sandy 496 Rhodora [NoveMBER and loamy woods along Three Creek, northwest of Carey Bridge, no. 10,448. See pp. 356 and 362 and MAP *Ruppeckia Heniopsipis T. & G. Prince Grorce County: borders of dry pine and Benn 2-3 miles north of Disputanta, nos. 10,837, 11,196 and 1 Very extensive ak a first known outside eastern Ala- bama and western Georgia. See pp. 375 and 382 and map 10. *R. Heuiopsipis T. & G., forma villipes, f. nov., petiolis imis pag gion —Prince George County, VirciniA: border of dry pine oak woods 2-3 miles north of Disputanta, July 24, 1939, Fernald & Long, no. 10,838, August 17, 1939, no. 11,197 (TYPE in Herb. Gray., IsoTYPE in Herb. Phil. Acad. :; Most of the plants in the large colony belong to Rudbeckia Heliopsidis a. of Torrey & Gray, “almost glabrous”, originally from Columbus, Georgia. Forma villipes is relatively scarce. It was described from Cherokee County, Alabama, by Torrey & Gray as var. B., “stem stouter, pubescent below with spreading, above with appressed hairs.” In Prince George County it is at best a minor form. CoREOPSIS ONIScIcARPA Fern. The greatest area in Virginia of this species is in southwestern NANSEMOND COUNTY. There in sphagnous swales and springy, peaty slopes it is often 9 dm. high, single plants bearing 30-60 heads. See p. 399. Dr. Sherff, in Field Mus. Nat. Hist. Bot. Ser. xvii. 609 (1939), states that “A careful comparison of each specimen [of C. onis- cicarpa lent him from the Gray Herbarium] with Floridan and other specimens of C. Linifolia [capitalization Sherff’s] Nutt. showed no constant differences in foliar or capitular characters to warrant separating C. Oniscicarpa [capitalization Sherfi’s] even varietally from C. Linifolia.” The involucre with lanceo- late outer phyllaries and a series of achenes of C. oniscicarpa were shown, from photographs, in Ruopora, xl. pl. 534, figs. 1 and 8, the habit of the plant, « 14, in plate 533. Coreopsis linifolia Nutt. Journ. Acad. Phila. vii. 75 (1884) was one of several species described “from the dried specimens in the herbarium of the Academy of Natural Sciences in Phila- delphia.” It came from Alabama; and since other Alabama specimens cited in the same paper were collected by Dr. Gates (Liatris pauciflosculosa and L. squamosa) it is reasonable to 1940] Fernald,—Additions to Flora of Virginia 497 take the series of Alabama sheets from Gates labeled by Nuttall as his own C. linifolia and preserved at the Philadelphia Acad- emy as the Type series. Other Nuttall types described “from the dried specimens in the herbarium of the Academy” are there preserved and it is not probable that the Philadelphia Academy would have transferred them to the British Museum. Sherff, in Field Mus. Nat. Hist. Bot. Ser. xi. 436 (1936), gave for C. linifolia “Type specimen: Collected in Alabama (Brit.?),” the queried “Brit.” standing for British Museum. On the seemingly safe assumption that the Typx series is the one at Philadelphia labeled by Nuttall, but with similar specimens of the Gates material also in the Gray Herbarium and in the herbarium of the New York Botanical Garden, I am showing details of the TYPE series, as represented by the Gates specimen in the Gray Herbarium. Since Sherff cites none of the Gates material (ex- cept through his note on C. callosa), I have felt it justifiable to include some details from other specimens actually cited by him as C. linifolia. In Pate 649, ric. 1 is an 1sorype (Gates), X %%, of C. linifolia, labeled by Sherff; ric. 2, a flowering head, X 2, from the he margined; rig. 3, two plants, x %, from ateswood, Alabama, Tracy, no. 8565, labeled and cited by Sherff as C. linifolia; Fias. 4, 5 and 6, achenes x 10, from tidal marshes on Bilo ississippi, Septem- ber 16, 1885, John Donnell Smith (labeled see eked by Sherff as C. linifolia) . In C. linifolia the relatively large involucre has the ovate white-margined outer phyllaries, to use Sherff’s phrase, “ovatae -- . lateribus scariosae, apice subacutae vel rotundatae”. The much smaller involucre of the more northern C. oniscicarpa (shown, also < 2, in Ruopora, xl. pl. 534, fig. 1) has the outer phyllaries lanceolate, the inner more oblong and not conspicu- ously white-margined. In the more southern C. linifolia the bodies of the achenes are 2.5-3.2 mm. long and 1-1.2 mm. broad; In C. oniscicarpa the bodies of the achenes are 1.8-2.2 mm. long and 0.6-0.9 mm. broad. These details of C. linifolia, shown in PLATE 649, and the photographic illustrations of C. oniscicarpa, already referred to, clearly speak for themselves. The photographie reproductions of other species, in Coreopsis and Bidens, all reduced by Sherff in the above cited paper, also speak with complete definiteness. 498 Rhodora [NoveMBER *BIDENS POLYLEPIS Blake. CHarLes Crry County: clearing in woods and roadside gutter, southeast of Sturgeon Point School, no. 11,203. Sherff cites no coastwise material from south of Maryland. *B. coronata (L.) Britton, var. typicaA Fernald in RHopora, xl. 349, pl. 506, figs. 4 and 5 (1938). Kine Wiiti1am County: fresh tidal shore of lacey: gigi Horse Landing, near King William Courthouse, no. 11,637; similar habitat, northwest, of ing William, no. t1 638. KinG sx SEEN County: similar habitat, Walkerton, no. 11,639. See First from south of lower ae leaves simple to com- pound; awns of achenes very broad. *GAILLARDIA PULCHELLA Foug. rsa sen County: sandy roadside south of Emporia, no. 11,205. Hrnrico County: waste ground and railroad ballast, Rakatons, no. 2498. Spread from original cultivation. *LapsaANA communis L. SourHampton County: cinders of freight siding, Branchville no. 10,454. Henrico County: waste ground, Richmond, no. 12,206; embankments and cinders of Chesapeake and Ohio Railroad west of Elko Delos no. 12,207. ERINIA OPPOSITIFOLIA (Raf.) Ktze. Add a station in GREENS- ey Cou ce fallow clay field west of Vincent School, no. 1 Reems AMPTON County: sandy roadside ditch, abet of Windia? s Mill, no. 1 LacTUcA CANADENSIS L., var. LONGIFOLIA (Michx.) Farwell. GreensviLte County: clearing on wooded bottomland of Fon- taine Creek, southeast of Taylor’s Mitpond, no. 10,850. One of the more northern varieties, not represented in the Gray Herbarium from Atlantic States south of New York and singularly out of place with Taxodium, Crataegus Marshallu, Sagittaria Weatherbiana, ger ee! cernua var. odorata and Amsonia Tabernaemontana. See p *L. rLormpana (L.) Gaertn., forma leucantha, f. nov., —— albidis—Sussex County, Vircinta: wooded botto mland, Jone Hole Swamp, west of Coddyshore, September 17, 1939, Feral & ae no. 11,484 (Type in Herb. Gray; 1soryPE in Herb Ac PRENANTHES ALTISSIMA L. Local range extended southward into SouTHAMPTON County: rich marly woods along Three Creek, northwest of Casey Bridge, no. 11,486. See p. 363. (To be continued) A CENTURY OF ADDITIONS TO THE FLORA OF VIRGINIA M. L. Fernaup (Continued from page 498) Part III. PuHyvrogrogrRaPpHic CONSIDERATIONS In 1937 I published”! a brief analysis of the diverse geographic affinities of the flora of the Coastal Plain of Virginia. At that time I suggested seven primary types of relationship there dis- played; these should perhaps be reduced to six. The main groupings, however, seem to hold. During the succeeding years considerable additions have been made to most of the groups; and during the five trips from June to October, 1939 (excluding obviously introduced weeds), we were able to extend into Vir- ginla a great number of ranges: 11 plants of the upland (chiefly Blue Ridge, Appalachian Valley and Alleghenies) new to the Coastal Plain; 20 heretofore unrecorded from north of North Carolina; 9 unknown north of South Carolina; 6 unknown north of Georgia; and 1 heretofore known only in Alabama. Giving bs 4 station or stations to fill in the previous broad gap between tetas Carolina and isolated areas in Delaware, Maryland or “ie Jersey are 6 species; while 2 supply an intermediate area nellahate South Carolina and the North, and 4 species were found Bais had heretofore been known only from the Mississippi ma or the Gulf States. New southern limits were established Species: 13 the first from south of New Jersey, Delaware, 21 Rxopora, XXxix. 465-489 (1937). 504 Rhodora [December Maryland or the Potomac near Alexandria, 1 the first from south of southern New England. The Asiatic Aneilema Keisak 1s new to the American flora. Desmodium glabellum has ap- parently been unknown since its original collection in the 18th century in South Carolina. One species, erroneously identified, drops from the Virginia list; and 22 plants new to science, 7 of them endemic in Virginia, were discovered or worked out from earlier collections. Still others await fuller study. It is needless here to discuss in detail most of the geographic relationships of these plants. They fit, for the most part, into the groupings already discussed. It is gratifying, however, to see regular increase in the group of species which apparently radiated out from the Appalachian Upland as it became ele- vated from its coastal plain status of Cretaceous time; and to see the gradual reduction in number of the species isolated from North or South Carolina in Delaware or New Jersey. One group of species, those of the fresh or but slightly brackish tidal shores and marshes has not previously been considered in this series of papers. It may, therefore, be specially discussed. Tue Fiora or Fresu Tinan Estuaries AND SHores.—The peculiarity of the fresh or barely brackish tidal estuary”® is the regular action of tide, alternately flooding and leaving bare the inner shores of streams and inlets twice a day with essentially fresh water. The plants which can tolerate such daily changes are a limited number. Besides the regular and somewhat in- different species of reed-marsh they consist of a remarkable group of species of a few paludal genera, the species or the genera usually exhibiting as a regular feature of their geographic distribution extreme localization. Although a few plants found in our more brackish estuaries, like Spartina cynosuroides and Scirpus robustus, tolerate considerable salinity and usually fol- low the outer coast, the more typical estuarine species are in- tolerant of much salinity in the waters and confine themselves to the fresh to but slightly brackish reaches of streams, pools and inlets. This group is, then, of peculiar interest, since the plants have apparently mostly attained their present habitats and extreme isolation in the past, at periods when they could Page Fassett, os ee The Vegetation i om Estuaries of Northeastern North America. Proc. Bost. Soc. . Hist. xxxix. no. 3 ( 1940] Fernald,—Additions to Flora of Virginia 505 migrate from river to river along fresh or brackish (not strongly saline) shores. Such conditions today prevail in a region like Back Bay in southeastern Virginia, where the off-shore bars and continuous dunes of the outer shore shut in a shallow body of tidal waters, perpetually renewed by fresh streams or seepage, and with the nearest inlet from the open sea far below the Virginia-North Carolina line. Rarely, during wild storms, sea-water dashes from the outside Atlantic into sheltered Back Bay; and sufficient salinity has been preserved in some of the marshes to maintain a few specially tolerant halophytes. The shores of Back Bay and the fresh pools and ponds on its margins or on Long Island are, however, the homes of such notable plants of the fresher marshes and waters as Cyperus haspan var. americanus (MaP 9), the American representative of a pantropical species (tropical and warm-temperate North and South America, Africa and warmer regions of Asia and Australia). The world-range of the species is suggestive of that of Cyperus brevifolius (map 21). Other typical plants of the freshish marshes of Back Bay are en- demics of the two Americas; such, for instance, as Typha truzil- lensis (Map 8), with details of distribution somewhat different from those of the two species of Cyperus and an ability, through its coma of perianth-bristles, to spread, locally, away from the slightly brackish marshes (where it rarely persists). Others are endemics of the southeastern United States. A good example is Juncus megacephalus (map 5), a very distinct species which °ccurs in the fresh to brackish marshes from Texas to the shores of Back Bay. Its habitats, as given on such labels as clearly Indicate them, are as follows: ditch (La.), wet sandy shores (Fla.), mucky ground (Fla.), flatwood ponds (Fla.), prairie (Fla.), pineland (Fla.), low pinelands (Fla.), moist pine barrens (Ga.), low ground back of sand-dunes (Ga.), lime-sinks (Ga.), Savannah (S. C.), dune-hollows (S. C.), river-marsh (S. C.), salt Meadow (N. C.), marsh (N. C.), dune-hollow (Va.), swale back of dunes (Va.), inner border of brackish to fresh marsh (Va.). Th brief, Juncus megacephalus, like its associates, is not a pro- nounced halophyte; we do not find it, like J. Roemerianus, for instance, following the salt marshes. It is not a plant of the Saline outer coast but rather of the fresh to barely brackish inner 506 Rhodora [December margin of the coast, sometimes in fresh inland habitats. With great stretches of fresh to slightly brackish inner shore, now extending from below Cape Henry to Cape Fear and, formerly, doubtless more continuously to Florida, it has been able to follow more or less without interruption its most favorable habi- tats; but it does not follow north along the saline outer coast. Not all the American plants of Back Bay shores have so continuous a range on the inner coast as do Cyperus haspan var. americanus and Juncus megacephalus. As pointed out on p. 371, the remarkable little genus Lilaeopsis (formerly called Crantzia) is a living relic of a very ancient dispersal (map 7), with its species variously scattered in New Zealand, southeastern Aus- tralia, Tasmania, temperate and subtropical South America, the Andes, Mexico and southern Arizona, and Pacific and Atlantic temperate North America. Apparently this disruption of the genus is the result of breaking down of old connections, with Eurasia and Africa omitted. It is difficult not to consider it a remnant of an old spread northward from ancient Antarctica. A frequent species of the Atlantic margin of North America is L. chinensis (name reflecting a geographic misconception by Lin- naeus), with dispersal in fresh to brackish estuaries from Florida to western Nova Scotia. It is in many estuaries of Virginia. The specialty of Back Bay, however, is a much larger species with matted creeping and floating stems, L. carolinensis (MAP 6). On the southeastern side of Long Island in Back Bay 4 small pond at the head of fresh marsh is filled by this species, a very distinct member of the genus, with its chief center on the lower reaches of La Plata River in temperate eastern South America, but with four remote stations known in North America: near New Orleans; shallow water near Myrtle Beach, South Carolina; an unidentified station (presumably near Wilming- ton), North Carolina; and this pond on Long Island. Lilaeopsts carolinensis doubtless occurs in other shallow waters along OT near the inner coast, back of the outer rim of sand dunes, but its two areas, one in latitudes 30°-36° 35’ north, the other be- tween latitudes 25° and 35° south, were presumably derived from a former more continuous large area. In this connection the range of Triglochin striata (map 19), with a subantarctic dispersal, on Chatham Island and in New Zealand, Australia, eee ae ee pe ee 1940] Fernald,—Additions to Flora of Virginia 507 South Africa, temperate South America and warm-temperate North America (in eastern South America found between lat. 22° and.40° south, in eastern North America between . lat. 22° and 38° north), comes to mind, for Triglochin striata is known in Virginia only from the tidal marshes of Back Bay. It here accompanies Ammannia Koehnei (map 4). This is one of the most localized of estuarine plants. Discovered prior to 1840 on the marshes of Hackensack River in northern New Jersey, the species is now known on the estuary of York River and on Back Bay (an endemic variety on one river-estuary), with two known stations (doubtless many more) on the off- shore bar of North Carolina, the other known stations in Florida and locally along the Gulf of Mexico. Like so many estuarine species its range is interrupted. It is, therefore, significant that it belongs to another genus of somewhat general pantropical and warm-temperate range. As stated by Koehne in his treat- ment of the Lythraceae (in Engler’s Pflanzenreich), Ammannia has 20 species, occurring in Australia, the Malayan region, P olynesia, southeastern and southern Asia (extending thence into southeastern Europe and the Mediterranean), Africa, sub- antarctic Sandwich Islands, South America and the warmer regions of North America. Ammania is not a holarctic genus. Like so many others which stretch northward into warm- temperate North America the genus today is primarily tropical and it also shows an austral disruption suggesting an old Ant- arctic dispersal. In brief, the species which characterize the fresh to but slightly brackish shores and pools about Back Bay are largely plants of highly restricted and localized occurrence, and they belong for the most part to genera or species with the characteristically Severed geographic occurrence of all pantropical and subant- arctic groups. Their primary dispersal, considering the fact that they are incapable of succeeding in highly saline habitats, Such as prevail on most coasts, has been a phenomenon of the Past. Only on landlocked coastal shores can it now go on, €xcept when the plants are transported by man or by the rarest or natural agencies, Of course, when a plant like “Wild Celery”, Vallisneria americana, intentionally transplanted as a food for waterfowl, is placed in so favorable a habitat as Back 508 Rhodora [December Bay it will prosper. But the plants which give significance to the flora of the region, Cyperus haspan var. americanus, Juncus megacephalus, Lilaeopsis carolinensis, Triglochin striata and Ammannia Koehnei, are largely unnoticed by those who look upon conservation of the wild life of such an area primarily as the attraction and maintenance of waterfowl. There is no probability that the typical estuarine species have been recently introduced by man. The peculiarly significant indigenous plants of the fresh to but slightly brackish shores of Back Bay are not alone species of tropical and austral groups. Some plants of boreal dispersal are also isolated there. On the seeping, springy sands border- ing Back Bay west of False Cape there is a remarkable turf, occupying the fresh springheads and saturated sands. It con- sists of a close mat of a few species: Eleocharis albida (tropical American, here near its northern limit), 2. halophila Fern. & Brackett (Newfoundland and Gulf of St. Lawrence southward, here at its isolated southern limit), E. Lindheimeri (Clarke) Svenson (northern Mexico and San Bernardino Mountains, with distant stations eastward and northeastward to Texas and Michigan, here remotely isolated) and other local plants Among them is the very definite Ranunculus hederaceus, a matted herb of springy ground. R. hederaceus occurs in western Europe, and its American stations?* are remote. By early authors, with no field experience with the plant, it was assumed to be an introduction from Europe. This may sometimes be the case. In Newfoundland?!, however, it occurs with the regular indigen- ous species of wet sands; and, surely, on the shores back of False Cape, where it forms part of the mat, it seems as indigenous 4s the endemie American species of Eleocharis with which it grows or as the highly localized and endemic Ludwigia brevipes (New Jersey and southeastern Virginia) which abounds, along with the subtropical Bacopa Monnieria, on the neighboring flat. In the Great Dismal Swamp, likewive; it is not in an area where man would presumably carry it. The significance of Ranunculus hederaceus in the problem will later be discussed. 23 See Drew, W. B.: The North American Representatives of Ranunculus, § Batrachium, RHOpORA, xxxviii., especially pp. 12-14 (1936). *4 See Fernatp, Some Relationships of the Floras “4 the Northern Hemisphere, Free. Internat. Congr. Plant Sci. ii., especially p. 1506 (1929 1940] Fernald,—Additions to Flora of Virginia 509 Back Bay, which today presents ideal conditions for the local spread of species of fresh to brackish shores, is not like the typical fresh river-estuary. The fresh estuaries are found far up-stream from the mouths of the eastern Virginian rivers. Those which have been explored by us are on the James and the Chickahominy and their tributaries, on the Pamunkey, the Mattaponi, and, to a lesser extent, on the Nottoway and the Blackwater. They are highly developed but only slightly ex- plored (and that some years ago) on North Landing River. On the main River James the best of the fresh tidal marshes begin about 50 miles from its mouth and thence extend 25 miles up- river, but the smaller tributary creeks and the Chickahominy have good tidal marshes from near their mouths well up-stream, on the Chickahominy at least to Windsor Shades. The York is salt for more than 30 miles, up to Westpoint, at the confluence of the Pamunkey and the Mattaponi. The fresh tidal marshes on the Mattaponi follow that stream (often with broad tidal- marsh islands) about 30 miles, to the region of Walkerton. On the Pamunkey they are finely developed but we have not determined their extent. The Nottoway and the Blackwater unite at the North Carolina line to form the Chowan in North Carolina, reaching the sea via Albemarle Sound, which itself merges into Pamlico Sound, these sounds mostly cut off from the open Atlantic by a wonderful development of off-shore bars, Sometimes 4 miles broad. Tidal conditions extend slightly into Virginia on the Nottoway; on the Blackwater they extend above Franklin. North Landing River, a sluggish tributary of Curri- tuck Sound, thence to the sea through Albemarle and Pamlico Sounds, is bordered nearly its whole length by broad and fresh tidal marshes, Its mouth is more than 60 miles from the first pening to the Atlantic, at Oregon Inlet, on the outer coast of Dare County, North Carolina. The fresh tidal estuaries to Which I am referring, are, then, anywhere from 30 miles (York River system) to 125 miles (Blackwater River) from the open Atlantic. They illustrate, very typically, the estuaries of Atlantic orth America north to the St. Lawrence. On several or all of the Virginian estuaries examined a few of the species of Back Bay are found: Sagittaria falcata Pursh (Guatemala to Dela- Ware and Maryland) ; Cyperus haspan var, americanus (Map 9, 510 Rhodora [December already discussed); Eleocharis albida (see p. 508) and Lippia nodiflora (tropical America, north to Texas, Oklahoma, south- eastern Missouri and southeastern Virginia). The river-marshes, however, have a considerable restricted flora, which we do not know on Back Bay. This includes IsoBTES sAccHARATA (tidal mud of Delaware, Maryland, Dis- trict of Columbia, and Potomac waters to Alexandria and vicinity.—Pfeiffer, Mon. Isoétaceae; southeastern Virginia, closely approaching North Carolina. See p. 406). SAGITTARIA SUBULATA ie mud, Alabama pe: Florida to south- eastern Massachusetts) . ZIZANIOPSIS MILIACEA ly America, north to Maryland and southeastern Miss ECHINOCHLOA PUNGENS, var. ‘coarcrata Fern. & Grise. (tidal marsh, North Landing River, endemic CLADIUM JAMAICENSE (tropical America, ‘north to marshes of North Landing Rive CYPERUS BREVIFOLIUS entiipiial: north to Florida and south- ern Georgia; Chickahominy and Delaware Rivers. See PP. 395 and 419 and m RHYNCHOSPORA MACROSTACHYA var. COLPOPHILA (tidal marshes of Maryland and Virginia ERIOCAULON PARKERI tidal a ud, St. Lawrence River; Penobscot River, Maine, to Blackwater River, Virginia. See. p. 432 and MAP 17). approaching North Carolina. Se ee p. 441 and map 20). F CASSIA FASCICULATA Var. MACROSPERMA (fresh tidal marshes an shores, southeastern Virginia, endemic. See p. 455 and PLATE 635). AESCHYNOMENE virginica (fresh to brackish tidal marshes and shores, southern New Jer ersey, southeastern Pennsylvania and eastern Maryland to the valley of the James. See RHopora, xli. 466 and map 1). HyYPERicuM M sitet ae var. LATISEPALUM (Florida to Texas; fresh tidal marshes of Mattaponi River. See p. 402 and 466). ELATINE AMERICANA (chiefly on tidal mud, St. Lawrence River and interruptedly to Virginia: See p. 466 and map 18). es ee KorHNEI var. EXAURICULATA Fern. (marshes of North Landing River, endemic inne ataTa Ell. (tidal marshes, span ‘e Florida, thence very locally to North Landing Riv ERYNGIUM Pobpaln (Texas to Florida, north to New Jersey. See pp. 386 and 467). LILAEOPSIS pac (tidal marshes, Florida to Nova Scotia. See pp. 391 and 470). 1940] Fernald,—Additions to Flora of Virginia 511 Bacopa CYCLOPHYLLA (tidal mud, Florida to Maryland, with apparent gaps of hundreds of miles. See p. 402 and MAP 22). B. opovara Fern. (tidal mud, Chickahominy river, very rare. See map 24), LoBELIA ELONGATA Small (tidal marshes, very localized, Georgia to Delaware and Maryland. North Landing River and tributaries) . OLTONIA ASTEROIDES Var. GLASTIFOLIA (fresh tidal marshes and shores, southern New Jersey to Louisiana. See pp. 396 and 6 and PLATE 641). sik BIDENS MITIS (Michx.) Sherff (tidal marshes, Louisiana to Flor- ida, thence, very interruptedly, to Maryland. North Landing River). These 22 plants, which, in Virginia at least, are strictly estuarine, are for the most part members of wide-ranging genera; but, whereas some of the more notable plants of the fresh to brackish shores of Back Bay have relatively continuous ranges northward, though others are with strikingly isolated stations, practically all the truly estuarine plants are highly localized. Two of them perhaps persist as relics from former semi-cosmo- politan ranges. Cyperus brevifolius (Map 21) has as wide a Tange as C. haspan (see p. 419), occurring rather generally in the warmer parts of Asia (even north to southern Kamtchatka), the Malayan region, islands of the Indian Ocean, eastern Australia, New Zealand, Oceanica, locally in Africa, on islands of the South Atlantic, and from La Plata River in eastern South America northward to Bermuda and southern Georgia, With isolated stations on the Chickahominy and the Delaware and, westward, in Central America, Mexico and southern Cali- fornia. The world-range is definitely of the pantropical order, with the Suggestion of radiation out of ancient Antarctica. Those who know the plant in eastern Asia and the Malayan region, however, state that it is there inclined to become a weed. On the Delaware it has not long been recognized and we have only a Single station as yet on the Chickahominy. That is below an old ferry-landing, where it is not impossible that the Plant started from oriental packing or straw thrown away. It heeds further watching before we can surely assert that it is Indigenous on the Chickahominy and the Delaware. € other species of remote geographic relationship is Aneilema Keisak (Map 20). It is so definitely a part of the regular vege- 613 Rhodora [December tation of river-shores and fresh tidal marshes throughout the area from the Mattaponi to the Blackwater, always with the endemic and highly conservative eastern American estuarine species, that it is most difficult to think of it as a possible intro- duction. It seems as indigenous as the local Cassia, Aeschyno- mene and Rhynchospora with which it associates and as Phryma, Liriodendron, Carya and the other woodland genera which occur only in eastern North America and eastern Asia. When map 20, showing the range of Aneilema Keisak, is compared with map 21, giving the range of Cyperus brevifolius, one can not fail to recognize that it is like two small segments of Map 21, with the rest of the world eliminated. It must be noted, however, that the very recent discovery of Aneilema Keisak in America at first seems like an argument against its being indigenous. When, however, we consider that such an abundant and very conspicu- ous plant of drier (therefore more accessible) areas of south- eastern Virginia as the gigantic sunflower-like herb, Silphiwm compositum (up to 10 feet high and with leaves often a foot broad) was long overlooked as a Virginian, until within the past decade*®, that the regular estuarine companion of Aneilema Keisak, Aeschynomene virginica (up to 8 feet high, with orna- mental pea-like flowers), was not known as a living Virginian from the time of its discovery by Clayton two centuries ago until Long and I found it in 1938, or that the very conspicuous and usual companion of these, the abundant endemic large- fruited Cassia (up to 6 feet high and with showy orange-yellow flowers) was undetected until 1939, the fact that the Aneilema has only recently been discovered in America becomes a very unimpressive point. Most of the phytogeographically significant plants of tidewater Virginia were unknown there a decade ago. It is most interesting, therefore, that A. Keisak occurs exclu- sively with conservative and endemic American estuarine species in southeastern Virginia and not in rubbish, waste spots, road- sides or man-made ditches. If it is not a native it has assumed a remarkable resemblance to one. It should not be overlooked, on the other hand, that another Asiatic species of Aneilema is found on the Connte Plain from Florida to South Carolina. This is A. nudiflorum (L.) Wallich?* of southern Asia, which 5 See Fernald, Ruopora, xxxix. 329 (1937). * For statement of nomenclatural situation see Merrill, Journ. Arn. Arb. xviii. 65 (1937). 1940] Fernald,—Additions to Flora of Virginia 513 Small (Manual) cites from ‘“Roadsides, woods, and orange- groves, Coastal Plain, Fla. to Ga. Nat. of E. Indies.” Not only is A. nudiflorum native of the East Indies; Clarke, in his monograph, says “India Orientalis, Malaya, China; ab Himalaya ad Zeylaniam, Borneo, ins. Philippine et Loo-Choo; alt. 0-2000 met., vulgatissima”. I do not know its status in Florida and Georgia, except from the statement of Small; but the only Flori- dan label in the Gray Herbarium with statement of habitat reads “low flat woods”, while Neil Hotchkiss, writing of its occurrence on Minim Island, Santee Delta, Georgetown County, South Carolina, said “the plant appeared to be at home along the margin of a marsh’?7. That suggests the behavior of A. Keisak in southeastern Virginia. Both plants must be watched. Tidal marshes are scrupulously avoided by all except the most hardened of botanists; and even though A. nudiflorum may be a recent adventive from Asia which is rapidly spreading, A. Keisak may prove to be, as its behavior suggests, a conservative and ancient member of our flora. There is no question that the remaining 20 species which in Virginia are restricted to fresh river-estuaries are indigenous. They include many phytogeographic types: some are tropical American species, like Cladiwm jamaicense (“Saw-grass’’), reaching their northern limit on streams entering Currituck Sound, or Zizaniopsis, which comes farther north; others, like Ludwigia alata, are strictly North American but unknown in Virginia except along North Landing River; others, like Eryngium aquaticum or Lobelia elongata, are primarily southern but reach New Jersey, Delaware or Maryland; while some, such as Sagit- taria subulata and Lilaeopsis chinensis, are scattered from the Southeastern states to southern New England or Nova Scotia. Another series is prevailingly northern. Eriocaulon Parkeri, MAP 17 (member of a pantropical group of probably ancient dispersal from Antarctica), is on the fresh tidal mud of the St. Lawrence from above to far below Quebec, on tidal marshes of New England, New York, New Jersey, Delaware, Maryland and Virginia; while Elatine americana (map 18, excluding re- Ported stations in the interior of the continent) is on the tidal Teaches of the St. Lawrence, and on remote tidal muds from * Hotchkiss, Ruropora, xlii, 21 (1940). 514 Rhodora [December Northumberland Strait, New Brunswick, to the James, the known areas often 100 to 200 miles or more apart. Others, like [soétes saccharata and Aeschynomene virginica are chiefly on tidal marshes of the Delaware system and those confluent with Chesa- peake Bay; Rhynchospora macrostachya var. colpophila belongs in the tidal marshes of the Chesapeake area in Maryland and Virginia; Cassia fasciculata var. macrosperma is endemic in tidal marshes from the Mattaponi to the James; and three others are known only from a single series of tidal marshes each: Echinochloa pungens var. coarctata and Ammannia Koehnei var. exauriculata on North Landing River and Bacopa obovata Fer- nald (map 24) on the Chickahominy. If we were to follow northward, investigating the specialized floras of the different fresh tidal estuaries, we should find these conditions repeated: Bidens bidentoides on the Hudson, Dela- ware and Maurice (New Jersey) Rivers; B. mariana Blake on Northeast River and the lower Susquehanna in Maryland; B. Eatoni (with many localized and recognizable varieties along separate rivers) in the marshes of far-distant rivers from the St. Lawrence to the Hudson; B. infirma Fernald endemic on the St. Lawrence; Micranthemum (or Hemianthus) micranthemoides from the lower Hudson to the Potomac; Cardamine Longit Fer- nald (map 24) on tidal mud of Cathance River, Maine; Gentiana Victorinii Fernald (map 20) and Cicuta Victorinii Fernald on the St. Lawrence; and so on with several others. Whether primarily southern and failing to reach north to Virginia, or barely entering the state, or known northward to the Potomac, the Susquehanna or the Delaware; or more northern and known from the James or the Blackwater to the Penobscot or the St. Lawrence, the estuarine flora shows undoubted localization within the narrow ecological limits in which it thrives. Further- more, specific or varietal endemism is a regular feature of this flora. Restricted endemics, known from no other area, are found in the marshes of more than a dozen rivers from south- ern Virginia to the St. Lawrence; and from Maine to Virginia they show a marked preference for the smaller rivers and creeks with extensive swales, rather than the larger. In Virginia, 50 far as we yet know, the endemics of a single (rarely also on an adjacent one) river are on the North Landing River, the Chick- 1940] Fernald,—Additions to Flora of Virginia 515 ahominy and the Mattaponi, not on the James, the Rappahan- nock and the Potomac. “This is evidently due to the much greater development of marsh along the small streams (the larger rivers, like the James, having more open wave-washed and unstable shores) and to their naturally more circumscribed areas (the larger rivers tending to have a more generalized flora). When we consider the genera to which the species belong it will be seen that they are all wide-ranging or subtropical or tropical groups. The strictly holarctic genera are not repre- sented. * This fact, that the estuarine species belong in genera of semi-cosmopolitan, subtropical, tropical or extreme austral oc- currence (Hriocaulon and Lilaeopsis, for instance) is of im- portance, for we do not get estuarine floras well developed on the more northern areas of eastern North America. The St. Law- rence and the streams entering the Gulf of St. Lawrence are, ap- parently, the northernmost rivers with well developed estuarine floras, but from there to North Carolina‘ and beyond the estuarine floras become significant to the student of the flora. I have sufficiently emphasized the extreme isolation of these plants and their very limited tolerance of other conditions than those in which they grow. \ By some their dispersal, whether they be pantropical types or endemic American species, is satisfac- torily explained by saying “the birds did it”, just as by a geolo- gist of some renown I am told that the famous isolation of Coastal Plain plants about the head of Lake Michigan is wholly explained by. the presence there of a bustling commercial center, Gary, freight=cars:'and railroad-engines, to his mind, having transported the seed:. Not having the imagination to visualize railroad-trains dipping down into the Coastal Plain bogs and pools to secure the seeds of rare species of Psilocarya and other highly localized paludal and aquatic conservatives, in order to Plant them (many milleniums before Gary was ever thought of) in the bogs and pools of northwestern Indiana, I can hardly be Satisfied by so simple an explanation. So, cognizant of the many studies showing that migrating birds fly clean and that they are such expert aviators as not to carry on their long flights adhering chunks of mud to unbalance them, and that they sig most plants as food, not as altruistic spreaders of remote *Pibioties, I can accept the superficial and too easy explana- 516 Rhodora [December tion that birds are the chief agents which have brought about the present ranges of many plants, only in case of their very short flights and flutterings from one spot to another in close proximity. If birds have been the primary agents in dispersing our most conservative estuarine species, it seems very strange that we should have so many limited endemics, known only from the shores of single or of few rivers. As has been repeatedly shown, the overworked bird is scarcely to be taken seriously in this problem So, likewise with winds. The plants which characterize the estuary-flora are those of wet mud, inundated shores and drowned marshes. They are not plants of dry habitats. If seed-bearing portions get stranded and sufficiently dry to be picked up by wind, this must be a rare exception and not enough to account for the regular occurrence in so many fresh estuaries of the same species. The seeds of estuarine species rarely, if ever, have modifications to favor wind-dispersal. I have shown how, along such an extensive landlocked area as Back Bay, spread of the shore- and marsh-plants is a simple mechanical process, and that such a species as Juncus mega- cephalus, intolerant of much salinity, occurs back of the off-shore bars, islands and dunes, very regularly from Pamlico Sound to Back Bay. Although its two nearest relatives, J. scirpoides and J. brachycarpus, plants of stable sands, peats and clays and of more inland occurrence, follow north, in the former case to southern New York, in the latter to the local Tertiary beds of Massachusetts, J. megacephalus of fresh to but slightly brackish marsh stops its northern spread abruptly at Back Bay, beyond which the coast becomes open and exposed to the full saline in- fluence of the Atlantic?’. Right here, I believe, is the explanation of the great isolation in our fresh tidal river-estuaries of the dis- tinctive plants of warm-temperate, subtropical, tropical and subantarctic relationships. These plants are intolerant of the extreme salinity of outer coasts; they thrive in the area between high and low tide where the waters are at most only slightly brackish. They are an extremely conservative and fastidious element in our flora. The wide latitudinal range of this special- Extensive landlocked bays farther north, like en Sinepuxent and sawoman Bays, extending from Accomac County, Virginia, to Sussex County, _— may, when peopel explored, yield many of these so ie ies. know 1940] Fernald,—Additions to Flora of Virginia 517 ized flora along the margin of the northeastern United States and Canada, from southeastern Virginia (some of the species from Florida and the Tropics) to New Jersey, southern New York, the tidal rivers of Maine or even of eastern New Bruns- wick and, in some cases, the St. Lawrence from Lake St. Peter to below Quebec, calls for a condition comparable with that of Pamlico, Albemarle and Currituck Sounds and Back Bay today. Most geologists are agreed, I believe, that such a condition existed, all the way from Florida to the Gulf of St. Lawrence, when the continental shelf, now submerged off our Atlantic coast, was elevated as a nearly continuous outside ridge. That would make a tremendous southwest to northeast landlocked sound along the borders of which plants of fresh to merely brackish tidal shores could freely travel, just as today they are swash- ing and spreading on the changeable marshes and shores of Back Bay. The shores need have been no more stable than are those of Back Bay today; the exact stations of the plants need not have been fixed. The quality of the shore, tidal and fresh to brackish, not strongly saline, was the essential to success. To me this seems the obvious explanation. With the depression of the continental shelf the coast, especially north- ward, lost its outer fringe, the shores were bathed directly by Seawater and the long stretches of country between the fresh tidal reaches of the rivers and creeks lost the estuarine flora. It today exists as a relic of the period before the continental shelf became depressed. In considering when this migration northward along the land- locked sounds which extended to the Gulf of St. Lawrence, took Place it is pertinent to quote from the thoughtful study of our Coast by Professor Douglas Johnson. From his New England- Acadian Shoreline2® I quote: In Georgia and Alabama, exclusive of the Florida projection, that re of the Atlantic coastal plain exposed above sealevel has a breadth of 150 to 175 miles; in the Carolinas and Virginia it narrows to 125 “ Jounson, Dovetas, The New England-Acadian Shoreline. New York, John Wiley Sons. 1925. See especially pp. 296-302. 518 ‘Rhodora [December broadens off the Carolinas and: Virginia to 50-80 miles, reaches a breadth of 100 miles off northern New Jersey, and where wholly sub- merged off the increta ed Maine has a width of 150 miles or more. The increase is not u m, however, for the submerged part of the plain is unusually nathan appt the bight where Florida and Georgia meet, and unusually narrow in the Cape Hatteras region. At the southwest d as in an ce og so in the Gulf of Maine. Could we have a more striking eben of a single great ‘Gaia belt 150-200 miles broad, sub- rged progressively deeper and deeper toward the northeast, one of Pa pe et after another sre) gion from view, until all are com- aid buried under the ocea + wil appear from “this table ‘taat fore included! that ay margin of the Atlantic continental shelf (excluding the Bahama banks) is only a few fathoms below sealevel off Florida, is from 25 to 35 5 fathoms by the aervane of the e med ys oastal plain toward the northeast, is Sree confirmed ss the attitude of the = ayes of the se asimaaenats a from the B and to material removed from their summits by waves and cue Rogie doubtless flowed from — — = we ore 1940] Fernald,—Additions to Flora of Virginia 519 support from the fact that in the course of their operations on the Banks fishermen bring to the surface fragments of fossiliferous sandstone and limestone, A series of these collected and described by Upham, and e end of the Tertiary or still later in post-Tertiary time; for after the deposition of the late Tertiary sediments we must allow time for the figs of the lowland prior to its submergence. If the bevelled top of the cuest Plain beds (and perhaps also on the crystallines of the oldland), then resultant topography. Thus we should expect the subsidence to be at least post-Miocene, and more probably post-Pliocene. With this picture, so graphically presented by Douglas John- Son, of the northward depression of the Coastal Plain until, from Massachusetts eastward, it was completely submerged (except for the relatively slender and rapidly disintegrating Sable Island at the outer rim, more than 100 miles south of Cape Breton Island), it is easy to see what happened to the flora of fresh to but slightly brackish shores which, as I view the problem, freely Spread along the margin of the landlocked sound which, by the nal submergence of the Banks Cuesta in “nost-Miocene, and more probably post-Pliocene” was finally severed into scattered remnants The interpretation that the conservative plants of the estuaries of New York, New England, New Brunswick and the St. Law- rence from Lake St. Peterito*below Quebec can have persisted Somewhere in those regions* through the Wisconsin glaciation 8 distasteful to many who still hold to the archaic idea that 520 Rhodora [December Wisconsin glaciation eliminated all life from these areas. So many evidences exist indicating that with us the Wisconsin was a relatively weak phase of Pleistocene activity as compared with earlier ice-accumulations, and so many conservative animals and plants are now found in regions where it is most improbable that they have arrived, without leaving traces of their migra- tions, since the Wisconsin, that I find myself not at all dis- turbed. My views and much of the evidence on this question and the phenomenal bulk of parallel evidence accumulated by Nordhagen, Hultén and others in Scandinavia and elsewhere in Eurasia and in Alaska are presumably well known. They need no expansion here. The evidence added by the isolated colonies of estuarine plants lingering in New York, New England, New Brunswick and Quebec is a slight but important addition to the whole story. I have referred to Ranunculus hederaceus of wet sands of western Europe acting like a native on wet sands, tidal shores and about spring-heads in southeastern Newfoundland and on scattered points southward to Back Bay. It is simply one of many species which share western Europe and eastern America, especially Newfoundland. That considerable flora includes plants of mossy woodlands, acid bogs and other strictly natural habitats, plants which do not tolerate and can not spread by means of salt water. In addition to these plants numerous freshwater and land snails of native and undisturbed habitats show similar ranges. These and other cases, including some higher animals, are so numerous that it is absurd to imagine that they have been swimming the Atlantic in post-Wisconsin time, to find the natural habitats of Newfoundland, Gaspé and other areas within the latitude of Wisconsin glaciation. Their occurrence and their pre-Wisconsin spread has elsewhere been discussed and need not now divert us. Ranunculus hederaceus as well as Carex arenaria on the sands of Cape Charles may well be members of this illuminating group. With this discussion of the disrupted floras of fresh tidal shores I close the paper. Their study is only begun. Thousands and thousands of miles of shores of fresh to merely brackish sounds and bays and hundreds and hundreds of fresh tidal river-estuaries from Florida to Delaware are botanically un- 1940] Fernald,—Additions to Flora of Virginia 521 known. They will yield many new endemics. As I have re- peatedly said, there is plenty to do; there are few botanically equipped and with energy or initiative to do it. R} odora Plate 626 G. Fe a 5 FOGON PRAEMATURUS: FIG. 1, TYPE, X 14; FIG. 2, raceme, 1, Irom TY! ‘- 3, Upper hs { alf of raceme, * 5. from TYPE a DIVER iIENS { : | \ MARITIMUS FIG ay raceme. X 1. Rhodora Plate 629 Photo. H. G. Fernald. COMMELINA ERECTA: FIG. 1. mature spathe, « 2, Forma INTERCURSA: FIG. 2, flow ering tip, x 1; Fic. 3, int the 2, from C. ERECTA, Var. AN‘ GU STIFOLIA, a, forma CRISPA AC. ertspa) : FIG. 4, a x western Texas: Fi spathes, « 2, from Vi irgini a. C. ERECTA, var. eine A: FIG. 6, spathe, 1. Rhodora Plate 630 ’ SateAofyZ, irr ONO [4 oR Photo. H. CG. Fernald. (" é . ° 5 ws TY YMMELINA ERECTA, Var. ANGUSTIFOLIA: FIGS. 1 an portions of Michaux’s 4 Y ed . e ’ Md ee OF .C. angustifolia, < 1; FIG. 3, spathes, X 2, from Georgia. Rhodors Plate 631 Photo. H. G. Fernald. COMMELINA ERECTA, var. DEAMIANA: FIGS. 1-3, portions of TYPE, X 1; FIG 1, an inflorescence, < 1. Rhodora Plate 632 W. H. Hoda Ms LiumM C \TESBAEI: FIG. base, and Fic. 6, flower, X 4%, from aie al Cou nty, co orida » FIG. 7, capsule, i. Fcses South Carolina; Fic. 8, seeds, x 9; », trom. South Arolina ; FIG. 9, seed, 5. from South C her ag FIG. 10, seed, Koo, from Florida. Var. Lonen FIG. 1. lan folde 1 (with eee in lower left-hand corner), x 72. fom hse a FIGS. 2 and 3, capsules ah TOPOTYPE; FIG. 4, seeds, x 5, iTOm Rhodora Plate 633 Photo. H. G. Fernald : * " . gk and tUBUS PERNAGAEUS: FIG. 1, floricane and tip of primocane, x 1; FIG. 2, bud a a xed c aly r'X- lobes Rhodora Plate 634 Ph Oto, FH. Gq, Fernald. Rusvus PERNAGAEUS: FG. 1, floricane, * 1; FIG. 2, tip of primocane, x 1. Rhodor: Plate 6: ee Photo. W. H. Hodge SIA FASCICULATA: FIG. 3, fruits, x 1, from Tennessee: FIG. 4, seeds, X 3; siete Delon : : Var. MACROSPERMA: FIG. 1, pore - TYPE, X 1: FIG. 2, seeds, X 3, of TYP Var. ROBUSTA: FIG. 5, frui oe ak ae ‘OL X ‘dvBJANS-]Ro, LOMO] JO gouaosaqnd ‘eo ‘s “ULSABUU-JBO| ¥ ~ Sold » WINLEBGLO FY uBouul’y ul "T X qd AL jo uoTy..od ‘| ‘DId 3: VOS HAV] SILT A 7S ex dee 5 ee ie. i Peaapraviy sy Rhodora Plate 637 Photo. H. G. Fernald. ; m ‘tt1s Lasrus ar, SUBEDENTATA: FIG. 1, aaghote . lbiges x 1; Fic. 2, leaf see Maryland; Fic. 3 bene surface of leaf, x 10, from T Rhodora Plate 638 Photo, Ww. H. Hodge ~ SIDA INFLEXA: FIG. 1, TYPE, y 9 - ~ ~ IG. 3, lower surface of leaf. G. 5 FIG. 2, portion of stem 10, from TYPE; 1rOM TYPE: Fy Cape Fs ; 10, from TYPE; FIG. 4, calyx from the side. 0, ring of carpels, < 4, from above, trom ryPt Rhodora Plate 639 | H. H SIDA INFI ressed flowe1 2, fro DD sridge, Virgu FIGs. 2 ane 3 ripe carpe! 10, from Typ} ELLIOTTII: FIG. 4 ripe carpe 10, trom Florida a S. RUBRO-MARGINATA: FIG. 5. calvx and ring of rpels f, from ISOTY Sols e carpel 10, trom Florid:; S. NEO-MEXI f} » LINDHEIMERI Fi S, Tipe carpe 10, froy Rhodora Plate 640 Photo. H. G Fernald ee sha ASTEROIDES: Fics. 1 and ‘6 portions of TYPE, i; . 3, two heads, fre m egg ge soied FIG, oO a head, 2, to aheree ‘lit rece} ttacle and mature fri Wi 5. ripe ache Rhodora Plate 641 CRC OER RR Ne Te ae fe | Fn > LAMELEAEiAUAL EN LL . a inv variable’ ASTHROIE S, Var. GL SSTIFOLIA ; PIG. lant. * 2/5, from (¢ hickahoma) Ri FI nd 3, he: ads, 2. Kia. 4. » } lls achene, « 10. Rhodora Plate 642 Boron ty { "a AROLINIANA: FIG. 1, plant, « 2/5; Fic. 2, flowering head, x 2: Fic. ” Iruiting heads, » 2; Fic, 4, mature achene, « 10 Rhodora Plate 643 I oH. G2 ld BoLttoNta RAVENELIL: FIG. é original Ravenel material, « 2/5, TYPE at leit; rics. 2, 3 and 5, heads, * 2; ric. 4, mature acher ne, ). Rhodora Plate 644 NTERIOR: FIG. 1. TYPE, X 2/5; FIGS. 2 and 3, heads, « 2; ) Bouron1a DIFFUSA, Var. I ©. 4, mature achene, x 1( Le] Rhodora Plate 645 Vite ceereemmenee eI : - FIG. 3, flowering ~ Bout TONIA Let ae AMA: Fics. 1 and 2. paren of TYPE, Re head, . 4, fruiting head. « 2; Fig. 5, mature achene, < 10. Plate 646 Rhodora to. H. G. F rnald, Bouron 1, aia AMA, Var. RECOGNITA: FIGs. 1 and 2, portions of TYPE, X 1; FIG. 3, he: ad, — eae eines V; a. eto ae ALA: FIG. 4, group of mature heads, X 2; FIG. 5, expand ad, FIG, ature leak «x 10 Rhodora Plate 647 a. Gi 2 ASTER UMBELLATUS: FIGS. 5-7, head and involu “ s “4, : luer’ : : Var. BREVISQUAMUS: Fria. l, portion of TYPE, Fics. 2-4, heads and invé Var. PUBENS: FIGs. 8 and 9, involucres, « 4. 648 Plate hodora rYPe, { portions LATIFOLIUS: FIGS. 1 and Ey Va 1. ASTEE “AR YT MBELLATI 5, o~—, INnvoluers 5 Rhodora Plate 649 COREOPSIS LINIFOLIA: FIG. 1, IsorypE, X %: Fic. 2. head. & 2. from ISOTY FIG. 0, portions of two plants (perhaps topotypes), X 14. from Alabama; FIGS ichenes, < 10. a PE, 6 INDEX New scientific names are printed in full-face type Acalypha caroliniana, 461 ostyraefolia, 369, 461 Acerates floridana , 476 Aconitum, 356 uncin: natu um, 356 Aeschynomene, 394, 400, 402, 512 virginica, 385, 391, 393, 457, ee 512, 514 Aletris au a, 445 Alisma subulatum, 408 Amarant Amara ari Torey 369, 450 Amelanchier, 379 ee hium Muscaetoxicum, 362, Ammannia, 07 “on 367, 370, 467, 507, r. exauriculata, 510, 514 Ampelopsis arborea, 363, 370, 462 Amsonia , 475 Tabernaemontana, 359, 475, oe 475 Anacharis dé densa aon Sopra is, pen casinefolin, 471 nuda, 471 pulverulenta, 473 builvernicnte, 472 a, var. 8. pulverul Andropogon ec P enta, 473 eapillipes 384, 403, 415 divergens, 415, pl. 626 glomeratus tenuispatheus, 416 macrourus, y. hirsutior, 416 cei! 415, pl. 626 ie aturus, 413-415, pl. § Schizachyrium, 413 Heaps 357, 413, 414 ar. divergens, 413, 415 ubsp. maritimus, p. di- vergens, 4 , 384 virginicus, var. glaucus, 384, 403, 41 var. hirsutior, 416 enuispatheus, 416 . weer moe il tior, 416 " ec ttathenk. '384, 16 a 392, 394, 400, 5 , 391-394, 399, 100, 441, 0-513 nudiflorum, 512, 513. Anonymos uniflorus, 481 Antennaria solitaria, 363 Apocynum hypericifolium, 475 sibiricum, 397, 475 Arabis canadensis, 361, 452 Arisaema, 430 atrorubens, 430 f. viride, 430 f. zebrinum, 430 pusillum, f. pallidum, 430 Aristida lanosa, 369 Arum virginicum, 43 Arundo Dona, ton, 410 Asarum a Asclepias ate age 387, 476 var. pauper reula, 476 paupercula, 362 tataricus, 492 umbellatus, 493, 494, pl. 647 523 524 var. brevisquamus, 492, pl. f. flexicaulis, 493 var. latifolius, 493-495, pl. var. pubens, 493, pl. 647 Bacopa cyclophylla, 401, 402, 479, 511 Monnieria, 370, 508 obovata, 401, 511, 514 rotundifolia, 402 Baptisia villosa, 362 speoieie 39 trichosperma, 402 a, 498 asteroides, 396, 482-486, 491, 92, pl. 640 var. decurrens, 492 glastifolia, 486, pl. 9-491 interior, 490, pl. 644 easily ic 485, 486 urTens, 492 latisquama, 484, 490-492, pl. var. decurrens, 492 . “act aaa 484, var. paciciale: 491 recognita, 491, pl. paste aig 491 nelii, ‘484, 488, 489, pl. Breweria humistrata, 476 Bulbostylis ciliatifolius 362, 379, Burmannia biflora, 384, 446 INDEX Calopogon pallidus, 378, 380, 446 anna flaccida, 446 Cardamine, 357 Longii, 401, 514 Carex, 35 arenaria, 520 crus-corvi, var. virginiana, 359, Frankii, 371, 429 intumescens, 429 x louisianica, 429 louisianica, 359, 430 striatula, 363 Carphephorus bellidifolius, 365, 399, 481 tomentosus 362, 384, 481 r. Walteri , 481 362, Carya, olla var. pubescens, 361, 446 a, Cuceia, 394, 400, 512 Chamaecrista, 393, 456 fasciculata, 393, 456, pl. 635 10, 5 rand 456, pl. 635 Cartan } Marmuette, var. arcuata, pumila, var. habe 398, 447 os Margaretta, 398, 446 Ceanothus Fgh comm var. inter- medius, 4 intermedius , 462 eee '385, 471, 472 thyoi res 0 Chamaeliri uteum, 363 Chalcnthie le lanosa, 383, 404 Chelone Cuthbertii, 384, 479 abra, var. elatior, 479 bliqua, 363, 399, 479 Cherokee Gentian, 375 Chingapin, Chondrilla j juncea, 369 Chrysanthemum carolinianum, 482, Clei . Commelina, 386, 388, 435 INDEX angustifolia, 435, 436, 438, 439, 441, pl. 630 »~P crispa, 403, 435-437, 440 diffusa, 388, 391, 434 erecta, 435-438, ‘441, pl. 629 var. angustifolia, 435- 439, 441, - 630 f. albin f, crispa, "440, pl. 629 var. crispa, 435-437, var. Deamiana, 438, 440, pl. var. hamipila, 438, 440, var. typica, 438 ft ENG ee 439, 441 9 hamipila, 438, 440 longicauli s, 434 Nashii, 435, 437, 439 nudiflor a, 34 saxicola Swin iaenia, ” 436, 437, 439 eo ca, snare 439 7 inifoi, 496, 497, pl. 649 niscica: Con niseicarpa, 399, 496, 497 Kage 456 spidul. te rat rs a, ta Crantzia, 506 Crataegus s Marshallii, 498 Cunila origanoides, 363, 478 ren muralis, 479 51 96, 419, 505, globul ne ceo 419 erythrorhizos, 369 haspan, 420, 511 var. americanus , 376, 386, 419, 420, 441, 505, 506, retrorsus, var. Nashii, 369 rivularis, 394 f. elutus, 394, 419 var. elutus, 419 Cyrilla racemiflora, 379 Desmodium, 399 ‘a bellum, 399, 457, 504 usum, 399, 457 Disheonenn colorata, 370, 371, 420 _— teres, var. hirsutior, ’389, var. hystricina, 402, 481 Diospyros, 3 Di Be ay maritima, 370, 387, 409 Diplopappus comnifolius, 494 Doellingeria amygdalina, 494 umbellata, var. lexical 493 Drosera intermedia, rotundifolia, ee bien. 358 Eatonia aristata, 357 Echinochloa colonum, 385, 413 longearistata, 383, Wal Echinodorus subulatus, Elatine americana, 378, 386, 401, 466, 510, os Eleocharis, 5 albida, 370, 508, 510 halophila, 370, co ge 370, 508 parv cundrangulat tenu a 360, 420 PR arta phot se, , 404 hyemale, var. affine, 363 ayer hirsuta, 390, 410 aevivaginata, 391 hypnoides, 4 eregrina, Ericaceae, tribe ak 385 Erigeron bonariensis, 389, 495 vernus, . 0, 9, Eriocaulon oe ‘378, 386, 391, 399, 401, 432, 13 Eriogonum tomentosum, 449 Eriophorum virginicum, 390, 421 Eryngium _ ” 386, 441, 467, 510, 5 Eupatorium sald tics; 384, 399, 4 Euphorbia dentata, 461 heterophylla, 390, 461 526 polygonifolia, 400, 461 prostrata, 369, 461 Euphorbias, 369 Fox Grape, 390 Fraxinus {canine 475 microcarpa, 361, 475 poses ie var. Austini, 47 a r. lanceolata, 475 Froelichia saeaihe. 374, 450 Fuirena breviseta, 390, 421, 442 p Gaillardia pulchella, rg Galega spicata, 456, 457 Galium, 366 obtusum, 366 Gaura biennis, 467 Gentiana cherokenciic 375 linearis, 3 Stoneana, 362 icto ictorinii, , 514 Geranium carolinianum, 458 dissec Gerardia purpurea, 480 racemulosa, ne Gillenia trifoliata < Gladiolus sss ar 446 Glyceria arkansana, 360, 410 Grape, Fox, 390 psa bes S77, 382 Grass 383 Habenaria se oH agg 381, 446 cristata, 381, 44 abenaria Canby, 381, 446 ftanaaon, 380 virginiana, var. parvifolia, 380, Heliops Heliotropium curassavicum, 369, 477 ercaa ine ag aaa 382 375 Hemianthus micranthemoides, 514 Herpestis rotundifolia, 402, 479 lanceolata, 362 INDEX tenuifolia, 362 Humulus japonicus, 390, 447 scandens, 390, 447 Hydrolea meanders, 477 mondii, 466 466, Hypoxis hirsuta, 445 f; villosissima, 445 subgenus Ianthe, 374 leptocarpa 397, 445 Longii, 372-374 sessilis, 374 Hyptis mutabilis, 400 var. spicata, 400, 478 llex coriacea, 368, 400, 461 gla Ipomoea hederacea, var. integrius- cula I seats rhizomatosa, 370, 371, 450 Isoétes, 358, 364 Engelmanni, var. caroliniana, 358, 367, 375, 406 saccharata, 364, 391, 393, 405, 406, 510, 514 Juncus, js Se 379, 442 diffusiss apeiron ee 367, 370, 371, 5, 506, 508, 516 Roemerianus, oo seirpoides, 5 Kalmia latifolia, 471 v — a 471 Kicksia Elat Kuhnia espn Co 398, 481 Kyllingia brevifolia, 396, 419 Labiatae, 400 lasaeanien anceps, 381, 4382 Lactuca canadensis, var. longifolia, 368, 4 a f. leucantha, 498 INDEX Lapsana communis, 498 aes aaa "397, 400, 411 mna perp pias Pg fasciculin 390 Lespedeza angustifolia, 458 aah capitata, subv ee cress 457 oblon cate lia Leucothoé axilla aris, oc —— bellidifolia, ‘481 uciflose ulosa, 4 496 alteri, 3 Lilaeopsis, 371, 373 506, 515 carolinensis, 367, 371, “470, 506, ee 391, 470, 506, 510, Lilies, 381, 444 Lilium , 403 Citesaci, 381, 448, 444, pl. mae Longii, 443, 444, pl. os Michau Lipp a nodior, 370, 389, 478, 510 Laroauattas Litsea piace 362, 397 Lobelia elongata, 511, 513 62 ine, 365 Ludwigia alata, 510, 513 - revipes, 508 om chinense, 479 Lycopodia, 359 , 405 Lycopodium, 358 adpressum, f. polyclavatum, a 358, 359 te ssum, f. poly- um, 405 carclinisnten, 358, 359, 405 Inundatum, var. adpressum, 358, f. polyclavatum, 405 var. er pee 405 Lona ee ~Ste 358, 359 ysimachia deals 364, 474 quadrifolia, radicans, 366, 367, 397, 474 527 strict = 364 Lythraceae, 507 ya nineare: 370 Manisuris, 382 Matricaria apr 482, 485, 486 glasti Medicabo ren ad var. glandulosa, Melanthium virginicum, 381 Micranthemum pe ies Fed 514 Najas quadalupensis, 360 Nelumbo, Nicotiana Tabacum, 400 Nuphar fluviatile, 451 canes aquaticum , 475 Nyssa vatica, var. dilatata, Onosmodium virginianum, 362, 477 Orchid, 359, 363 Orchids, 380, 381 Orontium, 393 Osmanthus americanus, 374 Panicum agrostoides, 386, 412 var. condensum, 391, 412 albomarginatum, 364 amarum, 396 caerulescens, 412 capillare, condensum, 386, 412 cryptanthum, 364, 383, 412 laxiflorum, 383 — 379 strigosum, 35 Paronyehia ine 450 fastigiata, var. ener 361, 450 riparia, 450 Paspalum Urvillei, 411 83, 396 aris carolinia: a, 370 Phlox x Hentai 309, 476, 477 nivalis, 4 6, 477 528 Phryma, 512 Physalis angulata, 371, 478 pubescens, 479 Pin Oak, 361, Pinus palustris, 365 i 2 60, 461 cae ong 362, 363, 458, 459 obtusata, 363, 459 61 vertcllat, 4 459-4 ambigua, 46 ssoeyela, 459 Fob eoaila, 403 articulata, 379, 401, 403, 449 paces var. asperifolium, 448 = ye vic: iliatum, 448, var. eich 448, 449 ‘c m, piccuantyues 448 sanibalense, 448 polygamum, 449 alps 389 Sad vod nthieva ehh 363, 446 Potamogeton bupleuroides, 387 ep: ay ce , 409 var. Nuttall, 400, 407 : 00, 407 Spirillus, Potentilla satin 377, 453 rect: Prenanthes altissima, 363, 498 Psilocarya, 515 8 eitpoides, var. Grimesii, 384, Pycnanthemum, 390 vertici Pyxidanthera, barbulata, 474 Quercus Bushii, 383 Catesbaei, ‘447 cinerea, 362, 365, 403 x falea ata, 403, 447 nevis, 362, 365, 44 447 INDEX -prarsty 2mm 383 velutina, 447 palustris 361, 447 velut x bet be hii, 447 < Quercus subintegra, 403, 447 Guillwort 357 einer bulbosus, 451 r. dissectus, vote 452 typicus, valde pubens, 451, 452 dabaibcte f. ae 5 er 508, Lane Rhesis cles , 319 Rhododendron canescens, 361, 363, 367, 512 axillari 422-425, 427, 428 biceps, “ys caduca, 363 eapiteliata. 427, 428 cephalantha, 366, 422, 423, 425, 427, var. pleiocephala, 423, 424, var. typica, 423, 425 chalarocephala, 426-429 dodecandra, 429 § Eurhynchospora, ser. Glo- meratae, 4 fascicularis, sgereiene: 427, 428 “ "minor, 426, 427 a 386, 393, 401, 422, 4 var. colpophila, 421,422, 441, 510, 514 mierocephla, 366, 422, 423, pallida, 378, 381, 429 perplexa, 382, 429 Rhynchosporas, 38 ntralis, 455 INDEX Enslenii, 454, 455 geniculatus, 453 Grim: i, 45 x ann, 361, 454 pauxi 8, 454 parnaenous, 454, 455, pls. 33, 634 § Procumbentes, 454 2 oan 454 Rubdeckia Helionsidis 375, 378, 382, 496 6 a, var. a, 375 “Bp, 375, 496 f. villipes, 496 Ruellia strepens, f. dleistantha, 480 Rumex oe la, var. tegri- folia, 448 os a 389, 474 cue , cleistantha, 474 aris Saciolepis ¢ shay 393 Sagittaria, 387 alata, 391, 509 lor. a, 409 “oa 408 var. gracillima, 408 lorat; var. lorata, 408, 409 pusilla, 4 subulata, = 399, 407-409, os relia 408 natans, 108 var. natane, 387, 407, 408 Wea:heshiana, "359, 400, 409, indices te a minor, 390, 4 Sanicula canadensis, 481-470 amplissimo, etc. 468 var. floridana, 468 enuina, 467 “ee “ gran typica, 467 europaea, 469 florid 68 marilandica, 468-47: Sarracenia, 379, 389 : ava, 359, 362, 374, 389 Sassafras albidum, 452 529 Saw-grass, 513 Schnetaa syillaris, 422, 423 Scirpus atrovirens, var. folius, 359, 420 debilis, 390, 420, 442 divaricatus, 400, 420 fontinalis, var. virginiana, 386, flaccidi- robustus, 370, 504 Scleria minor, nitida, 365 Sedges, 381, 383 edum alboroseum, 453 Serinia oppositifolia, 4 98 Sesuvium maritimum, 369, 387, 450 Setaria viridis, var. Wei inmanni, 413 Seymeria cassioides, 379, 445 sia conan, 382, 383, 464, 465, B. To 465 graci inflex a, 463-465, 638, 639 iindheneer 465, - 639 neo-mexicana, rubro-marginata, 164, 465, pl. Reathae ss Jauriolis 37 t Solidago Elliottii, 482 fistulosa, 482 ulmifolia, 482 Soy Bean, 3 Spartina ‘alterniflora, var. pilosa, cynosuroides, 370, 411, 504 patens, hee Lage 396, 411 Spergularia ma Sikeoophole intermedi, 357 , 396, 357 Spiranthes cernua, 394 rar. odorata, 359, 394, 1, 446, 498 odorata, 359, 446 Sporobolus asper, 400, 411 Steironema radicans, 366 Stellaria media, var. glaberrima, prostrata, 451 530 Stewartia Malachodendron, 372, tlegi Stillin ca aed Stiptlieds re Se Strophostyles ie f. ochro- leuca, 4 Styrax americana, 377, 397 Taxodium, 406, 498 ascendens, 406 distichum, var. imbricarium, 7 eanuges 308, 456, 457 mitonsa, 456, 457 Petwigenothers eee 362 elianthoides, 356, 367, 495 Tillandsia ee es, 369 Tofieldia racemosa, 381, 442 Torilis japonie 4 0 Tradese: rosea, wat graminea, 379, 397, 44 ‘Pelueenit OE 397, 478 setaceum, 397, 478 Trifolium pratens e, f. pilosum, 456 Triglochin striata. 387, 388, 407, Trillium lanceolatum, 445 Trilliums, 375 Triodia Chapmani, 378, 383, 410, 375, 378, shia! i 410, 411 hapmani, 410 Typha, 374 iste 374, 376, 407, 505 Umbelliferae, 371 Utricularia, 384 biflora, 387, 480 cleistogama, 384 INDEX fibrosa, 403, 480 juncea, 384, 480 subulata, 384 cle eistogama, 384 virgatula, 384, 480 Vaccinium, 379 ar reum , 863 crassifolium, 378, 379, 474 Vallisneria ee 387, 409, 507 Verbena officinalis, 4 m, 363 Viola emarginata, var. acutiloba, culenta, 400, 466 striata, oba, ’ var. dilatata, 363, 466 Vitis may - 0, 463, pl. 636 var. ede dant 462, 7 ? Water Lilies, 360 Wild Celery, 507 Woodsia obtusa, 383 i bie ambigua, 432 renicola, 379, 434 eercisinias 396, 399, 433, 434 f. flacci ida, 433, 434 difformis, 369, Prd 434 flexuosa, 379, 4 Zenobia, 385, 397, 403, 472 cassinefolia, 403, 471, 473 pulverulenta, 378, 385, 397, 471, 473 ies 38 gla berrimus, 374, 381, 442 Zanini, 363, 513 cea, 510 _ CONTRIBUTIONS S FROM THE GRAY HERBARI OF HARVARD UNIVERSITY Reprinted from Ruopora, Vol. 43, January and February, 1941 CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY No. CXXXIV REVISION OF THE GENUS PTERIDIUM R. M. Tryon, JR. DatTEs oF IssuE Pages 1-31 and Plates 650-662. .............---++.s2+>: 15 January, 1941 \: SPOlend Pisses 8 February, 1941 CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY—NO. CXXXIV A REVISION OF THE GENUS PTERIDIUM R. M. Tryon, Jr. (Plates 650-653) INTRODUCTION THE only species of the monotypic genus Pteridium, P. aqui- linum, is one of the most familiar plants in the world. Its distri- bution is world-wide in temperate and tropical regions and it is Common to the point of becoming a weed in most of the regions in which it grows. Many authors complain of its weedy char- acter in the British Isles, northwestern United States and New Zealand. In these regions and probably in others it invades fields and most especially pastures and it is probably the most obnoxious native weed with which the farmers have to deal. Apparently, it immediately assumes the role of a weed as soon 48 man tries to turn to his own uses any land that it finds suit- able for its growth. However, the weedy character seems to vary With the variety, or region, for var. latiusculum and var. pseudo- Caudatum in eastern North America only rarely, if ever, become 4 serious menace to farmers. Among the many species of ferns that spread by means of thizomes or runners, the Bracken is particularly well adapted to the role of a weed. In favorable soil conditions there is a widely creeping, deep, main rhizome from which lateral, ascending branches arise. These branches usually rebranch and produce 2 Rhodora [JANUARY the fronds fairly near the surface. A dormant bud is produced at the base of each frond so that each year the plant has a po- tential second set of fronds. If, through burning or plowing, the lateral branches are destroyed, the main deep-seated rhizome can continue the life of the plant. A few excerpts from a recent article on the Bracken in Britain’ will illustrate its economic importance as a weed and some of the methods taken to control it. ‘Bracken ranks as the most plentiful and widespread of British ferns. It is generally dis- tributed in most rough grazings and in many old permanent pastures in almost every part of Britain. It is plentiful in most open woodlands and on the sites of formerly wooded areas. .. . Many of the older writers have dealt with the loss of grazing due to Bracken. . . . ‘I have seen the Roots of it in some Grounds, eight Foot deep. The best cure is often mowing of it while in grass. If you plow it up, plentiful dunging of it and Ashes are very good: but the certainest cure for it is Urine.’ J. Mortimer, The Whole Art of Husbandry, 2nd Ed., 1708.” At the present time, the authors recommend cutting the fronds when they are fully mature and following this with heavy grazing; and also the use of phosphate and lime fertilizers to build up the ground and make it unsuitable for the growth of the Bracken. Sodium chlorate and sulphuric acid are recommended as killing sprays. “Tn closing this paper, however, the writers desire very definitely to record their view that the bracken menace has grown so rapidly, and has now attained such proportions, that it is of real national importance: at the International Grassland Congress of 1937 one speaker referred to it as ‘the ghost stalking silently at our side, which nobody dares to discuss.’ . . . A million pounds sterling devoted to the reduction of the area infested by bracken would be well spent.” Many species of fungi are parasitic upon Pteridiwm aquilinum? and attempts have been made to use them in its control. Fusarium sp. and Corticium anceps (Bres. and Syd.) Gregor have been tried? but due to the fact that the host seems to be 1 Long and Fenton, The Story of the Bracken Fern. Journ. Royal Ag. Soc. Eng. 99: 15-36 (1938). 2See Seymour, Host Ind. Fungi N. Am. (1929) and Faull, Gen. Urediniopsis. Contrib. Arn. Arb. 11 (1938). 3 See Gregor in Verdoorn, Man. Pterid. 146 (1938). 1941] Tryon, Jr..—Revision of the Genus Pteridium 3 more durable under varying environmental conditions than the fungus, they have not met with success. To compensate for its economic liabilities, the Bracken has made itself available for many uses, though at the present time none are of real importance. The fronds have been used for stock-food, when mixed with grass; as silage; for packing fruits; in the place of hops in brewing; medicinally; in glass making and as litter for both man and stock. The fronds and rhizomes have been used in thatching and the green fronds burned as a source of potash. Nakai* says that in Java the fronds are laid on racks over the Cinchona beds to protect the young plants from the sun and that: “This shelter is not liable to rotten and decays harmoniously with the growth of the Cinchona. When Cinchona reaches to the height of the roof, it want the shelter no more, and the fronds are so good and perfect as they then become bones only, dropping all of the leaves.” The rhizomes have been used in brewing a beer and also as swine-food. It is reported that the swine prefer the rhizome plus soil. The hairs rom the base of the petiole of var. caudatum are used for stuffing pillows and the fronds are used as packing in the tops of char- coal bags. A solution of the boiled fronds has been used effec- tively as a rose spray. Bracken is often accused of being Poisonous to stock but apparently this is true only when it is eaten exclusively and in large amounts. The rhizomes, which contain a considerable amount of starch, have been used as food, at one time or another, in Switzerland, France, Canary Islands, western North America, Society Islands, New Zealand, Tasmania, New South Wales and China. It is said that formerly it was 4n Important food-item of the Maoris of New Zealand. One doctor js reported as using the starch in place of arrowroot. However, the ordinary rhizome would undoubtedly be poor fare and Probably was usually eaten only to stave off actual starva- tion. Forster’ says that: “Radices ab incolis inopibus et famelicis exsuguntur, insipidae, parum nutrientes lignoso-fibrosae.” The ‘roziers have been used as “asparagus” in Japan and western North America. A number of botanists have reported eating the broiled croziers and say the flavor is quite agreeable. I can , go. ‘Nakai, Critical Notes on Japanese Ferns. Bot. Mag. Tokyo 39: 110 (1925). orster, Pl. Escul. 74 (1786) 4 Rhodora [JANUARY attest this myself but must add that the dense pubescence is rather troublesome. In the Canary Islands, Orobanche trichocalyx (Webb & Berth.) G. Beck and O. Schultzii Mut. are parasitic upon Pteridiwm aquilinum. REPRODUCTION Vegetative propagation by means of the creeping rhizome is the predominate method of reproduction of the Bracken, al- though in favorable localities ordinary sexual reproduetion is quite as effective. Although the Bracken habitually grows in rather dry places these are apparently not suitable for its normal reproduction, the prothallial stage developing only in relatively damp situations. Long and Fenton® say that in the eastern side of Britain, the prothallial stage is rare or unknown, while in the western side, where there is considerably more rainfall, young plants are not uncommon. In eastern North America young plants are only rarely seen. The young stages are not common as herbarium specimens. It is an indication of their relative rarity that at least two species and one variety have been described on the basis of young plants, although their authors no doubt little suspected they were redescribing the familiar Bracken. If, in the past, the Bracken has reproduced largely by vegetative means, this might well explain its present vigor and the lack of specific segregation within the genus. Judging biological age by the number of generations, it would still be a relatively young species. NECTARIES Nectaries are present at the base of the lower pinnae. These were apparently first noticed by Charles and Francis Darwin’ but they have been mentioned by few authors® since then. I had an opportunity to observe them on var. latiusculum in the Har- vard University Botanic Garden in the spring of 1940. On the sixth of May, the basal pinnae were just starting to unroll and 6 Op. cit. 7 Francis Darwin, Journ. Linn. Soc. 15; 407, Tab. S P deg : (1876). 8 The only other original observation I have seen Victorin, Les Filicinées du Québec. Contrib. Lab. Bot. Univ. Montréal No. 2, et adh who says: “Au printemps les nectaires basilaires des segments sont presque toujours visités par une grosse fo urmi noire.” 1941] Tryon, Jr.—Revision of the Genus Pteridium 5 the nectaries were clearly visible at the base of the pinnae as dark brown, smooth, somewhat swollen areas on the dorsal side of the then densely pubescent rachis. As the fronds continued to unroll proportionately smaller nectaries could be seen at the base of the second and third pair of pinnae. These nectaries were functional in that they exuded a liquid in the form of droplets from small pores. The nectaries at the base of the second pair of pinnae were not as active as those below and those at the base of the third pair of pinnae seemed to be entirely inactive. Large red and black ants were attracted by the exuda- tion and observed in considerable numbers feeding upon it. Al- though an analysis was not made, the fact that ants feed upon the liquid undoubtedly indicates that it contains sugar. A free- hand section of the nectary area showed parenchymatous tissue but there was no evidence of glands. For these reasons I think it is likely that the exudation is neither a guttation of nearly- solute-free water nor a modified secretion but rather an excretion of the stipe-sap. The nectaries are active in the sun and also in the shade. They were quite active on a cloudy humid. day. They were observed to remain active until the fronds were almost completely unrolled and may be active for a longer period of time. By midsummer they have become quite insignificant and entirely inactive. I believe this phenomenon is physiological, perhaps related to root pressure. PRIMITIVE AND ADVANCED CHARACTERS P hylogenetically considered, Pteridium has a curious combina- tion of relatively primitive external characters and relatively ad- vanced internal characters. In the Pteridoideae it is considered to be a primitive genus? on the basis of the following characters: Presence of a more or less vestigial inner indusium; initial basi- Petal succession of sporangia; hairs, but no scales, present on the rhizome; equal dichotomy of the axis in its early develop- ment; and open venation. On the other hand, the highly de- Veloped vascular structure of the stem and petiole and the Presence of true vessels! indicate an advanced condition. The highly developed internal structure, however, is undoubtedly, at least in part, related to the relatively large size of the plant. Fy Bower, The Ferns. 3: 41-45 (1928). liss, The Tracheal Elements in the Ferns. Am. Journ. Bot. 26: 620-624 (1939). 6 Rhodora [JANUARY HistoricaL AccoUNT Most of the taxonomic work on the Bracken has been done by authors working in a limited area or on a limited number of forms. There have been only a few comprehensive treatments. This has resulted in a general tendency to treat the various groups as species and to recognize, in one rank or another, trivial varia- tions, a condition which naturally has led to a multiplicity of names. Not only has it been unnecessary to describe any new varieties in the present treatment, but in 1839 Agardh likewise found all of the groups he recognized already named. Most of the critical work has been done by local authors. As in the case of so many groups, Linnaeus, Sp. Pl. 2 (1753), laid the foundations for the present treatment. He recognized, in the genus Pteris, two species, P. aquilina and P. caudata, which represent the ssp. typicum and caudatum of this treatment. Willdenow, Sp. Pl. 5 (1810) recognized, besides Pteris aquilina and P. caudata, some additional species described since 1753: P. capensis, P. esculenta and P. lanuginosa. Under P. aquilina he recognized, in var. 8, a combination of var. latiusculum and var. pseudocaudatum of this treatment, but most later authors did not take up this segregate. Agardh, Ree. Pterid. (1839) was the first author really to study material from most of the regions of the world and his treatment is in close agreement with the present one although he regarded most of the groups as species. He followed Linnaeus in treating the Brackens under the genus Pteris, but set them apart from the other species as the section Ornithopteris. He recognized P. esculenta, P. arachnoidea, P. caudata and P. decomposita and applied the names in essentially the same sense as in the present treatment. In P. semihastata and P. recurvata he recognized, respectively, var. yarrabense and var. Wightianum. Under P. aquilina he mentioned, but did not describe, var. pseudocaudatum. In the main his treatment differs from the present one only in the recognition of P. lanuginosa and in uniting var. latiusculwm with P. aquilina. Later authors would have done well to follow Agardh’s treatment more closely. Hooker, Sp. Fil. 2 (1858) presented the next, and really the last, comprehensive treatment of the genus. He followed Agardh in treating the species under section Ornithopteris of Pteris, but ss 1941] Tryon, Jr..—Revision of the Genus Pteridium “g except for P. coriifolia and P. psittacina, which he says he did not understand, having seen no specimens, he reduced all names under P. aquilina and recognized only varieties. The present treatment agrees with his in this respect, but his varietal lines were rather poorly drawn. All varieties in the present ssp. typicum he treats under var. glabra, fronds glabrous or nearly so beneath, and var. lanuginosa, fronds quite pubescent beneath. In the present ssp. caudatum he recognized var. caudata, apply- ing it in the present sense, and var. esculenta, including the present var. esculentum, var. arachnoideum and var. yarrabense. This treatment was long followed and led especially to a confusion of var. esculentum of Polynesia, Australia and New Zealand and var. arachnoideum of South America. Diels, Nat. Pfl. 14 (1899) and Christensen, Ind. Fil. (1905) recognized the single species Pteridiwm aquilinum. Nakai, op. cit. (1925) in a rather rambling and non-critical discussion, made several new combinations under Pteridiwm, reviewed much of the literature and discussed the status of various species and varieties. Under various names, Scopoli, Fl. Carn. (1760), Gleditsch, Syst. Pl. (1764), Newman, Phytol. 2 (1845) and John Smith, Hist. Fil, (1875) separated the present genus Pteridium from teris, but it was not until Kuhn in v. d. Decken, Reis. Ost.-Af. 3° (1879) took up and defined Pteridium that the Brackens were widely treated as a distinct genus. TERMINOLOGY Certain terms have been used in the descriptions in a strictly limited sense and certain unusual terms and characters have been used. These are explained below: The vernation of the frond equal, all of the pinnae becoming Unrolled at essentially the same time; swbgleichenioid, the tips of the upper pinnae still inrolled while the lower pinnae are com- pletely unrolled; or gleichenioid, the basal pinnae being com- pletely unrolled while the third or fourth pair of pinnae are ‘ompletely inrolled. _ Acosta is the midnerve of a pinna. : A costule is the midnerve of any segment of lesser order than ot with the exception of the midnerve of the ultimate seg- *nts, which is called the midnerve. 8 Rhodora [JANUARY In some varieties there are lunate or sublunate, entire lobes along part of the rachis, costae and costules, between the divisions of the blade. These are called free lobes. Piater 650, Fic. 10. The ultimate segments are adnate, broadest, but not dilated, at the base (pL. 650, ric. 7) ; decurrent, dilated at the base toward the base of the midnerve (pu. 650, Fic. 16); surcurrent, dilated at the base toward the tip of the midnerve (PL. 650, FiG. 3); or narrowed at the base (PL. 650, FIG. 6). The upper surface of the ultimate segments includes the mid- nerve but excludes the margin. The margin, although morphologically on the upper surface of the segment, is described separately. It is the chlorophyll-bear- ing tissue closest to the base of the outer indusium. Peculiar, irregular, wing-like epidermal outgrowths occur on the veins and midnerve on the lower surface of the ultimate seg- ment in some varieties. These are called membranous wings. PuaTe 650, Fia. 14. Some varieties have a farinaceous appearance on the lower surface of the ultimate segments. This is a minute epidermal outgrowth, occurring between the veins, which, under high magni- fication, resembles a sparsely branched coral fungus. It may be an excretion but it is not affected by alcohol or dilute acid. PuaATE 650, Fic. 13. The term indusium is applied to the outer, functional indusium, the modified margin of the segment. The outer indusium is pres- ent in both fertile and sterile fronds and throughout partially fertile fronds. If it covers sporangia it is spoken of as the fertile indusium, if not, as the sterile indusium. The inner i- dusium is non-functional and usually nearly obsolete. When the inner indusium is meant it is called such. The costules and lower side of the midnerves in ssp. caudatum are described as pubescent with white, dark or bicolorous hairs. The white hairs are long, thin, terete, whitish, multicellular hairs with oblique cross-walls. These are essentially of the same type that occur on the tissue on the lower surface of the segments. The dark hairs are rather large, stout, dark or reddish brown, usually terete, multicellular hairs with straight cross-walls, which are similar to those that occur on the rhizome. As these mature they often become moniliform, with whitish walls, only the septae 1941] Tryon, Jr.,—Revision of the Genus Pteridium 9 remaining brown. These are called bicolorous hairs. When one of these dark or bicolorous hairs breaks off, it often leaves a swollen base that is dark reddish brown and has the appearance of a sessile gland. SYSTEMATIC TREATMENT Although Pteridium is world-wide in distribution and the vari- ous populations show very strong phytogeographic affinities, and although it is undoubtedly a rather old genus, three lines of evi- dence lead me to believe that specific segregation has not taken place. First, in all areas where two varieties of the same sub- species overlap in range intermediate specimens are found (with the exception of var. africanum in ssp. typicum). Such inter- mediates are not common but they are of such frequency that, assuming indiscriminate rather than critical collecting, they must form a percentage of the Pteridiuwm population of the area that cannot be overlooked. Second, in areas occupied exclusively by one variety, individuals, perhaps representing local populations, occur that have one or more characters of some other variety. That is, characters of one variety crop up occasionally in other varieties. In some cases a specimen will resemble another variety 80 closely that it would certainly be identified with it except by using the geographic “character.” In other words, the critical characters of the varieties are not thoroughly stable. Third, the critical characters show little constancy within the group as a whole. That is, ciliation of the indusium in var. Feei, or non- ciliation in var. africanum, are critical characters but both con- ditions are commonly found in var. caudatum; pubescence on the undersurface of the blade in var. pubescens and the lack of it in var. latiusculum are critical characters, but var. Wightianum, While most commonly pubescent, is not infrequently nearly glabrous; pubescence on the margin of the segments in var. Feet and the lack of it in var. typicum are important characters, but var. Wightianum frequently exhibits both conditions; and the free lobes in var. arachnoideum and their absence in var. cauda- tum are diagnostic characters but both conditions are found in var. yarrabense. It seems to me that, when the critical charac- ters of a group show such inconstancy, they are materially weak- ened for use in specific segregation. 10 Rhodora [JANUARY The treatment of each variety follows a fixed order somewhat different from the conventional one. First, the accepted name and its synonymy and then any necessary discussion of the nomenclature. Second, the description of the variety followed by a discussion of taxonomic problems and a comparison with closely related varieties. Third, a short phytogeographic discus- sion, if desirable, a statement of habitat preference and a general statement of range. Last, a citation of specimens. In this ar- rangement, the critical discussion of nomenclature and taxonomic problems follows the synonymy and description, respectively; and the usually long citation of specimens comes at the end. This study is based on the 2,500 sheets of Pteridiwm in the Gray Herbarium, the Herbarium of the New York Botanical Garden, the United States National Herbarium and the Her- barium of the Field Museum of Natural History. DESCRIPTIONS The critical characters in the descriptions have been italicised. The descriptions are based on mature specimens and do not take into account juvenile forms. The cutting of the blade is the- same in the juvenile state of all varieties. The margins of the segments are flat, not revolute, and there is no sterile indusium. None of the various epidermal outgrowths are present except pubescence, and that is only weakly developed and not distinctive. Mars The maps of the varieties have been compiled primarily from the cited specimens. These localities have been augmented by localities taken from other specimens examined and in a few cases by localities taken from the literature. These are used only when a specific locality was mentioned and where there was no doubt as to the identity of the variety. Localities based on specimens are indicated by a dot and those based on a printed record by a cross. The map of the species includes all of the localities on the varietal maps and in addition several localities taken from the literature where the identity of the variety was in doubt. On this map no distinction is made between localities based on specimens and those based on the literature. No at- tempt has been made to make the maps absolutely complete. 1941] Tryon, Jr..—Revision of the Genus Pteridium 11 The series of specimens studied has in most cases been sufficient to plot the general range of each variety. The maps used have been selected from Goode’s Series of Base Maps, Henry M. Lep- pard, Editor, Copyright by the University of Chicago, published by the University of Chicago Press. CITATIONS OF SPECIMENS The citations of specimens have been limited, when necessary, to two or three from each minor political division, i. e. state, province or department. The most widely distributed collections are cited in preference to the more locally distributed ones. Full citations are given in the discussions only for specimens that are not listed in the formal citations. Herbarium-sheet numbers, such as, Herb. Field Mus. no. 47805, are used in cases where the data on the label is insufficient to identify the specimen accu- rately. The usual abbreviations for the herbaria are used: F, for the Herbarium of the Field Museum of Natural History, Chicago, lll.; G, for the Gray Herbarium, Harvard University, Cambridge, Mass.; NY, for the Herbarium of the New York Botanical Gar- den, ae Park, N. Y.; and US, for the United States National Herbarium, Heshean tan Institution, Washington, D. C. In the eh where other herbaria are cited, abbreviations are not used. ACKNOWLEDGEMENTS I wish to thank Dr. P. C. Standley, Dr. H. A. Gleason and Dr. W. R. Maxon for the generous loan of material and for privileges extended at the herbaria under their care; and es- pecially to express appreciation to Mr. C. A. Weatherby, without Whose constant help and encouragement this study could scarcely have been completed. SYSTEMATIC TREATMENT PTERIDIUM Gled. ex. Scop. Fl. Carn. Ed. 1, 169 (1760), nomen wl eee ; Kuhn in v. d. Decken, Reisen in Ost- 879). Trevis fey Soe. Ital. sc. nat. 17: 239 (1875). Asplenium Bernh. chrader’s Journ. 1799": 309 (1799), in part. Allosorus Bernh. 12 Rhodora [JANUARY Schrader’s neues Journ. 1*: 36 (1806), sensu Pr. Tent. Pterid. 151 (1836), in part. Paesia St.-Hil. Voy. Brésil 1: 381 (1833), sensu Moore, ward, Meets 878 (1858), in part. HEupteris Newm. it 296 (1899), in a: Pustoowina Farwell, Am. Mid. Nat. 12: 290 (1931). Although it is generally considered that Scopoli took up Pteridium Gled. for the Bracken, as a segregate genus from Pteris L., and it would therefore be the earliest valid generic name under the International Rules, some have argued that he proposed it as a substitute name for Pteris L. In order to meet this argument, Pteridium has been conserved (International Rules of Botanical Nomenclature, 131 (1935) ). Rhizome subterranean, extensively creeping, repeatedly ified margin of the segment, and the inner indusium;" re- ceptacle a vascular anna connecting the vein-ends, the inner ve i arisin ng at its inner side ; spores brown, very finely spinulose, t obo 1 by a single world- wide sori found in er laniperate and tropical regions. TERIDIUM AQUILINUM (L.) Kuhn in v. d. Decken, Reisen in oat aoe 3°: 11 (1879). Marl. Bracken. Characters of the genus. Key To THE SUBSPECIES AND VARIETIES” a. Ultimate segments adnate, or equally decurrent and sur reur- low nose or absent. (ssp. typicum) b. b. Fertil hee sterile indusium ciliate e, or pubescent on the surface, or both; ultimate segments densely or pean pubescent beneath between the margin and the midnerve; pinnules nearly at right angles to the costae. Cc. 11 The inner indusium is quite variable in its structure. In sus specimens it is @ continuous membrane, as figured by Luerss. in Rabenh. Krypt. Ed. 2, 3: 101, fig. 81 and Bower (after Luerss.), The Ferns, 3: 43, in others it is an pulsars broken mem- some dire: only of a few hairs 12 The user can expect that a — ember of intermediate and non-typical speci- mens, especially aus ke will not run down satisfactorily. 1941] Tryon, Jr.,—Revision of the Genus Pteridium 13 c. Pinnae and pinnules long-acuminate; a Ppa faleate or pagel: pon indusium 0 less wide, the sterile 02 r less wae “andi Java, Ne fee Wightianum, p. 22 c. Pane sand pinnules short-acuminate to Pastis ; ulti mate segments straight; fertile indusium abrous, or if pubescent, the margin gla- e. e. Rachis more or less pubescent; blade lanuginose be- ath (Europe, Africa and adjacent islands). var. typicum, p. 1 e. apa blade subappressed-Ianuginose be- h (Hawaiian Islands) . r. decompositum, p. 40 d. ones | surface of the ultimate segment fh aeclenl, at least near the margin, and the marin pubes cent. f. f. Sterile Tadualiin mm. or more wide; the fe oe portion of the indusium ce bronds er than t sterile on the same segment (Mexico, Gate: mala and ae ai PP eV ay gn ar. Feet, p. 37 f. Sterile indusium 0.2 r less wide; the ae tion of oe sotiisaiisn broader than the sterile gg the same segment heed Nuk h and northern Mexico)...3. var. pubescens, p. 26 b. F oe and stelle "taaeatiten glabrous; ha buigan'e segments argl t - or ve glabro TOR) oi ce a. r: +: aiioaaa, p. 51 g. Blade Gaklly bipinnate-pinnatifid or emanate. often broadest at the base, the midnerve be- th Bey derately pubescent, sometimes glabro h. agi os the ultimate ———- moderately y 5-8 mm. wid sae Severe th —, sia . Var. ‘latiusculum, p. 41 h. Margin : the ultimate segments griewin or subglabrous; longest entire segment or (so and + ao vas United ‘States) . 7. Var. "peudocandatum, p. 48 me of the ulti ir segments decurrent only, or mo ro [JANUARY Rhodora 14 ee pus Sg ecergtaer: — ‘WONTTINOV - gears 3 jo — — a 4 Dh A RG STC PP | et a aMEGe aE00 a abe ie Pia 4 ™ £h : is } ot: oft Ts) 4 12, Hi sats ea, ee ae Bi, : “I, ‘. amr : s ce2en < AOS Neiinianet ay ,%, gah > cm $30 ashe x Cates Apa 1941] Tryon, Jr..—Revision of the Genus Pteridium 15 t. Ultimate segments having a iorinaren tt appearance beneath, except in some glabrous iia per eee straight, ap- ssed or ar ty ac j. No free lobes present on Se Fane is, costae and costules; the midnerve the lower sur ere ace of the ultimate e in ; and northern South Am ONG edu 9. r. caudatum, p. 54 of ee t on the eta costae and costules the mi im segments usually dark-pubescent; fertile portion of the inttostun usually not broader than the sterile on the Piped cells of the ap iene —— irregularly ti ed, small, averaging 4 by i: margin, the indusium at i 6, caaay 8 Pa k, Unilctpurties - the segments with membranous wings n the veins and midnerve, arachnoid-pubescent; pee rgin of ie segments often pu geen Wet eae Mexico, Central America and Sou Elda Caicnre ee Sto e en een yr ere es VeVArS T okidnin p. 57 k, Undereurface of the segments without membranous wings on the veins and ee appressed-pubes- cent with straight hai margin of the segments glabrous (Australia, Now Zealand and oo a). 11. r. exelent, p. 61 7. Ultimate segments not having a farinaceous Habit neath, sublanuginose-pubescent beneath (India \ to Sumatra, Philippine Islands and stbbiasigate MAUR ee Slee p. 63 oe AQUILINUM = typicum es at oon angles or at an oblique angle ei the co nae. ultimate segments broadly ovate to ce sometimes linear, ‘adnate or Surcurrent, or equally surcurrent and decurrent, or narrowed at the base , the upper surface glabrous to quite pubescent, the lower re ace lanuginose- pubescent or glabrous, not having a farina- cous appearance; cells of the sterile indusium of medium size, usually irregularly arranged. Map 1 (dots . 1. PreripiuM AQUILINUM var. typicum,'? Pate 650, FIG. 2, “Only the more important, the more obscure and scattered, and the more widely used a _ Tepresent trivial variations. Gandoger’s treatment (op. cit. 27: 180 (1891), the Pening portion of which is here quoted, amounts almost to a naming of individuals 16 Rhodora [JANUARY PLATE 651, Fic. 4, Map 4. Illustrations: Fl. Dan. 13: t. 2303 (1840) ; Waldner, Deutschl. Farne t. 17 (1883); Schk. Krypt. Gew. t. 95 ( 1809) ; Lowe, Native Ferns 2: 407, t. 61 (1867); Marloth, FI. 8. Africa 4: pl. 37 (1915), habitat. Pteris aquilina L. ~ Pl. 1075 (1753). Cincinalis aquilina (L.) Gled. Verm. Abh. 1: 24 (1765), reference taken from Woy- nar, Hedwigia io 381 (1915). Pteris nudicaulis Giild. Reisen Russ. 1: 421 (1787), according to C. Chr. Ind. Fil. 603 (1906). Pteris Sori Salisb. Prod. 402 (1796). Asplenium aquilinum (L.) Bernh. Schrader’s Journ. 17991: 310 (1799). Pteris capen- sis Thunb. Prod. 2: 172 (1800) ; a a of type at NY. Pteris lanuginosa Bory ex Willd. Sp. Pl. 5: 403 (1810); fragment of type at NY. Pteris aquilina a ee ciliata Opiz, Kratos 2': 19 (1820).14 Pteris aquilina L. var. aru Opiz, Kratos 27: 19 (1820). Pteris foemina Gray, Nat r. Brit. t. Pl. 2: 16 (1821). Allosorus aquilinus (L.) Pr. Tent. Pierid 153 (1836). Allosorus lanuginosus (Bory ex Willd.) Pr. Tent. Pterid. 154 (1836). Allosorus hottentottus Pr. Tent. Pterid. 154 (1836). Allosorus acutifolius Pr. Tent. Pterid. 154 (1836), nomen nudum. AI- losorus villosus Pr. Tent. Pterid. 154 (1836), nomen peer re Allosorus tauricus Pr. Tent. Pterid. 154 ( 1836) , nomen nudum. Pteris brevipes Tausch, Flora 19°: 427 (1836). Pteris pene Bory ex Willd. us adipic: (Thunb.) Ag. Rec. Pterid. 51 (183 Pteris plebia R. Br. ex Ag. Rec Page 52 (1839), in Hatt Pteris sordid oat riaea 182: 1 pit at piers aquilina L. var. vera Moore, Handb. Brit. foea Ed. 1, 134 (1848), refer- ence taken from Moo re cose Hist. Brit. Ferns, 167 (1851). Pteris aquilina L. var rima Moore , Handb. Brit. Ferns, Ed. 1, 134 (1848), eae pa from Moore, Pop. Hist. Brit. **20. Pteris L. 1, PTERIS rk L. a tota. ‘Lot ty integri 4 mm. longi 2 bi sup. integri 5 mm. longi 5 Lobi sup. integri 6 mm. longi 8 Lobi sup. integri 7 mm. lon 17 1 Lobi sup. in i 8 mm. longi 28 \ Lobi sup. integri 9 mm. longi 34 Lobi su tegri 10 mm. longi 40 i integri 11 mm. | 45 Lobi sup. integri 12 mm. longi 47 Lobi sup. integri 14 mm. longi— Dordogne, Payzac (Gdgr.) Pteris attenta Gdgr. 3 i 2 1 Pinnae contiguae 4 Lobi obtusi- Pedem., ag Sere se (Rostan) P. polystichoides Gear. 3 Lobi acuti- Basses-Pyr . Lestibette (Gdgr.) P. oreodora: Gdgr %4T am indebted to eae oi ‘Adal = for the references to Kratos, a periodical apparently not in the United Sta 1941] Tryon, Jr.,—Revision of the Genus Pteridium 17 Ferns, 167 (1851). Pteris aquilina L. var. pubescens Spreng. ex Liebm. Vid. Selsk. Skr. s. 5, 1: 225 (1849), as to basinym, Pteris lanuginosa Bory ex Willd., not as to plant. Pteris aquilina L. var. multifida Moore, Handb. Brit. Ferns, Ed. 3, 226 (1857). Pteris aquilina L. var. crispa Moore, Handb. Brit. Ferns, Ed. 3, 226 (1857). Pteris aquilina L. var. pubescens Afz. fil. ex Hook. Sp. Fil. 2: 198 (1858), in synonymy. Pteris aquilina L. var. lanuginosa (Bory ex Willd.) Hook. Sp. Fil. 2: 196 (1858). Al- losorus capensis (Thunb.) [incorrectly attributed to Presl by] Pappe & Raws. Syn. Fil. 32 (1858). Allosorus coriifolius (Kze.) Pappe & Raws. Syn. Fil. 31 (1858). aesia coriifolia (Kze.) Moore, Gard. Chron. 1858: 878 (1858). Aquilina vulgaris Pr. ex Milde, Fil. Eur. 45 (1867), in synonymy. Pteris Heredia Clem. ex Colm. Enum. crypt. Esp. y Port. 1: 16 (1867), refer- ence taken from C. Chr. Ind. Fil. 599 (1905) and Colmeiro, Enum. Plantas Hisp.-Lusit. 5: 437 (1889). Paesia aquilina (L.) Keys. Pol. Cyath. Hb. Bung. 22 (1873). Ornithopteris aquilina (L.) J. Sm. Hist. Fil. 298 (1875). Cincinalis lanuginosa (Bory ex Willd.) Trevis. Atti Soc. Ital. sc. nat. 17: 239 (1875). Pteris gracilis Paterson in Hennedy, Clydesd. Fl. Mem. Ed. 255 (1878). Pteridium aquilinum (L.) Kuhn var. lanuginosum (Bory ex Willd.) Kuhn in v. d. Decken, Reisen in Ost-Afrika 3°: 11 (1879). Pteris aquilina L. var. abbreviata Gillot, Bull. Soc. Bot. France : June, xxii (1882). Pteris abbreviata (Gillot) Gerard, Bull. Soc. Bot. France 29: June, xxii (1882). Pteris aquilina L. var. transsilvanica Schur, Enum. Pl. Transs. 841 (1885), nomen % Uu Ktze. Rev. Gen. 2: 820 (1891). Pteris aquilina L. f. lanuginosa (Bory ex Willd.) O. Ktze. Rev. Gen. 2: 820 (1891). Pteridium Schweiz 12: 54 (1900). Pteridium aquilinum (L.) Kuhn var. Capense (Thunb.) Christ, Beitrige Krypt. Schweiz 1°: 55 (1900). Pteridium capense (Thunb.) Krasser in Zahlbruckner, Ann. Nat. Hofm. Wien, 151: 4 (1900). Pteridium lanuginosum (Bory ex Willd.) Clute, Fern Bull. 8: 38 (1900), as to name-bringing ‘Synonym, not as to plant. Pteridium ceheginense Barnola, Bol. Soe. Aragonesa Cienc. Nat. 11:35 (1912). Pteridiwm aquilinum (L.) Kuhn var. crispulatum Barnola, Bol. Soc. Aragonesa Cienc. Nat. 11: 35 (1912). Pteridiwm aquilinum (L.) Kuhn ssp. fapense (Thunb.) Bonap. Notes Ptérid. 2: 66 (1915). Filix aquilina (L.) Woynar, Hedwigia 56: 383 (1915). Filix-foemina “quilina (L.) Farwell, Am. Mid. Nat. 12: 290 (1931). Tn 1848 Moore published Pteris aquilina L. var. vera but he 18 Rhodora [JANUARY did not base it on Pteris aquilina L. He described it as having the secondary pinnules pinnatifid, while the Fuchs plate, the type of Pteris aquilina, has the secondary pinnules entire (in fact the pinnules are only pinnatifid). In contrast to his var. vera he had var. integerrima, secondary pinnules entire. He apparently thought of var. vera as the more common and hence more “typical” phase and his description of it excludes the type of Pteris aquilina. Kuntze published a Pteris aquilina L. var. normalis in 1891 but did not actually base it on Pteris aquilina L. and cireum- scribed it much more broadly than the typical variety in the present treatment. “U. St.: Oil City” includes var. latiusculum and “Hongkong” includes either var. latiusculum or var. Wightianum. Because of the above objections to taking up either var. vera Moore or var. normalis O. Ktze., I am designating the typical variety as var. typicum, definitely based on the type of Pteris aquilina L. Ppt kas! tip of the rhizome with a tuft of dark hairs; frond usually about 1—1.5 m. high, the taller fronds scandent, peat _leicheniid; stipe longer or shorter than the blade; blade 2-20 dm., pally about 6-10 dm. lon Angee te ovate- mate segments usually pinnatifid, less often pinnate or tale: pinnatifid; longest entire segment or part of a segment from glabrous or pubescent only on the midnerve; fertile and sterile indusium ciliate and acne also pubescent on the canis sur- face, the fertile usually 0.3 mm. or more wide, the sterile usually 0.2 mm. or more wide, the fertile oon broader than the sterile on the same segment, or no broader 1941] Tryon, Jr..—Revision of the Genus Pteridium 19 Type: Figure of Filix femina Fuchs, Hist. 596, misprinted 569 (1542). Linnaeus did not have a specimen of Pteris aquilina in his herbarium in 1753 and the only illustration cited by him is taken as the type. TYPE Locauity: Europe. I have not been able to separate the African Bracken from the European. Extreme specimens from Africa differ from the typical European plant (Tidestrom 13295, France) in having a densely pubescent, flexuous rachis and the fertile indusium no broader than the sterile on the same segment. However, these characters are not always correlated and they show no definite Seographic segregation. Specimens with the rachis densely pubescent and flexuous have been seen from Mauritius, “Ex. Herb. Wm. Boott” (G) and Madagascar, Herb. Field Mus. no. 595745. Specimens from British East Africa, EZ. Heller; Natal, Abraham 28; Cape of Good Hope, U. 8. Nat. Herb. no. 22419; St. Thomas, Moller 33 and Sierre Leone, Johnston 7 have the rachis densely pubescent but straight. Distributed over es- sentially the same area are specimens that have a straight glabrous rachis: Egyptian Sudan, Lynes 186; British East Africa, Mearns 1229; Kilimanjaro, Abbott; Madagascar, Webb (?) 113 (G) ; Bourbon, Halsey (U.S. Nat. Herb. no. 1285441) ; and Cape of Good Hope, Sim 1579. Most of the European and about half of the African specimens have the fertile indusium broader than the sterile on the same segment. Specimens from the Azores, Madeira and Canary Islands and from northernmost Africa are distinetly of the European type. Shade forms often approach var. latiusculum in some charac- ters: they may be almost completely glabrous (Ziesché, Herb. Field Mus. no. 764960; Prov. Westfalen (G); Kaulfup, Bavaria (US)) or only slightly pubescent with the sterile indusium con- siderably less ciliate than normally (Ziesché, Reisengebirge, Silesia ; Richter, Hungary). Intermediates between var. typicum and var. latiusculum are only slightly pubescent beneath except on the midnerve and have the fertile and sterile indusium only slightly ciliate. Such are Danzig, July 24, 1876, Baenitz (F); Pl. Scand. ex insula Got- landia, June 16, 1894, Johansson (US); Plantae Goreanae, Taquet 2317 (US); and Hungary, Aug., 1893, Marton (F). 20 Rhodora [JANUARY Some specimens, here referred to var. latiusculum, approach var. typicum in one or more characters. The sterile indusium may be slightly ciliate and the blade pubescent beneath between the margin and the midnerve and the blade ovate: Pl. Sibiricae Exsice. 1288, Tomsk; Honig, Bavaria; near Berlin, 1844. In A. O. Olson, Vestrogothia, Sweden, Aug. 17, 1915 (F) the in- dusium is slightly pubescent. Several minor variations have been named, some of which may be worthy of recognition as forms, but I do not feel justified in recognizing them in view of the limited amount of material I have seen and the lack of field experience. Some of the varia- tions that have been described certainly do not deserve recogni- tion. Pteris aquilina var. abbreviata Gillot, for instance, is a much stunted plant found growing under extremely unfavor- able conditions, on rocks and bathed in a sulphurous vapor at 20° C. Pteridiwm ceheginense Barnola is a young plant with ab- normal leaf-cutting and Pteris gracilis Patterson is also a young plant. Var. typicum differs from var. latiusculum in its gleichenioid rather than equal vernation, in having the undersurface of the segments densely pubescent rather than pubescent only on the midnerve, in the glabrous rather than pubescent margin of the segments and the ciliate rather than glabrous fertile and sterile indusium. Also the frond is considerably taller on the average and the blade is ovate to ovate-triangular and evenly pinnate rather than broadly triangular with the basal pinnae considerably longer and broader than the second pair. The pinnules are usually at right angles to the costa rather than at an oblique angle and the penultimate segments are usually evenly pinnatifid rather than pinnate or pinnate-pinnatifid. The critical differences between var. typicum and vars. Wightianum, pubescens and Feei, with which it is also closely related are discused under those varieties. It is perhaps most closely related to var. pubescens of western North America— probably representing another example of the well-known re- lationship of the floras of western North America and Europe. Var. typicum grows in open woods, in clearings, at the edge of woodlands, in thickets, in pastures, in abandoned fields, in grassy places, on open hillsides, in forests and in recently burned- 1941] Tryon, Jr..—Revision of the Genus Pteridium 21 over areas. It is most common in dry places and in acid soils. It grows from sea-level to about 1800 m. in the Alps and to 3000 m. in the mountainous regions of Africa. It ranges throughout Europe and all but the drier regions of Africa, and adjacent islands. Evurasia.—SwEben : vier tes igh 1862, Suetberg (Herb. Field. Mus. no. 335698). Irex Holywood, Oct. 20, 1884, (G). Great Brirain: Saeiioke® (Wales). July 19, 1905, Pease 8031 (G) ; Isle of Man, Sept., ne Underwood (NY). NerHER- Tily 20, 1903, Ziesché (G): Spandau, ‘Brandenburg, ‘a : Ruhmer (F. Schultz herb. norm. no. 975) (G, NY). Niirnberg, Bavaria, Aug. 26, 1902, Kaulfup (us). CZECHOSLOVAKIA: Aug., 1911, Petrak (FI. Bohemiae et Moraviae exsicc. no. 603) (G). Huncary: Klausenberg, Aug. 16, 1901, Richter (G) ; Piliseaba, Sept., 1916, Filarszky & Kiimmerle (Fi. Hung. exsice. no. 43 1) ‘CG, US) ; Orawicza, Wierzbicki (Herb. Field Mus. a. bebe SwirzeRLANp: Wilazek (G). France: Bures, Sept. 26, 1909, Jeanpert (Herb. Field Mus. no. 81553 1); Perignac, fiareate Sept. 4-11, 1893, Guillon (Mangier, FI. selecta exsicc. no. 3395) i; 1934, Tidestrom. 13295 (F, G, US); er ih 9, 1935, 09 ( Barbour (NY). Porrucat: ‘Porto, lee Puchtion (U. S. ‘Nat. Herb. no. 1095027). Corsica: July 20, 1933, Aellen 1306 (F, ge June 17, 1933, Aellen 1316 (F, US). Iraty: Veneti tia, Nov. T Gamece: Olympia, April 2 1908, Pease 9081 1G), oo: Taurus Mts., Aug. 12, 1907, ‘Wankow (NY); ‘Plantae auricae,” Wankow (US). Unton or Soviet Soctanist ReE- me of Chernomose era Sea), Rouséasivos (G) ; ating Expedition) e Cavan Istanps: Grand oe Feb., He Cooke 107 (F, G, NY, US). Srerra Leone: Oct., 1875 erb. Field Mus. no. 69396) ; Freetown, April 8, 1882, H. H. 22 Rhodora [JANUARY Johnston 7 (G). Liperta: Grand Bassa, Dinklage 1628 ae along Dukwai River near Firestone Plantatio ons, 1928, Cooper 21 (F, NY, US). Cameroon: Jan., 1919, Gocker 133 ia), St. THOMAS (Is.): 1885, Moller 33 (US). FRen NCH EQUATORIAL Arrica: Fernand Vaz, French Congo, 1917, Aschemeier 19 (US). Union or Soutu Arrica. Capr or Goop Horr: Vicinity = Cape Town (U.S. South Pacific Exploring Expedition) (U. S. Nat. Herb. no. 22419); King Williamstown, 1892, Sim 1579 (G). Nata: Mapumulo, 1867-69, Abraham 28 (G, US); Buchanan 534d (US). Transvaau: Bokfontein, Aug. 3, 1934, ’Mogg 4755 (US); Rustenburg, (District of) Rustenburg, Sept., 1910, Col- lins (Herb. Field Mus. no. 653635) —MAaApDAGASCAR: 1897 (He rb. Field Mus. nos. 595715, Hence 595745, U. S. Nat. Herb. nos. 1431040, 1431121) ; Webb(? ) ois (G3. ‘BouRBON: (Herb. Field Mus. no. 830986); Halsey ha S. Nat. Herb. no. 1285441). Mauritius: 1867, Pike (U. S. Nat. Herb. no. 593158); (U. S. Nat. Herb. no. 516236). TANGANYIKA (GERMAN East Arrica): Kilimanjaro, 1889-90, Abbott (U. 8. Nat. Herb. nos. 22433, 22472) ; Uineamy Bunduki, Aug. 2, 1935, Bruce 96 (US). ZANZIBAR: Oct., 1873, Hildebrandt (NY). "UGANDA: Mt. De- basien, Jan., 1936, Eggeling 2703 (G); Vicinity of Kigomma, Dec. 30, 1909, Mearns 2617 (US); Kikayo, Diimmer 551 (US). Kenya (Britis East Arrica): Mau Range, July 27, 1923, A. G. Curtis 848 (G); Lake i ent Pee 17 to Aug. 15, Ren adi 929 (US); Wambugu, Sept 1909, Mearns 1229 (G, US) ; Mbololo, Nov. 7-11, 1911-— 12, i Meller (U.S. Nat. Farb’ no. 634399). ‘EcyPTian Supan: Jebel Marra, Dafur, Dec., 1921, Lynes 186 (US). Apyssrnza: 1842, Schimper 856 (US). 2. PTERIDIUM AQUILINUM var. Wightianum (Ag.), n. comb. Pate 650, Fic. 1, PLATE 651, Fic. 3, MaP 2. Illustration: Bedd. Ferns 8. India, t. ‘42 (1863). Pteris revoluta Bl. Enum. Pl. Jav. 214 (1828). Pteris excelsa BI. Enum. Pl. Jav. 213 (1828). Pteris lanigera Bl. Enum. Jav. 214 (1828). Pteris densa Wall. List no. 99 (1829), nomen nudum, isotype at US. Pteris firma Wall. List no. 100 (1829), nomen nudum. Pteris rp Wall. List no. 101 (1829), nomen nudum. sh recurvata Wall. List no. 113 (1829), nomen nudum; ex Ag. Rec. Pterid. 50 1839), isotype at US. Pteris Wightiona Wall. List no. 2178 (1829) , nomen nudum. Allosorus recurvatus (Wall.) Pr. Tent. Pterid. 154 (1836), nomen nudum. Pteris recurvata Wall. ex Ag. var. Wightiana Ag. Ree. Pterid. 50 (1839), as Wigtiana, epithet taken from Pteris Wightiana Wall. Pteris villosa Fée, Gen. Fil. Mém. Fam. Foug. 5: 128 (1850-1852). Cincinalis villosa (Fée) Trevis. Atti Soc. Ital. sc. nat. 17: 239 (1875). Pteridium aquilinum (L.) Kuhn 1941] Tryon, Jr.,—Revision of the Genus Pteridium 23 var. osmundoides Christ ex Léveillé, Bull. Acad. Geogr. Bot. 20: no. 243, 9 (1910), nomen nudum. Pteridium capense (Thunb.) Krasser var. densa Nakai, Bot. Mag. Tokyo 39: 109 (1925), epithet taken from Pteris densa Wall. Pteridium revolutum (Bl.) Nakai, Bot. Mag. Tokyo 39: 109 (1925). Growing tip of the rhizome with a tuft of dark hairs; frond 0.6-4 m. high, the taller fronds scandent, vernation gleichenioid; stipe usually shorter than the blade; blade 0.3-3 m., usually about 1.1-5 m. long, ovate-triangular, or less often pentagonal, not ternate, usually tripinnate-pinnatifid, less often bipinnate- pinnatifid or tripinnate; rachis usually densely or very densely heath and less so above; penultimate segments usually pin- natifid, less often pinnate or pinnate-pinnatifid; longest entire segment or part of a segment from three to seven, usually about four, times as long as broad; ultimate segments usually falcate or subfaleate, adnate or broadest at the base, the upper surface usually glabrous or slightly pubescent, the margin glabrous to moderately pubescent, the lower surface usually densely pubes- cent, less often slightly pubescent, rarely glabrous or pubescent only along the midnerve: fertile and sterile indusium ciliate and sometimes also pubescent on the outer surface, rarely gla- brous, the fertile usually 0.3 mm. or less wide, the sterile usually 0. mm. or less wide, the fertile portion no broader than the sterile on the same segment. Type: Wallich 2178, at Kew (not seen). Type Locauity: Dindygul, India. Specimens from the Himalayas (R. R. & I. D. Stewart 4761, 4048; Stewart 6492) are almost completely glabrous and Stewart 6492 has the ultimate segments narrowed at the base. How- erer, they have the long-acuminate pinnae and pinnules of ypical var. Wightianum and probably represent only a local variation, Other material from the same region is entirely Pical. Some specimens approach var. typicum and_ var. latiusculum in having a nearly glabrous rachis (Henry, China; Fang 3281, China) or acuminate to obtuse pinnules (Bartsch 153, Philippine Is.; Sallet, Indo-China). Var. Wightianwm can be separated from var. typicum, with Which it is apparently closely related, by its much more densely pubescent rachis, its long-acuminate rather than subacute or 1941] Tryon, Jr.—Revision of the Genus Pteridium 25 obtuse pinnae and pinnules, its faleate rather than straight ultimate segments and the narrower fertile and sterile indusium. The critical differences between var. Wightianum and var. latiusculum, with which it intergrades, are discussed under that variety. Var. Wightianum grows on dry hillsides, in jungle clearings, on waste slopes, on craters and in grassland, usually in sterile, often dry, soil. It occurs from 700 m. to 3300 m. in India and up to 2500 m. in China and 2800 m. in Papua. It ranges from northwesternmost India east to Formosa, south to Ceylon, Java and British New Guinea. Inpia: N. W. India, 1871, J. L. Stewart 3626 (NY); Nazara, N. W. Himalayas, May 16, 1896 (US Nat. Herb. no. 1274966) ; Simla Region, Simla, Sept. 18, 1883, Blanford (NY); Sonamarg, Kashmir, July 28, 1921, R. R. Stewart 6492 (NY Kashmir, July 31, 1919, Rk. R. & I. D. Stewart 4761 Ne er “ 2182 (NY, US): Neva 1820, Wallich 113° (US), isotype of Pteris recurvata Wall.: ; Khasi Hills, Shillong, Assam, Sept., 1888, Mann (NY, US) ; Sikkim, Hooker (G); Nilghiri, Thomson (G) ; 40 miles north of Darj jeeling, Dec., 1884, native collector (NY). EYLON: Beckett 182 (G ); (U.S. Nat. Herb. no. 81 6885) ; (U. 8. Nat. oie no. 22423) ; G. Wall (NY). Sram: Doi Sutep, Dec. 15, 1928, H. M. Smith 410 (G, US); Doi Chang Mt. near Hue San, Chiengmai Prov., Jan. 10, 1922, Rock 1720 (NY, US); Jan Khien, Doi Sootep, Chiengmai, Aug. 14, 1931, Cunniff 7 (NY). Frencn Inpo-Cui HINA: Annam, Massif de Bah-Na, 30 Sale Southwest of Tourane, Aug., 1920, Sallet (G, NY, US) ; Cochin- china, mays 1870, Pierre 5793 (US) : Angkor, Cambodia, Jan. 14, 192 : Sm ith 295 (US). Fep DERATED Mauay STATES: my ihe Wray 1486 (US); Penang, 1822, Wallich 99 (US), isotype of Pteris densa Wall.; Pulan Tulai, Pahang, May 27, 1927, Henderson Te (US). Cutna. Kwanerune: Canton and vicinity, Nov. 1 1, 1917, Levine 1834 (G). Harnan: Hainan, Dec, 18, 1933, Wine 35562 (NY). Kwetcuow: Kyingtenshan, Sunyi, Jan. 3. 1930, Tsiang 5241 (G, NY); Liang Feng Yah, Tsunyi Hsien, July 31, 1931, Steward, Chiao & Cheo 53 (F, NY, US). YUNNAN: Between Mohei and Moakai, March 20—April 7, 1922, Rock (G). Szecuuan: Mt. Om oo Hsien, 1928, Fong 3281 (G); Patung District, Feb., 1887, Henry (G).-—For- MOSA: Arisan, Dec: 4, 1933, Kanehira 2995 (NY, US) : Shakko, 26 Rhodora [JANUARY Dec., 1913, Faurie (E. Rosenstock exsicc. no. 51) (G); Taihei, Taiheisan, Sept. 28, 1926, Bartlett ae (US). PHILIPPINE IsLaNDs: Dum aguete, Island of Negros, Prov. Negros Oriental, June, 1908, Elmer 10349 “¢ NY, US); Port f Dos Amigos, Twai "Twai ‘Island, Feb. 19, 1908, Bartsch 153 (G, NY, US); Twin Peaks, Benguet Ro ad, Luzon, March 2, 1908, Bartsch 310 (G, NY, US) ; Bosoboso, Prov. Rizal, Luzon, June, 1896, Ramos (Herb. Phil. Bureau Sci. no. 1054) (US) ; Bukidnon, Mindanao, July—-Aug., 1913, Escritor (Herb. Phil. Bureau Sci. no. 21414) (US). Borneo: Korthals (NY). Sumatra: Aels Kanopan, Loendoet Concession, Koealoe, March 19, 1927, Bartlett 6993 (US) ; Toba, eae 204 (G). JAVA: Tjiboeroem, Preanger Prov., April 12, 1909, Palmer & Bryant 171 (US); vicinity of Goenoen S onaee Batavia Prov., May 16, 1909, Pa lmer Bryant 529 aa 541 (US). Papua (Brivis New Gurnea): Murray Pass, Wharton Range, Central Division, . June—Sept., 1933, paee 4634 (G, NY). 3. PTERIDIUM AQUILINUM var. PUBESCENS Underw. Our nae Ferns, Ed. 6, 91 (1900). Pate 650, ric. 3, PLATE 652, F MAP 3, Illustration: Ashton, PI. Rocky Mt. Nat. Pk. 22 (1933), abitat. Pteris aquilina L. var. set caba Bong. Mém. Acad. St. Petersb. s. 6, 2: 176 (1832), isotype at G. Pteridium lanuginos- um (Bory ex Willd). Nts, Fem Bull. 8: 38 (1900), as to plant, not Pteris lanugi Bory Willd.; nomen provisorium. Pteris lanuginosa ope Clute, i. Bull. 8: 37 (1900), not Bory ex Willd. Pteris aquilina pubescens (Underw.) Clute, Fern Bull. 15: = (1907). Pteridium latiusculum (Desv.) Hieron. ex Fries var. pubescens (Underw.) [comb Sager incorrectly at- ‘jclimtet = Underw. by] Seymour, Host Ind. 25 (1929). Filix- foemina aquilina (L.) ae var. lanuginosa (Bong.) Farwell, Am. Mid. Nat. 12: 290 (1931). Pteris aquilina L. var. pubescens (Underw.) [combination apparently incorrectly attributed to O. Ktze. by] Hanna, Am. F. Journ. 22: 6 (1932). Pteridium coud ane (L.) Kuhn var. lanuginosum (Bong.) Fernald, Rho- dora 37: 247 (1935), not (Bory ex Willd.) Kuhn (1879). Me et tip of the rhizome wae with a tuft of dark hairs; ond 0.3-5 m., usually about 0.8-2 m. high, the taller fronds ip fe vernation Speke. stipe aout shorter than the blade: blade 0.2-4 m., usually about 0.6-1 m. long, usually ovate- triangular, less often ace or ovate, not ternate, usually tripinnate or tripinnate-pinnatifid ; rachis slightly to moderately pubescent; pinnae and pinnules subacute to obtuse; pinnules usually nea arly or quite at right angles to the costa, sometimes somewhat at an oblique ie: eer He socally slightly 1941] Tryon, Jr..—Revision of the Genus Pteridium 27 to moderately pubescent beneath and less so above; penultimate segments often pinnatifid, less often pinnate or pinnate-pin- natifid; longest entire segment or entire part of a segment from three to five, usually about four, times as long as broad; ultimate margin, the midnerve usually glabrous, the margin moderately pubescent, the lowe pcp yea densely pubescent, less often slightly pubescent, rarely glabrous except along the midnerve; fertile indusium ai "slightly Poliate and slightly pubescent on the outer surface, rarely becoming glabrous with age, sterile indusium ciliate and sometimes also pubescent on the outer litince, the fertile usually 0.25-0.3 mm. wide, rarely up to 0.5 mm. wide, the sterile 0.25 mm. or less wide, very rarely 0.4-0.5 mm. wide, ae fertile portion broader than the sterile on the same ia Typ 4 type designated, specimens labeled by Underwood in ‘ty Herbarium of the New York Botanical Garden typify the variety. Tyre Locauiry: “Utah, California and northward.” Shade forms approach var. latiusculum in having the sterile indusium glabrous or nearly so and the lower surface of the blade nearly glabrous, or in some cases more glabrous than in typical var. latiusculum—for example, Merrill & Wilcox 923, Wyoming; Nelson & Nelson 6733, Wyoming. Intermediates between var. pubescens and var. latiusculwm °ceur along the eastern border of the range of var. pubescens. Hanna remarked that the Wyoming Bracken was “rather in- termediate between the eastern P. aquilina [var. latiusculum] and the western P. aquilina, var. pubescens (Underw.) Kuntze.” In a more recent article, on the ferns of Colorado where var. latiusculum also occurs, Wherry!’ says that “The relations be- ard to ecological variation I think it is worthwhile to put on record the results the bellowing experiment. As part of the Carnegie Institution transplant experiments, Prof. Hall sent two plants of var. pubescens from Mather, California, to Mr. C. A. e) im nee, exposed to the een of age of the ns of the two plants were similar. de did not Phe ae ‘oe ue did not become noticeably pon a support m a 6H i ’ er ntetene and ope a egy tgp higher latitude at the former. "Deters pat Distribution “er the Ferns of Wyo A . Journ. 2 (1932). herry, Colorado Ferns. Am. ?. Journ. i 28 Rhodora [ JANUARY tween these two Brackens need further study, for their inter- gradation in this region is so marked as to cast doubt on their specific distinctness.” Some of the intermediates have the general characters of var. latiusculum except that they are pubes- cent beneath or have the pinnules set nearly at right angles to the costa. Others have an ovate blade and the sterile in- dusium slightly ciliate as in var. pubescens but are glabrous beneath except along the midnerve; or have the general char- acters of var. pubescens but have the pinnules at an oblique angle to the costa. Intermediate specimens are: Miller’s Canyon, Huachuca Mts., Arizona, July 10, 1909, Goodding 170 (G, NY); San Francisco Mts., Arizona, Aug. 17, 1889, Knowlton 14 (G, US); Buckskin Mts., Arizona, June 30, 1909, Tidestrom 2336 (US); Central, New Mexico, Aug., 1895, Mulford 398 (NY); Rindoso Creek, Lincoln Co., New Mexico, July 1, 1895, Mearns (US); Buffalo Pass, Colorado, Aug. 11, 1898, Shear & Bessey (US); Southwest of Franktown, Douglas Co., Colorado, June 17, 1937, Wherry (US); Silver Reef, Utah, May 5, 1894, M. E. Jones 5176 (US); Gibbon Canyon National Park, Wyom- ing, Carleton 204 (F); Centennial Valley, Wyoming, Aug. 18, 1896, Nelson 2662 (NY). Some specimens of var. latiusculum approach var. pubescens in characters of indusium and vestiture. A specimen from 4 burnt-over hillside, Moore 2221, West Virginia, is pubescent beneath between the margin and the midnerve and the fertile and sterile indusium is slightly ciliate and pubescent. The fol- lowing specimens from New York, Ohio, Massachusetts and Indiana are pubescent beneath between the margin and the midnerve: Washington Co., New York, July 30, 1890, Burnham (Herb. Cornell U.)18; Otsego Co., New York, Frost (Herb. Cor- nell U.); Cleveland, Ohio, Aug. 15, 1875 (Herb. Cornell U.); Line Station, Indiana, Aug. 2, 1876, Grassly (F); Granville, Hampden Co., Massachusetts, Sept. 19, 1913, Seymour 60 (G). An interesting phase, approaching var. pubescens in outline of blade and pubescence is discussed under var. latiusculum. Var. pubescens differs from var. latiusculum in having the tip of the rhizome with a tuft of dark hairs rather than nearly 18T am indebted to Dr. R. T. Clausen for bringing to my attention this material in the Cornell University Herbarium. 1941] Tryon, Jr.—Revision of the Genus Pteridium 29 naked, and an ovate-triangular, fairly evenly pinnate blade rather than a broadly triangular, ternate blade. Also the pin- nules are nearly at right angles to the costa rather than at an oblique angle, the lower surface is densely pubescent between the margin and the midnerve rather than pubescent only on the midnerve and the fertile and sterile indusium is ciliate and sometimes also pubescent rather than glabrous. The frond is a little taller on the average and certainly reaches an extreme that var. latiusculum never approaches. Var. pubescens differs from var. typicum in the pubescent rather than usually glabrous margin of the segments and the markedly less ciliate and pubescent fertile and sterile indusium. The penultimate segments are usually not as evenly pinnatifid or pinnate. Var. pubescens is also closely related to vars. Feei and de- compositum and the differences are discussed under those varieties, In Michigan, Ontario and Quebee var. pubescens probably occurs as a pre-glacial relic on or related to local nunatak areas.!® Var. pubescens grows in pastures, open forests, burnt-over areas, on open slopes, in thickets, and in woods, in damp or dry places; mostly in the Transition and Canadian Zones, from sea level up to 3000 m. It ranges from southern Alaska to Mexico, east to Wyoming, Colorado and western Texas; isolated eastward in South Dakota, oS Michigan, Bruce Co., Ontario and Megantic Co., uebec. ) Dominion or Canapa.—Quersec: Caribou Hill, Black Lake, Megantic Co., Aug. 26, 1915, Fernald & Jackson 11960 and 11961 (G). Onranto: Tobermory, Bruce Peninsula, July 28, 1933, Krotkov 6319 (US) ; Dunk’s Bay, Tobermory, Bruce Co., Aug. 20, 1933, T. M. C. Taylor 6100 (G). Atserta: Vicinity of Banff, Aug. 29, 1899, McCalla 2434 (NY). BririsH Cotum- ® See Fernald, Critical Plants of the Upper Great Lake Region of Ontario and Michigan. Rhodora 37; 247 (1935). 30 Rhodora [JANUARY : Tulameen River, 1900, Kemp (NY); Revelstoke, Selkirks, iar 3, 1905, Shaw 786 (G NY, US) ; Near Victoria, Vancouver Island, July’ 28, 1908, Macoun 84096 (F) NITED STATES OF ‘Amertca —MICHIGAN: Lake Manganese, 1 mile southeast of Copper Harbor, Keweenaw Co., July 23, 1936, Hermann 8231 (US); West Bluff, 314 miles west of Copper Harbor, Keweenaw Co., June bs 1936, Hermann 7520 (NY); de Bluff, Keweenaw Co., July , 1934, Fernald & Pease 3033 G NY); "Mackinac Island, lily’ 2 8-29, 1898, Millspaugh 85 OF G). Sour Daxkora: (undoubtedly from the Black Hills) Gifford (G). Texas: Mt. Livermore, Davis Mts., Jeff Davis Co., Oct. 6, 1926, EF. J. Palmer 32010 (G). Monrana: Near Apgar, Glacier National Park, July 30, 1937, T: G.@ £. C. Yuncker 7006 (F); Belton, Aug. 22, 1903, Umbach 732 (F, NY, US). ag Trinity, Elmore Co. , Aug. vy 1910, Macbride 599 (F, G, NY, US) ; Lake Pend Oreille, Kootenai Co., Aug. 23, 1892, Sandberg, MacDougal & Heller 943 (US). Wyomtne: Gibbon & Wilcox 923 (G, NY, US) : Gees Lincoln Co., ‘July 13, 1923, Payson & Armstrong '3486 (G); Bradley Creek, Grant Teton National Park, July 17, 1932, L. Williams 887 (G, NY). Cot- orapo: Near Pagosa Peak, Aug. 25, 1899, C. F. Baker 128 (G, NY, US); Park Range above Steamboat Spa, Aug. 11, 1898, Shear & Bessey (NY). Utan: La Motte Peak, July 24, 1926, E.B. & L. B. Pues 5078 (G); Granite Canyon, Deep ’ Creek Mts., Juab Co., June 20, 1933, Maguire & Becraft 2466 (G). Nee Washoe Co., Aug. 14, 1902, C. F. Baker 1460 ie 1923, Eggleston 18881 (NY). Arizona: Chiricahua Mts., Bare- foot Park, Oct. 19, 1906, Blumer 1450 (F, G, NY, US); Washall Gulch, Santa Catalina Mts., Oct. 2, 1917, Shreve 5397 (F, G). CALIFORNIA: Avalon, Santa Catalina Island, Mar., 1897, Trask (US); Near Forest Ranch, pare Foothills, Butte Co., Sept. 23, 1916, A. A. Heller 12660 (F,G NY, US); Plains of Mendocino, Aug. 10, 1882, Pringle (F, G, NY, US). OREGON: Portland, June 25, 1920, Fisher 14 (G): Salem, 1871, E. Hall 678 (F, G). Wasx- INGTON: Wenatchee, Chelan Co., July 6, 7, 1916, Eggleston 12930 (F) ; Seattle, King Co., Sept. 19, 1931, J. W. Thompson 8106 (G). Mextco.— Basa CALIFORNIA: Sierra San Pedro Martir, 1923, Gallegas (US). CHrmuanua: Majalea, Aug. 20, 1935, LeSuer Mex-498 (F); Southwestern Chihuahua, Aug. _Nov., 1885, E. Palmer 447 (G, US); Madera, May 27—June 3, 1908, E. Palmer 288 (F, G, NY, US) ; Majarachic, April 24, 1938, Knobloch 1941] Tryon, Jr.,—Revision of the Genus Pteridium 31 5114 (F). Duranco: Metates, north of Cueva, Aug. 29-30, 1934, Pennell 18410 (US). (To be Continued) A REVISION OF THE GENUS PTERIDIUM R. M. Tryon, JR. (Continued from page 31) 4. Preriprum aquiLinum var. Fret (Schaffn. ex Fée) Maxon ex Yuncker, Field Mus. Pub. Bot. 17: no. 4, 308 (1938). PLaTE » FIG. 4, PLATE 651, Fic. 1, MAP 7. Pteris aquilina L. var. pubescens Kze. Linnaea 13: 142 (1839), as to plant, not as to basinym, Pteris lanuginosa Spreng. Pteris aquilina L. var. pubescens Spreng. ex Liebm. Vid. Selsk. Skr. s. 5, 1: 225 (1849), as to plant, not as to basinym, Pteris lanuginosa Bory ex Willd. Pteris Feet Schaffn. ex Fée, Mém. Fam. Foug. 8: 73 (1857). Pteridium Feei (Schafin. ex Fée) [combination Ncorrectly attributed to Maxon by] Faull, Contrib. Arn. Arb. 11: 87 (1938). Pteris aquilina L. var. pubescens Spreng. ex Liebm. based on Pteris lanuginosa Bory ex Willd. and Pteris aquilina L. var. pubescens Kze. based on Pteris lanuginosa Spreng. are both earlier varietal names than var. Feei, and were originally applied to this variety, but their basinyms refer them, respectively, to var. typicum and var. latiusculum. Also, of course, they could hot to be used under Pteridiwm because of var. pubescens Underw. 0 Growing tip of the rhizome with a tuft of dark hairs; frond “ip m., usually about 0.5-0.7 m. high, vernation subgleichen- ped stipe usually shorter than the blade; blade 1-5 dm., usually ee dm. long, usually broadly ovate or pentagonal, less olten ovate or broadly triangular, not ternate, usually bipinnate- Pinnatifid to tripinnate, less often tripinnate-pinnatifid; rachis "ually slightly pubescent, sometimes strongly pubescent or 38 Rhodora [FEBRUARY glabrate; pinnae and pinnules short-acuminate to obtuse; pin- nules usually nearly at right angles to the costa, sometimes at an oblique angle; costules slightly to moderately pubescent be- neath and less so above; penultimate segments pinnatifid, often pinnate, or pinnate-pinnatifid; longest entire segment or entire part of a segment from three to eight, usually about four, times as long as broad; ultimate segments usually straight, rarely sub- falcate, adnate or broadest at the base, the upper surface slightly to moderately pubescent, at least near the margin, the midnerve labrous or slightly pubescent, the margin usually moderately also pubescent on the outer surface, the fertile usually 0.3 mm. or more wide, the sterile usually 0.4 mm. or more wide, the fertile portion no broader than the sterile on the same segment. Type: Schaffner 138, 141. Probably at Rio de Janeiro (not seen). Tyre Locauity: Huatusco, Mexico. Some specimens of var. Feei are not entirely typical, having some character of var. pubescens : Heyde & Lux from Guatemala and Mohr from Vera Cruz, Mexico have the sterile indusium narrow, 0.2-0.3 mm. wide; Rose 2212, Tepic, Mexico has the sterile indusium only slightly ciliate; and Palmer 67, San Luis Potosi, Mexico has an ovate blade. Ortega 7400, Sinaloa, Mexico, 1934 (F) is intermediate between the two varieties. Var. Feei may be separated from var. pubescens by its con- spicuously ciliate rather than only slightly, if at all, ciliate fertile indusium, the usually much more ciliate sterile indusium, which is also almost twice as broad, and the fertile portion of the indusium no broader than the sterile on the same segment rather than broader. On the average, the fronds are consider- ably smaller. It differs from var. typicum in having the margin of the seg- ments pubescent rather than glabrous, the sterile indusium usually considerably broader, and in its usually smaller size. Its differences from var. decompositum, with which it is also closely related, are discussed under that variety. Var. Feei grows in the mountains of Mexico, Guatemala and Honduras, up to 2800 m. Mexico.—San Luts Potosi: San Miguelito Mts., 1876, Schaff- ner 925 (G); Alvarez, Sept. 5-10, 1902, E. Palmer 67 (F, G, NY, 39 Tryon, Jr..—Revision of the Genus Pteridium 1941] :(9 deur) WANVoruaV ‘4 se oe ‘(Z dvul) iggy “ava BA 5(¢ dew) WaALvanvoodnasd “IBA os oo! on c (fp deut) WOOMdAL ‘1B A WANITINODY WalaHaLg Jo saduvy 40 Rhodora [FEBRUARY US). Terr. Tepic: near Santa Teresa, Aug. 12, 1897, Rose 2212 (G, NY, US). Guanasvato: 1905, Duges 6 (US). HIDALGO: Durango, Aug. 13, 1937, Fisher (NY, US) ; El Chico, July, 1927, Lyonnet 98 ; between Somoriel an as Lajas, Aug. 5, 1905, Rose, Painter & Rose 9204 (NY, US). Vera Cruz: Huatusco, April, 1857, Mohr (U. S. Nat. Herb. no. 724103) ; Cordoba, 1889-91, Fink 18 (G, NY, US); nee Jalapa, May 12, 1900, Pringle 8342 (US). Mexico: Oct, 1875, Schaffner 59 and 116 (NY); below Ajusco, Sept. 1 9, 1908, Rose & Painter 7214 (G, US). Pusrsua: Teziutlan, Sept. 7 , 1910, Orcutt 4029 US). Oaxaca: Cerro de San Felipe, Sept. 26, 1897, Conzatti & Gonzales 487 (G); Cuicatlan, June 16 and 22, 1898, Conzattr & Gonzales 747 (G, US) —GuaTeMALa: Between Solola and Chiducadenango, Aug. 13, 1936, Hatch & Wilson 322 (US); Chichavac, Chimaltenango, Nov—Dec., 1930, Skutch 12 (US); Laguna de Avarza, Jalapa, Sept., 1892, Heyde & Lue (J. D. Smith no. 4080) (G, NY, US). Honpuras: Near Siguatepeque, Dept. Comayagua, "July 3, 1936, Yuncker, Dawson & Youse 5600 (F, G, NY, US). 5. PTERIDIUM AQUILINUM Most decompositum (Gaud.) comb. Puare 650, Fic. 5, PL 651, ric. 2. Illustration: St. John & Hosaka, bi Pisestale Fields. Haw. Is., Univ. Haw. Res. Pub. 6, 24 (193 Pteris Pr kate ing Gaud. in Freyc. Voy. Bot. 393 a Pteridium capense (Thunb.) Krasser var. decompositum (Gaud.) Nakai, Bot t. Mag. Tokyo 39: 110 (1925). Croving ‘tip of the rhizome with a tuft of dark hairs; frond 5-2 m., usually about 0.7-1 m. high, vernation subgleichenioid; LN usually shorter than dis ‘blade: ade 2-10 dm., usually also eh Fe on the outer surface, pete becoming ster with age, the fertile usually about 0.3 mm. wide, the sterile 1941] Tryon, Jr.,—Revision of the Genus Pteridium 41 usually about 0.2 mm. wide, the fertile ip ea broader than the sterile on the same segment, or no broa Tyre: Gaudichaud, in Herb. Muséum d'Histoire Naturelle, Paris (not seen). TYPE LocaLity: Hawaiian Islands. Var. decompositum is closely related to vars. Feei and pubes- cens. However, it has a nearly glabrous rachis rather than a pubescent one as in those two varieties and the upper surface of the segments is glabrous or rarely slightly pubescent only along the midnerve rather than pubescent and usually with a glabrous midnerve. The sterile indusium is only about half as broad as that of var. Feet and on the average it is not as large a plant as var. pubescens. The margin of the segments is some- times glabrous or nearly so rather than pubescent. The sub- appressed pubescence on the lower surface of the segments is characteristic though not always well defined. Var. decompositum is a part of the small element of the Hawaiian flora that is related to the American flora. It occurs exclusively in the Hawaiian Islands, where it grows along field borders, in virgin land, on craters, on bare eroded slopes, on open grassy slopes and in thickets from 300 m. up to 2700 m ae ISLANDS. ane June 22, 1895, A. A. Heller 2416 (F, G, NY, US). Oanvu: Waianae Mts., Honouliuli, May 2, i937 Fosberg 13810 (G): : Koalau Mia, March 29, 1933, Fosberg 9320 (G); Masel June 3, 1909, Forbes (NY). Motoxat: Kahuaawi, May 30, 1928, Degener 3558 (NY, US). Lanar: Kaokahi, Nov. 28, tee Fosberg 12429 (G). Maut: Aug. 22, 1933, Fosberg 9943 (G). Hawam: Kilauea Bird Park, Aug. 5, 1925, Neal (NY). ©: PTERIDIUM AQUILINUM Var. LATIUSCULUM (Desy.) Underw. ex ate Cat. N. Am. Pl. Ed. 3, 17 (1909). _ 650, FIG. 8, PLATE 652, FIG. 1, map 8. Illustrations: Tilto , Fern ‘Lover’s Comp. 48, 49 (1922); Tryon et al., Ferns of Wis. "18, 19 (1940), habitat; Svensk Bot. 2: t. 90 (1 803 Ff Pteris caudata L. sensu Schk. Krypt. Gew. 88 (1809), in part, Pl. 96b, a. Pteris ciliata Willd. ex "Schk. Krypt. Gew. 89 (1809 “4 i synonymy. Pteris lanuginosa Spreng. Nova Acta 10: 2 (1821). “Pteris Sprengelii Steud. Nom. Bot. 2: 358 (1894). Pterig latiuscula Desv. Mém. Soc. Linn. 62: 303 (1827). Pteris Aquilina L. var. pubescens Kze. Linnaea 13: 142 (1839), as to [FEBRUARY Rhodora 42 ‘(OT dew) asNaavuuva “ava £(6 dvul) WOLNaINOSa ‘AWA 5(Q Avul) WOTNOSAILVI ‘IBA WOANTTIOdY WoldIHaLg jo setuy — Al te | | Be aes Bee ae” RES THA SSS et See ee Ti eA Ne ARB a) = HAAR Ran EE =e a ee an PR oe A RE 1941] Tryon, Jr.,—Revision of the Genus Pteridium 43 hodesia-Kongo Exp. 1': 7 (1914). Cincinalis latiuscula (Desv.) Vict. Contrib. Lab. Bot. Univ. Montréal no. 2, 71 (1923), Fl. Gen. Indo-Chine 7?: 138 (1939). Pteridium japonicum (Nakai) Tardieu-Blot and C. Chr. in Lecomte, Fl. Gen. Indo- Chine 7?: 138 (1939), in synonymy. The earliest varietal name, Pteris aquilina L. var. pubescens Kze., cannot be transferred to Pteridium aquilinum because of Pteridium aquilinum var. pubescens Underw. Growing tip of the rhizome usually naked, or with a few whitish hairs, rarely with a tuft of dark hairs; frond 0.3-1.5 m., usually pubescent only along the midnerve, rarely slightly pubescent between the margin and the midnerve; fertile and sterile indusium usually glabrous, rarely the fertile slightly pubes- ent on the outer surface or ciliate, and the sterile slightly ciliate, the fertile 0.25-0.4 mm. wide, the sterile 0.1-0.2 mm. wide, the fertile portion broader than the sterile on the same segment. _ Type: Sheet labeled Pteris latiuscula Desv., Herb. Desvaux In Herb. Muséum d’Histoire Naturelle, Paris (not seen). Photo- is of type in U. 8. National Herbarium and Gray Herbarium seen). Tyre Locatrry: Newfoundland and St. Pierre. 44 Rhodora [FEBRUARY In northern Europe, Kamtchatka and occasionally through- out its range in North America, plants of var. latiusculum occur that have the sterile indusium slightly ciliate and the lower sur- face of the blade somewhat pubescent between the margin and the midnerve. Such plants in North America are discussed under var. pubescens. Also, occasionally, the blade is ovate rather than broadly triangular. These are apparently normal variations in any large population of var. latiusculum. In northern Wisconsin and adjacent Michigan, and perhaps more widely distributed, plants with an ovate blade, pubescent beneath between the margin and midrib, and with the sterile indusium ciliate are not uncommon (PLATE 652, Fic. 4). Repre- sentative specimens are: Boulder Junction, Vilas Co., Wisconsin, July 3, 1938, Tryon 3914 (G); Hersey, Osceola Co., Michigan, June 25, 1938, Fassett 19244 (G); Northwest of L’Anse, Baraga Co., Michigan, Fassett 19251 (G). They constitute a rather noticeable proportion of the var. latiusculum population. An attempt to identify such plants led me into this study of Pteridium but I am still unable to give a satisfactory interpreta- tion of them. In the summer of 1940 I made an effort to study them in the field more closely than I had in 1938?° but heavy late frosts had killed or mutilated most of the Bracken. They may be regarded as a scattered population intermediate between var. pubescens and var. latiusculum, closely related to the former in the characters given above but, I believe, derived from the latter by rhizomes or spores. Or, there is considerable evidence that they are merely the result of adverse growing conditions such as burning, pasturing, and extremes of exposure and soil sterility. They are found in especially dry, sunny places, often in pastures, fields, railroad rights-of-way and recently burnt- over land. There is a certain amount of intergradation between var. latiusculum and var. Wightianum. Some specimens with the leaf-cutting of var. latiusculum are slightly pubescent beneath between the margin and the midnerve and have the sterile in- dusium slightly ciliate, while others have tapering pinnules set at right angles to the costa, as in var. Wightianum, and are pubescent beneath only on the midnerve. Such intermediates 20 See Tryon, Notes on the Ferns of Wisconsin. Am. F. Journ. 29: 1 (1939). 141] Tryon, Jr..—Revision of the Genus Pteridium 45 are: Kwangtung, China, Jan. 4, 1928, Tsang 16704 (F); Kwangtung, China, Lau 2353 (G); Canton, China, 1874, Poli (Herb. Field Mus. no. 593622); near Kau Fung, Loh Ch’ang Dist., Kwangtung, China, Nov. 2-30, 1932, Tsang 20872 (NY, US) ; Foochow, Fukien Prov., China, Metcalf 7406 (US) ; Wang Shan, Anhwei Proy., China, Aug. 28, 1923, Ip (US); Mt. Reng- ger, Java, Sept. 25, 1907, Buysman (US). Var: latiusculum may be separated from var. Wightianum by its equal rather than gleichenioid vernation, its broadly tri- angular and ternate rather than ovate-triangular and evenly pinnate blade, its nearly glabrous rather than densely pubescent rachis and its subacute or obtuse rather than long-acuminate pinnae and pinnules. Also the ultimate segments are straight rather than faleate, the lower surface is glabrous except along the midnerve rather than densely pubescent, the fertile and sterile indusium are glabrous or nearly so rather than quite ciliate and the pinnules are at an oblique angle to the costa rather than at right angles. Vars. typicum, pubescens and pseudocaudatum are also closely related to var. latiusculum and the critical differences are dis- cussed under their treatments. Var. latiusculum, in eastern North America and eastern Asia, is another example of the well-known relationship of the floras of those two areas. The localities in the Black Hills of South Dakota and the mountains of Wyoming, Colorado and Nuevo Leon undoubtedly represent relics of a once continuous range, the intervening population perhaps having been wiped out by aridity in the Great Plains Region. I do not have enough data at hand to interpret the occurrence of var. latiusculum in north- ern Europe. It may have survived glaciation in local nunatak areas in Scandinavia and spread since the disappearance of the ice, or it may have spread, since glaciation, westward from un- glaciated areas in western Siberia. Var. latiusculum is probably more widely distributed in central Asia than Map 8 indicates. Several of the localities in central Asia on Map 1 probably represent var. latiusculum. Var. latiusculum grows in pastures, open woods, thickets, on Open slopes, in woods, on grassy slopes in abandoned fields and in burnt-over areas, in damp or more often dry, usually sterile 46 Rhodora [FesruarRy soil; from sea level up to 1500 m. in eastern North America, 2300 m. and 2700 m. in the mountains of Wyoming and Colorado and up to 2000 m. in China. It ranges from Newfoundland to Minnesota, south to Okla- homa and Tennessee; isolated in Mississippi, Wyoming, South Dakota, Colorado and Nuevo Leon; Sweden south to Germany, east to western Russia; Siberia; Kamtchatka to Amur, south to Formosa, Hainan and Szechuan. NortH AMerIcCA——SainT Pierre: Cape Noir, July 10, 1901, Arséne 6 (G); July 10, 1909, Arséne 5 (NY). Ite Miqueton: July 27, 1882, Delamare 364 (NY). NewrounpLanp: Holy- rood, South Arm River, Aug. 23, 1894, Robinson & Schrenk 4 (F, G, NY, US) ; Grand Falls, July 10, 1911, Fernald & Wiegand 4281 (G). Dominion or Canapa. Quesec: Boishébert, Mut- ton Bay, Saguenay Co., Aug. 15, 1915, H. St. John 90010 (G); Seven Islands, Saguenay Co., Aug. 12, 1907, C. B. Robinson 873 NY); Bic, Rimouski Co., Aug. 15, 1927, Rousseau 26884 (G); Northwest of Three Rivers, Champlain Co., Aug. 1, 1923, Cham- berlain & Knowlton (G); Grindstone, Grindstone Island, Mag- dalen Islands, Aug. 23, 1912, Fernald, Long & St. John 6645 (G). Prince Epwarp Istann: Dundee, Kings Co., Aug. 26, 1912, Fernald, Long & St. John 6646 (G). New Brunswick: Shediac Cape, July 23, 1914, F. T. Hubbard (G). Nova Scorta: Brazil Lake, Yarmouth Co., July 16, 1921, Bartram & Long 23003 (G) ; Digby, July 2-7, 1901, Howe & Lang 258 (G, NY). Onvarto: Moore Lake, Bruce Peninsula, Aug. 26, 1934, Krotkov 9606 (G, US) ; east end of Timagami Lake, Timagami Provincial Forest, Aug. 4-11, 1935, E. C. & T. G. Yuncker 5499 (F) ; Laurier, Parry Sound District, Aug. 13, 1905, Moyer (NY). Unrrep States OF America. Marne: Mt. Katahdin, July 14, 1900, Fernald (G); Boundary Lake, Aroostook Co., Aug. 12, 1902, Eggleston & Fernald (G). New Hampsutre: Randolph, Sept. 1, 1903, A. H. Moore 1454 (G) ; Jaffrey, July 25, 1897, B. L. Robinson 287 (G). Vermont: Manchester, July 27, 1898, Day 260 (G); Brandon, May 23, 1908, HZ. F. Williams (G). Massacuuserts: Sharon, July, 1905, S. F. Poole 307 (G); Tisbury, Martha’s Vineyard, Dowell 4506 (G, US). New Jersey: Budd’s Lake, Sussex Co., Aug. 12-14, 1890, Small (F). Prennsytvania: Wissahickon 1941] Tryon, Jr..—Revision of the Genus Pteridium 47 Ravine, Philadelphia Co., July 19, 1924, Lang 626 (G); Reading, erks Co., Sept. 11, 1929, E. J. Palmer 36311 (G). Maryann: between Oakland and Thayerville, Garrett Co., July 5, 1913, Tidestrom to (G); Cumberland, 1894, Shriver (NY). Dis- TRICT OF CoLumBIA: Terra Cotta, June 17, 1888, Holm (G). WEstT ela White Sulphur Springs, Greenbrier Co., May 14-17, 1914, “rpaneases (G) ; hig ree Re Co., Sept. 13, 1904, ‘A. 1(G is VireiniA: Bull Run Mts., Fau- quier Co., June 0, ee Allard 598 (G. NY); ee Smyth Co., June, 1892, Britton, Britton & Vail (NY). Nortu Caro- LINA: near Waynesville, Sept. 5, 1910, Standley 5529 (US). oe 4 miles northwest of Calumet, ay awe Co., ag! croft (Herb. Field Mus. nos. 140093, 14 0094) ; Hiram, Portage Co., Aug. 15, 1897, Webb 265 (G). Inprana: Millers, Lake Co., July Gq 1908, Lansing 2759 (F, G); Lake Oliver, July 16, 1933, Shoop - (Herb. Field Mus. nos. 907912, 9079 922). KENTUCKY: Burnt Bridge Ridge, Moree Co., July 7, 1937, mre Hodgdon & Brown 3625 (G); e Mt., Bell Co., Sept., 1893, Kearney (US). TrNNEssER: Mill of Craggie Hope, Cheatham Co., Aug. 20, 1922, Svenson 342 (G); Henderson, mare 1892, Bain (NY). Mississippr: Biloxi, June 3, 1898, Tra y 5171 (F, NY, UB). Wisconsin: Delavan, July 13, 1919, Hollister 146 (G, US); Solon Springs, Douglas Co., Sept. 7, 1930, Somerville 41 (G). Ituvors: Starved Rock, La Salle Co. Sept. 7, 8, 1914, Lansing 3786 (F); Joliet, Sept. 20, 1904, Skeel 549 (F) ; Pine Hills, Union Co. , May 6 1902, Gleason 2899 (G). Minnesora: St. Clo ud, Tuly, 1896, Campbell (F); Itasca Park, Clearwater Co., July 16, 1933, Ma yle 654 (G, NY). Iowa: Fayette Co., July, i894, 0 Fink 444 G, "8) Lebanon, daly 5, 1897, Sample 502 (G, 8). Missou : Monteer, Oct. 1907, Bush 1146 (G, NY US) ; Ironton, phe 23; 1897, tle & Stull 328 (F). ARKAN- SAS: Jasper, Newton Co., June 18, 1932, D. M. Moore 32503 (G). 1192 (NY, US): Pinecrest Camp, Deadwood, "1927, Haywood 1115 (F). Wyomrna: Jackson’s Hole, Lincoln Co., Aug. 11, 1920, L. B. & E. B. Payson 2275 (G); Laramie Peak, chan Co., July 10, 1900, Nelson 7518 (G, NY). Cotorapo: Rabbit Bar Range, Routt Co., July 18, 190 3. Goodding 1595 (G, NY. US) ; Brush Creek, Sept. 9, 1910, Tidestrom 4166 (G). —Mexico: Sierra Madre Mts., Monterey, Nuevo Leon, July 6, 1933, C. & M. T. Mueller 366, in part (G). Evropr.—Norway: Bygdé, June 28, 1907, O. Anderson (US). SWEDEN. (StocKHoLM): near Stockholm, Sept. 18, 1887, A. F. Carlson (US). Osrercbruanp: Aug. 8, 1915, A. O. Olson (Herb. 48 Rhodora [FEBRUARY Field Mus. no. 821335) ; reg July 21, 1911, A. O. Olson (NY); Asunden, Aug. 17, 1915 O. Olson (NY). KALMAR (Sacha) Kalmar, 1882, Linddorf (G).—Finutanp: Nyland, Aug. 20, 1908, H. Lindber g 409 (NY). GERMANY: near Berlin, 1860 (Herb. Fie 1d Mus. s. 29377, 162025). ASIA.— SIBERIA. ona near Titovka, Aug. 29, 1928, Proto- popova (G). KamrcHatKa: Savoiko, Aug. 29, 1928, Eyerdam (G, NY, US); Petropavlovsk, Aug. 6, 1928, Eyerdam (F,-G, NY, + U8): Primorsk: Vladiv ostok, May—Oct., 1919, Top- ping 2343 (US). Amur: Bibeinjeachtadliehok 1906, Karo (G, US) Acorn 1872, Augustinowicz (G). Japan: Maries (U. S. Nat. Herb. no. 022422) - June, 1896, Halbrook 40 (NY); Aigo, Pee 13, 1894, agers (Herb. Field Mus. no. 825006) Kano San, Kadsusa, Sept., 1888 (U. S. Nat. Herb. no. 22432) ; Mt. Kano San, Kadsusa, ‘Sept. 7, 1908 (U. S. Nat. Herb. no. 1095365) ; Yase near Kyoto, June, 1921, Husimi (U. 8S. Nat Herb. no. Bibi A Nanokawa, Tosa, July 3, 1892 (U.S. Nat. Herb. no. 22439). Formosa: Hancock 56 (US). CuHtna. MANCHURIA: ee Chen 494 (NY). Curuui: Tungling Mts., May 18, 1921, Cowdry 1214 (US). Kianc Su: Poa Wha Mt., hu u-Yung, Oct. 10, 1915 (U. S. Nat. Herb. no. 1094030). CHEKIANG: Mo Kan Shan, June 28, 1926, Cheo & Wilson 12663 G). Annwer: Chiu Hua Shan, June 28, 1925, Ching 8478 (G, NY). Hupex: 1885-88, Henry 3146 (G); Wuchang, June, 1932, Chung 9058 (F). Kranest: Lu Shan, Sept. 1 1922 Steward 2724 (US). Kwanctune: Hong Kong, ieee Le Wright (G, US). Harnan: Ka Chik Shan, “April 25, 1933, Lau 1637 (G). YuNnNAN: Ping-pien Hsien, June 5, 1934, Tsai 60128 (G). SzEcHUAN: Mt. Omei, Omei Hsien, 1928, Fang 3034 (G), 3231 and 3317 (G TS) Nanchuan Hsien, 1928, Fang 5841 ( Frencu Inpo-Cuina: Nov., 1921 (U.S. Nat. Herb, no. 1505970) . 7. PTERIDIUM AQUILINUM Var. PSEUDOCAUDATUM (Clute) Heller, Cat. N. Am. Pl. Ed. 2, 12 (1900). Prats 650, ric. 7, PLATE 652, eo - MAP 5. Illustration: Blomquist, Ferns of N. Car. 42 Pteris caudata L. sensu Schk. Krypt. Gew. 88 (1809), in part, Pl. 96b, b. Pteris etal Bory ex ibe Sp. Fil. 2: 197 (1858), in synon Pteris aquilina L. var. pseudocaudata Clute, Fern Bull. "8: *30 (1900). Pteridium Meo pate pseudo- caudatum (Clute) Clute, Fern Bull. 8: 39 (1900), nomen pro- 1941] Tryon, Jr..— Revision of the Genus Pteridium 49 visorium. Pteris pseudocaudata (Clute) Anon. in Index, Proc. Biol. Soc. Wash. 14: 200 (1901). Pteris latiuscula pseudo- caudata (Clute) Clute, Fern Bull. 11: 62 (1903), nomen pro- visorium. Pteridium latiusculum pseudocaudatum (Clute) Maxon, Am. F. Journ. 9: 44 (1919). Filix-foemina aquilina (L.) Farwell var. pseudocaudata (Clute) Farwell, Am. Mid. Nat. 12: 290 (1931). Pteris latiuscula Desv. var. pseudocaudata (Clute) E. P. St. John, Am. F. Journ. 25: 40 (1935). Growing tip of the rhizome usually with a tuft of dark hairs; frond 0.3—1.5 m., usually about 0.5-1 m. high, vernation equal; stipe longer or shorter than the blade; blade 2-7 dm., usually about 5 dm. long, usually broadly triangular, rarely broadly ovate or ovate, sometimes ternate, bipinnate-pinnatifid or tri- pinnate, rarely tripinnate-pinnatifid; rachis glabrous; pinnae ertile and sterile indusium glabrous, the fertile 0.3-0. ; Wide, the sterile 0.1-0.2 mm. wide, the fertile portion broader than the sterile on the same segment. = Clute 339, isotype in Herb. New York Botanical Garden seen). Tyre Locauity: Babylon, Long Island, New York. It is interesting to note that in 1899, one year before Clute described var. pseudocaudatum, Maxon identified Ball 511 as Pteris aquilina L. var. nov.” “not typical—approaching P. caudata Linn.” Var. pseudocaudatum intergrades to a considerable extent with var, latiusculum. Intermediate specimens have the leaf- cutting of var. latiusculum but are nearly glabrous or have the leaf-cutting approaching var. pseudocaudatum and are either glabrous or have a pubescent margin and midnerve. Such Specimens are: Hammonton, New Jersey, May 30, 1919, Killip 2260 (Us) ; Bladensburg, Maryland, July 31, 1919, Mazon 6461 (G); Table Rock, North Carolina, June, 1879 (Herb. Field Mus. ho. 315115, U. S. Nat. Herb. no. 22450); near White Sulphur Springs, Greenbrier Co., West Virginia, Aug. 29, 1903, Mackenzie 50 Rhodora [FEBRUARY 381 (NY); Henderson, Tennessee, June, 1892, Bain 162 (G); Wasioto, Bell Co., Kentucky, Sept., 1893, Kearney (NY); Bowl- ing Green, Kentucky, July, 1891, Price (NY). Lansing 513, West Pullman, Illinois, Sept. 18, 1898 (F) approaches var. pseudocaudatum. This strongly suggests that typical var. pseudo- caudatum occurs at the southern tip of Lake Michigan, and it has been reported from the Dunes Region,”! but I have not seen any specimens. Var. pseudocaudatum may be separated from var. latiusculum by the glabrous or subglabrous rather than pubescent margin of the segments, the usually glabrous rather than pubescent mid- nerve and the long and narrow rather than relatively short and broad segments. Also the growing tip of the rhizome usually has a tuft of dark hairs. Although not closely related, var. africanwm approaches var. pseudocaudatum in some characters and the differences are dis- cussed under the former variety. Var. pseudocaudatum grows in open woods, pastures, thickets, in burnt-over areas and abandoned fields, usually in dry sterile soil but sometimes in fairly damp or rich places. It is primarily of Coastal Plain distribution: Cape Cod, Massachusetts, and Long Island, New York, to Florida and Texas; also inland in North Carolina, Tennessee, West Virginia, Ohio, Indiana, Missouri, Arkansas and Oklahoma. UnitTep States or AMERICA——Massacuvusetts: Harwich, Barn- Eng. Bot 1898, Clute 339 (NY), isotype of Pteris aquilina var. pseudo- caudata Clute EW JERSEY: Hammonton, Aug. 19, 1879, 21 Peattie, Fl. Ind. Dunes, 29 (1930): ‘‘acc. to Clute’. 1941] Tryon, Jr..—Revision of the Genus Pteridium 51 Manning 21 (G); Goldsboro, Wayne Co., June 21, 1935, Cor- rell 1382 (G). SouTH CAROLINA: near Navy Yard, Charleston, May 4, 1912, B. L. Robinson 198 (G) ; Moa. Beach, Horrey Co., June 12, 1936, Correll 5218 (G); Laurel Hill, July 6, 1936, Tar- bor 735 (NY). Grorcia: Sumter Co., July 24. 1901, Harper 1110 (F, G, NY, US); Near Darien, MelInto sh Co., June 20, 1936, Correll 5458 (G). Fuoripa: Warrenton, May 23, 1903, Tracy 8633 (F, G, US); Eustis, Lake Co., May - 15, 1894, Nash 638 (F, G, NY , US). KENTUCKY: Mammo th Cave Road, Edmonson Co. , July 2, 1916, King 121 (F). TENNESSEE: Look- out Mt. , Eggleston (NY). ‘Ouro: Salem Township, Meigs Co., Oct. 10, C. H. Jones (Herb. Ohio U.). Inpiana: 1 mile east of aswell Crawford Co., Aug. 17, 1913, Deam 13976 (Deam Herb.) ; %4 mile south of Emison, Knox Co., Sept. 2, 1939, Tryon 4268 (G). ALABAMA: Mone Co., June, 1905, Dukes (G): near Fairfax, Chambers Co., Aug. 7, 1936, Correll 6562 (G) ; Auburn, Lee Co., Oct. 14, 1897, arts re (NY). Mississtepr: West of Kosciusko, Attala Co. , May 17, 1933, C. A. & U. F. Weatherby 6300 (G, NY, US); French Camp, Apri 28, 1899, I. M. Clute 54 (F, N NY). Missourt: Monteer , May 1 , 1901, Bush 474 (G); Chadwick, Christian Co., Oct. '6, 6, 10, 1915, Eggleston 12187 nie Webb City, Jasper Co., Aug. 22, 1920, H. J. Palmer 18788 NY). ARKANSAS: Nashville, Howard Co., Oct. 19, 1932, ae 9952 (G) ; West Otis, Sevier Co., July 26, 1937, Brinkley 256 (F) ; Wilmar, Drew Co., Oct. 17 1936, Demaree 14008 (NY). Lovistana: Alexandria, May 31,1 899, C: B. Ball 511 (F, G, NY, US); Chapin, Natchitoches Parish, Oct. , 1915, E. J. Palmer 8845 (NY). OxLAHoMa: Page, LeFlere Co., Sept. 9, 1913, Stevens 2715 (G, US); Idabel, McCurtain Co., May 29, 1916, Houghton 3909 (G, NY). Texas: 10 miles south of Yellow Pine, Sabine Co., Oct. 3 3, 1934, Cory 10750 (G) ; Huntsville, June 3-12, 1908, Dizon 122 (F) Houston, Harris Co., May 18, 1917, E. J. Palmer 11942 (NY er bp AQUILINUM Var. AFRICANUM*? Bonap. Notes Ptérid, I: 2 (1 915). Pare 650, ric. 6, PLATE 652, ric. 3, MAP 6. ikon *centrali-africanum ‘Hieron. ex Fries, Wiss. Ergebn. Schwed, Rhodesia-Kongo Exp. as 7 (1914). Pteridium aqui- -) Kuhn var. caudatum (L.) Sadebeck f. africanum (Bonap.) Bonap. Notes Pierid. A 321 (1923). Growing tip of the rhizome with a tuft of dark hairs; frond 0.6-1 m high, vernation gleichenioid; stipe shorter than the blade; ad 4-8 dm. long, ovate to broadly ovate, not ternate, tripinnate- -pinnatifid, or more often quadripinnate ; rachis gla- brous or subglabrous; pinnae and pinnules acute to obtuse; * Described under ssp. caudatum. 52 Rhodora [FepRUARY pinnules at an oblique angle to the costa; costules glabrous to slightly pubescent; penultimate segments pinnate; longest entire segment or entire part of a segment from five to eight times as long as broad; ultimate segments usually straight, at least some, often many, narrowed at the base, the upper surface glabrous, the margin glabrous, the lower surface glabrous or very slightly pubescent on the midnerve; fertile and sterile indusium glabrous, the fertile usually 0.3 mm. wide, the sterile 0.1-0.2 mm. wide, the fertile portion broader than the sterile on the same segment, or no broader. Type: Busse 944, in Herb. Muséum d’Histoire Naturelle, Paris (not seen). Type Locauiry: “Afrique Orientale allemande. Magaba- Thal.” Var. africanum may be separated from var. typicum by its more finely cut, often quadripinnate, blade, its glabrous rather than pubescent rachis and its glabrous rather than ciliate fertile and sterile indusium. Also the pinnules are at an oblique angle to the costa rather than at right angles and the ultimate segments are usually narrowed at the base and glabrous or subglabrous rather than adnate or broadest at the base and densely pubescent beneath. It differs from var. pseudo- caudatum in its gleichenioid rather than equal vernation, its ovate or broadly ovate rather than broadly triangular blade and in that the ultimate segments are usually narrowed rather than adnate or broadest at the base. Var. africanum grows in dry moderately light woods, in virgin forest and in open grassland in tropical Africa, up to 1400 m. Arrica.—Be.o1an Conco: Stanley Pool, Sept., 1883, H. H. Johnston (U. 8. Nat. Herb. no. 22427) ; Elizabethville, Dec. 30, 1919, Shantz 513 (US) ; Albertville, 1931, Lugen 89 (G). Portu- curse West Arrica: Near Cuanza River, Sept. 24, 1923, A. G. Curtiss 358 (G). Union or Sourn Arrica: Ndola, northern Rhodesia, Stevensen 400 (US). Tancanyrka (GERMAN East Arrica): N’Gano-N’Gano, Urundi, March 17, 1919, Schantz 757 (US). PTERIDIUM AQUILINUM ssp. CAUDATUM (L.) Bonap. Notes Ptérid. 1: 62 (1915) Growing tip of the rhizome with a tuft of dark hairs; rachis glabrous to slightly pubescent, rarely densely pubescent; costae glabrous to moderately pubescent; pinnules usually nearly at right angles to the costa, rarely at a somewhat oblique angle; 53 of the Genus Pteridium TRICO ORT AA. Jr. ,—Revision 1941] Tryon, PAP o\ NN, ef l 65, Oe mam: Pe eR RNa aS ae aa a Z| 54 Rhodora [ FEBRUARY surface glabrous to moderately pubescent, the lower surface usually appressed-pubescent with straight hairs or arachnoid- pubescent, rarely sublanuginose or glabrous, usually having a farinaceous appearance (PLATE 650, Fic. 13), fertile indusium 0.3-0.5 mm. wide, the sterile 0.1-0.35 mm. wide. Map 1 (stars). 9. PTERIDIUM AQUILINUM var. cAUDATUM (L.) Sadebeck,”* Jahrb. Hamb. Wiss. Anst. 14: Beiheft 3, 5 (1897), as to indicated basinym Pteris caudata L., not as to plant. Pate 650, Fics. 9, 12, PLATE 653, Fic. 4, MAP 11. Illustrations: Plumier, Pl. Amer. t. 22 (1693) ; Jacquin, Ic. Pl. Rar. 3: t. 645 (1786-1793) ; Brit- ton, Fl. Bermuda, 419 (1918). Pteris caudata L. Sp. Pl. Ed. 1, 2: 1075 (1753). Pteris aqui- lina L. var. caudata (L.) Link, Hort. Berol. 2: 33 (1833). Al- losorus caudatus (L.) Pr. Tent. Pterid. 154 (1836). Pterws aquilina L. var. mexicana Fée, Mém. Fam. Foug. 8: 114 (1857). Pteris caudata L. var. mexicana Fée, Mém. Fam. Foug. 9: 8 1857), nomen nudum. Ornithopteris caudata (L.) J. Sm. Hist. Fil. 298 (1875). Cincinalis caudata (L.) Trevis. Atti Soc. Ital. sc. nat. 17: 239 (1875). Pteridium caudatum (L.) Maxon, Proc. U. 8. Nat. Mus 631 (1901). Pteridium aquilinum (L.) Kuhn var. caudatum (L.) Sadebeck f. glabratum Hieron. Hed- wigia 48: 246 (1909). Pteridium aquilinum (L.) Kuhn var. caudatum (L.) Sadebeck f. pubescens Hieron. Hedwigia 48: 246 (1909). Filix-foemina aquilina (L.) Farwell var. caudata (L.) Farwell, Am. Mid. Nat. 12: 290 (1931). Frond 0.6-7 m., usually about 1.2-2.5 m. high, the taller fronds scandent, vernation not clearly observed, apparently gleichenioid; stipe usually about as long as the blade; blade 0.3- 4 m., usually about 0.6-1 m. long, triangular to broadly ovate or long-triangular in large plants, tripinnate, or more usually tripinnate-pinnatifid or quadripinnate; costules of the penulti- mate segments usually glabrous, sometimes slightly pubescent above and beneath with long white hairs, rarely pubescent above with short white hairs, or beneath with dark or bicolorous hairs; free lobes not present along the rachis, costae and costules; ulti- mate segments usually linear or long-linear, sometimes oblong- ovate or ovate, the margin glabrous or infrequently slightly pubescent, the lower surface usually densely appressed-pubescent with long straight hairs, rarely arachnoid-pubescent, sometimes glabrous, having a farinaceous appearance except in the glabrous 23Qn many of my annotation labels the combination is accredited to Underwood (1900). 1941]. Tryon, Jr..—Revision of the Genus Pteridium 55 forms, the midnerve usually glabrous, rarely pubescent with dark or bicolorous hairs, only rarely membranous wings present along the veins and midnerve; fertile and sterile indusium usually glabrous, sometimes slightly or densely ciliate, the fertile portion broader than the sterile on the same segment; cells of ee sate 12). indusium large, in fairly definite rows (PLATE 650, FIG. Type: Specimen in the Linnaean Herb. (not seen). Linnaeus had two specimens labelled Pteris caudata in his herbarium in 1753.24 One is apparently var. pseudocaudatum and the other is clearly var. caudatum. The latter specimen is taken as the type. Since Linnaeus’ description in the Species Plantarum was taken from his Hortus Cliffortianus, a specimen, if there is one, in the Clifford Herbarium might be considered to be the type. However, since there is a perfectly good specimen available in Linnaeus’ own herbarium, it seems best to designate that as the type. Although one of Linnaeus’ specimens is apparently var. pseudocaudatum, the application of his name is perfectly clear from the figures and localities cited by him, Tyre Locatiry: West Indies. Linnaeus, Sp. Pl.: “Jamaica, Dominica.”, Hort. Cliff.: “Santo Domingo, Jamaica etc.”*5 The typical phase of var. caudatum, with the segments long and narrow and remote, occurs mainly in the West Indies, Florida and in the coastal regions of northern South America, Central America and Mexico. At the higher altitudes, mostly in Cen- tral America and Mexico, there is a phase with the segments relatively short and broad and approximate. The extremes of this phase (Lago San José, Porto Rico, July 15, 1912 (U.S. Nat. Herb. no. 566772) ; San Jose, Tamaulipas, Mexico, 600-1100 m., 1902, Kemp (NY) ; San Rafael de Norte, Nicaragua, March 25, 26, 1917, Miller & Griscom 157 (US) ; Costa Rica, 1800 m., Aug. 2, 1933, Solis 277 (F); Columbia, Charetier 33 (NY, US); El Salvador, 1200-1500 m., Standley 21537; Nicaragua, 850 m., Mazon, Harvey & Valentine 7421) are well marked, but there 1S @ great deal of intergradation from one phase to the other. In fact, almost a third of the specimens I have examined are intermediate, and the ranges overlap considerably in individual Me. C. A Weatherby has kindly made his notes on the Linnaean Herbarium avail- able to me. * Linnaeus, Hortus Cliffortianus, 473 (1737). 56 Rhodora [FeBRvARY cases. I do not believe that this variation can reasonably be given varietal status. Occasional specimens, especially of the “compact” phase men- tioned above, but also of the typical phase, have the segments only slightly decurrent but these can usually be placed in var. caudatum by the characteristic pubescence. Anthony 400, Lower California, has the lower surface of the segments appressed-pubescent with short hairs as is often the case in var. esculentum. Var. caudatum can be distinguished from var. esculentum by its lack of free lobes along the rachis, costae and costules; and the fertile indusium is broader than the sterile on the same seg- ment rather than usually no broader. Also the glabrous phases of var. caudatum do not have a farinaceous appearance. Var. caudatum is most closely related to var. arachnoideum; the differences are discussed under that variety. It grows in clearings, rough pastures, on dry hillsides, in cut- over forest land, in fresh-water marshes, in pinelands, scrublands and in shady rocky places, mostly at the lower altitudes but up to 2000 m. in Central America and Mexico, and 3000 m. in Venezuela; and from 1000 m. to 1300 m. in the Revillagigedo Islands. It occurs from Bermuda to Florida, West Indies, Mexico, Central America and northernmost South America. Unirep States or AMERICA. a Rit Myers, Lee Co. June 1, 1916, J. P. Standley 213 {F. NY, US) ; Cape Sable, July, A. i Curtiss giua tr, GG, NY ERMUD Sete Marshes, Aug. 31-Sept. 20, 1905, Brown & Britton 159 (F, G, NY, US). Wesr Inpres:—Banama Istanps: Near Nassau, Feb. 11, 1903, A. H. Curtiss 74 (F, G, NY, US); Orange Creek and vicinity, Cat Island, Feb. 27, 28, 1907, Britton & Millspaugh 5754 (F, NY). Cua: Monte Verde, Jan —July, 1859, Wright 872 (F, G, NY, US); Josephina, north Ay Ja aguey, Yateras, Oriente, April 23, 1907, Mazon 4129 (G, NY, US). Jamatca: Mulgrove, north of Ipswich, St. Elizabeth, April 1, 1920, Mazon & Killip 1 1488 (¥,.G, US). Hispantoua: Anse Galette, Seat Island, Haiti, ‘ds 3-14, 1920, Leonard 3210 (F, G, US), 3 208 ( Vicinity of Mission, Haiti, sr ps ey 4, 1920, Leonard 3916 (US) ; San Lorenzo Bay, south ec of Samana B y, Dominican Republic, April 5-11, 1921, Abba: 1275 (G, US); Province of 1941] Tryon, Jr..—Revision of the Genus Pteridium 57 Barahona, Dominican Republic, July, 1911, Fwertes 1053 (F, G, , US). Porto Rico: Santurce, Jan. 22, 1903, A. A. Heller 6446 (F, G, NY, US). Monvserrat: Turner (U.S. Nat. Herb. no. 428409). Awntiacua: (US). Mextico.—Islands off the coast of Lower California and on the adjacent mainland, March—June, 1897, Anthony 400 (F, G, , US). Nuevo Leon: Sierra Madre, July 6, 1933, C. H. & M. T. Mueller 366, in part (G). Tamautipas: La Vegonia, San Jose, July 5, 1930, Bartlett 10096 (US). Srvatoa: Sierra de Chabarria, 1927, Ortega 4079 (US). Nayarit (Terr. TeEpic): Jan. 5-Feb. 6, 1892, E. Palmer 1948 (G, US); Jalisco, Nov. 11, 1925, Ferris 5958 (G, US). Vera Cruz: Mt. Orizaba, Aug. 21, 1891, Seaton 110 (F, G, NY); near Jalapa, May 12, 1900, Pringle 8342 (G, , US). Lima: Socorro Island, Revillagigedo Is- lands, May 8, 1925, Mason 1662 (G, US). Gurrrero: Montes de Oca, San Antonio-Buenos Aires, May 5, 1938, Hinton 14083 (G, US). Oaxaca: Tolosita, June, 1937, L. Williams 9614 (F). Cutapas: Tacnalpan, July 28, 1890, Rovirosa ; 43 Yucatan: Tuxpena, Campeche, March 23, 1932, Lundell 1431 (F). Terr. Quivrana Roo: Cozumel Island, Feb. 20, 1899, Millspaugh 1551 (F, G). ENTRAL AMERICA.—BritIsH Honpuras: Big Creek, April 27, 1929, Schipp 190 (F, G, NY, US). Guaremaza: Vaxactum, Dept. Peten, April 7, 1931, Bartlett 12521 (F, US); Vicinity of Puerto Barrios, Dept. de Izabal, June 2-6, 1922, Standley 25028 (US). (G Nicaracua: Casa Colorado and vicinity, south of Managua, June at. Herb. nos. 22440, 154190) ; San Jose, Prov. San Jose, 1887, J. J. Cooper (J. D. Smith no. 6018) (US). Panama: Pedro Miguel, Jan. 27, 1918, Killip 2825 (US); vicinity of El Boquete, Chiriqui, March 2-8, 1911, Mazon 4926 (US). outH America.—Co.tumsiA: Santa Marta, near Onaca, Aug. 22, 1898-1901, H. H. Smith 1088 (F, G, NY, US); Vicinity of Medellin, 1911, Charetier 33 (NY, US). VeENezueta: 214 miles east of Merida, State of Merida, Jan. 23, 1931, Reed 210 (US). 10. Preriprum AQUILINUM var. ARACHNOIDEUM (Kaulf.) Her- o Rev. Sudam. Bot. 5: 21 (1937).2¢ Prater 650, rics. 10, 13, fi PLATE 653, Fic. 3, MAP 12. [Illustrations: Vellozo, Fl. Flum. : t. 80 (1827); Christ, Geog. Farne, Fig. 9 (1910), habitat. 28 eee Combination incorrectly attributed to Baker. 58 Rhodora [FEBRUARY Pteris psittacina Pr. Delic. Prag. 1: 185 (1822). Although I have not seen Presl’s type, his description clearly refers his name to Pteridium idee Hk and the locality, Rio de Janeiro, to var. arachnoideu Material recently referred to Pteridium a represent Shlpsae leaf-forms of var. arachnoideum and var. caudatum. eaf-cutting of young plants of these varieties is eraevably different from the — ae and the typical types of pubescence are not eloped. Pteris Cane Schrad. Gott. gel. Anz. 1824); a “(1804). Pteris arachnoidea Kaulf. Enum. Fil. 190 (1824). Allosorus psittacinus (Pr.) Pr. Tent. Pterid. 153 (1836), as A. psitaccinus. Allosorus arachnoideus (Kaulf.) Pr. Tent. a 153 (1836). Pteris aqui- lina L. var. arachnoidea (Kaulf.) D. C. Eaton, Proc. Amer. Acad. n.s. 8: 203 (1861). Pteris Gardneri Pr. ex Ettingsh. Denkschr. Ak. Wiss. Wien, 23: 42 (1864). Aquilina Gardneri Pr. ex Et- tingsh. Farnkr. 1 ( Sati in synonymy. Pteris aquilina L. var. psittacina (Pr.) Baker in Martius, Fl. Brasil. 12: 404 (1870). Cincinalis pale eaters (Kaulf.) Trevis. Atti Soc. Ital. se. nat. 17: 239 (1875). Pteridium aquilinum (L.) Kuhn var. esculen- tum (Forst.) Kuhn f. arachnoideum (Kaulf.) Hieron. Hedwigia 48: 246 (1909). Pteridium arachnoideum (Kaulf.) Maxon, Journ. Wash. Acad. Sci. 14: 89 (1924). Filix-foemina re gece (L.) Farwell var. arachnoidea (Kaulf.) Farwell, Am. Mid. 12: 290 (1931). Pteridium psittacinum (Pr.) Maxon, Pi Biol. Soc. Wash. 46: 141 (1933). Frond 1-3 m. high, vernation not clearly observed, apparently gleichenioid; stipe usually shorter than the blade; blade 0.5-2 m. long, ovate-triangular to long-triangular in large plants, tripin- nate to quadripinnate; costules of the penultimate segments usually pubescent beneath and less so above with short white and also dark or bicolorous hairs, sometimes glabrous; free lobes present along the rachis, costae and costules; ultimate segments arachnoid-pubescent, rarely appressed-pubescent with short straight hairs, or glabrous, nearly always having a farinaceous appearance, the midnerve usually pubescent with dark or bi- colorous hairs, membranous wings usually present along the veins and midnerve (PLATE 650, Fic. 14); fertile and sterile indusium ciliate and sometimes also pubescent on the outer surface, or glabrous, the fertile portion no broader than the sterile on the same segment; cells of the sterile indusium small, irregularly arranged (cf. PLATE 650, Fic. 11). Type: Chamisso, probably at Berlin (not seen). TYPE LOCALITY: Braz 1941] Tryon, Jr..—Revision of the Genus Pteridium 59 Riedel, Brazil, “Ex. herb. hort. Petropolitani” (G) and Curran 128, Bahia, Brazil (G, US) differ from typical var. arachnoideum in having no free lobes along the rachis, costae and costules. The following are intermediate between this and var. cauda- tum. Hitchcock 17031, Penal Settlement, British Guiana, Dec. 3-9, 1919 (G, US) has free lobes present and a farinaceous ap- pearance, even though glabrous, as in var. arachnoideum but has the fertile indusium broader than the sterile as in var. caudatum; Heller 4468, 14 miles northeast of Mayaguez, Porto Rico, Feb. 1, 1900 (G, F, NY, US) and Mazon 4075, San Piedra, Oriente, Cuba, April 14, 1907 (G, US) are similar to var. caudatum in characters of pubescence and of the indusium but they have a few free lobes as in var. arachnoideum; Rose & Painter 7595, Jalisco, Mexico (US) is arachnoid-pubescent but. has no free lobes; and Pennell 5162, La Cumbre, El Valle, Colum- bia (US) and Ariste-Joseph A207, Bogota, Columbia (US) are arachnoid-pubescent as in var. arachnoideum but have the fertile indusium broader than the sterile and have no free lobes as in var. caudatum. Var. arachnoideum differs from var. caudatum in having free lobes along the rachis, costae and costules rather than not having them; the midnerve is usually pubescent with dark hairs rather than usually glabrous and the fertile portion of the indusium 1S no broader than the sterile on the same segment rather than broader, Also the lower surface is arachnoid-pubescent rather than appressed-pubescent with straight hairs, the cells of the indusium are smaller and irregularly arranged and membranous Wings are usually present on the veins and midnerve beneath rather than usually not present. Even the glabrous phases have 4 farinaceous appearance beneath. he differences between var. arachnoideum and var. esculen- tum, with which it has often been confused in the past, are dis- cussed under the latter variety. _ ‘40. arachnoideum grows on open slopes, in open rocky places, I thickets, forests, grassland, in cleared land and on the edge of forests from the lower elevations up to 3000 m.; and from m. to 700 m. in the Galapagos Islands. It Tanges from the West Indies, Cuba to Trinidad, to southern Mexico, Central America, Galapagos Islands and throughout 60 Rhodora [FeBsruaRY South America except the southern portion; also it is apparently absent from most of the Amazon Basin. West Inpres:—Cvusa: Monte Verde, Jan.—July, 1859, Wright 985 (G); Loma del Gato and vicinity, Sierra Maestra, Aug., 1923, Hioram & Clement 6497 (US); Santiago, Santa Ana, March 23, 1902, Hamilton 240 (NY). JAMAICA: Vicinity of St. Helens Gap, St. Andrew, March 4, 1920, Mazon & Killip 619 (F, G, NY, US). Huspantowa: vicinity of Furey, Haiti, May 26—June 15, 1920, Leonard 4339 (G, US); Prov. Monte Cristi, Santo Domingo, June 24, 1929, Bhincn 12990 (NY, US). Porro Rico: Aug. 28, 1885, Sintenis 2658 (G, US). Santa Lucia: Ventine Sulphur Springs (Soufriére), May, 1935, Box 449 (US). Trinipap: St. Ann, March 17, 1920, Britton, Hazen & Mendelson 676 (G, NY); 1877-8, Fendler 77 (G, NY, US). Mex! co.—Mext1co: Nanchititla, Temascaltepec, Feb. 14, 1935, pk 7371 (G, eee Vera Cruz: Zacuapan, Dec., 1912, Pur- pus 6191 (F, G, NY, US). Guerrero: Montes de Oca, San Antonio-Buenos Aires, May 3, 1938, Hinton 14069 (G, US). Oaxaca: Cuicatlan, June 16-22, 1898, Conzattt & Gonzalez 748 (G). CeNTRAL AMERICA.—GUATEMALA: Volcan de Fuego, , t L S). July 21, 1938, ea rae & Wagn 55 NICARAGUA: San Rafael de Norte, March 25, 26, 1917, Miller & Griscom 152 (US). Costa Rica: from Vara Blanca to La Con- eit ay 23, eet Mazon & Harvey 8400 (US) ; 1901-1905, (Herb. Field Mus. no. 404457 7). ew aNaMa: Vici ity of Monte ey oo Chiriqui, June 27—July 13, 1935, Seibert 234, in par Sourn America.—Cotumsta: Palmira, Dept. El Valle, May 27, 1922, Pennell & Killip 6100 (G, NY, US): Santa Marta, Aug. 26, 1898-99, H. H. Smith 1091 (NY). VENEZUELA! Maracai, Vogl (G); Island of Margarita, Aug. 28, 1903, J. R. Johnston 177 (G); Tovar, 1854-5, Fendler 104 (G); 1917, Curran & Haman 1111 (G, NY). Britiso Gutana: Malali, Demerara River, Oct. 30-Nov. 5, 1922, de la Cruz 2658 (F, G, NY, US); Mt. Tramaikpang, northwest part of Kanuku Mts., April 22, 1938, A. C. Smith 3657 (G). teal GUIANA: Near Cayenne, Oct. 8, 1830, Leprieur 100 (F, G, US). Ecuapor: Western San Miguel Mts., Oct. 21, 1933, Schimpff 247 (F); Andes, 1857-9, Spruce 5601 (G, NY); Wreck Bay, Chatham Island, Galapagos Islands, July 6, 1905-6, A. Stewart 996 (F,G, NY, US). Psrv: 1941] Tryon, Jr.—Revision of the Genus Pteridium 61 Estrella, = Ayacucho, May 8, 14, 1929 Killip & cece 23095 NX; U; ero de Cusilluyoe, Dept. Cusco, May 1925, Pennell cs (G, NY, US); Tarapoto, Dept. San “ine Dec., 1929, L. Williams 5971 (F de Braziu: Para, Nov., 1913, Petelot (Herb, Field Mus. no. 593026) ; Near Petropolis, July 10-16, of Rio de Janeiro, Dec. 31, 1928, L. B. Smith 1587 (F, G, NY, US) ; Sao Leopoldo, Rio Grande do Sul, Rick 24 (G). Boutvia: Ineachaca, Dept. Gee Pro v. Chapare, Jan. 24, 1929, Steinbach 8927 (F, G); L acotal, Dest Cochabamba, Prov. Chapare, Feb. 25, 1929, Steinbach 9363 (F, G, NY); Tipuani, April, 1920, Buchtien 5271 (F, US). PARAGUAY: Y-acé River, Cordillera Centralis s, 1900, Hassler 6997 (G); Paranda River, 1909-10, Fiebrig 6138 (G s Urucuay: Catalan, Dept. Artigas, Nov., 1927, Herter 996 (NY); Pan de Azucar, Dept. rarer Jan, 21, 1912, Osten 5688 (US ). ARGENTINA: Fonta Re- sistencia, Chaco, Feb., 1933, Schulz 727 (G) ; Dept. Punilla, Prov. Cordoba, March 16, 1939, Dawson 588 (G) ; Prov. de Catamarca, Nov. 11, 1910 (U. S. Nat. Herb. no. 111 3401). 11. PreripruMm AQUILINUM Var. ESCULENTUM Shain eho oath 347 (1882). Puate 650, rics. 11, 15, PLATE 653, F 1, Map 9. Illustrations: Domin, Bibl. Bot. 85: figs. 33, 34 (1914) ; Schk. erg te Gew. t. 97 (1809); Dobbie, ae Zealand Ferns, » Ed 3, 183 (19. Hit : ; 8C 2 fh: FLAN. (ss Cincinalis esculenta (Forst.) Trevis. Atti Soc. Ital. se. 17: 239 (1875). Ornithopteris esculenta (Forst.) J. Sm. Hct, Fil. 298 (1875). Pteridium esculentum (Forst.) Diels in Engl. «& Prantl, Nat. PA. 14: 296 ( Pork Pteris aquilina L. f. ography or Sin gael Pteris aquilina L. f. caudata Christ in arb. Monsunia 1: 68 (1900), without bibliography or reference. r teridium aquili en (L.) Kuhn var. pt loa Domin, Bibl. Bot. 851: 162 (1914). Pteridium aquilinum (L.) Kuhn var. pseudocaudatum Domin, Bibl. Bot. 851: 161 (1914), not oe) Heller. Pteridium aquilinum iL) Kuhn ssp. esculentum Sort.) Bonap. Notes Ptérid. 4: 116 (1917). Frond 0.6-3 m high, vernation subgleichenioid; stipe about as long as the laa: blade 0.3-1.5 m. long, ovate to triangular, tri anal to quadripinnate; costules of the penultimate segments glabrous above, glabrous to slightly pubescent beneath with White and often’ also dark hairs; free lobes usually present along 62 Rhodora [FEBRUARY the rachis, costae and costules; pp ares segments oblong or usually linear, the margin gla brou the lower surface densely appressed-pubescent with long, or ee straight hairs, always aving a farinaceous appearance, the m idnerve usually glabro Us, sometimes moderately pubescent with white and sometimes also dark hairs, no membranous wings present on the veins and mid- nerve; fertile and sterile in Ate abt usually glabrous, rarely a ciliate, a fertile portion usually no broader than the sterile on the same segment; cells of the sterile indusium small, irregularly adler (PLATE 650, Fic, 11). orster, location unknown. A fragment of the type “ex Forster Herb.” “collected” L. M. pecan t is at Herb. New York Botanical Garden TYPE Locality: Society ae ‘Copeland?? says that: “The sole Tahitian record is that of Forster . . . its absence from all later collections cee that it does not beseem a wild plant to be edible. Var. esculentum differs from var. arachnoideum in its pubes- cence,—appressed with straight hairs rather than arachnoid,— the absence rather than presence of membranous wings along the veins and midnerve and the midnerve usually glabrous or with white hairs rather than pubescent with dark or bicolorous hairs. De la Cruz 2658, British Guiana and Gleason 423, Tumatu- mari, British Guiana, June 18—July 8, 1921 (G), var. arachnoide- um, approach var. esculentum in being appressed-pubescent with short hairs. The differences between var. esculentum and vars. caudatum and yarrabense are discussed under those varieties. Var. esculentum grows in open places, pastures, thickets and clearings, from sea-level up to 1300 m. It ranges from Australia to the Society Islands. AUSTRALIA: eceaites Victoria, July 3, pei a (G); Hall’s Gap, Grampi n Mts., Victor ia, Dec., 2, Tilden 848 7 G); Port pet Victoria, Oct -Nov., 191%, Tilden 762 G): Bondi Bay, Sidn ney, New South Wa les, Sept., 1912, mtn 570 (F, G); near Mareton Bay, 1850-51, Strange (G); east coast (Nouvelle Hollande), 1845, Verreaux 267 (G, US). TasMANIA: Gunn (G). Norrotx ISLAND: 1884, Metcalfe (U. Pe Nat. Herb. no. 22443). New Zra.anp: Whakarewarew2, ov., 1909, Leland, Chase & Tilden 143 (F, G); South Island pee Field Mus. no. 355839) ; Craig (Herb. Field Mus. no. 27 Copeland, Pteridophytes of the Society Islands. Bishop Mus. Bull. 93, 57 (1932). 1941] Tryon, Jr..—Revision of the Genus Pteridium 63 596860, G) ; Taranaki, Heywood 56 (G); Mt. Ngongotaka, May— July, 1898, Prince (G); North Island (Ex. Herb. T. Kirk) (G). CuatHaM IsLtanp: Oct., 1874, Kershner (US); Dec., 1874, Scott (US). New Cateponia: (Herb. Field Mus. no. 596487) ; 1861— 67, Deplanche 1563 (G); 1874-76, Germain (NY). New Hesripes: Aneiteum, Feb., 1859 (Herb. Field Mus. no. 596565). Fist Istanps: 1860, Seemann 809 (G) ; 1877-78, Horne 601 (G); (Herb. Field Mus. no. 593802). Navicator Istanps (Samoa): 1873, McAlesber (NY). Society Isuanps: “Ex. Forster Herb.’ (NY), fragment of type. 12. PrERIDIUM AQUILINUM Var. YARRABENSE Domin, Bibl. Bot. 851: 161 (1914). Pare 650, ric. 16, PLATE 653, FIG. 2, Map 10. Illustration: Domin, Op. cit. fig. 32. Pteris lorigera Wall. List no. 103 (1829), nomen nudum; iso- type at US. Pteris semihastata Wall. List. no. 102 (1829), ance, the midnerve usually densely pubescent with white and infrequently also dark hairs, no membranous wings present on the veins and midnerve; the fertile and sterile indusium rather densely ciliate and usually also pubescent on the outer surface, the fertile portion broader than the sterile on the same segment, or oa ahaa cells of the sterile indusium small, irregularly ar- ranged. Typr: 1910, Domin, probably at Praha (not seen). YPE LocaLity: ‘“‘Nordost-Queensland bei Yarraba.” Var. yarrabense differs from var. esculentum in having the lower surface of the blade sublanuginose rather than appressed- pubescent with straight hairs, the midnerve on the lower surface of the segments pubescent with white hairs rather than usually glabrous and the fertile and sterile indusium ciliate and pubes- cent rather than glabrous. Also it does not have a farinaceous appearance beneath and usually does not have free lobes along the rachis, costae and costules. 64 Rhodora [FEBRUARY Strange, Australia, and Seemann 809, Fiji Islands, var. esculen- tum, approach var. yarrabense in having the fertile and sterile indusium ciliate and the midnerve of the segments slightly pu- bescent beneath with white hairs. Ching 5360, Kwangsi Prov., China (NY, US) is intermediate between var. yarrabense and var. Wightianwm and _ therefore represents an intermediate between ssp. caudatum and ssp. typicum. Var. yarrabense grows in clearings, thickets, open slopes and at the edge of woods, up to 2500 m. from northern India to Sumatra, east to the Philippine Islands and northeastern Australia. Inp1A: Kumaon, Blinkworth, Wallich, 103 (US), isotype of Pteris lorigera Wall. Frencu Inpo-Cutna: Cochinchine, 1862- 66, Thorel (Herb. Field Mus. no. 540736); Bokor, Cambodia, Jan. 18, 1926, H. M. Smith 288 (G, US). Siam: Koh Chang, April 2, 1924, H. M. Smith 197 (US). Feperatep Matay States: 9 son (US); Singapore (U. S. Nat. Herb. nos. 22437, 1097181). Anampa IsLanps: Jemaja, Nov. 4, 1928, Henderson 20306 (US). Sumatra: Vicinity of Rantau, Parapot, Bila, March 28—May 10, 1932, Toroes 1832 (NY). Puriwirprne Isuanps: Mindanao, Clemens 166 (F) ; Bucas Island, Oct. 4, 1906, Merrill 5264 (US) ; number” (G, NY). British NorrH Borngo: Mt. Kinabalu, Kundasang, April 7, 1932, J. & M. S. Clemens 29107 (NY, US); Dvusious AnD ResEcTED NAMES Pteris aquilina L. var. mexicana Fée, Mém. Fam. Foug. 9? 8 (1857), nomen nudum.—Fée in Mém. Fam. Foug. 8: 114 (1857) described Pteris aquilina L. var. mexicana, which is clearly Pteridium aquilinum var. caudatum. In Mém. Fam. Foug. 9: 8 (1857) he lists Pteris caudata L. var. mexicana, with- out reference, but it undoubtedly represents a transfer of his earlier var. mexicana. However, he also lists Pteris aquilina L. var. mexicana which, while it apparently is not the same as his var. mexicana, Mém. Fam. Foug. 8: 114 (1857), cannot be defin- 1941] Tryon, Jr..—Revision of the Genus Pteridium 65 itely placed without an examination of the collections cited by him: “Orizaba, W. Schaffner (1834) No 136 et (1856) No 468.” Pteris aquilina L. var. lanuginosa Fée, Mém. Fam. Foug. 9: 8 (1857), nomen nudum.—Although probably a synonym of Pteridium aquilinum var. Feei, this name cannot certainly be placed without an examination of the collection cited by Fée: “W. Schaffner, No 137 Orizaba (1854).” Pteridium aquilinum longifolium, Am. F. Journ. 1: 88 (1910) —The publication of this name was an error. It was 4 new combination based on “Pteris aquilina longifolium Hook.”, a name taken from sheet number 583 in the Herbarium of the American Fern Society. Dr. L. 8. Hopkins informs me that the sheet is actually labeled Pteris aquilina lanuginosa Hook. and that lanuginosa was misread longifolium. Pteris aquilina L. var. decipiens Lawson, Edinb. New Phil. Journ. n. s. 19: 110 (1864). Pteridium aquilinum (L.) Kuhn var. lanuginosum (Bong.) Fernald f. decipiens (Lawson) Fernald, Rhodora 37: 248 (1935) —Lawson’s name cannot be definitely Placed without an examination of the type, which is apparently lost. Lawson says he sent a specimen to D. C. Eaton, but an examination of Eaton’s Herbarium at Yale University failed to reveal such a specimen. The name has been placed under var. pubescens and if this is correct would take precedence over it as an earlier varietal name. However, there is considerable doubt that the plant in question is var. pubescens and I am rejecting the name. Although the plant was collected in the Gaspé, a likely place for var. pubescens to occur as a preglacial relic, and described 48 lanuginose, authentic material has never been collected in Gaspé Co., Quebec and the remainder of the description: “frond bipinnate, thin and membranous, . . . barren.” indicates that the specimen was taken from a young plant. “Lanuginose”’ may apply to some part of the frond other than the lower surface between the margin and the midnerve. In a footnote Lawson himself says that: “Since the above was written, I have had &n opportunity of studying the forms and development of Pteris aquilina [this would be var. latiusculum] and am quite satisfied that the doubtful plant [var. decipiens] is a state of that species, hot old enough to be mature.” 66 Rhodora [FEBRUARY Pteris aquilina L. var. glabra Hook. Sp. Fil. 2: 196 (1858). Pteridium aquilinum (L.) Kuhn var. glabrum (Hook.) Luerss. Rabenh. Kr. Fl. Ed. 2, 3: 107 (1889) —This name includes such a mixture that it cannot be definitely placed in any one variety. Hooker placed the following names under it in synoymy: Pteris aquilina L. (= var. typicum), Pteris caudata Schkuhr (= var. latiusculum and var. pseudocaudatum), Pteris recurvata Wall., Pteris firma Wall. and Pteris excelsa Bl. (= var. Wightianum) and Pteris latiuscula Desv. (= var. latiusculum). His range- citations include the following localities: Europe, Cape of Good ope (= var. typicum), Java (= var. Wightianum), Massa- chusetts (= var. latiusculum), New Orleans (= var. pseudo- caudatum) and Brazil (= var. arachnoideum). EXPLANATION OF PLATES Puiate 650. Fig. 1, Segment of var. Wightianum, X 4, pubescence on one half nat shown, on part of this half the sporangia and venation are not shown; Fig. 2, Segment of var. typicum, X 4, pubescence on one half not pubescence are not shown; Fig. 5, Segment of var. decompositum, X, 4, pubescence on one half not shown, on this half the ri ia and venation are not shown; Fig. 6, Segment of var. africanum, ; pubesc “ke re- : wn venation and farinaceous appearance are also not shown; Fig. 1%, Seg- ment of var. arachnoideum, combination based on Ixia anemonaeflora Jacq., is an Ixia nomencla- turally and a taxonomic mixture as to specimens cited. Ten years later, Baker, in Journ. Bot. xiv. 239 ff. (1876), described six new species, of which two should have been described in Hesperan- tha. His résumé of the genus, dividing it into two sections based on the length of the perianth-tube in relation to spathe-length, appeared in Journ. Linn. Soc. xvi. 93-95 (1877), listing twenty-six species and two varieties. A series of comments upon and additions to Baker’s paper was published by Klatt in 1882,° in the course of which he described several new species. Baker’s treatment of the genus in his Handbook of the Irideae (1892) included thirty species and eleven varieties. As was his custom, when he reduced a specific name to varietal status, he cited the binomial for the third part of the trinomial, so that most of his hew combinations in varietal status were invalid; the same practice was followed in the Flora Capensis. In the index to the latter book, some of these were properly cited as trinomials, with the correct page Teferences, so that it should probably be assumed that the new combinations were made in the index, which may well have been the work of some publisher’s clerk. In the following year (1893), Klatt published a synopsis of the genus in Durand and Schinz, Conspect. Fl. Afr. v. 176-181; this was merely a list of forty species, with synon- ymy and an occasional citation of specimens, retaining nearly all the errors which he himself had ever made, with most of Baker’s in addition. The treatment of the genus in Thiselton-Dyer, Fl. Cap. vi. 65-76 (1896),* differed little from the treatment in Baker’s Handbook, but specimens were cited and a key to the species added. The principal division of the key was made on length of perianth-tube, a highly unreliable character where large groups of species are concerned. Actual errors occur in the key, also; e. g., G. secunda is found opposite the category “Stem and leaves glabrous.” In point of fact, the Specimens which Baker cited have densely short-villous stems, a very prominent character which can hardly be overlooked. Within recent years, a few more species have been described, par- ticularly by South African botanists. One appears to be an old and ost” species rediscovered and redescribed as new, while others are distinct additions to the genus. In the present paper, a number of hew species and varieties will be described. *In Abhandl. Naturf. Ges. Halle xii; reprinted with separate pagination, and cited a. this —" under the abbreviation Ergiinz., since only the reprint is available to me The. se volume was the work of J. G. Baker. 6 FOSTER From this brief survey it is clear that Geissorhiza was erected by Ker-Gawler as a segregate from Jzia, and, for a considerable period of time, its members were largely transfers from that genus. A year after the segregation of Geissorhiza, another segregate, Hesperantha, was proposed by Ker-Gawler.> The line of demarcation between the two closely related genera was clear to Ker-Gawler, but has not always been equally clear to later workers. To this later blurred concept of generic characters, a further source of error has been added in failure to make the essential dissections of floral structures. The types of several species described by Baker in Geissorhiza show no signs of dissection. Had this been done, the plants would undoubtedly have been described in Hesperantha, where they properly belong. The chief, and easily recognizable, character separating the two genera is the length of the style relative to the perianth-tube: Style longer than perianth-tube, the stigmatic branches usually short, recurved, and exceeding the anal Beckie ON EEL is GA Geissorhiza. Style shorter than or equal to the perianth-tube, the branches ong, ascending, often not equalling the anther-apex........ Hesperantha. For the most part, it is not difficult to distinguish Geissorhiza from Ixia. In all but five of the species whose corms are known, the tunics are solid (either imbricate or concentric), not finely fibrous as in Ixia. The spathes are herbaceous, or membranous in the upper portion, but not entirely membranous, scarious, or chartaceous, as in Ixia. The five species of Geissorhiza known to have at least the outer corm-tunics fibrous or fibrous-reticulate have herbaceous spathes, and the seeds, where known, appear to be quite distinct from those of Ixia. These five species are regarded in this treatment as constituting a separate subgenus of Geissorhiza. To discuss the relationships of these three genera adequately can be done only when much more information has been accumulated about their cytology, genetics, and comparative anatomy. At present, nevertheless, some of the salient facts may be outlined. Of the three groups, Geissorhiza is found in the southwestern and western portions of Cape Province in South Africa, with two species known to occur in Madagascar. One species, G. gracilis Baker, has been described from Griqualand East, and another, G. macra Baker, from the Transvaal. A photograph of the type of the latter has been seen, and [ doubt very much if it belongs in Geissorhiza; it strongly suggests Hesperantha or Acidanthera, I have been unable to find any material which can be assigned to G. gracilis, which, to judge from the 5 Ker-Gawl. in Koenig and Sims, Ann. Bot. i. 224 (1804). STUDIES IN THE IRIDACEAE—II 7 description, might well be an Ixia. These two species will be discussed more fully at a later point. With the exception of these two doubtful species, the genus as it occurs in South Africa is found chiefly in the region shown on Pole Evans’s vegetation map® as the evergreen sclerophyll bush. A few species are to be found on the western edge of the Karroo and Upper Karroo, and in Little Namaqualand, desert areas, where they tend to occur at higher altitudes.? Geissorhiza, then, is primarily a component of the non-tropical flora of western and southwestern Cape Province, where it is differentiated into easily distinguishable subdivisions, with numerous endemic species. One of the two species in Madagascar, it should be noted, is an element of the high montane flora, which has marked affinities with the flora of South Africa.’ On the other hand, Hesperantha is well-developed in the eastern portion of Cape Province, in Natal, in the Transvaal, with one species in western tropical Africa (in the Kamerun mountains), and two in eastern tropical Africa, in the high mountains from Kenya Colony south to Lake Nyassa; it has no representatives in Madagascar. Of the species listed by Baker in Fl. Cap. vi. 58-65 (1896), half are com- ponents of the tropical flora of the eastern portion of the Union of South Africa, and only rarely is a species of the tropical flora present also in the western part of Cape Province. So far as can be seen from incomplete studies, Hesperantha is perhaps less easily divisible into well-marked subgroups than is Geissorhiza. Although many species are highly localized in range, there is perhaps less sharp dis- tinction between many species of Hesperantha than is true of Geisso- rhiza With regard to [xia, the situation is still far from clear. The dis- tribution of that genus in South Africa is similar to that of Geissorhiza, with two species described from Griqualand East, and one from Hereroland in the former German Southwest Africa. The description of this last species, I. Dinteri Schz., seems to suggest that it is a apeyrousia, since the stigmas are said to be briefly bifid, a character- istic found in several species of Lapeyrousia in that part of Africa. Like Geissorhiza, Ixia is clearly a part of the non-tropical flora of Cape Province. The question of whether Geissorhiza developed out ° Bot. Surv. 8S. Afr., Mem. no. 15 (1936) 7 See Pearson in Ann. 8S. Afr. Mus. ix. 9, 27 (1911); on the latter page he notes that at i a Cape flora predominates at elevations above 3800 ft. , Baker in Journ. Bot. xix. 362-365 (1881); Baron in Journ. Linn XXv, 246-304 (1889); Humbert, Les Composées ‘he Madagascar, esp. pp. Saecant (1923). 8 FOSTER of Ixia through the small group of species containing G. Bojeri Baker, or vice versa, is one which cannot be answered at present. All that can be said is that the genera are closely related, and that the connec- tion lies in G. Bojeri and its relatives on the one hand and Ixia subgen. Morphixia on the other. In the latter group, the perianth-tube be- comes funnel-form as in Geissorhiza, although the spathes, and probably the seeds, are different from those of G. Bojeri. DraGNostic CHARACTERISTICS Corms. The corms are more or less globose, or ovoid, or even conic in shape, always with a flattened base, although in some this is barely perceptible. It is unfortunate that corms are not known for all members of the genus, but most have been described. On the whole, the shape of the corm seems rather less significant than the nature of the corm-tunics. Of those known, all but five have solid tunics of two types: imbricate or concentric. Imbricate tunics have the basal edges regularly and deeply notched, with the notches smoothly margined. The exception to this is G. ornithogaloides, in which the basal edges are irregularly frayed and serrate, as in some species of Hesperantha. The second type is exemplified by G. juncea, with the tunics completely concentric, tending to split downward from the top, and with the sections of the tunic cusped at the apex. There are, in addition, slight differences in size, color, texture, and hardness which are useful on occasion. The five known exceptions which do not have solid tunics have at least the outer coats fibrous. The limitation is necessary because G. Patersoniae L. Bol., with coarsely fibrous-reticulate outer tunics, has solid, papery, light yellow-brown inner coats, suggesting the inner tunics of some species of Gladiolus. I regard this differentiation of corm-tunics as of great importance, especially since the corm differences are associated with others which will be noted in the discussion of the proper characters. It is this difference in corms, primarily, which has led to the division of the genus into two subgenera. LEAVEs. Thorough study of the available material shows that leaf- venation occurs in two types. In one, the veins are large and rounded, with the edges thickened and downward-revolute. In such plants, the leaf-margin is revolutely incurved, often almost touching the mid- rib, when no other veins are present. A cross-section of such a leaf has an elaborately cruciform outline, if only one conspicuous vein occurs init. One large group of species has only one such vein present STUDIES IN: THE IRIDACEAE—II 9 in the basal and cauline leaves. The basal and cauline leaf-sheaths contract abruptly into a straight, linear-filiform blade, which often appears terete or subterete through the approximation of the leaf- margin and midrib. On the other hand, a second group of species possesses a similar type of venation, but often with more than one vein prominent in the basal leaves, and always several in the cauline leaves. The sheath of the cauline leaves in this group is always ventricose, presenting a heavily striate appearance. The midrib is scarcely more outstanding than the others, when more than one is present. Contrasted with this, a second major type is found with leaves more like those of Jzia in appearance. The veins do not have a rounded appearance, with the edges thickened and downward- revolute, but are more nearly simple ridges. The midrib is far more conspicuous than the others when more than one is present. In cross- section, a single-ribbed leaf in this group presents a simple cruciform outline. Rarely, the cauline leaf-sheaths in this group are ventricose, but if so they are seldom as heavily striate as in the preceding group. In view of the comparative stability of these venation types, this character is the one upon which rests the sectional divisions proposed in this study. An occasional transition form occurs (G. subrigida L. Bol. is the most puzzling of these), but there is far less uncertainty in regard to venation than there is concerning which of several categories should receive the rather variable length of perianth-tube. Leaf-shape is not remarkably diversified. The linear-filiform type and ventricosely-sheathed cauline leaves have already been men- tioned. A few species, such as G. ovata (Burm. f.) Aschers. & Graebn. and G. ovalifolia Foster, have distinctly, if sometimes narrowly, oval leaf-blades. Most of the rest have linear-attenuate leaves, often with the blade contracted above the sheath in a subpetioloid manner, as in G. setacea (Thunb.) Ker and G. subrigida. G. Dregei Baker and G. rupestris Schltr. have bulbils in both cauline and basal leaf-sheaths, while some of their relatives have bulbils in the cauline leaf-sheaths only. The leaf of G. corrugata Klatt is highly distinctive in its extreme flexuosity. STEM AND INFLORESCENCE. If the stem is not simple, the branching may occur near the base within the sheath of the cauline leaf, as in G. monantha (Thunb.) Eckl., or from the very base, as in G. foliosa att. In more cases, branching occurs near or above the middle of the stem. When branches occur near the top of the stem, they are likely to emerge from a pair of spathiform bracts. G. subrigida has 10 FOSTER: such markedly divergent branches for its group that it is possible to utilize this character in the key. For the most part, however, the branches are rather upright. Stems are glabrous, on the whole, al- though a few species such as G. splendidissima Diels have densely short-villous stems. This tends to disappear on the axis of the inflorescence. The inflorescence varies from a single terminal flower to a many- flowered spike, with the floral arrangement distichous to secund. In such species as G. Louisabolusae Foster, the inflorescence is bent at an angle, although the angle is never so great as in the genus Freesia. SpatHes. The spathes vary in shape from lanceolate-acute to oblong-ovate and obtuse. In the latter case, there are sometimes further variations, with the apex somewhat trilobulate, the middle lobe occasionally longer and apiculate. In the past, as in Baker’s key, Fl. Cap. vi. 66 (1896), much stress has been laid on the difference between entirely herbaceous spathes and those which are partially brown-membranous. It cannot be denied that the differences exist and are often pronounced, but they are too inconstant to permit one to lay great stress on them. In the case of G. Bellendeni MacOwan (a synonym of G. monantha), the type and several isotypes have been seen. On the type the spathes are herbaceous, but on some of the isotypes the upper portion of the spathe is quite brown-membranous. The inner spathe is almost always membranous or scarious, bicarinate, with the nerves green, and usually slightly bifid at the apex. Pertantu. The perianth-tube is a character which may be of considerable value in differentiating species if it is used with discre- tion. As between very short tubes, infundibuliform almost from the very base, e. g. G. quinquangularis Eckl. ex Klatt, and long, straight tubes, ampliate near the top, as in G. violacea Baker, there is no question of diagnostic value. Between related species with a com- paratively small difference in tube-length, the character seems valueless. The tubes of the lowermost flowers in an inflorescence are almost always longer than those of the uppermost flowers. Conse- quently, it is necessary to measure the tubes on flowers of approxi- mately the same position in the spike, if the comparative measure- ments are to be significant. Another possible source of error lies in the fact that in a number of species the tube appears to lengthen as the flower matures and withers. Without studying living plants it is impossible for me to be certain of this, but it appears sometimes to be the case. In any event, care has been taken to make comparisons STUDIES IN THE IRIDACEAE—II 11 of tube-length on mature but not withering flowers in approximately the same position in the spike, whenever this could be done. Scanty and poor material has sometimes made it impossible. To use the length of the perianth-tube relative to the length of the outer spathe as a means of separating groups of species, as Baker has done, Fl. Cap. vi. 66 (1896), and even as a basis for creating sections, appears to me to be highly questionable. For example, Baker, in Journ. Linn. Soc. xvi. 95 (1877), placed G. ovata (@. excisa) in Section Weihea, on the ground that its perianth-tube greatly exceeds the outer spathe. A casual inspection of even a few specimens of this species will show a high degree of variability in this respect, the tube varying from almost equal to the spathe to much longer. It must be admitted that in this species the tube is ordinarily very long, but the variability in this and many other species is too great to permit sections to be distinguished on the basis of tube-length alone. The perianth-segments, or tepals, are usually subequal in size, there being little difference between the exterior and interior ranks in this respect. More often, there is a difference in color, a difference most frequently noted when the perianth is yellow or white. In such species, the outer tepals are often suffused or striped with red or pink on their exterior, so that a bicolored effect is produced. Here again the study of living plants would be of value, for the suspicion arises that in a number of such instances the reddish suffusion is the result of aging of the flower or an after-effect of drying in the process of pressing the plant. Within limits, however, this red color appears to be constant. In fact, in most cases, color seems to be a fairly constant character. Certain blue- or red-purple species, such as G. secunda, are more variable in this respect than other species. Again, such yellow-flowered species as G. bicolor (Thunb.) N. E. Br. and some of its relatives vary in the amount of purple found on the tube and base of the tepals. Size and shape of the tepals are naturally of great value in dis- tinguishing not only individual species but also groups of species in some instances. If it were possible to deal with living material in this study, the nature of the flower-shape, whether rotate or open- campanulate or, more rarely, somewhat turbinate, would be of considerable value, but too often this character is obscured in the pressed plant. Sramens. Aside from a few cases, the stamens offer little of diagnostic value. Such species as G. inaequalis L. Bol., G. rosea (Ker.) Foster, G. ramosa, and several others, have unequal stamens, 12 FOSTER one being distinctly shorter. The relative lengths of stamens and tepals is often a point worth consideration, although it is hardly of major importance. By definition of the genus, the stamens are sup- posed to be equilateral in arrangement, but the studies of Dr. Louisa Bolus on living material indicate that this is by no means invariably true. The exceptions will be dealt with in the taxonomic treatment of species. SryLe AND Stigma. The length of these two characters is often useful. G. corrugata has stigmas so long that the plant might easily be mistaken for a Hesperantha, except for the fact that the style exceeds the perianth-tube by several millimeters. In G. ixioides Schltr. ex Foster, the style is short and the stigmas do not exceed the anthers. Generally, however, the long style nearly equals the anther apex, while the relatively short, recurved stigmas exceed the anthers. The style is rarely somewhat unilateral. Two exceptions to the usual type of narrow, conduplicate stigma are found. In these, described by Dr. Bolus as G. eurystigma and G. Mathewsii, the stigma is broad, d group with fibrous corm-tunics, the capsules are small and either turbinate or subturbinate, with numerous small, more or less globose seeds, which are often angular from pressure within the capsule. This statement must be qualified by the admission that, in too many cases, mature capsules and seeds are not available for study. The three exceptions mentioned, G. Bojeri, G. ambongensis H. Perr., and G. Patersoniae, have rather large, ovoid or obovoid, or even almost cylindrical capsules, with winged seeds like those of Gladiolus. ACKNOWLEDGMENTS At the conclusion of this study, it is pleasantly inevitable that I should find myself under considerable obligations to members of the Gray Herbarium staff, above all to Professor M. L. Fernald and Mr. C. A. Weatherby. Despite the numerous demands upon their time, they have always been most generous in their willingness to discuss and to advise upon difficult points. Mr. Weatherby increased my indebtedness by photographing several types at the British Museum and at Paris during the summer of 1939. In addition to the material in the Gray Herbarium (G), material has been seen from the Royal Botanic Gardens, Kew (K), the Botanic Museum, Berlin (B), and the Geneva Conservatoire (Gen). To the Directors of these institutions, Sir Arthur W. Hill, Professor Ludwig STUDIES IN THE IRIDACEAE—II 13 Diels, and Professor B. P. G. Hochreutiner, I am deeply grateful for their kindness in lending specimens. Synoptic TREATMENT OF THE GENUS GEIssorHIzA. Corms ovoid, subglobose, or somewhat conic, the tunics entire and imbricate or concentric, or rarely at least the outer ones fibrous: Agayes basal and cauline, few, erect, sometimes recurved, or (in one instance) to oe mostly linear-attenuate, glabrous or pubescent all sheathing at the base; the cauline leaf- Sheath often ventricose: stem cake or branched, erect, usually flowered, spicate, secund or distichous, usually lax and somewhat flexuose: outer spathe herbaceous or partially membranous, lanceolate to oblong-ovate, acute or obtuse, or trilobulate; inner spathe usually shorter, membranous-scarious (rarely green), hyaline, bicarinate, the nerves green, pel ane tly bifid at the acute apex: ovary small, equal, or sometimes une, inserted at the throat of the tube, the filaments free, usually flattened, the anthers linear or lanceolate, basifixed, Sagittate: style es from the perianth-tube, unilateral in a few s sear, san i scene or proceed Gite the anthers; op. von 22 (1: 827), 1 nomen nudur han Type-spectes: G. obtusata Pectaw in Bot. Mag. xviii. t. 672 803). Kry* a. ee solid, imbricate or concentric; seeds (so far as ~~. = winged; capsules small, turbinate or sub- Hi te popes b. aves not flexuose-tortuose. . . . with conspicuous ve ine; “the midrib hardly more pane icuous than the rest, veins rounded and the edges — narrow, filiform or subterete, the basal leaf- * Since so few species are known to have fibrous corm-tunics, it has been assumed that in unknown cases the corm-tunics were solid. 14 FOSTER yh Pere | 1 concentric, not este peng: in nos. 7 and a .e e. Stamens equal.... 7: —— several-flowered; flowers yellow, at q. Toaven glutin owers a yellow, wnbathed 2... 5 83 . G. humilis. Flowers rl gua r -flushed on the LOO outer s. Leaves me hispidulous....... G. nesta var. bicolor. Leaves hispidulous on midrib ca marg 8. G. hecweliies var. hispidula. g. Leaves not glutin Plan Page ah oat ‘11 em. tall, 2-3-fld.; upper of cauline leaf-sheath subnaviculat te; al leaves not exceeding inflorescence. ..4. G. Pappet. Plant 20-50 cm. tall, several—-many- fid.; cauline sheath not naviculate; basal leaves often exceeding inflorescence. times deeply flushed red-brown on ex- aed al senoaig te pare slg or sub- ge NEUE 6. juncea var. our tik J. infaccebeans & a singh red-purple flower........... . G. furva. ‘ ——* unequal; corm-tunics setter dolysk or subim- cate. eset numerous, small, pink to ad purple; stem Deually bra — (ie ee aA d. Leaves narrow or potlvat 8 baat tease contracted above sheath, widening into blade; cauli a Seat taash G. Mathewsii var. eurystigma. i. Stigmas narrow, ig ii eonduplicate Leaves rise stem glabro sigs 1-2( ~3)-fid., fatheknahed, usually — Se UN eres at ca et ss. oa = G. Rochensis. ; ona var. spithamaea. ves and stem hirtellous.......... 14. G. splendidissima. base dee lant usually large, tepals + 3 em. Outer tepals clear yellow; cauline leaf-blade ane or slightly exceeding inflores- als Maas Gisela ued ours gi 15. G. bicolor. Out a aoe striped rose-purple on exterior; cauline leaf-blade usually much exceedin me _— CNCG . bicolor var. Macowani. Plan ar gpesioe & + 1.5 em. 1 17. G. o-lutea. *. Perianth-tabe’ and tepal base n Sones oo dae STUDIES IN THE IRIDACEAE—II 15 k, Stamens unequal. aves not exceeding inflorescence...18. G. Louisabolusae. At least one basal leaf and the cauline leaf exceeding oe . G. Louisabolusae var. longifolia. k. Stamens equal... .1. l. Perianth-tube to 5 mm. long; tepals to 2.2 cm. long; flushed pinkie. Fs es 20. G. rubicunda. l. bby ies o-wk not over 3 mm. long; tepals ong m. Perianth bieolored, pink and yellowish _ ry a uniform, dull, NSS by Po ae ees et 21. G.W: rightit. m. Perianth P haath or yellowish... .. n. Be rianth yellow. . . .0. . Perianth pale sulphur-yellow; leaves w-nerved. — faa 1-1.5 mm. wide, 1(-3)- ; flowers unmarked... .22. G. sulphurea. Basal eaves 2-3 mm. wide; ou _ tepals striped a on exteri _sulphurea var. arenicola. 0. fers deeper yellow, not sulph eS se sually wider oe Deter vents reddish on exterior.24. G. imbricata. Outer tepals unmarked. Basal leaves cat cauline leaf- blades 3-7-nerved, apically ob- tuse, often rec urved. 5. G. imbricata var. concolor. Basal leaves and cauline leaf- blades net a fabs secs acute, exceedin orescen G. a irl var. Brehmii. n. Perianth almost white, unmark G. sulphurascens. c. Leaves seldom with caer veins other than the mid- rib (except in no. 37); if so, the midrib more prominent than the others; veins ae wit ownward-revolute edges, often pubescent; corm-tunics imbricate, rarely concentric... .p. Pp. Ange pease ne or both, pubescent... . . Leaves glabrous, or sparsely pepilidonr on the sub- tioloid ones of the bla peek ns 1-3-fid. ; branching (if present) near e stem hehe flowe Stem several-fid. ; praten dark blue- or red- purple, the base lighter and pellucid...... 28. G. monantha, Stem and each branch 1-fid.; tepals nearly white, the base pa red-purple...... 29. G. tulbaghensis. Inflorescence many-fid.; if branched, from the middle of stem, or above; flowers relatively i i ee OG SS 30. G. secunda. q. Leaves pubescent... oe r. Stem puberulent.... oe s. Stamens equal; Sahisoleicds aphiee. 3.5... 31. G. ixioides. 16 FOSTER ed 8. ‘ pore nk in er a 1 filament ray shorter than the othe epals purple, Tepals lavender-purple.......... 82. G. a Bric Tepa Wis she hs e le rosy-lilac, obtuse. ....... 34. G. Roperadl Perce-obh sicke-eltiptic; iis: blue-violet 35. G. inaequalis. i. Anthers unequal in length, 1 fi t much shorter, its anther much Shae han the pe pe i hy eG oe a 6 36. G. Leipoldtiv. se bgp brous.... u. Stem divacioately leaked Aor ee 37. G. subrigida. u. Stem simple; if br anced, the branches more or Porianth tube + 1S mm. long; ee and stigma exceeding anther apex. . .39. G. pusilla. Perianth-tube to 9 mm. long; style an nd stigma hardly exceeding reaper oo _ G. namaquensis. w. sg psrng oh imbricat Flowers deep red- ~ ‘blue-purp! le. Inflorescence 3-7-fid.; tepals to 2.4 ong Infloresence 2-3(—4)-fld.; tepals to 1.5 Ws TOME oe a 42. G. erosa var. kermesina. Howees not deep purple. Flowers whitish or yellowish; outer tepals usually red-s of ag externally; perianth-tube light in G. quinquangularis. heen yellowish; eet cn ‘striped rnally; peri ert -tube p. pg and stem — ee MOURA ee ar cere, 45. G. montana. inquangularis var. atrofaux. x. Stamens equal.... y. Leaves bulbilierous; corm-tunics imbricate, thin, rather soft Basal and eauline oe bulbilliferous. blue; haves SEs te es ss. 46. G. Dreget Inflorescence loose and flexuose; flower con- ses asal leaves litiohie, sacha eee Ruek se Pugin. Coeeis Co Nee 47. G. cab Only ait cauline leaves bulbilliferous. Leaves to 10 em. long, linear-attenuate....48. G. B m. long, elliptic-oval....... 49. G. sola y. Leaves not bulbilliferous; corm-tunies concentric (where known), except in n an z. Basal leaves broadly _elliptic-ovate, thick-tex- STUDIES IN THE IRIDACEAE—II 17 z. Basal Seite: not broadly elliptic-ovate, rather hin-textured....A. sabiced not over 6-7 mm Flower light yellow , concolor DASE: 1. G. parva. pier s ite A. Tepals at € ioust 1 em. long, usually neo oe. B. Flowers light facie or violet all; Fen leaves iolet, sometimes pale.......... 53. G. violacea Plant seldom over 25 all; basal leaves short, gna usually falcate; flower light sy-purple, wit eee WARES. Ress Ls oobi 54. G. foliosa. B, toa ala. iat or suffused with red n), or Selina ie C. Inflorescence 2-several-fid. ... . 58. G. inconspicua. C. Inflorescence 1- (rarely 2-)fld...D. — leaves to 8 mm. wide, numer- SES SS ne en ee rence ge 55. G. bracteata. D. Basal leaves 0.5-1.5 mm. wide, few E. ish at Pe oe tall. . 56. G. ‘geminata. E. Plan rely 10 cm. tall (to cm. by a F, Tepals 2.5 cm. ‘or more lon 59. G. eiscibad iikads: F; sepeie li -5 em. or less in length. m. Flower deep yellow; leaves 0.5 mm. or less in Width cc. 2 ee 60. G. Marlothii. fale ——— the basal aguaraiaed frayed and se Basal ne wide; tepals to 1. ger em eed rons 61 G. ornithogaloides. zon leaves wide; seg to 1.5¢ long .62. G. crutealosies var. aon Db: lakves (attadens Meanie ss eS 63. G. corru a. Corm-tunies, at least thay ties ones, fibrous or fibrous-reticu- te; capsules sreagr - e, ovoid to oblong-ovoid; seeds (so far as 2g own) alate. G. Perianth-tube ecelgnt: flowers relatively small. Perianth-tube shorter than outer spathe. Perian th pale lilac-gray, the tepals somewhat — mies South PRON ras i RS ale bag pe e or yellow; Madagascar.......... . G. Bojeri. Peri P ae gate than outer s the; a als p erianth-tube longer than pa p «pp G. Perianth-tube curved; flowers large, yellow, ena red.. .67. G. pre 18 FOSTER Subgenus Eugeissorhiza, subgen. nov. Cormi tunicae omnes solidae, imbricatae vel concentricae; capsulae parvae, turbinatae vel subturbinatae; semina exalata. Type: G. imbricata (DelaR.) Ker- Gawl. It is possible and convenient to divide this subgenus into three sections, retaining the names which Baker first used for two, but radically altering the bases on which they were founded. Originally, the length of the perianth-tube relative to the outer spathe was used, but the nature of the leaf-venation seems to provide a more stable character upon which to base such divisions. Section Rocuea Baker, emend. Baker in Journ. Linn. Soc. xvi. 93 (1877). Nerves of leaves thick, the midrib hardly more prominent than the others, if, indeed, more than one is present, the nerves rounded and their edges somewhat revolute downward; leaf-margins thickened and inwardly revolute; leaves and stems mostly glabrous. cata. Standard-species: G. imbrica This section, again on the basis of leaf-types and venation, almost automatically divides into two subsections. Subsection Filiformes, subsect. nov. Folia angusta, filiformia, vaginae basales abrupte in lamina lineari-filiforme contracta; laminae omnes uni-nervatae; vagina folii caulini non ventricosa; cormi tunicae concentricae, non imbricatae (G. ramosa et G. Burchellio exceptis). ype: G. juncea (Link) A. Dietr. 1. G. numiis (Thunb.) Ker. Plant small, 10-14 em. high; corm ovoid; the coats concentric, hard, brittle, dull brown, long-cusped at the apex (the cusps to 4 mm. long), to 1.2 em. high and + 1 cm. wide: basal leaves 2, linear, recurved, acute, glabrous but glutinous, the midrib and edges thick, to 15 em. long and 1 mm. wide, equalling or exceeding the inflorescence; the single cauline leaf sheathing for 4-4 its length, to 5 em. long: stem simple, terete, glabrous, pero 8 mm. wide, obovate, acute; both sets clear unmarked yellow: stamens half as long as the tepals, the anther and filament each 5 mm. long: style 1.3 em. long, almost equal to the anther-apex; the stigmas 4 mm. long, exceeding the anther: capsule and seeds not seen.—Ker- Gawl. in Koenig and Sims, Ann. Bot. i. 224 (1804). Ixia humilis unb., Diss. Ixia 8 (1783). G. humilis var. grandiflora Baker, STUDIES IN THE IRIDACEAE—II 19 Handbk. Irid. 153 (1892). G. humilis var. G. juncea (Link) Baker, Handbk. Irid. 153 (1892), as to plant. Type: sheet alpha of Ixia humilis in Herb. Thunb. (not seen). Hasirat: sandy places, locally abundant in Cape Div SPECIMENS SEEN: South Africa, without precise locality, Bowie, no. 402 (Gen); W. M. pp isis (K); Zeyher, no. 3964 (KK): Cape Div.: Cape Flats, Pappe (K); Wynberg, Schlechter, no. 1559 (B, Gen); Fish Hoek, Aug. 30, 1896, Wolley Dod, no. 1616 (KK); Klein Slangkop, Sept. 26, 1897, Wolley Dod, no. 3303 (K); Simons Bay, C. Wright, no. 253 ae K): Paarl Div.: by the Berg River, near Paarl, Drége, no. 8472 K, Gen). On oe sheet of Wolley Dod, no. 1616, N. E. Brown has noted: “Matches and is quite identical with the type of Ixia humilis Thunb., sheet a! in Thunberg’s Herbarium.” In view of this resemblance, the description has been based primarily on this collection. True G. humilis has been seen only from Cape Division; Drége, no. 8472, from Paarl, lacks the glutinous leaves and may well be a hybrid in the ancestry wi lai G. imbricata var. concolor Baker has been involved. 2. , var. BICOLOR Baker. Differs from the species in having hie’ cssunligrata aller flowers suffused or striped with red on the e VECR becoming wider.—Baker, Handbk. Irid. 153 (1892). G. setacea @ Ker-Gawl. in Bot. Mag. xxxi. t. 1255 (1810). TYPE: no specimens cited in the original descriptio Hasirar: as in the species, but extending into Caine and Stellen- bosch Divs. SPECIMENS SEEN: South Africa, without precise locality, in 1822, Dr. Thom, no. 723 (K): icy Div.: Capetown, Spielhaus (B); Claremont, ae no. False Bay, ee 1883, MacOwan, no. 2474 (K): Caledon Div. Zwartberg, near Caledon, at 2500 ft., Oct. 21, 1897, Galpin, 0% 4672 (K); eastern side of the Hottentotsholland Mts., at + 350 met., Oct. 10, 1900, Diels, no. 1355 (B); Shaws Mt., Sept. 13, 1931, W. F. Ba rker, no. 8 (K): Stellenbosch Div.: near Somerset in Hottentotsholland, Ecklon & Zeyher, Irid. no. 295 in part (B, Gen). Except for the smaller and more numerous flowers, it would be difficult to consider this as anything but a color form of G. humilis. 3. HUMILIS, var. hispidula A G. humile et var. bicolore nervo primario et foliorum paiiches hispidulis recedit. 20 FOSTER hie Zeyher, no. 1599, Cape Flats, Cape Div. (K, 1soryPes in B, cers chiefly in Cape Div., in sandy plac SPECIMENS SEEN: South Africa, without sie locality, Lalande in Hb. Kunth, no. 1018 (B), Leibold, no. 567 (B), Leibold, no. 574 (B), pacha (B), Niven (K): Cap e Div.: Wynberg Flats, Sept. 16, 1846, ior (K, G); eastern side oF Table Mt., eee Constantia, E Rcklon ph Trid. no. 296 (B , Gen); Table Mt., 00 ft., Feb. 21, 1892, ree | no. 412 (B in part); Claremont, Sept 1892, Schlechter, no. 1568 (B, Gen); Simons Bay, C. Wright, no. 248 in part (G, K), n 265 (G): Stellenbosch Div.: near Sites et in a seats ashaae Ecklon & Zeyher, Irid. no. 295 in part (B, Gen): Caledon Div.: Caledon, in 1862, W. M. Rogers (K). This variety differs from typical var. bicolor in having the leaf- margins and midrib shortly hispidulous, the hairs being appressed against the leaf-blade. 4. G. Papper Baker. Plant about 11 em. tall: corm ovoid, 3-5 mm. wide and 4.5-7 mm. high; the tunics concentric, long-cusped at the apex, light brown: 2 basal leaves to 9 cm. long, closely superposed, subterete-filiform, acute, glabrous; the aula leaf sheathing for 4 branch 1-flowered: outer spathe as long as the perianth-tube, to 5 mm. long, oblong-ovate, abruptly acute or subobtuse, herbaceous, hyaline- edged and membranous at the tip; the inner spathe to 4.5 mm. long: ovary 1-2 mm. long, subturbinate: perianth-tube to 2.5 mm. long, ae in color than the tepals: outer tepals to 6.5 mm. long and 2 mm. wide, obovate-elliptic, obtuse; the inner to 5 mm. long, 2.5 mm. wide, obovate, obtuse, concolorous (?): stamens about es 24 as long as the outer tepals; the anthers 2.5 mm. long; the filaments 2 mm. long: style 6 mm. long, equal to the anthers; the stigmas about 0.75 mm. long: capsule and seeds not seen.—Handbk. Irid. 154 (1892). Type: Zeyher, no. 3965, from the Zonder Einde Mts. (?), Cape Prov. (K, 1soTYPE in es. SPECIMENS SEEN: Cape Province: in arenosis prope Sir Lowry’s Pass, 300 ft., July 16, 1892, Schlechter, no. 1186 (B). Although Baker named this species for Dr. Pappe, and cited one of his collections in the original description, I have not seen this speci- men. The Kew specimen of Zeyher, no. 3965 (cited in the original description) is marked “type,” in the handwriting of N. E. Brown, and the Berlin sheet of this number is marked “'Typus.”’ Consequent- ly, I am accepting this as the type collection. Baker, in Fl. Cap. vi. 68 (1896), states that the leaves are dilated at the base, but I am STUDIES IN THE IRIDACEAE—II 21 unable to see that this is the case. Baker also referred to the flower as “concolorous, reddish.”” From an examination of the type and isotype, it appears to me that the flowers were probably whitish or very light yellow, with the outer tepals red-flushed on the exterior. If the identification of Schlechter, no. 1186 as this species is correct, the flower was undoubtedly somewhat bicolored. 5. G. JUNCEA ates A. Dietr. Plant ieee em. tall: corm ovoid or Shes, 1 em. high (excl. cusps) and 8-9 mm. wide; the tunics Sela smooth, heey brown, the saat cusps to 1 cm. ‘long: basal leaves , the lowermost reduced to a dark brown sheath; the others saan long-sheathing, 10-36 cm. long, 1 mm. wide, Givers. sub- quadrangular or subterete, glabrous, acute, the edges and midrib very thick; the single cauline leaf inserted on the stem below the middle, long-sheathing, 5.5-7 em. long, glabrous: stem simple, terete, glabrous, to 30 cm. long, shorter or longer than the leaves; the inflorescence a lax fester secund, 3—6-flowered spike: outer spathe to 6 mm. long, oblong-lanceolate, obtuse or mucronate, herbaceous below, white- or brown-membranous above, longer than the perianth-tube; the inner spathe as long as the outer: ovary to 3 mm. long, turbinate: perianth- tube 1-2 mm. long: outer tepals to 8.5 mm. long (1 cm. on the type), 2.5 mm. wide, elliptic-obovate, subacute or obtuse, white or pale yellow, flushed red on the exterior; the inner tepals to 7.5 mm. long, whitish: stamens 14-24 as long as the outer tepals; the anthers to 3 mm. long; the a ge lar to 2.5 mm.: style about 6 mm. long, equalling the anther ape ; stigmas 1 mm. long: capsule and seeds not seen.— A. Dietr., Sp. PPL, ii. 587 "(1833). sp ae oe Enum. i. 50 (1821). Ixia scillaris Thunb., Diss. Ixia 13 (1783), aL: Ixia ramosa Ker- Gawl. in Bot t. Mag. xvi. sub t. 549 (1802). Tein Jahobiiproldes Roem. tel ag Syst. i. 385 (1817), nomen illegitimum. Ge pins setacea sensu Klatt in in Linn. xxxiv. 656-657 (1865-66), non (Thunb.) Ker- Gawl. G. humilis var. G. juncea (Link) Baker, Handbk. Irid. 158 (1892), as to name but not as to plant. G. secunda var. G. setifolia -_~ QO za Nee we ve) & qr a>) Lar j AS Com i=) SS La | 5 =a j=} L = ~ s co aa pm —" =e ~I ~] < ~ os id 9) et ° a | b~] 5 fa") > ne : ei e. ramosa (Ker.) Klatt in Durand & Schinz, Conspect. Fl. Afr. v. 180 (1893), as to name. G. humilis var. juncea (Link) Baker ex Klatt in Durand & Schinz, Conspect. Fi. Afr. v. 178 (1893), as to name. G. setifolia Eckl. ex Klatt in Durand & Schinz, op. cit. v. 180 (1893). Tyre: Bergius, from Cape of Good Hope, without precise locality, collected Sept. 19, 1815.(?) (B). Hapirar: abundan t on hills and mountain slopes in Cape Div., occurring elsewhere infrequently. PECIMENS SEEN: South Africa, without precise locality, Burmann f. (Gen), Leibold (B), Lichtenstein (B), Hb. Link, no. 1205 (B), ort Maire (B), Otto, no. 13, in Hb. Spr. (B), W. M. Rogers (K), H. Schlechtendal, coll. in 1826 (Gen), Verreaux (Gen), Zeyher, no. ie 22 FOSTER (B): Cape Div.: top of Table Mt., Ecklon, no. 315 (K, Gen), Table Mt., at 2300 ft., Dec. 25, 1891, Schlechter, no. 92 (B), at 2500 ft., Dec. 25, 1891, Schlechter, no. 130 (B, K), Platte Klip Gorge, Table Mt., 3300 ft., Nov. 18, 1897, Galpin, no. 4688 (K), Platte Klip Gorge, at 2500 ft., Nov. 18, 1897, Galpin, no. 4689 (K); Devil’s Peak, Aug. 25, 1895, Wolley Dod, no. 520 (K), wet rocks beyond Waterfall, Devil’s Peak, Oct. 24, 1897, Wolley Dod, no. 3368 (K); eastern side of Lion’s Head, 300 met., Aug. 28, 1900, Diels, no. 133 (B), Lion’s Head, Aug. 27, 1883, Wilms, no. 3702 (B); Lion’s Rump, Sept., Dr. Pappe (K); Muizenberg, Sept., 1846, Prior (K); Kalk Bay, Sept. 1, 1846, Prior (K); Simons Bay, C. Wright, no. 248 in part (G, K): George Div.: George, Sept., 1827, Verreaux (Gen): near Worcester, Ecklon & Zeyher, Irid. no. 217 (B). A glance at the synonymy of this species will show that there has been much confusion. It is the plant which is commonly known as G. setifolia Eckl., but the original use of that name was invalid. Before the name was finally validated, others had been used for the species, preventing the use of Ecklon’s name. The first valid name which I can find is Ixia ramosa Ker-Gawl., a new name given in 1802 to Ixia scillaris Thunb., non L. So far as can be determined, this name was not transferred to Geissorhiza until Klatt did so in 1893. As his reference to an earlier publication in Linnaea xxxiv. 657 (1865-66) shows, he was under the impression that he was dealing with another plant, the species which will be treated later in this study as G. ramosa Ker ex Klatt. Unfortunately, the presence of this earlier G. ramosa bars the use of this trivial for the plant here called G. juncea. In 1898, also, Klatt validated Ecklon’s name, G. setifolia, by reference to G. setacea sensu Klatt, non (Thunb.) Ker-Gawl. Inasmuch as Link described Ixia juncea in 1821, the name being transferred to Geissorhiza in 1833 by A. Dietrich, most of the con- fusion of nomenclature seems unnecessary. The identity of Ivia scillaris Thunb., non L., and hence of J. ramosa Ker-Gawl., with I. juncea and G., setifolia is shown by N. E. alpha, not sheet beta of Thunberg’s Herb.”’ This specimen, most of the specimens cited under G. setacea sensu Klatt, and the Bergius specimen (B) which I take as the type of I. juncea Link are all con- specific. The details in the original description of I. juncea, especially the measurements, are apparently all taken from this Bergius speci- men, except for the details of the corm. These were probably taken from Hb. Link, no. 1205 (B), which is labelled “J. juncea miihi].”’ STUDIES IN THE IRIDACEAE—II mo An additional note of confusion was added by Baker, in Journ. Linn. Soc. xvi. 94 (1877), who regarded G. setifolia as a variety of G. secunda, and, as usual, made the new combination illegitimately. This was continued in Handbk. Irid. 155 (1892) and in Fl. Cap. vi. 69 (1896); in the index of this last work, the combination G. secunda var. setifolia was validly made. Baker’s brief description does not fit G. juncea, nor do the specimens cited belong to that species. As previously noted, Baker regarded G. juncea as a variety of G. humilis. No other species in the genus, I think, has been so completely misunderstood and confused as G. juncea. 6. G. JUNCEA, var. pallidiflora (Schltr.), comb. nov. Plant 20-50 em. tall: corm m ovoid, 1 cm. wide, 1.5 em. high; the concentric tunics long-cusped at the apex: basal pondiaer to 33 cm long, but usually shorter than the base of the inflorescence; the cauline leaf ov er 4 sheathing, 5-11 cm. long, details as in the species; inflorescence a 2-9-flowered lax, Seicterheaee spike: outer spathe exceeding the perianth- tube, to 7 mm. long; inner spathe as long: ovary + 3 mm. long: outer tepals to 9 mm. long, 3 mm. wide, obovate, obtuse, yellow, flushed red-brown on the exterior; the inner tepals to 7.5 mm. long, obovate, obtuse, yellow: stamens about 34 as long as the outer tepals; the anthers 4 mm. long; the filaments 3 mm. long: style to 7 mm. long, not quite equalling the anther; the 3 mm. stigmas recurved, e ing the anther: capsule and seeds _ present.—G. pallidiflora Schltr, in Engler, Bot. Jahrb. xxvii. 98 (18 99). Type: Schlechter, no. 9088, in arenosis humidis inter Ceres Road et Bainskloof, at 1000 ft. alt., Nov. 11, 1896 (B). Hasrrar: Ceres, Malmsbury, Piquetberg, Tulbagh, and Paarl Divs. SPECIMENS SEEN: South Africa, without precise locality, Burmann f. (Gen); Drakenstein Mts., 2000-3000 ft., Drége (B, Gen): Piquetberg Div.: in collibus ad flum. Ber erg Rivier pr. Piquetberg, Sept. 10, 1892, Schlechter, no. 5273 (B): pentane wie : in the es ger of H. Bolus, no. " 5388 (K), top ‘of Witsenberg Mt. ‘near Steendahl, 2006 ft., Oct., 1884, H. Bolus, no. 547 7 (K), Winterhoeksberg, 1000-5000 ft., E cklon & Zeyher, Irid. no. 227 in part (B, Gen), W interhoeksberg, 000-5000 ft., Zeyher, no. 3963 (B): Paarl Div.: between Paarl and Pont, Drege, no. 8460 (K, Gen). If only the type and other specimens from the more northern portion of its range were seen, it might be possible to maintain this plant as specifically distinct from G. juncea. Unfortunately, in the more south- 24 FOSTER ern parts, it intergrades rather badly with G. juncea. On the whole, it tends to have deeper yellow flowers, often, but not invariably, dis- tichous rather than secund in arrangement. The reddish flush on the exterior of the outer tepals may be lacking in some cases. Generally, it is somewhat taller and fewer-flowered than G. juncea, but specimens have been seen with more flowers than are found in the species. mosa Ker-Gawl. ex Klatt. Plant to 48 em. tall: co ovoid to sahelebehe, to 1.3 em. wide and 1.3 em. high; the rabboiekts tunics cusped at apex: basal leaves 2, to 30 em. long and 2 mm. wide, nearly equal to or shorter than the base of the inflorescence, linear, sheathing cauline leaf to 24 em. long: stem terete, glabrous: the in- florescence branched (rarely simple); the branches subtended by short spathiform bracts, erect or divaricate, each branch 3-8-flowered in a upper flowers in the inflorescence obscurely tridentate, the lower portion herbaceous, the upper part membranous, oblong-ovate, some- what obtuse, the inner spathe about as long as the outer: ovary turbinate, 2-3 mm. long: perianth-tube shorter than — 2-3 mm. aes light straw-brown in in color: tepals subequal, to 1.3 em. long and m. wide, but usually smaller, oblong-obovate, obtuse, reddish- on stamens often nearly equal to the tepals in length, one stamen 0.75 1 ae long, da rk-brown. ee Geel x Klatt in Linn. xxxiv. 657 (1865-66). G. ramosa Sa erg Klatt in Desa & Schinz, Conspect. Fl. Afr. v. 180 (1893), as to nee G. secunda, var. G. ramosa —) Baker, Handbk. Irid. 155 (1892). Type: Ecklon & Zeyher, Trid. no. ” in part (B; 1soType in Gen), ae near Worcester, Cape Provi HapitaT: sandy places, tea di Kees Div. to Riversdale Div. SPECIMENS SEEN: Tulbagh Div.: Tulbagh, Pappe (kK); Winterhoeks- rg, 1000-5000 ft., Ecklon & Zeyher, Irid. no. 227 in part (B); on the Witsenberg ®, Ecklon & Zeyher (G, B); Witsenberg, 2000-3000 ft. Leyhet, no. 1596 (B, Gen): Worcester Div.: Dutoits Kloof, Drege, no. 848la (K, B, Gen): Cape Div.: Rietvalley, Mund & Maire i in part (B): Cale Div.: Vousleat Dec. 2, 1896, Schlechter, no. 9577 (K, Gen): Swellendam Div.: Swellendam Mt., ca. 2500 ft., Oct. 16, 1897, re no. 4680 (K); Tradouw Pass near Zuurbraak, 1200 He, Oct. 15, 1897, Galpin, no. 4697 (KX): Riversdale Div.: Valley River, on the Lange Bergen near Kamscheberg, Burchell, no. 7044 (K). "2. So far as can be determined, Ker-Gawler never described a G. ramosa, nor did he use it as a new combination. He did create [xia STUDIES IN THE IRIDACEAE—II 25 ramosa as a new name for I. scillaris Thunb., non L., both names occurring in the synonymy of G. juncea of the present treatment. Klatt described this species, cited specimens, and gave no synonymy, so that as it stands it is a new species, incorrectly ascribed to Ker- Gawler. In 1893, Klatt cited it (in Durand & Schinz, Conspect. FI. Afr. v. 180) as G. ramosa (Ker) Klatt, giving I. ramosa Ker as a basinym, and the place of publication as Klatt in Linn. xxxiv. 657 (1865-66), showing that he probably intended it as a new combina- tion in the first place. It seems to me that a time-lapse of nearly thirty years is too long to permit treating this as a new combination dating from 1865-66. Unless G. ramosa Ker-Gawl. ex Klatt is treated as having been aduenibed as a new species, G. juncea would become a nomenclatural synonym of G. ramosa (Ker) Klatt, as would the name G. setifolia Eckl. ex Klatt, forcing the use of the trivial ramosa for an unbranched plant, and leaving the plant here treated as G. ramosa nameless. I see no point in upsetting the nomenclature in this manner. In his description in 1865-66, Klatt cited three specimens, all of which have been seen. The first, a Drége collection, is not G. ramosa as that plant is generally understood, and the third, Ecklon & Zeyher, Irid. no. 234, is a mixture of two species of Hesperantha. The second, Ecklon & Zeyher, Irid. no. 229, now at Berlin, agrees reasonably well with the description, despite the fact that the two left-hand specimens appear to be G. juncea. Consequently, I have selected this sheet as the type of G. ramosa Ker-Gawl. ex Klatt, excluding the two left-hand specimens Superficially, this species bears a close resemblance to G. juncea, but it has imbricate corm-tunics, purplish flowers, unequal stamens, and is usually well-branched, so that the two species should be fairly easily distinguishable. Baker, possibly in despair, treated this plant as a variety of G. secunda, Handbk. Irid. 155 (1892). Most of the specimens cited for the variety in Fl. Cap. vi. 70 (1896) are actually G. ramosa, except for Mader in Herb. MacOwan, no. 2165, which seems to me to be G. secunda 8. G. Burchellii, ape nov. Cormus incompletus, ovoideus vel subglobosus, + 9 mm. altus, 7 mm. latus; tunicae imbricatae (?), brunneae, nitidae, aa - cuspidatae: folia basalia 2-3, 10-20 em. longa, 2 mm. lata, basi vaginantia, lamina filiformis, longe attenuata, acuta, glabra, nervo primario prominente, marginibus incrassatis; folium caulinum 4-9 em. longum, 14-24 vaginans: caulis simplex, teres, glaber, 2-3-fl., 23-38 cm. longus, bracteis 2 brevibus, 1-2 cm. longis, 26 FOSTER rik cir rasp spatharum valvae aequales, ad 1 cm. longae, exterior oblongo-ovata, s Se vel obtusa, herbacea, apice i dat suffusa: ovarium ad 4 mm. longum, turbinatum: sueatal tubus tenuis, ad in mm. longus: tepala subaequalia, ad 1.8 em. longa, 5 mm. lata, elliptico-obovata, apice obtuse rotundata, vix retusa, concolora, ilacino-purpurea: antherae ad 5 mm. longae, filamenta inaequalia, 2 ad 1.3 em. longa, 1 ad 1 em. longum: stylus ad 1.2 em. longus, quam aie Toineiors parum brevior; stigmata ad 2 mm. longa, stamina excedentia: capsula seminaque non visa. Type: Schlechter, oe Gre (B), from Swellendam Div.: Langebergen near Zuurbraa shee an HaBirTaT: ap eG in moist places, at rather high altitudes (3500 ft., itis ae to the data on the t SPECIMENS SEEN: South Africa, without _ precise locality, Burchell, no. 7322B (K). This species bears a deceptive resemblance to G. hesperanthoides Schltr., but the corm-tunics are not fibrous and the stamens are very unequal. It resembles G. ramosa in its unequal stamens, but the flower-shape and -color, as well as the size and number of the flowers, differentiate it easily from that plant. Although this species has been named for the great collector, W. J. Burchell, whose no. 7322B was the first specimen seen, it has seemed advisable to designate Schlechter, no. 2160 as the type, since that has full data as to locality and habitat. The Burchell specimen may be regarded as a co-type, if the locality-data can be ascertained. 9. G. rurvA Ker. ex Baker. Plant about 10-11 cm. tall: corm small, + 1 cm. high, 5 mm. wide, ovoid; tunics concentric, hard, — ivicalte — basal leaves 2, filiform m, ina — pie the anthers: capsule an ode not presen nt.—Baker, Handbk. Irid. 155 (1892); Ker-Gawl. in Koenig ee ces Ann. Bot. i, 224 pei nomen nudum. Tyre: Drége, no. 8478 a Vines in Gen), collected between Paarl aa Pont, Cape Provi Hasitat: low (under 1000 rs ), stony hills, covered with shrubs, according to Drége’s label. STUDIES IN THE IRIDACEAE—II1 27 Although Baker attributed this species to Ker-Gawler, the latter’s G. furva was a nomen nudum, and apparently remained so until Baker validated the name by a description in 1892. Of the three specimens cited in the original description, I have seen only one. Since this fits the description reasonably well, and even has the remnant of a corm, it may be designated as the type. Baker stated that the anthers were longer than the filaments, but in the type, they are obviously only half as long. G. furva can be regarded as a connecting link between subsection Filiformes, where its leaf-structure would place it, and the next sub- section, for its flower has definite similarities to that of G. Rochensis. Subsection Ventricosae, subsect. nov. Folia angusta vel lata; folia basalia super vaginas angustata, in lamina dilatata; vagina folii caulini ventricosa; laminae omnes plu ee cormi tunicae cers Type-species: G. seis (DelaR. wl. G. Matuewst L. Bol. Plant 10-22 cm. tall: corm ovoid, to 1 cm. ane gh, 8 mm. diam.: basal leaves 2, 7-15 cm. long, 2-3.5 mm. wide, Nigahenattotens the blade narrowed above the sheath then =e multinervate (7-9), acute, glabrous; the cauline leaf 8-12 with a ventricose sheath, the blade remanent indented eihvally above the sheath, multinervate: stem 1 1-branched within the cauline leaf-sheath, 7-21 em. long, ee eae wae pening cence a 1—5-flowered flexuose, secund spike: outer spathe cm. long, ovate, somewhat membranous in the upper half, Sieh ts, the middle lobe acute or mucronate; the inner spathe nearly as long: ovary + 3 mm. long, subturbinate: perianth-tube about 2 mm. long: tepals subequal, to 1.6 cm. long, 9 mm. wide, obovate-cuneate, reddish on the lower half, the remainder dark blue-purple: anthers 4-5 mm. long; filaments 9 mm. long: style 7 mm. long; the stigmas to 4 mm. long, broad, flat, thick, the edge heavily papillate. —L. Bol. in Ann. Bol. Herb. iv. 42 (192 6). Type: J. W. Mathews, Actes —Sept., 1923, near Darling (in Bolus Herb., no. 18502, not see Haprrat: damp and ahs places SPECIMENS SEEN: Malmesbury Div. Malmesbury, Sept., 1892, Th. Kiissner (B). Although I have not been able to see the type, the single specimen seen agrees so well with the excellently detailed description given by Dr. Bolus that there seems no doubt of its identity. Certain portions of the description given here have been drawn from the original = G. Matuewsu L. Bol., var. eurystigma (L. Bol.), comb. nov. Differs from the species primarily in its longer perianth-tube (6-7 28 FOSTER mm.), shorter filaments (4.5 mm.), and longer style (1.4 em. eurystigma L. Bol. in S. Afr. Gard. xxi. 281-282, fig. E (1931). pee Be es DelaR.., Doar 17 (1766), non Bergius, Pl. fin 6 (1767). he Bolus, Sept., 1931, near Mamre, Malmesbury Div. (in Herb. ea no. 19876, not seen; ISOTYPE in K), Hasirat: known to me only from Malmesbury Div., in damp or gegen sandy places. SPECIMENS SEEN: South Africa, without precise locality, Burmann f. (Gen), Hb. Ventenat (Gen): Malmesbury Div.: Groen a) = 500 ft. - Sept., 1884, MacOwan in Herb. Norm., no. 506 i &; Ks Ge a: between Mamre and Darling, Sept., 1932, pot ae in Nat. Bot. ’Gard., no. 3316/32, J. Steytler (K). With the exception of Ker-Gawler, in Bot. Mag. xvii. t. 597 (1802), authors have generally assumed that Ixia secunda De la Roche and I. secunda Berg. are identical. Such an assumption was even made by Bergius himself in the Addenda to his work, Pl. Cap. 360 (1767). Nevertheless, the two, as Ker recognized, are not the same. De la Roche’s plant was apparently glabrous, for no mention of pubescence was made; the flowers were described as having a purple throat and a very dark blue limb, not concolorous; the perianth-tube was longer than the spathes; and the stigma was broad and membranous, with acrenate margin. J. secunda Berg., on the other hand, was reported as having a shortly villous stem, long, glabrous leaves, several to many red-purple or blue-purple concolorous flowers in a secund spike, with the perianth-tube shorter than the spathes, and the stigma narrow and apparently conduplicate. This latter plant, common and wide- spread in western and southwestern Cape Province, is the one trans- ferred to Geissorhiza by Ker-Gawler in Bot. Mag. xviii. sub t. 672 (1803). Therefore, it serves to block the transfer of J. secunda DelaR. to Geissorhiza under that trivial name. As it happens, Dr. Louisa Bolus has described two species with stigmas similar to that described by De la Roche, and they are the only plants in the genus which do fit that description. As they appear to me to be conspecific, although varietally distinct, the long- tubed plant, G. eurystigma, which can be identified with I. seeunda DelaR. tae some confidence, is here treated as a variety of G. Mathew pina these plants have a superficial resemblance to G. monan- tha, they can readily be distinguished by their smooth stems, usually 3-4-flowered inflorescences, trilobulate spathes, longer perianth-tubes (at least in var. eurystigma), and, above all, by the broad flat stigmas. In the variety, there is a tendency for the style and stamens to be STUDIES IN THE IRIDACEAE—II 29 unilateral, but this is difficult to distinguish in pressed material. “soto it should be noted, the basal leaves are very few-nerved. C. RocHENsIs Beds Ker-Gawl. Plant rather slight, a em. tall: corm ovoid, + 1 em. high; the imbricate tunics finely cusped at the apex: ‘weal pee tag 2, glabrous, filiform-subulate (or almost quadrangular), exceeding the inflorescence, + 1.5 mm. wide; the cauline leaf with a very ventricose sheath, to 8.5 cm. long: stem simple, terete, glabrous, 1—2-flowered: outer spathe to 1.1 cm. long, ovate-lanceolate, longer than the perianth-tube, herbaceous, mem- branous at the truncate, trilobulate apex, the middle lobe somewhat longer than the others; the inner spathe somewhat shorter, sometimes subherbaceous: ovary to 3 mm. long: perianth-tube to 6-7 mm. long: tepals subequal, to 1.8 cm. long, + 1.1 cm. wide, co vathcannate obtuse, dark purple, with a basal foveole: stamens nearly as long as the tepals; the anthers 4-5 mm. long; the filaments + 8 mm. long: style equalling the anthers; the recurved stigmas exceeding them: capsule and seeds not seen.—Ker-Gawl. in Koenig and Sims, Ann. Bot. i. 224 (1804). Ixia anc Ker-Gawl., var. palmaris Ker- : wm ‘ A. Dietr., Sp. Pl. 586 (1833). G. Tilikna Sweet, Hort. Brit. (ed. 1) 399 (1826-27). G. ar hit hae Ergiinz. 56 ey) in part. Rochea venusta Salisb. in Trans. Hort. Soc. Lond. i. 322 (1812). G. Rochensis, var. G, brat yA (Sweet) aie Handbk. ted, 156 (1892), as to plant, but not nam P described and figured from living “reas with a sheet labelled Ixia azurea in Herb. Banks. cited in additi Hapirat: aero places, Malmesbury Div. to Ca See e Div. SPECIMENS SEEN: Malmesbury Div.: Groenekloof, 500 ft., Sept., 1884, MacOwan in Herb. Norm., no. 506 in part (G, B, K ; near Yeerfontein Sept. 13, 1931, ne M. pat no. 1345 (K); ad ae pe Malmesbury, 250 met., Oct. 1, 1892, Schlechter, no. 1610 i Sart t (B): Paarl Div.: near Paarl, Drége, no. 8486 in part (B, type of G. tulipifera; Gen): Cape Div.: Table Mt., Ludwig in part (B); sand dunes at Saldanha ty Ecklon & Zeyher, Irid. no. 219 in part (Gen). Since Ixia Rochensis var. palmaris was the first to be described and the only variety figured by Ker-Gawler in 1802, it must be taken as the typical variety of Geissorhiza Rochensis. In its extreme forms it appears distinct, but it intergrades markedly with the following beinstt occurs with it, and possibly should not be separated from it. G. RocHENsIs, var. SPITHAMAEA (Ker-Gawl.) Ker ex Baker. ei tall, 10-30 em.: corm ovoid, 1 em. wide and 1.2 cm. high; the imbricate tunics hard, dark brown, finely cusped at apex: basal leaves 2, 8.5-24 cm. long, 1-2.5 mm. wide, thick, sometimes almost quadrangular, linear, acute, glabrous, edges, midrib, and sometimes 30 FOSTER 1-2 other nerves thick and prominent; the cauline leaf with a very ventricose sheath, multinervate, glabrous, the blade extremely narrow, 7-18 cm. long, usually equalling or exceeding the inflorescence: stem simple or branched, terete, glabrous, slender or stout, flexuose; the inflorescence a 2-—7-flowered lax, secund, flexuose spike: outer spathe to 2 cm. long, usually 1-1.5 em. long, ovate-lanceolate, longer than the tube, herbaceous except at the truncate, trilobulate apex; the inner to 1.5 cm. long, usually shorter, sometimes mea: ovary to 4 mm. long, piles See perianth-tube 6-8 mm. long: outer tepals to 2.3 cm. long, 1.3 mm. wide, sea te-suscata, obtuse, dark purple; the inner like the outer, but 3 mm. shorter: stamens 34 as long as the tepals; the anthers 6 mm. long; the filaments to 1.2 cm. long: style to 2.5 em. long, nearly as long as tepals; fee 4 mm. long, dark purple: capsule turbinate, 6 mm. long; seeds small, brown, ae er —Baker, Handbk. Irid. 156 ine fie eeetly attributed o Ker. Izia Rochensis, var. spithamaea Ker-Gawl. in Bot. Mag. Xvii. ee t. 598 (1802). G. Rochensis, var. multiflora A. Dietr., Sp. PL 586 (1833). I2ia radians Thunb. in Hoffm., Phyt. Blitt. i. 3 (1803). G. radians (Thunb.) Diels in Engler and Prantl, Nat. Pflanzenfam. Hort. Brit. (ed. 1) bh oG. monantha Sweet, Hort. Brit. (ed. 2) 503 (1830), non (Thunb.) Eckl. G. tulipifera Klatt in pert her 56 (1882). Jia secunda sensu Houtt., Nat. Hist. Deel II, xii. 33-34, t. 78, f. 1 (1780), non DelaR. G. Rochensis , var. G. acai (Sweet) faker Handbk. Irid. 156 (1892), as to name, but not as to plan ‘PE: original description cited a sheet in Herb. Banks., labelled Ixia bet In the British Museum, if the specimen has been pre- not see Poca with the species. ECIMENS SEEN: South Africa, without precise locality, Burmann ie (Gen) Leibold (Gen): Malmesbury Div.: near Malmesbury, 250 met., Oct. 1, 1892, Schlechter, no. 1610 i in part (B); Groenekloof, 500 ft. Sept., 1884, MacOwan in Herb. Norm ., no. 506 in part (G, K, B, Gen) Sept. 23, 1883, M acOwan, no. 2281 (K); Malmesbury, Sept., 892, Th. Kiieomer i in part (B); near Groenekloof, 300 ft., Oct., 1878, i. Bolus, no. 4340 (KK): Paarl Div.: near Paarl, Drege, no. 8486 in part (B, K): Cape Div.: Table Mt., Ludwig (B); dunes at Saldanha Bay, below 1000 ft., Ecklon & Ze ryhe er, Irid. no. 219 (B); Caapsche Vlakte, Oct. 14, 1815, Bergius (B). Like G. monantha, this species and its variety might be confused with G. Mathews, var. eurystigma, but the stigma and cauline leaf of G. Rochensis are quite different. The latter has the cauline leaf adaxially incised above the sheath, then widened into a relatively broad, many-nerved, tapering, linear-ensiform blade. G. Rochensis STUDIES IN THE IRIDACEAE—II 31 has the cauline leaf sheath very ventricose, narrowing abruptly into a narrowly linear, sometimes almost subulate, blade. Its stigma is narrow in comparison with that of G. Mathewsii, var. eurystigma. On the Kew sheet of MacOwan in Herb. Norm., no. 506, N Brown has noted: “Matches the type specimen of Ixia radians, Thunb. in Thunberg’s Herbarium.” If this is the case, it is rather surprising to find Diels, in his treatment of the family in the revised edition of Engler and Prantl, making J. radians synonymous with Geissorhiza hirta. It becomes even more surprising in view of the fact that the basinym, Iria hirta, was described in 1783, while I. radians dates from 1803. The Burmann specimen cited above, it might be noted, is apparently that from which Houttuyn made his drawing of the plant which he called [xia secunda. His drawing is so faithful that it even reproduces one or two tears or breaks in tepals and spathes, which are plainly visible in the pressed specimen. 14. G. sPLENDIDISSIMA Diels. Plant 15-22 cm. tall: corm ovoid or subglobose, basally somewhat truncate, to 1.3 cm. high, 1 em. wide; long, 4 mm. wide, the numerous prominent ribs hirtellous: stem 1- branched, or simple, the main axis to 13 cm. long, terete, near the hairs recurved upwards, slightly hooked; the inflorescence a 4—5- flowered, lax, secund spike: outer spathe to 1.9 cm. fee ovate- lanceolate, acute, herbaceous, subglaucous, membranous at the apex, very much longer than the perianth-tube, carinate, many-nerved, the nerves slightly hirtellous; the inner spathe to 1.4 cm. long: ovary 2-4 mm. long, erat iy gi perianth-tube to 5 mm. long: tepals brilliant dark purple, deeper at the base, subunguiculate, the claw and tube yellowish-green; tals tepals to 2.5 cm. long, 1 cm. wide, obovate, acute; the inner to 2.2 cm. long, 1.3 em. wide, ovate-obovate, subobtuse: anthers 7 mm. long; filaments to 1 em. long: style 1.5 em. long; the stigmas recurved, papillose, 3 mm. long, a trifle exceeding the anthers: capsule and seeds not present.—Diels in Engler, Bot. ire baste 117 sang : Diels, 627, Sept. 13, 1900, from Oorlogskloof, on the heat Calvinia Div., at 750 met. alt. (B). Hanirar: hard clay-loam, i in scanty shrubbery, Calvinia Div. er on Calvinia Div.: Loeriesfontein Road, Nieuwoudt- ville, Sept., , L. Bolus, in Herb. Bolus, no. 19439 (K); Nieu- MA aa a neg 1931, H. Buhr, cult. in Nat. Bot. Gard., no. 823/30 Kk) oe FOSTER This beautiful plant is well-deserving of its name, for it is probably the most striking member of the genus, as to color. Diels speaks of its relationship to G. hirta (to use the name then current), saying that it can be distinguished by its green spathes and by the perianth-color. Still more distinctive are the facts that it has a hispidulous or hir- tellous stem, while G. hirta has a glabrous stem, its basal leaves are of a wholly different type, and the pubescence on the leaf nerves is quite different. It appears to me to be much closer to G. Rochensvs var. spithamaea. 15. G. stcotor (Thunb.) N. E. Br. Plant about 12 cm. tall: corm not seen: basal leaves 2, to 10 or 12 em. long, basally sheathing, narrowed above the sheath and widening into a linear-attenuate, sub- obtuse, glabrous, subfaleate blade, 3-4 mm. wide, 3-7-nerved, the edges thickened; the cauline leaf to 10 em. long, with a basal ventri- cose, multinerv ed sheath, adaxially incised at the apex, widening into a linear-falcate, attenuate, + 7-nerved, glabrous blade: stem simple, or 1-branched within the cauline leaf-sheath, terete, glabrous, 4-11 cm. long; the main axis 1—3-flowered, the branch 1-flowered: outer spathe to 1.3 cm. long, oblong-ovate, obtuse or abruptly acute, herbaceous, brown- or cenlesnine! at the apex; the inner spathe nearly as long: ovary 2-4 mm. long, subturbinate: perianth-tube to 5 mm. long, shorter than the spathes, dark purple: tepals subequal, + 3 cm. long, 1 em. wide, obovate, obtuse, yellow, with the base dark purple: stamens little more than half the length of the tepals; the anthers to 6 mm. long; the filaments to 1.3 cm. long: style to 2.5 cm. long, about equal to the anther apex; the stigmas to 5-6 mm. long, exceeding the anthers: capsule to 8 mm. long, obovoid-turbinate; the seeds immature.—N. E. Brown in Journ. Linn. Soc. xlviii. 44 (1928). Ixia bicolor Thu nb. in Hoffm , Phyt. Blatt. i. 3 (1803 Type: sheet alpha, labelled Ixia bicolor i in Herb. Thunb. (not seen). Hasirat: sandy places; known to me only from Malmesbury Div. SPECIMENS SEEN: Malmesbury Div.: near Malmesbury, 400 ft., Oct. 2, 1892, Schlechter, no. 1625 (B); near Mamre, Sept., 1888, MacOwan, no. 2488. This little-understood species was treated as identical with G. purpureo-lutea Baker, by Baker in Fl. Cap. vi. 68 (1896), but it eems to me that N. E. Brown was correct in regarding it as distinct. By error, Brown gave the reference for the basinym, Ixia bicolor, as Thunb., Diss. Ixia 15 (1783), but no such species is included in the work; the correct reference is given above. In his description of the type, Brown stated that the perianth-tube was nearly twice the length of the spathes. Except in the case of G. ovata, in which this condition is sometimes true, I know of no species of Geissorhiza in which the STUDIES IN THE IRIDACEAE—II 33 tube is anywhere nearly twice the length of the spathes. At times a flower becomes detached from the ovary in pressing, with the result- ing gap concealed by the spathes, and unless a complete dissection is made, or unless the specimen is held up against a powerful light, the gap will not be discovered. This may be the case with G. bicolor, but in any event, I regard the length given as exceptional, and feel that the length given in the above description it probably more character- istic of the species. On the whole, it appears to me that the plant is a state of a species better-characterized by the following variety with which it intergrades. BICOLOR, var. Macowani, var. nov. A specie floribus numer- osioribus, folio basale uno et folio caulino inflorescentiam excedentibus, et lamina perlonga folii caulini differt.—G. inflexa sensu Baker, Fl. Cap. vi. 73 (1896), for the most part, non (DelaR. ) Ker-Gawl. Type: MacOwan in Herb. Austr.-Afr., no. 1568, near Malmesbury, 400 ft., Sept., 1894 (G; IsoyTpEs in K, B, Gen Hasrrat: sandy places, Malmesbury and Tulbagh Divs. SPECIMENS SEEN: Tulbagh Div.: Tulbagh, Dr. Thom (K). This differs from the species chiefly in having more numerous flowers, and at least one basal, as well as the cauline, leaf exceeding the inflorescence. The corm is ovoid, subglobose, the imbricate tunics hard, dark brown. 7. G. PURPUREO-LUTEA Baker. Plant 6-17 em. high: corm ovoid- conic, 1 em. high, 7 mm. wide; the imbricate a oa dull gray- brown, apically cusped: basal leaves 2, to 13 ecm. long, 3 mm. wide, shorter than the stem, narrowed above ‘the sheath seit expanding into a linear-faleate, glabrous, acute blade, the edges, midrib, and some- times 2 other nerves thickened; the a cauline leaf with a subven- tricose basal sheath, heavily ribbed, 2 —10 em. long, incised above the sheath, the blade few- to many-nerved : stem cae or once-branched, terete, glabrous, terminating in a 1—4-flowered lax, distichous spike, the branch 1-2-flowered: outer spathe to 1 cm. long, exceeding the tube, lanceolate, herbaceous, purple at the acute apex; the inner spathe nearly as long: ovary to 2 mm. long, turbinate or subglobose: perianth- 2-4 mm. long, dark purple: tepals subequa 1, 1.5-1.3 em. long, 5 mm preg ovate, obtuse, entire, purple at the base, the selaubatics yellow, often with a broad purple or red strip on the exterior of the outer series: stamens about 24 as long as the tepals; the anthers to 4.5 mm. long; the filaments as long, purple-based: style to 1.3 em. long, about equalling the anthers; stigmas 3 mm. long, ee so anthers: immature capsule turbinate, 5 mm. long; seeds not seen Baker i in Journ. Bot. xiv. 238 (1876). E: Drége, no. 8476, between Paarl and Pont, below 1000 ft. (K; ict in Gen). 34 FOSTER Hasirat: the type was collected on low, stony hills. SPECIMENS SEEN: Malmesbury Div.: ee Post, Sept. 13, 1934, T. M. Salter, no. 4756 (K, in part): Paarl Div.: near Wellington, Sept. 17, 1932, L. Bolus, in Herb. Bolus, no. 20327 (K); between Paarl and Pont, on low, stony hills, in sparse shrubbery, in Sept., Drége, no. 8475 (Gen). The species is undoubtedly close to G. bicolor and is perhaps only doubtfully distinct from it. In the original description, Baker stated that the species was close to G. humilis; I do not understand this. 18. G. Louisabolusae, spec. nov. Cormus ovoideo-subglobosus, 1 em. altus, 1 cm. latus; tunicae megs apice cuspidatae: folia basalia 2, basi breviter vaginantia, ad 1 onga, super vaginam paullum angustata, in lamina 1.5 mm. 4% “Sivormis, acuta, glabra, dilatata; folium caulinum ad 13 em. longum, vagina longa, multi- s caulini 1-ramosus paullum super basim, axis primarius ad 20 cm. Hea aie eet inflorescentia spica laxa, secunda, flexuosa, 4-6 fl., ra 1-3- patharum valva exterior ad 1.2 em. longa, oblongo- ovata, - estan apice obtusa, esha espana interior ad 1 em. longa: ovarium subturbinatum, ad 3 mm. longum: perianthii tubus ad 3 mm. longus: tepala subaequales, ad 2. 6-2. ne em. longa, 1 em. lata, subunguiculata, lamina ovata, subobtusa, pallide flava, clara: stamina inaequalia; filamenta 2 ad 1.2 em. longa, 1 ad 8-9 mm. longum; antherae ad 7 mm. longae: stylus ad 1.7 cm. longus, quam apicem antherarum parum brevior, fortasse nhilateeslin stigmata 5-6 mm. longa, antheras excedentia: capsula valde immatura, semina non visa. TYPE: M. Salter, a 2790, 6 miles east of Graafwater, Clan- william Div., Sept. 30, 1932 (K). This species is in all respects a member of the imbricata group, but is distinguished easily by its unequal stamens, and by the rather medium- sized, light, unmarked yellow flowers. It gives me much pleasure to name this distinct and attractive species for Dr. Louisa Bolus, who has been deeply interested for many years in South African [ridaccae, although possibly her work on Mesembrianthemum is more widely known in this country. 19. G. LouisaBoLusag, var. longifolia, var. A specie floribus paucioribus, folio basale uno et folio caulino inflorescentiam excedentibus, foliis omnibus perangustis, 1(-3)-nervatis, ad 35 ¢ longis, planta plerumque altiora, differt. Typ achmann, no. 1091 , on the road to Moorreesburg, in the neighborhood of Hopefield, Malmesbury Div., Sept., 1 SPECIMENS SEEN: South Africa, without locality, Pi ei f. (Gen) apparently belongs here. STUDIES IN THE IRIDACEAE—II 35 The details given in the diagnosis are enough to distinguish this plant from the species. Both are unlike other members of the group to which they belong. 20. G. rubicunda, spec. nov. Cormus ovoideus, 1 em. altus, 8 mm. latus; tunicae imbricatae, apice cuspidatae: folia basalia 2, super vaginam subpetioloideo-angustata, in lamina lineari-attenuata, 16-20 cm. longa, 3 mm. lata, glabra, subacuta, prominente 3—5-nervata, marginibus incrassatis, dilatata: folium caulinum 1, vagina subven- tricosa, multi-nervata, lam mina super vaginam adaxialiter incisa, in secunda, flexuosa, 2-5-fl.: spatharum valva exterior ad 1.4 cm. longa, oblongo-ovata, ‘sbmipke acuta, quam perianthii tubum longior, herbacea, apice rubro-purpurea; valva interior ad 1 cm. gmat ovarium ad 3 mm. longum, subturbinatum: perianthii tubus ad 5 mm. longus, summo ampliatus: tepala subaequalia, ad 2.2 cm. longa, 6-7 mm. lata, obovata, obtuso-rotundata et apice retusa, exteriora pallide lutea clara, nahin rubro-lineata, interiora non-maculata: antherae ad 7 mm gae; filamenta ad 1.1 cm. longa: stylus ad 2.3 cm. longus, HE We Bi (2), antheras aequans vel stnpaat bccn 4 mm. longa, percide: excedentia: capsula seminaque Schlechter, no. 8931, from Mitchels Poss, a 1500 ft., Sept. 10, 1896 B. ISOTYPE in Gen). Hasirat: damp places. This species is close to G. Louisabolusae, but can readily be dis- tinguished by the fact that its stamens are equal and the flower a deeper yellow, with reddish or purple stripes on the exterior of the outer tepals. From G. bicolor it is separated easily by the lack of the purple base and purple perianth-tube. Possibly MacOwan in Herb. Austr.-Afr., no. 1969, from Pakhuisberg, in Clanwilliam Div., is to be included in this species. 21. G. Wricutm Baker. Plant 30-40 cm. tall: free globose, 1 em. high and 1 cm. wide; the imbricate tunics dark n, hard, cus at the apex: basal leaves 2-3, to 29 cm. long, 1.5-3 mm. wide, exceeding the base of the inflorescence, narrowed above the sheath, linear, acute, glabrous, the midrib, 2—4 other ier eek: and the edges thickened; the single cauline leaf 17-18 cm. long, the basal sheath ventricose, many-nerved, adaxially indented a little above the sheath: stem simple or deeply branched, terete, glabrous; about 30 cm. long; the inflorescence a 6—11-flowe red lax, secund spike: outer spathe to 8 mm. long, ovate-lanceolate, herbaceous, membranous at apex, obtuse, longer than perianth-tube; the inner spathe to 6 mm.: ovary 2-3 mm. long, ellipsoid-turbinate: perianth-tube to 2 mm. long, lighter than the tepals: tepals subequal, to 1.3 cm. long, 5 mm. wide, 36 FOSTER the outer reddish, the inner white or light yellow, ovate-elliptic, obtuse: stamens 24 as long as the tepals; the anthers 4 mm. long; the filaments + 5 mm. long: style to 1.1 em. long, exceeding the anthers; the stigmas 3 mm. long, recurved: capsule immature, turbinate, to 8 mm. long; seeds not seen.—Baker in Journ. Bot. xiv. 238 (1876). Type: C. Wright, no. 243, from Simons Bay, Cape Div. (K; isoTyPE G) HasitaT: mountain slopes, rare in Cape Div., according to Bolus and Wolley Dod in Trans. 8. Afr. Phil. Soe. xiv. 331 (1903). SPECIMENS SEEN: South Africa, without locality, Bergius (B), in part; Drége, without number (Gen): Div.: foot of Table Mt., near Kirstenbosch, 400 ft. alt., Sept., 1879, H. Bolus, no. 4710 (K); by Orange Kloof swamp, Oct. 26, 1897, Wolley Dod, no. 3454 (K). Most of the material seen of this species is rather poorly preserved, with the result that it has been difficult to form a clear conception of it. Even the color of the flower is a little uncertain, although it is certainly not concolorous as Baker stated, Handbk. Irid. 154 (1892), thereby contradicting his own statement in the original description. The species clearly belongs in the imbricata group, but it is difficult to say more than that without actually studying it in the field. In the original description, Schlechter associated this species with G. ornithogaloides, but it appears to me to be much closer to the STUDIES IN THE IRIDACEAE—II 37 imbricata complex. It differs from G. ornithogaloides in its corm- tunics, in the shape of the corm, in its very ventricose cauline leaf- sheath, and in the shape of the spathes. The color was reported by Schlechter to be a brighter yellow than that of G. ornithogaloides. 23. G. SULPHUREA, var. arenicola, var. nov. G. sulphurea foliis basalibus latioribus (2-3 mm. lata), semper 3-nervatis, foliis omnibus plerumque brevioribus quam basis inflorescentiae (1— -fl.), floribus parum majoribus (tepala ad 1.4 cm. longa), pallide sulphureo- luteis, exterioribus externe rubro-lineatis differt. Type: Schlechter, no. 8992, Ceres Road, in arenosis, 800 ft. alt., Oct. 11-12, 1896 (B; 1sorypeEs in K, Gen). The differences between this variety and the species are sufficiently indicated in the diagnosis. Apparently, there is some difference in habitat. 24. G. mmpricaTa (DelaR.) Ker-Gawl. Plant 12-18 cm. high: corm ovoid, 1.5 em. high, 1.3 em. wide; the imbricate, tunics hard, slightly cusped at apex: basal leaves 2, to 23.5 em. long, 3.5 mm. wide, briefly subpetioloid above the sheath, the blade wider, linear, glabrous, acute, the edges, midrib and 2-6 other nerves prominent; the single cauline leaf to 15 cm. long, markedly striate, basally sheath- ing, but hardly ventricose, adaxially incised above the sheath: stem simple, glabrous, somewhat flexuose; the inflorescence a 3—6-flowered, lax, secund, flexuose spike, whose base usually exceeds the leaf-tips: outer spathe to 9 mm. long, longer than the perianth-tube, oblong- ovate, obtuse, mucronate, herbaceous but membranous at the apex; the inner spathe to 8 mm. long: ovary to 2.5 mm. long, subturbinate: perianth-tube to 2 mm. long, dark in color: tepals subequal, to 1.4 cm. long, 4 mm. wide, elliptic-obovate, obtuse, whitish yellow or white, with a broad red stripe on the exterior of the outer series, the inner completely white: stamens less than 74 as long as the tepals; the anthers 4 mm. long; the filaments 5 mm. long: style to 1 cm. long, about equalling the anthers; the stigmas 3 mm. long, exceeding the anthers: capsule and seeds not present.—Ker-Gawl. in Koenig and Sims, Ann. Bot. i. 224 (1804). Ixia imbricata DelaR., Deser. 17 (1766). G. obtusata Ker-Gawl. in Bot. Mag. xviii. t. 672 (1803), type-species of the genus. G. arenaria Ecklon, Top. Verz. 21 (1827), nomen nudum. G. sabulosa Klatt in Trans. S. Afr. Phil. Soc. iii (2). 203 (1885). G. imbricata, var. G. obtusata (Ker-Gawl.) Baker, Handbk. Irid. 157 (1892). Type: the location of the DelaRoche specimen, if indeed it still exists, is not known to me. Hasrrat: damp places, according to H. Bolus and Wolley Dod in Trans. S. Afr. Phil. Soc. xiv (3). 332 (1903). 38 FOSTER SpECIMENS SEEN: South Africa, without precise locality, Verreaux in me | aie gemeagt ee Sen Prior ae Leibold, me aN ’ 1882, MacOwan in Herb. Norm no. 2 , B, Gen in part); above Camp’s Bay, Aug. 29, 1897, Wolley pan no. 2767 (K), Van Kamp’s Bay, Sept. 10, 1881, MacOwan, no. Ae dae of G. sabulosa; no. eer : merset West Divs ‘near Sir Lowrys Pass, 25 met. Oct. 9, 1900, Diels, no. 1252 (B). The leaves and corm of this species show obvious relationship with G. Mathewsii and its variety eurystigma, but the shape and color of the flower, as well as the shape of the stigma, are utterly different. 25. G. IMBRICATA, var. CONCOLOR Baker. Differs from the species in often shorter leaves, which are much broader, with more numerous nerves, the stem often branched, the ne tube longer (to 4 mm.), peck flowers (tepals to 1.8 cm. long, 7 mm. wide), which are clear arked yellow, the style definitely shorter than the anthers, and the 4. 5 mm. stigmas exceeding them.—Baker in FI. Cap. vi. 72 (1896). Type: ‘ Bolus, no. 5567 (K), from Somerset West, below 100 ft. alt., 1880. SPE CIMENS SEEN: South Africa, without precise locality, Ecklon (B), Roxburgh (Gen), Verreaux oe ae entotsholland, near Somerset, Ecklon & Zeyher, Irid. no. 211 (K, B, Gen): Cape Div Devil’s Peak, Ludwig (B); He each Spice (B); Wynberg, rat C 3 (B 26. G. impricaTA, var. Brehmii (Klatt), comb. nov. Similar to var. concolor in color, but the basal leaves and cauline leaf-blade I-nerved, usually exceeding the a G. Brehmii Eckl. ex Klatt in Linn. xxxiv. 653-654 (1865- THE: Ecklon & Zeyher, Trid. no. oO (Bi in part; ISOTYPE in Gen), from Somerset, in Hottentotsholland. SPECIMENS SEEN: Cape Div.: Table Mt., 1887, Schinz, no. 92 (K). In his description, Klatt cited Ecklon & Zeyher, Irid. no. 211 in addition to no. 295, but no. 211 appears to me to be var. concolor, and no. 295 fits the description rather better. For this reason I have selected it as the type. STUDIES IN THE IRIDACEAE—II 39 27. sulphurascens Schitr. in herb., spec. nov. Cormus ignotus: folia basalia 2, 7-17 cm. longa, 0.5-1.5 mm. lata, acuta, glabra, super vaginam paullum angustata, in lamina dilatata, caulinum 1, 7.5-16 cm. longum, quam inflorescentiam parum brevius, longe ite paullum ventricosum, vagina gradatim in lamina lineari-attenuata, acuta, glabra, 1—3-nervata, transiente: caulis simplex vel Spactging glaber, teres, brunneo-purpureus, paullum geniculatus, ad 15-28 cm longus, inflorescentia spica laxa, - Ha secunda, 4-8-fl., ramus pauciflorus: spatharum valva exterior ad 1 cm. longa, oblongo-ovata, abrupte acuta vel obtusa, herbacea, apice purpurea, tubum excedens, valva interior ad 8 mm. longa, purpureo-membranacea: ovarium ad 3 mm. longum, ellipsoideo-turbinatum: perianthii tubus ad 3-4 mm. longus: tepala subaequalia, 1.2 cm. longa, 4 mm. lata, obovata, obtusa, alba vel pallide sulphurea: antherae ad 5 mm. longae, filamenta ad 7 mm. longa: stylus ad 1.3 cm. longus, antheras aequans, stigmata ad 2-3 mm longs, antheras exéodaate: capsula immatura turbinata, 4 mm. longa, semina non visa. Type: M alate no. 7658, on the aga west of Nieuwoudt- ville, in Calvinia Div., at 750 met. alt., Oct., 1916 (B). ABITAT: probably in rather dry, sandy ae At a first glance, this species appears to be a miniature version of G. Louisabolusae, but it can readily be distinguished by its smaller, whitish flowers, equal anthers, and its much smaller size. Section WerneEA Baker, emend. Baker in Journ. Linn. Soc. xvi. 95 (1877). Leaves 1-several-nerved, the midrib much more prominent than the others, the nerves more nearly simple ridges with the edges not revolute downward; stems or leaves, or both, often pubescent. Tyre-sPecies: G. ovata (Burm. f.) Aschers. & Graebn. In the original description of the section, Baker included two species, G. minima Baker and G. excisa (L. f.) Ker-Gawl. Since the former is a Hesperantha, the latter automatically becomes the type of the section, with the earlier specific epithet used, instead of excisa. This section, also, can be divided into two subsections, although the groups are in some respects, perhaps, less homogeneous than those of section Rochea. Subsection report pata: subsect. nov. Folia vel caules, vel utra- que, pubescentia; spatharum valva exterior in parte superiore valde brunneo-membranacea ; antherae saepe inaequales. Type-species: G. erosa apni R. C. Fos 28. G. MONANTHA (Thunb. ) Eckl. Corm ovoid-conic, 8 mm. high, 7 mm. — the tunics imbricate: basal leaves to 15-20 cm. long, 3-4 wide, subpetioloid above the sheath, mostly glabrous but some- nies very sparsely hispidulous at the edge near the base of the blade, 40 FOSTER somewhat lax, each branch 1-flowered: outer spathe to 1.5 em. long, ovate-lanceolate, acute, entire, hispidulous along the basal portions of several nerves, herbaceous, colored at the tip, becoming brown after the flower has withered ; the inner spathe to 1.1 em. long; both spathes exceeding the perianth-tube: ovary to 5 mm. long, subturbinate: perianth-tube 2.5 mm. long, tubular: tepals subequal, to 2 cm. long, subunguiculate, 1 em. wide, obovate-spatulate, subacute, dark blue- purple except for 6 mm. at the base above the — this being pellucid; the outer tepals somewhat longer and narrow anthers to 6 mm. long, the filaments to 1 cm. iaiies style 1.5 cm. ar the recurved stigmas ‘about equalling the anther apex: capsule and seeds not seen.—Ecklon, Top. Verz. 21 (1827). Ixia monanthos Thunb., FI. Cap. (ed. b 226 (1811), non DelaRoche. Geissorhiza Bellendeni MacOwan in Journ. Linn. Soc. xxv. 393 (1890). Type: sheet labelled Ixia monanthos in Herb. Thunb, (not seen). HaBiraT: in es damp sandy or grassy plac SPECIMENS SEEN: South Africa, without locality, Leibold (B): Malmesbury Div.: i eeecticat at about 300 ft. alt., Oct., 1878, H. Bolus, no. 4341 (K); Groenekloo f, + 500 ft. alt., Oct., 1887, 'M acOwan in Herb. Norm., no. 810 (K type of G. Be allendeni, G, B, Gen); near Yserfontein, Sept. 13, 1931, T. M. Salter, no. 1350 (K); between Darling and Yserfontein, Sept. 23, 1932, 7. M. Salter, no. 2709 (K). Ixia monanthos Thunb. is a later homonym of I. monanthos DelaR., Descr. 21 (1766), but was transferred to Geissorhiza by Ecklon in 1827, at which time, so far as I know, no synonym was available. Since G. Bellendent was not described until 1890, it appears legitimate to continue the use of the specific epithet monantha. On the sheet of H. Bolus, no. 4341 in Kew, N. E. Brown has noted: “These specimens of 4341 match the type specimen of Ixia monanthos Thunb. in Thunberg’s Herb.”” He also wrote on the sheet “G. monantha Sweet” and “G. Bellendeni MacOwan.’ The latter is correct, but the identification of this plant with G. monantha Sweet isnot. This was merely a new name given by Sweet to Ixia secunda sensu Houttuyn, non DelaRoche, a plant which is G. Rochensis. As pointed out in the discussion of that species, I have seen the plant in the Burmann Herbarium from which Houttuyn made his very exact drawing; it is certainly not G. monantha (Thunb.) Eckl., but is G. Rochensis. 29. G. TULBAGHENSIS Bol. f. Corm ovoid-subglobose, 7 mm. high; the ‘tabitente tunics hard, smooth, dark brown: basal leaves 2-3, STUDIES IN THE IRIDACEAE—II 4] PRS IE por Ce gg narrowed above the sheath, broadening a linear-falcate blade, acute, glabrous (the edges of the leaves aliphaly pilose in some cases, especially the narrow basal portion), the edges and midrib thick, to 15 cm. long and 3 mm. wide; the 1-2 cauline leaves hardly ventricose at the base, to 5 em. long, mostly free, many-nerved, the edges and keel of the sheath somewhat pilose: stem 1—2-branched, flexuose, terete, densely short-pilose, the lower branch usually occurring not far above the base, the main axis and the branches each 1-flowered: outer spathe to 1.2 cm. long, oblong- ovate, acute, herbaceous at the base, brown-membranous above, slightly pilose or ciliolate near the edges at the base, many-nerved, exceeding the ee _ inner spathe almost as long, mem- branous-herbaceous at the base, brown-membranous above, with = least 4-6 Latur tee es aire ovary subturbinate, 2-3 m long: perianth-tube 4 mm. long, dull purple: tepals subequal, & git 6 em. long, 8-6 mm. wide, obovate-spatulate, subacute, the base dull purple, the upper portion white: stamens about 34 as long as the outer tepals, unilateral; the anthers 5 mm. long; the filaments to 1 em. long, purple: style to 1. 7om. long, almost as long as the anthers; the stigmas to 2.5 mm. long, exceeding the anthers: immature capsule to 6 mm. long, turbinate, the seeds not seen.—Bolus f. in Ann. Bolus Beck. il. 160 (1918). Type: L. Bolus in Herb. Bolus, no. 14851 (in Herb. Bolus, not seen; ISOTYPE in K), near the village of Tulbagh, about 490 ft. alt., Oct., 1915 Hasirat: sandy places. SPECIMENS SEEN: South Africa, without locality, J. C. Breutel (B): Piquetberg Div.: near Piquetberg Road, 80 met. alt., Aug. 30, 1900, Diels, no. 1164 (B). Mr. Bolus, in the original description, differentiated this species from G. Bellendeni (= G. monantha) by its pubescent stem, but G monantha likewise has a pubescent stem. It is undoubtedly most closely related to G. monantha, but its color is completely distinctive, the cauline leaves are less ventricose, and its stamens are unilateral, thus approaching Tritonia, as Mr. Bolus pointed out. 30. G. secunpa (Berg.) Ker-Gawl. Corm ovoid-conic, 1.2 em. high, 9 mm. wide; the imbricate tunics hard: basal leaves 2, to 28 em. long, 4-6 mm. wide, attenuate below, linear, acute, gabrous, the edges, midrib and occasionally 2 other nerves prominent; the 3 cau- line leaves 17-2.5 em. long, the longer sheathing the stem for several em., the uppermost not sheathing, slightly hispidulous along the lower edges of the sheath and the basal portion of some nerves: stem many-branched, slender, terete, flexuose, shortly hispid, exceeding the leaves; the inflorescence a 3-6-flowered lax, secund spike: outer spathe 42 FOSTER to 1.2 em. long, oblong-ovate, herbaceous, the upper portion brown- membranous, the apex acute or somewhat truncate, the inner spathe to 8 mm. long: ovary 2-3 mm. long, ovoid-subturbinate: perianth- tube 1-2 mm. long, lighter than the tepals in color: tepals subequal, 1.2-1 em. long, 3.5-4 mm. wide, elliptic-obovate, subacute, blue- purple, the inner slightly emarginate: stamens 14 as long as the tepals; the anthers 3 mm. long, the filaments 4 mm. long: style 8 mm. long, equalling the anther apex; the stigmas 1.5 mm. long, recurved, ex- ceeding the anthers: immature capsule to 7 mm. long, the seeds globose-angular.—Ker-Gawler in Bot. Mag. xviii. t. 672 in textu (1803). Ixia secunda Berg., Pl. Cap. 6 (1767), non DelaRoche. Gladiolus junceus Burm f., Prodr. 2 (1768). Type: unknown; may no longer be in existence. HasitraT: common, widespread, in sandy places. SPECIMENS SEEN:* South Africa, without locality, Burmann f. (Gen, type of Gladiolus junceus Burm. f.): Clanwilliam Div.: Modder- Zeyher, Irid. no. 228 (G, B, Gen); Capetown, Prior (K), F. A. Rogers, no. 11318 (K); Tartville (?), near Capetown, 1879, Spielhaus (B); north base of Devil’s Peak, 35 met. alt., Oct. 3, 1900, Diels, no. 1139 no. 3705 in Brinkwater Ravine, Oct. 31, 1920, F. T. McClean, no. 65 (G); Table Mt., Ecklon, no. 310 (K, B, Gen); Kirstenbosch, Oct. 19, 1931, W. F. Barker, no. 23 (K); Red Hill, Mrs. Hugh Jameson (K); Signal Hill, Oct. 3, 1883, Wilms, no. 3739 (B); Rosebank, near Capetown, below * With one see exception, no material from unknown locality has been cited, although much was n. STUDIES IN THE IRIDACEAE—II 43 Sept., 1879, H. Bolus, no. 4804 (K); Camp Ground, near nent # below 100 ft. alt., Sept., 1880, H. are no. — G, A; Simons Bay, C. Wright (G); Hout t Bay, Sept. 2, 1846, r (K); Caapsche Vlakte, Oct 14, 1815, Bergius (B); Bieciee die & Maire (B); Constantia, Sept., 1827, Verreaux (Gen): Paarl Div.: be- tween Paarl and Pont, Drége, no. 8483c (B, K, Gen), no. 8483b (B); Paarlberg, 1000-2000 ft., Drége, no. 8483a (B, Gen): Swellendam, Sept., 1827, Verreaux (Gen): Stellenbosch Div.: near Somerse Hottentotsholland, Ecklon & Zeyher, no. 226 (K, B, Gen). Almost without exception, as was pointed out in the discussion of G. Mathewsii, this species has been attributed to (DelaR.) Ker-Gawl., but actually the basinym is Ixia secunda Berg., not I. secunda DelaR. The only synonym, so far as is known to me, is Gladiolus junceus Burm. f., which cannot be transferred to Geissorhiza without creating a later homonym in that genus. Ixia pusilla Andr., given by Ker- Gawler as a synonym of this species, is actually distinct. ” Conse- quently, it seems proper to retain this old and well-established specific name, despite the fact that its basinym is a later homonym In the large number of specimens studied, a very considerable range of variation was found in the size and shape of the leaf, in the size and color of the flower, as well as in the number of flowers in an inflorescence, in the size of the plant, and in the degree of branching of the inflorescence. No one character or combination of characters remained sufficiently constant to warrant recognition as a variety. From the very dwarf, few-flowered, narrow-leaved group to the very large, many-flowered, broad-leaved group, it is possible to establish an almost perfect series of intergrades. Since most of the specimens seemed to be in the middle of the range of variation, a group of which MacOwan in Herb. Norm., no. 807 is a good sample, that has been chosen as typical and the description based primarily upon it. The glabrous leaves and densely short-villous stem serve to distinguish this species easily from its immediate relatives which have glabrous stems. In his key to the genus, in FI. Cap. vi. 66 (1896), Baker incorrectly states that the stem is glabrous. It is possible that some of the more extreme departures from the norm are the result of hybridization, but it seems more likely that much is due to difference in habitat, or as a response to temporary climatic changes. A thorough study of the species in the field would do much to answer questions here left unanswered. 31. G. ixioides Schltr. in herb., spec. nov. Cormus ovoideus, 1 ¢ altus, 8 mm. latus, tunicae imbricatae, apice cuspidatae: folia haeiiia 2, ad 20 cm, longa, 4 mm, lata, superius quam inferius longius, super 44 FOSTER — vicar gn lamina lineari-attenuata, acuta, s primarius marginesque incrassati, sparse pilosa, pilis appressis; foliam caulinum 1, ad 12 em. longum, pluri-nervatum, vaginae mar- gines et nervus primarius sparse F cpases biapidull: caulis simplex vel l-ramosus, persparse pilosus, ad 14-20 cm. altus; inflorescentia = laxa, disticha, flexuosa, 2—4-f1.: soebam valva exterior ad 1.5 c¢ subglobosum: perianthii tubus 0.5-1 mm. longus: tepala subaequalia, 1.2-1 cm. longa, 5-4 mm. lata, obovata, apice retusa, pallide Syiaad purpurea (in siccis): antherae ad 5 mm. longae; filamenta ad 3 m longa: stylus et stigmata 6 mm. neg sherae valde breviora: a sula beers ad 6 mm. longa, semina 2 mm. longa, angulato-globosa. PE: tn no. 10866, from Agtertuin, northwestern Cape Puinde 900 met. alt., Aug. 15, 1897 (B; 1sorypEs in K, Gen). SPECIMENS SEEN: r South Africa, without locality, Prior, in 1847 (K). In its short style and open flower, this species has a real resemblance to Iria. Nevertheless, the spathes and corm are those of Geissorhiza. The isotype at Kew appears to have light red-purple flowers with a darker purple blotch at the base of the tepals, but this may be an artificial result of drying. G. HETEROSTYLA L. Bol. Corm not known to me: basal leaves 2, me a third reduced to a brown-membranous, brown- and red-striate sheath, to 36 cm. long, 2 mm. wide, narrowed but hardly subpetioloid Typ . Bolus is in ‘Herb. "Bele no. 19148 (in Bolus Herb.; not seen), near Whitehill | in Laingsburg Div. , Aug., 1929. SPECIMENS SEEN: Laingsburg Divs : Tweedside, Sept. 25, 1932, G. J. Lewis and W. Barker, cult. in Nat. Bot. Gard., no. 2694/32 (K). STUDIES IN THE IRIDACEAE—II 45 Although it has not been possible for me to see the type, the single specimen seen was sent to Kew by the Bolus Herbarium, and it may therefore be presumed to be authentic material. According to the original description, the style is unilateral, and highly variable in length, being shorter than, equal to, or longer than the stamens. The stamens themselves appear to be unequal, this being due to different filament lengths. In general, the species resembles a large G. secunda. 33. G. Lewisae, spec. nov. Cormus ignotus: folia basalia 2-3, 2-3 em. basi vaginantia, ad 19 cm. longa, quam inflorescentiam prima- riam breviora, super vaginam angustata, in lamina lineare, acuta, 1.5— 2.5 mm. lata, dilatata, nervo primario et marginibus incurvatis folio- rum renin villosis: folia caulina ramos subtentia, vagina basale non ventricosa 3-1.5 ¢ onga, multinervata, nervis breviter pubescenti- bus, lamina 10-3 ct cm. longa: caulis 1—3-ramosus, paullum angulatus (?), dense puberulus, ad 25 cm. longus; inflorescentiae axis primarius spica laxa, 3-4-fl., secunda, flexuosa, ramis 1—3-fl.: spatharum valva exterior ad 1.5 cm. longa, perianthii tubum valde excedens, oblongo- ovata, obtusa, mucronata vel abrupte acuta, vel apice trilobulata, herbacea, pars superior pallide brunneo-membranacea; valva interior m. , mm. lata, obovata, obtusa, atrocaeruleo-purpurea, basi ochracea: antherae ad 6 mm. longae; filamenta subinaequalia, basi pallide lutea, supra pallide purpurea: stylus ad 1.1 em, longus, apicem antherarum vix aequans; stigmata ad 3 mm. longa, antheras excedentia, purpurea: bo chee seminaque non visa. Type: G. Lewis in Herb. Bolus, no. 20303 (IX), from Langebaan in Sialesines Div., Sept. 7, 1932 (distributed as G. furva). In this species, the difference in filament lengths is not so pro- nounced as in some of its relatives. The spathes are like those of G. Rogersii N. E. Br., but the brilliant, deep blue-purple flowers and the ne itself are more like those of the Rochensis group. . G. InAEQuALIS L. Bol. Plant 30-40 cm. tall: corm ovoid-conic, bran ched, at sGanencctine of the main axis ited t 8-flowered, the spike lax, distichous, secund: outer spathe to 2.3 cm. long, 15-nerved, the nerves hispidulous, obovate, acute, herbaceous, with the apex purple-membranous, much longer than the tube, the inner spathe to 46 FOSTER 1.7 em. long, 4-nerved: ovary 3.5 mm. long, subturbinate: outer tepals to 2.6 cm. long, inner to 2.2 ecm. long, 8 mm. wide, oblong-ovate, obtuse, emarginate, yellow-white at the base, the remainder blue- violet: stamens a the anthers to 8 mm. long, the filaments unequal, 2 being 1 cm. long, the third, 5-6 mm. long, the longest stamen about half te length of the outer tepals: style to 1.4 cm. long, not quite equalling the anthers, unilateral, stigmas 3-5 m exceeding the anthers: immature ‘capsule 5 mm. long, shleiigvellipeetd: the viata nt seen.—L. Bolus in S. Afr. Gard. xx. 346 (1930), and Fig. B on p. TYPE: t. eas in Herb. Bolus, no. 19256 (in Bolus Herb., not seen), from Calvinia Div., near Nieuwoudtville, Sept., 1 SPECIMENS SEEN: Calvinia Div.: near Nieuwoudtville, Sept., 1931, H. Buhr, cult in Nat. Bot. Gard., no. 376/31 (topotype, K) As was true of G. heterostyla, the type has not been seen, but in this instance, too, the single specimen seen was sent out from the Bolus Herbarium so that it may well be regarded as authentic, especially since it iv apparently a topotype. A note on the label indicates that this cultivated specimen is stouter than the type. Superficially, this species resembles G. namaquensis Barker, but it has a very short tube, a — stem, and unequal stamens; the corm, too, is distinctive. . G. Rogersn N. E. Br. Corm not present: basal leaves 2, 5- Hee em. long, 1.5-4 mm. wide, narrowed above the sheath and enlarging into a ‘aes: acute blade, the midrib and margins thick, with ap- pressed hispid hairs; the 1-2 cauline leaves to 10.5 cm. long, sheath- ing for about half their length: stem terete, _puberulent, foi 9 1—2-branched, the inflorescence of the main axis a 3—5-flowered, lax, secund spike, the branches 1-3-flowered: outer spathe to i em. long g, apparently lengthening in fruit, ov i rie subtrilobulate or abruptly acute, herbaceous at the base, reddish-brown-membranous above; the inner a trifle shorter: ovary + 2 mm. long, subturbinate: perianth-tube to 1.5 mm. long: tepals subequal, to 1.5 cm. long, 3mm. wide, elliptic, obtuse, light reddish-lilac; the inner tepals probably white: pa unequal, about half as long as the tepals; the anthers 4 mm. long; 2 filaments to 4 mm. long, the third to 2 mm. long: style to 1.3 em. Ga the stigmas 2 mm. long: capsule to 6 mm. long, turbi- nate; the immature seeds with a purple funicle—N. E. Brown in Kew ae 1931: 452 (1931). Typ . A. fogers, no. 16275, from Worcester Div., between Os- plaats pe Tunnel Sidings, 2000-3000 ft. alt., Aug., 19 915 (K). This species is close to G. pusilla (Andr.) Klatt, but differs as follows: it is more openly branched, the stem is pubescent, the hairs on the leaves are hispid, appressed, the spathes are longer, and the stamens are unequal. From G. secunda it can be distinguished by its STUDIES IN THE IRIDACEAE—II 47 hispid leaves, the unequal stamens, the very short tube, and the color of the flowers. 36. G. Leipoldtii, spec. nov. Cormus ovoideus, 9 mm. altus, 7 mm. latus, tunicae imbricatae: folia basalia 3, basale infimum pallide brunneo-membranaceum cataphyllum; basalia altera ad 25 cm. longa, super vaginam adaxialiter incisa, lamina linearis, ad 7 mm. lata, at- . tenuata, acuta, marginibus dense pubescentibus (pilis non adpressis), nervo primario prominente et sparse hirtulo (pilis ad laminam ad- pressis); folium caulinum ad 10 cm. longum, vagina ad 6.5 cm. longa, non ventricosa, nervis 2-3 prominentibus, hirtulis, pilis plerumque patentibus: caulis simplex vel in vagina folii caulini l-ramosus, teres, dense breviterque puberulus, ad 35 cm. altus; ages yout CBee is spica laxa, subflexuosa, secunda vel subsecunda, 3-fl., ramus 1-3-f1.: spatharum valva exterior ad 1.9 cm. longa, ese oonianthis tubum valde longior, ovata, trilobulata, lobulo medio longo et acuto, herbacea sed pars superior pallide brunnea; valva interior pallide roseo-brunnea: ovarium ad 4 mm. longum, turbinatum: perianthii tubus + 1-1.5 m. longus, a aan colore in basim tepalum diffuso: ae. subae- auntie, ad 2.6— m. longa, 7 mm. lata, obovata, obtusa, pallide longus, antheras aequans vel excedens; a ad 4 mm. longa, antheras excedentia: capsula seminaque non v Type: Letpoldt in Herb. Bolus, no. 20778 (K), from the Pakhuis- i ag Hasirat: “in abies wale according to Schlechte SPECIMENS SEEN: Clanwilliam Div.: Pakhuihars. 2900 ft., Aug. 23, 1896, Schlechter, no. 8621 (K, B, Gen). This rather attractively colored species is clearly most closely related to G. Rogersii. In general, it is a larger plant, less branched (even in large specimens), with the leaf pubescence spreading on the edges, not inwardly appressed. The flower is larger, different in shape and color. The Schlechter specimen is smaller than the type, and the flower has a little more blue in it; it undoubtedly comes from a higher altitude. One rather striking aspect of the stamens is that one anther is definitely longer than the other two, and the filament of this stamen is shorter than the other two filaments. G. susricipa L. Bol. Corm not seen: basal leaves 2, 15-19 em. aie 8 mm. wide, shortly sheathing, widening above the sheath into a linear-elliptic, 7-S-nerved blade, the nerves and edges densely short-pilose, the subobtuse apex brown; the cauline leaf 13.5-15.5 em. long, sheathing for half its length, ventricose, nerves pilose: stem once-branched, each of these again branched, the branches very 48 FOSTER strongly divergent, terete, es subrigid, flexuose, brownish; the inflorescence a lax, secund, 4-flowered, occasionally 3- flowered, spike: outer spathe to 9 mm. long, oie g-ovate, emarginate or subtrifid, red-purple to the middle, the remainder bright brown, membranous; the inner spathe to 7 mm. long, bro cates» Aine the nerves red- purple: ovary + 2 mm. long, turbinate: perianth-tube + 3 mm. long, yellowish: outer tepals to 1.7 cm. long, inner to 1.4 cm. long, 56 mm. wide, elliptic-ovate, acute, bright, dark, blue-purple: stamens a little more than half as long as the tepals; the anthers 6 mm. long, the purple filaments 4 mm. long: style 9 mm. long, not quite equalling the ane the stigmas about equalling them: immature capsule turbi- + 5 mm. long, the seeds not seen.—L. Bolus in S. Afr. Gard. Xxii. 270-77 el 932). Tyre: Buhr, in Herb. Bol., no. 19943 (in Bolus Herb., not seen), from Calvinia Div., near Nieuwoudtville. SPECIMENS SEEN: Calvinia Div.; Nieuwoudtville, Sept., 1930, H. Buhr, in Herb. Bolus, no. 19995 (K), These two specimens were collected from cultivated plants in the aig of Dr. Louisa Bolus, in September, 1933, and September, 1 This is a very striking species, both in form and color, and not likely to be confused with any other. The broadly linear, obtuse, pilose leaves, the dichotomous, divergent branching, and the bright reddish-brown spathes cause it to stand out with great clarity. The two specimens seen, both marked no. 19995 in Herb. Bolus., are labelled “Type,” but the original description gives the type number as Bolus Herb. no. 19943. 38. G. rosea (Klatt), comb. nov. Plant 15-40 em. tall: corm ovoid-subglobose, 1.2 cm. high, 1 em. wide; the imbricate tunics hard, ull gray-brown, apically cusped: basal leaves 2, to 22 em. long, 2-5 mm. wide, linear, acute, the upper with a longer sheath than the lower, the edges and midrib thick, edged with rather stiff hispidulous hairs appressed against the blade; cauline leaf 2-16 cm. long, sheathing for inflorescence: stem 1-branched above the middle, terete, glabrous, flexuose; the inflorescence a 3 lowered, lax, secund or subsecund spike, the branch 1-flowered: outer spathe to 1.3 cm. long, much ex- ceeding the tube, oblong or ov ate-lanceolate, the truncate apex some- what —— incised, herbaceous in the lower portion, brown- membrano ove; inner spathe nearly as long; both spathes brown- striate: poate er 2-3 mm. long, turbinate: er very short, mm. or less: tepals subequal, to 1.4~1.2 em. long, 4 mm. wide, elliptic-ovate or -obovate, apically subobtuse or ‘lightly core white (if so, the outer tepals slightly flushed with pink), pink, or reddish- STUDIES IN THE IRIDACEAE—II 49 purple, usually concolorous: stamens unequal, about half as long as the tepals; the anthers 3-5 mm. lon g; the filaments unequal, 3-7 mm. long, but one always shorter than the other two: style as long as the anthers, to 8-9 mm. long, the stigmas 3 mm. long, exceeding the ane capsule and seeds not seen.—H. esperantha rosea Klatt, Ergiinz. 1 (1882). Geissorhiza graminifolia Baker, Handbk. ne 155 (1892). Hieperakthe ciliata E. Mey. ex Klatt, Ergin 60 (1882). Tyrer: Miss Nigdepats, Pete Knysna (B). Hasitat: stony ground and open sacs banks, from Swellendam and Worcester Divs. east to Port Elizabeth Div SPECIMENS SEEN: Worcester Div. (?): Hexrivierskloof, 1000-2000 ft. alt., in Dec., Drége, no. 525 in part (B; type o of Hesperantha ciliata Klatt): Swalctelnas Div.: Swellendam, July, 1900, Harold Fry (K): eit Div.: near George, Oct. 6, 1814, Burchell, no. 6088 (K): Knysna Div.: near Keurboom’s River, 100 ft. alt., Sept. 17, 1897, Galpin, no. 4681 (K): Riversdale Div.: Riversdale, 1891-93, C. Rust, no. 13 (B), no. 555 (B): Humansdorp Div.: Oudebosch Flats, 800 ft. alt., Oct., 1927, Fourcade, no. 3364 (KK); near Humansdorp, 500 ft. alt., Sept., 1888, W. Tyson, no. 3067 (K): Port Elizabeth Div.: Baaken’s cyt near Port Elizabeth, 150 ft. alt., Aug. 15, 1931, F. R. Long, no. 441 (KK); Patons Farm, on Baaken’s River, Aug. 22, 1931, Pods. es no. 470 (K). The original description of G. graminifolia cited no specimens, so that Burchell, no. 6088 (K), marked as being the type of that species cannot actually be so. This specimen and Drége, no. 3496 (K) were cited as G. graminifolia by Baker in FI. Cap. vi. 70 (1896). The two specimens are not conspecific, and I am unable to separate the Burchell specimen from the plant described by Klatt as Hesperantha rosea. The type of Klatt’s species is rather narrow-tepalled and is more delicate than other specimens seen, but there are no real points of difference between it and them. Consequently, the transfer has been made to Geissorhiza. Although Ecklon, Top. Verz. 20 (1827), included a G. rosea in his list of species, the name was a nomen nudum; it has never been taken up and validated, and therefore cannot bar the transfer of Hesperantha rosea. Hesperantha ciliata Klatt, the type of which is Drége, no. 525 (B), was based on a mixture. Of the three specimens on the type-sheet, the two on the left have sparsely pubescent leaves, thus having the character for which the species was named, and are actually Geisso- rhiza, conspecific with the type of G. rosea (Klatt) Foster. The third specimen is a Hesperantha, allied to if not identical with H. radiata (Jacq.) Ker-Gawl. Since it is thus a mixture, it has seemed to me preferable to retain the specific epithet rosca and transfer that name to Geissorhiza. 50 FOSTER The species is well-characterized by its unequal stamens, concol- orous flowers, rather narrow tepals, and the rather short, stiff hairs appressed against the leaf-blade. Only the midrib is prominent. In his key, Fl. Cap. vi. 66 (1896), Baker places his G. graminifolia under the heading “Stem and leaves hairy.’’ Actually, the stems are glabrous. 39. G. pusmua (Andr.) Klatt. Corm ovoid-globose, to 8 mm. high and 8 mm. wide; the tunics concentric, apically cusped, smooth, light chalky or clay-brown: basal leaves 3, the lowermost reduced to a brown sheath, the whee normally 9-14 cm. long (rarely much a, and 2-3 mm. wide, linear-ensiform, adaxially indented above the sheath, acute, the midrib and edges thick and densely pilose, 2 other nerves sometimes subprominent and slightly pilose; the lower cauline leaf halfeheathitg, subventricose, 6-12 cm. long, many- nerved, the nerves pilose; the A 9) cauline leaf half-sheathing, 1-1.7 cm. long, sometimes with a short, 1-flo branch in its axil: stem simple ts : |-branched, terete, flabrows; 58 pe reoneie a 1-4-flowered secund spike: outer spathe to 6 mm. long, oblong-ovate, herbaceous at the _ the eat beight red-brown, membranous, the apex somewhat crenate or even tridentate; the inner spathe nearly as long, pant faces at the base, the rest hyaline, aly mage ovary to 3 mm. long, psig perianth-tube to 1.5 mm. long: tepals sara) or subequal, to 9 mm. long and 3 mm. wide, ovate- elliptic, somewhat gactaee acute, blue-purple, che | inner lighter: stamens little more than half as long as the tepals; the anthers 3 mm. long, the filaments 2.5 mm. long: style to 6 mm. long, about equal to the anther; the stigmas recurved, exceeding the anthers: immature capsule to 5 mm. long, turbinate, the seeds ae nae Repo Ergiinz. 58 (1882). Ixia pusilla Andr., Bot. Rep. 5 (1 802). Geis- sorhiza secunda var. G. pusilla (Andr.) Fel " Handbke Irid. 155 (1892), nomen illegitimum. G. pubescens Wolley Dod in Journ. Bot. Xxxvill. 171 (1900). Tyre: Andrews’s plate Hasirart: locally Sastauk on hill slopes around Capetown, among shrubs or rocks SPECIMENS SEEN: without precise locality, Bergius (B), Lalande (B), Hb. Paes, no. 1204 (B), Mund (K), Mund & Maire (B), Verreaux va. Cape Div : Devil’s — 250 ft. alt., oe 1882, MacOwan 1602 (K, type of G. pie 5 scens Welle, Dod; B); Lion Mt., 400 ft. alt.., Sept. 1, 1892, Schlechter, no. 1360 (B); Table Mt. , Zeyher (K); Cape- town, in 1877, Spielhaus (B). STUDIES IN THE IRIDACEAE—II ae | For this plant, frequently and unnecessarily confused with G. secunda, I have taken up Klatt’s name, based on Ixia pusilla Andr. The description given by Andrews was, as usual, somewhat inade- quate, and his plate is not clear on certain points, but from the text and figure enough can be learned to make it impossible to subordinate the plant to G. seeunda. The corm was described as having a “ hard, smooth skin,” is colored a light brown, and shows no sign of imbricate tunics. In this respect, it is in marked contrast to the corm of G. secunda, which is dark brown, with imbricate tunics, these being regularly and smoothly notched at the base. G. secunda has a villous stem and glabrous leaves, while no mention of either is made by Andrews. Nevertheless, G. pubescens Wolley Dod has a corm like that of Andrews’s plant, has blue flowers very similar to it, and in general appearance seems identical with [xia pusilla. Except for G. pubescens, there is no plant of similar appearance and identical corm, with which J. pusilla could be identified. It might be noted that the original description of G. pubescens describes it as having a subgla- brous stem, but an examination of the type and an isotype shows that the stem is completely glabrous. 40. G. Nn a Barker. Plant about 26-27 cm. high: corm globose-ovoid, 1.4 igh, 1 cm. wide; the tunics concentric, kag shiny, dark an orn cusped: basal leaves 2, to 16 cm. long, 3 m. wide, basally long-sheathing, not especially indented se the eek linear, acute, apically somewhat incurved, the five nerves edged with appressed hairs; the cauline leaf to ie em. long, long- sheathing but mostly free, not ventricose, pilose: stem glabrous, terete, 1-branched from below the middle; the inflorescence a 6- flowered lax, distichous, flexuose spike, the branch 2—3-flowered: outer spathe to 1.6 cm. long, ovate-lanceolate, abruptly acute, saber the apex membranous reddish-brown; the inner spathe cm. long: ovary to 3 mm. long, subturbinate: oan ee to On mm. long, ampliate at the top, greenish: tepals subequal, 2-1.8 cm. long, 4-6 mm. wide, the outer linear-elliptic, subunguiculate, the inner elliptic- obovate, both emarginate at the apex, dark blue-purple, yellow at the base: stamens a little over half as long as the tepals; the anthers 6 mm. long, the filaments to 9 mm. long: style unilateral, 1.2 cm. long, equalling the filaments of the stamens; the short stigmas recurved: capsule (slightly immature) globose-turbinate, 5 mm. long, the seeds not seen.—Barker in Pole Evans, Fl. Pl. S. Afr. xviii. pl. 688 (1938). Tyee: Phillips in Herb. Bolus, no. 22163 (in Bolus Herb., not pis SPECIMENS SEEN: Little Namaqualand, Klipfontein, Joly, 11 Mrs. E. Phillips, cult. in Nat. Bot. Gard., no. 1471/30 (K); Kiunkopt in the Richtersveld, August, 1925, Marloth, no. 6782 (B). ps FOSTER Although the type has not been seen, it appears probable that the Kew specimen was prepared from the living plants sent in to the National Botanic Gardens by Mrs. Phillips in 1930, as was the actual type. This distinct species is well-separated from its relatives by its long perianth-tube, virtually herbaceous spathes, and by its unilateral style. The figure shows the flower to be open-spreading, with the tips of the tepals incurved. The leaf-tips, likewise, are slightly incurved, almost uncinate. he G. erosa (Salisb.), comb. nov. Corm ovoid-subglobose, 1.3 igh, 1 em. wide; the tunics imbricate, dull pagal short- cape at hes apex: basal leaves 2, 12-23 em. long, to m. wide, subpetioloidly narrowed above the sheath, on are faa a rather broadly linear blade, the edges, midrib, and 2 other nerves thickened and pilose, the inner ribs with the hairs appressed against the blade; the cauline leaf to 14 cm. long, the sheath ventricose, many-ribbed, all ribs pilose: stem simple, to 34 em. tall, terete, glabrous; the in- florescence a 3-7-flowered, lax, secund, flexuose spike: outer spathe to 2.3 em. long, oblong-ovate, slightly trilobulate or entire and obtuse, basally herbaceous and slightly hispidulous on a few nerves (usually glabrous on most flowers), brown-membranous above, much pueene the perianth-tube; the inner spathe nearly as long: ovary to 4 m pple sg ra 8 perianth-tube 2-4 mm. long: tepals subequal, te ong, 1 cm. wide, obovate, obtusely rounded at the apex, ie ia pte stamens 2% as long as the tepals; the anthers to 7 mm. long, the filaments to 7 mm. long: style to 1.4 em. long; stigmas 3 mm. long, exceeding the anthers: capsule and seeds not seen.— Tria erosa Salisb., Prodr. 36 (1796). yPE: unknown to me; Salisbury, cult. in Britain, 1816 (Gen), and sent to page we Salisbury, may be taken as the standard ABITAT: dam $s, more northerly than the commoner variety. SPECIMENS cae ses th Africa without precise locality, Prior (KX): near the tone at Tulbagh, Sept., 1885, MacOwan, no. 2678 (K); =, mira ares 500 ft. alt., Sept., 1886, MacOwan in Herb. Norm., 90 (G, K, Gen, B) The plant here treated as G. erosa is conspecific with, but varietally distinct from, the plant ordinarily known as G. hirta. It is unfor- tunate that this well-known plant must be eaamied, and since I. erosa, is the earliest available synonym, it has been transferred to Geissorhiza. When Thunberg described I. hirta, he seemed to regard it as a new species, but in his list of synonyms, he gave I. inflexa DelaR. as a synonym, this name having been validly published in 1766. Consequently, by Article 60, part 1, of the current Rules of STUDIES IN THE IRIDACEAE—II 53 Nomenclature, J. hirta is invalid because it was superfluous when published. That J. hirta and I. inflera are not actually conspecific, that, indeed, they do not even belong in the same genus is immaterial. Thunberg so circumscribed his J. hirta as to include the earlier J. inflexa; hence, his name is invalid. 42. G. BROSA, var. kermesina (Klatt), comb. nov. Differs from the species in being generally siege the stem simple or rarely 1- branched; the inflorescence a 2-3 (-4)-fl owered spike: the flowers deep blue- or red-purple, the tepals to 1 yp cm. long, 8 mm. wide: the anthers to 7 mm. long; the filaments 3 mm. long: the style to 7 mm. long; the stigmas sometimes to 4 mm. long, exceeding the anthers.— Hesperantha kermesina Klatt, Ergiinz. 61 (1882). Ixia hirta Thunb., Diss. Ixia 9 (1783). G. hirta (Thunb.) Ker-Gawl. in Koenig and Sims, ing Bot. i. 224 (1804). E: Drége, no. 8480, except the two “prog marked a (B; ISOTYPES in K, Gen), from between Paarl and Pont. Hapirtat: sandy or stony places, sometimes dam See Reis: South Africa without locality, Burmann f. (Gen W. M. Rogers (K): Ceres Road, at 700 ft. alt., Sept. 12, 1896, Bsa ter, no. 8975 (B); Gordon’s Bay, False Bay, 50 ft. alt., Sept., 1902, H. Bolus, no. 9939 (K); near the Bergrivier and in Zwartland, Ecklon & Zeyher, Irid. no. 215 (B, Gen). This plant, which I cannot separate specifically from G. erosa, has been given, in varietal status, the name which Klatt gave to it in 1882, when he was under the misapprehension that it belonged in Hesper- antha. The type of H. kermesina proves, on examination, to consist of material of two species. The two large specimens (which I have marked a) in a more or less central position on the sheet appear to be G. quinquangularis, a species closely related to G. erosa. A study of Klatt’s original description will show that virtually all the measure- ments and details were drawn from the other specimens on the sheet, probably because the two large specimens lacked good flowers. With this limitation, then, as to the type material, Klatt’s species is clear, and is unquestionably the same as that heretofore known as G. hirta. For that reason, I have transferred his name, altering the category, despite the mixture of plants on the type sheet. The Rogers specimen (K), cited above, has been annotated as follows by N. E. Brown: “ Matches the type of Ixia hirta Thunb. in Thunberg’s Herbarium.” 3. G. qurnquaneutLaRis Eckl. ex Klatt. Plant 20-30 cm. tall: corm ovoid, 1.2 em. high, 1 em. wide; the tunics imbricate, apically cusped: basal leaves 2, to 16 cm. long, 2-5 mm. wide, narrowed above the sheath, widening into the linear, recurved, acute blade, the edges, midrib, and 2-5 other nerves thick and edged with appresssed hairs; 54 FOSTER the single cauline leaf 5.5-11 cm. long, the sheath ventricose, with many pilose ribs: stem simple or 1-branched from near the base, glabrous, ridged; the inflorescence a 3-5-flowered, lax, secund spike, the branch 2-3-flowered: outer spathe to 1.3 cm. long, longer than the perianth-tube, oblong to ovate, herbaceous at the base, red-brown- membranous above; the inner spathe nearly as long: ovary to 3 mm long, subturbinate: perianth-tube not over 3 mm. long: tepals sub- equal, to 2.3 cm. long, 7-8 mm. wide, elliptic-ovate, apically blunt, the outer tepals externally red-striate, the inner tepals white, or creamy-yellow: stamens less than half as long as the tepals; the anthers 5 mm. long, the filaments + 5 mm. long: style to 1 em. long, equalling the anthers; the stigmas to 5 mm. long, much exceeding the anthers: capsule to 8 mm. long, eno obovoid; the seeds immature.— Klatt in Linn. xxxiv. 654 (1865-66). G. hirta, var. G. inched (Eckl. YI Baker, Handbk. Irid. 156 (1892) at least as to name; the c bination invalid and based upon a nomen nudum. Hocptialb quinquangularis Eckl., Top. Vers 23 (1827), nomen nudum. Geis- sorhiza graminifolia var. bicolor Baker, Handbk. Irid. 155 (1892). Hesperantha quinquangularis (Eckl.) E Eckl. ex Klatt in Durand and Schinz, Conspect. Fl. Afr. v. 176 (1893). Type: Ecklon & eda cies no, 214 (B; 1soTyPE in Gen), from the Zwarteberg, in Caledon Hasitat: rather ate in damp sands, or grassy or rocky places. PECIMENS SEEN: South Afri rica, without ae locality, Burmann f. ik . cklon (B), Ludwig (B), Nine iven hee pe Div.: near Capetown, 9, H. 805 r(K): Lion Mt., Ecklon, no. ae (K, Gen, B); ty Bay, + Cet alt., Sept., 1889, 9, MacOwan in Herb. Norm., no. 261 (G, B, Gen): Caledon Div.: shave vlei, 1 mile from turnout to Villiersdorp from Caledon, Sept. 14, 1931, W. F. Barker, no. 6 (K); Zwarteberg, 1000-2000 ft. alt., in Aug., Ec tion & Zeyher (G, B); Zwarteberg, 3200 ft. alt., Oct. 17, 1894, Schlechter, no. 5561 (B); Caledon, Ecklon (KK): Paarl Div: between Paarl and Pont, Drége, no. 8480 in part (Gen, B): Swellendam Div.: on hills by the Breede River, near Swellendam, Drége, no. 3496 (K, Gen, B). Like a few others in the genus, this species has a rather confused history. Apparently the first name given to it was Hesperantha quin- quangularis Eckl., which was a nomen nudum. The specimens cited by Ecklon came from the same locality as his no. 312, which he labelled and distributed as G. guinquangularis Eckl., likewise a nomen nudum. The first valid publication came when Klatt described the species in Geissorhiza, in 1865-66. Of the specimens there cited, which actually belong to the species, Ecklon & Zeyher, no. 214 seems most characteristic, and therefore I have chosen it as the type. In 1892, Baker, Handbk. Irid. 156, attempted to reduce it to varietal STUDIES IN THE IRIDACEAE—II 55 status under G. hirta, but as usual he did so illegitimately (G. hirta var. G. quinquangularis Eckl.), and, moreover, based his attempt upon the nomen nudum, G. quinquangularis Eckl. This attempt was not improved by the fact that as synonym he cited Hesperantha ciliata Klatt, a very different plant, the species treated in this study as G. rosea (Klatt) Foster. His brief description and the synonym cited make it quite probable that he had this other species in mind, but in Fl. Cap. vi. 71 (1896), he cited two specimens for the variety, one of which I have seen. It is Ecklon, no. 312, which is certainly not G. rosea. In Handbk. Irid. 155 (1892), Baker made no improvement by describing G. graminifolia var. licolor, which is perfectly character- istic G. quinquangularis. To complete the confusion, Klatt, in the following year, decided that the species was, after all, a Hesperantha (he was quite incorrect in his decision), and published 7. quinquangu- laris, attributing it to Ecklon, Top. Verz. 23 (1827), a nomen nudum, and then validated it by citing in the synonymy his own G. quin- quangularis of 1865-66. Comment seems unnecessary. Beyond any doubt, the plant is closely related to G. erosa, and is possibly not separable from it. It differs chiefly in the number of flowers in the inflorescence, markedly in flower color, and somewhat in leaf pubescence, for the hairs on the leaf edges are spreading and not appressed against the leaf blade. 44, G. QUINQUANGULARIS, var. atrofaux, var. Flores et folia G. quinquangulare similis, sed flores pallide lutei clans, perianthii tubo et tepalis basi atrocaeruleo- vel brunneo-purpur Type: Schlechter, no. 5210 (B), from Pigustbers se "1000-1500 it. alt., Sept. 6, 1894 (1soryPEs in Gen). Perhaps this is no more than a color form of G. a scape but it occurs much to the north of the range of that specie Subsection Foliosae, subsect. nov. Folia us glabra; spa- tharum valva exterior herbacea; folia basi dense superposita. Type: G. foliosa Klatt. 45. G. montana, spec. nov. Cormus ovoideo-obconicu — 1 em. latus; tunicae imbricatae, apice cuspidatae: folia, ok 3, infimum vagina sine laminam, 4.5 cm. longum, apice breviter liberum, basalia altera ad 15 cm. longa, 2 mm. lata, superposita, vaginis basalibus 4.5 cm. longis, super vaginam angustata, in lamina lineari-attenuata, acuta dilatata, marginibus nervoque primario incras- ramus e bracteis duabus spathiformibus originans; inflorescentia spica disticha, flexuosa, 6-9-fl., ramis 3-4-fl.: spatharum valva exterior ad 56 FOSTER mm. longa, oblongo-ovata, obtusa, subtrilobulata vel re herbacea, apice margine purpurea, perianthii tubum aequan vel fue excedens; valva interior exteriorem aequans: ovarium a 2 m longum, subturbinatum: penne tubus 2 mm. longus: tepala ih. aequalia, ad 7-8 mm. longa, 2 mm. lata, elliptico-obovata, obtusa, integra, pallida (in siccis): genom 2-3 mm. longae, filamenta + 2 mm. longa, unum quam altera brevius: stylus ad 7 mm. longus, antheras aequans; stigmata 1 mm. longa, antheras exeedenta capsula immatura 4 mm. longa, subturbinata, semina non v TYPE: Dees s. n. (B; ISOTYPES in K, Gen), from Punaienial, 3000-— 4000 ft., Hasirat: stony and rocky places. SPECIMENS SEEN: Genadendal, 3000-4000 ft., Drége, no. 1553 (B). Superficially, this plant is very similar to small plants of G. ramosa, but it has a broader leaf with a different type of venation; the corm tunics are highly imbricate, rather thin and soft. I place it with the Foliosae because of the corm-tunics and because of the crowded, superposed basal leaves. 46. G, Drecet Baker. Plant to 14 cm. He Ltera/ much pouton corm ovoid, 6 mm. high, 3-4 mm. broad; t cs imbricate d at the apex: leaves in a ‘basal cluster ficdek siipebnsad tli, bit 3, to 4.5 em. long and 4 mm. wide, linear-lanceolate, somewhat recurved, acute, glabrous, the midrib and edges thickened, and at least one bulbil contained near the base of the su perposed sheaths; the cauline leaf to 3 cm. long, the sheath soniowlet ventricose, and sometimes lacking a bulbil: stem 1-branched, to 7.5 cm. tall, filiform, glabrous; the inflorescence a 5-7-flowered distichous spike, rarely with a bulbil in the spathes of the lower flowers: outer spathe to 6 mm. long, oblong, truncate, mucronate, herbaceous, membranous at the apex; the in- terior spathe as long: ovary + 2 mm. long, subturbinate: perianth- tube to 3 mm. long, about equal to the outer spathe: tepals salen ne to 1.6 cm. long, 3 mm. wide, ovate-elliptic, acute, yellow with a blue green blotch at the hake. continued downward on the tube: saraae about 14 as long as the tepals; the anthers to 3 mm. long, the filaments to 1.5 mm. long: style about 8 mm. long, as long as the anthers; stigmas recurved, exceeding fon anthers: capsule and seeds not seen.— Baker, Handbk, Trid. 158 (189 Type: Drége (K), from the Paar Mts. (isoTyPE in Gen). HaBitat: il oe not known SPECIMENS SEEN: ig Afvicn, without precise locality, Burmann f. (Gen), Bergius (B): Paarl Div.: Paarlberg, Drége, no. 1552 (B): Cape Div.: Table Mt., Kelle. no. 311 (Gen): Swellendam Div.: near the Zondereinderivier, at Appels kraal, and in the mts., Ecklon & Zeyher, rid. no. 222 (B, Gen): Uitenhage Div.: Olifantshoek, between STUDIES IN THE IRIDACEAE—II 57 Boschmansrivier and Zondagrivier, Ecklon & Zeyher, Trid. no. 221 (K, B). In no description known to me have I seen mention made of the bulbils in the sheaths of the basal and cauline leaves. In this species, they are sometimes absent from the cauline leaves. 7. G. RUPEsTRIS Schltr. Corm ovoid, 6 mm. high, 4-5 mm. wide, the tunics imbricate: basal leaves 3-4, to 4 em. long, 4 mm. wide, slightly indented adaxially above the sheath, widening into a linear- ensiform, or ligulate, or even subfalcate, acute blade, glabrous, the midrib and edges thick; cauline leaves 2-3, 1.5-0.7 cm. long, bract- like, with bulbils in the axils or bases of the sheaths: stem simple, terete, slender, glabrous, purplish, rarely 1-branched; the inflorescence nearly as long, subherbaceous, purple-membranous at the bifid apex: ovary to 2 mm. long, subturbinate: perianth-tube 2.5-3 mm. long: tepals subequal, 8-7 mm. long, 2 mm. wide, elliptic-obovate, acute, probably white: stamens half as long as the outer tepals; the anthers to 2 mm. long, filaments to 2 mm. long: style to 7 mm. long, exceeding the anthers; stigmas 1.5 mm. long, recurved: capsule and pte not present. —Schlechter in Braler, Bot. Jahrb. xxvii. 98 (1899). G. recurvifolia (Poir.) Klatt in Linn. xxxiv. 655 (1865-66), as to speci- mens cited, in part, but not as to name. Tyre: Schlechter, no. 9167 (B; IsoTYPES api sg from Bainskloof in Wellington Div., at 2400 ft. alt., Nov. 19, Hasitat: damp pla ces SPECIMENS SEEN: South Africa, without precise locality, Bergius in part a ee Link, no. 1197 (B): Dutoit’s Kloof, at 2300 ft. alt., Jan H. Bolus, no. 5247 in part (K). On the a en the anthers are about 3 mm. long and the fila- ments about 5 mm. long. It is undoubtedly close to G. Bolusii Baker, and possibly should be made a variety of that species. On the whole it is rather smaller, the basal leaves are smaller and much less long-sheathing in proportion to their size, and the style exceeds the anthers. The material seen is too inadequate, however, to permit final judgment on this point. Certainly, it is curious that Bolus, no. 5247 appears to contain a mixture of the two species, for this is the type of G. Bolusiv. 48. G. Botusu Baker. Plant 16-28 cm. high: corm small, ovoid, 1 em. high, 5 mm. wide; the tunics ya dt rather chartaceous, apically cusped: basal leaves 2 2-3, 3-10 cm. long, 3-5 mm. wide, long- sheathing, acute, the blades ensiform or sick ligulate, glabrous, the midrib prominent; cauline leaves 2-3, 4 cm. to 4 mm. long, all cauline 58 FOSTER leaves with bulbils in yrage sheaths: stem simple, terete, glabrous; the inflorescence a 3-8-flowered spike: outer spathe to 5 mm. long, oblong, truncate, miata: = the apex, an membranous-tipped, equalling the perianth-tube; the inner nearly as long: ovary ovoid, to 2 mm. long: perianth-tube + 2 mm. long: tepals subequal, to 8 mm. long and 2.5 mm. wide, elliptic-ovate, pe probably white: stamens half as long as the tepals; the anthers to 1.5 mm. long, the filaments to 3 mm. long: style 7 mm. long, as tone’ as the dntnay, the stigmas ex- ceeding the anthers: capsule and seeds not present. —Baker, Handbk. Trid. 1 (1892 ere: H. Bolus, no. 5247 in large sep (K), in saxosis montium circa Dutoit’s Kloof, 2300 ft. alt., Jan. In this species, so far as can be ak nud the exceedingly scanty material available, the basal leaves are not bulbilliferous, a point in species, the differences are slight and the two species should probably be united. 49. G. ovalifolia, spec. nov. Cormus parvus, ovoideus, altus, 4 mm. latus; tunicae imbricatae, apice cuspidatae: folia basalia 3, superposita, longe vaginantia, a ad 2 cm. longa, 5 mm. lata, super vaginam adaxialiter incisa, lamina Lphlem arrears acuta, glabra, nervus primarius incrassatus: caulis simplex, teres, glaber, 1-3-fl., ad 5 em. altus; bractea caulina 1(—2), bulbillifera amie spatharum valva exterior a mm. longa, plerumque periant ii tubum brevior, ovato- oblonga, obtusa, integra, herbacea; valva interior aequalis vel parum longior, subherbacea, integra, acuta: ovarium ad 2 mm. longum, sub- turbinatum: perianthii tubus ad 3 mm. longus, summo ampliatus, So ae (in sicco): tepala subaequalia, ad 7 mm. longa, m. lata, oblongo-ovata, obtusa, albida (in sicco): antherae ad 2 mm. "eedne: Rafienes ad 4 mm. longa: s tylus ad 7 mm. longus; stigmata 1 mm. longa, antheras parum excuenta: capsula turbinata, ad 4 mm. longa; semina minuta, 0.5 mm. lon TYPE: Heh from the Divkeustetiubargen, "4000-5000 ft. alt., October (B; IsoTyPE in G Sp tia a SEEN: Tulba gh, by the cy nr pipet ft. alt., November, Ecklon & Veuhiy, Irid. no. 223 (B, This small species of relatively aie a is 5 lies to G. Dreget in flower and tube color, and to G. Bolusii and G. rupestris in its eauline bulbilliferous bracts and in the absence of bulbils in its basal leaf-sheaths. Its general appearance and in particular its short, broad, almost oval leaves give it so distinctive an appearance that it seems worthy of specific rank. Certainly, there is more to separate it from its relatives than there is to separate G. Bolusii and G. rupestris. STUDIES IN THE IRIDACEAE—II 59 To judge from the specimens cited by Klatt for his G. recurvifolia, it is probable that G. recurvifolia in part, at least as to some specimens cited, but not as to name, should be regarded as a synonym of this species. 50. G. ovata (Burm. f.) Aschers. and Graebn. Corm ovoid, to 1 cm. high, 4-5 mm. wide; the tunics concentric, crustaceous, apically cusped: basal leaves 2-3, so closely superposed as to appear opposite, to 7 cm. long (usually shorter) and 1 em. wide, thick-textured, erect, the midrib prominent, the edges slightly déekehed and densely short- ciliate, the blade ovate-elliptic to ensiform, abruptly acute, send or heavily black-dotted, with a pronounce adaxial excision abov sheath; the single cauline leaf half-sheathing, the free portion re 1 8 e ciliolate: stem usually simple, sometimes 1- branched, terete, glabrous, 5-23 cm. long, averaging about 10-12 cm. long; the inflorescence a lax, distichous, 1—7-flowered spike: outer spathe i 1.4 cm. long, oblong-ovate, acute or eine at the apex, herbaceous, but edged and tipped with red-purple, the nerves incon- spicuous; the inner spathe almost as long: ovary to 7 mm. long, ovoid to turbinate: perianth-tube to 2 cm. long, usually much shorter, but exceeding the outer spathe as a rule, slightly ampliate at the top: tepals subequal, to 1.8 cm. long, 6 mm. wide, the outer red-purple, with deeper veining on the exterior, the inner white, oblanceolate- ovate, subacute or rounded at the apex: anthers to 3 mm. long; fila- ments nearly as long as the anthers: style to 2.8 em. long; the stigmas .5 mm. long, recurved, equalling or slightly exceeding the anthers: capsule and seeds not seen.—Ascherson and Graebner, Synops. r. 1* (176 Txia excisa L. f., Suppl. 92 (1781). Geissorhiza excisa (L. f.) Ker- Gawl. in Koenig and Sims, Ann. Bot. i. 224 (1804). Weithea excisa (L. f.) Eckl., Mg verz. = (1827). vl ged ta latifolia Spreng. e Steud., Nom. (ed 2)1.7 53 (1840), nomen — TYPE: no longer i in Burmann’s Herbari Hasirtat: stony places, SPECIMENS SEEN: Worcester Div.: Bainskloof, Sept. 27, 1936, Hafstrim & Lindeberg (G): Cape Div one Sty Zeyher soe Table Mt., 2100 ft. alt., Oct., 1879, H. Bolus, no. 4 04 (G, K); Table Mt., Ecklon, no. 831 (B); Table Mt., Aug. 26, ei. Tie | no. 2271 (K, Gen); Table Mt., near rivulets, Aug. 27, 1846, Prior GS); Devil’s Peak, Aug., 1882, MacOwan in Herb. Norm., no. 259 (G, B, Gen); Devil’s Peak, Sept. 27, 1895, Wolley Dod, no. 518 in part (K); Stanford Valley, Aug., 1901, A. Engler (B); a 1500 ft., Sept. 17, 1899, Schlechter, no. 1458 (B, Gen); Capetown, Spielhaus (B): New- lands, Sept. 9, 1883, Wilms, no. 3707 (B); Beans Bay, C. Wright (G, K): near Somerset in Hottentotsholland, Ecklon & Zeyher, Irid. no. 294 (Gen, B); Hottentotsholland, Zeyher, Fl. Cap. no. 39 (Gen, 60 FOSTER B); Stellenbosch, Aug. 20, 1846, Prior (KK); Stellenbosch, J. Sanderson, no. 980 (K): Paarl Div.: between Paarl and Pont, below 1000 ft., Drege (B, Gen): Caledon Div.: Hermanus, July 24, 1031... 7,.M, Salter, no. 1184 (K); Kleinriviersberg, 1000-3000 ft. alt., Ecklon & Zeyher (B); Kleinriviersberg, Zeyher, no. 3959 (K, B); Shaws Mt., Sept. 13, 1931, W. F. Barker, no. 11 (K). Note: in view of the plenti- ful material available, specimens lacking locality data have not been cited. This distinctive species seems almost never to be misidentified, unlike so many other species in the genus. Its long perianth-tube, thick-textured, ovate to lance-ovate leaves, and its color cause it to be easily recognizable. Its leaf-shape and arrangement lead me to associate it with the Foliosae. 51. G. PARVA Baker. Plant 6-10 cm. tall: corm small, ovoid, 7 ; , 4 mm. wide; the tunics concentric, hard, 1 apically “cusped basal leaves 2, long-sheathing, the blades 1-2.5 ¢ , 1- wide, thin-textured, subfalcate, acute, glabrous, the ¢ edges and tnidrib thick; the cauline leaf sheathing at the base but mostly free, 8-16 mm. long, inserted slightly below the middle of the stem: stem simple, terete, glabrous, flexuose; the inflorescence a 3-7-flowered, distichous, flexuose spike: outer spathe to 5 mm. long, oblong-ovate or sub- orbicular, as long as or slightly shorter than the perianth-tube, her- baceous, the truncate or subtrilobulate apex red-membranous; the inner spathe nearly as long: ovary 2 mm. long, subturbinate: perianth- tube 2-3 mm. long: tepals subequal, 7-5.5 mm. long, 2.5 mm. wide, obovate-spatulate, obtuse, light yellow: anthers 1.5-2 mm. long; filaments 3 mm. long: style 7 mm. long, exceeding the anthers; the stigmas short, recurved, exceeding the anthers: capsule turbinate, 4 prio long; the seeds globose, 14 mm. long.—Baker in Kew Bull. 1906: 26 Pees: Penther, no. 723, Houw Hoek in Caledon Div., Oct., 1894 (K). In his original description, Baker suggested that this plant was close to G. rupestris, from which it differed by having the outer spathe orbicular in shape. To me it appears to have the spathes of @. incon- spicua Baker and the corm of G. selacea. 52. G. NANA Klatt. Plant 4-9 em. tall: corm ovoid, 5 mm. high, 3 mm. wide; the hard tunics concentric, but sometimes giving the ap- ronan of being slightly imbricate: basal leaves 3, to 5 cm. long, 1.5 e, sheathing at the base, slightly narrowed above the sheath the blade iy ees acute, glabrous, the edges and mid- rib thick; the single cauline leaf subventricose at the ase: 1-2.5 cm. the main axis 1—2-flowered, the branches 1-flowered: outer spathe to 5 mm. long, ovate, abruptly acute or subobtuse, herbaceous, mem- STUDIES IN THE IRIDACEAE—II 61 ao at the apex; the inner x a as long as the outer: ovary ob- ong, 1-2 mm. long: perianth-tube to 2 mm. long, equal to or slightly shorter than the spathes: tepals tee to 4 mm. long, 1 mm. wide, ovate-elliptic, acute, the outer tepals dark (purple?), the inner white: stamens a little over half as long as the tepals; the anthers to 1.5 mm. long, the filaments 1. 5m m. long: style about 4 mm. long, as long as the filaments; the stigmas 1 mm. long, recurved: ‘pas capsule 4.5 2) long, sblotigcollipentd seeds not seen.—Klatt, Ergiinz. 57 (1882 Type: Zeyher, no. 3967, from atte Kraal by the Zondereinde Ri: in Caledon Div. (B; 1sorype in K). This is the smallest and most ‘dhe species in the genus; its inconspicuousness is probably the explanation for its not having been collected since the time of Zeyher. G. vioLtacEA Baker. Corm unknown: basal leaves 2, shortly sheathing, linear, —— glabrous, the edges and midrib thick, with ther nerves somewhat prominent, 18-22 cm. long, 4 mm. wide; the 2 cauline leaves sabteading branches, 12-3 ¢ em. long, almost en- the base, hyaline-edged, the acute apex membranous, lavender-tipped; the inner spathe nearly as long, the two nerves red-tinged: ovary 2-3 mm. long, subturbinate: perianth-tube about 6 mm. long, yellowish, sometimes exceeding the spathes by 1-2 mm.: tepals subequal, 1.5- 1.3 em. long, 4-5 mm. wide, oblanceolate, obtuse, rather dark violet, unveined: stamens 24 as long as the tepals; the anthers 3 mm. long, filaments 7 mm. long: style 1.4 cm. long, about equal to the anthers; the stigmas 2.5 mm. long, exceeding the anthers: capsule 7 mm. long, herein the seeds subglobose, on. rate 0.75 mm. diam., 906: 1906). dark-brown.—Baker in Kew Bull. 1 TYPE: Padiet no. 677, from Blau pti River, Knysna Div., Nov., 1894 (K). Hasrrar: along rivers at low altitudes. SPECIMENS SEEN: Knysna Div.: Keureboom’s River, 100 ft. alt., Nov. 11, 1894, Sohieihter, no. 5949 (B). This plant, large for its group of species, is at present known only from Knysna Div. Its leaves suggest those of G. Bolusii, but are longer and narrower. Like those of several species in this group, they seem not unlike the leaves of some species of Hesperantha. 54. G. rottosa Klatt. Corm ovoid-subglobose, 8 mm. high, 5 mm. wide; tunics concentric, oak-brown: basal leaves 4-5, closely super- to 8-9 cm. long, 5 mm. wide, linear-falcate, acute, glabrous, aeveral-nerved: the 2-3 cauline leaves basally sheathing for about 62 FOSTER half their length, 5-2 cm. long: stem 1—3-branched, terete, glabrous, 16-18 cm. long; the inflorescence a 4-flowered, lax, secund spike, branches 1-several-flowered: outer spathe to 1.2 cm. long, oblong- ovate, somewhat obtuse, herbaceous, membranous at the apex, several- nerved, longer than the perianth-tube; the inner spathe nearly as long: ovary subturbinate, 2-3 mm. long: perianth-tube to 4 mm. long: tepals subequal, to 1.4 cm. long, 4 mm. wide, elliptic-obovate, acute, lilac, with darker veins: stamens 7 as long as the tepals; the anthers 4 mm. long, the filaments 5 mm. long: immature capsule to 6 mm long, turbinate; seeds not seen.—Klatt in Linn. xxxiv. 658 (1865-66). Tyre: Ecklon & Zeyher, Irid. no. 218, from the mountains near a pan Swellendam Div. (B; 1soryrs in G). Hasit een ft. alt. Sansa EN: Robertson Div.: Buffeljagt’s River, near Rietkuil, in October, Zevies. no. 3961 (K). On occasion, the stem is branched from the very base; the corm is very similar to that of the following species. G. BRACTEATA Klatt. Corm subglobose, strongly flattened at the base; the tunics concentric, 8 mm. high, 8 mm. wide: basal leaves numerous, to 10 cm. long, 6-8 mm. wide, indented adaxially above the midrib prominent: stem branched at the base and then branched again several times, the branches glabrous, terete, 1(—2)-flowered, to 22 cm. tall: outer spathe to 9 mm. long, equalling the perianth-tube, oblong-ovate, obtuse, several-nerved, herbaceous; the inner spathe nearly as long, apparently herbaceous, noe ovary to 4 mm. long, subturbinate: perianth-tube 4 mm. long: tepals subequal, to 1 cm. long, 3 mm. wide, ovate or obovate, subobtuse, whitish (?) or yellow: stamens 14 as long as the tepals; the an ong, ments to 2 mm. long: style about 9 mm. long; the stigmas exceeding the anthers: capsule to 9 mm. long, fadblnnia: the seeds flattened- globose, 1 mm. in diam., dark-brown.—Klatt, Ergiinz. 57 (1882). Type: Ecklon & Seyber, virgin ‘a at Olifantshoek, between the mouths of the Boschmansrivier and a le helow 300 ft. alt., in September; Uitenhage Div. (B; isotype i SPECIMENS SEEN: South Africa, without te locality, Verreaux (Gen), Krebs, no. 339 (B): Riversdale, in the hills, Oct., 1892, Schlechter (B); Humansdorp, hillside, 400 ft. alt., Sept. 10, 1897, Galpin, no. 4674 (K); E. Olifants River, Oudtshoorn Div., Dr. Gill (K); woods at Olifantshoek, between the mouths of the Boschmansrivier and the Zondagrivier, in September, Ecklon & Zeyher, Irid. no. 221 (B, Gen); Uitenhage, Zeyher heed brickfields at Grahamstown, Oct. 25, 1905, Miss M. Daly, no. 861 (B). STUDIES IN THE IRIDACEAE—II 63 Not unlike G. foliosa, this species can readily be distinguished by its rather broader leaves and yellowish or whitish flowers. G. GEMINATA E. Mey. ex Baker. Plant 20-30 cm. tall: corm ova (incomplete on all specimens): basal leaves 2—3, superposed, sheathing the stem for at least 5 cm., to 12 cm. long, 1.5 mm. wide, glabrous, linear, , setaceously acute, midrib prominent, ld equalling the base of the inflorescence; cauline leaves 2, 3-9 cm. long, sheathing for half their Teagih: stem simple or 1-branched, sank axis sometimes 1-branched again, each 1—3-flowered, terete, glabrous: outer spathe to 8-9 mm. long, much shorter than the perianth-tube, oblong-ovate, herbaceous, but purple-membranous at the acute or obtuse apex; the inner spathe nearly as long as the outer: ovary 3 mm. long, oblong- oie perianth-tube dark green, to 8 mm. long: tepals subequal, 1.4 m. long, 5 mm. wide, lance-ovate, subobtuse, light yellow, the outer rank with red pencillings near the base: stamens little more than half as long as the outer tepals; the anthers 3-4 mm. long, the filaments 4 mm. long: style 1.2-1.5 cm, long; stigmas 3 mm. (7 mm in one instance) long, somewhat exceeding the anthers: capsule ovoid- turbinate, to 9 mm. long; pacar seeds dark brown, angular- globose to subpyriform, 0.75 mm. in diam.—Baker, Handbk. Irid. sg ae 2). = geminata E. Me sg ex « Drage i in Flora ii, Besond. Beigab. 83 (1843), nomen nudum. G. geminata (Vahl), incorrectly attributed to E. ay. « ex Baker by Klatt in Durand and Schinz, oe Fl. Afr. v. Bi 893). icin geminata Vahl, Enum. ii. 68 (180 Tyre: Drége, between Slangheuvel, Frenchhoek, and abides (K; ISOTYPE Gen). HapitaT: apparently in marshy ground SPECIMENS SEEN: Slangheuvel, on the Breede River, Drége, no. 1540 (B); Bergrivier, near Paarl, under 500 ft. alt., Drége, no. 8472 (B); near Brand Vlei, Worcester Div., in marshy ground, Sept. 23, 1932, . J. Lewis, in Herb. Bolus, no. 20414 (K). Through carelessness, this species has an unnecessary nomenclatural history. The first publication of the name in the genus Geissorhiza by E. Meyer in 1843 was not only without description, but also with- out any indication that it might be a new combination. It was, therefore, a nomen nudum. Apparently the first valid publication of the name in this genus was that of Baker in 1892, who not only gave a description but cited a single collection; this must therefore become the type. Actually, however, it is probable that Meyer intended his name to be a new combination, based on Ixia geminata Vahl. The sheet of Drége, no. 1540 in Meyer’s herbarium (now at Berlin) bears the label “ Geissorhiza geminata mihi,” followed by “Ixia geminata Vahl sed non Geissorhiza obtusata Ker.”’ This, I think, shows that a 64 FOSTER new combination was intended, but the combination G. geminata (Vahl) was not validly published until Klatt did so in 1893. As usual, he added a blundering touch by attributing the new combiretion te E. Mey. ex Baker. From the description of I. geminata Vahl I am inclined to believe that the identification of it with G. geminata may well be correct, but the proper authority for the specific name remains E. Mey. ex Baker. All but one of the specimens seen by me are in very poor condition, and it is hardly surprising that the species has remained little-known and misunderstood. Miss Lewis’s collection was distributed as G. humilis; from it the details of the flower have been drawn for the description given above. No complete corms remain on any of the specimens, so that it is uncertain whether the tunics are imbricate or concentric, 57. G. seracea (Thunb.) Ker-Gawl. Plant 7-12 em. tall: corm somewhat tubular at the base, enlarging to become ovoid; tunics con- centric, hard, dull-brown, long-cusped at t the apex: basal leaves about 3, to8 em. long, 1.5 mm. wide, subpetioloid above the sheath, widening into a linear or recurved portion, acute, glabrous, the midrib ¢ ‘and edges thick; the 1-2 cauline leaves 1.5-3 em. long, sheathing at the base. narrowed above the sheath, widening into a 1.5mm. wideac: * «le stem usually Morente oceasionally 1—2-branched, terete, filif iad glabrous, 1(—2)-flo : outer spathe to 1 cm. long, ovate-lanceolate, acute, aiabeea, herbaceous many-nerved, equal to or shorter than the perianth-tube; the inner spathe as long as the abe herbaceous: ovary 2-3 mm. Tees subturbinate: perianth-tube to 7 mm. long, ampliate near the apex: tepals subequal, to 1.4 cm. 1 ong. 4-5 mm. wide, subacute or obtuse, obovate-elliptic, whitish or pale elge, the outer rank reddish on the exterior: stamens less than half as long as the tepals; the anthers 3 mm. long, the filaments 2 mm. long: style to 1.4 cm. long, i: the anthers; the stigmas 24 mm. long, recurved: capsule and seeds not seen.—Ke r-Gawler i in Koenig and =. Ann. Bot. i. 224 (1804). 0 setacea (Thunb.) Baker in Journ. Linn. So-. me “e (1877), non Ker-Gawl. Izia setacea Thunb., Diss. are 13 1 Type: sheet labelled Ixia setacea in herb. Thunb. (not seen). Hasirar: rather damp sands, or in shrubber SPECIMENS SEEN: South Africa, without locality, Bergius (B in part pies no. 8493 (Gen in part): near Capetown, 1880, H. Bolus, o. 4803 (K); Table Mt., 2300 ft. alt., Feb. 2, 1892, Schlechter, no. 412 (B in in part): in arenosis ‘inundatis pr. Sir Lowry’ s Pass, 300 ft. alt., July 5, 1892, Schlechter, no. 1154: along the rivulet at Grahamstown, Burchell, no. 3546 (K). STUDIES IN THE IRIDACEAE—II 65 The corm of this species is rather remarkable for its almost tubular base, above which it expands to become ovoid; the inner spathe seems to be herbaceous, a condition not at all usual in this genus. For the most part, the description given has been based on Bolus, no. 4803, which has been annotated as follows by N. E. Brown: “These match the type specimens of [xia setacea Thunb. in Thunberg’ s Herbarium . . Thunberg collected it near Capetown.” 58. G. inconspicua Baker. Plant to 18 em. high, usually ne shorter: corm unknown: basal leaves 3, to 8 cm. long, 1-3 mm. slightly indented adaxially above the sheath, the blade sie re- curved, glabrous, acute, the edges and midrib thick; the cauline leaf inserted on the stem near the middle, 1.5—2.8 cm. long, half-sheathing, subtending the branch: stem simple or 1-branched, terete, glabrous, much exceeding the leaves; the inflorescence a 2—5-flowered, lax, dis- tichous spike: outer spathe oblong-ovate, obtuse or abruptly acute, to m. long, herbaceous, purple-membranous at the apex and in part along the margin, about equal to the perianth-tube; the inner spathe as loug, herbaceous-membranous, red-purple at the bifid apex: ovary 2-3 mm. long, turbinate: perianth-tube + 3 mm. long: tepals subequal, 1.2—-1 cm. long, 4 mm. wide, obovate, ineoe pale yellow, the outer rank red- tinted externally: stamens 34 as long as the tepals, or nearly equal in some cases; the anthers 3.5 mm. long, the filaments 6 mm. long: style to 1.1 em. long, about equalling the anthers; the stigmas 1.5 mm. long, exceeding the anthers: immature capsule 5 mm. long, turbinate; the seeds not seen.—Baker in Kew Bull. 1906: 26 (1906). Tyre: Penther, no. 713, from Montagu Pass in George Div. (not SPECIMENS SEEN: South Africa, aoe locality, W. M. Rogers (KK): eorge Div.: Montagu Pass, 6 660 m alt., Nov. 2, 1894, Schlechter, no. 5796 (G, B, K, Gen), ie Although the type has not been available to me, Schlechter, no. 5796, from the type locality, appears to fit the original description reasonably well, except that it is larger in almost every respect. The determination of the Kew specimen, incidentally, was made by N. E. Brown. There seems to have been some carelessness in drawing up the original description, for Baker stated that the perianth-tube was as long as the spathe and also as long as the perianth-limb, which was 6.3 mm. (3 lines) long. Since the spathe also was only 3 lines long, the tube could hardly have equalled the limb unless the ovary was completely lacking. As this could not have been the case, and since the ovary was probably about 2 mm. long, the tube was proba- 66 FOSTER bly about 4 mm. long. Baker considered the plant to be close to G. humilis, but it appears to me to be much nearer G. sctacca. In fact, it seems only doubtfully separable as a species. Without seeing the type, however, no change in its status will be made. formis: tepala subaequalia, ad 2.7-2.5 em. longa, 6-7 mm. lata, elliptico-ovata, apice obtuse rotundata, flava: antherae ad 5 mm. alt., Oct. 3, 1892 (B). This new species is related to G. ornithogaloides in general habit, but its flowers, which are enormous for this group of species, as well as for the size of the plant, and its almost filiform leaves will enable it to be readily distinguished. Flowers seem in dried material to have a greenish tinge to the yellow, but this may not be true of the living plant. 60. G. Marlothii, spec. nov. Cormus ovoideus, 9 mm. altus, 6 mm. latus; tunicae concentricae m bs attenuata, ad 5 em. longa: caulis simplex vel 1—2-ramosus, teres, glaber, rubro-brunneo-suffusus (in sieco); axis et rami 1-fl.: spatharum valva exterior ad 8 mm. longa, lanceolato-ovata, acuta, vel subobtusa, integra, herbacea, pars superior saepe rubro-suffusa; valva interior aequalis, integra: ovarium 2-3 mm. longum, subturbinatum: edd anthii tubus ad 3 mm. longus, summo ampliatus: tepala subaequalia, ad 1.3-1.1 em. longa, 4-5 mm. lata, obovata, obtusa vel subacuta, lutea, exteriora aliquando externe rubro-suffusa: antherae ad 4 mm. longae, filamenta ad 6 mm. longa: stylus ad 1.2 cm. longus, stamina aequans; stigmata ad 1.5 mm. longa, recurvata, antheras excedentia: capsula ad 9 mm. longa, turbinata; semina subglobosa, 0.5 mm. diam., atrobrunnea. STUDIES IN THE IRIDACEAE—II 67 Type: Marloth, no. 10612, from Houdenbeck Farm on the Cold Bokkeveld, Clanwilliam Div., 850 met. alt., Oct. 13, 1921 (B). eerpsioach san ied oi SPEC : Malmesbury Div.: Oude Post, Sept. 13, 1934, T. M. “Salter, no. 5. 4756 (K, in par This species differs from G. eee in color, since it does not turn blue-green on drying; nor is it a sulphur-yellow, but is clear. Moreover, the ovoid corm has concentric rather than imbricate corm-tunics. This may appear to be a rather fine line of distinction, but the difference in corm-tunics alone would warrant the separation of this plant as a new species. It has been named for the late Dr. Rudolph Marloth, the original collector, whose contributions to South African botany were numerous and distinguished. The type bears an unpublished name written in the hand of Rudolph Schlechter, but as he had assigned the plant to the wrong genus, it has seemed advisable to adopt another trivial for the species. 61. G. ORNITHOGALOIDES Klatt. Plant 4—12 em. tall: corm strongly flattened at the base, ovoid-globose, to 7 mm. high an the tunics hard, dull dark brown, apically cusped, imbricate, the basal edges irregularly frayed and serrate, not regularly notched: basal leaves 3, to 11-12 em. long, equalling or exceeding “he inflorescence, narrowed subpetioloidly above the sheath, expanding into a blade about 0.5 mm. wide, linear-subulate, glabrous, the midrib prominent, but the edges less so; the single cauline leaf to 8 cm. long, usually much shorter, the basal sheath somewhat ventriiose: stem simple or occasionally 1-branched, terete, filiform, glabrous, 1—-2-flowered: outer spathe to 1 em. long, exceeding the perianth-tube, lanceolate, acute, entire, herbaceous, with a membranous margin in the upper part; the inner spathe almost as long: ovary + 2 mm. one, a: gclig cating: perianth-tube about 2 mm. long: tepals subequal, t m. long, 3 mm. wide, oblanceolate-spatulate, subobtuse, eo wo oo tending to dry blue-green: stamens little more than half as long as tepals; the anthers to 4 mm. long, the filaments 4 mm. long: ei to 8 mm. long; stigmas about 1 mm. long; capsule to 8 mm. long, obovoid- tnebinates seeds about 0.5 mm. in diam., angular-globose. — in Linn. xxxiv. 656 (1865-66). G. romeulcosdes Eckl., Top. Verz. 21 (1827), nomen nudum. Ixia ornithogaloides Lichtenstein in Roem. and Schult., Syst. i. 376 (1817 TYPE: Lichtenstein, collected on the Koude Bokkeveld (B). ee in hills and mountains, possibly in dry locations. SPECIMENS SEEN: near Worcester, Ecklon é Zeyher, Irid. no. 224 (B): Caledon Div.: fwitlece and in the neighborhood of the warm baths, Ecklon & Zeyher, no. 225 (B, Gen); Zwarteberg, 1000-2000 ft. 68 FOSTER alt., in August, Ecklon & Zeyher (G, B); Zwartberg, near Caledon, in September, Zeyher (K). The first specimen cited by Klatt in his description of G. ornitho- galoides, which was published as a new species, was the Lichtenstein collection from the mountains of the Cold Bokkeveld. As it happens, this specimen is the type of Ixia ornithogaloides Lichtenstein, of the existence of which Klatt seems to have been unaware. The result is that the binomial which should have been treated as a new combina- tion must be attributed to Klatt alone. Since the Lichtenstein speci- men fits Klatt’s description reasonably well, and since it is the first specimen cited, I am selecting it as the type of this species. One of the most remarkable things about this plant is the corm. If the corm alone were to be seen, there is little doubt that it would instantly be regarded as having come from a small plant of some species of Hesperantha. The entire aspect, in particular the irregularly frayed and serrate basal edges of the imbricate tunics, is that of many species of Hesperantha. Nevertheless, the style and the spathes are those of Geissorhiza, so that it can be left in that genus without difficulty. The color of the flower, a greenish or sulphury yellow, is also worthy of mention, especially since in drying it often turns to a rather dark blue-green. Baker has utilized this point in his key, FI. Cap. vi. 66 (1896), as a means of differentiating between G. humilis and G. ornithogaloides, but as the two species are unlike in so many ways, this seems a rather inadequate character, especially since it does not always hold true. 62. G. ORNITHOGALOIDES, var. flava (Klatt), comb. nov. Plant 11-15 em. high: corm and leaves as in the species; the basal leaves to 11 cm. long and 2 mm. wide: stem 1-2-branched from near the base, or simple, 1—2-flowered: outer spathe to 1.2 em. long; the inner spathe subherbaceous: ovary to 4 mm. long: perianth-tube 2.5-3.5 mm. long, yellowish-green : tepals subequal, to 1.5 cm. long, 5 mm. wide, greenish- yellow, veined brown-purple (in some specimens) on the exterior of the outer series, obovate-spatulate, obtuse: anthers and filaments each 4 mm. long: style to 1 cm. long; the stigmas 2 mm. long, exceeding the anthers.—Geissorhiza flava Klatt, Ergiinz. 58 (1882). Tyre: J. C. Breutel, from the Cape of Good Hope, without precise locality (B). Hasirar: dry hillsides. SPECIMENS SEEN: Clanwilliam Div.: Koude Berg, at 3400 ft. alt., 1, J. W. Mathews, cult. in Nat. Bot. Gard., no. 1419/30 (K): in Hottentotsholland, Prior (K): Caledon STUDIES IN THE IRIDACEAE—II1 69 Div.: hillside above a vlei on the Villiersdorp Road, 1 mile from the turnout from Caledon, Sept. 14, 1931, W. F. Barker, no. 13 (K); be- tween Caledon and Villiersdorp, Zondereinde River Bridge, Sept. 15, T. M. Salter, no. 4792 (K). A comparison of the type of G. flava with G. ornithogaloides shows that the former is a trifle larger in general, and is usually 2-branched instead of simple or 1-branched. Klatt described the stem as hirsute, but apparently he mistook a few fibres from the driers or some other foreign substance for hairs. Actually, the Breutel specimen, like its very close match, Barker, no. 13, is quite glabrous. So close is the resemblance between the two species that it seems impossible to retain specific rank for G. flava. Even in the matter of size and branching there is a tendency for the two to intergrade. Section Tortuosa, sect. nov. Folia _— teretia, tortuoso- flexuosa, quasi spiralia. Type: G. corrugata Klat In view of the very peculiar leaves, and the oa that the single species in this section has flowers closely resembling those of Hesper- antha, it seems advisable to recognize its distinctness from the other species of Geissorhiza. 63. G. corruGaTA Klatt. Plant 4-6 cm. tall: corm ovoid, to 1 sheathing, to 4-5 cm. long, nearly 1 mm. wide, recurved, strongly prominent, the uppermost leaf with a subventricose sheath: stem ge : spathe to 1.3 em. lon ng, much exceeding the perianth-tube, lanceolate- ovate, subobtuse, entire, the edges united at the base for nearly 2.5 m embranous-herbaceous, edged and tipped with dull red-purple; the inner spathe nearly as long, more membranous, entire: ovary to 5 mm. long, turbinate: perianth-tube to 3.5 mm. long, peak gg! at the is brown: tepals subequal, 1.6-1.4 cm. lon ng, 6-5 mm. wide, obovate, subacute, clear yellow: stamens over half the tenith of the tepals; the anthers to 6 mm. long, the filaments to 4 mm. long: style 6 mm. long, exserted from the tube for 3 mm.; the stigmas 7 mm. long, filiform, fimbriate, exceeding the anthers: capsule and seeds not present.—Klatt, Ergiinz. 57 (1882). Type: Dr. Meyer, no. 24, from the Hantam Mts., in Calvinia Div., 1869 (B). Hasirat: probably dry, sandy places, but data not available. This extraordinary species, of which the type collection is appar- ently the only one known, is so completely on the border-line between Geissorhiza and Hesperantha, that its actual disposition is one of some 70 FOSTER difficulty. The corm, or what is left of one, seems more like that of Geissorhiza, while the style protrudes from the perianth-tube, even if only for 3 mm. This being so, it seems desirable to leave the species in Geissorhiza, although the 7 mm. long stigmas are certainly those of Hesperantha, as are the spathes. The leaves are curiously twisted in the manner of some species of Romulea. A search through earlier literature produced the description of Ixia tortuosa Lichtenst. in Roem. and Schult., Syst. i. 375 (1817), a plant which Baker, in Journ. Linn. Soc. xvi. 88 (1877), transferred to Romulea. In Fl. Cap. vi. 37 (1896), he gave a description of it and cited two specimens, Burchell, no. 1343, and the Lichtenstein type of I. tortuosa (which he had not seen, incidentally), both from the same region, the Middle Roggeveld, near Jakhals Fontein. His description and that of Lichtenstein strongly suggest the plant here treated as G. corrugata. The outbreak of war in Europe has made it impossible to secure either the Burchell or the Lichtenstein specimen. It is only too probable that Baker did not dissect the Burchell specimen, and without doing so, it would be very difficult to tell whether or not the stigmas are bifid, the distinguishing mark of the genus Romulea. | quote Lichtenstein’s description and the amplification of it given by Roemer and Schultes: “ Flore solitario subsessili, spathis tubo longiori- bus, foliis filiformibus flexuoso-tortuosis. Lichtenstein Spicileg. FI. cap. M “Bulbus magnitudine pisi majoris. Scapi bipollicares plerumque bini, toti vagina laxa membranacea tecti, in terra reconditi, ut flos acaulis videatur. Folia plerumque sex cum vagina connata, teretia, flexuoso-tortuosa apice recto subulato. Spatha tubo longior, valvulis pellucido-membranaceis, costa media solummodo viridi. Limbi laciniae flavae, erectae aequales, tubo duplo longiores. Habitat in regione Middelroggeveld prope Jakhals Valley, cap. b. spei.”” There is hardly a word in this description which could not with equal accuracy be applied to the type of G. corrugata. Nevertheless, a final verdict cannot be given without actually seeing the type of Ixia tortuosa. Subgenus Ixiopsis, subgen. nov. Cormi tunicae (quidem exteri- ores) fibrosae vel reticulato-fibrosae; capsulae oblongo-ovideae, nec turbinatae nec subturbinatae; semina alata. Type: G. Bojeri Baker. In corm-tunics, capsules and seeds, this small group of species stands well apart from the rest of the genus. The spathes, indeed, are the deciding factor which leads me to retain the group in that genus, although the general aspect of the flowers is also that of Geissorhiza. STUDIES IN THE IRIDACEAE—II 71 At present, it does not appear to be a completely homogeneous group, and there is a suspicion that at least one species belongs in Acidanthera. G. PatrersoniAE L. Bol. Plant 16-74 cm. tall: corm globose, flattened, 1 cm. high, 1.5 em. wide; the outer coats rather coarsely fibrous-reticulate, the inner coats smooth and papery: basal leaves 2, to 40 em. long, shorter than the stem, long-sheathing, the she aths many-nerved and narrowing abruptly into a linear, acute, glabrous blade, + 1 mm. wide, the midrib and edges thick; cauline leaves 3-4, the upper ones reduced more or less to sheaths: stem simple or 1—2- branched, terete, glabrous, stout; the inflorescence a somewhat lax, distichous or equilateral, 3-18-flowered spike, the branches few- flowered: outer spathe to 1.5 cm. long, somewhat exceeding the tube, oblong-ovate, acute, entire, herbaceous, membranous at the apex and along the edges; the inner spathe 3-5 mm. shorter than the outer: ovary to 4 mm. long, oblong-ellipsoid : perianth-tube to 5 mm. long, narrow, ampliate at the top: tepals subequal, subunguiculate, to 1.7 cm. long and 6 mm. wide, a smaller, lanceolate or oblanceolate- ovate, acuminate, sometimes submucronate at the apex, lilac-gray with darker veins: stamens a little over half as long as the tepals; the anthers about 5 mm. long, the filaments about 4 mm. long: style capsule is Be aie about 1.5 cm. long; the seeds are ane about 5 mm n diam , with ms wing + 1.5 mm. wide.—L. Ann. Bolus Herb. i. 132 (1915). Tyre: Mrs. T. V. Paterson, no. 47, from Redhouse, Uitenhage Div., August, 1908 (in Bolus Herb., not seen; isoTyPE in K). ee ABITAT: in stony clay CIMENS SEEN: South Africa, without locality, Burchell, no. 4048 ): ge Div.: mountains near Buffeljagt’s River, 1000-2000 ft. alt., Zeyher, no. 3990 (K): Riversdale Div.: prope pagum Rivers- dale, in collibus argillaceis lapidosisque, 850 ft. alt., Oct., 1904, H. speak 11385 (K): Humansdorp Div.: Zuurbron, 800 ft. alt., Oct., 27, H. G. Fourcade, no. 3332 (K): Albany Div. — between Siuchenawiivies and Karegarivier, near Jage , below 1000 alt., in June, Ecklon & Zeyher, ge no. 286 6 (B): Mossel Bay Div.: Mossel Bay, Sept., 1862, W. M. Rogers (K): Uitenhage Div.: Zwart- kopsrivier, in Sept., Ecklon & oe Trid. no. 230 (Gen, B); Bethels- dorp, Sept. 28, 1930, F. R. Long, no. 242 (K). The original description cited five collections; material of two of these has been available to me. Since the species is named for Mrs. T. V. Paterson, I am selecting Paterson, no. 47 as the type In dried material, this distinctive species has a certain resemblance to Gladiolus, a resemblance heightened by the alate seeds. The Kew sheet of Zeyher, no. 3990 was, indeed, first determined as Gladiolus 42 FOSTER permeabilis DelaR. by Baker, and then as Gl. dichotomus Thunb. by Klatt. Dr. Bolus noted in the original description, however, that in living plants the perianth is regular and that the stamens are equi- lateral. 65. G. Bosert Baker. Plant 25-50 em. tall: corm ovoid-conic, the outer tunics finely fibrous-reticulate, 1.5 em. high, 1 em. wide: basal leaves about 4, the lower 2 being reduced to sheaths about 3 cm. long; e true leaves 11-25 cm. long, 1-3 mm. wide, basally sheathing 3-8 , narrowed ge si above the sheath, widening into a linear, acute, Hakistin blade, the midrib and edges thick; cauline leaves 2, 10-15 cm. long, dieathine basally for 1-3 cm.: stem usually simple, terete, glabrous, somewhat flexuose; the inflorescence 1—2-flowered, exceeding the leaves: outer spathe to 1 cm. long, exceeding the perianth-tube, oblong-ovate, abruptly acute, herbaceous; the inner spathe slightly shorter, subherbaceous: ovary to 6 mm. long, nas ee eae hee to 1.5-2 em. long: tepals subequal, to 9 mm. long m. wide, ovate-elliptic, acute, yellow (or pinkish?) : praca: a little. over half as long as the tepals; the anthers 3 mm. long, the filaments 2.5 mm. long: style about 7 mm. long; the stigmas recurved, exceeding the pedicellate, 1-3 cm. long; the seeds brown, 4 mm. in diam., alate. aker in Journ. Bot. xiv. 239 (1876 Type: Hilsenberg & Bojer, from ‘the Antoungoun Mts., Emirne Prov., Madagascar (K). Hasrrar: dry places, 1500-2500 met. alt., confined to Madagascar. SPECIMENS SEEN: Madagascar: in the interior, at Ankaratra, Scott Elliot, no. 2092 (B); Mt. Anteby above Ambositre, Dec. 11, 1894, C. J. Fo orsyth-Major, no. 662 (B); massif de sapoagt taks vallées de la Riambava et de I Antsifotra, 1600-2000 met. alt., Nov. 27—Dee. 8, 1924, H. Humbert, no. 3655 (G); central sieieg Sakoerintang, G. W. Parker (K). The only species with which this could be confused is the recently described G. ambongensis H. Perr., which is, like it, confined to the island of Madagascar. From the description of the latter, it appears to differ in having a longer perianth-tube, orange-yellow flowers, with two blood-red triangular marks at the base of each tepal, and in being a plant of lower altitudes. G. Bojeri has clear yellow flowers, which sometimes turn pink in drying, a relatively short perianth- tube, and is confined to the higher altitudes of the mountains in the more central portions of the island. The only capsules seen, on Parker (K), are very long and narrow. It has been suggested by Scott Elliot, in Journ. Linn. Soc. xxix. 59 (1891), that a broader capsule is more characteristic. In the original STUDIES IN THE IRIDACEAE—II io description, Baker suggested a relationship with G. alpina Hook. f., but this species is a Hesperantha. 66. G. HESPERANTHOIDES Schltr. Plant to 32 em. tall: corm ovoid, 1 cm. high, 7 mm. wide; the outer coats finely fibrous or fibrous- PET erg the i inner coats solid, probably becoming fibrous with age: basal leaves 2, 12-26 cm. long, 0. 5-1.5 mm. wide, much shorter than the stem, linear-recurved, acute, ats the midrib and edges very thick; the single cauline leaf 3-12 cm. long, sheathing for about half its length: stem simple, terete, sh getion 1-3-flowered: outer spathe to 1 em. long, shorter than the perianth- tube, oblong-ovate, truncately acute, herbaceous, becoming purple-membranous near the apex, the apical edge white-membranous; the inner spathe as long as the outer: ovary 2 mm. long, oblong-ovoid: perianth-tube to 9 mm. long, ampli- ate near or above the middle, pale in color: tepals subequal, to 1.6 cm. long, 4 mm. wide, obovate, obtuse, purple: stamens about 24 as long as the tepals; the anthers 4 mm. long, the filaments 7-8 mm. long, inserted at the middle of the tube according to Schlechter: style to 1.8 em. long, not quite equal to the anthers; the stigmas short, curved, exceeding the anthers: capsule and seeds not seen.—Schlechter in cia Bot. Jahrb. xxvii. 97 (1899). : Schlechter, no. 9555, from Vogelgat, in Caledon Div., at 3500 ft. ‘alt., Dec. 2, 1896 (B; 1sorypEs in K, Gen HaBitTaT: among rocks, at relatively high altitudes. SPECIMENS SEEN: Bredasdorp Div.: among rocks on top of Zout- melkskop (highest peak on range), Jan. 5, 1931, en ne no. 11334 (K). The Galpin collection lacks the corm, and without this it is impos- sible to state definitely that his plant belongs to this species, but in all other respects it appears to be G. hesperanthoides. On the isotype in Kew, N. E. Brown has noted: “ Schlechter 2160 from the Langebergen is the same as this.” Since Schlechter, no. 2160 is the type of G. Burchellii, a species in a different subgenus, it might be desirable to point out one or two differences, for the two plants superficially do resemble one another. In G. Burchellii all the corm-tunics are solid and imbricate, not fibrous; the stamens are markedly unequal. Neither of these characteristics occurs in G. hesperanthoides. 67. G. cranpis Hook. f. Plant about 25-30 cm. tall: - small globose; tunics finely fibrous-reticulate: basal leaves 2-3, 20 cm. long, 2 cm. wide, ng hapa the midrib prominent; the cauline leaves shorter: stem to m. tall, stout, terete, glabrous, simple; the inflorescence a isweel” riches lax, secund spike: outer spathe to 4.2 cm. long, elliptic- or lance-ovate, acute, herbaceous, much exceeding long, more or less ovoid: perianth-tube very slender, to 2.5 cm. long, slightly curved: tepals subequal, to 4.5 cm. long, 1 em. wide, elliptic- 74 FOSTER obovate, obtuse, yellow with a red midrib: anthers 7 mm. long, the filaments 7 mm. long: style about 4 cm. long, not quite equal to the anthers; the stigmas to 9 mm. long, much exceeding the anthers: ent and seeds not seen.—Hooker f. in Bot. Mag. xevi. t. 5877 1 ; Type: described from living material sent from Port Elizabeth, Cape Province, de ge in 18 SPECIMENS SEEN: Hort. Barr, in June, 1881 (K); Hort. Barr, in June, 1889 (K). Since the two specimens seen consist only of the inflorescence, and upper portion of the stem, the details of the vegetative parts have been taken from Hooker’s plate and the original description. Neither capsule nor seeds were mentioned by Hooker, nor does Baker mention on so that it cannot be determined whether or not the seeds were alat In “‘Haadbk Irid. 159 (1892), Baker stated that the style is “but little exserted from the tube.”” Hooker’s plate shows the style well- exserted, and this is true of the specimens seen by me. Baker's additional comment that the plant “recedes in the direction of Acidanthera’”’ seems quite correct. The material available is so scanty that final judgment cannot be passed on this point, but the species seems almost certainly not to belong in Geissorhiza. DoustFuL oR UNKNOWN SPECIES Mog oc H. Perr. in Notul. Syst. Mus. Hist. Nat. viii (2). This species, described in great detail by M. Perrier, is closely related to G. Bojeri, but seems, ex descr., to be distinct. No material has been seen, and Mr. Weatherby was unable to locate and photo- graph the type for me. Consequently, it has been placed in the category of species unknown to me, although there seems no doubt of its distinctness. G. FruiFoiA Baker in Journ. Bot. xiv. 238 (1876). No material has been seen which can certainly be assigned to this species. Through the kindness of Mr. Weatherby, a photograph of the type, Prior, collected in 1847, in the British Museum has been received. The narrow leaves are filiform, about 1 mm. wide, the longest about 25 cm. long, exceeding the inflorescence by about 2 cm. The inflorescence is a lax, 3-flowered, secund spike. In the original description, the flowers are said to be whitish, with a purple throat, and the relationship is said to be with @. secunda, although the spathes are reported to be entirely green. From the photograph, which is very STUDIES IN THE IRIDACEAE—II 75 clear, I should be inclined instead to relate it to G. juncea, but no definite statement can be made until the type or other material is seen. G. macra Baker in Bull. Herb. Boiss. (ser. II) iv. 1004 (1904). A photograph of the type, Jacottet, no. 1937, at Geneva is all that has been available to me. It was collected on Mt. Matchatchi, in the Transvaal. Since this region is so far from the normal range of the genus, it is hardly surprising to find that the plant has the aspect of Hesperantha or Acidanthera. The corm, which would assist materially in making a decision, is lacking. If Baker were correct in saying that the style is extruded from the perianth-tube and has short stigmatic branches, the plant cannot be a Hesperantha. The perianth-tube seems in the photograph to be somewhat bent, rather than “erectus” as Baker originally described it. This fact, the general appearance of the flowers, the shape of the leaves, and the locality—all make me think that the species should probably be in Acidanthera. Certainly, the 208 does not appear to be a relative of G. setacea, as Baker stated. G. QUADRANGULA (DelaR.) Ker-Gawl., Irid. Gen. 88 (1827). Izia adams De la Roche, Deser. 16 (1766 66). So far as I know, no one has seen any material which can certainly be assigned to this species since the time of its original description by Dela Roche. The type is probably no longer in existence. It has seemed possible that it may be close to or identical with G. splendi- dissima, but there are several points of difference between the descrip- tions of the two plants, and I should hesitate to make the identifica- tion, without seeing the type of J. quadrangula. G. eraciuis Baker, Handbk. Irid. 155 (1892). No material has been seen which can be assigned to this species, nor was the type, 7'yson, no. 1872, included in the specimens received from Kew. The locality given for the type, the Zuurberg in Griqua- land East, is in itself enough to raise doubt as to the correct generic attribution of the species. As it happens, another species, [xia brevi- folia Baker, Handbk. Irid. 165 (1892), was described from T'yson, no. 1892, this specimen likewise coming from the Zuurberg; in Fl. Cap. vi. 84 (1896), this number was reported as no. 1872. Consequently, it would appear that G. gracilis and I. brevifolia were described from the same collection. Baker reported that I. brevifolia belongs in Ixia, subgenus Morphixia, which, as pointed out previously, is very close to Geissorhiza in appearance. The descriptions are similar, differing chiefly in the lengths given for the perianth-tubes. For each species, too, the corm is unknown. To find two species in different genera, so closely resembling one another, ex descr., from the same altitude on 76 FOSTER one mountain, with probably the same collector’s number as type, and the locality well outside the range of one of the genera involved is rather unusual. Until further information can be received concerning the types of these two species, G. gracilis must be regarded with some scepticism. It appears quite probable that the plant does not belong in Geissorhiza. Species ExcLUDED FROM THE GENUS G. abyssinica R. Br. A. Rich., Tent. Fl. Abyss. ii. 308 (18 51?) as LAPEYROUSIA pecans (R. Br.) Baker in Journ. Linn. Soc. xvi. 155 18 G. an seemeciae (Jacq.) Klatt in Linn. xxxiv. 657 (1865-66). Jia anemonaeflora Jacq., Ic. Pl. Rar. il. t. 273 (1786-93) ; Collect. v. 10 (1796). The specimens ad by Klatt belong in Geissorhiza, but the plant figured and described by Jacquin does not. The spathes are shown as short, and are described as scarious, one being tridentate, and the other bidentate. It is probably an Ixia G. Briartii DeWild. & Th. Dur. in C. R. Soe. Bot. Belg. xxxix. 105 (1900). Originally described from a Briart specimen collected at Mussima (Haut-Lualaba), in the Belgian Congo, this plant is reported to be very similar in general appearance to Lapeyrousia Welwitschii Baker, but differs in having the stigmas entire as in the genus Geissorhiza. In recently studying a number of species of Lapeyrousia, I have found several which have entire or only slightly retuse stigmas. These in- clude such species as L. abyssinica, L. coerulea Schz., L. avasmontana Dinter, and probably others. Consequently, it appears proper to exclude G. Briartii from Geissorhiza and to consider it a Lapeyrousia. Without seeing the type, it would be improper to make a formal transfer or to identify the plant with any species now in Lapeyrousia. G. esate ve (DelaR.) Ker-Gawl. in ree and Sims, Ann. Bot. i. 223 (1804). Jaia inflera DelaR., Deser. 1 (1766). There has been some question concerning this species virtually since its original description. Baker, in Journ. Linn. Soc. xvi. 94 (1877), regarded G. vaginata Sweet, Brit. Flow. Gard. ii. t. 138 (1826), as synonymous with it, and retained this opinion in all his published work. To me, this seems quite correct. De la Roche described the spathes of his plant as large, erect, almost as long as the perianth, and the flower color as “ subviolacea luteoque insignitis,” a description with which Sweet’s plate agrees. Nevertheless, the specimens cited by Baker, Fl. Cap. vi. 73 (1896), agree neither with this description nor with Sweet’s plate. Sweet’s original description contains the signifi- STUDIES IN THE IRIDACEAE—II V7 cant statement that the style does not protrude from the perianth- tube, while the stigmas are relatively long. In other words, his plant was a Hesperantha. Upon discovering this, a study was made of Hesperantha, and it was found that H. Metelerkampiae L. Bol. is a perfect match for Sweet’s figure of G. vaginata. Although the type of H. Metelerkampiae has not been seen, I have seen a specimen sent to Kew from the Bolus Herbarium. Consequently, the species must be transferred to — so that it becomes ANTHA inflexa (DelaR.), comb. nov. Jia inflexra DelaR. G. infer “(Dela .) Ker-Gawl. G. ena Sweet. H. Metelerkampiae L. in Ann. Bol. Herb. iv. 114 (192 ars latifolia (DelaR.) Baker i sone Linn. Soc. xvi. 94 (1877). Ixia ng er ee Descr. 22 (1766). = TRITONIA LATIFOLIA (DelaR.) N. E w Bull. 1929: As 980 G. Daag tuba Klatt in Linn. xxxv. ag (1867-68) Fs eae Loneres (Klatt) Baker in Journ n. Soc. | DOMINICA BARGUOA SAl ° ost galaong ¢%> T MITT BS uevs. ANTIGUA on I t MARTINIQUE ADELOU + QMONTSERRAT Fe PE a po geren. omens Anan ‘onnere 4 DOMINICA % DOMINICA Dyrnwrngue Dy rnernave Osx LUCIA Osx LUCIA BARBADOS BARBADOS Ost VINCENT Pp , Ee VINCENT S da oe ?* * 2 GRENADINES 2 GRENADINES ere e%e Q Q _ Pt b 2 GRENADA fi) GRENADA It ad Map I. Distribution of CHARIANTHUS COCCINEUS, and its var. PARVIFOLIUS er I. Distribution of C. YMBOSUS and its varieties: var. TYPICUS (t); var. LATIF la); var. comarsontid (lo). Map III. peed sar of C. PURPUREUS and its varieties: var. TyPIcUS (t); var. CRINITUS (€); var. RUGOSUS (Tr); Var. BREVISETOSUS (b). show no proof of ever having been connected in the north with the Greater Antilles or in the south with the continent. Therefore, their biota, in the opinion of most Antillean biogeographers, is probably of waif-origin. The geological evidence thus suggests that the endemism of Charianthus in the West Indies is not relict but recent. § Eccharianthus in Jamaica is found at middle elevations, whereas, with the exception of C. corymbosus, var. typicus, the four species of § Eucharianthus in the Lesser Antilles are limited to the summits of the highest volcanic islands. Altitude with its concomitant factors is apparently decisive in the distribution of the genus, for on the low- 120 HODGE lying islands of the are—St. Eustatius, Antigua, Barbados, and the Grenadines—Charianthus, in spite of thorough collecting, has not been found. All four species of § Eucharianthus have close affinity, but the two setose species, C. purpureus and C. nodosus, and the two glabrous species, C. corymbosus and C. coccineus, are between them- selves even more nearly related. Species of highest altitudes (all but C. corymbosus) are the largest flowered; and since this habitat, in a region of active vulcanism, would be the last presumably available, it may be suggested that these species are the most recent in origin— and are possibly derivatives of C’. corymbosus by way of C. coccineus. In connection with this study, loans have been kindly forwarded from the herbaria of the Arnold Arboretum (A), and the New York Botanical Garden (NY). Collections bearing the symbol (G) are in the Gray Herbarium. To Dr. H. A. Gleason of the New York Botani- cal Garden, for helpful advice, and to Professor M. L. Fernald for invaluable suggestions and guidance, I wish to express my apprecia- tion. Synoptic TREATMENT OF THE GENUS CHARIANTHUS D. Don. Shrubs or small trees, generally much branched: leaves 3-5-nerved, decussate, lanceolate to ovate or el- youngest leaves usually covered with either close, brown or whitish- scurfy scales: flowers showy, in terminal (rarely lateral-axillary), paniculate or corymbose cymes, parts in fours: calyx-tube campanulate or urceolate, obscurely and obtusely 4-lobed or repand; the limbs per- sistent, each with a minute, bluntish, exterior tooth: petals 4, ovate to obovate or narrowly oblong-spatulate, glabrous, free, erect, usually forming an imbricate, bell-shaped corolla, bright red to crimson-purple, sometimes with creamy margins: stamens 8, in two equal series, inflexed in bud, much longer than the petals; anthers with longitudinal slits or rarely a terminal pore; connective not elongate and without any appendage: ovary completely fused to hypanthium, glabrous at apex, 2-4-locular; style very long, filiform, with simple, rounded stigma: fruit baccate, crowned by the persistent calyx, sub-globose or urceolate, juicy and black at maturity; seeds pyramidal—Mem. Wern. Soc. iv. 327 (1823); DeCandolle, Prod. iii. 196 (1828) and Mém. Melast. 81 (1828); Endlicher, Gen. no. 6263 (1836-1840); Naudin in Ann. Sci. Nat. sér. 3, xviii. 111 (1852); Grisebach, Fl. Br. W. I. Isl. 263 (1864); THE GENUS CHARIANTHUS 121 Fl. Trinid. & Tob. hei (1934). Type Specres: C. corymbosus (L. C. Rich.) Conia The sections epee and Eccharianthus are retained in this treatment and are separable by the characters under the two primary headings in the key. Sect. I. Cee om ee Fl. Br. W. I. Isl. 263 (1864); Cogniaux in DC. Monogr. Phan. vii. 714 (1891 Sect. II. foceiiania cepa Fl. Br. W. I. Isl. 264 (1864); Cogniaux in DC. Monogr. Phan. vii. 717 (1891). Key To SPECIES a. Petals broadly ovate; filaments not pruinose; anthers longitudinal slits; ovary 4-loculed; fruit subglobose with: a s reading — ealyx-limb. Sect. EucHARIANTHUS rise teas b. Young growth, i. e. re — and petioles, essentially glabrous, lacking setae... . c. Flowers large, ae ‘10-12. mm. long; filaments 15-18 eee long; style 15-30 mm, long; leaves usually pune- e beneath with persistent, appressed scales...... 1. C. coccineus. 6 Fiowue small, petals 5-7 i ong; filaments 5-6 mm Spek style 10 mm. long; leaves aie glabrous b. Young growth, i. e., twigs, leaves or petioles, setose.... d, Cymes ens 9 erect; leaves not punctate beneath with Pe eee yee ye ees ce 3. C. purpureus. ° a ° RD oO — ca i) Ko) 5 o i * wn aE eath, with closely coocine BiIOR ee ees . C. nodosus. a, Petals narrowly-oblong to spatulate; “fila aments pruinose a — = a pore; ovary 2-loculed; fruit urceolate, with an erect persistent calyx-limb. Sect. EccHARIANTHUS Girsebas Bae 6 oat ee es is a 5. C. Fadyeni. 1. CHARIANTHUS CoccrINEUs (L. C. Richard) D. Don, var. typicus. Shrub, up to 3 m, tall; twigs stout, terete to obscurely tetragonal, glabrous: leaves 5-nerved, ovate-elliptical, entire, abruptly short- acuminate, subacute to broadly rounded at base; blades 5-11 cm. long, 3-6 cm. broad, youngest leaves closely and densely brown-lepidote beneath, at length glabrescent with scattered, punctiform appressed scales, upper surface glabrous, coriaceous; petiole 0.5-3 cm. long, glabrous: cymes many-flowered, loosely-corymbose or -paniculate, 2-5 cm. long, the basal peduncles 1-3 cm. long; the i ip pedicels, hypanthia and calyx-tubes lightly lepidote: pedicels 1-5 m long; hypanthia 3-4 mm. long: calyx-tube 2-2.5 mm. long, nag a minute, blunt, exterior tooth just below the retuse tip of each lobe: corolla red; petals 10-12 aaa es ovate, obtuse: stamens inflexed in bud; filaments red, 15- . long; anthers arcuate, 2 mm. long: style red, 1.5-3 cm. ee te re rounded: berry subglobose, A mm. in diameter, juicy-black at maturity—Mem. Wern. Soc. iv. 328 ia. HODGE (1823) ; DeCandolle, Prodr. iii. 196 (1828); Naudin, Monogr. Melast. in Ann. Sc. Nat. sér. 3, xviii. 111 (1852); Grisebach, Fl. Br. W. I. L, 263 (1864); Triana, Melast. 99 (1871); Cogniaux i in DC. Monogr. vii. 714 (1891). Me lastoma coccinea L. C. Richard in Act. Soc. Hist. Nat. Paris. 109 (179 2). M. alpina Swartz, Prodr. 71 (1788); Swartz, FI. Ind. Oce. ii. 800 (1797); Willdenow, Sp. we 2, 597 (1800). C. tinifolius . Don in Mem. Wern. Soc. iv. 329 1823); DeCandolle, Prodr. iii. 197 (1828); Cogniaux in DC. Monogr. Phan. vii. 715 Soar hai & Rendle, Fl. Jam. v. 367 (1926). C. “coriaceus DC * Phan. Ant. Fran. 285 (1897), an error, presumably for C. poet neus ji Lesser Antilles. 1 have seen the following specimens. GUADE- : Madiana s. n. (NY); Savane a Mulets, chemin de la grande- ‘aie 1000-1200 m., July 15, 1892, Duss, no. 2259 (G, NY); sphagnetum de la Soufri riére, chemin de la seine 1100-1480 m., Feb. 16, 1936, Stehlé, no. 318 (NY); cones et savanes volcaniques, savane aux Ananas, 1100 m., April 25, 1936, Stehlé, no. 878 (NY); foret, sommets, voleaniques, Lac Flammarion, 1100 m., Oct. 11, 10, 1936, Stehlé, no. 1154, no. 1255 (NY); mountains above Bain Jaunes, Oct. 31, 1938, L. H. & E. Z. “Bailey, no. 161 (NY). Domrntica: middle winfocest-clad slopes, 800 m., Morne Anglais, Aug. 3, 1938, Hodge, no. 516, (G). PLate 6, Fic. 1; MAP I. C. coccineus is recorded by Cogniaux from Porto Rico (Stahl), Santa Domingo (Bertero; Eggers, no. 6916b), and from British Guiana (LeBlond, tyre). Eggers no. 6916b has been shown by Urban to be a misidentification; while the Bertero specimen, which DeCandolle de- scribed from fruiting material as Conostegia discolor, is very questiona- ble. The other extra-limital (outside the Lesser Antilles) citations are doubtful, particularly the LeBlond specimen which is discussed under C. corymbosus. Var. PARVIFOLIUS Cogniaux. This variety differs from var. ¢ in its smaller leaves, with petioles 5-10 mm. long; blades bee lanceolate to ears tere 4—6 cm. long, 1.5-3 em. broad.—Cogn. in DC. Monogr. vii. 1190 (1891).—Sr. Vincent; known only from this island: Guilding s. n. (NY); Soufri 6916 (p. p.) (A); Mt. St. Andrews, 2200 ft., Jen: 1890, Eggers, no. = (. 44 (IsoTyPE, in herb. Arnold Arboretum.) Puate 6, FIG. 2; pone coccineus and C, corymbosus are the only two glabrous (lacking long setae) species of the genus in the Lesser Antilles. They are apparently of very close affinity, and unless specimens are collected in full flower are difficult to separate in the herbarium. Richard’s descriptions are poor, and the only good clue (aside from an examina- THE GENUS CHARIANTHUS Bas tion of his types) to the identity of - species is the statement that C. coccineus is laxly flowered and “racemous’’, while C. corymbosus frequently approaches C. coccineus in this character as well as in leaf- shape. Grisebach, referring to the closely appressed scales as “ black dots beneath”’, states that C. coccineus always possesses them, whereas C. corymbosus (C. glaberrimus DC.) does not. This generalization is not infallible, for, although the scales on the leaves of C. coccineus generally persist for a very long time, nevertheless nearly all the species possess this type of scale for a time, at least, in their youth. C. coccineus always has much larger flowers; this difference, coupled with its distribution at the highest elevations, is sufficient to dis- tinguish it specifically from C. corymbosus. Since the original description by Don in 1823, the identity of Charianthus tinifolius has been very questionable. The type is an Anderson specimen (“ Hab. in India Occidentali’’), in Don’s time in the Lambert herbarium but cited by Cogniaux as in the herbarium of Delessert. It was not seen by DeCandolle, Naudin, Grisebach or Triana, the majority of whom placed the name in their lists of species dubious or unknown. Cogniaux, in his monograph, elaborated the description of the species for the first time, and cited an additional specimen of MacNab from Jamaica. In their Jamaica Flora, Fawcett & Rendle take up Cogniaux’s description and citations, but question the MacNab specimen, referring it to St. Vincent (?) rather than to Jamaica. This seems more logical, for section Eucharianthus is not known from Jamaica, and Anderson made his chief collections in the Lesser Antilles. In the absence of type material, I am unable to separate, from Cogniaux’s description, C. tinifolius from C. coccineus. If St. Vincent is the actual locality for the Anderson and MacNab specimens, it is probable that they both represent C. coccineus var. parvifolius. In his flora of the French Antilles, Duss published the name C. coriaceus DC., and this epithet also appears on many of his herbarium labels. Several specimens so named (Duss, no. 2259, G, NY), which I have seen, represent material of C. coccineus. Since the published description also applies to C. coccineus, it is safe to assume that C. coriaceus was published in error for C. coccineus. Coriaceus, as a specific or varietal epithet, has never been more formally applied in the genus Charianthus, and certainly not by DeCandolle. Duss, working alone, in Guadeloupe, apparently had to rely on his memory 124 HODGE for a great many of his names, for errors of this type also appear under C. corymbosus and C. nodosus. 2. CHARIANTHUS coryMBosus (L. C. Rich.) Cogniaux. Large shrub or small tree, up to 8 m. tall; crown spreading, much branched; bole fluted, bark finely furrowed; wood medium-hard, close-grained ; twigs terete to subtetragonal, glabrous: leaves 5-nerved, lanceolate to broadly ovate-elliptical, entire to minutely-crenulate or serrulate, abruptly short-acuminate or gradually long-attenuate, cuneate- attenuate to obtuse or subcordate at base; blades 7-16 cm. long, 2.5-9 cm. broad, the lower surfaces of youngest leaves minutely brown- lepidote, at length glabrous, upper surfaces glabrous, membranaceous to subcoriaceous; petioles 1.5-6.5 cm. long, glabrous: cymes generally closely corymbose, many-flowered, 5-10 em. long, the branches often reddish: pedicels 2-4 mm. long; hypanthia 2-3 mm. long: calyx 1-2 mm. long, minutely but sparingly lepidote, each calyx-lobe bearing 4 blunt exterior tooth near its obtuse apex: corolla red, often with buff- colored margins, petals 5-7 mm. long, ovate: stamens inflexed in bud; filaments 5-6 mm. long; anthers 2 mm. long, arcuate: style 10 mm. long, the tip rounded: berry globose, 5 mm. in diameter, black. a. Leaf-apices abruptly short-acuminate; leaf-bases rounded to subcordate, never attenuate b. . Cymes small, short-peduncled, closely corymbose; leaves : ovate, the blades averaging 8 cm. long, 5 em. wide...... Var. typicus. b. Cymes large, long-peduncled, loosely corymbose; leaves broadly elliptical, the blades averaging 13 cm. long, 8 cm. : WN ee erie te oe et ote Ae ig Oh aide acs Var. latifolius. a. Leaf-apices gradually long-attenuate-acuminate; leaf-bases : attenuate to narrowly rounded...................... Var. longifolius. Var. typicus. Leaves 5-nerved, ovate to ovate-lanceolate, entire to minutely-crenulate or serrulate, apex abruptly short-acuminate, base narrowly-obtuse to subcordate, blades 7-13 em. long, 2.5-6 em. broad, membranaceous to subcoriaceous; petioles 1.5-3 em. long: cyme Rich, ex DC. loc. cit., in synon. C. glaberrimus DeCandolle, op. cit. 196; DeCandolle, Mém. Melast. 82 (1828); Grisebach, Fl. Br. W. I. I. 264 (1864). C. corymbosus (L. C. Rich.) Cogn., “var. glaberrimus DC.” Duss, loc. cit., nomen nudum.—Woodlands and cleared slopes at lower and middle elevations, Lesser Antilles. I have seen the following speci- mens. GUADELOUPE: savanes ferugineuses, Lamentin, 1 m., THE GENUS CHARIANTHUS 125 Nov. 18, 1934, H. Stehlé, no. 38 (NY); plaines lateritiques, Capesterre, 100 m., Nov. 8, 1934, H. Stehlé, no. 48 (NY). Dominica: woodland borders at village of Laudat, 665 m., March 8 & 9, 1940, W. H. & B. T. Hodge, no. 1752 (G). Martinique: Sieber, Fl. Mart. no. 120 (G); Bois la Calebasse versant du Morne Rouge, 1878, 1880, Duss, no. 1166 (NY); above L’Alma, is 14, 1938, L. H. & E. Z. Bailey, no. 275 (NY). PLATE 4, FIG. 1; MAP II. Var. latifolius, var. nov., foliis 5-nervatis, ovatis vel late ellipticis, integris, apice abrupte v vel breve acumina atis, basi obtusis ve rotundatis; laminis 6-16 cm. longis, 4-9 cm _ latis, juventute squamis, fuscis, lepidotis, tandem glabrescentibus; petiolis 2-6 cm. longis, setis perpaucibus rare et inaequale; cymis grandis, 6-12 cm. longis, saepe laxe corymbiformibus, pedunculis longis. Rss orymbosus Cogn., var. “grandiflorus Cogn.” Duss, FI. Phan . Ant. Fran., 286 (1897), nomen nudum. C. corymbosus Cogn. ‘var. diffusus Cogn.” Duss. loc. cit. nomen nudum. GUADELOUPE: known only from this island at lower to middle elevations. Abondant a la Calebasse, des hauteurs du Précheur, des Fonds St. Denis, 1879, 1881, Duss, no. 664 (NY); hauteurs du Matouba, alt. 810 m., 1892, 1893, Duss, no. 2250 (TYPE in Gray Herb.; NY, 1soryPe); hauteurs des Trois Rivigres et Trou- aux-Chiens, alt. 320 m., March 1, 1892, 1893, Duss, no. 2267 Ny): Camp Jacob, Cascade de Vauchelet, alt. 460-500 m., 1894, Duss, no. 3204b, (G); same locality, 1896, Duss, no. 3204c (NY); foréts hinds a chemin Soufriére, 500-900 m., Jan. 20, 1935, H. Stehlé, no. 34 (NY). PuaTE 4, FIG. 3; MAP II. Var. longifolius (Cogniaux) stat. nov. Leaves 5-nerved, ovate to ovate-lanceolate, entire, with apex gradually long-acuminate and base narrowly-obtuse to cuneate-attenuate; blades 10-16 cm. long, 4-8 cm. broad; youngest leaves closely brown-lepidote beneath, at length glabrescent, waxy-subcoriaceous and glabrous above; petioles 1.5-7 cm. long, glabrous: cymes many-flowered, generally closely corymbose. C. longifolius Cogniaux in DC. Monogr. vii. 713 (1891). Dominica: known only from this island at middle to higher elevations. In silvis prope Lagunam Roseau, 1000 m., December 1882, Eggers, no. 654, (isotype in Gray Herbarium); Laudat, 1903, Lloyd, no. 331 (NY); Sylvania estate, 1500-1800 ft., June 21, 1933, Cooper, no. 31 (G, NY); same locality, Jan. 27, 1933, Cooper, no. 79 (G, NY); _ on the precipitous slopes of Morne Colla-Anglais, 610-732 m., Sylvania, Aug. 10-23, 1938, Hodge, no. 515, (G, be a cleared forest-land, Sylvania, 549 m., Feb. 16, 1940, W. H. & B. T. Hodge, no. 1310, (G BRI os woodlands about pial estate, 457 m., March 28, 1940, . Hodge, no. 2451 (G); woodlands near base of Morne Concorde Valley, May 6 & 7, 1940, W. H. & B. T. Hodge, no. 3436 (G); 126 HODGE mossy forest at the summit of Morne Nichols, 990 m., March 10, 1940, W. H. & B. T. Hodge, no. 1931 (G, NY); shrubby summit of Morne Megre Maron, near headwaters of Layou River, 700 m., Feb 11, 1940, W. H. & B. T. Hodge, no. 1066 (G, NY); steep, wet, north valley-walls near base of Roseau Valley Waterfalls, below Laudat, 500 m., March 12, 1940, W. H. & B. T. Hodge, no. 2021 (G). PLATE 4, FIG. 2; MAP Il, Charianthus corymbosus and C. coccineus were described by L. C. Richard at Paris from material sent him by LeBlond from Cayenne (French Guiana). Remarking on the LeBlond collection, Richard,’ in his “OBS.1”’ on p. 114, says, “ Pleraeque plantae Gallo-guyannenses: NONNULLAE Martinicensis.” (Capitals are the author’s.) e LeBlond Charianthus types apparently represent some of this Martin- ique material, yet all monographers of the genus have cited LeBlond’s collections as originating either in Cayenne or British Guiana. There are two citations of C. co orymbosus and C. nodosus from Trinidad based solely on Sieber specimens; but, despite the thoroughness of recent Trinidad collectors, no additional material has been accumu- lated from that island. Much of Sieber’s collecting, also, was done on Martinique and it seems, likely that this is a case of mixed labels. Grenada, the southernmost island in the true Lesser Antillean chain, is then the nearest station to South America from which Charianthus has been authentically recorded. Charianthus corymbosus, a close relative of C. coccineus, differs from it in its flowers which are about one half as large; and in its mature leaves which are generally glabrous on both surfaces. C. coccineus is a small shrub of the highest volcanic peaks, whereas the taller C. corymbosus is often a small tree, most abundant at lower and middle elevations. C. corymbosus is a variable species but the variants show a pro- nounced segregation by islands, and thus lend themselves well to treatment as geographical varieties. Richard’s original description states that the leaves are “subcordato-ovatis.”” I have not seen the type-material, but all Martinique specimens examined are ovate with slightly subcordate or emarginate bases. Until an examination of the type proves otherwise, C. corymbosus, as it is represented by all Martinique material, is here considered as typical. Var. typicus is apparently a plant of lower elevations, for all Guadeloupe and Mar- tinique specimens are from such altitudes. Duss, who collected chiefly in Guadeloupe, noted the variability of C. corymbosus. On some of his labels, as well as in his published ? Richard, Act. Soc. Hist. Nat. Paris. (1792). THE GENUS CHARIANTHUS 127 flora, there appear three varietal names without descriptions which are incorrectly attributed to Cogniaux and DeCandolle. Since this Guadeloupe material is sufficiently distinct to warrant its separation, var. latifolius is presented as a substitute for the illegitimate varietal names of Duss. Var. latifolius is easily distinguished from var. typicus by its long-petioled, broadly elliptical leaves with obtusely rounded, rather than subcordate bases. The cymes of var. latifolius are also longer and more loosely corymbose, with flowers which tend to be slightly larger. Examination of the isotype material of Charianthus longifolius Cogn. in the Gray Herbarium shows that it has insufficient characters to separate it as a species, and so I have reduced it to varietal rank. Its flower-characters are identical with those of C. corymbosus varieties typicus and latifolius. From var. typicus it can be separated only by its leaves, which are long-petioled and ovate-lanceolate, with long, gradually acuminate apices, and bases obtuse to attenuate. Base- and apex-characters are together sufficient to distinguish it from var. latifolius, with its broadly rounded bases and abruptly short-acumi- nate apices. Var. longifolius varies considerably: plants of open situations possess the narrowly lanceolate leaf-type while specimens from shaded habitats approach var. latifolius in size but not in shape. The stems of plants from higher elevations are frequently more nearly tetragonal than terete, but this morphological condition can apparently be cor- related with high altitude and exposed mountain summits, for other mossy-forest species of Charianthus, as well as of Miconia, invariably possess strongly angled twigs. The slow growth in such regions of species with a decussate phyllotaxy shortens the internodes, with the result that the quadrangular nodes strongly influence the shape of the internodal surfaces. Certain duplicate specimens of C. corymbosus, cited above and col- lected by Duss, possess herbarium-labels on which the field data varies considerably. Some of his numbers have two collection-dates on the same label; other numbers may show a certain date on one sheet, another date on the duplicate sheet. In all cases I have in- cluded in the citation of specimens all the collection-dates that appear on each label. 3. CHARIANTHUS PURPUREUS D. Don. Shrub, 1-10 m. tall; twigs slender to stout, obtusely tetragonal, the youngest often densely or lightly covered with brown, stiffish hairs or glabrate except for a setose ring at the nodes, the persistent hair-bases often remaining on 128 HODGE the older twigs as tubercles: leaves 5-nerved, broadly ovate to ovate- elliptical or subrotund, with entire or slightly undulate or crenulate margin, at times inrolled, and in young leaves of certain forms sparsely setose-ciliate, very shortly cuspidate, or often abruptly blunt-acu- minate, with base rounded to subcordate; blades 7-16 cm. long, 3-11 em. broad; the youngest leaves remotely scurfy-lepidote above, at length glabrous, below closely brown-lepidote or white-furfuraceous on veins and leaf-tissue, at length glabrescent, the main longitudinal and lateral veins at times sparsely-setose beneath with lax hairs 1-2 mm. long; petioles obscurely sulcate, 1-5 cm. long, the margins densely to sparingly-setose: cymes corymbose, few- to many-flowered, termi- nal, sessile, 3-8 cm. long; the branches, pedicels, hypanthia and calyx- tubes with scattered close or furfuraceous scales: pedicels 1-3 mm. long; hypanthia ca. 3 mm. long: calyx 1.5-2 mm. long, each lobe bearing a blunt often obscure exterior tooth near its middle: corolla bright to deep crimson-purple; petals oval, 7-12 mm. long, 6-7 mm broad at their middle: stamens inflexed in bud; filaments 12-14 mm. long; anthers arcuate, 2-3 mm. long: style red, 1.5-2.5 cm. long, rounded and yellowish at the tip: berry subg!obose, 5-7 mm. in diam., maroon, turning black at maturity. a. Youngest leaves not rugose, the main vein-system beneath, as well as the leaf-tissue, often thickly covered with close, u neath, especially on the youngest leaves; petioles of young leaves setose on channel-margins a ove, not white-scurfy below; setae on the young twigs both nodal and internodal.. .. b. i m. or more long), ovate; longest setae of petioles or twigs well over 2 mm. long; cymes o & R o 3 S < a | c. Petioles and young twigs lightly setose, the hairs fre- uently limited to an adaxial ring and to the upper third of the petiole; veins of the lower surface of young leaves usually devoid of setae; leaf-blades large, up to pex sho: uminate e internodes and petioles; veins of the lower surfaces of young leaves generally setose; leaf-blades smaller, ie 4 to 10 em. long, to 8 em. broa. } luntly cuspidate and often recurved, the margins setose-ciliate, frequently revolute.....°.__. Var. crinitus. b. Leaves small (averaging 6 cm long), subrotund; longest over 1 mm. long; cymes ; a plod cas 2 eo algae ar. brevisetosus. rugose, the main vein-system neath thickly covered with loose, whitish, stellate, por ns a white-scurfy beneath; setae on the yo twi i. €., limited to a narrow ring surrounding the a aca Var. rugosus. THE GENUS CHARIANTHUS 129 typicus. Twigs sparingly setose to glabrous, the hairs long- persistent at the adaxial region just above the nodes: Jeaves ovate to broadly elliptical, entire to slightly undulate, short-acuminate, rounded to emarginate or subcordate at base; blades 8-16 cm. long, 6-11 cm. broad, subcoriaceous; the young leaves often with scattered scales above, at length glabrous, often with closely appressed, brown scales on the veins and leaf-tissue below, at length glabrescent; petioles 2.5-5 cm. at oe cian it on the nea oe above —Charianthus pur pu . Wern. Soc. iv. we (1823); DeCandolle, Prodr. iii. ee hapa gine Fi. Br. W. I. (1864); Triana, Melast. 99 (1871); Cogniaux in DC. Monogr. vii. (1891). Melastoma coccinea Vahl. Eclog. i. 48 (1796); Vahl, Icon. Am. tab. 16 (1799), not Rich.; Willdenow, Sp. Pl. ii. 599 (18 00); Poiret in Lam. Encycl. Meth. Bot., Suppl. iii. 640 (1813). Melastoma Dodan- diana Hamilton, Prodr. Fl. Ind. Occ. 37 (1825).—In forests on the highest volcanic peaks of the northern Lesser Antilles. I have seen the following specimens. St. Kitts: edges of mountain pasture, Molyneux estate, Sept. 8-Oct. 5, 1901, Britton & Cowell, ee Bd (NY), p. p. RRAT: Gages, near the ‘Soufriare, Jan. 23, 7, Shafer, no. 193 (NY); Feb. 12, 1907, Shafer, no. 675 (NY); ane of Chaners Pond, 2800 ft., Jan. 26, 1907, Shafer, no. 676 (NY). Map ut. Var. crinitus Lage stat. nov. Twigs densely setose; the hairs 3-5 mm. long, especially abundant on the internodal area leaves ovate to obovate, entire to minutely undulate, with ee ae margin and subacute to bluntly-cuspidate, often recurved apex; base emarginate to subcordate; blades 6-12 cm. long, 5-9 em. broad, sub- coriaceous above, in you uth with scattered, closely appressed, brown scales, at length glabrous above, glabrescent below; the main veins of younger leaves setose below, at length glabrescent; petioles 1.5—5 cm. long, setose on the channel-mar rgins above.—Charianthus crinitus Naudin in Ann. Sc. Nat. sér. 3, xviii. 112 (1852); Triana, Melast. 99 (1871); Cogniaux in DC. Monogr. vii. 716 (1891); Boldingh, FI. Dutch W. I. I. i. 149 (1909).—In mossy forests on the highest. peaks of the northern Lesser Antilles. I have seen the following specimens. SaBA: mountain, 600-800 m., (1906), Boldingh, no. 1784b (NY). Sr. Kirrs: Mt. Misery, Sept. 8-Oct. 5, 1901, Britton & Cowell, no. Var. brevisetosus, var. nov. Ramulis leviter setosis; setis brevi- bus, ca. 1 mm. longis; foliis late ovatis vel subrotundis, integris vel minute crenulatis, apice obtuse cuspidatis, basi emarginatis vel sub- cordatis; laminis 4—-6.5 em. longis, 2.5-5.5 em. latis, coriaceis, utrinque leviter stellato-lepidotis juventute, tandem _glabrescentibus, subtus m. ] , 3 cm. longis, sessilibus, floribus paucibus. GRENADA: known only from dia island. 130 HODGE Morne au camp, 2400 ft. alt., Dec. 1889, Eggers, no. 6221 (TYPE in Herb. Arnold Arboretum). Puate 5, ries. 5 & 6; Map 11. Var. rugosus, var. nov. Ramulis ad nodos villosis, villis 2-3 mm. longis; foliis ovatis vel ellipticis, integris vel apicem versus crenulatis, valde rugosis; venis subtus squamis copiosis laxis albis_ stellato- furfuraceis persistentibus, petiolis 1-3 em. gis, supra sulcatis marginibus setosis, subtus albo-furfuraceis. Domt1 ndemic, in mossy forests at the summits of the highest peaks. In mossy forest at the summit of Morne Trois Pitons, alt. 1400 m., rainfall 762 cm., Feb. 23, 1940, W. H. & B. T. Hodge, no. 1421 (rype in Gray Herb., IsoTYPE NY); same locality, Aug. 15, 1938, Hodge, no. 509 (G, NY); rain-forest borders betweeen Laudat and Freshwater Lake, alt. 665 m., rainfall 902 cm., March 8 & 9, 1940, W. H. & B. T. Hodge, no. 1751 (G), a low altitude form; mossy forest on summit of Morne Plat Pays, 981 m., March 3, 1940, W. H. & B. T. Hodge, no. 1685 (G); rain- forest borders, Freshwater Lake, 665 m., March 8 & 9, 1940, W. H. & B. T. Hodge, no. 1850, (G). Piave 5, rics. 3 & 4; MAP III. Charianthus in the Lesser Antilles is represented principally by two species, one of which, C. corymbosus, is glabrous, while the other, C. purpureus, is distinctly setose. The varieties of the latter species, like those of C. corymbosus, are indistinguishable from each other in floral characters, but show well marked differences in size and shape of leaves, as well as in type and disposition of vestiture. Such differ- ences, I believe to be varietal rather than specific. Charianthus purpureus was first described and illustrated by Vahl from material collected by Ryan on the island of Montserrat. His description follows fairly well the specimens examined by me from that island, and on these I have based var. typicus. Naudin, from a plant (locality unknown) in the Bonpland herbarium, described C. crinitus—apparently without having seen material of C. purpureus, for he places the latter in a list of species “non omnino certae.” That the degree of villosity is variable was recorded by Ryan on the labels which Vahl later examined. It appears that specimens from the higher more exposed peaks generally possess heavily setose twigs and petioles; leaf-size is also reduced in such situations. These characters apparently can be correlated with habitat—altitude, exposure to constant wind-action, and high precipitation. For the present it seems proper to separate this crinite, mossy-forest form, not as 4 species, but as var. crinitus, while the broad-bladed, sparingly setose, lower-altitude form may be considered var. typicus. Future collec- tions may indicate, however, that these two varieties should be merged. THE GENUS CHARIANTHUS 131 The remaining varieties of C. purpureus show a definite geographi- cal segregation; var. brevisetosus is at present known only from Grenada—the most southern of the Lesser Antilles, while var. rugosus is represented only by Dominica collections. I have seen but a single specimen of var. brevisetosus, namely the type, Eggers, no. 6221, which has well marked differences warranting its varietal separation. Because of its very short (1 mm. long) sparse setae, the short, few- flowered cymes, and its small, distinctly subrotund leaves, it does not resemble any of its more northern relatives. Var. rugosus was first collected by the writer while prosecuting field-work in Dominica in 1938. In 1940, further, more abundant collections were made, showing that this variety is very distinct from its relatives in several vegetative characters. It is, therefore, recog- nized as a geographical segregant. Var. rugosus possesses the leaf-size and habit of var. crinitus, but the presence of abundant white scurf on the veins beneath, and the total lack of laminal setae on the strongly rugose leaves easily distinguish var. rugosus from its closest relatives. . CHARIANTHUS Nobosus (Desr.) Triana. Shrub, up to 2 m. tall; branches thick, tortuous, terete, nodose; the youngest twigs villose with rusty-colored setae, 2-4 mm. long: leaves borne at the ends of the branches, 5-nerved, coriaceous, bead to subrotund, with margins strigose-serrulate with setae curved towards the apex, obtuse to rounded, with ‘base obtuse to rounded or subcordate; blades 3-7 cm. long, 2-5.5 em. broad, glabrous above, punctate below with sunken, appressed scales; petiole 2-10 mm. long, glabrous: cymes pendulous, axillary on the old, leafless twigs, 2-4 cm. long, on slender peduncles, many-flowered; pedicels slender, 6-8 mm. long: hypanthium plus calyx-tube 7-8 mm. long, the tube short-lobed; the minute, exterior tooth near the tip of each triangular lobe extended into a seta ca. 1 mm. long: corolla red to purple, with petals 12-14 mm. long: anthers ca. 2 mm. long: style 2 cm. long: berries 4 mm. me sien at maturity.— Triana, Melast. 99 (1871); Cogniaux in DC . vii. 716 (1891), in part; Duss in Fl. Phan. Ant. Fran. 285 on *Willia ams & Chees- man, i Trin. & Tob. i. pt. 6, 367 (1934). Melastoma nodosa Desrous- se n Lam. Encycl. Mé th. Bot. iv. 55 (1796); DeCandolle, Prodr. iii. "302 (1828): Melastoma aculeata Presl. in Isis xvi. 272 (1828). Chari- anthus ciliatus DeCandolle, Prodr. iii. 197 (1828), and Mém. Melast. 83 (1828). Tetrazygos ciliatus Richard ex. DC. Prodr. iii. 197 (1828), in synon.—MaRTINIQUE: region 0 Paseo de la Montagne Pelée, August 1899, Mtoe no. 666 & 4044 (NY); Duss, no. 4044 (NY). PLATE G. 3 3 Duss’ cue caey een ncaa possess, as in this case, two 5 at ae and often two erent ; since he often collected during the same year on both Guadeloupe and somes it is a problem to know which ke a and “hich date is correct. loz HODGE Forma crinitus, (Duss) stat. nov., differs from the preceding in the abundant, dense, villose setae on the young growth, i. e., stem Charianthus nodosus (Desr.) Triana, var. crinitus, incorrectly attributed by Duss to Naudin in Fl. Phan. Ant. Fran. 286 (1897). Melastoma crinita Vahl, Eclog. iii. 28 (1807); DeCandolle, Prodr. iii. 199 (1828); Charianthus ciliatus Naudin in Ann. Sc. Nat. sér. 3, xviii. 112 (1852). I have seen the following specimens. Martinique: Montagne Pelée, m., Duss, no. 665 (TyPE in Herb. N. Y. Bot. Gard.); sylve rabougrie d’altitude 1300 m., Piton du Carbet, Morne Nert, Sept. 5, 1937, H. Stehlé, no. 2205 (NY). Pare 6, Fic. 4. This is another species of the genus the distribution of which, as cited in the Cogniaux monograph, is probably incorrect. I have seen only Martinique material, but according to Cogniaux this species also occurs on Guadeloupe (Purdie), Montserrat (Ryan), St. Chris- topher (Masson), Trinidad (Sieber) and British Guiana (LeBlond). It is possible that some of the Lesser Antillean citations are based on misidentifications of C. purpureus. Vahl, for instance, records only Charianthus purpureus from Ryan’s collections on Montserrat; and Duss, who spent many resident-years collecting both in Guadeloupe and Martinique, writes concerning C. nodosus, “Cette belle espéce avec sa variété n’existent pas 4 la Guadeloupe.” Sieber nos. 113 and 279, on which the Trinidad citation is based, are probably specimens from Martinique (see discussion under C. corymbosus). Desrousseaux, describing the type of C. nodosus in Lamarck’s Encyclopedie, states, “cette plante est originaire des Antilles, et se trouve au no. 93 de l’herbiere de Surian.” Triana, apparently citing the same specimen, lists it in error as no. 733. Cogniaux refers to the Surian herbarium-specimen, but gives no number; and, in addition, he cites, for the first time, LeBlond no. 93 from British Guiana. Desrousseaux probably did not realize that his type represented one of the three species of Charianthus (C. coccineus, C. corymbosus, and C. nodosus) collected by LeBlond on Martinique and that it was sent, mixed with British Guiana collections, to Richard at Paris. p- parently the material of C. nodosus was overlooked by Richard and so was described later by Desrousseaux as “ originaire des Antilles.” It is apparent, from Desrousseaux’s original description, that the ype-specimen of Charianthus nodosus (not seen) represents the glabrous-leaved form. Forma crinitus, growing intermixed with material of typical C. Contris. Gray Hers. CXXXV. PLATE 4 CHARIANTHUS CORYMBOSUS, Var. TYPICUS: FI , leafy r twig, “x 4 }. CORYMBOSUS, Var. LONGIFOLIUS: FIG. 2, Howe ring twig, LZ C. CORYMBOSUS, Var, LATIFOLIUS: FIG. 3, leafy twig, x lM. ContTrRIB. GRAY HERB. CX XXYV. PLATE 5 oe Wise. we ies te CHARIANTHUS PU RE UREUS, var. CRINITUS: FIG. 1, lower surface of young leaf and flowers, x 13 FI 2, twig and ates showing long crinite hairs, 1} PU ws UREI UGOSUS: FIG. 3, owering twig of plant growing at type lo- Ci ality. eo lower surface of young leaf, from ISOTYPE . PURPURE ‘ar. BREVISETOSUS: FIG. ! hue ring twig, x 16, from TYPE. FIG. 9, twig and ie biter s Seba short stiff setae, x 1X, from TYPE Contris. Gray Hers. CXXXYV. PLATE 6 CHARIANTHUS COCCINEUS: FIG. 1, flowering twig, X } C. COCCINEUS, Var. PARVIFOLIUS FIG, . 2, flowering ae «1 14, from ISOTYPE. NODOSUs: FIG. 3, leafy twig, x1 C. novos us, forma CRINITUS: FIG. Ag lea es, X 4. C. Fapyent: ria. 5, flowering twig, xd FIG. 6, fruits, x } y > mo 7 ConTrRIB. GRAY Hers. CX XXV. PLATE 7 HELICONIA ag Al: FIG. 1, plant oe vag peduncled inflorescence ‘cam & B. T. Hodge, no. 1086. Fic. 2. single flower, c: 1 (drawn from Bot. Reg. t. 37 i : H. c Ant AmA: FIG. 3, crimso 2 ‘od aiid ‘ orm showing sessile i nfloresce nce, W H. a B. T. Hodge, no. 2632. Fx bras Ss — form showing habit, W. H. & B. T. ) Hodge, no. 1442, Fa. 5, oe acoei THE GENUS CHARIANTHUS 133 nodosus, possesses dense, villous setae on both leaves and stems, and is thus readily distinguished. This epithet, validly published by Duss as var. crinitus, was incorrectly attributed by him to Naudin, who had previously published a C. crinitus (C. purpureus, var. crinitus). Duss probably had Naudin’s name in mind but misapplied it to C. nodosus. 5. CHARIANTHUS FADYENI (Hook.) Grisebach. Tall shrub or slender tree 2-9 m. tall; twigs slender and terete (stout, thickly-corticate on plants from exposed summits) ; young growth with scattered stellate rounded, or infrequently emarginate; base acute to rounded; blades 3-7 cm. lon , 1.5-4 em. broad; budding leaves densely covered with stellate scales, at length glabrous above, but — below with dark, sunken, appressed scales; petioles 5-16 mm. long, remotely stellate-lepidote i in youth, at length glabrous: esis terminal, 3-9 cm. long, with basal peduncle 1—3.5 em. long, the branches, pedicels, buds, and fruits remotely sllatorepidote: pedicels 1-3 mm. long; hypanthia 2-3 mm. long: calyx-tube 1.5-3 mm. long, cup-like, indis- tinctly lobate with a minute, obscure, ise ‘tooth near the top of each lobe: corolla bright to dark crimson; petals narrowly oblong to spatulate, 10-15 mm. long, 4-5 mm. broad: stamens inflexed in bu filaments pruinose, 10-15 mm. long, narrowly linear to aaseunalacs at their summit; anthers 4—5 mm. long, slightly arcuate, with a single apical pore: style 2-2.5 em. long, rounded at the tip: tgs Try oe nie ule “ 5-6 mm. in diameter.—Grisebach, Fl. 5) Bo PES & 1864); Triana, Melast. 99 (1871); Cogniaux i in DC. Monogr. 2% 713 (1 ioe Fawcett & Rendle, FI. Jam v. 366, t. 135 (1926). Tetra- script in the British Museum. Jamaica: endemic on this island. Alexander s. n. (G, NY); pened Mount near Ewarton, 2600 ft. alt., Aug. 11, 1896, Harris, no. 6512 (NY); same locality, Feb. 15, 1905, Harris no . 8883, (NY); Holly Mount, Mt. Diablo, Aug. 29, 1905, Harris, no. 8994 (NY); Dolphin Head, te ft., May 18, 1906, Harris, no. 9234 (A, NY); Peckham Woods, U per Cisendon, 2500 2 alt., July 7, 1911, Harris, no. 10997 (NY); Sept. 27, 1912, Harrie, os. 11181 & 11190, (NY); Mulgrave, St. no Pied it. alt, ee 14, 1916, Harris, no. 12373 (NY). PLaTE 6, FIG Charianthus Fadyeni is perhaps the most roan species of the genus. Endemic to Jamaica, it stands alone in a section unrelated to the species of the Lesser Antilles. Its slender petals, single, apical- pored anthers with pruinose filaments, and a 2-loculed ovary, at once set it off from its congeners. 134 HODGE 4. THE TYPE SPECIES OF HELICONIA W. H. Hopar (Plate 7) During several seasons of field work on the island of Dominica in the British West Indies I have had excellent opportunity carefully to observe the two species of Heliconia which abound in the well-drenched mountainous interior. One of these species has been known as Heliconia Bihai L., the other H. rutila R. F. Griggs. The two species are very similar in vegetative habit but are readily separable in the field by the shape, color, and size of the inflorescence- bracts as well as by the shape and size of the perianth-members. The color-differences in the inflorescence-bracts are striking for in the species known until now as H. Bihai the hue is invariably a deep crimson-red, frequently almost maroon (or sometimes solid green, OT yellow), whereas in H. rutila the bracts are bright red, nearly scarlet, with yellow margins (variegate). That two species of Heliconia growing on this island—as well as on the other neighboring Lesser Antillean islands—should possess just these readily recognizable color-differences is significant in the interpretation of the type-species: For from this very area (particularly from neighboring Martinique was collected the herbarium-material from which Plumier, Linnaeus, Jacquin and other authors described many of their West Indian species. The collections from which Linnaeus described the type-species, Heliconia Bihai L., Mant. ii. 211. 1771, probably came from this region as R. F. Griggs suggests (Bull. Torr. Bot. Club. xxx. 656. 1903). If we refer to Linnaeus’ first description of this plant, in Sp. Pl. 1043. 1753, under the name Musa Bihai L., we learn that his species had “amplissimis foliis, florum vasculis variegatis.”” Linnaeus listed two varieties under the species, var.6“ . . vasculis coccineis”, and var.y“ . . . vasculis subnigris”; with var. y we need not deal. Linnaeus’ species, Musa Bihai and its variety @ represent, without doubt, the most common Lesser Antillean species of Heliconia. How- ever, Griggs (Bull. Torr. Bot. Club xxx. 655. 1903), believing that Linnaeus’ species, with “ vasculis variegatis”’, was not determinable, rejected it and took up, as his concept of Heliconia Bihai, var. %: “vasculis coccineis”. Var. 8 was figured by Plumier (N. Pl. Am. Gen. 50. 1703 and Am. ed. Burm. 49, pl. 59. 1756) yet, even with these supplementary illustrations, there is little excuse for setting up this variety as true H. Bihai when the succinct and sufficient diagnosis THE TYPE SPECIES OF HELICONIA 135 “vasculis variegatis”’ exists. The latter Linnaean description is even more significant when we realize that there is but one variegate- bracted species of Heliconia inhabiting this geographical area from which the early collections were sent, hence little reason to doubt that the plant now passing as H. rutila is actually the species which Lin- naeus had in mind when describing H. Bihai. Grigg’s conception of H. Bihai as a plant possessing solid red inflorescence-bracts should be replaced by Linnaeus’ own concept of H. Bihai, a plant with varie- gate bracts. Thus to the plant at present known as H. rutila Griggs, should be applied the correct name, H. Bihai L., and for the species until now passing as H. Bihai must be resurrected the next valid name, H. caribaea Lamarck (Encycl. i. 426, 1783), which was based on the solidly red-bracted var. $ figured by Plumier. The revised synonymy of the two species involved in this discussion is as follows: Heticonia Brnat L., Mant. ii. 211 (1771); Lodd. Bot. Cab. pl. 252 (1818); Edwards, Bot. Reg. t. co ed Bihai florum vasculis variegatis, Plumier Nov. Pl. Gen. 50 (1703), in part. Musa Bthai L. Sp. Pl. ii. 1043 (1753) oe ed. 2. ii. 1477 (1763). H. variegata acq. Pl. Rar. Hort. Schoen. i. 25 (1797). H. rutila Griggs in Bull. Torr. Bot. Club xxx. 657 (1903). Bihai rutila (Griggs) Griggs, op. cit. xxxi. 445 (1904). B. Bihai (L.) Griggs, loc. cit., as to source of name only. HELICONIA CARIBAEA Lamarck, Encycl. i. 426 (1783). Bihai nib vasculis coccineis, Plumier Nov. Pl. Am. Gen. 50 (1703) and Am. ed Burm. 49, pl. 59 (1756). Musa Bihat var. @., L. Sp. Pl. 1048 (1753). Heliconia Bihai sensu authors: Jacq. Pl. Rar. Hort. Schoen. i. 24 (1797), L. C. Richard in Nov. es xv. suppl. 22. pl. 10 (1831); Griggs in Bull. Torr. Bot. Club xxx. 657 (1903); non L. H. bori rinquena Griggs, op. cit. 658 (1903). Bihai Bihai sensu Griggs, op. cit. xxxi. 445 (1904), not as to typonym. Heliconia, like many another large succulent tropical genus, is represented in most herbaria by only a few, often poorly made speci- mens with insufficient field-data. It is to the interest of a better knowledge of the genus that field-notes be full and accompanied by ample photographs particularly of the flowering portions, for, as Griggs has pointed out, the color as well as the shape of the inflores- cence-bracts are diagnostic features which are easily lost in shrunken herbarium specimens In Dominica, Heliconia Bihai and H. caribaea are familiar species in shade or sun along the streams and wet ravines of the interior or even becoming gregarious on drenched mountainsides at higher ele- 136 HODGE vations. When, due to the clearing of the surrounding forest, the species are forced to grow in full sunlight, the plants are usually much dwarfed. H. caribaea is perhaps the more common of the two and, in addition, is a more robust plant as far as height (5 m.,), size of leaves and inflorescence-bracts is concerned. In leaf-characters the species are practically inseparable. The leaves of H. caribaea, though somewhat rounded at the tip, always possess an abruptly acute tip while those of true 1. Bihai are more gradually acute. The majority of plants possess leaves which are green on both surfaces, although now and then an individual in a colony will exhibit leaves which are glaucous below—one of the characters on which Griggs based H. borinquena. In H. caribaea the inflorescence-bracts show much variation in color, a condition not observed in H. Bihai. The most common form of H. caribaea is one with deep crimson bracts (not scarlet as described by various authors), but sometimes mature individuals, in full flower, are seen with solid green, solid yellow, or green and yellow inflores- cence-bracts. Since such individuals differ only in the color of their bracts they can be considered simply as forms. In his key to the species confounded with H. Bihai, Griggs uses as a distinctive charac- ter the presence of a “peduncle—long, stiff, erect” —in H. Biha (i. e. H. caribaea) and the absence of such in H. rutila (i. e. H. Bihat). My field observations of these two species, as illustrated in the ac- companying photographs, PLATE 7 (plant-specimens here illustrated are in the Gray Herbarium), show that exactly the opposite holds— true H. Bihai possesses a short peduncle (the lowest inflorescence- bract is separate from the sheathing petioles), while in H. caribaea the inflorescences are always sessile (the lowest inflorescence-bract is in contact with the sheathing petioles). Although occasional plants of each species may be found in flower throughout the year, the greater proportion of Dominica’s Heliconias exhibit a somewhat seasonal anthesis. H. Bihai blooms from February to April, H. caribaea, with more tardy flowering, from April to June; the bright blue fruits appear throughout the summer months. Jac- quin, visiting Martinique a century and a half ago, observed that the ocal species of Heliconia were known as “balisiers.” Today the natives of the patois-speaking islands still use the same name and make use of “ balisier” leaves for thatch and as waterproof lining in baskets. The following key may be used to separate H. Bihai and H. caribaea. All measurements used are from fresh material, now dried, and the herbarium specimens deposited in the Gray Herbarium. THE TYPE SPECIES OF HELICONIA pg gradually acute: inflorescence short-peduncled; the bracts always variegate, scarlet with bract-tips, upper margins ing ner surface bright yellow (see Bot. Reg. t. 374), up to 6 in number, shallowly boat-shaped (each bract ca. 2.5-4 cm. wide), narrowly triangular, widely separate and never over- 1tBOlE on Ci pe es ee i Pe — sey ad acute: pre gneeat sessile; the bracts deep crimson (but sometimes yellow or er een), 6 to 15 in number, deeply hoacehannd (each bract ca. 5-6.5 cm wae. broadly triangular, everapoinn at the ba: —_— and gece 2 ties greenish; the free sepal ca. 5 mm. wide, with somew downcurved margins, "the sepal never F Goned Wack on iteclf 137 ..H. caribaea. INDEX New scientific names are printed in full-face type Acidanthera, 6, 71, 74, 75, 77; bellum, 89, 98, 102, 103, 105, 115, Fourcadei 73, pl. 3;canaliculatum ,89, ,104— 106, 1 4, pik: costaric cen se, 99; Bihai Bihai, 135; florum vasculis ace. 101, 102; Ghiesbreghtii, coccineis, 135; florum vasculis 89, 98, 109, 110, ‘113; Hartwegi- variegatis, 135; rutila, 135 anum, 88, 92, 96, var. amans, 88- 90, 93-95, 97, 112 114, pl. 2, var. Charianthus, sect. one = eccai i typicum m, 88-91, 94, 95, 113, 114, 117-119, 121, sect. Eucharian pl. 2; hispidum, 106; jalise canum, thus, 137 121, 123; Berteroanus, 110, 115, pl. 3; Jo onii, 124; ciliatus, 116, 131, 132; coc- 90, 95, 96, rid, pl. 2; Maxonii, 89, cineus, £16; > 1 7, 119133, 126, 97-100, 115, pl. 3; Mexiae, 111, 132, var. parvifolius, 119, 122, 114, pl. 1; oblongifolium, 84, 85; 123, pl. 6, var. typicus, 121, pl. procumbens, 97, 112, var. longipes, 6; coriaceus, 122, 123; corymbo- 12; usianum, , 98, sus, 117, 119-124, 126, 127, 130, 106; radicans, 81; reptans, 81; 132, var. diffusus, 125, var. gla- Schindleri, 89, 98, 105, 106; berrimus, 124, var. grandiflorus, Schlechtendalii 101, 102; spirale, 125, var. latifolius, 119, 124, Ne stoloniferum, 86; stoloni- 125, 127, pl. 4, var. lo lius, r “rye ge 92; 119, 124, 125, 127, pl 4, var. eubascaile: 8-91, 4, pl. 2; sub- typicus, 119, 124, 126, 127, pl. tile, 102, 108; el i Rg 89, 97, ‘4; crinitus, 116, 129, 130, 133; 98, 100-103; Wydlerianum, 80, Mh ben 116, 117, 121, 133, pl. 6; glaberrimus, 116, 117, "123, 124; Ensatae, 4 longifolius, ; nodosus, Z 116-118, 120, 121, 124, 126, 131— Freesia, 10 Ri wa forma crinitus, oy per apony: subg. Eugeissorhiza, pl. 6, var. crinitus, 132, 133; purpureus, 116, 118-121, 127, sect. Rochea, 18, 39, sar 129-132, var. brevisetosus, 119, ormes, 18, 27, subsec 128, 129, 131, pl. 5, var. crinitus, Ventricosae, 27, sect. Tortuo- 119, 128, 129, 130, 133, pl. 5, var. sa, 69, sect. Weihea, 11, 39, sub- osus, 118, 119, 128, 130, 131, sect. Foliosae, 55, 56, 60, subsect. pl. 5, var. typicus, 119, 128, 129, Pubescentes, 39; subg. Ixiop- 130 - tinifolius, 116, 1 122, 123 sis, 70; abyssinica, 76; alpina, Chaenanthera mucronata, 124 beter ge 12 72, 74 es 37, 38; Bellendeni, 10, 40, 41; Desmodium Alamani, 89, 98, 102, bicolor, 11, ‘ 35, var. 107, 108, 110, 113, 115; amans, M , 33; i, 8, 12, 4; axillare, 79-81, 83, 85, var a7, 10; 74; Bolusii, 16, 57, 5 acutifolium, 80, 83, 85, 86, 88, 89, 61; bracteata, 1 i, 4 , pl. 1, var. acutifolium, forma 38; Briartii, 76; ) 14, robustius, 87, var 8. acutifolium, 18, 25, 73; corrugata, 9, 12, 17, ar. 6. acutifolium, forma 69, 70; Dregei, 4, 9, 16, 56, 58 robustius, 84, var. statum, stair ’ ? ? 55, r r. 8. angustatum, 84, 85, 16, 53; eurystigma, tatum, forma ro- 12, 28; excisa, 11, 39, 59; foliosa, bustiu 4, 85, var. genuinum, 9, 17, "5B; filifolia, 74; flava ‘ 80, 83, 86, 88, 89, st = 1, var. 69; foliosa, 61, 63; Fourcadei, 78; « genuinum, 81, obtusi- furva, 4, 14, 26, 27, 45; geminata, foliola, 86, var. a. obtusifoliola, 17, 63, 64, 78; gracilis, 6, 75, 76, 81, var. Sintenisii, 80, 84, 87-89, 73; grandis, 17, 73; graminifolia, 114, pl. 1, var. y. Sintenisii, 87; 49, 50, var. "bicolor, 54, 55; 139 140 INDEX hesperanthoides, 17, 26, 73; tarts. 16, 44, 46; ee 31, , §2, 55 peesthon Fe 54, 55; humilis 14, 18, 19, 28, 34, 64, 66, 68, sae, 15, 39, var. fonitfotia, 15, 34; lutea, OY gt macra, 6, 75; malmesburi riensis, 17, 66; Marlot mil, 17,. 66: Mathewsii, 12, 14, 27, 28, 38, 43, .e tigma 14, , 28, ‘ , 55; namaquensis, 16 46, 51; nana, 17, 60; obtusata, 4, 13, 37, 38, nod omithogalides, 4 4, 8, , 66-69 i ‘ealie , pusilla, 16, 46, 50; adrangula, 4, 75; quinquangularis, 4, 10, 16, &3, 55, va trofaux, 16, 55; dians, 30; ram 18, 21, 22, 24-26, 56; recurvifolia, 4, 57, By of fae? na, 4, och sis, 4, 14, 27, 29, 30, 40, var. monantha, var. multi- ora, 30, var. pauciflora, 29, va spithamaea, 14, ; romu- leoides, 67; Rogersii, 16, 45-47; » Ll, 16, 48, 49, 55; rubi- eunda, 15, 35; ru is, 9,16, 7, 58 ; Sabulosa, 37, 38; Schlechteri, 77; secun a, 4, 5, a3, 15, 23, 25, 41, 45, 46, 31, 74, v G. pusilla, 50, var. G. pcg 24, var. setifolia, 23, var. G. setifolia, 21; setacea, 4, 9, Li,.21, 22, 60, 64, 5, setacea B., 19: setifolia, 21-23, 25, 78; splendi- dissima, 10, 14, 75; sublutea, 4; oak gery: 9, 16, 47; sulphura- Dig 39; sulphurea, 15, 36, arenicola, 15, 37; teretifolia, 78; tulbaghensis, 15, 40; ae era, 29, 30; mbrosa, 78: Vv vagina ta, 76, 77; SOR Be 10, 17, 61; Wrightii 15, 35 Gladiolus, 8, 71; dichotomus, 72; junceus, re 43; permeabilis, 72; subaphyllus, 78 Hedysarum, 112, 113; Alamani, 108; axillare, 81; frutescens, 113; longifolium, minus, 113; obinaata lium, 84; reptans, 81, ea stoloniferum ot 80; irk orum Heliconia, 134, he Bihai, 134-136, pl. rinque 6; caribaea, 135, 136, - 7 Pala. 134-136; variegata isoerce. 38, 12, 25, 39, 49, 53, 55, 61, 68-70, re 75, 77, 78; ciliata, 49, 55; ‘infle xa, 17; insipida, 78; kermesina, 53; lati- folia, 59; lon ngituba, 77; lutea, 77; Mete lerkampiae, 77; ma, 77; ora, 77; Benther, 74,78; @ nitciaaienie tt S, radiata, 49; rosea, 49; ee ikiet 77 2 a sl 3, 4, 34 Ixia, 4-9, 44, 76; subg. Morphixia, Sy: Jor 76; cisa, 59; geminata, 63, 64; — 31, 52, 53; Raitt 18, 19; bricata, 37: inflexa, 52, 53, 76, 77; juncea, 21, 22; La R hei latifolia, ornithogaloides, 4, 67, 68; 59; phalangioides, 21; pusilla, 43, 50, 51; quadrangula, 75; page 30, 31; ramosa = wis ear gg 4, 77; ad sani var. palma ris, 29, var. i , 40, 42, 43; setacea, 64, 65; splendidissima, 31; tortuosa, 70: violacea, 30 es, 7, 76; abyssinica, 76; avasmontana, 76; coerulea, 76; Welwitschii, 76 Meibomia, 101, 106; Alamani, 108; INDEX 141 huronoe 102; albida, 87, 88; Musa we 134, 135, var. B., 134, , 94; andin na, 81, 83; axil- 135, v , 134 Soria 81, 86, var obtu ifoliola, 81, 82, var. B. acutifolia, 84, 86; Ne a subg. Swartziella, 79; bella, 103, 110; Blakeana, 97, 98; axillaris, 81, 86, var. 8. acuti folia, costaricensis, 99, 100; elegans, 97, : 5, v 6. acutifolia, forma 101, ; Ghi esbreghtii, 109 robust acutifoliola, Hartwegiana, 92; hispida, 97, 84, 85, var gy eo song var 106; jaliscana, 110; Maxonii, 99, a. obtusifoliola, 81, 83, : ee. ;_Purpusi- Sintenisii, ana, 107; rep 81; Sintenisii, 87; sahesaatas ‘OL: subti lis, 102, Rochea, 13; venusta, 29 108; umbrosa, 84 86; venusta, Romulea, 4, 70, 77 101, Melastoma aculeata, 131; Berteri- Sisyrinchium, 3 anum, 124; coccinea, ; corymbosa, 1 16, 124; crinita, 132; Tetrazygia, 115, 116; Fadyeni, 116, Dodandiana, 129; nodosa, 11 6, 117, 133 31; splendens, 124 mnetralyaee ciliatus, 131 Melastomaceae, 115-117; tribe Mi- Tritonia, 41; latifolia, 77; pauci- conieae, 115 ora, esembrianthemum, 34 Miconia, 115, 116, 127 Weihea, 13; excisa, 59 Reprinted from Ruovora, Vol. 43, A CONTRIBUTIONS FROM THE GRAY rages 1 OF HARVARD UNIVERSITY. CXXXVI EUPHORBIA SUBGENUS. CHAMAESYCE IN CANADA AND UNITED STATE EXCLUSIVE OF SOUTHERN FLO. Pap * Reprinted from Ruopora, Vol. 43, April—June, 1941 CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY CXXXVI EUPHORBIA SUBGENUS CHAMAESYCE IN CANADA AND THE UNITED STATES EXCLUSIVE OF SOUTHERN FLORIDA Louts CuTTER WHEELER Dares or Issuz Pages 97-154 and Plates 654-656..................-. sues, . April, 1941 Pages 168-205 and Plates 657-664.........................9 9 May, 194 Pages 220-286 and Plates 665-668... 2... ieee eee 13 Fane: 1941 CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY—NO. CXXXVI EUPHORBIA SUBGENUS CHAMAESYCE IN CANADA AND THE UNITED STATES EXCLUSIVE OF SOUTHERN FLORIDA Louis CutTER WHEELER (Plates 654-668) INTRODUCTION From an early curiosity about the peculiar ‘“flower’”’ of Eu- phorbia, the writer’s interest developed into a serious study of the genus, which soon led to the realization that not only was the structure of the “flower” involved but that the classification of the genus was in a remarkable state of confusion. Boissier in DC. Prod. 15 (2): 3-188. 1862, published the first revision of the entire genus since Linnaeus’ time. Since 1862 the number of species of Euphorbia published has about doubled. Whereas Boissier recognized about seven hundred species as valid, Pax & Hoffmann in Engler & Prantl, Nat. Pflanzenfam. 2 Aufl., 19¢.: 209. 1931, guess that there are about one thousand six hundred species. With this great increase since an organized treatment of the genus the need of taxonomic revision is evident. I. GENERAL DiscussION History The following remarks apply only to the subgenus Chamaesyce except where expressly stated to be of broader scope. Wiman, in Linnaeus, Amoen. Acad. 3: 102. 1756, lists Chamaesyce as a 98 Rhodora [APRIL synonym of Euphorbia. S. F. Gray, Nat. Arr. Brit. Pl. 2: 260. 1821, in describing Chamaesyce as a genus, credits it to Dios- corides. Thus it is evident, if Gray’s interpretation was correct, that the group was recognized in antiquity. Persoon, Syn. Pl. 2: 12. 1806,' used Chamaesyce as a synonym of a defined but unnamed subdivision, of subgeneric rank, of Euphorbia. Rafin- esque, Amer. Mo. Mag. 2: 119. 1817, independently published Chamaesyce as a subgenus. Haworth, Syn. Pl. Suce., 159. 1812, proposed the genus Anisophyllum for the same entity but the name was preoccupied. Quite independently, Duby in A. P. DeCandolle, Botanicon Gallicum ed. 2, 1: 412. 1828, took up the formerly unpublished Euphorbia section Anisophyllum of Roeper. Authors who have assumed that nomenclaturally this was based on Anisophyllum Haworth would do well to note this point. This sort of confusion of the independent use of the same name in either the same or different categories by different authors is common under Euphorbia and the genera segregated from it. This is due to the fact that many of these names are either traditional, as Chamaesyce, or suggested by some striking characteristic of the group, as Anisophyllum. For the more de- tailed history of the uses of the name applicable to this subgenus see beyond under the taxonomic section. The workers who have made the principal contribution to our knowledge of Euphorbia subgenus Chamaesyce in the area under consideration are Boissier, George Engelmann, Millspaugh, and Small. Boissier revised the entire genus Euphorbia in 1862. Engelmann described many of the species discovered in the United States. Millspaugh’s contribution was mainly new species, a multitude of new combinations, and the accumulation of invaluable fragments of types. His work is notable for the number of worthless new species that he proposed. Small either contributed the actual text for several of the principal floras ° the region, or exerted a strong influence, as in the case of Ryd- berg’s works. My own publications relating wholly or partly to Euphorbia subgenus Chamaesyce are listed in the bibliography at the end of this section (pp. 106 and 107). ‘As noted by Blake, Ruopora 17: 134, footnote 1. 1915, pages 1-277 of —_ tarum secunda”’ of Persoon’s Synopsis Plan were by published in 1806, as shown review of these pages in Botanische Zeitung, Regensburg 5: 321. 1806 (Nov. 15)- 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 99 MorPHOLOGY Roots: The lateral roots are ordinarily slender and branching in the manner common in small annual and perennial herbs. Occasionally the taproots become thickened and even subfarina- ceous. Certain of the longer-lived perennials often develop buds well beneath the surface of the ground and the stems arise beneath the ground and come to the surface more or less scattered instead of originating from a common point at the surface. TEMS: The stems ordinarily have an interrupted main axis. After the first pair of primary leaves the main axis ceases to elongate. Lateral branches arise from the apex without any particular relation to the leaves. In species developing a single erect stem one branch only arises or, if more than one, then one soon dominates. If this stem attains any appreciable diameter the interruption of the main axis becomes completely concealed in the manner illustrated by Croizat in Degener, Fl. Hawaii, fam. 190, Chamaesycez-3. 1937. Whether this type of axial growth obtains in Zuphorbia gracillima and revoluta is uncertain from her- barium specimens. The origin of this type of branching was first suggested by Roeper, Enum. Euph., 30. 1824. He suggested that the branches arising from the crown of the main stem in subgenus Chamaesyce were homologous with the rays of the pleiochasium of Euphorbia subgenus Hsula Pers. (Euphorbia sect. Tithymalus). This seems to be a very reasonable explanation. It is only necessary to add that, whereas the main axis in subg. Esula is terminated by a cyathium, in subg. Chamaesyce it is not. This would seem to be a fundamental difference but it is not. When by progressive reduction of the main axis subg. Chamaesyce finally arrived at the habit of branching after the first pair of true leaves appeared, the plant was obviously too small to produce all the elaborated food necessary for the production of a cyathium with its repro- ductive structures requiring abundant protein, fats, and carbo- hydrates; so we find that the cyathium which would otherwise terminate the main axis is omitted. The often apparently lateral type of branching of the stems of subgenus Chamaesyce appears to differ from the plainly dichoto- mous or sometimes trichotomous type of the rays of subgenus Esula. But Roeper was equal to this difficulty. While poten- 100 Rhodora [APRIL tially dichotomous, the branching in subg. Chamaesyce often, by reduction, becomes apparently ordinary lateral branching though it is really sympodial as shown by the position of the cyathia. But in species with this type of branching the stem-tips often exhibit symmetrical dichotomy with actually terminal cyathia. All the cyathia are morphologically terminal even though they appear axillary when the suppressed branch of the scorpioid sympodium is absent. Collateral branching, i. e., subsidiary branching in addition to the main and more or less suppressed branches of the dichotomy, is common. Puate 655B, rigs. 1-4, shows a possible reductional series from a member of either Euphorbia subg. Esula or Agaloma. Fig. 1 represents the ancestral type with stem-leaves alternate between the whorled leaves subtending the terminal pleiochas- ium and the first pair of true leaves. Both E. Peplus L. (subg. Esula), and E. marginata Pursh (subg. Agaloma) have this habit. The cotyledons are represented by the lowermost pair of lateral outgrowths of the stem. Fic. 2 shows the apparent type of axial growth of E. gracillima 8. Wats. and E. revoluta Engelm. (subg. Chamaesyce). Whether this is actually the type of growth ob- taining in these two species awaits confirmatory study of their seedlings. In any event, the type of axial growth shown in FIG. 2 represents a plausible intermediate between Fic. 1 and the type commonly found in subg. Chamaesyce. Fic. 3 is a diagrammatic representation of a branch of E. maculata L. showing the easy transition from a sympodium below to a subsymmetrical dichot- omy at the apex. Fic. 4 shows a branch of the sort commonly found in members of subg. Chamaesyce, e. g. E. supina i. Here is shown the gradual suppression of alternate branches of the dichotomy until there is derived a scorpioid sympodium simu lating an ordinary monopodium, In prostrate species such “ E. supina the branches radiate from the sides of the short main axis. Whereas in Fig. 2 the distance from the cotyledons to the first pair of true leaves is represented as about 1 cm., in more reduced species, as EH. supina, it is of the order of a millimeter. Conclusion: The branches of members of subg. Chamaesyce are homologous with the pleiochasial rays of members of subgenera Esula and Agaloma. Whether subgenera Agaloma and Chamaesyce are independent 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 101 developments from subgenus Esula, or subg. Esula gave rise to subg. Agaloma, and that in turn to subg. Chamaesyce, or subg. Agaloma gave rise to two independent lines represented by sub- genera Hsula and Chamaesyce, is an obscure question which will occupy the attention of phylogenetic speculators for many un- profitable hours. Some members of Euphorbia subg. Agaloma, e. g. EH. marginata Pursh and £. corollata L., have a pleiochasial inflorescence, and similar species, rather than any members of subg. Hsula, may be the ancestors of subg. Chamaesyce, though the occurrence of subg. Agaloma only in the New World, as con- trasted with the occurrence of subg. Chamaesyce on all continents, suggests that such ancestry is improbable. Leaves: The leaves are simple, thin to thick. The margin is entire to variously toothed. The base is usually markedly in- equilateral. The leaves, when sufficiently thin, exhibit a curiously mottled appearance when viewed by transmitted light. The venation is reticulate and the chlorophylliferous cells, instead of being distributed in the usual manner, are around the veinlets, thus leaving clear areas between. Veh, Ann. Jard. Buit. 38: T. XV, figs. 32-34. 1928, illustrates this. Strputes: The stipules are small, variously membranous, lacerate, connate, or distinct. VESTITURE: The trichomes, which constitute the only ves- titure, are always simple. Most of them, unless very short, are multicellular, with the cells end to end. Occasionally, as in Euphorbia hirta, the distinction in size between hairs and capil- lary multiseriate segments disappears. Some of the divisions of the involucral lobes in E. hirta are no thicker than the trichomes on the lobes. The walls of the hairs are usually microscopically rugulose. In a few species such as E. arizonica and setiloba the cell-wall is quite smooth. Cyaruia: The cyathia are compound inflorescences simulating simple flowers yet consisting of few to numerous staminate flowers and a central pistillate flower, these surrounded by a cuplike involucre composed of commonly five foliar structures united by their margins. Alternating with the tips (lobes of the involucre) of these modified leaves are nectariferous glands. Each gland is supplied by two traces, one from each of the ad- jacent modified leaves. The glands often bear petaloid append- 102 Rhodora [APRIL ages. Usually one of the glands is missing and the pistillate pedicel is commonly deflexed into the interval thus created. Piate 655B, Fics. 5 and 6, illustrate the structure of the cyath- ium and involucre. The voluminous literature relating to the interpretation of the cyathium can be found by examining the bibliographies given by Pax 1884, Schmidt 1906, Denis 1921, Haber 1925, Bodmann 1937, and Schoute 1937. The staminate flowers are pedicellate, naked and monandrous. Each staminate flower is primitively subtended by a bracteole. These bracteoles may be entirely free but are generally more oF less connate and often adnate below to the involucre. The staminate pedicels are termed androphores in this paper. : The pistillate flowers are pedicellate and naked or with a rudi- mentary 3-lobed calyx. The pistillate pedicels are termed gyn0- phores in this paper. The ovary is 3-locular and sessile. The styles are usually bifid. The structure of the dehiscent 3-locular capsule is illustrated in PLATE 655B, ria. 7. The embryo and endosperm are surrounded first by a light tan to nearly black structure here called the testa. Outside the testa is a layer of dried mucilage which is white and, depending 0D the thickness, makes the seed more or less white by obscuring the testa beneath. This dried mucilage is here called the coat. Pammel, 1891, has considered the structures surrounding the embryo and endosperm in the seeds of Euphorbia. RELATIONSHIPS OF SUBGENUS CHAMAESYCE The subgenus Chamaesyce has probably, as suggested in the discussion above under stems, been derived from either subgenus Esula or Agaloma. It is with the greatest difficulty that, aside from the supposedly constant difference in the development of the main axis, subgenus Chamaesyce can be defined so as to &* clude all members of those subgenera and at the same tume include all members of subgenus Chamaesyce. GrocraPnicaL DisrriBuTION oF THE SUBGENUS CHAMAESYCP Subgenus Chamaesyce occurs native in the warmer parts of all the continents including Australia. It reaches its greatest development in subtropical regions. It has attained a remark- able development in Hawaii where some species are arborescent- 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 103 An analysis of the geographical distribution of Euphorbia sub- genus Chamaesyce in Canada and the United States exclusive of southern Florida has yielded the following data: Of the 48 species occurring in this area, 17 are originally endemic though 3 are now established in other parts of the world. Of the 17, 11 are found only west of the Mississippi River, 1 only east, and 5 on both sides. Of the 31 species not native solely within the specified area, 21 are native in this area plus Mexico. Few of these 21 range south of the northern half of Mexico. Of the 10 species not included in the two preceding categories, 3, viz. E. glomerifera, hirta, and hyssopifolia, are predominantly subtropi- cal or tropical, and, within the area under consideration, are found native only in southern U. S. Most of the remainder are species with wide ranges in North and South America. While no introduced species has become established in the United States or Canada there have been local introductions of species native within the area. Notable examples are E. maculata and supina. On the other hand, some of the species native solely within the area have been introduced into Europe and elsewhere, e. g. E. polygonifolia and supina. In order to determine what region had the largest number of species the distributional maps were examined to ascertain where a circle with a fifty-mile radius would include the largest number of species. Such a circle centering at Marsh, Pima County, Arizona, includes 25 species. Of this number probably only £. supina is introduced. This area is fairly well known botanically and it is not likely that there will be many additional species discovered in it. Curiously enough no species is endemic within the area circumscribed by this circle. A second center of density is Elephant Mountain, Brewster County, Texas. Within a fifty- mile radius of this point 21 species occur. Probably all of these are native. In view of the fact that this region is not well known botanically and that there are several more species in adjacent regions, it is highly probable that additional species will be found here. Two species and one very distinctive variety are endemic in this region. It may be protested that both of these centers are merely points adjacent to the Mexican deserts and owe their density to & greater density southward. If an equal area is chosen any- 104 Rhodora [APRIL where in Sonora it is not likely to surpass the Arizonan center much if at all since several species which are common in Arizona are absent in Sonora. Similarly, several species have their south- ern limit included in the Big Bend area of Texas and the loss of these and the species endemic in the Big Bend area would scarce- ly be offset by species occurring in Chihuahua or Coahuila but not in Texas. Probably a third center of density is in southern Florida but determination of that must await examination of all the species found there. Economic VALUE The members of Euphorbia subgenus Chamaesyce are generally of little economic value. The species are often weedy but are rarely if ever classed as noxious weeds. Occasionally stock are supposed to be poisoned by accidentally eating some of these plants mixed with other herbage. Ordinarily not even grass- hoppers will eat these plants. One Kansan correspondent a formed me in 1938 that grasshoppers ate nearly everything except the spurges! Some use is made of Euphorbia hirta as @ drug plant. Aboriginal peoples often utilize the latex in their medicine. There are persistent reports that these plants are remedies for snake bite. H. M. Hall, Yosemite Flora, 151. 1912, mentions that E. serpyllifolia is often used for this purpose by “Indians and others”. C. R. Orcutt, 1890, also discusses the supposed virtues of these plants. No serious study of the alleg- edly theriacal qualities of the group seems ever to have been made. It has been stated by Standley, Field Mus. Pub. Bot. 3: 1930, and doubtless others, that Euphorbia hirta harbors the organisms causing tropical leg-ulcers. Dr. J. C. Bequaert of the Harvard Medical School assures me (in conversation) that there is 2° proof of this. Nevertheless I consider it entirely possible. Flagel- lates are common in many herbaceous species of Euphorbia and are by no means confined to even subtropical regions, for they have been found in Europe. No study seems to have been made of the species occurring in the United States, to determine Whether they, too, harbor these organisms. The literature Te lating to flagellates in Euphorbia and other laticiferous plants be voluminous. Mesnil, Ann. Sci. Nat. ser. 10, Bot. 3: xlii-lvil. 1921, gives an interesting resumé up to that date. 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 105 ACKNOWLEDGMENTS The completion of the work on this problem has been made possible by the cooperation of various persons to whom the author wishes to extend his thanks: Professor M. L. Fernald, Director of the Gray Herbarium, under whose direction the study was prosecuted; Mr. C. A. Weatherby, Senior Curator of the Gray Herbarium, who corrected the Latin descriptions and rendered opinions in matters nomenclatorial; Miss Rut Sanderson, Librarian of the Gray Herbarium, whose cheerful assistance in bibliographical matters has been constantly helpful; Mr. R. K. Godfrey, who, while at the Gray Herbarium, copied and transmitted essential data; the curators of the herbaria of the following institutions: Dr. William R. Maxon, United States National Herbarium; Dr. H. A. Gleason, New York Botanical Garden; Dr. P. C. Standley, Field Museum, Dr. J. M. Greenman, Missouri Botanical Garden; and Dr. F. W. Pennell, Academy of Natural Sciences, Philadelphia, for permitting me to visit their institutions and examine their collections. The author’s wife has assisted in many ways throughout the study. The drawings were made under the writer’s direction mainly by Mr. G. W. Dillon with a few by Mrs. Frances M. Fay and Miss Elsie Herbold. The drawings were made possible by funds from various sources; among them were the Department of Biology, Harvard University and the Department of Botany, University of Mis- souri. A grant in aid of research from the Research Council of the University of Missouri made available essential clerical assistance. In addition to those listed above, the writer is indebted to the curators of all the herbaria listed under abbreviations for making the specimens at their disposal available for study. BIBLIOGRAPHY Battion, H. 1858. Etude Générale du groupe des Euphorbiacées. were G. 1880. Notes on Euphorbiaceae. Journ. Linn. Soe. Bot. 17: BopMann, HELENr. 1937. Zur'Morphologie igh ee von Euphorbia, eet rT. , Ba. aa poy 5 IssteR, E. 1862. Euphorbieae in DC 5 (2): 3 ear L. 1936. On the classification of fig ceo T ie important orr. u : 4 Glands of the Euphorbiaceae and of Euphorbia. Chron. Bot. 4: 512-514 106 Rhodora [ APRIL Decener, O. & Croizat, L. 1936. Chamaesyce in Degener, Fl. Hawaii, fam. 190, Chamaesyce; &4 eee TOT. Chamacsyce ; in Degener, Fl. Hawaii, fam. 190, Chamae- €2 & x, fi Daksa M. 1931. Les Euphorbiées des Iles Australes d’ Afrique. GaucuER, L. 1898. Etude anatomi mique du genre Euphorbia L. E a, 99. Etude anatomique des glandes du cyathium des ae phorbes et de leurs seep colorantes. Journ. Bot. Paris 13: Del e des laticifares. ee Sci. Nat. (Bot.) 241-260. Gray, 8. F. 1821. Natural rirpmniagi pec of British plants. 2: 255-260. P Haber, J. M. 1925. The Anat omy and the Morphology of the Flower 0 Euphorbia. Annals of Botany 39: 657-707. Haworts, A. as gio She ad cap dae succulentarum, 126-164. a MEsnIL, F. 192 La “flagellose”’ “‘Leptomoniase”’ des une e Asclé dices hits. Sci. hae. Pa aris. sér. 10 Bot. 3: xlii— tt CUTT, & - 1890. The Golondrina Plant. West pedenie ‘Scientia PaMMEL, L. H. 1896. On the seed coats - the genus Euphorbia. Trans. Acad. Sci. St. Louis 5: 543-568, Pl. 12-1 an , F. 1884. Die Anatomie der Eu “ods ansioe in ihrer Beziehung 2 System derselben. Engler’s Bot. Jahrb. 5: 384-421, Taf. VI-VII. aia ———.. 18 Euphorbieae in cr ’& Prantl, Nat. Pflanzen 3 (5): 102-112. ———— & Horrmann, K. 1931. Euphorbieae in Engler & Prantl, Nat. Pflanzenfam. 2 Aufl. 19¢: 207-233. um PERs oon, C. H. 1806. Synopsis plantarum seu enchiridium botanicum, RaFINIsquE, C. 8. a8} 7 Second decade of undescribed American plants. mer, Monthly M ag. RoEPER, J. 1824. Enumeratio Euphorbiarum quae in Germania et Pannonia gignuntur. Scumipr, Henri H Lupwic. 1906. Uber die Entwicklung rs ryt: yy lite atinds von Euphorbia L. und Diplocyathium n. 21-69, gs. I-VI, Taf. I-IV; preprint dents Beih. Bot. Centralbi, 22: figs. I-VI, Taf. I-IV. 1907. horbia Scuoure, J.C. 1937. On the Aestivation in the Cyathium of Eup f the fulgens, with some remarks on the morphological interpretene” : cyathium in general. ueil Travaux Bot. Néerl. 24: 168- 4: 30-32. ; - a. Cyathium glands of Euphorbia. “Chron, aang shorbia. -_ Reply to the above remarks by Croizat on Bp Chiesa. ron, ba 4: 514-51 Arten be pret A. 1907. Die n Europa bis jetzt si meio gt Euphorbia- tion “Anisophyiln Bull. Herb 2, T: 741 Hage hor: Veu, Ronni v 1928. Beitrag zur ewig ace Anisophy llum- see Hiche einige vergleichende und _ entwicklungsges¢ neg mee Untersuchungen iiber die Driise des Cyathiums. Ann. Jar a Bot. “100, 1. & - : # WHEELER, L. C. 1934a, California Euphorbia notes. Bull. So. Cal. Acad. 33: 105-11 128. 1934b, Euphorbia on Guadalupe Island. Leafl. West. Bot. 1: eens 3. Range and synonymy of Euphorbia capitellata. Bull. Torr. he Ghab 62: 537-538. al. Acad. Euphorbia i in the Pacific States. Bull. So. C 35: 127-147, aa ee: pamomec runes, 36b. Revision of the Euphorbia polycarpa gro southwestern United ee and adjacent Mexico. Bull. Torr. Bot. Club xt figs 63: 397-416, 429-450, 3 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 107 —_————-. fe Note on Euphorbia. Rxuopora 39: 496. ee een ee reign of New World Euphorbiaceae. Contr. _ Gray Herb, 124: 35-4 9b. A miscellany of New World Euphorbiaceae II. Contr. ” Gray Herb, abt 48-78, Pl. III-IV. Se wo ornamental Mexican Euphorbias. Journ. Cactus & Succulent Soc. 11: 44-47. ——_——, 194, Dichapetalacea et Euphorbiaceae novae. Proc. Biol. Soc. Wash. Agee? Js 1756: ‘Buhoe Linnaeus’ Amoenitates Academicae 3: 100— 131. Il. Taxonomy Score oF THIS TREATMENT This paper is intended to include all native or naturalized species of the subgenus Chamaesyce occurring in Canada and the United States, exclusive of Southern Florida. The species oc- curring in the United States only in southern Florida are mostly closely related to, or even conspecific with, the West Indian species and can be satisfactorily treated only in relation to them. The line of demarcation in Florida is at about 26°-27° north latitude. This conclusion was drawn from examination of the abundant Floridan material of the subgenus at the New York Botanical Garden. After I had reached this conclusion, Mr. K. J. Alexander stated, in conversation, that 26° north latitude is the usual northern limit of subtropical species in Florida. It is intended to treat all members of subgenus Chamaesyce growing in the area specified. For species occurring primarily within the area all synonyms or alleged synonyms have, so far as possible, been considered and referred to their proper position. In order that the exact basis of all conclusions may be evident, names are treated according to their types. All names based on one type are included in one paragraph. In this paragraph the type, its location, and whether it has been examined are all concisely indicated. An exclamation point (!) is used, as is cus- tomary, to indicate that the type, photograph of the type, or isotype has been examined. MEASUREMENTS All measurements have been made on dried material with the exception of those of involucres, glands, androphores, bracteoles, and styles which were made after the parts were boiled in water. If in some cases it appears that the seeds would be a rather tight 108 Rhodora [APRIL fit for the capsules, remember that fully mature capsules dehisce on drying. Consequently the capsules measured on dry her- barium specimens are not only slightly immature but, in addition, must have shrunk a little in drying. This method of measuring, while open to some theoretical objections, is eminently practical since these plants are nearly always dry before identification is attempted. = The tangential dimension of the seed is the dimension in _ the plane normal to the radius when the seed is oriented as it is in the capsule. The radial dimension is comparable and is measured along the radius. A possible source of confusion is the statement as to the “radial” shape. This is the outline seen when looking along the radius and viewing the inner (raphal) face of the seed. The raphal face or the ventral side of the seed is the side toward the center of the capsule; the dorsal side is that away from the center. ABBREVIATIONS The herbaria from which material has been studied and me which specimens are cited in this revision are indicated by following abbreviations: A—United States Field Station, Sacaton, Arizona. B—Berlin-Dahlem, German C—University of California: Berkeley. CA—California Academy of dees, San Francisco. CL—University of California, Los ae eles Cl—Clokey Herbarium at Los Ang eg reane D—Dudley Herbarium, Stanford Daire ersity, "Cali forni t or in Deam—private herbarium of C. C. Deam, Blufton, ne ai now & process of transfer to cag University. F—Field Museu um, Chicago, Illinois, G—Gray Herbarium, Harvard Piaversity Cambridge, Mass. Ge—Geneva, Switzerla nd. {Intermountain Hoteran Utah State Agricultural College, Loga”, U J—Herbarium of W. L. Jepson at ig tec! of California, Berkeley. K—Royal Botanical Gardens, Kew, England zam—Los Angeles Museum, Los Angeles, California. Missouri Botanical Garden , St. Louis, M Mia—Maria Mitchell Society, Nuntuskee fea 1 ta. N_—National Arboretum Washi eS AE tm Fs 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 109 —New York Botanical Garden, New York City. iy se rium of Frank W. tech, 1077 New York Ave., Altadena, alif Ph—Academy Natural Sciences, Philadelphia. PhB—Philadelphia Botanical Club at Acad. Nat. Sci., Philadelphia. RS—Rancho Santa Ana Botanic Garten, Anaheim, California. SB—Santa Barbara Museum, Californ Sh—Forrest Shreve, ge hechasbuiss Tucson, Arizona. T—University of Ariz ucson. US—United States Nationa Serica. Washington, D. C. —Herbarium of the aut Since completing the manuscript, the herbarium of Catholic University of America, Washington, D. C., and specimens from Southern Branch, University of Idaho, Pocatello, have been ex- amined and important extensions of ranges incorporated in the maps, but no specimens from these herbaria are cited. Kry To SUBGENERA OF EvPHORBIA OCCURRING IN CANADA AND THE UNITED STATES In order that it may be ascertained whether the Euphorbia in hand is to be found in this paper the following key to the sub- genera in this area is offered. 1. Glands of i involucre Phe Lig te mi ndages; glands eeply ouprs if lea — e below the inflorescence; leaves essentially symm 2. Glands of ey involuere ether deeply cupped or concealed by the inflexe segments of the margin; stem never — into a symmetrical 3-several-rayed inflores- Sn soils ately Seba coe hee pee a ee I. Poinsettia. 2. Glande of the involucre flat or convex, never concealed; e symmetrically forking inflorescence (floral te aor ES ad PRUE II. Esula. 1. Glands of the involucre with petaloid appendakes or, if rf ndages oe leaves all strictly opposite and with in equilateral 3. Leaves alters opposite, or even whorled, their bases sym saneteloal= ye glandlike OP MONG Fi in Caen III. Agaloma. 3. Leaves all strict y opposite, their bases usually strongly inequilateral; stipules rae well developed, always evident in species with symmetrical leaves........ IV. Chamaesyce. Subg. Poinsertra (Graham) House, N. Y. State Mus. Bull. 254: 473. 1 1924. (Poinsettia Graham, New Philos. Journ. 20: 412. es se ks additional synonymy see Contr. Gray Herb. 127: 110 Rhodora [APRIL II. Subg. Esuna Pers., Syn. Pl. 2: 14. 1806. (Huphorbia sect. Tithymalus of various authors. Tithymalus Adans., Fam. Pl. 2: 356, 611. 1763, et al. Galarhoeus Haw., Syn. Pl. Succ., 143. 2 812.) III. Subg. AGatoma (Raf.) House, N. Y. State Mus. Bull. 254: 471. 1924. (Agaloma Raf., Fl. Tellur. 4: 116. 1838; Lepa- ena Raf., op. cit., 113; Dichrophyllum Klotzsch & Garcke, Monatsb. Akad. Berlin 1859: 249. 1859; Tithymalopsis Klotasch & Garcke, |. ¢.; Zygophyllidium (Boiss.) Small, Fl. SE U. 8. 715. 1334. 1903). For additional synonymy see Contr. Gray Herb. 127: 57-58. 1939. IV. Subg. Cuamarsyce Raf., Amer. Mo. Mag. 2: 119. 1817. Type: Euphorbia supina Raf—Chamaesyce Raf. 1. ¢. momen provisorium under the above name. Both proposed without reference to Persoon. a defined subdivision of Euphorbia of subgeneric rank). Type: L Anisophyllum Haw., Syn. PI. Succe., 159. 1812, not Jacquin, Select. Stirp. Amer. Hist., 283, T. CLXXX, fig. 5. 1763. 1 A. Peplis (L.) Haw.; based on Euphorbia Peplis L. nies cyathium n. g., 16. 1906, preprint from Beih. Bot. Centralbla® : 32. 1907, published merely as “Untergattung Anisophyllum Haw.”’ Without basinym, hence not valid. : Chamaesyce S. F. Gray, Nat. Arr. Brit. Pl. 2: 260. 1821. bbe Chamaesyce maritima 8S. F. Gray, an actually and avowe oy superfluous name for Euphorbia Peplis L. Small, Fl. SE U. 1. 707. 1903; Millsp., Field Mus. Pub. Bot. 2: 300. 1909; aug Amer. Midl. Nat. 1: 204. 1910 ; Small in Britton & Brown, ; fl. No. U.S. & Canada, ed. 2, 2: 462. 1913; Millsp., Field Mus Pub. Bot. 2: 384. 1914; Millsp., op. cit., 401. 1916; eae : 0 aesyce (S. F. Gray) House, Bull. N. Y. State Mus. : : ; ; : De Euphorbia section Anisophyllum Roeper in Duby, A. P. ¥ Candolle, Bot. Gall. ed. 2, 1: 412. 1828. Type: Euphorbia eer ven en 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 111 Peplis L., — here. G. D. J. Koch, Syn. Fl. Germ. Helv., 627. 1837; Ledebour, Fl. Ross. 3 (2): 587. 1849-51; Baillon, «& : r 1862; Muell. Arg. in Martius, Fl. Brasil 1 (2): 669. 1874: Bentham & Hooker, Gen. Pl. 3: 258. 1883; Chiovenda Bull. Soc. Bot. Ital. 1895: 61. 1895; Thellung, Bull. "Herb. B att: 2, 4% 746. 1907; Thellung in Ascherson & Graebne rg tg iicicars Fl. 7: 422. 1917; Denis, Euphorb. Iles Asset: ’Afr., 27. 1921 Tithymalus (Tourn, ) aig Anisophyllon Gomez de la Maza, Fi. l. Habanera, 152. Eupho rbia A. Chonan Reichenbach, Fl. Germ. Excur, 755. 1832.1. Type: Euphorbia Chamaesyce L. Reichenbach Repert. Herb. Nom., 193. 1841, as Euphorbia “1. Chamaesyce Caesalp.”?; Nyman, Consp. Fl. Eur. 3: 655. 1881, Srl as Buohotea ‘““g. Chamaesyce Rchb.” Xamesike Raf., Fl. Tellur. 4: 115. 1838. Type: X. vulgaris Raf.; based on Euphorbia Chamaesyce L. ee ich Aut. Bot 96. 1840. —Xamesike subgenus a aneeske Raf., op. ou Xamesike subgenus ny Raf., Fl. Te ilur. ri "15. 1838; Type: Xamesike supina (Raf.) Raf., based on Euphorbia supina Raf.; ; Rafinesque, Aut. Bot., 97. 1840. Raf., New FI. No. Amer. 4: 98. 1838. Type: A. eter pe Raf. propos Bhan without reference to Euphorbia prostrata isto .—Xamesike subgenus A plarina (Raf.) Raf., Aut. Bot., 97. Euphorbia section Anisophyllum § Acutae Boiss. in DC. Prod. 15 (2): 18. 1862. Type: Euphorbia acuta Engelm. Pax in Engler & Prantl, Nat. Pflanzenfam. 3 (5): 104. 1891; Pax & Hoffmann in Engler & Prantl, Nat. Pflanzenfam. 2 Aufl, 19¢: 210. 1931. uphorbia — Anisophyllum § Chamaesyceae Boiss. in DC. Prod. 15 (2): 1862. Type: Euphorbia — syce L. Pro- posed without gL ks to earlier uses of similar names though Boiss., op. cit., 11, cites in synonymy under ge Anisophyllum, genus “Chamaesyce Haw., and ‘Sect. Chamaesyce Reichb. Fl. — exc. p. 755. Hatten in Engler & Prantl, Nat. Pflanzenfam. 2 Aufl. 19¢c: 1931 Euphorbia subgenus Bea a Gaucher, Etude anat. genre Euphorbia, 123. en proposed independently and escribed, no species assi Glabrous to vestite herbs or sometimes sub-shrubs; leaves See page preceding title page between 434 & 435 for dates. 112 Rhodora [APRIL opposite, simple, mostly with inaequilateral bases, dy short to none; stipules small, often united; cyathia solitary a pean unculate, smooth to variously sculptured, with a ela t of varying thickness.—This description applies only to pai included in this paper. Key To THE Species or SUBGENUS CHAMAESYCE 1, Ovary and capsule vesti 2. Perennial (except no. 3b with capsule ca. 2.5 mm. in diam.); or rarely as few as 15 in nos. 25 and 26; retonasty never urceolate. : 3. Cyathia born ense cymose glomerules, or a few in addition pominti aes aatary | in the upper biiaronen) Webven OIG WOT es ele tac 3. Cyathia solitary at the vate fai at the tips of a anches; leaves always en — s scarcely an angled, carne ovoid, une” by 4. Seeds eds quadrangular, variously smooth to se . E. capitellata. E. pediculifers. 5. Hawes hag phi spreading hairs 6. Capsules 1.3 mm. in diam ae sharply hirtella. ate a es ee eS 26b By nant t var. 6. Capsules ca. 2.5 mm. in diam.; seeds qr uadrangular.. 6.54. 3b. EF. ocellata var. Rattanit. 5. Herbage casinly age with appressed, long and weak, or matted hai if Leaves acuminate, tins mostly over 1 cm. long. 8. Stems strigose; seeds 1.6-1.9 mm. long; cap- : 1 wee 2. a EA wns lone... 5. 6.4; - i 7 hs E, angust@ 8. At least the young stems villous; seeds 2.2- seule mm. long; capsule ca. 3 mm. long........--- 18. E. 7. a senate blunt, shen more than sub-acute, arely pg as 1 ¢ m. long 9. Capsules . 2.5 ing ones seeds ca, 2 mm. 1.7 mm 10. or wider than the glands and short spreading hairs beneath and 9. Capentos ne over 2 mm. long; seeds not over ong. a oe WN 24. E. vallis-mortae- 10. AEDOSa EN wide to absent, glabrous or ely with a few hairs beneath next to the involucre 11. Appendages usually conspicuous; styles nr cE ieee 11. Appenciages absent or very naITOW} peer 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 113 2. Annual — no. 39 with urceolate involucre) ; ; staminate flo o 12, or ws gt oyay 0 15 in nos. 385b and 43; cies 5 s than ae j airs tapering; annual................. 40. E. setiloba. 12. Involucres ‘oboonieal to campanulate ce orne in dense axillary and terminal leafless TUB 6s ee eel et as 16. E. hirta 14. Cvathia solitary or on short leafy lateral branchlets. 5. Proximal a Spo greatly prolonged, often wa pubis Ae CRDONIG as oe ee . E. indivisa. 15. a and. distal eg without ied isparity in -truncate, apex sharply acute; styles entire, sometimes emarginate........... eco E. stictospora. 17. Glande oithious: appendages or ik but the merest rudimen tire, not over 8 mm. long; vestiture — straight, and ‘spreading ee ag epee . micromera. or granular; leaves often s i some often over 8 _ ong; entities long and weak, cris 18. pcr: “sparsely tae gy not t ndged... 6.6. E. serplifolia var. hirtula. 18. Gaseite's strigose, or if with spreading ha ansverse 19. Seeds with low rounded transverse rid te) on summit or with pride granular surfaces; capsules 20. ri, ge ca. 0.4-0.5, rarely to 0.6, long, re seed ie not aa: nodes nev r rootin 21. a all stunted from the in- erie Soe ae ey ae 45. E. supina. 21, ce ules ostly half included and disten nding “he involucre at ma- AN pieces ee ea 44. E. thymifolia. 20. Styles oe O7 mm. long, slender; seed t granular; nodes often rooting 46. E. humistrata. 19. Seeds with narrow sharp transverse ridges, or rounded transverse ridges whitened on the langemare capsules to- menhtose or with crisped mavating hairs Haw 1.1-1.3 mm. A ee the Sanat ridges -- nag same width as the cereals era “si ae ee ei aa E. laredana. er age nea r greenish, less densely vestite to ieee: seeds 0.9-1 114 Rhodora [APRIL mm. long, the minute sharp ridges narrower than the intervals between 48. E. Chamaesyce. 1. Ovary and capsule glabro 23. Pn ‘er es united into a white; glabrous, membranous scale. Hae aor nnual; staminate neg aly P10. ee oa ee 30. er pens. a Perenn ial; staminate flowers 12 or more....... E. albomargsaaity 23. — not onted te a white, glabrous onan 25. Styles entire, casi very short and capitate or as long e and very slender 26. Styles ogi as long as the snail, slender; leaves a prolate: MME 6.6 ech His Gee pases Styles very short, capitate; —<— entire; perennial. . 33. E. asty 25. ‘Stylos bifid or if entire neither ex — short and cap- itate nor about as lo rhe hoe apsu 27. Face of seed virtually flat, the iene apical mucro overhanging the raphe. 2c. . 6. i i ee i 2. E. platysperma. 27. Paces of seed not flat; st He mucro wanting or minute . Seeds transversely wrinkled or ridged; involucre -9-1 mm. in diam.; appendages no longer than ai ——- ultimate branchlets 0.15- 0.2 n diam.; ongest leaves rarely oc Se Oe ew ee eee ee ee Oe Oe Oe Oe Kee . E. revoluta. timate “lanes ets ca. mm. in pg : 30. Se m. lon ds sharply rales — leaves 2 TR on mm " trachysperma. E. polygonifolia. 30. Capea pie than 3 mm. : Oe, aye ae es Hack oo with eee ‘seed mm. were = ndages aon icuous; see 5 subalatel ‘an led es “ ; 10. E. florida. nute rudiment; leaves always — never linear, never more than 10 34. Roots with ce OMe 38. E. thervaca. 34 s smooth to rugulose but never trans- versely ridged. 4 en: oe “sae ss ; V3, B. oll elena nneenmnnenle sineenm eae T nee 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 115 37. Glands discoid; staminate flowers Die oar, sae eee ys ees 27. E. Parishit. 37. Glands transversely oblong; stam- inate flowers 15-32......... 26. E. polycarpa. 33. sepa with appeniaece leaves serra al e, sometimes more t 0m on 38. Robust erect annuals with the larger anche ingiey over 15 mm. long, margins 39. Cae 1.6-2.3 mm. long, wider below e i fa cymes not very dense and of mostly few to several cyathia. 40. Stems “usually crisply hairy at least e young tips, rarely pilose; mel nely wrinkled.......... 13. E. maculata. 40. Stems mostly 1 a sometimes ilose; seeds with broad very shal- sce depressions separated by low a ey ‘2. . hyssopifolia. 39. Capsule 13-1 ee rea long, widest at the equator or nearly so; cymes sa very dense with numerous Ps so Se ak We 15. E. glomerifera. 15 mm. hee. if robust and erect, leaves all entire, ba completely glabrous, and seeds sm 41. Cyathia in caine sien inal cymes, a few in addition sometimes in the upper bifupentionas | stems not pilose 17. E. capitellata. 41. pss solitary, or if in leafy cymes, pilose 42. _ Seed with ele transverse ridges. radially “cin, vhs to a long; caps widest at — at cant the stems oben PRUE US ee an 37. E. Abramsiana. 43. Geode ovate radially ; capsule widest well below the equator; ce e wars glabrous, «3... 3s 6. E. glyptosperma. 42. Seeds smooth to rugulose but never wi tran: miei 44. Herbage variously vestite. 45. th and mostly chalky white, sharply angled; leaves entire and plant perennial. 46. Leaves entire; pono: — age pu ee ee . E. polycarpa. 46. Leaves not entire, or if ed ope perennial, stem 47. Carpele mostly prolonged in 116 Rhodora [Aram 47. Carpels never prolonged; 7 stems never white with long tapering hairs. 48. Leaves 5-15 or rarely 19 tapering; involucres 1.2 -1.6 mm. in diam.; an- 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 117 55. —— P crite into numer- segments; leaves entire ak mostly ordate bases. 28. E. cordifolia. 55. Stipules. ia as above; leav serrulate or entire, not pani as 56. pacing, el than 2 Rens eat ar 32. E. Fendleri. 56. Capsules inti than 2 mm. 57. aaa with aes —— seeds 1.6— on 57. Pontinlel: or leaves mostly i neg on 58. Perennial leaves en- tire; stipules ciliate (except in var, sim- ulans); stems nev- ee 26. E. polycarpa. 58. “ys leaves most- pens often winged 35, E. serpyllifolia. . 1. EvpHorsia POLyGonIFoLia L., Sp. Pl. 1: 455. 1753. Type: Habitat in Canada, Virginia’, Kalm (Linnaean Herbarium, not seen; photograph G!). Boissier in DC. Prod. 15 (2) : 28. 1862; g i > 439. Grae 1917. Anisophyllum polygonifolium (L.) Haw., Syn. Pl Succ., 16 1 amesike ¢ pee Raf., Aut. Bot., 98 based on “Eu es}, i. e. doubtless EH. polygonifolia of everybody, hence, by ines nee, Linnaeus.—Chamaesyce polygont- folia (L.) Small, Fl. Se. AE S., 708, 1333. 1903. . maritima Nutt., Trans. Amer. Philos. Soc. 5: 1837. Type: ‘Sea coast, L{ittle?] Egg Harbor’, New Tersey, N ‘uttall?] (Ph!, possibly only isotype; photographs G!, Gl: abrous annual; stems few to several, Aen prostrate, sometimes res to erect, 1-25 cm. long, 1-2 mm. in diam., internodes 0.5-5 ¢ sos mide r 6-16 mm. 0 4 oblong- 118 Rhodora [APRIL RA . ; LOMERIFE Map 1, range of EupHorsia ABRAMSIANA in the U. §.; 2, E. ¢ ar. b..:3. OLYC: in the U.S ie ;U.S.; 2 9 : YCARPA E. POLYCARPA var. HIRTELLA in the U.§.; 4, E. Pou 5 sE ‘CARPA V TYPICA in the U.S P : var A grades, circles E. pouycar var. SIMULANS; 6, E. PEpI ;ULIFERA a: 8, dots in the U. §.; 7, s E. serrvua in the U S., circles E, THERI ona in i. REVOLUTA, circles E PLATYSPERMA; 9, E RRYI; 10, dots E. M fe : Canada he U.S., ] GRACILLIMA; 11, Npivyins 1 ERMEDIA; 2, E. MELANADENIA ots sae between E circles E. mMissurica var. Typica, half filled circles intergrades . HIRTULA MISSURICA Vars ICA a INTERMEDIA; 14, IFOLIA V da: 16, .8.; 15, E. serpyiurronra var. GENUID nd : VILLIFERA var. TYPIca in the U. § - saith indiana 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 119 a slightly exceeding the glands, proximal often narrower; glands 0.3-0.4 mm. long, broadly transversely oval to subcircular or acres nearly double and figure-eight-shaped, shallowly cupped; appendages rudimentary or absent; fifth gland small or absent, the involucre irregular and the parts difficult to identi- fy in this interval; sinus narrowly U-shaped to V-shaped, little depressed; bracteoles glabrous, united below, free above, forming a three- to several- parted tangential structure between and outside the fascicles, shorter than the rabies cased stami- nate flowers 1-3 per fascicle, 5-14 per cyathium, or some of the uppermost cyathia with staminate flowers abortive or sone ng; androphores 0.8-1.3 mm. long, glabrous; gynophore glabrous, soon exserted and reflexed; ovary roundly 3-lobed; styles 0.7—1 mm. long, bifid to the middle e, mostly Aaely reflexed; capsules 3-3.5 mm. long, roundly 3-lobed, broader below the equator; seeds compressed-ovoid, base subtruncate, apex subacute, back stone rounded, face slightly rounded, 2—2.6 mm. long Me 6-1.9 tangentially, 1-1.6 m radially, coat white, secdieataaty celbilanahenlal: beiuroaneina i the brown microfavose macro- scopically smooth testa showing through.—P.La 54A, Sandy or gravelly beaches above high tide aie sand dunes, both maritime and lacustrine, Magdalen Islands, Quebec, south to Georgia; shores of Lake ‘On tario, Erie, Michigan and the southern tip of Lake Huron; introduced in — oun (M \P ). pete ie specimens seen: Hementh Long & St. par 7720 (G, Ph). Prince Epwarp IsLanp. Prince Co.: near Campbell’s Pond, Malpeque, Fernald & St. John 11114 (G); Lower vn Cow Pond, ’ Fernald, Long & St. John 7719 (G, Ph, US). Queens Co.: Grand Tracadie, Fernald, Long & St. John 7722 (G, Ph, “US): Brackely poets Fernald, Long & St. John 7721 (G, Ph). Nova Scorta. Cape Breton Co.: Sydney Mines, Bissell & ‘Linder 217565 (G, Ph, NY). Seen Co.: Little Cariboo Island, C. B. Robinson 187 (NY). Ouiabestana Co.: Linden, Aug. 11, 1885, Trueman (G). Queens Co.: Central Port Mouton, Graves, Long & Linder 21754 (G, Ph, US). New Brunswick. Charlotte Co.: Long Pond Beach, Grand Manan, eee 5700 (G). Gloucester Co.: Trac adie Beach, Blake G, NY, US). Marne. Knox Co.: Matinicus, C. A. soe 74 (NE). C rland Co.: Brunswick, railroad . Sept. 21, 1907, Kate Furbish (NE). Sagadahoc Co.: Popham each, Aug 1894, Fernald (G, NE). Oo Segoe 13994 (NE, Ph); Old Orchard, Fernald 2783 (NE). tee Hampsuire, Rockingham Co.: Little oonge Rye, Sept. 19, 1901, E. F. Williams (G); Hampton, Oct. 5, 1901, Knowlton (G, NE). Massacuuserts. Essex Co.: Baca D. White 296 120 Rhodora [ APRIL (NE); Plum Island, Newbury, Pease 2716 (G); Rockport, Aug. 14, 1898, E. F. Williams (NE). Suffolk Co.: Revere Beach, Pease 9858 (NE). Norfolk Co.: Quincy (Rufe’s Hummock), M, NE, NY, Ph, US). Nantucket Co.: Wauwinet, Churchill 545 (G, M); Tuckernuck Island, Aug. 6, 1909, Cushman (Ma); Nan- tucket, 1912, Brewster (NE); Coatue, Aug. 12, 1933, Wyatt « Franklin (Ma); southwest beach, Nantucket Island, July 20, 1910, Cushman (Ma). RuopE Istanp. Newport Co.: Prudence Island (Portsmouth), Sanford 10384 (NE); Block Island, Fernald & Long 9817 (G, NE, Ph). Kent Co.: Greens Island, Warwick, Aug. 13, 1921, Hope (NE). Washington Co.: Quo- nochontaug, Oct. 12, 1919, Hope (NE); Westerly, Aug. 31, 1919, Weatherby & Collins (NE). Connecticut. New London Co.: Old Lyme, Sept. 29, 1917, Woodward (NE). New Haven Co.: New Haven, Blewitt 849 (NE); Orange, Bissell 528 (NE) ; Guilford, July 30, 1906, Bartlett (G). Fairfield Co.: Bridgeport, Sept. 4, 1898; Eames (G). New York. Suffolk Co.: Fishers Island, St. John Nashawena, Elizabeth Islands, July—Aug., 1901, Northrop oie G sex Co.: Long Branch, 1852, Short (Ph). Monmouth Co.: Sandy Hook, Aug. 15, 1887, Stabler (G). Ocean Co.: below Mantolok- ing, sand dunes, Aug. 11, 1902, Lyons (US). Atlantic ©o-' Atlantic City, Gross 2498 (NY). Cape May Co.: Cape May; Pennell 2214 (US). Salem Co.: along Delaware River, betwee? Straight & Black Ditches, 3.75 miles west-northwest of Hancocks Bridge, Fogg 7753 (G). Punnsyivanta. Erie Co.: Presque Isle, Erie, Sept. 3, 1868, Porter (US); Presque Isle, Erie, Sept. 4, 1868, 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 121 Garber (Ph). Drtaware. Kent Co.: 4 mile south of Smyrna iver, Larsen 1044 (US). Sussex Co.: Ellendale, along railroad, Aug. 29, 1908, Williamson (Ph); Rehoboth, Larsen 425 (G, M, Ph); south of Bet thany Beach, Fogg 11208 (G). MARYLAND. Worcester Co.: Ocean City, Killip 7343 (US). pei ib Co.: Gunpowder River, Plitt 857 (G). Calvert Co.: sandy shore, Chesapeake Beach, House 371 (NY, US). St. Marvé 6" beach. Tall Timbers, Killip 32209 (G, US). Vrrerta. Elizabeth City Co.: Fortress Monroe, 1879, Vasey ie Northampton Co.: Savage Neck, Fernald & Long 5350 (G, NY). Princess Anne Co.: Kearney 181 NortH Caroutina. New Hanover Co Carolina Beach (below Wilmington), Biltmore cate 3838" (G, , NY, US). Carteret Co.: Atlantic Beach, R. K. Godfrey 6490 (G); Sold, Salter Path, Bogue Island, July 15, 1926, Anonymous Ph). SoutTH CAROLINA. Charleston Co.: Isle ‘of Palms, Biltmore Herb. 3838 (US). Beaufort Co.: Bulls Point Beach, St. Helena Island, Cuthbert 707 (NY). Groroaia. Chatham Co.: Tybee Island, Harper 786 (NY, US). Glynn Co.: Saint Simons Island, Biltmore Herb. 3838° (US). Onrario. Prince Edwar : Wellington, Sept. 3, 1902, oe (G, US). Welland Co.: Point Abino, Biltmore Herb. 38384 (US), McCalla 287 (US). Lambton Co.: Point Edward, Macouwn 88089 (G). a St. Clair Co.: shore of Lake Hu ron, near Port Huron, Aug. 8, 1 Dodge (G, US); Fort Gratiot, along St. Clair River, July 20, 1870, Gillman (G). Leelanau Co.: 34 mile east of shore of Lake Michigan, Glen Haven, Hermann 2248 (G). Berrien Co.: Harbert, Johnson 1132 (US ); lake shore, St. Joseph, Aug. 10, 1838, Houghton (NY). Onto. Lake Co.: Lake Erie, Salida Beach, Webb 1 6 (G). Cuyahoga Co.: Lake Erie, Cleveland, 1840, Sullivan (Ph). Erie Co.: lake shore, Vermilion, Sept. 16, 1895, k (US); Lake Erie, Cedar Point, July 17, 1914, MacDaniels o InpIANA. La Porte Co.: frequent at Michigan City, Deam 5206 (G, NY). Porter Co.: Lake Michigan, Mineral 4 te Lansing 3859 (G, US); Dune Park, V. H. Chase 205 (G, M), A. Chase 2111 (US), Umbach 1917 (Ph). Lake Co.: Lake Michigan, Miller’s, Aug. 14, 1897, Umbach (US), Sept. 4, 1911, Sherff (G); Lake Michigan, Whiting, Aug. 18, 1897, A. Chas e (Ph). Wis CONSIN. Door Co.: Lake Michigan, Sturgeon Bay, “Bassolt 18156 (G, M). Kewaunee Co.: Algoma, Fassett & Wilson 14674 (G, M). Milwaukee ser Nbweukees 1865, Lapham (G). Racine Co.: Racine, Sept. 1878, oe s (G). Intrnors. Lake Co.: Wau- kegan, Gleason & "Shobe 3 36 (G). Cook Co.: near shore, Sheridan Park, Chicago, A. Chase 1896 (us) FRANCE: Gironde, sables maritimes, Soulac, Oct., 1903, Pitard (G eo UPHORBIA PLATYSPERMA Engelm. ex 8S. Wat Calif. 2: 482. 1880. Typr: Near the oa of chs teas 122 Rhodora [ APRIL River, southern Arizona, 1869, Ed. Palmer 2 (M 144649!; photo- graphs G!, W!; isoryprs F!, G!, NY!, US!). } son, Man. Fl. Pl. Calif., 600, 1925. TYPE: marginate, margin light color than the brownish inner portion but scarcely differentiate 2 mm. long, exserted; gynophore glabrous, long-exserted @ reflexed at maturity, slightly angled; ovary glabrous, scarcely lobed, carpels evidently channeled on back; styles stout, Par ‘4 to the base, 0.5 mm. long; capsule rotund-ovoid, pene 3-lobe ; l dially, .é tangentially, broadly oblong radially, base truncate obliquely inward, apex with an inflexed mucro, back rounded, smooth, face with two smooth, flat, nearly approximate, slightly de- pressed facets separated by the elevated raphe.—PLaTE 6 5D. California to Arizona? (Map 8). Only the two above-cited collections of this species are know? Jepson found but one plant and Palmer appears likewise to have | | 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 123 found only one. The exact locality of Palmer’s collection is un- certain. The data on labels vary. Some read merely ‘‘Southern Arizona”’, others ‘‘Near mouth of Colorado R.’’, and Engel- mann’s own ‘“‘near the mouth of the Colorado River, Ariz.” This species is of particular phytogeographic interest in that it is seemingly an originally littoral species closely related to the common £. polygonifolia of the Great Lakes and Atlantic Coast. The rarity of this plant may well be due to the change from littoral to inland desert habitat caused by the geologically recent recession of the sea from the Salton Sink. The sea covered this area so recently that the old beach-line is plainly discernible and small delicate gastropod shells left lying on the desert have yet to disintegrate. Perhaps examination of the old beach line, par- ticularly in sandy areas, will yield further collections. 3. Eup Sei. ser. 2,3: 4 nnual: peor prostrate, few to numerous, long, often to 1.5 mm. diam., slightly loeeahen wally ead gla- brous or daha. internodes 1—3 cm. long; leaves glabrous or pubescent, blades entire, Heats Aiea talento, 4-10 mm. long, apex blunt or mucronulate, base oblique, midrib elevated be- neath and lateral — pr rominent, margin revolute, or ovate- UPHORBIA OCELLATA Durand «& Hilgard, Journ. Nat. Acad. 46. 1854. lanceolate, 7-15 long, a acuminate, base obtuse and only slightly oblique, thidrib not elevated below and lateral veins mostly obscure, margin plane; petioles 1.5-2 mm. long, amplexi- caul on lower side of stem; stipules distinct or lower slightly united at the base toward the stem-tips, ew or broader, entire or parted, pean or pubescent, 1—-1.3 mm e- duncles stout, 2-4 mm. long, glabrous or ee Comat ceyathia Peron ary at the Bea verse involucres turbinate to campanulate, 2mm. diam., five-lobed especially in the late season, epee is pubescent without, pubescent above within; lobes br ciadlly inne or less united below and adnate to the Peeiint ca. "3-10 per group, tips heavily or sparsely beset with short stout hairs; na flowers 8-13 per fascicle, 40-60 per involucre; andro- Phores 1.7-2 mm. long, glabrous, barely equaling or mostly 124 Rhodora [APRIL shorter than the glands; gynophore glabrous or with short hairs above, long-exserted and usually reflexed at maturity; ovary three-angled, glabrous or pubescent; style ca. 0.5 mm. long, parte ed to the middle, glabrous throughout or pubescent below, divisions terete; capsules strongly three-lobed, 2—2.3 mm. long, broader than long, glabrous or with short spreading hairs, carpels roun on the back and mostly with a low channeled ridge on the back; seeds ovoid, with lateral pie oe barely visible or wanting, or turgidly quadrangular, nlp en to ovate-acute radially, 1.4-1.7 mm. long, 1.1-1.3 mm. radially and tangentially, smooth to rugose, coat white, amicroreticulate, sometimes little obscuring the brown to gray t Key To VARIETIES saat ba glabrous. n leaves ovate-lanceolate, not at all or very slightly aieats, acuminate, usually without evident lateral veins; | eels SoS te ee a eee c. var. arenicola. Median leaves ovate-deltoid-faleate, blunt or mucronulate, lateral veins evident below; seeds often rugulose or rugose. < Bhs: hates : Fiorbage. pabement. 050 ae b. var. Rattan 3a. E. ocettata D. & H., Journ. Nat. Acad. Sci. ser. 2, The 1854, var. Typica L. C. Wheeler, Bull. a Bot. Clu ier 1936. Typn: Poso Creek, Kern Count pesca . W)). Heermann (Ph!, possibly only isotype; Tho iveranbl Gi, R An average member of the species. Durand & Hilg ard, i Pree ocellata (D. & H.) Millsp., Field Mus. Pub. 2: 10. 19 Chamaesyce obi Millsp., Field thot ~~ Bot. 2: 405. 1916. Typn: hills near Big Chico Creek, east of Chico, Butte County, California, Sept. 16 1913, A. ea Heller 11 He a te A photographs G!, W!; isotypes Ch, ! M!, Ph!, US!). local race with i Bass ide A Poe, seemingly too defined for recognition.—E. ocellata D. & H. var. sulfuren : - 1936. A. Cieencitate valleys. of California from Shasta County south to San Bernardino County (Map 22). Representative specime : ir, Jepson 6 River, in the low hills, Jepson'15279a (J). Contra Costa Co. Mount Diablo, Jepson 13926 (J). Stanislaus Co.: La sp Jepson 13925 (J). Monterey Co.: Jolon, Vasey 577 ( 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 125 Fresno Co.: Rancho Cantua, es Creek, Sept., 1908, Lillis _ Tulare Co.: Halstead’s Rane avis Ranch, north fork o Kaweah River, Sierra Nevada, Jepson 577 (J). Kern oe Cottonwood Creek on River Road, between Bakersfield and Bodfish, Abrams 5841 (NY); near Havilah, Coville & Funston 1081 (US). we Bernardino Co.: sandy plains, Colton, 1882, Pringle (G, NY, Ph). For oe ee additional specimens see Bull. Torr. nae Club 63: 403. 1936. This entity is very uniform throughout most of its range. The plants of the San Bernardino region, though probably iso- lated at the same time as those of the Mohave Desert, show no significant differences from those of the Central Valley. Inci- dentally, collectors would do well to search for this plant near San Bernardino. Parish made several collections in this vicinity before 1900 but recent collectors have not found it. The uncul- tivated brushy plain northwest of Slover Mountain and north of Jurupa Mountains (Hills) is a likely place. The specimen from Dunsmuir in the Sacramento River canyon is probably a chance introduction. The small leaves, long internodes, and slender stems support this supposition. At the extreme north end of its range H. ocellata breaks into local races. Var. Rattanii, q. v. infra, is the best marked. Var. sulfurea (Millsp.) Jepson is one of these ill-defined local races. I have no particular quarrel with anyone wishing to recognize it. Collectors will do well to secure a close series of E. ocellata at the north end of the Sacramento Valley around Chico and Oroville in order to determine the stability and range of var. sulfurea and the race north of Oroville. This race, which is usually readily recognizable by color, occurs at Table Mt. Olive Ranch, 7 miles north of Oroville, Butte Co. and is well represented by A. A. Heller 11143 (C, G, M, NY, Ph, US). Another collection made at apparently the same place and season three years later, Heller 12645 (F, G, M, NY, Ph, US) largely lacks the reddish cast but the foliage agrees fairly well in shape. The foliage and long internodes of this Oroville race bear a close resemblance to the var. sulfwrea of the foothills near Chico. But the seed coats of the Oroville plants are only very slightly wrinkled instead of Strongly rugose as in var. sulfurea. The plants of this Oroville race are usually reddish in color. Otherwise the species is yellow- green throughout except for some individuals of the pubescent 126 Rhodora [APRIL var. Rattanii. Besides the elongate internodes the Oroville facies differs from var. typica in the slightly larger and less falcate leaves. 3b. E. ocenuata D. & H. var. Rarranu (S. Wats.) L. C. Wheeler, Bull. So. Calif. Acad. Sci. 33: 107. 1934.—E. Rattanit S. Wats., Proc. Amer. Acad. Arts & Sci. 20: 372. 1885. TyPE: Stony Creek, Glenn County (formerly part of Colusa Co.), California, June 1884, V. Rattan 57 (G!, fragment F!).—Chamae- syce Rattanii (S. Wats.) Millsp., Field Mus. Pub. Bot. 2: 411. 1916. Like variety typica but the herbage beset throughout with short stout hairs; upper half of the gynophore, the ovary, and styles pubescent in like manner ; glands often with very narrow white appendages; seeds turgidly quadrangular, ovate-acute radially.— PLaTE 665B. ; Local in the Lower Stony Creek drainage, Glenn Co., Cali- fornia (Map 24). Additional specimens seen: CALIFORNIA. Glenn Co.: Stony Creek, two miles north of Orland, L. C. Wheeler 4041 (CL, Peir, P, W); gravelly bed of a large winter stream 5 ny east of Newville, A. A. Heller 11555 (C, CA, D, F, G, M, This is certainly an incipient species. With only the type Watson was fully justified in believing it a distinct species for, im addition to the presence of appendages and pubescence, the seeds of the type are all dark gray and evidently angled. But in my collection made at the type locality some seeds were distinct as in the type but others matched those of var. typica. 3c. E. ocentara D. & H. var. arentcoxa (Parish) Jepson, Man. Fi. Pl. Calif., 600. 1925.1 E. arenicola Parish, Erythea 7: 93. 1899. Typ: Camp Cady (Sink of the Mohave River on some labels), Mohave Desert, San Bernardino County, Califorma, S. B. &. W. F. Parish 1370 (D!; 1sorypes C!, F!, G!, M!, a US!).—Chamaesyce arenicola (Parish) Millsp., Field Mus. ae Bot. 2: 408. 1916.—#. cuspidata Engelm. ex Parish, Erythea ‘* 93. 1899, in synonymy; not A. Bertoloni, Misc. Bot. 2: 9. i Leaves ovate-lanceolate, acute, to 15 mm. long, base slightly oblique, midrib straightish, not elevated below, lateral ver mostly obscure; seeds strictly ovoid or the back and latera angles slightly evident, very smooth.—Puate 665C. h ohave Desert, California, east to Nevada, Utah, and north- western Arizona (Map 22). Carirornta. Inyo Co.: Searle Lake, Jepson 7144 (J). San Bernardino Co.: sand hills, Soda ‘icles oe date questionable but here accepted as no conflict as to priority 2 volved. AP RAE TS, SEAL cat Oe OETA LR ene Ae tae TE a a Oe ne RT nT ADEE eae eer ER en eR ae Oe ee i 1941] Wheeler—Euphorbia Subgenus Chamaesyce 127 Lake, Parish 10875 (CA, F, J). Nervapa. Washoe Co.: Wads- worth, aoe 918 (C, D, M, NY, US). Churchill Co.: Fallon, Headley 42 (US). Truckee Desert t, S. Watson 1077 (US). Uran. near Pahvant Butte, J. A. Harris 253 584 (G). Arizona. Mohave Co.: Virgin River, Purpus 6187 (C, NY). For rae of addi- tional specimens see Bull. Torr. Bot. Club 63: 1936. EvurHorsBia Parryr Engelm., Amer. ae vs 350. 1875. Tyre: loose drifting sand, St. George, Washington County, Utah, 1874, C. C. Parry 27h (M 144658!; photographs G!, W!; ISOTYPE G ty, Rather small plants. ere omni Parryi (Engelm. ) Rydb., Bull. Torr. Bot. Club 40: 53. 1 E. petaloidea goaceg 8 pope bese Engelm. in Emory, U. S. & Mex. Bound. § 2 (1): 185. 1859. Typr: valley of the Rio Grande eft Frontera eer ray sandy soil, El Paso County, Texas, July 26, 1851, C. Wright 1826 (M 149817!; photographs G!, W!; 1sorye GN. Large plants.—E. zygophy _— Boiss. var. flagllformis (Engelm.) Engelm. ex Boiss. in DC. P rod. 15 (2): 29. 1862.—E. cog eees (Engelm.) ba ere mt: S. Brande- l. : 1906, Bull. Colo. Agr. Exp er. Sta. 100: 223. 1906 (FI. Colorado). E. longeramosa S. Watson, Proc. Amer. Acad. Arts & Sci. 25: 1890. : leaves 5-28 mm. long, linear, entire, aequilateral, shortly petio- late; stipules distinct, linear, entire or parted; cyathia long- 1.2 mm. te 3-angled, Fr fememr wile yr mm. long; weeds. a triangular, 1.8 mm. long, narrowly ovate radially, mottled saat and white.—PuaTr 654D. Southwestern Colerae: Utah and Nevada; San Bernardino County, California ; Arizona: New Mexico; Western Texas; and northern Chihuahua, Mexico (Map 9). Representative speci- mens seen: CoLorRapo. Montezuma ee sandy plains, ~ Juan Valley, 1875, T. S. Senate ee (M). Uran. Grand vee miles west of Moab on desert, Maguire & Redd 1 958 (I, M). ne Co.: Kanab, in sand, Jones 6044 (M, NY). bi larae ssel Co. “a miles west of Toquer ville, B. Maguire, Ruth Mag Piranian 12315 (G, I): Anderson’s Ranch, Maguire £ viele ae 128 Rhodora [APRIL Goodding 759 (G, M, NY, US). Coconino Co.: Moqui Village, Aug., 1891, Owens (G); cindery soil near Tolchaco, H. C. Hanson A210 (M, NY). Navajo Co.: Holbrook, Rusby 827 (F, NY, Ph, US); Winslow, Peebles 9599 (US). Apache Co.: Adamana, Sept. 1, 1909, Rusby (NY); Billings, on Puerco River, Jones 4720 (I, NY, US). Graham Co.: Camp Goodwin, Gila Valley, Rothrock 389 (M, US). Cochise Co.: Bowie, Sept., 1884, Jones (US); near Wilcox, Sept. 8, 1914, Shreve (W). New Mexico. Dofia Ana Co.: 2 miles northwest of San Miguel, Fosberg $3785 (G, O); Straus’ Station, 1912, Stearns (M). Socorro Co.: Sabinol, Wooton 349(M). Texas. El Paso Co.: Cory 1922 (G). MEXICO. CuIHuAHUA: 36 miles south of Ciudad Juarez, Shreve 7922 (W); Sapio, Sierra Madre, Sept. 10, 1903, Jones (M, US); Colonia Diaz, Nelson 6455 (US). 5. EUPHORBIA AMMANNIOIDES HBK., Nov. Gen. & Sp. 2: 55 (quarto), 44 (folio). 1817.1. Typ: In arenosis, Cumana, Vene- zuela, Bonpland 406 (Herb. Mus. Paris ; fragment F!, photo- graph G!).—E. maritima Willd. ex Boiss. in DC. Prod. 15 (2): 28. 1862, in synonymy.—Chamaesyce ammannioides (HBK.) Small, Fl. Se U. S., 709, 1333. 1903. Chamaesyce Ingallsii Small, Fl. Se U. S., 708, 1333. 1903: Tyre: New Orleans, Louisiana, 1835, Dr. Ingalls (NY!, photo- graphs G!, W!).—E. Ingallsii (Small) Cory, Ruopora 38: 406. Glabrous annual; stems usually prostrate, 4-40 cm. long; mostly 1-2 mm. thick, internodes 0.5—4.5 cm. long; leaf-blades 4-15 mm. long, mostly narrowly to broadly oblong, occasionally elliptic-oblong, apex sometimes mucronate, usually obtuse, ase obtuse, inequilateral ; petioles 1-2 mm. long; stipules segments; peduncles short; cyathia solitary at the nodes; 10 volucre obconical-campanulate, 1.2-1.6 mm. in diam., glabrous g; : : 0.4-0. h diam., slightly cupped; appendages mostly narrower than the ‘See J. H. Barnhart, Dates of the “Nova Genera” of Humboldt, Bonpland -~ Kunth, Bull. Torr. Bot. Club 29: 585-598. 1902. The name, originally printed E. amannioides, corrected to E. ammannioides by HBK. op. cit. 7: 294 (quarto). 1825: 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 129 glands, rarely up to twice as wide; 5th gland linear, rarely of two parallel linear segments, 4% as long as to equaling the lobes; sinus U- to V-shaped, slightly depressed; bracteoles numerous, linear, often somewhat united below, glabrous below, short- hairy above, between and outside the staminate flowers, some adnate to involucre beneath the glands, a little shorter than the androphores; staminate flowers 5-16 per cyathium; androphores —2 mm. long, glabrous; fee Es pees soon exserted and reflexed; ovary ro oundly 3-lobed; styles 0.35-0.5 mm. long, thick, bifid to about the middle, mostly clavate; capsule 2-2.1 mm. long, broader than long, wider below the equator, strongly roundly to subacutely lobed; seeds ovoid sub- ge Syren 1.5-1.9 mm. long, 1.3-1.5 mm. tangent ially, hes reel a m. radially, coat mottled white, microreticulate—PLa Coastal sands: southern Virginia, orth Gaile, Florida west to Texas; epics South America (Map 39). Representative specimens seen: VrreintA. York Co.: York River northwest o ee Fernald & Long 7510 (G); York River above York- tow ernald & Long 12703 ao Surry Co.: inner border of ance of Cobham Bay, Jam 2 Rice northwest_of Chip- pokes, Fernald & Long 12705 (W). Isle ied Wight Co.: inner inner border of sand be ach along fetinen co Ragged Tetandl northeast of Carrollton, Fernald & Long 12704 (W). Princess - u (NY). Fioripa. Duval Co.: Pablo aa Bs ia! gt (NY). Volusia Co.: Mosquito Inlet, Moldenke 1 (NY). Dade Co.: beach opposite Miami, Small 2113 (NY). Lee Co.: seashore, Punta Rassa, A. S. Hitchcock 320 (F, G, M, NY). Manatee Co.: Palmetto, Nash 2448 (G, ne Per Tsland, oe 6376 (G, M, NY). Pinellas Co.: St. tersburg, Aug. 5, 1894, Lewton (NY). Mussissiprr. Jackson Gn. ‘fork arise Trdg 471 7 (NY); Dog Key, Tracy & Earle 2882 (NY). Harrison Co.: Cat Island, Lloyd & Tracy 207 (NY). LovistANa. Quenit rina Co.: Battledore Island, Lloyd & Tracy 214 G, M, NY). Jefferson Co.: south central Grande Isle, Cang & Andrus 1 (NY). Texas: Cameron Co.: Boca Chica, € Clover 356 (NY). Boissier in DC. Prod. 15 (2): 28. 1862 included E. bombensis Jacquin, Enum. Pl. Carib., 22, 1760 (not seen, but ed. 2, 1762 examined), Select. Stirp. Amer. Hist., 151. 1763, as a possible Synonym, judging by the description, of E. ammannioides. I draw the same conclusion. However, until some of Jacquin’s speci- mens of this species are located, I hesitate to use the name since 130 Rhodora [APRIL the description is too vague to apply to any one species with certainty. E. ammannioides presents a very puzzling problem, not only in its marked similarity to and close relationship with E. Geyert but also in its internal heterogeneity. Floridan, Mississippiaa, and Louisianan specimens examined present, with but two ex ceptions, a reasonably uniform unit characterized by very narroW appendages, styles markedly clavate, and staminate flowers generally 5-7. The two exceptional collections are: FLORIDA. Manatee Co.: Snead’s Island, Sept. 10, 1899, Tracy 6376 (NY). Dade Co.: Sand dunes opposite Miami, Mar. 7, 1915, J. K. & E. W. Small 5869 (NY). The first has 10-15 staminate flowers and appendages sometimes as wide as the glands. The second has the glands virtually obsolete on some involucres and the lobes strongly inflexed, making the involucre 5-lobed. The Vir- ginian specimens are rather stunted and can be fairly well matched in aspect, though not in technical details, by some specimens of E. Geyeri. The Texan plants, of which there are but two collections available, have 7, 9, 9, 15, and 16 stam! nee flowers per cyathium in the five counted. Clover 356 is particty larly marked in its general coarseness and its appendages uP ' twice as wide as the glands. It has the styles thickened below and slightly tapering. The fragment of the type of Euphorbia ammannioides has 15 staminate flowers in the one cyathium dis- sected, appendages from narrower than to a little wider than the glands, and thick, very slightly clavate styles. The type; from Venezuela, seems about intermediate between the Texan and Floridan plants. Heller 1231 from Virginia is Euphorbia polygonifolia in some herbaria. Fernald & Long 7510 was also probably E. polygon™ folia in part, for seeds of that species were intermixed. Professor Fernald informs me that at all stations in Virginia where E. ammannioides has been found E. polygonifolia is also present. The latter is paler in color and more inclined to occupy the lower border of the beach, near high-water mark, while EZ. ammannr oides characterizes the sands farther back from shore. 6. EvpHorsia Gryert Engelm. in Engelm. & Gray, Bost Journ. Nat. Hist. 5: 260. 1845. Typr: Beardstown, Cass ! Illinois, Aug., 1842, C. A. Geyer (M 47878!; photographs Gt 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 131 W!; rsotypes G!, NY!).—E. ce Engelm. var. microsperma A. Gra ay, an. Bot. No. U. , 386. 1856.—A nisophyllum yl cme tora Klotzsch & rere Abh. Akad. Berlin, Phys. 1859 —Chamaesyce Geyeri (Engelm.) Small, Fl. SE U. S., 708, 1388. 1903. E. polyclada Boiss., Cent. Euph., 10. 1860. Typ: ‘‘ Collection du Texas oriental, faite en 1848-49, recue en 1850,” C. Wright (Ge!; photographs G!, W!). The Texan phase with slightly more blunt seeds. --Chamacsyce polyclada (Boiss.) Small, Fl. Se. U. S8., 711, 1333. chateue poset stems ee mostly prostrate, occasionally sabe 6 to 37 ¢ m. long , 0.4-1.4 mm. thick, internodes up to 3 cm. g, average inca T m.; leaf blades oblong to ovate-oblong to sHiptio- obisng 4-10 mm. lon ng, margin entire, base — aoe obtuse or emarginate, usually mucronate; petioles 1-2 m : stipules distinct, or the ventral nae ed, glabrous, ca 3 long, mostly parted into 3 filiform segments; evailes solitary, most internodes; peduncles up to 2 mm. long; oe broadly campanulate, slightly tapering to the peduncle, 1-1.1 mm. in diam., glabrous outside, with a few hairs at the eee inside; lobes triangular, acuminate, tips glabrous, slightly exceeding the glands; pe broadly ov val to subr otund, slightly cupped to folded, 0.2-0.6 mm. long; appendages aie , from one wice as wide as the gland, rounded o bietinoes pointed, entire or —_ toothed; fifth gland linear, pecan below inside, 24 as ng as, to equaling the lobes; s U-shaped, not depressed; beatin forming a radial ictiten peace each gland, adnate to the involucre nearly up to the base of the stipe of the gland but with inwardly ee ing linear segments scattered along the inner edge, segments linear, bearing a few to many hairs at the n side each fascicle; staminate flowers 1-5 per fascicle 5-17 per cyathium; androphores ay aaa 1—1.3 mm. long; gynophore gla- brous, soon exserted and reflexed; wen © 3-lobed; styles 0.3-0.6 mm. long, 44-4 bifid; capsule 2 mm. long, roundly ‘and deeply 3-lobed, wider below the equator, Sane truncate; seed ovoid-sub- opintaher od ovate to narrowly ovate radially, 1. 3-1.6 mm. long, tangentially and radially, coat smooth, microreticulate, white but the brown testa often showing thr ough. —PuaTE 654C. Sandy barrens or dunes mostly along rivers, Wisconsin, Illi- hes irae Iowa, North Dakota, Nebraska, south t to northern Texas, west to eastern Colorado and New Mexico (Map 20). Representative specimens seen: WISCONSIN. Pepin Co.: open dunes, upper terrace, Pepin, Fassett 10260 (G). Trem- pealeau Co.: Trempealeau, Fassett & Wilson 5808 (G). La 132 Rhodora [ APRIL Crosse Co.: La Crosse, 1861, Hale (M). Lancaster Co.: Boseo- bel, 1861, Hale (G, M, NY). Invurnors. Lee Co.: Dixon, Vasey G). Henderson Co.: banks of Mississippi River, near Oquawka, Sept. 18, 1871, Patterson (G). Mason Co.: Havana, Aug. 18, 1904, Gleason (G). Minnesora. Anoka Co.: Moore Lake, sand Dakota. M Stevens & Kluender (G). Nespraska. Holt Co.: Paddock, Clements 2784 (G, NY). omas Co.: Dismal River, sand hills, Rydberg 1504 (G, NY). Brown Co.: Long Pine, Aug. 13, 1898, Bates (G). Kansas. Hamilton Co.: sand hills, Hitchcock 466 (G, M, NY). Oxzanoma. Payne yen sandy woods, Oliver 137 (NY). Creek Co.: Sapulpa, Bush 52 (G, M, NY). Texas. Dallas Co.: Dallas, 1878, Reverchon (M). Parker Co.: Weatherford, Tracy 7860 (G, M, Mo, NY). Anca Co.: Shafter Lake, Cory 16614 (W). New Mexico. Roose Co.: 2 miles west of Bledsoe, Cochran Co., Texas, Cory te (W). Conorapo. Morgan Co.: sand hill 4 miles south of Brush, Ramaley & Ewan 16326 (W 2: 7. EUPHORBIA MissuRICA Raf., Atlantic wot 1: 146. 1832; 3: 10.1 linear segments from 24 as long as to about equaling the pie pubescent below; sinus U-shaped, little or not at all depress je bracteoles from partly united below into a radial partition 4 3-lobed; styles ca. 4 bifid, 0.7-0.9 mm. long; capsule strongly 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 133 3-lobed, 2-2.5 mm. long, from about as broad as long to gon than long; seeds ovoid to broadly ovoid-triangular, 1.5- long, 1.1-1.4 mm. tangentially and radially, rotundly Sone to ovate-acute radially, angles faint to evident but blunt, coat mottled whitish and bro Key To VARIETIES Leaves broadly oblong to linear, apex obtuse to slightly retuse; ultimate branchlets mostly more os = capillary; seeds scarce- ly angled, micropylar area not m Ae mag and trun- eatee peduncles up to 11 mm. nk CAPIUBTY 05264 6s a. var. lypica. Leaves linear, apex truncate, mostl pee araia nate; ultimate branchlets not capillary; seeds mostly definitely angled, mi- saree area obliquely truncate; peduncles up to 4 mm. long, WOR ic oa he yikes Sain cea es aN ee ee b. var. intermedia. 7a. E. missurtcaA Raf., Atlantic Journ. 1: 146. 1832, var. typica. Typr: ‘Arkansa’’, Nuttall eae Big G!, W!). A fairly typical, though young, specim E. zygophylloides Boiss., Cent. Euph., 10. 1860. Type: None of the original specimens from hich Boissier drew his descrip- tion has been seen. Cons equently it seems inadvisable to select any type. The interpretation is eh since two of the numbers cited by Boissier have been seen at M.—Chamaesyce zygophylloi- Se .) Small in Britton & ook Ill. Fl. No. U.S. & Canada 161. 191 E. arenaria Nutt., gets preis erg Soc., n. 1837, not H.B.K., Nov 57 (quarto), 45 “Golio). 1817. TYPE: Akane”, | Nuttatleh Phi, perhaps only isotype; oe G!, W!). Engelm. & Gray, Bost. Journ. Nat. Hist. a ste (Pl. Lindh. Y ook: id hag N a Engelm. in Em U. 8S. & Mex. Bound. Surv. 2 ‘wD: 1859.—-E. N uttalli Bgelm,) Small in Britton & Brown, in. TL No. U.S. Ca sae nlp a ca Nuttallii (Engelm. ) Small, Fl. —PLaTE 666B Missouri, northw ae peated as, Kansas, south to Texas (Map 13). Representative specimens seen: Missouri. Barry County: Eagle Rock, Bush 76 (M, NY). Cass Co.: July, 1865, Henry Co. mark 15972 (M, Mo). fan ane Co.: Greenwood, Bush 10336A » NY); Dodson, Bush 506 (M, NY); Cockrel, Bush 6487A (M); Westport, Sept. 13, 1896, K. K. Mackenzie (M, Ph). Stone : along east side of White River, south of fisoth of Big Creek, sioeth of Mill Creek, Steyermark Bn (M, Mo). ARKANSAS. pi Co.: Beaver, E. J. Palmer 33 (M). ” Kansas. Ellis Co. : near Hays, Bondy 291 (M, Mo). Geaey Co.: Junction City, A. Brown 178 (NY). Miami Co.: Paola, Oyster 7306 (NY). Osage Co.: Osage City, fae: 1890, Bodin (NY). Riley Co.: Stony hills, Norton 468 (M, N Y). OKLAHOMA. Comanche Co.: Fort Sill, 134 Rhodora [APRIL Clemens 11671 (M). Johnston Co.: Tishomingo, E. J. Palmer 6426 (M). Kiowa Co.: Wichita Mountains, July 27, re Sheldon (M). Love Co.: Marietta, E. J. Palmer 10406 ( } Murray Co.: Platt National Park, Bromide Hill, G. M. — 676 (NY). Ottawa Co. eo Miami, G. W. Stevens 2344 ( , Pawnee Co.: Cleveland, E. J. Palmer 6879 (M). Payne Co.: 18 files eoutheaat of ieationstan, Stratton 206 (M). TEXAS. Austin Co.: Industry, het 1844, Lindheimer (M, Ph). bas Co.: near Bracken, Groth 55 (N Y). Burnet Co.: 3.1 miles OME of Burnet, Cory 15617 Suk Dallas Co.: Dallas, Bush 1140 ( N y): Gilles pie Co.: Squaw Creek, Jermy (M). Grimes o> Anderson, G. L. ihe 3791 7 (F). Hays Co.: San ae ange (NY). Houston Co.: and, E. J. Palmer Grapel M). Parker Co.: Weatherford, Tracy 8124 (M, Mo, NY, Ph). Tarrant Co.: Fort Worth, Tracy 8168 (M, Mo, NY). Travis re pts Tharp 2854 (U 8). Washington Co.: Mill Creek, Lt heimer 186 (M). The Nuttallian rather than the James specimen is taken as type of E. missurica because the latter is a sterile seedling while the former had at least one cyathium from which Torrey drew the essential characters. E. zygophylloides var. cymulosa Engelm. ex Boiss. in DC. Prod. 15 (2): 29. 1862. This probably belongs here but the only ap- parently authentic material seen bore this varietal name but was placed under another species. The intergradation between var. typica and var. intermedia complete and there are plants which cannot be definitely assig™ to either. Examples of such intergradation are the following collections: 7b. BUPHORBIA MISSURICA what var. pavers ig na eeler, Proc. Biol. Soc. Wash. 53: 1940; based M petaloidea Engelm. attees Engelm. in hairy, _&M le ound. Surv. 2 (1): 1859. Typr: Fort Pierre, Standley mies South Dakota ae 21, 1853, F. V. Hayden (M 14 photographs G!, W!) E. petaloidea Engelm., l.c. Typ: Forks of the ier Riv tineca County, Nebraska, July, 1858, He abd Engen 28. 149966!; photographs G!, W)). Boiss. in DC. Prod. 15 2): 38 1862; Small in Britton & Brown, Ill. Fl. No. U. g. 7 Canada all, 371. 1897, seeds poor. —Chamaesyce petaloidea (Engelm.) Sm Fl. Se. U. S., 711, 1333. 1903. 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 135 E. petaloidea « Nicolletii Engelm. in Emory, U. S. & Mex. Bound. Surv. 2 (1): 185. 1859. Typr: mouth of Powder River, Yellowstone, Prairie County, Montana, July, 1854, F. V. Hayden (M 202539! Minnesota, west to eastern Montana, south to Texas and New Mexico (Map 13). i ee aren an seen: MINNESOTA. Ottertail ie Clitherall, July, 1 J. E. Campbell (M, Mi, US); sandy beach of Ottertail “seth ape 2333 (Mi); sandy shore of Battle Lake, Rosendahl 4177 (Mi). Norta Dakota. Slope Co.: Bad Lands, Marmarth, L. R. Moyer 741 (Mi). Sourx Dakota. Fall River Co.: Hot ‘Springs, Hayward 546 (NY). Potter Co.: Forest City, Sept., 1892, T. A. Williams (M). Nesraska. Banner Co.: Rydberg 353 (NY). Brown Co.: Long Pine, Conklin (NY). Cherry see Valentine Lakes Refuge, Aug. 10, 1937, Tolstead tac Custer Co.: Anselmo, Webber 11 (NY). Dawson Co. : Nea 7» othe ee Heller 14300 '(M). Deuel Co.: July, 1890, Rydberg (NY). Hooker Co.: on Middle Loup River, near Mullen , Rydberg ee ies Scott’s Bluff Co.: Kiowa Valley, Rydberg 3853 (NY). Sioux Co.: Kramer 152 (M). Thomas .: Dismal River, Webber r 2 (NY). Kansas. Clark Co.: 10 miles south of Ashland, Rydberg & Imler 747 (NY). Graham Co.: Bogue, Imler 63 (M, NY). Grant Co.: Ulysses, C. H. Thompson 60 (M, NY). Hamilton Co.: Syracuse, Rose & Fitch 17099 (NY). Kearney Co.: 13 miles south west of Lakin, (M, NY). Osborne Co.: ear weld Shear 126 (NY). Reno Co.: Hutchinson, Smyth 45 (US). Riley Co.: Norton 467 (M, NY). OKLAHOMA. Alfalfa’ Co.: near Geckos G. W. Stevens 622 (M). Beaver Co.: north of Beaver, E. J. Palmer 41887 (M Cleveland Co.: Norman, Emig 525 (M). Harper Co.: north of Rosston, G. J. Goodman 2197 (M). Payne Co.: 18 miles south- east of Stillwater, ape 656 (M). Trxas. Bailey Co.: 2 miles south of Muleshoe, Ferris & Duncan 3422 (M, NY). Callahan Co.: Baird, E. J. pal liner 14542 (M). Eastland Co.: 31% miles east of Ranger, Cory 13130 (W). Grayson Co.: Denison, Stuart 147 (M). ll Co.: Estelline, Heverchon 8799 a= Hartely Co;: 5.3 miles uke of Middlew er, ni Pe 16313 (W). Hemphill prairie north of Canadian, rene 1900, og ert (M). Hood : Granberry, a 3797 GD. ’ Mitchell Co.: Colorado, Pratig 8121 (M, Mo, NY). Motley Co.: 16.4 miles north of Matador, Cory 16037 (W). Wilbarger Co.: Pease River near Vernon, Ferris & Duncan 33538 (M). Wyomina. Natrona Co.: Alcova, Goodding 161 (M, NY). Cotorapo. Boulder Co.: Boulder, Penard 301 (NY). Cheyenne Co.: $1, T148, R51W, Owenby 1850 0 (M mete Denver Co.: Denver, Jones 292 (M). Huerfano Co.: La V eta, Clements 152 (N Y). Morgan Co.: 4 miles south of Brish, Ramaley & Ewan 16825 (W). Weld Co. 136 Rhodora [APRIL Crow Creek, Pollard 92 (NY). New Mexico. Chaves Co.: 20 miles south of Roswell, F. S. & E. S. Earle 279 (M, NY). Lea Co.: sandhills near Loving, Standley 40862 (US). Union Co.: on the Cimarron, Wislizenus 464 (M). 8. EupHorpia Gracituima S. Watson, Proc. Amer. Acad. Arts and Sci. 21: 438. 1886. Type: Hacienda San Miguel, near Batopilas, southwestern Chihuahua, Mexico, Aug., 1885, Hd. Palmer 68 (G!). (Watson published the locality as ‘“ Hacienda San Jose”? which seems to have been a lapsus calami.)—Chamae- syce gracillima (S. Wats.) Millsp., Field Mus. Pub. Bot. 2: 409. 1916. Glabrous erect annual 8-18 cm. tall; main stem up to 2 mm. thick at base, soon forking repeatedly into branches with inter- nodes up to 2 cm. long below, progressively shorter toward the tips, ultimate branchlets ca. 0.1 mm. thick; leaves 2-15 mm. long; blades narrowly linear, margin entire, revolute; petioles 0.3-0.6 mm. long; stipules entire, distinct, glabrous, linear- subulate, 0.3-0.5 mm. long; cyathia solitary in the bifurcations; » peduncles 0.3-0.9 mm. long; involucre turbinate, tapering to the peduncle, 0.5-0.7 mm. in diam., glabrous outside, with short he : scarcely obscured, with longitudinal rows of reticulations— PLATE 667A. Arizona, south to Sinaloa, east to the Pacific slope of Chihua- So (Mar 10). Representative specimens seen: ARIZONA. Pima 8142 (A, US). MEXICO. Sonora: Bajada south of Las i cheras, Shreve 6375 (Sh); hills and mesas near Altar, Aug- oak 1884, Pringle (G, US); San Bernardo, gravelly soil, alt. 250- ; m., Pennell 19730 (US); Hermosillo, M. E. Jones 22604 (possibly #1605) (G); Chorijoa, Rio Mayo, in the sand, river bank, Gentry 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 137 1608 (G). Srnauoa: Culiacan, Sept. 8, 1904, T. S. Brandegee (G); Culiacan, Oct., 1904, T. S. Brandegee (US). 9. EUPHORBIA REVOLUTA Engelm. in Emory, U. 8. & Mex. Bound. Surv. 2 (1): 186. 1859. Typr: detween Santa Fe and Moro (Mora?) River, in the mountains about the base of trees and shrubs, New Mexico, Aug. 10-16, 1847, A. Fendler 789 (M 200216!; photographs G!, W!; 1soryPE G)). Boiss., Icon. Euph., t. 13. 1866. -, Chamaesyce revoluta (Engelm.) Small, FI. SEU. 8., 711, 1383:..1903. Glabrous erect annual 3-20 cm. tall: main stem up to 2 mm. thick at the base, soon josie repeatedly into branches with internodes up to 3 om. lon g below, ipanees rie shorter toward the tips, ultimate breaeobietet 15-0. niin thick; leaves 2.5-26 mm. long, largest leaves on any plan ripe tly over 1 cm. blades narrowly linear, margin entire, revolute; petioles 0.5—-1. 5 m. long; stipules entire, distinct, glabrous, linear-subulat attenuate, 0.3-0.8 m one ng; cyathia solitary i in the hacen. peduncles. 0.5-1.4 mm. long; involucres ee obconical to turbinate, tapering ro ‘the peduncle, 0.9-1 m n diam., gla- brous outsi e, with short hairs around cg pecan inside or glabrous; lobes triangular, acuminate, ey ee markedly exceeding the glands; glands subcircular, 0.15-0.3 n slightly cupped; appendages from a mere aang: beneath the gland to radially elongated and nearly as long as the gland; 5th gland linear, ca. half as long as the lobes; sinus V-shaped, little depressed; bracteoles linear, entire, or bifid, with a few short hairs above or glabrous, ca. 24 as long as the androphores, one united to the involucre pagenr each gland; staminate — 1-3 per fascicle, 5-10 per cyathium; androphores 0.7-0.9 m long, glabrous, include : api cor aia glabrous, soon eS and usually reflexed ; ovary 3-angled, glabrous; styles 0.5 tips; capsule glabrous 1.3-1.4 mm. lo ong, sharply pated slightly wider below the equator, base eee gene triangular- pyramidal to ——— sharply angled, 1-1 m. long, 0.9 mm. tangentially, 0.6-0.8 mm. radially, eaially: ovate-acute with truncate base, ventral facets concave and nearly smooth or face rounded, traversed by two (rarely 1) transverse rounde ridges, angles sharp, the ridges and valleys passing through them but slightly, or not at all, coat white, sgt ha aa the brown testa sometimes showing through. —Pia Bi olorado, New Mexico, Chihuahua, ia Arizona (Map 8). Representative specimens seen: CoLorapo. Fremont County: Canyon City, Clements 272 (NY): ; rocky hills near Canyon City, Biltmore Herb. 6415 (US). New Mexico. San Miguel-Guadalupe County: between Anton Chico and Las Ve egas, Rose & Fitch 17638 (M, NY, US). Sandoval-Bernalillo County: Sandia 138 Rhodora Mountains near La Luz Mine, Ellis 335 (US). Sante Fe C Santa Fe, Rose & Fitch 17699 (M, NY, US). Lincoln Gray, Skehan 118 (NY, US). Sierra County: 1 mile west Hillsboro Foomenyed 1267 G, M, NY, US). Dofia Ana Leiberg 5969 (US). Yavapai en Clarksdale, W. aa J . C287 (G). Pima County: Davidson’s Canyon, Sept. oo Pringle (US). Gila County: Pinal Creek, Harrison & Kearn 8291 (G). ——— Cruz County: Nogales, Peebles, Harrison - Cochi i Garcia in the Sierra Madre gate & Barber + 284 (M, hills near Chihuahua, Pringle 999 (M); 6 miles east of Pilonet! road from Jimenez to Camargo, via El Arroyo del Fierro, i Johnston 7874 (G). a Chas. Wright no. 1830 would have been taken as the type ¢ lection, as it was more widely distributed in herbars i Fendler 789, if it were not for the fact that Asa Gray, in up Wright’s sets, combined two of Wright’s collections 1830. This fact j is inferred from the presence in the pocket one of the two sheets of this no. 1830 at the Gray Herbarium two of Wright’s original collection-numbers. Omitting tails of the method of elimination, the data for the pune | “298. Stony hills at the Copper Mines, Aug. 18, 1851”, Mexico; and “524. On the Sonoita near Deserted Rancho, § 15, 1851 ”, Santa Cruz County, Arizona. 10. Evrnorsta riortpa Engelm. in Emory, U. 8. & | Bound. Surv. 2 (1): 189. 1859. ~sdl Valley west of the Cl Frat 18 Mountains i i us erect annual 15-65 ¢ bag — simple, 1-6 1 mm. thick at she — . graduall Brom ggeen toward tips; leaves linear, 1 fgets em, long or the Soe ne : 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 139 linear lobes near the base, glabrous or sometimes with a few cilia; cyathia solitary at the nodes or somewhat clustered at the branch-tips: peduncles up to 6.5 mm. ong but mostly much shorter; involucre campanulate, 1.4-2 mm. in diam., glabrous without, densely pubescent above inside; lobes slightly exceeding the glands, triangular, acuminate, entire or the distal with one or Ge erect teeth o n the sides and the proximal sometimes many-toothed; glands circular or nearly so but often folded to- gether, 0.5 mm. in diam.; appendages conspicuous, white, often rubescent, obawat to elliptic to ee ee (radially) 1-2.8 mm _long, glabrous; 5th gland filiform, ca. 24 as long as to nearly equaling the lobes; sinus helia ee slightly depressed ; bracteoles forming a radial partition opposite the glands, adnate to the involucre for about half its length, united to the gynophore at the very base, upper half of 3-6 filiform ciliate segments equaling the androphores: staminate flowers 9-12 per fascicle, 37-58 per cyathium; androphores 1.9-3 mm. long, glabrous, included; gynophore glabrous, soon exserted and reflexed; ovary glabrous, roundly 3-lobed; styles bifid 4-34 distance to the base, 1-1 mm. long; capsule glabrous, oblate-spheroidal and nearly sym- metrical, rounded-triangular i in horizontal cross section, ca. 2. mm. long: seeds subtetragonal, lateral sibel dorsal kaneis rica winged, ventral angle very low, 1. long, 1.3-1.6 m radially, 1.4-1.7 mm. tangentially, a alle: fonds ovate, eee truncate, facets smooth se i for two — ly three on dors al facets) low transverse ridges which do ass through the angles, occasionally the rides irregular; white gee wera micro- reticulate coat mostly thick enough to make the seeds a dull white to sordid —PLatr 667C outhern Arizona, south to Sinaloa (Map 17). me poner specimens seen ZONA. Yava Co.: Fo ipple, Rio Verde, Coues & Palmer 531 (M). Gila Co.: Rock Creek, Collom 314 S) m Co.: Camp Grant, Ed. Palmer 237 (M) 1 O.: near n, es, Harri & aboquivari Canyo e Kearney 2773 (M, US); foothills of “fom Catalina Mountains, July 30, 1881, C. G. Pr ringle (M, US); Baboquivari Mountains, M. E. Jones 24862 (G, NY); Santa Rita Forest Reserve, Griffiths & Thornber 8 (NY, US); small range reserve near Tucson, mph gre 6172 (US); common on range reserve, Wilmot, Thornber 6 (M). Cochise Co.: San Bernardino Ran ch, Mex. Boundary nee Mearns 611 beg Rucker Valley, Lemmon 469 (G); ns M. E. Jones 4263 (I, NY, US). MEXICO. Sonora: arroyo, Conejos, Rio Mayo, Gentry 1119 (G, W); La Cruz de las E haias. Lloyd 463 (G); high mountains, Guaymas, Sept., 1887, Palmer 209 (G, US); Alamos, Palmer 640 (G, US); San Bernardo, Rio Mayo, Gentry 1148 (Sh); 5 Zits east of Garumbu llo, Wiggins 6124 (US). Stnatoa: Cofradia, vicinity of Culiacan, Oct. 21, 1904, T. S. Brandegee (G). 140 Rhodora [APRIL 11, KupHorsia TRACHYSPERMA Engelm. in Emory, U. 8. «& Mex. Bound. Surv. 2 (1): 189. 1859. Typr: low damp soil near the San Pedro River, Cochise County, Arizona, Sept. e pe Wright 1832 (M 200493!; photograph G!, W!; 1soryPss Gl, e : —Chamaesyce trachysperma (Engelm.) Millsp., Field Mus. Pub. 1916 into 3 or 4 linear, shortly hairy segments about equaling is androphores; staminate flowers 10-13 per fascicle, 54-6: Ra: cyathium; androphores glabrous, 2-2.8 mm. long; ayn i: glabrous or rarely with a few hairs, om exserted and refiexe¢, truncate, apex obtuse; Fally, dorsal and lateral angles sharp, ovate to narrowly ovate de base truncate, apex acute, 2-2.3 mm. long, tangentially i coll mm., radially 1.3-1.6 mm., ventral facets plane or slightly be cave, dorsal facets convex, surface of seed slightly roug yr scattered irregular pits, angles irregularly notched, or “6670. facets with two or three faint transverse grooves.—PLATE saeitl Southern Arizona, south to Sonora (Map 21). Repreese™ 1. specimens seen: Arizona. Yuma Co.: west of La Paz, 186 ( an enclosure, Guaymas, Oct., 1887, Ed. Palmer 319 (G, US); (G). UPHORBIA HYSSOPIFOLIA L., Syst. Nat. ed. 10, 2: ere 1759. Type: Probably from Jamaica, Patrick Browne (Linna 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 141 Herb., not seen; photograph G!; rephotograph W!). The narrow-leaved extreme. Thellung in Ascherson & Graebner, Syn. Mitteleur. Fl. 7: 432. 1917.—Anisophyllum hyssopifolium (L.) Haw., Syn. Pl. Suec., 161. 1812; Klotzsch & Garcke, Abh. Akad. Berlin, Phys. 1859: 35. 1860.—E. hypericifolia L. ‘@ hys- sopifolia, L.”’, Griseb., Fl. Brit. W. I. Is., 54. 1859; published as quoted.—E. brasiliensis Lamarck var. hyssopifolia (L.) Boiss. in DC. Prod. 15 (2): 24. 1862.—Chamaesyce hyssopifolia (L.) Small, Journ. N. Y. Bot. Gard. 3: 429. 1905. E. brasiliensis Lamarck sensu Boiss. in DC. Prod. 15 (2): 24. 1862 and most subsequent authors.—Chamaesyce brasiliensis (Lamarck) Small, Fl. Se. U. S., 712, 1333. 1903, sensu Small, |. c., and subsequent publications in part. E. Jonesiit Millsp., Pittonia 2: 89. 1890. Typx: Bowie, Cochise County, Arizona, Sept. 17, 1884, M. E. Jones 4247 (F 196592!; photographs G!, W!; 1sorypEs O!, US!). The large- seeded western facies of the species—Chamaesyce Jonesii (Millsp.) Millsp., Field Mus. Pub. Bot. 2: 410. 1916. E. stenomeres Blake, Contr. U. 8. Nat. Herb. 24: 13. 1922. Type: Open hillsides wooded with pine, trail from Los Amates to Izabal, Department of Izabal, Guatemala, May 31, 1919, S. F. Blake 7776 (US 989628!, photograph G!, W!). Appears to be a good match for the type of E. hyssopifolia L. nnual or sometimes perhaps overwintering; stems mostly erect, mostly glabrous, sometimes sparsely pilose in Arizonan plants, simple below or sometimes branched from the base, 8-60 cm. tall, mostly 1-3 mm. thick, internodes mostly 1-3 cm. long, rarely up to 6; leaves 5-30 mm. long, lanceolate and often falcate to oblong, all leaves often small in over-wintering plants, mostly glabrous, sometimes sparsely pilose especially toward the base, base inequilateral, margin usually serrate; petioles 1-1.5 mm. long; stipules mostly united, for the most part triangular and as broad as high with slightly lacerate and occasionally ciliate margin, toward the branch-tips sometimes partly distinct, narrowly triangular, up to ca. 1 mm. long, with erect linear teeth; cyathia solitary in the upper bifurcations and in few-flowered eafy cymes; peduncles 0.5-2 mm. long, glabrous; involucres ob- conical to obconic-campanulate, tapering to the peduncle, 0.7— 0.9 mm. in diam., glabrous outside, glabrous inside except at the base of the lobes and on the stipes; lobes triangular, attenuate, slightly exceeding the glands, distal mostly entire, proximal often With erect linear teeth; glands circular to broadly elliptical transversely, 0.15-0.3 mm. long; appendages white or some- times reddish in age, glabrous, entire, 144 as wide as to twice as wide as the gland, semilunate; fifth gland linear, ca. 4% as long as the lobes; sinus U-shaped, slightly depressed; bracteoles form- ing a very narrow radial partition adnate to the involucre below 142 Rhodora [APRIL each gland, free portion parted into few linear subglabrous segments; staminate flowers 4-15 per cyathium; androphores glabrous, 0.9-1.3 mm. long; gynophore glabrous, well-exserted and usually reflexed; ovary glabrous, strongly 3-lobed; styles glabrous, 0.5-0.9 mm. long, bifid 14-24 to base; capsule glabrous, broadly ovoid, wider below the equator, 1.6-2.1 mm. long, strongly and subacutely 3-lobed; seeds ovoid-subquadrangular, 1—-1.4 mm. long, 0.7—1.1 mm. tangentially and radially, radially ovate to elliptical-ovate, ventral facets slightly concave to slightly convex, dorsal facets convex, both with very shallow depression from half as wide as to as wide as the facet, separated by very low smooth ridges, chocolate-brown to grayish-white.—— PLATE 656D. South Carolina, Florida, Louisiana, western Texas, southern ew Mexico, and Arizona; generally distributed south to tem- perate South America (Map 38). Representative specimens from the United States: Sour CaroLina. Colleton Co.: 3 miles south of Walterboro, Wiegand & Manning 1817 (G). FLORIDA. Alachua Co.: campus, Gainesville, O’ Neill 458 (US). Lake Co.: near Eustis, Nash 930 (G, M, NY, US). Orange Co.: Orlando, Curtiss 6670 (G, M, US). Dade Co.: Miami, Demaree 10214 M, ; Arch Creek Prairie, J. K. Small, Mosier & G. K. Small 6800 (NY, US); Buena Vista, Moldenke 328 (M, NY, US). Lee Co.: Myers, A. S. Hitchcock 324 (G, M, NY, US); near Mountains, Mulford 10374 (M, NY). A Camp Lincoln, 1869, Ed. Palmer (US). Maricopa Co.: Mesa, 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 143 & Kearney 8045 (US). Santa Cruz Co.: Nogales, Peebles, Harri- son & Kearney 4590 (US); Nogales, Peebles & Harrison — (US}i Ge chise Co.: near Fort Huachuca, Wilcox 320 (US Portal, Chiricahua wba Eggleston 10725 (US); Fort Lowell, Phen 209 (M, NY). E. brasiliensis Lamarck, Encye. Meth. Bot. 2: 423. 1786, has generally been considered as conspecific with EH. hyssopifolia L. However, the ovary and capsule of E. hyssopifolia and what has been known as E£. brasiliensis are noted for being quite glabrous. Examination of two fragments of FE. brasiliensis at F, one sup- posedly from the type, collected by Commerson in Brazil, and the other from an isotype both at Herb. Mus. Paris, shows that Lamarck did not err when he wrote “‘ L’ovaire est chargé de poils blanes.”’ The hairs are mainly on the backs of the carpels and seem to be at least partially deciduous. The rather young seeds do appear to be identical with those of E. hyssopifolia. The type collection of EZ. brasiliensis may represent merely a trivial variant of E. hyssopifolia but whatever the entity, it does not occur in the United States or Canada. The plants of Texas, New Mexico, and Arizona are generally larger-seeded than those of South Carolina, Florida, and Loui- siana. The western plants also tend to be slightly vestite. The absence of the species in southern Texas and its presence in western Texas further suggest that there are two distinct races involved. However, since the differences are so trifling when many specimens are examined and Mexican specimens seem to be intermediate, it seems futile to try to separate as named categories the northern tips of two lines of northward migration of a widespread and variable tropical species. Examination of - the total variation of the species is sasiea before a proper evaluation of any of its facies can be ma 13. EvPpHORBIA MACULATA L., Sp. Pl. 1: 455. 1753. Type: “America septentrionalis’’ (Linnaean Herb., not seen; photo- graph G!; rephotograph W!). A good representative of the species as here interpreted. Moncoue eras he Altera, 392. 1771; Elliott, Sketch Bot. So.-Car. & Geo gia 2: 654. 1824, and probably other early writers.—E. hypericifolia L. 6. BE. maculata (L.) Lam., Enye. Meth. Bot. 2: 422. 1786.—Tithymalus macu- pom (L.) ‘Moench. Meth. Pl., 666. 1794.—Anisophyllum macu- um (L.) Haw., Syn. Pl. Suce., 162. 1812.—E. hypericifolia L. i maculata (L.) Raf., Med. Fl. U. S. 1: 183. 1828 (basinym 144 Rhodora [APRIL given above in synonymy, and the plant described belongs here). Xamesike maculata Raf., Aut. Bot., 97. 1840, basinym indicated only as ‘‘Euph. d[itt]o. O.[mnes].”’—Z. trinervis Bertoloni, FI. It. 5: 37. 1842, a renaming of ‘‘E. maculata Mant. alt. p. 392” under the misapprehension that Linnaeus’ use there was a misapplica- tion.—Chamaesyce maculata (L.) Small, Fl. Se. U. S., 713, 1338. 1903. e E. nutans Lag., Gen. et Sp. Pl., 17. 1816. Type: “Habit. [at] in N.{ova] H.[ispania]” (Perhaps at Madrid judging by A DC., Phytographie, 426. 1880). See discussion after citation 0 specimens.—Chamaesyce nutans (Lag.) Small, Fl. Se. U. 8. 712, 1333. 1903. E. Preslii Guss., Fl. Sic. Prod. 1: 539. 1827. Typs: Palermo, Italia, Todaro (Praha?; fragment F!). h material at best but marked in Millspaugh’s hand “From the type material Herb. Praag.’’. The specimen belongs in — entity. Boiss. in DC. Prod. 15 (2): 23. 1862.—Chamaesyce Prestt 912. stands for Euphorbia. E. hypericifolia L. sensu American authors in large pee Many extralimital synonyms listed by Boissier in DC. : 15 (2): 23. 1862, and Thellung in Ascherson & Graebner, 5Y?- Mitteleur. Fl. 7: 426~432. 1917, are omitted here since no. authentic material has been available. mi nnual; stems mostly erect, mostly simple below, 8-80 sak tall (or taller according to collector’s notes), 1-4 mm. yee internodes mostly 1-4 em. long, crisply pubescent at least 0D e yi g oblong-lanceolate to oblong or even occasionally falcate- lane late, 8-35 mm. long, mostly glabrous above, sometimes sparse base, base inequilateral, margin serrate; petioles 1-1.5 mm. long; stipules mostly united, mostly broadly triangular-acv" ong, margin ciliate, slightly toothed uppermost often nearly distinct, narrowly triangular, ma esirepunnomeanyrnese nee ON TIE EERE LATE SESE . 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 145 with erect linear teeth; cyathia both solitary at the nodes and clustered in cymes; peduncles 0.5-5 mm. long, glabrous; involu- cres glabrous outside, sparsely pubescent inside on the bases of the lobes and stipes, obconical to obconical-campanulate, taper- ing to the peduncle, 0.7—1 mm. in diam.; lobes triangular, slightly exceeding the glands, the distal mostly entire, the proximal often with erect linear teeth; glands she Deh ak circular to broadly oT rede, 0.1-0.3 mm. in dia appendages rudi- mentary to 0.5 mm. long, oval, white to morally glabrous, entire or bluntly wes fifth gland linear, ca. 24 as long as lobes; sinus U-shaped, slightly depressed; bracteoles forming a slender seg- ment below each gland, united to the involucre below, entire or with two to four filiform pet niet or sparsely hairy divisions above; a ep flowers 5-11 per cyathium; androphores glabrous, 1—1.1 mm. long; gynophore glabrous, well-exserted and usually ellexad: > glabrous, roundly 3-angled; styles 44—-\% bifid, 0.6-1 mm. long; ae glabrous, broadly ovoid, wider below the Reber 1.9-2.3 mm. long, strongly and subacutely xf ohtaken seeds ovoid-subquadrangular 1.1-1.6 mm. long tangentially and radially, elliptic-ovoid vatlaliy, ‘facets ae with a finely rippled surface, coat dark grayish, brown or some- times pale gray.—P.LatsEs 656B an Ontario; all states east of the Mississippi River except Maine; Minnesota and South Dakota south to northern Florida and Texas; introduced in California and Wa shia eton: southward to South America; and introduced in the Old World (Map 10). Representative specimens seen from the United States and Canada: New Hampsuire. Belknap Co.: Meredith, Aug., 1894, Carter (NE). Strafford Co.: meg Dover, Hodgdon 677 (NE). Hillsborough Co.: Milford, Sept. 5, 1902, J. A. Wheeler (NE). VERMONT. Rutland Co.: Brandon, Sept. i, 1909, Duin (M); Wallingford, Aug. 3, 1907, Kennedy (G). MASSACHUSETTS. Middlesex Co.: Concord, Sept., 1857, Hoar (NE); Somerville, Aug. 31, 1899, Flynn (NE). Hampshire Co.: South Hadley 1887, Cook (US). Berkshire Co.: Pittsfield, Sept. 23, 1899, Hoffmann (NE). Connecticut. Hartford Co.: ghee Fp Bissell 529 (NE). Litchfield Co.: North Canaan, Aug. 17, 1910, ae (NE). Middlesex Co.: Middletown, Blewitt 850 (NE). w Haven Co.: Oxford, Sept. 28, 1916, Harger (Ph). Fairfield Co.: Fairfield, Aug. 1, 1893, Eames (Ph). New York. Orange Co.: Newburgh, Sept. 8, 1895, Pollard (US). West- chester Co.: Worthington, panei 828. 2 (P h). Warren Co.: Lake George, Sines Vasey (US). Tompkins Co.: Ithaca, July 20, 1893, Wiegand (US). Dutchess Co.: — Clove, Standley & Bollman 12243 (US). New York Co.: nx vee ago 3, 1896, Nash (Ph). Suffolk Co.: Wading as “Miller 1 9 (US). New Jersey. Middlesex Co.: Stelton, Mackenzie 9867 (M, US). 146 Rhodora [Apa Somerset Co.: Peapack, Perry (M). Hunterdon Co.: Stockton, on Delaware River, Benner 7008 (PhB). Ocean Co.: near Bay Head, Sept. 21, 1914, Lighthipe (M). Camden Co.: Oaklyn, Darlington (M). Detaware. Queen Anne’s Co.: Centreville, 1868, Commons (Ph). Marytanp. Montgomery Co.: above Cabin John, Painter 1027 (M). Prince Georges Co.: Laurel, Sept. 12, 1897, Knowlton (US). Calvert Co.: Chesapeake Beach, House 1451 (US). Wicomico Co.: Salisbury, Tidesirom 7460 (US). Worcester Co.: Snow Hill, Moldenke 6632 (NY). > S RS par) i Disrricr or Cotumpra. Washington, Rock Creek, near west gate of zoo, Pollard 648 (US). West Virarnta. Ritchie Co. Laurel Junction, Sept. 12, 1879, Smith (US). Upshur Co.: neat Buckhannon, July 29, 1895, Pollock (US). Virarnta. Page ©0-' near Luray in the Blue Ridge, Steele 37 (US). Alexandria Co.: Glencarlyn, Dewey 40 (US). Fairfax Co.: Clifton, Oct. 10, 884, Ward (US). James City Co.: Williamsburg, Grimes 4433 (NY)- 900, U Greensville Co.: Emporia, Tidestrom 6902 (US). NorTH Cx LINA. Durham Co.: Durham, Aug. 18, 1931, Blomquts Iredell Co.: Statesville, Hyams (US). Swain Co.: Sway Sept. 10, 1913, Mooney (US). Buncombe Co.: Biltmore, Bi more oo 403° (NY). Haywood Co.: Lake Junaluska, ONY): Canal, Sept., ?, Ravenel (M). Grora1a. De Kalb Co.: Stone A. S. Hitchcock (M). Brevard Co.: Okeechobee region, Fredholm 6014 (NY). Leon Co.: Tallahassee, Nash 2519 (NY, [yy Ontario. Kent Co.: Chatham, Macoun 5898 and 24716 ( NCO NERS LOE EEE RE AS Tee eee ee oe Ne Oh ee eet Ee tr eae 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 147 MicuiGan. Saint Clair Co.: Port Huron, Dodge 7 (US). Ingham Co.: East Lansing, Aug. 19, 1891, C. "F. Wheeler (US). Van Buren Co.: South Haven, Lansing 3339 (US). Washtenaw Co.: 3 miles northwest of Ann Arbor, Hermann 9073 (NY). ayn Co.: River Rouge south of Detroit, Aug. 17, 1916, Chandler (Us). Onto. Lorain Co.: Oberlin, Aug. iH 1894, Ricksecker (US). Cuyahoga Co.: Berea, June, 1897, Ashcroft (US). Holmes Co.: Salt Creek Township, Drushel 9 470 (US). Shelby Co.: Sept., 1904, Clevenger (US). Miami Go. Fletcher, Aug. 20, 1897, Clevenger (US). Franklin Co.: Columbus, Oct. 24, 1898, Clevenger (US). Pickaway Co.: a Aug. 25 , 1926, Dreisbach (Ph). oe Co.: Jamestown, Sept., 1896, Wooton (US). Hamilton : Cincinnati, Aug., seas a (Ph). Scioto Co.: Friendship, Donen 10809 (US). eons Marshall Co.: Lake Maxin- kuckee, Evermann 1021 (US). Lake Co.: Gary, Lansing 4028 (Ph). Ha milton Co.: Mattsville, Wilson 27 (US). Lake Co.: Clarke, Lansing 4020 (Ph). Dubois Co.: 2 miles east of Clear Springs, Deam 26793 (Ph). Spencer Co.: 4 miles southwest of Chrisney, Deam 87495 (Ph). Warrick Co.: 2 miles southeast of Yankeetown, Deam 37598 (Ph). Kernrucxy. Franklin Co.: isa _ Biltmore Herb. 4034 (US). Scott Co.: Stamping 136 (Ph " Bell Co.: Aug. 15, 1888, F. B. Lloyd (NY). TENNES- SEE. Carter Co.: Roan Mountain Station, Sulla 8223 (NY). Coffee Co.: Tullahoma, Biltmore Herb. 408% (US). Knox Co.: Knoxville, Ruth 2498 (NY). Davidson Co.: Nashville, Svenson 139 (G). Hamilton Co.: Chattanooga, Lippincott 98 (Ph). ALABAMA. Jefferson Co.: Birmingham, H. E. eee 1207 (NY). Lee Co.: Pkaeies F. S. & E. S. Earle 36 (G, US). Hale Co.: south of Rosemary, fete! 3246 (G, M, NY, US). Baldwin Co.: Tensaw, en 9018 (NY, US). Mobile Co.: Mobile, Sept., 1892, Mohr r(U i; Mississirrt. Tunica Co.: west of Dundee, Anderson 4479 (Ph). Warren Co.: Vicksburg, Demaree 14104 (M). ee Co.: Raymond, Holt 74 (US). Wisconsin. Saint Croix : Hudso on, Sept. 1896, Mayer (U S). Inuroris. Cook Co.: Chinuae Lansing 2780 (NY). Stark Co.: Wady Petra, Aug. 21, 1894, V. H. Chase (Ph, ile McLean Co.: Bloomington, Robin- son in Pl. in x8. eae 9 (M, NY, Ph, US). Vermilion Co.: Catlin, Lansing 3494 US). Richland Co.: near Olney, Ri 'dgway 3260 0 (Ph). Minnesota. Hennepin Co.: Fort Snelling, Mearns 549 (US). Chippewa Co.: Myers, Moyer 2319 (Mi). Washing- ton Co.: Lake St. Croix, Afton, Sept., 1919, Butters (Mi). Goodhue Co.: Red Wing, July, 1886, Sandberg (Mi). Scott Co.: Prior Lake, Moyle 2084 (Mi). Wabasha Co.: Lake City, Man- ning 3019 (Mi). Winona Co.: Dresbach, Arthur 79 (Mi). Houston Co.: Crooked Cree k, W. A. Wheeler 336 (Mi). Iowa. Fayette Co.: Fink 451 (US). Marshall Co.: LaMoille to Mar- 148 Rhodora [ APRIL shalltown, Aug. 21, 1927, Lounsberry (Ph). Johnson Co.: Coralville, Somes 3569 (US). Mussourt. Pike Co.: Aberdeen, Davis 955 (US). Nodaway Co.: Burlington Junction, Singleton 213 (Mo). Clay Co.: Randolph, Mackenzie 380 (M). Jackson Co.: Eton, Bush 7749 (US). St. Louis Co.: Allenton, Oct. 1, (M). Jasper Co.: Joplin, Trelease 874 (M). Barry Co.: Roaring a Trelease 1178 (M). McDonald Co.: Butler Creek, Noel, rs ARKANSAS. Carroll Co.: Oakgrove, Bush 15918 (M). Benton Co.: Bush 15743 (M). Crittenden Co.: West Memphis, Demaree 097 8239 (M, US). Garland Co.: Hot Springs, Runyon 1495 (US). Jefferson Co.: Bayou Bartholomew, Demaree 13972 (M). _ Co.: Murfreesboro, Demaree 9480 (M). Howard Co.: Minera Springs, Demaree 9736 (M). Hempstead Co.: Fulton, Bush 949 Nevada C Ouachita Co.: Bastrop, Demaree 1 4116 (NY). Natchitoches Co.: Natchitoches, E. J. Palmer 8729 (NY, US). Rapides Co.: a Alexandria, Ball 405 (NY, US). Caleasieu Co.: near Lake ). Bon Homme Co.: Running Water, Sept., 1892, Tho g (M). NeprasKa. Knox Co.: Pishelville, Clements 2751 (US): Saunders Co.: Ashland, July, 1888, Williams (US). Cass = Weeping Water, June, 1887, Williams (US). Otoe Co.: Nebras h City, Oct. 14, 1900, Hedgcock (M). Saline Co.: Crete, Dr we 6025 (Ph). Kearney Co.: Minden, Aug., 1892, Milligan Ui ) Nuckolls Co.: Aug., 1899, Hedgcock (M). Kansas. Riley ob Norton 478 (US, M). Osborne Co.: Osborne City, Sept. 2; a Shear (US). Graham Co.: Bogue, Imler 46 (M). Geary \°- along Otter Creek, Gates 18947 (M). Douglas Co.: Lawrence, July, ?, W. C. Stevens (US). Morris Co.: Dwight, Oct, water, Waugh 381 (M). Oklah -: Oklahoma City, Waugh pn (us) Ts (M) ahoma Co.: Okla — bes 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 149 Paso, M. E. Jones 4174 (I, NY, “aoe Jeff Davis Co.: Mus Creek Canyon, Sperry T243 (US). Travis Co.: Austin, Schulz 720 (US). Montgomery Co.: Willis, Aug., ?, Warner : Gillespie Co.: Cherry Springs, Jermy (M). Val Verde Co.: 48 miles south of Ozona, Ferris & ioe 3012 (M, NY). Kerr Co.: Kerrville, Heller 1922 (Ph, US). Austin Co.: Industry, Wurzlow 19 (US). Harris Co.: Houston, Fisher 607 (US). Bexar Co.: San Antonio, Wilkinson 56 (M). Comal Co.: Brack- en, bed of Cibolo River, Groth 208 (US). Wharton Co.: Pierce, Tracy 7432 (US). Catirornia. Placer Co.: Dutch Flat, Mac- Fadden 12754 (NY, W). Tuolumne Co.: near Bear Creek, fest 242 (NY). Amador pe Middle Fork, fa ansen 121 6 US). Butte Co.: east of beiteei, Heller 11139 (F, G bee's US). Spleen Skamania Co.: near Hood, Suksdorf 12327 (G, M, NY, Ph, US). According to Thellung in a postal card of Feb. 6, 1916, filed at Gray Herbarium, Dr. E. Bonnet examined an authentic specimen of E. nutans in the Cosson herbarium at Paris and found it to be identical with EZ. Preslii. (How the identity of E. Preslii was ascertained is not explained.) Dr. Bonnet even communicated a fragment of the specimen in the Cosson herbarium to Thellung that he might judge for himself. However, Lagasca described his plant as “‘Caule . . . villoso” and Thellung verifies the fact and terms it “fortement poilue”. Having satisfied himself that EH. nutans and E. Preslii were identical and the type of E. nutans had villous stems, Thellung became concerned over the fact that Millspaugh had somewhere characterized E. nutans of the United States as glabrous. As a consequence Thellung de- cided that the plant of the United States differed from E. nutans and therefore required a new name which he proposed in the ambiguous manner quoted above near the end of the synonymy. A more confusing way of proposing a new name can hardly be imagined but is matched by “Gerbera (viridifolia var.?) Con- rathii Thell., spec. vel var. nov.” Vierteljahrsschrift der Naturf. Gesell. in Zurich 68: 454. 1923. Not only is the type of the new name practically impossible to determine but the position and tank of the new name are exceedingly uncertain. Nevertheless, the problem for me is fairly simple. I am satisfied that the plants known to American authors as either E. Preslii or E. nutans are E. maculata L. Consequently Thellung’s new name is a synonym of E. maculata. 150 Rhodora [ APRIL There is a specimen at F which probably represents the plant which Lagasca described as EH. nutans. It bears the data: “ex antiquo herbario generali, Herbarium Horti Botanici Matri- tensis, EZ. Preslii, E. nutans Lag., in hort Madrid.”’ The stems are pilose-tomentose on the young portions and subglabrate at maturity. This agrees with the observations of Thellung. It appears that HZ. nutans was based on unusally vestite plants of E. maculata. The history of the misapplication of the name £. maculata L. to the small-leaved prostrate plant properly known as E. supina Raf. has been amply reviewed in Contr. Gray Herb. 127: 74-76. 1939. Evidence that E. hypericifolia L. is properly applicable to a plant ranging from the West Indies and Mexico south to South America is presented, op. cit., 73-74, and Proe. Biol. Soe. Wash. 1 ripe 1940. EvPHORBIA VERMICULATA Raf., Amer. Monthly Mag: 2 (2): 119. 1817, op. cit., 2 (3): 206. 1818. Typr not known to exist, “Found in August, 1816, near Sandyhill aud Glen’s Falls, state of New York in fields.” House, N. Y. State Mus. Bull 254: 471. 1924—NXamesike vermiculata (Raf.) mat, Aut. “ 97. 1840. Peoar p i vermiculata (Raf.) House, N. Y. Sta Mus. Bull. 233-234: 8. 1922 E. hypericifolia L. 82 ercule Torr., Comp. Fl. No. & vel States, 331. 1826, fully accepted in this category by 17 C State N. Y. 2: 176. 1843. Typn: : (NY?, not found). Perha reference to synonymy here is erroneous since Torrey, |. ¢.; 843, ’ states “capsules even, sometimes pubescent.” As a pplied er the capsules are always glabrous.—E. hirsuta (Torr.) Wie Bot. Gaz. 24: 50, Pl. LI. 1897 July); Robinson & Fernald, Gray’s ~ Man. Bot. ed. 7, 54 E. Rafinesquat Greene, Pittonia 3: 207. 1897 (Sept.) hake Rafinesquii (Greene) Arthur, Torreya 11: 260. 1912 (Jan.)? Heller, Muhlenberg 2: coe 1912 acl Small in Britton & Brown, Illus. Fl. ed. 2 fe Rothrockii Millsp., Field Mus. Pub. Bot. 2+ hy 1913. Typx: Camp Crittenden, Santa Cruz Co., Arizona, oe 1874, J. T. Rothrock 672 (F 197499!; ‘photographs GI, lata. M!). Intergrade between E. vermiculata and mac ual; stems prostrate to suberect, sparsely pilose, few x pies from the base, 8-40 em. long, 0.5-1.5 mm. thick, inte ' Title-page date 1921, but cited by House, N. Y. State Mus. Bull. 254: 471. 19? as Pres pose received at Gray Herb. Feb. 24, 1922. of issue fide Torreya 12: 1, 912, ea te ee eee eae ee ee eS Te Oa He Le St ee ee ete hn ee i 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 151 nodes up to 6 em. long, mostly 1-2 cm. long; leaf-blades ovate to aaibealaes, 5-15 or sometimes 19 cm. long, upper surface glabrous or very sparsely “ lower surface more or re ne especially ard tow the base, margin serrulate; petioles 1 mm. long; tule distinct or united, up to 1 m m. long, ciliate or glabrous, margin fimbriate to divided into a ‘aw linear segments ; cyathia solitary at the nodes or the uppermost terminal in the bifurca- — ns of the small leafy cyme; aestr 0.5-2 mm. long, g gla- panulate, tapering to the peduncle, 0.7-1 mm. in diam.; lobes triangular, slightly exceeding the glands, the distal mostly entire, the proximal often with erect linear teeth; glands —_ stipitate, circular to broadly y transversely elliptical, 0.2-0.3 m iam.; appendages white, glabrous, to 0.6 mm. long, en ue or eblente toothed; fifth gland linear, ca. 24 as long as the lobes; sinus U-shaped, slightly depressed; bracteoles forming a slender segment below each gland, united to the involucre below, entire or with two to four filiform glabrous or sparsely hairy —— —— staminate flowers 5-15 per cyathium; androp gla- brous, 1—-1.1 mm. long; gynophore glabrous, exserted Hg eaoaies reflexed; ovary glabrous, roundly 3-angled; styles ca. 24 bifid, 0.6 mm. long; capsule glabrous, broadly ovoid, donee below ‘ds equator, 1.6-1. nae m. long; seeds quadrangular, 1 .1-1.3 m — 0.9-1 mm. taneentialle. 6 0.7-0.9 mm. radially, radially dite, ventral facets mostly slightly concave, dorsal facets to to slightly convex, smoothish to slightly wrinkled, coat dark grayish brown or sometimes pale gray.—PLATE 656A Nova Scotia, New Brunswick and Quebec to oo and New York, west to Date. Michigan and Ohio; also in British Columbia, acer oes New Mexico (Map 41). Repre- sentative ~ taonge en: Scotia. Hants Co.: Windsor, Co.: North Sydney, Bissell & Linder 21757 (G). New Bruns- WICK. Victoria rest ’ Aroostook Junction, Andover, Fernald 1983 (G). QueEBEc. —— Co.: Matapedia, on —— i 10, 1904, Collins & Fernal rye Megantic Co.: east of Blac Lake, Fernald & Jackson 12118 (G). Deu Sua ee ep La Trappe, Louis-Marie 198 (G). Longueuil Co.: shores of the St. Lawrence, Marie-Victorin 1062 (G). D’Argenteuil Co.: — M arie-Victorin & Rolland-Germain 34052 (G). Ponti- o.: Fort-Coulonge, Marie-Victorin, Rolland-Germain Meilleur 48720 (G). Onrarto: Kingston, Sept. 23, 1901, Fowler 22 ; Mi bri ridge, Ottawa, Macoun 87831 (G); Southampton, Macoun 88079 (G). British Couumsta: Englishman’s River, Parksville, 152 Rhodora [APRIL July, 1915, Carter (G). Marne. Penobscot Co.: Bangor, Fernald & Long 13996 (NE); Orono, Aug. 27, 1908, Fernald (NE). a Nye (G); Winslow, Fernald 2737 (G). Sagadahoc Co.: Topsham, Furbish (NE). Oxford Co.: Woodstock, 1887, Parlin 52 (G); Bethel, 1897, Furbish (NE). Cumberland Co.: Portland, Fernald, Long & Norton 18995 (NE). York Co.: York Beach, 1879, Furbish (NE). New Hampsuire. Cheshire Co.: Hinsdale, Raup & Weatherby in Pl. Exs. Gray. 565 (G, I, M, Mo, NE, NY); Hinsdale, Robinson 578 (G, NE). Sullivan Co.: West Clare- mont, — §81 (G). Grafton Co.: Haverhill, ee 14469 : Cog (G). Litch- field Co.: Canaan, Aug. 19, 1910, Woodward (NE). New London 1917, Hunnewell (G). St. Lawrence Co.: Canton, Phelps 634 (G). Washington Co.: Big Hollow, Bakers Falls, Hudson ree Aug. 3, 1914, Burnham (G). Rensselaer Co.; Tombhannoe reservoir, House 10618 (G). Albany Co.: Hudson River, Albany, Sept. 10, 1910, Burnham (G). Chenango Co.: Oxford, July 7 aa Coville (US). Oswego Co.: Cooperstown, Hunnewell 6876 North Fair Haven, Sterling, sand bar of Lake Ontario, Whetzel 12388 (G). Tompkins Co.: Ithaca, Wiegand 755 (GQ). . o.: Montour, Wiegand 8408 (G). New Jersey. Hunterdon Co.: southeast of Linvale, Long 51421 (G). Ocean Co.: Lakewood: Hunnewell 6977 (G). PEnnsytvanta. Lancaster Co.: betwee? 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 153 Churchtown and Beartown on Welsh Mountains, Sept. 7, 1892, Small (G); Chikis Bu Aree. between Churchtown Road an nd Beartown, Heller 670 (G nee. BGP Colainhia Co.: Aristes, Fosberg 1 5764 (G). aati Washtenaw Co.: 2.3 miles east of nn C. K. Dodge (G). Inp1ana. Steuben Co.: 5 miles northeast of Angola, Deets 32533 (Deam, G). Onto. Lake Co.: Painesville, Aug. 24, 1886, Werner (G). Erie Co.: Perkins, Aug. 20, 1895, cpus’ (G). ‘Lucas Co.: Swanton, Sept. 26, 1926, Moseley (G). ExIco. Grant Co. : Mangas Springs, Aug., 1901, Met- ae (us, Mogollon Creek, 2 miles above falls, Maguire, Rich- ards & Moeller 11982B (G). Socorro Co.: Frisco, July 25, 1900, Wooton (US). Arizona. Cochise Co.: Tanner ‘Canyon, Hua- chuca Mountains, Goodding 798 G, NY, IT, US), After a careful consideration of Rafinesque’s description, I am forced to conclude, as did House, that it applies to this entity. The plant is probably introduced in British Columbia. In Arizona and New Mexico it may be native and represent the fringes of a puzzling Mexican complex centering around E. maculata. Most of the Arizonan and New Mexican plants ap- proach or even intergrade with EZ. maculata. To identify plants of Arizona and New Mexico with those of New England when they are not found between may seem fantastic but even worse is the fact that some Argentinian specimens seem identical. In New England and vicinity the entity here recognized as E. ver- miculata is distinct; elsewhere it is vague and seems to intergrade. A complete study must be made of E. maculata and its satellites in order to evaluate properly E. vermiculata. A collection from Bangor, Penobscot Co., Maine, Sept. 13, 1898, Fernald 2741 (NE) has one small plant with overlapping leaves, capsules widest at the summit, and peculiar seeds half sordid white and half dark gray. It appears to be an abnormal plant. E. hirsuta Schur, Verh. Siebenb. Ver. Naturw. 4: Abhang, 66. 1853, has been given as preoccupying E. hirsuta (Torrey) Wie- gand 1897. Schur published his name as a subdivision, of un- stated rank, of Z. Esula L. The fact that he published it as a binomial does not make it capable of preoccupying an identical Specific name but rather makes it invalid in the rank proposed 154 Rhodora [ APRIL since binomial nomenclature is not admissible for subdivisions of species (International Rules of Botanical Nomenclature ed. 3, Art. 28.1935). Boissier in DC. Prod. 15 (2):—1862, omits this name of Schur’s. Unless someone can find a validation of Schur’s name in specific rank prior to 1897, Wiegand’s name 1s not preoccupied though it is a later synonym. Fernald, RHopORA 36: 417-420. 1934, has discussed a case in which a similar name of Schur’s was validated in specific rank. The listing of Z. hirsuta Schur in Index Kewensis as a synonym of E. Esula obviously does not validate Schur’s name. 168 Rhodora [May EUPHORBIA SUBGENUS CHAMAESYCE IN CANADA ND THE UNITED STATES EXCLUSIVE OF SOUTHERN FLORIDA Lovis CurreR WHEELER (Continued from page 154) 15. EKUPHORBIA GLOMERIFERA (Millsp.) L. C. Wheeler, Contr. Gray Herb. 127: 78. 1939; based on Chamaesyce glomerifera Millsp., Field Mus. Pub. Bot. 2: 377. 1913. Typx: El Rancho, Dept. Jalapa, Guatemala, alt. 1000 ft., Jan. 20, 1908, W. A. Kellerman 8053 (F 224827! [a “6” is pencilled after the stamped of this name. labrous, annual, or perhaps sometimes of slightly longer duration; stems erect, with occasional branches below, 12- cm. tall, from 1 mm. thick above to 4 mm. thick and slightly woody below, internodes 2-4.5 em. long on the main stem; leaf- 0.6 to rarely as much as 4 mm. long, glabrous; involucre obconl- cal, tapering gradually to the peduncle, 0.4-0.9 mm. in diam., in some cyathia; androphores glabrous, 1-1.1 mm. long; gY2° phore glabrous and sometimes reflexed; ovary glabrous, 3-lobed; 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 169 styles nage bifid to about the middle, subclavate, ca. 0.4 apsules depressed-globose, roundly 3 3-lobed, glabrous, 1.3-1.4 m ae: widest at about the equator; seeds ovoid- retinal eae: 0. 9-1 mm. long, ca. 0.5 mm. tangentially and radially, radially ovate, facets with slight irregular depressions separated by very low smooth ridges, gelatinous coat so thin as to little obscure the light brown testa.—PLaTE 6 Southern Florida and extreme southern Texas; Bermuda, West Indies, Central Aiiceee northern South America (British Guiana, Venez uela, and Colombia), and Hawaii Representative specimens seen from the United States: FLoripa. Dade : Ft. Lauderdale, Small & Carter 644 (NY); Miami, Tracy 91: 27 (G, M, NY, US); Old Rhodes Key, Small & Mosier 5699 (NY, US); ; Brickell Hammock south of Miami, Small 4036 (NY); Buena Vista, Moldenke 330 (M, NY, US); Soldier’s Key, Britton 330 (NY); hammocks between Miami & Cocoanut Grove, J. K. & G. K. Small 4619 Ses Monroe Co.: Ten Thousand Islands, apie 388 (G, US); Doctor’s Arm, Big Pine Key, Simpson $10 (NY, US): Big Pine Key, Killip 31582 (US); lower portion of Key ales Small & 8209 (NY); No Name Key, Pollard, Collins & Morris 126 (NY, US); Upper Metacombe Key, Curtiss’ 2486 (G, M, NY, US); West Summerland Key reer J. J. Carter & G. K. Small 3627 (NY). oat Hidalgo commonly pilose with Nee yellow tapering hairs Say toward the stem-tips, internodes up to 7 cm. long but mostly 1-4; leaf-blades prevailingly broadly rhombic-lanceolate, varying from narrowly lanceolate to ovate, 4-40 em. long, sparsely strigose and vateeee above, with appressed to spreading crisped hairs beneath, base strongly inequilateral, apex acute, margin sharply to bluntly serrate; petioles 1-2 mm. long; stipules tri- angular, long and slenderly attenuate, ca. 1 mm. long, distinct or barely united at the base, mostly with linear divisions below, with short scattered hairs; cyathia i in dense pedunculate cymose heads of numerous cya athia; peduncles glabrous to sparsely Strigose; involucre obconic-campanulate, 0.6-0.9 mm. in diam.., upwardly strigose outside, glabrous inside or a few hairs on the inside faces of the stipes; ‘lobes ciliate on the outer margin, tri- angular, mostly about equaling the glands, the margins lacerate into erect. filiform segments; glands on long stipes, . to patelliform, circular to transversely oval, 0.15-0.3 m n 170 Rhodora [May diam.; appendages white, glabrous, entire, from narrower than to twice as wide as the gland, or sometimes wholly absent; fifth gland ca. 4 as long as the lobes; sinus U-shaped, scarcely isi pressed; bracteoles sometimes reduced to one filiform segmen the glands, free portion parted into few linear shortly hairy segments shorter than the androphores; staminate flowers 2-8 per cyathium; androphores glabrous, 0.9-1 mm. long; gynophore glabrous, shortly exserted and mostly reflexed; ovary shortly strigose upwardly, 3-lobed; styles glabrous, bifid 14-24, 0.2-0.4 mm. long, slightly clavate; capsule 1-1.15 mm. long, sharply 3- angled, wider below the middle, shortly strigose, base truncate; seeds sharply quadrangular, 0.7-0.9 mm. long, 0.5-0.6 mm. tangentially and radially, ovate-acute radially, base truncate, facets with sub-regular to quite irregular low smooth wrinkles, ventral facets concave, dorsal concave to plane, microreticulate white coat often so thin as to little obscure the light brown to tan testa. Key To VARIETIES a. var. typica. Bie W-6 ee Oe WS O-8 Re ee SES wee ing freely, often forking symmetrically (or nearly so) at the tip; mostly low plants with small leaves.............. b. var. procumbens. E. globulifera HBK.., Nov. Gen. et Sp. 2: 56 (quarto), 45 ease 1817. Type: Cumana, Venezuela, Bonpland 403 (Herb. & Mex. [Creek} near Deserted Rancho,” Santa Cruz County, Arizona, Sept. 1 1851, C. Wright 1842 (M 144667!; plstoauahe G!, W!; 1soTyPEe 1941] Wheeler—Euphorbia Subgenus Chamaesyce 171 G!, US p. p.!).. Merely plants with red-spotted leaves.—E. discolor Pave ex Millsp., gues Mus. Pub. Bot. 2: 402, 440. 1916 (without basinym) by e r, aS synonym of Chamaesyce hirta.—E. pilulifera L. 1. [“‘Spielart neateat capt ) Ho ipo in Ascherson & Graebner, Syn. Mit Chamaesyce Rosei Millsp., Field sa ech a 2: 102, oie. Type: along an arroyo in the vicinity of Alamos, Sonora, Mexico, Mar. 13, 1910, Rose, Standley, & Russell 12728 (NY!; fragment F!; 1soTyPE FI). A rather stunted and perhaps overwintering plant probably belonging to a race found in Sonora and Sinaloa, rather intermediate between E. hirta vars. typica and procumbens. . pilulifera L. var. guaranitica Chodat & Hassler, Bull. 679. 1905 Herb. Boiss., ser. 2, 5: 05. TyprE: in regione cursus superioris flumin is Apa, Pesci Nov. 1901/2, #. Hassler 7735 (Ge?, not seen; Or E G!). A low plant with smaller leaves than usual for var. dei E. pilul ifera L. sensu Jacquin, Icones Pl. Rar. 3: t. 478. 1786— 93; Boiss. in DC. Prod. 15 (2): 21. 1862; A. M. Marselt, Contri- bution 4 l’Etude Botanique, Physiologique, et Therapeutique de Euphorbia _pilulifera, eo pour le Doctorat en Medecine, Année 1884, No. 36, pp. VI, 62 [2], 2 plates; J. D. Hooker, Fl. Brit. India 5: 251. 1887; Thellune in Ascherson & Graebner, ba ing Mitteleur. Fl. 7: 1917: Farwell, Roopora 38: 331-2. 1936; and many other Hoy te under Euphorbia, Anisophyllum, Cham- aesyce, and Tithym e following Australian forms foots fs belong here: E. pilulifera L. forma rubromaculata, f. humifusa, & f. viridis K. Domin, Bibliotheca Bot. Band. 22, Heft 89 (4): 866. 1927. Plate 657A. Casual and not persisting in Michigan and New York; South Carolina, Florida, Alabama, Arizona, West Indies, Mexico, south to Argentina; widely introduced in the Old World (Map 24). Representative specimens seen from the United Bente: MicuicgaNn. Wayne Co.: Detroit, Farwell 8756 (G). Sourn Garourna. Charleston Co.: Charleston, Fernald & Long 9747 (G). Fuorma. Brevard Co.: Indian River region, F tadhokn 5517 (G). Hillsborough Co.: Fredholm 6348 (G). Lake : near Eustis, Nash 157 (G). Lee Co.: Myers, AS. Hitchcock 326 (F, G). Manatee Co.: near Bradentown, June 2, , Simpson (F). Marve Co.: Upper Metacombe Key, H. Curtiss 2496 (F, G). Orange Co.: Fredholm 5429 (G). Palm Beach Co.: Palm Beach, A. H. Curtiss 5395 (G). Pasco Co.: St. Leo, Mar. 24, 1927, O'Neill (M). Pinellas Co.: near St. Petersburg, Deam 2762 (F, G). Seminole Co.: Sanford, Oct. 8, 1892, Leeds (F). ALABAMA. Mobile Co.: Mobile, Dukes 6 (G); Mobile, Sept., 1878, Mohr (G). Arizona. Cochise Co.: near Fort Huachuca, Huachuca Mts., J. G. Lemmon 2875 (F, G). Santa Cruz Co.: Tumacacori, Harrison & Kearney 6022 (G, US); 172 Rhodora [Mar base of Patagonia Mts., Peebles, Harrison & Kearney 4653 (US); Nogales, Harrison & Kearney 6026 (US); near Patagonia, Kearney & Peebles 10172 (US). : 16b. E. urrta L. var. procumBrens (DC.) N. E. Brown 10 Thiselton-Dyer, Fl. Trop. Afr. 6 (1): 497. 1911; L. C. Wheeler, Contr. Gray Herb. 127: 69, Pl. IV, C, fig. 2. 1939; based on £. procumbens DC., Cat. Pl. Hort. Monsp., 111. 18138, not Miller, 8 E. pilulifera L. var. procumbens (DC.) Boiss. in DC. Prod. 15 (2): 21. 1862.—Chamaesyce pilulifera (L.) Small var. procumbens (DC.) Small, Fl. Se. U. S., 714, 1334. 1903. Since no authentic material has been seen it has been necessary to accept withou confirmation the interpretation of Boiss. in DC. Prod. 15 (2): 21. 1862 from the garden at Montpellier, France, (Geneva? ir seen).— E. obliterata Jacquin, Enum. Syst. Pl. Carib., 22. 1762, & Select. Stirp. Amer. Hist., 151. 1763, at least in the sense ID which it was used: E. pilulifera L. var. obliterata (Jacq.) A. ¥. Hitchcock, Ann. Rep. Mo. Bot. Gard. 4: 127. 1893. No authentic material has been seen. ; E. opthalmica Persoon, Syn. Pl. 2: 13. 1807. Typx: Rio de Janeiro, Brazil, July, 1767, Commerson 238 (Herb. Mus. Paris, not seen; fragment F!). A small-leaved plant. in N.fova] H.[ispania]’; perhaps at Madrid judging by Alph. C., Phytographie, 426. 1880. Supposed by Boiss. in DC. Prod. 15 (2): 21. 1862, and others, to be the same as E. procumbens DC.—Chamaesyce gemella (Lag.) Small, Fl. Miami, 110, 200. 1913. Florida; adventive in Pennsylvania; Louisiana; Mexico, West 8). Co.: Columbia, April, 1876, Garber (F). Frorrpa. Brevard Co.: Meritt’s Island, Indian River, A. H. Curtiss 2496 (F, G). Brow- ard Co.: Pompano, Pease 26455 (G). Dade Co.: Miami, Tracy 9115 (G); hammocks between Miami and Cocoanut Grove, J: Gy: & G. K. Small 4694 (G); Elliott’s Key, Simpson 505 (F; i Miami, . A. Curtiss 5849 (F, G). Monroe Co.: Pine Crest, Moldenke 865 (M, NY). Palm Beach Co.: Kelsey City, Fanmét R. Randolph 135 (G); Palm Beach, May 20, 1895, A. H. Cu (G). Lovrstana. Terrebonne Co.: Houma, Sept. 6, 1914 Wurzlow (F). For a review of the evidence supporting the application of the name Euphorbia hirta to the species described above see MY dis- cussion in Contr. Gray Herb. 127: 71-72. 1939. Also see OP» cit., 78, for reasons for applying E. pilulifera L., which has been applied to the concept here called E. hirta, to an Old World plant. 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 173 Aap 18, ra inge of EuPHORBIA HIRTA Var. PROCUMBENS . U.S. but Pennsy!- Vanian telloet plants omitted; 19, E. — in U.S.; 20, E. Gryertr; 21, TRACHYSPERMA; 22, dots E. OCELLATA Var. TYPICA, aoe E. OCELLATA var. ARENICOLA; 23, E. gr nar sais oP in U. S.; 24, dots E. crypto eg circles E. iret yar, yale n U.S. but Michigan sand Nev York waifs mitted, Square E. oce.nat Ratranu; 25, E. Partsui; 26, E. scisaplitiocs in U.S.; 27, E. CINBRASCENS in U. S.; 28, E. Dacian WU. 8. ots E. pe var. a n U.S., cir cle E. FEeNDLERI — _TRILIGULATA 30, dots E. corp FOLIA, circ ny ge 31, dots E. Fenp.eri var. pnb OCALYX, circles intergrdes 8 n E. Fenpuert vars. nate and CHAETOC: 32, dot s E. Hoo gh ne E. VALLIS-MoRTAE; 33, dots E. ieciwes, circle E. serra i in U. s. 174 Rhodora [May 17. EvpHorBIA CAPITELLATA Engelm. in Emory, U. 8. & Mex. Bound. Surv. 2 (1): 188. 1859. Type: low valley at San Bernardino, Cochise County, Arizona, Oct. 3, 1851, C. Wright 1849 (M 149810!; photographs G!, W!; 1sorypzs G!, NY). Rather lax and long-leaved; leaves nearly glabrous. Boissier m DC. Prod. 15 (2): 22. 1862; L. C. Wheeler, Bull. Torr. Bot. Club 62: 537. 1935.—Chamaesyce capitellata (Engelm.) Millsp., Field Mus. Pub. Bot. 2: 408. 1916.—E. capitellata var. typwa L. C. Wheeler, Bull. So. Calif. Acad. Sei. 35: 127. 1936. E. pycnanthema Engelm. in Emory, U. 8S. & Mex. Bound. Surv. 2 (1): 188. 1859. Type: on mountainsides near Lake Chamaesyce pycnanthema (Engelm.) Millsp., Field Mus. Bot. 2: 411. 1916. E. geminiloba Millsp., Proc. Calif. Acad. Sci. ser. 2, 2: 228, 1889. Typ: Pozo de Los Dolores, Lower California, Mexied : Apr. 5, 1889, 7. S. Brandegee (F 196142!, photograph G!, Wi A specimen with the tips of the branches missing and the leaves coarsely serrate. ; 92 Euphorbia pycnanthema forma serrata Millsp., op. cit-) rs 1889. ‘‘Pozo de Los Dolores, April 5th” 1889, Lower Califor T. S. Brandegee (?). The typx has not been located. At pie an isotype may be expected at C. The description suggests t “a it was based on the same minor variant and perhaps even ¢ same specimen as EL. geminiloba. & E. capitellata var. laxiflora 8. Wats., Proc. Amer. Acad. Arts Sci. 242 74. 1880. ‘Tyre: high mountains, Guaymas, Sonot Mexico, 1887, Ed. Palmer 210 (G!, 1soryrr US!). Erect WI long internodes and narrow glabrous leaves. : 4 E. Chamberlinii 1. M. Johnston, Proc. Calif. Acad. Sei. set-.* 12: 1066. 1924. Typx: Escondido Bay, Lower Caliform® Mexico, June 14, 1921, I. M. Johnston 4136 (CA 1288!). Has distinctive appearance due to an abundance of white coccids. E. gladiosa M. E. Jones, Contr. West. Bot. 15: 144. 192. ymas, Sonora, Mexico, Nov. 2, 1926, M. E. Jone 22613 (P!). The same variant as E. capitellata var. laxiflora. 1941] © Wheeler,—Euphorbia Subgenus Chamaesyce 175 Perennial; stems few to numerous, ascending to erect, 5-40 em. long, 0.5-1.5 mm. thick, glabrous to pubescent, internodes up to 4.5 em. long, mostly about 1 cm. long; leaf-blades ovate- acute to linear-lanceolate, 4-25 mm. long, glabrous to pubescent, base markedly inequilaeral margin entire to coarsely sharply serrate; petioles 1 mm. long; eee mostly etn triangular- to sdbulate-ettciiet parted into a few erect segments, ciliate to pubescent, 1.5-2 mm. ar evathin gested in cymose glomerules of several to many, or a few pinay in the upper bifurcations; pedun cles pubescent to glabrous, 0.5-1 or rarely to 3 mm. lon ng mae - ag tee sy broadly obconical-campa sala tad 1,3-1.7 dia glabrous to pubescent outside, hairy on the nai of the leat and stipes; lobes narrowly tria angular, acuminate, slightly to smi od exceeding the glands; glands circular to transversely oval, 0 0.45 mm. in diam., on long stipes; appendages white to Sil glabrous, entire, usually conspicuous; fifth gland linear, pubes- cent, 1434 as long as the lobes; sinus U-shaped, sli ehtl tly de- scent e, or brous, 1.6-1.9 mm. long; ovary slightly 3-lobed, mostly pubes- cent, sometimes spatter’ styles 144-24 bifid, glabrous or some- times with a few hairs at the bas we 0.6-0.7 mm. long; capsule pubescent to glabro 1.3-1.9 m o ong aabeeaias 3-lobed, widest at the econtOr me slightly apes eg quadrangular, 1. 2- mm. long, 0.6—-0.8 mm. tangentially, 0.6-0.7 mm. radially, narrowly ovate to very narrowly oblong-ovate radially, apex acute, base obtuse to truncate, facets with small shallow depres- mew or even sub-regular faint transverse wrinkles.—PLATE 657B. zona, western Texas, Chihuahua, Coahuila, Sonora, Sina- loa, pe California (Map 45) Recteeniatce specimens seen Texas. Brewster Co.: along Blue Creek, foot-hills of Chisos M G). Arizona. Coconino Co.: Ashfork to Williams, Kearney & Peebles 12077 (G). Yavapai Co.: Prescott, Rusby 317 (NY, U ila Co.: 6 miles east of Cassadore Spri rings, Maguire, Richards & = oeller 13068 (1); Roosevelt Dam, Haskeeod 8668 (G). Pinal Co.: 2 miles seg Coolidge Dam, Maguire, Richards & M fons is (G, I); Picacho Mountains, Peebles a Maguire, Richards & Moeller 10375 (G, 1). Pima Co.: near Colossal Caves, Tucson, Maguire, Richards & Moeller 11699 (G, 1); east of Ranger Station, Baboquivari Mountains, Wiegand, Maguzre, Richards & Moeller 10778 (I); near Tucson, May 3, 1883, 176 Rhodora [May Pringle (NY, US), Apr. 8, 1881, Pringle (G, M, US), Oct. 27, 1905, Tracy 8987 (G, M, NY, US), Oct. 29, 1905, Tracy 8958 , M, NY, US). Santa Cruz Co.: Ruby to Nogales, Peebles & Fulton 11446 (NY). Cochise Co.: near Fort Huachuca, neat Huachuca Mountains, Lemmon 3112 (G). MEXICO: CHIHUA- nua: near Chihuahua, June 5-10, 1908, Ed. Palmer 370 (G, US); near Chihuahua, June 5-10, 1908, Ed. Palmer 376 (G); neat Chihuahua, Pringle 699 (G, US); 7 miles north of Charco Piedra, Johnston 7931 (G). Coanvrta: Monclova, Aug., 1880, Hd. Palmer 1211 (G). Sonora: Guaymas, June, 1887, Hd. Palmer 83 (G, US), Aug., 1887, Ed. Palmer 142 (G, US), Oct., 1887, Ed. Palmer 317 (G, US). Badebuache, C. E. Lloyd 457 (G). SINA- LOA: Culiacan, Aug. 27-Sept. 15, 1891, Ed. Palmer 1517 (G, US). Lower Carirornia: Concepcion Bay, Johnston 41 73 (G, US); 30 miles south of Mulege, Shreve 7096 (G). In Bull. So. Calif. Acad. Sci. 33: 105-6. 1934, I had provision- ally accepted the data on a specimen of this species which claimed to have come from an altitude of 6000-8000 feet in the San Jacinto Mountains, Riverside County, California. In view of the fact that this is the only collection purporting to have come from California and that it is far above the life zone in which it occurs in adjacent regions I am now refusing to accept as valid the data of this collection. This species is as polymorphic as E.pediculifera. Both have a linear-leaved variation centering about Guaymas, Sonora. Perhaps the linear-leaved variation in one is of as much con sequence as in the other. However, with some hesitation I have concluded that E. capitellata var. linearifolia is too vague &™ ill-defined for recognition due to the great number of intermedi- ates in all characters. 18. Evpnorsia acuta Engelm. in Emory, U. 8. & re Bound. Surv. 2 (1): 189. 1859. Boiss., Icon. Euph., t. 6. oa Typu: ““N. Mex.”, 1851, C. Wright 1839 (M 149791}; fragm ood representative of the species.—Chamaesyce , C. Wright 1840 (Ge!; photographs G!, W!; aoe !, M!, US!). A narrow-leaved extreme intergrading comple Perennial from a farinaceous taproot as much as 1.5 cm. thics, stems annual, erect or ascending, numerous, 10-30 cm. long, : 1.5 mm. diam., with long weak hairs partially deciduous 12, Be : internodes 1-4 em. long; leaves sessile or subsessile, spat! 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 177 long-villous to densely be esgat tomentose below, less so and glabrate above, mostly m. long, ovate-lanceolate to lanceo- late, base ieveamnetieal pints long-acuminate, cartilaginous at the extreme tip, margin sometimes strongly revolute; stipules apparently wanting; peduncles stout, to 2 mm. long, sparingly to densely villous; cyathia solitary. at the nodes; involucres turbinate, 1.7-2.5 mm. diam., sparsely to markedly villous with- out, with fine hen hairs within; lobes narrowly deltoid and entire, or broader and two-toothed, "equaling or slightly exceeding the glands s; glands a ae slightly concave, ochroleucous, the proximal 1-1.5 long, the distal shorter; appendages a as wide as to mek ssid longer than the glands, glabrous, white, margin with irregular short blunt sar fifth aad totally absent; sinus somewhat depressed, U- shaped; bracteoles 3-5 opposite each gland, united at the base and sometimes throughout, of various lengths, some usually nearly equaling the glands, long-hairy; staminate flowers 4-5 per fascicle, 20-25 per cyathium; androphores 2.2-2.5 mm. long, slightly exserted, with from few to numerous slender hairs throughout; gynophore hairy, exserted and reflexed at maturity; ovary densely white-hairy, styles glabrous, ca. 1 mm. long, parted to the middle, Belen somewhat flattened, recurved; capsule sharply three-lobed, 3 mm. long and in eager ne ort~ap- pressed-hairy; seeds quadrangular, ovate radially, 2.2—2 long, 1.5-1.7 mm. tangentially, 1.5 mm. radially, base pe ae coat white, microreticulate-—PLATE 6598. Southern New Mexico, western Texas, and northern Coahuila (Map 35). Representative specimens seen: New Mexico. Sierra Co,: Lake Valley, 1915, Beals (US). — ae ies = Carlsbad, pretees* 40287 (US). Texas. Bre r Gos Fria road, Cory 1915 (G); south of Alpine, oe. 18593B “8 Crockett Co.: Ozona, M. 'E. Jones 26015 (M). Edwards Co.: Barksdale, E. J. Polis 10984a (US). Kinney Co.: Cory 508 HUILA: near Diaz, Pringle 8278 (G, M, NY, US); 100 miles north of Monclova, Sept., 1880, Ed. Palmer (G); 17 miles south of Allende, Aug., Johnston 7028 (G); at foot of eastern slope of the Sierra de Puerto Santa Ana, Wynd & Mueller 243 (M, NY, US). For citation of additional specimens see Bull. Torr. Bot. Club 63: 435. 1936. There seems to have been some confusion about the collections 178 Rhodora [Mar referred to this species. Engelmann in Emory, U. 8. & Mex. Bound. Surv. 2 (1): 189. 1859, states “‘Stony prairies western Texas, along the San Pedro and Pecos river, &c.; Bigelow, Schott. (No. 1739 and 1749, Wright.)” There are before me what are presumably all the specimens of this species from the United States in the Herbarium of Missouri Botanical Garden. Yet none of the collections cited is in the suite. However, there 1s Wright 1839 which I am taking as type. The most plausible explanation which occurs to me is that someone made a mistake in numbering some of these collections. All the other Euphorbiae of Wright’s collections of the years 1851-2 bear numbers in the eighteen hundreds. (They were numbered phylogenetically by Asa Gray.) There are sheets at G, NY, and US bearing the number 1739. At both G and US someone has queried this number and added 1839. Just how Engelmann managed to cite No. 1739 when the sheet in his herbarium bore only the number 1839 is not clear. Engelmann’s citation of number 1749 is even more puzzling. He did have a number 1840 which by a combina- tion of poor handwriting and perhaps unknown circumstances was evidently converted into 1749. However, the problem can be dealt with very simply after the obvious assumption is made that there were errors in the numbers. Wright 1839 is taken as type since it is a good specimen and entirely representative of the species. Wright 1840 is the type collection of Euphorbia acuta var. stenophylla Boiss. and represents a narrow-leaved extreme of the species. The usual methods of elucidating the source of Wright’s collections fail completely in this case. Of the three original numbers found, all came from Western Texas. It aP- pears very likely that the material distributed as No. 1839 (oF 1739) may have been from more than one of Wright’s collections: . Evpworsia anausta Engelm. in Emory, U. 8. & Mex Bound. Surv. 2 (1): 189. 1859. Typs: Rocky bluffs at camp Icon. Euph., t. 7. 1866.—Chamaesyce angusta (Engelm-) Small, Fl. SE U. S., 711, 1333. 1903 : 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 179 up to 8 cm. long but mostly not over 2-4 em. long and gradually shorter upward; leaf-blades of three completely intergrading sorts, (1) the basal, ovate to elliptic-oblong, the median, elliptic-linear to linear, acuminate, 2-4 cm. long, (3) the upper, linear, often involute on drying, 4-10 mm. long, all entire (with the exception of a few serrulate basal leaves on on petioles 0.5-1.5 mm. long; stipules tardily deciduous, distinct, consisting of brown segments arising from interpetiolar portion of the amplexicaul petioles, bearing a few short hairs, longest segments 0.6-0.9 mm. long; peduncles 1-3 mm. long, strigose; cyathia at the upper nodes, solitary; involucres narrowly cam- panulate to obconical, tapering to the peduncles, 1.3-1.6 mm. in diam., strigose outside, strigose snide except on the lower half beneath the lobes; lobes short, triangular, little exceeding the glands, small and "densely hairy; glands narrowly recy ri oblong, 0.4-0.6 mm. long, strongly depressed in the middle, ipiidn dion ascending, white, 0.3-0.7 mm. wide, ee than the glands, with a few short appressed hairs beneath at the base, outer margin truncate, shallowly and irregularly toothed; fifth gland absent, sinu s U-sha ped, strongly depressed; bracteoles united at the base er a tuft adnate below to the involucre be- neath the sar rnc hese a little shorter than the andro- pee staminate flow gay per cyathium; androphores 1.3-1.4 mm. pena os irae some in each cyathium, with sparse fine hairs; p Tapear tnt pres g shortly exserted and usually reflexed ; ovary 3-lobed, densely appressed-hairy; styles obliquely sprea ading to erect, 0. 4-0. 6 mm. long, bifid only at the stigmatic apex to 14 to the base, i Bs ae ‘appressed hairs at the base; capsule strigose, 2.1-2.4 mm. long, wider than long, deeply roundly to easier es < 3-lobed, wider below the equator; seeds quadrangular, 1.6—-1.9 mm. long, 1.1-1.4 mm. tangentially, 1. 1.4 mm. radially, ovate to broadly ovate radially, base cite to truncate, angles blunt but definite, ventral facets plane or concave, dorsal slightly convex, both traversed by few to several low irr egular transverse ridges, coat off-white to chalk-white, microreticulate, testa dark gray.—PuaTE 659 Local in the Edward’s Plateau region, western Texas (Map 33). Reprdsnitaie specimens seen: TEXAS: Poe Co.: Medina Lake Hills, Tharp 6013 (US). Bexar : on the Cibolo and Sabinas (near San Antonio), Lindheimer oe (G, M). Comal Co.: bed of Cibolo River, Bracken, Groth 131 (G, NY, US). Comanche Co.: Comanche Spring, June 1849, Lindheimer (G, M). eke ge = Ranch Expt. Stetina, Cory 3188 (G). Hays Co.: San Marcos «& vicinity, May, 1897, Stanfield (NY). Kerr Co.: Turtle Creek, Bry 276 (US); Kerrville, Heller 1738 (G, M, NY, US). Llano Co.: Llano, E. J. Palmer 10287 (US). Tom 180 Rhodora [May i Tweedy 258 G Co.: Knickerbocker Ranch, Dove _ Creek, (US). Travis Co.: Mt. Burnell, Austin, Hall 559 (G, M, He US). Valverde Co.: Devil’s River, Orcutt 6040 (M); mouth o Pecos River, Cory 26701 (G). Wilson Co.: Sutherland Springs, Aug., 1879, Ed. Palmer (G). * if tonum data “Ex coll. Geo. Thurber, Texas, Pope’’).—Alectoroctonu ciliiotam (T. & G.) Klotzsch and Garcke, Abh. Akad. Berlin, Perennial, herbage with short appressed hairs; stems ascend- ing or erect, 10-15 em. long, 0.5-1 m a pr long; leaf-blades ovate-deltoid-falcate to longeliee Be virtually linear in some cases by revolution of the margins, above; gynophore hairy, shortly exserted and reflexed at a ity; ovary with short appressed hairs, three-angled; sty pr -75 mm. long, parted to or below the middle, with very § pe hairs below; capsule sharply three-lobed, with appressed hate ‘5 mm. long, 2.5 mm. in diam.; seeds quadrangular back long, 1 mm. radially and tangentially, long-deltoid radially, si rounded, raphal ridge straight in tangential silhouette, pe obtuse-truncate, angles sharp, facets smooth, depressed, ¢ white, microreticulate.—PLatTE 6. 2 qd New Plains of Kansas, south to Texas, west to Colorado an * Date fide Pritzel, Thes. Lit. Bot. ed. 2,240. 1872. 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 181 Mexico (Map 36). Representative specimens seen: KAN Morton Co.: on Cimarron ri north of Elkhart on Point ock, Rydberg & Imler 944 (M, NY). Texas. Brewster o 0.2 seed ee south of Alpine, Cory 9294 (G). Coleman Co.: Coleman, April, Reverchon 1855 (M, NY). Coryell Co.: mee hills, Eagle rings, Bigelow (NY). Culberson Co.: near Kent, E Tracy 381 (NY); Signal i Guadalupe Mountain, Whitehouse 502 udspeth : Cory 1921 (GQ). Jeff e Davis, Small & Wherry 12047 (NY). Martin Co.: near Stanton, June 12, 1900, s nagert (M). Mitchell Co.: north of eae me June 8, 1900, Eggert (M). Potter Co.: prairies, Amarillo, May 28, 1902, Reverchon (M). Presidio Co.: “Cory 1919 (G). Banda Co.: we st Canyon City, Aug. 12, 1900, hae (M). Reeves Co.: plains went of the Pecos, Earle & Tracy 104 (Mo, NY). Taylor Co.: north Abilene, June 7 , 1900, Eggert (M). Tom Green Co.: San Angelo, EZ. J. Palmer 10810 (M). County?: near the Sabi- nal, May 13, 1851, C. Wright 1841 (G, M, NY). Cotorapo. aca Co.: 19 milés north Boise City, Oklahoma in Colorado state, Stratton 441 (M). New Mexico. Chaves Co.: Rosswell, alt. ca. 3800 ft., Earle 343 (NY). Dofia Ana Co.: my rgan n Moun- tains, Vasey (M). Lincoln Co.: Carrizozo, —— 92 (NY). For Bull. Tor Club 63 citation of ‘seaitional specimens see Bot. Clu 4 434. Formerly, 1. c., I included some “collections aba Coahuila i in this species. "Phey are not at hand now. Palm 1205 in 1880 at G is EZ. fruticulosa Engelm. Wright no. 1841 would have been preferable as type as far as locality-data are concerned. However, the collection chosen as type is accompanied by drawings and notes of diagnostic charac- ters and the plants are in far better condition. Consequently, since the description seems to have been drawn from these plants, I have taken them as type. The specimens were very likely collected on the Pope Expedition in Texas somewhere near the thirty second parallel of north latitude. 21. EupHorsra GoLonprRINA L. C. Wheeler, Proc. Biol. Soc. ash. 53: 8. 1940. Typ: along sandy beach ne sy gre to Boquillas oe Chisos Mountains area, Brewster County, Texas, Aug. 5, 1937, B. H. Warnock 998 (US 1726028!: t Sauce G!; dcr ai W!). nnual, glabrous; stems oki to 15 em. long, 0.7-1.5 mm. thick, internodes s up to 2 em. long; leaf-blades mostly 6-9 mm. long, o oblong to narrowly oblong sd even narrowly elliptic, entire, base inpccnusioetsls petioles ca. 1 mm. long, amplexicaul ; stipules 0.7-1 mm. long, mostly glabrous, ventral united into a median, cubits often bifid structure, dorsal distinct, linear; peduncles 182 , Rhodora [May 1—1.5 mm. long; cyathia solitary at the nodes; involucres turbin- ate, 1.1—-1.3 mm. diam., glabrous without, glabrous within except for short hairs at base of lobes, gland-stipes and a line extending half-way down below the stipes; lobes slenderly deltoid-attenuate, not quite equaling the glands; glands subcircular or a little longer than wide, deeply concave, sometimes folded together, ; . long, : nearly to the base, slightly clavate; capsule broadly ovoid, gla- brous, 3-angled, ca. 1.8 mm. long; seeds 1.6-1.8 mm. long, sub- quadrangular, narrowly ovate radially, base truncate, ca. 0. mm. radially and tangentially, facets slightly convex, irregularly wrinkled, dorsal and lateral angles blunt, raphe so low and blunt as to scarcely separate the front facets —PLaTe 664A. nown only from the type (Map 37). 22. EUPHORBIA PEDICULIFERA Engelm. in Emory, U.S. & Mex. Bound. Surv. 2 (1): 186. 1859. Perennial from a taproot stout in age; stems prostrate to erect, appressed-pubescent, glabrate in age, up to 2 mm. in diam. toward the base, internodes up to 5 em. long, often very short toward the stem-tips thus congesting the cyathia; leaves closely appressed-pubescent, or sometimes closely tomentose, to subgla- brous, blades 2-37 mm. long, 1-10 mm. wide, ovate with oblique base, oblong with subsymmetrical base to spathulate and even narrowly linear with symmetrical base, petioles 1-2 mm. long, amplexicaul on ventral side of stem; stipules mostly less than 0.5 mm. long, the ventral united, the upper distinct; peduncles clothed as the leaves, up to 1.5 mm. long; cyathia solitary at the nodes, sometimes congested at the branch-tips by shortening of the terminal internodes but not. strictly glomerulate; involucres campanulate, 1.5-2 mm. long, closely appressed-pubescent bes sub-glabrous without, more or less short-hairy within above; lobes deltoid, hairy, equaling the glands; glands transversely oblong, 0.5 mm. wi 0.75—-1. or slightly lobed, glabrous; fifth ‘gl nd ve short or usually 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 183 um; androphores ca. 1.25 mm. long, included or shortly exserted, glabrous or with few hairs above; gynophore nearly glabrous or hairy nearly throughout, exserted and reflexed at maturity ; ovary very slightly lobed, densely hairy, tapering upwar styles ca. 1 mm. long, slender, parted to the base, short-hairy on the ses side to — tip; capsule appressed- pubescent, widest below the middle, 2 mm. in diam. and long, markedly three- lobed, the lobes vn seeds slenderly ovoid, 1-1.3 mm. long, diam., encircled by 4 or 5 rounded ridges with V- shaped channels between, coat white Key To VARIETIES ae ovate to lanceolate with obtuse oe rarely over 2 ¢ ong; bracteoles 6-8, united only at base and conspicuous. a var. lypica. Leaves strictly linear, often over 2 cm. fone, up to 3.7 cm. lon bracteoles usually inconspicuous, if conspicuous united caviar bi aitial 4 UG bale Gre ee DAS seca ee er a EE Oy UL ale b. var. linearifolia. EK. a re Engelm. in Emory, U. ex. io. Surv. 2 (1): . 1859, var. TYPICA : C. Wheeler, Bull. Torr. Bot. Club 63: ret 1936. Type: ‘“‘On the Sonoita [Creek] near Deserted Rancho,’ Santa Cruz County, Arizona, Sept. 15, 1851, C. Wright 1848 (M 144671!; photographs G!, W!; isoryPEs Gi; NY !), A very good representative of the species with short broad leaves and medium-sized appendages.—Chamaesyce pedic- poet Finer Rose & Standley, Contr. U. 8. Nat. Herb. 16: “a. aoluat Millsp., Proc. Calif. Acad. Sci., ser. 2, 2: 227. 1889. Type: Comondu, Lower California, eon 1889, T. S. Brandegee (F 196145!; phot G!, W!; IsoTyPE Cl). A plant from which the larger leaves have fa ile —Chamaesyce involuta (Millsp.) Millsp.. Field Mus. Pub. Bot. 2: 410. 1916.—E. pedicu- lifera Engelm. var. page ei I. M. Johnston, Proc. Calif. Acad. Sci., ser. 4, 12: E. conjuncta Millsp., Piee ‘Calif Acad. Sci., ser. 2, 2: 227. 1889. Type: Purisima, Lower California, Feb. 12, 188 9, T. S. Brande- gee (F 196147!: - photographs G!, W!; ISOTYPES Cl, G!). “eth somewhat narrowed at the bas ase, which is not unusual.—Cham syce conjuncta (Millsp.) Millsp., Field Mus. Pub. Bot. 2: 408. 1916. E. pediculifera Engelm. var. inornata T. 8S. Brandegee, Zoe 5: 209. 1905. Type: Cofradia, vicinity of Culiacan, Sinaloa, Mexico, Oct. 23, 1904, T. S. Brandegee (C!; isotypes F!, G! "A minor variant with appendages of glands lacking and rather short internodes. E. vermiformis M. E. Jones, Contr. West. Bot. 16: 23. 1930. Type: Ajo, Pima County, Arizona, Sept. 18, 1929, M. E. Jones 24856 (P!; isotypes G!, NY!). A variant with long internodes 184 Rhodora | [May and long narrow leaves approaching E. pediculifera var. lineari- folia.—PuatE 664C, Fics. 1-9. 3 , Colorado Desert, California, southern Arizona, Baja Cali- fornia, Sonora, and Sinaloa (Map 6). Representative specimens seen: CALIFORNIA. Imperial Co.: upper end of Painted Gorge, Carisso Mountains, Ferris & Rossbach 9624 (G). Arizona. Yuma Co.: near Mohawk, Peebles & Harrison 5021 (US); Dome to Castle Dome, Peebles & Kearney 10939 (US). Yavapai Co.: Castle Creek, Bradshaw Mountains, Toumey 260 (US). Mar- copa Co.: Black Cafion Road, 23 miles north of Phoenix, ees 8665 (US); Camp Creek, Harrison 1938 (US). Pinal Co: re soil, 44 mile north of Mammoth, Maguire, Richards & ee 10834 (G, I); Oracle, Newlon 699 (J); near Maricopa, Pee “ti, Harrison & Kearney 4909 (US). Pima Co.: Picture Rocks, Tucson Mountains, Bartram 326 (US); sandy wash-bed, D: miles east of Tucson, Maguire, Richards & Moeller 11222 (G, )); Quitovaquito, Mearns 2746 (US). Santa Cruz Co.: epee and Nogales, Peebles, Harrison & Kearney 5628 (US); hi between Calabasas and Nogales, Tidestrom 802 (US). Coe Re 0.: Bowie, Lemmon 283 (G). MEXICO: Lower Cu San Marcos Island, Johnston 3641 (C, G, US); San Luis - a Bay, Johnston 3331 (C, G, US); near El Marmol, Wiggins *G (G, US); Carmen Island, Nov. 1-7, 1890, Ed. Palmer 836 ‘ } US); Cocopa Mountains, MacDougal 122 (NY); Santa nee north of flying field, Ferris 8697 (US); San Felipe, Goldman Us) (US); Isla Partida, Collins, Kearney & Kempton 145 ie i Sonora: Hacienda Oquito (Cutting’s Ranch) 6 miles eas i: Altar, Wiggins 5967 (US); granitic hills, 5 miles east of Garu a bullo, Wiggins 6125 (US); Bacum Station near Rio Yaqui, Penn 13 20214 (US); 12 miles east of Libertad, M acDougal & Shreve ee ); New Year’s Mine, 20 miles south of Hermosillo, M. E. J wis 3 22617 (G); 7 miles west of Mina San Jose on road to Misa, Won 6311 (US). Sryatoa: Topolobampo, Rose, Standley & i 13276 (US). For citation of additional specimens see Dull. Torr. Bot. Club 63: 442-443. 1936. Some of the plants of Arizona, particularly M. E. Jones 24856, approach the linear-leaved var. linearifolia. If, however, the leaf variations here included under var. typica were all named little but confusion would result. Some of the plants from the hottest and driest parts of the deserts have very small whitish- tomentose leaves resembling E. melanadenia from which the very different seeds distinguish it. The specimen chosen here as type was left unnamed by Engel- mann. Nevertheless, this is taken as type in preference to the 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 185 collection of Schott which Engelmann did name, for Schott’s specimen lacks seeds and has but few cyathia. Too much weight need not be given the fact that Engelmann did not name the cited sheet of Wright 1848 in his herbarium for there is another case in which Engelmann failed to name his specimens. Of the three numbers cited by Engelmann as E. glyptosperma var. tenerrima none was named! 22b. E. P agg a Engelm. var. LINEARIFOLIA 8. Wats., Proc. Amer. Acad. Arts & Sci. 24: 76. 1889. Typr: high moun- tains, Guaymas gi nade Merion, Sept., 1887, Bd, Palmer 216 ATE Local about Guaymas, treat Additional eens seen: Guaymas, 1893, 7. S. Brandegee (C); among rocks at foot of hills, gia mas, Nov., 1887, Ed. Palmer 627 (C, F, G, “Us). San Pedro Bay, T. Craig 671 (FI. This variety, though extralimital, is included here for com- pleteness since some of the Arizonan plants approach it. 23. EUPHORBIA CINERASCENS Engelm. in Emory, U. 8. & var. Subinappendicult Loa ove Proc. “Amer. Acad. Arts & Sci. 5: 172. 1861.1. Boissier in DC. Prod. 15 (2): 32. 1862. “Chamaesy et ihe (Engelm. ) ‘Small, Fl. Se. U.S. are 1333. 1903. ers ial, forming mats up to 50 cm n diam.; dogg Wy 30 ss tee mania or decumbent, mostly andes (1 mm. diam.), slothics with crisped, short, mostly appressed hairs, segs p .5 em. long, average ca. 1 cm.; leaf-blades 2-9 mm. long, ovate with oblique base to oblong with slightly oblique base, usually glabrous above, closely tomentose to sees a genta petioles tomentose, 1-2 mm. long; stipules hairy, ca. 0.6 mm. long, — united, linear, dorsal distinct, linear; Saraeles less m. long, wi with shor appressed hairs; ¢ waitin solitary at the nodes; involucres turbinae 1.2-1. 5 mm. aa appressed- narrow or usually wanting; ‘fifth gland absent; sinus U-shaped, not depressed, densely hairy; bracteoles forming a radial —— age opposite each pn eiery often united only below, with 5 or 6 very slender short-hairy free segments above; staminate ower 3-4 per fascicle, 15-20 per cyathium; androphores 1.5-2 m 1 Date according to Trelease & Gray, Bot. Works Geo. Engelmann, 439. 1887. 186 Rhodora [May long, glabrous; gynophore shortly appressed-hairy, exserted and reflexed at maturity; ovary copiously hoary-tomentose, roundly three-lobed; styles parted nearly to the base, 0.5-0.8 mm. long, short-hairy below, clavate; capsule 1.5-1.75 mm. long, ovoid, sharply angled, very short-tomentose; seeds quadrangular 1.2- 1.5 mm. long, ca. 0.6-0.9 mm. radially, ca. 0.6-1 mm. tangen- tially, facets smooth or faintly wrinkled, oblong or often deltoid- oblong radially (i. e. wider below), base obtuse or truncate, apex acutish, coat white, microreticulate—PLaTr 663B. outhwestern Texas, Chihuahua, Coahuila, Nuevo Leon, San Luis Potosi, and Tamaulipas (Map 27). Representative specl- mens seen: Texas. Brewster Co.: Lechuguilla Flats out of Green Gulch, Chisos Mountains, Sperry 449 (US); Chisos Mountains, Mueller 8080 (M, NY, US). Kimble Co.: 5 miles west of Roose- velt, Cory 21208 (G). MEXICO. Curnvanva: limestone hill- side, pass 19 miles east of Jimenez, Johnston 7851 (G); silty plain 8 miles northwest of Cruces, Johnston 7987 (G); gravelly benches, pass between Chilicote Station & Las Animas, Johnston 7997 . Coanuma: Juarez on the Sabinas River, 100 miles north of Monclova, Sept., 1880, Ed. Palmer 1204 (F, G, US); rocky slopes of canyon, 5 miles north of Saucillo, Johnston 7911 (G); on desert plain, 7 miles south of Hipollito, J ohnston 7240 (G) ; desert 41 miles west of Saltillo, Johnston 7694 (G); Municipio de Ramos Arizpe, dry mountain slope east of Hacienda la Rosa, Wynd & Mueller 37 (G). Nurvo Lron: Monterey, Feb. 17-26, 1880, Ed. Palmer 1197 (G, US). San Luts Porost: Estacion de Catorce, Sierra Madre Oriental, gravelly bed of arroyo, Pennell 17570 (US); Chareas, Lundell 5196 (US); rocky slopes of a hill, 11 miles south of Matehuala, Johnston 7577 (G). ‘TAMAULIPAS: Victoria, May 1—June 13, 1907, Ed. Palmer 548 (US); Cerro de la Tamaulipeca, near San Miguel, Sierra de San Carlos, Bart ett 10559 (US). For citation of additional specimens see Bull. Tort. Bot. Club 63: 439-440. 1936. 24. EUPHORBIA VALLIS-MoRTAE (Millsp.) J. T. Howell, Madrofio 2: 19. 1931. Typ: a few kilometers north of Indian Wells, between Mohave and Keeler, Kern County, California, June 21, 1891, Coville & Funston 1008 (US 16203!; fragment F!)— eet paid vallis-mortae Millsp., Field Mus. Pub. Bot. 2: 403. _ Perennial, usually forming a dense rounded plant up to 15 em. high; herbage hoary-tomentose throughout; stems usually brown and glabrous, aérial portion to 1 mm. diam., internodes SRR a ae ee eae eT aa 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 187 filiform, ca. 0.7 mm. long, densely hairy; cyathia solitary at the nodes; peduncles stout, to 1.5 mm. long, densely hairy; involu- cres campanulate, ca. 2 mm. diam., densely hairy without, with long erect hairs extending halfway down within opposite glands; lobes with long ascending hairs within, equaling or slightly exceeding the glands, deltoid, ae - glands s yellowish or reddis transversely oblong, to 1 mm. long, the distal slightly shorter: appendages white, as wide as and a "little longer than the glands, entire or crenulate, with numerous short hairs beneath and on the margins and a few above; fifth gland absent; sinus U-shaped, with long erect hairs at the bottom; bracteoles mostly united into one group of 6-10 bracteoles opposite each gland, more or less united below, adnate to the involucre, sometimes with shorter bracteoles outside the fascicle, all hairy above; staminate flowers 3-5 per fascicle, 17-22 per involucre: androphores ca. 2 mm. long, slightly exserted, sometimes with a few short hairs ers gynophore densely hairy, long-exserted ae reflexed at maturity; ovary three-lobed, densely hairy; styles ca. 0.5 mm. long, parted to the middle, short- hairy below; capsule tomentose, three-angled, 2 mm. long and in diam.; seeds ‘sharply quadrangu- lar, 1.4-1.7 mm. gee ca. 0.7 mm. tangentially and radially, ovate radially, raphe straight, back rounded in tangential silhouette, base obtusely truncate, facets smooth or nearly so, ventral facets concave, dorsal facets slightly convex, coat white, microreticu- late.—PLATE 663C. Eastern pi of the Sierra Nevada from northwestern Mohave M, P). 617 (P), Purpus 5473 (F, G, J, M, US); 6 miles north of Freeman, Hoffmann 585 (CA, SB); Dove Springs, 1931, Hoffmann (SB); Red Rock Canyon, J. T. Howell 4973 (CA, Peir). The particular locality-data for the type collection are lacking on the label but are given by Coville, Contr. U. 8. Nat. Herb. 4: 256. 1893. 25. EvUPHORBIA MELANADENIA Torrey, Rep. Expl. & Surv. Miss. R. to Pacific Ocean 4: 135. 1857. Types: “‘ Low places near San Gabriel’, Los Angeles County, California, 1853-4, J. M. Bigelow (NY!; photographs G!, W!; 1sorype G!). A good representative of the species. Munz, Man. So. Calif. Bot., 89, fig. 153. 1935, good wie styles s should be 3.—Anisophyllum melanadenium (Torr.) Klotzsch & Garcke, Abh. Akad. Berlin, Wats., Bot. Calif. 2: 73. 1880.—Chamaesyce melanadenia (Torr. ) Millsp., Field Mus. Pub. Bot. 2: 410. 1916. 188 Rhodora [May E. cinerascens Engelm. var. appendiculata Engelm. in Emory, U.S. & Mex. Bound. Surv. 2 (1): 186. 1859. Types: San Felipe, San Diego County, California, May, 1852, Geo. Thurber 628 (M 46715!; tsoryprs G!, NY!). Differs in no consequential respect.—E. polycarpa Bentham var. appendiculata (Engelm.) Munz, Bull. So. Calif. Acad. Sci. 31: 68. 1932. Chamaesyce aureola Millsp., Field Mus. Pub. Bot. 2: 406. 1916. Type: Azusa, Los Angeles County, California, alt. 800 feet, May 3, 1912, H. H. Smith 4933 (F 389282!, photographs Gl, W!). A good match for the type of E. melanadenia. . polycarpa Bentham sensu Thurston, Wild Flowers 50. Calif., 181, fig. 274. 1936 (photograph). Perennial from a taproot as much as 5 mm. in diam.; stems ascending or erect, to 20 em. long, sometimes stout (1.5 mm. diam.) below, closely tomentose, glabrate; leaf-blades 2-9 mm. long, ovate to ovate-lanceolate, base oblique, closely and often hoary tomentose on both surfaces, petioles clothed, as the leaves, 1-2 mm. long; ventral stipules mostly united, linear, hairy, to 1 mm. long, dorsal stipules distinct, linear, hairy, to 1 mm. long; peduncles less than 1 mm. long, with short appressed hairs; cyathia solitary at the nodes; involucres open-campanulate, 1.2-1.5 mm. diam., appressed-short-hairy without, glabrous with- in except below the glands; lobes narrowly deltoid, copiously hairy, equaling the glands; glands transversely oblong, da? reddish; appendages usually conspicuous, twice as wide as and longer than the glands to rarely wanting, white, margin crenate to subentire, glabrous or rarely with a few short hairs beneath next to the gland; fifth gland absent; sinus U-shaped, not de- pressed, densely hairy; bracteoles more or less completely united reflexed at maturity; ovary copiously hoary-tomentose, roundly three-lobed; styles parted nearly to the base, 0.5-0.8 mm. long; short-hairy below, slender throughout; capsule 1.5-1.7 mm. long ovoid, sharply angled, very short-tomentose ; seeds quadrangular, 1.2-1.5 mm. long, ca. 0.6 mm. radially and tangentially, facets smooth or slightly wrinkled, apex acutish, coat white, mict® reticulate —PLatr 663A. : Southern California, southern Arizona, northern Baja Cali- fornia including Guadalupe Island, Sonora (Map 12). Repr® 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 189 slopes of Sierra Madre Canyon, San Gabriel Mountains, July 29, 1927, H sets (NY); Lone Hill, near Glendora, Munz & Eggles- ton 196 2 (G). San Diego Co.: San Felipe Valley in Agave aatohien Reed 5833 (O); Yaqui Wells, Colorado Desert, Eastwood 2778 3 (G, NY). Imperial Cott mile east of Mountain Springs, Wiegand & Upton 3742 (G). Arizona. Yuma Co.: Pass, Lemmon 296 (G). Yavapai Co.: Copper Basin, Towmey 251 (NY); on dry mesa, Big Bug, July 21, 1891, Toum mey (US). Maricopa Co.: Agua Fria, Coues & Ed. Palmer 264 (M); among the rocks, Canyon Lake, A. Nelson 11216 (1); road banks along Apache Trail, west end of Canyon Lake, A. & R. Nelson 1709 (M, NY). Pinal Co.: rocky south slopes, 5 miles west of Superior, Maguire, Richards & Moeller 10263 (G, 1); Oracle sored Sta- tion, Coronado Forest, Eggleston 15967 (G, US). a Co.: La Osa, Mearns 2688 (US); Canyon Diablo, Ajo Ricuntis Peebles & Kearney 10886 (US); Santa Catalina Mountains, Shreve 5154 (G, US); Fresnal, Thackery 83 (US). Gila Co.: Collom’s camp at foot of Matzatzal Mountains, A. & R. Nelson 1955 (G); Globe, Kearney & Peebles 12060 (NY); peat slopes of sandstone, Cassadore Spring Canyon, San Carlos Indian Reser- vation, Maguire, Richards & Moeller 10301 (1); Collom he ek Matzatzal Mountains, Collom 33 (M, NY, US). Nava ee Fort Apache, 1892, Hoyt (NY). Graham Co.: ro ocky s oil, miles east of Coolidge Dam, US Highway 180, Maguire, Rickards & Moeller 13024 (G, 1). Cochise Co.: Pinery Creek, Chiricahua Sida ary Aug., 1896, Fernow (US). MEXICO. "BAJA ni Jepson, Nan. Fl. Pl. Calif., 600. 1925" includes under Euphorbia polycarpa var. vestita three entities, judging by the range given: E. melanadenia “Santa Monica; Glendora; Cahuenga Pass”’’; E. polycarpa var. hirtella, at least in part, ‘Colorado Desert”; E. vallis-mortae, “Inyo Co.”. However, Jepson, Fl. Calif. 2: 429. 1936, has the entities correctly delimited and named except that the proof of the statement that E. melanadenia occurs in “western Nevada” has yet to be supplied. I find neither explana- tion nor support for it in the Euphorbiae of Jepson’s herbarium which he so kindly loaned to the Gray Herbarium for my use. ‘ ‘ Title-page date questionable but here accepted as no question of priority is Nvolved, 190 Rhodora [Mar That the type of H. melanadenia did not come from “Low places near San Gabriel” is highly probable since this plant Is confined to a narrow zone on the foot of the mountains in this region. Probably Bigelow collected it in the vicinity of Sierra Madre. 26. EvPHORBIA POLYCARPA Bentham, Bot. Voy. Sulphur, 50. 44 : us at the branch-tips, campanulate, 1-1.5 mm. in diam., glabro . attenuate, equaling or slightly exceeding the glands, hoe glands maroon, transversely oblong, 0.5-0.75 mm. long; ae a ages up to three times as wide as the glands to absent, as woes or longer than the gland, white or reddish, entire or ¢ rangular, 1-1.3 mm. long, ovate in radial outline, 0 btuse; radially and tangentially, apex acutish, base truncate % 0 se angles sharp, back curved, ra aight, micropy 7 ng of slightly truncated, facets smooth or slightly wrinkled, Pk ntly concave, the back facets lower than the angles, i. ee © hat depressed, coat micro-reticulate, white, opaque, or so thin the brown testa shows through. 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 191 Key TO VARIETIES ad te Ligeti petioles ca. 14 as long as lea Appen e to narrow and herbage ene essentially Py A ead oe odie ee wick Soe Oe Le ie a ee a. var. typica Appendages narrow and herbage pubescent............... b. var. hirtella, Appendages absent, petioles ca. 14 as ie i ‘iia Se oe c. var. simulans 26a. E. potycarPa Bentham, Bot. Voy. Sulphur, 50. 1844, var. Typica L. C. Wheeler, Bull. Torr. Bot. Club 63: 408. 1936. Type: Bay of Magdalena, Lower California, Mexico, 1841, Hines (K!; photographs G!, W!; fragment F!). Boissier ‘in DC. 25 ;—Chamaesyce hes ages se (Benth.) Millsp. ex Parish, Cat. PL Salton Ber 6. 1913 (preprint from Carn. Inst. Wash. Pub. 193: 110. 1914. \—PLATE 657 California and Nevada, a“ to Lower California and Sonora (Map 4). Representative specimens seen: CALIFORNIA. Los Angeles Co.: Eagle Rock foothills, Rockwell 300 (J). San Bernar- dino Co.: the Needles, M. E. Jones 6178 (I, O); Dunes, Needles, Parish 9608 (G, M).. Riverside Co.: Elsinore, Apr. 1 892, Mc- Clatchie (NY); slopes of Box Springs Mo ountains, Riverside, Nov. 12, 1919, Barrus 7 (QO); between Cottonwood Mountains and Mecca, McKelvey 5038 (G); near Desert Center, M. E. as ones 24860 (G). San Diego Co. Epos Diego, salage 615 (G, NY); Sweetwater valley, Apr. 30, 1883, G. C. Deane (G); Yaqui Wells, Colorado Desert, Pees: 2766 (G); aoe flea Canyon, Jepson 12475 (J); i we Meyer 230 (J); Del Mar grade from a Jolla, Newlon 312 (J). Nevapa: 8 miles above Rioville, M. E. Jones 5085 (MD; hawt River, Goodding 708 ube ARI- reset ah & Peebles (US); Yuma, Apr. 21, 1913, Wooton (US) Maricopa Co.: Phoenix, June 20, 1891, Dewey (US); "a Tempe, Gillespie 8415 (U Black Canyon Road, 23 miles north o 8); Phoenix, Gillespie 8666 (US); Hyder, Peebles Sn (US); near Phoenix, Peebles, Harrison & sergio 2461 (US); 5 miles east of 2300 ‘ Pim Apr. 8, 1 1884, Pringle (U gy. For citation of additional specimens se ¢, Bull Torr. Bot. Club 63: 408. 1936. 26b. E. potycarra Bentham var. HIRTELLA Boiss. in DC. Pro 15 (2): 44. 1862. Typx: California, Emory (Ge; fragment F)— Chamaesyce eres ye var. hirtella (Boiss.) Millsp. ex Parish, Cat. Pl. ae —_ 6. 1913 (preprint from Carn. Inst. Wash. Pub. rire 14)—C. tonsita Millsp., Field Mus. Pub. Bot. 2: iste. Beene of California and southern Nevada, south to lower 192 Rhodora [May California and Sonora (Map 3). Representative specimens seen: CALIFORNIA. San Bernardino Co.: Soda Lake Mountains near Borego Valley, Duran 3176 (G, I, O); Signal Mountain, Colorado Desert, Abrams 3187 (G). Imperial Co.: upper end of Painted Gorge, Carisso Mountains, Ferris & Rossbach 9605 (G). NEVADA: the Muddy Range, Goodding 2222 (G). Yuma Co.: Yuma, Feb., 1881, Vasey (US); Yuma, Nov. 6, 1909, Mowry (US); Aatec, Helena Canyon, Aug. 8, 1938, Warnock C506 , in part (US); ei Chisos Mountains, Young 139 (M); between Goat and Trap Mountains, E. J. Palmer 34207 (NY); near San Vincente, Sper?y 1358 (US); Santa Helena Canyon, June 7, , Wa Ff (US). Presidio Co.: Presidio, Sept. 27, 1937, Warnock 79 (US). (Map 5). 27. Evpnorsia Parisui Greene, Bull. Calif. Acad. 2: 56. 1886. Tyrer: Warm Springs, Mohave Desert, San Bernardino i _ B. & W. F. Parish 1384 (probably lost when the herbarium of California Academy burn 9 in 1906, for Dr. Theodor Just states in letter of Feb. 25, 199% filed at Gray Herbarium, that there is only a fragment Herbarium Greeneanum at Notre Dame University; 1soTTPe D!, M!, NY!).—Chamaesyce Parishii (Greene) Millsp. ex Paris Cat. Pl. Salton Sink, 6. 1913, preprint from Carn. Inst. Pub. 193: 110. 1914.—E. polycarpa Bentham var. P arishit (Greene) Jepson, Fl. Calif. 2: 429. 1936. ‘tenet een 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 193 E. patellifera J. T. Howell, Leafl. West. Bot. 1: 53. 1933. Typr: Palm Wash, sii Colorado pons San Diego County, California, J. T. How 488 (CA!; IsoTYPE F!). Perennial, orming pire mats 20-50, cm. “across, or a low bush 15-20 em. high; stems slightly woody below in age, slender, glabrous, internodes 5-15 mm. long; leaf-blades mostly ovate, 2-4 mm. long, entire, pom. or very rarely tomentulose be- neath, base oblique, apex mucronulate, midrib evident at least in lower half of Blades etka 0.5-1 mm. long, glabrous, am- plexicaul on ventral side of stem; — stipules distinct, mostl entire, ciliate, broadly linear, 1 mm. long, lower sti ipules often more or less united, ciliate, linea mre m. long; peduncles up to 1 mm. long, glabrous; chad on Thies at the nodes; involucre campanulate, tapering to the peduncle, 1—-1.2 mm. diam gla- brous without, with many short hairs within above; ed broadly deltoid, mostly age: eon on inner face, equaling the glands; glands discoid, 0.5 mm. diam., pale ‘yellow or reddish, on stipes ca. half as w mor 5 as t ee gland; stipes ciliate on inner side: appendages absent; fifth gland ciliate on inner side, linear, mostly shorter than t e lobes; sinus U-shaped, not depressed ; bracteoles united for half their length, forming a membranous radial appendage ca. 1.3 mm. long, adnate for half its length to the involucre opposite pci gland, glabrous agit) ciliate above; staminate flowers 8-10 per fascicle, 40-50 involucre; androphores glabrous, ca. 15 mm. long, slightly exserted at maturity; gynophore glabrous, long-exserted and usually reflexed at maturity; ovary glabrous, vicoe dalled: styles ca. 0.5 mm. long, bifid to the middle, glabrous; ci glabrous, sharply ‘three-angled, oblate-spheroid, ca. 1.75 m long; a. ca. 1.5 mm. long, ca. 0.75 mm. tangentially, ca. 0. 65 mm. radially, Rey aeaprie ee covets in radial outline, raphe straight, slightly truncated above, back sharply angled, facets faintly wrinkled, coat white, microreticulate. —P.aTE 658B. Deserts of Inyo, Kern, San Bernardino, Riverside, and San Diego Counties, sen east to Nevada (Map 25). Repre- sentative specimens se n: CALIFORNIA. Inyo Co.: Furnace Creek Ranch, Death Valley, yr 30, 1917, alt. to 100 ft., W. L. Jepson (J); Stove Pipe Wells, Death Va lley, P. A. Mun z2&C.L. Hitch- cock 11032 (P); near Triangle Bessie: Death Valley growing in dense brown mats on pebbly wash-fan, Apr. 17, 1917, J. Grinnell (J); Surprise Canyon, S. ne Parish 10217 (Cc, J); Pe oy aay Canyon, Panamint Mountains, R. S. Ferris, F. M. Scott & R. Bacigalupi 3998 (D); Baga Springs a B. Hoes 10190 Motaies Desert, 1926, M. E. Jones (P). Nevapa. Lincoln Co.: Muddy Valley, alt. i, 700 ft., Kennedy & Goodding 77 (NY, US). 194 Rhodora [May For citation of additional specimens see Bull. Torr. Bot. Club 63: 406. 1936. 28. EUPHORBIA MICROMERA Boiss. ex Engelm., Proc. Amer. Acad. Arts & Sci. 5: 171. 1861; Boiss. in DC. Prod. 15 (2): 44. 1862. Typn: bed of a creek descending to the San Pedro River, Co- chise County, Arizona, Sept. 8, 1851, C. Wright 1854 (M 149918}; photographs G!, W!, isotypes F!, G!, Ge!). Small and essentially glabrous, a good representative of the species.—E. polycarpa Bentham var. micromera Millsp. ex Orcutt, West Amer. Scientist 10: 134. 1901; with neither basinym nor description, identity inferred from the coincidence of names.—Chamaesyce macromera (Boiss.) Wooton & Standley, Contr. U. 8. Nat. Herb. 16: 144. 19138. E. pseudoserpyllifolia Millsp., Pittonia 2: 87. 1890. TYPE: Bowie, Cochise County, Arizona, Sept. 15, 4, M. E. Jones 4228 (F 196599!; photographs G!, W!, 1sorypss G!, I!, P!, US!) Differing little from the type of E. micromera which was not considered.—Chamaesyce pseudoserpyllifolia (Millsp.) Millsp., Field Mus. Pub. Bot. 2: 411. 1916.—E. pseudoserpyllifolia Millsp. forma typica J. T. Howell, Leafl. West. Bot. 1: 52. 1933. E. podagrica I. M. Johnston, Univ. Calif. Pub. Bot. 7: 44¥: 1922. Type: washes at Gold Mountain, Esmeralda County, Nevada, 1898, C. A. Purpus 6437 in part (C 110920!). Differing in no essential respect from the type of E. micromera which was not considered. E. pseudoserpyllifolia Millsp. forma villosa J. T.. Howell, t. 1: 53. 1933 Y E. setiloba Engelm. var. nodulosa Jepson, Fl. Calif. 2: 427. 1936. Type: between Brawley and Salton Sea, Colorado Desert, Imperial County, California, Oct. 15, 1912, 115 feet below wie level, S. B. Parish 8301 (J!; photographs G!, W!; 1soTYPES D:, F!,G!). This is the nodulose vestite variant local in the Colorado Desert. Some of the glands bear minute appendages. Possib f with more numerous collections this variant may prove itsel Prostrate annual: stems glabrous or pubescent, extremely variable, one extreme very straight, thick, with thickened nodes and internodes up to 1 cm. long, the other extreme flexuous OF straightish, slender, nodes not thickened, internodes up to 2 ¢™ long; leaves glabrous or short-pubescent, blades 2-7 mm. long; ovate and base markedly oblique in the larger, oblong and ast slightly oblique in the smaller, petioles ca. 0.5 mm. long; stipules ca. 0. - long, or shorter in pubescent plants, triangular, ciliate, upper distinct, lower often united toward stem-tlP 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 195 peduncles glabrous or pubescent, up to 1 mm. long; cyathia solitary in the axils; involucres ca. 0.9 mm. in diam., very short- campanulate, narrowed above, more or less cuneate to the peep glabrous or pubescent without, glabrous within except e lobes, green-veined beneath the lobes; lobes deltoid, equaling or slightly exceeding the glands, hairy within; ; glands pink or red, strictly discoid or transversely oblong, especially the proximal, 0.1-0.15 mm. diam.; appendages absent or, in some egies 58 n margins; fifth gland absent; sinus broadly V-shaped, hairy, little — ressed ; bracteoles reduced to a solitary linear hairy appendage PO. 5 mm. long, adnate for most of its length to the involucre aes the glands; staminate flowers 2—5 per involucre; andro- phores glabrous, included, 0.7-0.9 mm. long; gynophore glabrous throughout or short-hairy above, long-exserted and usually reflexed at maturity; ovary three-angled, csbeots to pales cent, carpels slightly grooved on the back; styles bifid, glabrous, ca. 0.2-0.3 mm. long, clavate; capsule ‘three-angled, parr to pubescent, spheroid, ca. 1. 3 mm. long; seeds quadrangular, 1.1— l. m. long, 0.5 mm. tangentially, 0.4 mm. radially, narrowly ovate "yadially. angles sharp, facets smooth or with very faint wrinkles, convex, especially the front, base truncate, raphe straight or slightly concave, shortly truncate at a slight angle above, microreticulate white _— thin, with the brown of the testa showing through.—PuatTE 658C. eserts from Inyo County aod to Imperial County, Cali- fornia; Esmeralda and Clark Counties, Nevada; San Juan County, Utah; Arizona; Grant and Dofia Ana Counties, New exico; Reeves and Brewster Counties, Texas; Chihuahua; and Coahuila and Peru (Map 26). Representative specimens seen: CALIFORNIA. San Bernardino Co.: Daggett, Mohave Desert, Oct. 13, 1933, Beal (J). Nevapa: Clark Co.: near Boulder City, Eastwood & Howell 6292 (G). Uta: Sa ee Co.: along San Juan River near Ces Rydberg & Garrett 9896 NY). Arizona: Yuma Co.: south of Quartzsite, rinse Peebles 1 iii — Mohs wk, Pesblee, Harrison & Kearney 1976 (US). nal Co.: Sacaton, Peebles 5000 (US). Pima Co.: Wil- mot, i yang? reserve, — 341 (US). Gila Co.: Sierra Oct. 4, 1879, Zuck (NY, US in part). NEw Mexico: Grant Co., of Charco Piedra, Johnston 7925 (G). Coanuria: 3 miles soutl of Pefia, Johnston 7729 (G). For citation of additional iia: mens see Bull. Torr. Bot. Club 63: 432-433. 1936. 196 Rhodora [Mar This species is nowhere abundant and, while wide-ranging, occupies only scattered stations. Formerly, |. ¢., I included Orcutt 1331 from Socorro, northern Baja California here. That collection, the basis of the nomen nudum, Euphorbia baja cali- fornica Millsp. ex Orcutt, West Amer. Scientist 10: 134. 1901, differs in having styles ca. 0.6 mm. long and scarcely clavate; involucres ca. 1.1 mm. in diam.; staminate flowers 7-8; seeds strongly and irregularly ridged and glands often appendaged. (The specimen which is the type, if nomina nuda are worth typ fying, is F 197073! for this came from Millspaugh’s herbarium and is labeled ‘‘E. Baja-Californica sp. nov.” There was some mistake made since the plants on the sheet are E. cordifolia; only the fragments in the pocket are the Lower Californian plant. Probably a mixture occurred during mounting.) This entity may be only worth varietal recognition, but, being extra-limital, is excluded here. Likewise Ed. Palmer 789 (US), Baja Califorms, Lagoon Head, is excluded as it seems to be the same as Ore 1331. The reason that the specimen at M rather than the specimen at Ge is taken as type is that in the loan from Ge there is included only the merest fragment which is mislabeled as ‘‘ Fendler 0. 1854” when it should have been “ Wright no. 1854”. Further- more the label accompanying this fragment bears no name. The piece sent may be a portion of Boissier’s specimen but that 18 not certainly known. In view of the fact that Engelmann publish 7 the species first, attributing it to Boissier, and left a good speck men with a label bearing the name of the plant, it seems justi- fiable to take Engelmann’s specimen as type. The following new example of common identities betwee? North and South America is to be noted: Shale cliff above 5° alt. 0-20 m., Paita, Dept. Piura, Peru, July 4, 1925, F. W. Pennell 14815 (G). EvPHorBIA corpIFoLIA Elliott, Sketch Bot. So.-Car. & Georgia 2: 656. 1824.1. Typn: “ Grows in cultivated land, comma around Beaufort [South Carolina] in dry soils.”’ (Charleston, Sout Carolina, Museum).—Ch : cordifolia (Ell.) Small, Fl. 8¢ U8, 100, 1588, Oe ee _ Annual, glabrous; stems prostrate, or occasionally ascendiié in small plants, 4-35 em. long, 0.5-2 mm. thick, inter? o~ J. H. Barnhart, Dates of Elliott's Sketch, Bull, Torr. Bot. Club 28: 680-6" — — 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 197 rarely up to 6 cm. long, mostly 2 em. long or shorter; leaf-blades elliptic-orbicular to oblong and ovate-oblong, 4-12 mm. long, base more or less inequilateral, often cordate, margin entire; petioles ca. 1 mm. long; stipules parted to the base into few to several filiform segments up to 1.4 mm. long, mostly with short scattered hairs at least when young, pe distinct, ventral often united; peduncles 0.4-4 mm. long; cyathia solitary at the nodes and at the branch-tips but often congested by the marked one: of the upper internodes; involucre broadly campanulate, 1.6 mm. in diam., glabrous outside, glabrous inside except at the base of the thes. and beneath the glands; debe subulate, pubes- cent below, glabrous — posting exceeding the glands; glands transversely elliptical to nee e, hee Segre ia 0.5-0.9 mm. ong; neste pre fro wide gland, to 1.3 mm. wide, the wider "adially broadly elliptical to reniform, glabrou , entire or with two or three low blunt teet fifth gland pre pe of 1 or 2 Case filiform segments, daleans above, equaling the lobes; sinus U-shaped, not depressed; or ca. rel low, reflexed; ovary glabrous, S-angled: sare parted to the base, 0.6-0.9 mm. long; capsule see ase para 3-angled, hepa Lorn the equator, 1.7—2.1 mm. lon eeds ovoid-triangu 1.2-1.5 mm. long, 0.7-0.9 mm. tact gentially and radially, sy ovate to oblong-ovate, usually acute, or with lo t wrinkles, slightly concave to slightly convex, angles hes pete white, microreticulate, mostly — “ee to little obscure the pale, (NY). Sours Caroma: “ Sand hills of S. C.’’. Oct. ?, Ravenel (G). Grorera. Richmond Co.: Augusta, Aug., 1902, Anon. NY). Macon Co.: on ‘ae of Andersonville stockade, Sept. 5, 1897, Harper (NY). Dooly Co.: near Flint River, Harper 574 (G, NY). Dougherty Co.: Albany, Tracy 4710 (NY); pine barrens bordering the Altamaha River, Curtiss 2469 (G, NY). hegre Lake Co.: near Eustis, Nash 1070 (G, NY). Orange : 1894. Lewton (NY). Polk Co.: Haines Ci ity, Curtiss 5959 G, NY), Hillsborough Co.: west coast, 1886, Curtiss (G). alton Co.: summer, 1885, Curtiss Paes Escambia Co., Heoteeos Herb. 5895? (NY). Suwanee Co.: 5 miles west of Live Oak, Wiegand & Manning 1801 (G). Gilchrist Co.: Hammock 198 Rhodora [May along the Suwanee River east of Old Town, J. K. Small, J. W. Small & DeWinkeler 11470 (NY). Mussissrppi1. Jackson Co.: Horn Island, Tracy 6370 (NY). Harrison Co.: Cat Island, F. E. Lloyd & Tracy 208 (G, NY). Lovistana. Rapides Co.: Alex- andria, Hale (NY). Texas. Tarrant Co.: in field, Ruth 686 (NY). Medina Co.: 30 miles west of San Antonio, Sept., 1879, Ed. Palmer 1212 (G). Hays Co.: San Marcos, Stanfield (NY). Colorado Co.: 6 miles northeast of Alleyton, Cory 25096 (G) Walker Co.: near Huntsville, Dizon 335 (G, NY). Waller Co.: Hempstead, Hall 547 (G, NY). Jefferson Co.: Sabine Pass, July, 1884, Neally (G). Nueces Co.: near Corpus Christi, Mar., 1894, Heller (NY). The type was recently examined by Professor Fernald who reports that it is identical with the usual interpretation of the species as exemplified by the two following collections from Georgia with which he compared it: A. H. Curtiss 2469 (G); R. M. Harper 574 (G). There are two races of this species. This was discovered by the counts of the staminate flowers. In ten collections from west of the Mississippi River the number of staminate flowers Pe cyathium was 29-44; in ten collections from east of the Missis- sippi River the number per cyathium was 9-27. Having made this discovery a reexamination of the collections from these two areas was made in order to ascertain whether there were any more obvious differences between the two races. While there 1s ® tendency in the eastern plants to have smaller leaves and shorter internodes there are too many exceptions to make any practical or certain division on these characters. Examination of @ more ample suite of specimens would very likely produce intermediate numbers of staminate flowers. Since the species forms an 4 ceptable unit as an undivided aggregate no attempt will be made here to further distinguish the two races. Boissier in DC. Prod. 15 (2): 30. 1862 identifies Euphorbia ludo- viciana Raf., Fl. Ludovie., 111. 1817, with E. cordifolia. Since Rafinesque described his species as having leaves other than entire this identification must be erroneous. 30. EupHorsta serpens HBK., Nov. Gen. et Sp. 2:, (quarto), 41 (folio). 1817... Type: Cumana, Venezuela, Bye pland 407 (Herb. Mus. Paris; photograph G! fragment F!). ' See hart, Bull. Torr. Bot. Club 29: 585. 1902 as to date and for discussio? of the quarto and folio editions. Ss he tC a I Pe 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 199 Average as to habit and leaf-size, nodes rooting. Boissier in DC. Prod. 15 (2): 29. 1862; Millsp., Bot. Gaz. 25: 18. 1898, fifth gland broken in figure; Thellung, Bull. Herb. Boiss. ser. oy 755. 1907; N. E. Brown in Thiselton-Dyer, Fl. Trop. Afr. 6 (1): 511. 1911: Thellung i “ Ascherson & Graebner, Syn. Mitteleur. Fl. 7: 440-443. 1917; L. C. Wheeler, Bull. So. Calif. Acad. Sci. 33: 108. 1934. ey phi ws serpens (HBK. ) Klotzsch & Garcke Abh. Akad. Berlin, Phys. 1859: 23. 1860.—E. radicans Moricand ex Klotzsch & Garcke, op. cit., 24, as synonym of Anzsophyllum serpens.—Chamaesyce serpens (HBK.) Small, Fl. Se. U. 8., 709, 1333. 1903.—E. serpens HBK. A genuina Thellung in Ascherson & Graebner, Syn. Mitteleur. Fl. 7: 442. 1917. E. herniaroides Nutt., Trans. Amer. Philos. Soc. n. s. 5: 171. 1837. Tyrpr: Arkansas, ee Nuttall (PH!, or perhaps — photographs GI, Ww! E. flexicaulis Scheele, Linnaea 22: 153. 1849. Type: “nord- lich von Neubraunfels: Lindheimer. August.” Comal County, Texas (?). ei places it here.—E. serpens HBK. var. —— (Scheele) Coulter, Contr. U. > Nat. Herb. 2: 388. 1894 (the pier a E. serpens var. radic ey in synonymy). —E. serpens A genuina III flexic aie: (Scheele) Thellung in Adchertdhe: & Graebner, Syn. Mitteleur. Fl. 7: 442. 1917. E. serpens var. Tac Engelm. ex Boiss. in DC. Prod. 15 (2): 30. 1862. Typr: Tam oe ae Berlandier 140 (Ge!; photo- graphs G!, W!; pice “US 1 9354!). Merely a vegetational phase with roots at some of he nodes.—E. radicans Morican ex Boiss synonym nec Millsp. Field Mus. Pub. Bot. 2: 411. E. serpens var. imbricata Boiss. in DC. Prod. is (2): 30. 1862. Type: Texas, F. Lindheimer 693 (Ge!; photographs G!, W!). Internodes shortand leaves thick, presumably due to a dry habitat. Sat is the Chamaesyce H artwegiana (Boiss.) Small sensu Small, Se. U. ed. 2, 1349. 1913).—E. “‘herniaroides Nutt. var. imbrieata Engelm, ex Blankinship, Ann. Rep. Mo. Bot. Gard. 18: 149. 1907, as synonym of E. serpens HBK.; indexed in op. cit., E. serpens A genuina a Ul green b p oe Thellung, : c. Typu: Illinois (Zurich?). Referred lo from the descrip- = E. forbuserpens HBK. ex Wood & McCarthy, Journ. Elisha Mitchell Soc. 1885-6: 119. — epee gg description) is doubt- ess a lapsus calami for E. serp rostrate annual, often rt eats: herbage glabrous through- out; stems slender, up to 50 cm. long, internodes to 3 ¢ m. long, no des sometimes rooting; leaf-blades 2-7 mm. e ovate- orbicular to oblong, base oblique except in the smallest, one 200 Rhodora [May entire; petioles mostly less than 1 mm. long; both upper and lower stipules united into a white, membranous, glabrous scale less than 1 mm. long with more or less lacerate margin; peduncles up to 2 mm. long, expanding upward, glabrous; cyathia solitary at the nodes; involucres turbinate, 1 mm. long, 1 mm. diam., tapering to the thick upper end of the peduncle, glabrous with- out and within; lobes deltoid, sometimes with a few hairs on the margins below, apex acute to attenuate, equaling or slightly exceeding the glands; glands transversely oblong, ca. 0.2 mm. long, ochroleucous, concave; appendages mostly present, little wider than the glands, white, glabrous, margin mostly crenate; fifth gland linear, exceeding the glands, with a few marginal hairs at the base; sinus slightly depressed; bracteoles forming & single, linear or broader, entire or parted, slightly hairy append- 0.6 mm. long or shorter, adnate below to the involucre opposite each gland; staminate flowers 5-10 per cyathium; andro- phores glabrous, 0.9-1 mm. long, equaling the glands; gynophore glabrous, exserted and mostly reflexed at maturity; ovary lobed, glabrous; styles glabrous, markedly clavate, 0.2 mits long, usually parted to below the middle, rotately spreading; capsule glabrous, 3-angled, ca. 1.2 mm. long; seeds smooth, ovoid with rounded angles or somewhat turgid-quadrangulat with more prominent angles, ca. 1 mm. long, ca. 0.5 mm. radially and tangentially, coat microreticulate, white, with the brown testa showing through.—P.Larte 661C. Casual introduction in New Hampshire, New Jersey, Pennsy!- to Colorado and Montana, east to North Dakota, Iowa, 4? d Illinois; south to South America, introduced in the Old Worl (Map 37). Representative specimens seen from the Unit ” States and Canada: Onrarto. Kent Co.: Chatham, Macoun 589 : ” ker ice Co.?: on ballast sand, Petty’s Island, Aug. 27, 1866, Par arves, Brunswi rper 1520 (G, NY, US). Franklin Co.: Apalachicola, Biltmore Herb. 3848a (G, NY, ve. ALABAMA. Hale Co.: Rosemary, Harper 3248 (G, NY, U®): Sees ee eee ae er cL SE Ah Eh ae ian 3 Palette ae Pep tale bee er span ped ee e : 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 201 Perry Co.: Uniontown, 1889, Newman (US). Mobile Co.: Mobile, July 14, 1891, Mohr (US). Musstssrpp1. Lowndes Co.: Artesia, Tracy 31 25 (NY). Iowa. Webster Co.: Somes 3841 (US). Woodbury Co.: Sioux City, Pammel 192 (NY). Missouri. Jackson Co.: Courtney, Bush 412 (NY, US). St. Clair Co.: 34% miles north of Ieconium on Osage River, Steyermark ge he ARKANSAS. Johnson Co.: Pine Bluff, Demare e 8778 US). Pulaski Co.: Little eis Coville 10 (US). Louisiana. ra Co.: Shreveport, Cocks 3625 (NY). Orleans Co.: New Orleans, J uly, 1837, Riddell & Carpenter (US). Felciane, Carpenter (US). Nortu DaKoTA. Morton Co.: Glen Ullin, Ber oe M). SoutH Dakota. Mellette Co.: White River, Fan 6131 (US). Cheyenne River bottom, July 16, 1896, Wallace CNY): eee Creek, Aug. 1891, Walliams (US). ‘Nepraska. Knox Niobrara, Clements 717 (G, NY, US). Lancaster Co.: ener Aug,, , Williamson (Ph). Kearney Co.: Minden, Sept. 18, 1929, Hane ig Franklin Co.: Franklin, Aug. 25, 1929, Hapeman (Ph). ansas. Ellis Co.: near Hays, Rydberg & Imler 1193 (NY). ecinin Co.: Osborne City, Shear 176 (G, NY ere sri Co.: Norton ste (G, NY, US). Saline Co.: ’ Miami rhe Aug. 10, 1885, Oyster (NY). Hamilton Co. : Syracuse, Rose & Fitch 17011 (NY, US). Kiowa Co.: Belvidere, Sept. 26, 1897, Ward (US). OKLAHOMA. Woods Co.: Alva, G. W. Stevens 1604 (G, NY, US). Kay Co.: Tonkawa, G. W. Stevens NY). Cleveland Co.: Norman, Emig 527 (US). Texas. Austin Co.: Industry, 1846, Lindheimer 300 Fasc. III (US). Bexar O.: near Bracken, Groth 15 (G, US). Brazoria Co.: Columbia, Bush 136 (NY). Cameron Co.: near Point Isabel, Tharp 1180 (US): Dallas Co.: Dallas, Stephenson 191 (US). Harris Co.: rere Thuron 4 (US). Howard Co.: Big Springs, Tracy 8127 (Mo, NY, US). Jackson Co.: Carancahua Pt., Tharp 1416 (US). Spesiee Co.: Amarillo, Ball 1275 (US). Lubbock Co.: Lubbock, Reed 3005 (US). Nueces Co.: Corpus Christi, eogget 1463 (NY, US). Red River Co.: Clarksville, feels 25, 1894, Plank (NY). Starr Co.: 5 miles north of Rio Grande City, Clover 1864 (NY). Tarrant Co.: Grapeland, Tharp 881 (US). Taylor Co.: Abilene, Tracy 7844 (Mo, NY, US). Travis Co.: near Austin, Armer 5890 (US). Valverde Co.: bank of te Rio Grande, Del Rio, June 13, 1891, Dew (US). Walker Co.: near Huntsville, Dizon 376 (NY). Waller Co.: Hempstead, 1872, Hall (US). Webb Co.: Laredo, Reverchon 3787 (G, US). Wharton Co.: Pierce, Tracy 743 5 (G, NY, US). Wilson Co.: Sutherland Springs, Aug. 22-30, 1879, Ed. Palmer 1191 (G, US). Montana. Cascade Co.: Great Falls, Williams 160 (US). CoLoRraDo. Weld Co.: New Windsor, Aug. 8, 1901, Osterhout 2345 (NY). 202 Rhodora (Mar New Mexico. Valencia Co.: near McArty’s Ranch, Rusby 378% (M, NY). Chaves Co.: 20 miles south Rosswell, 7. S. & B.S. Earle 308 (M, NY, US). Eddy Co.: Carlsbad, Tracy 8170 (NY, US). Dofia Ana Co.: west of Organ Mountains, Aug. 26, 1899, Wooton (US). Arizona. Santa Cruz Co.: Santa Cruz River at La Noria, Mearns 1192 (US). For citation of additional spetl- mens see Bull. Torr. Bot. Club 63: 448-9. 1936. 31. EvPHORBIA ALBoMARGINATA T. & G., Rep. Expl. & Surv. Miss. R. to Pacific Ocean 2 (4): 174. 1855. Type: Rio Pecos, Texas, Nov. 1850, Thurber 98 (NY!; photographs G!, W!; 1soTYPE G!). The common late-season phase with small leaves and short internodes. L. C. Wheeler, Bull. Torr. Bot. Club 63: 446-7. 1936.—E. stipulacea Engelm. ex Boiss. in DC. Prod. 15 (2): 30. 1862, in synonymy.—Anisophyllum albomarginatum (T. & G.) Klotzsch & Garcke, Abh. Akad. Berlin, Phys. 1859: 53. 1860.— Chamaesyce albomarginata (T. & G.) Small, Fl. Se. U.S., 710, 1333. Glabrous prostrate perennial from a taproot woody 10 age, stems several to numerous, up to 40 cm. long, internodes uP . 6 cm. long but usually much shorter, nodes often rooting; leal- blades mostly 3-6 mm. long, sometimes as much as 15 mm. long; orbicular to oblong, sometimes with a red spot in the middle on the upper surface, midrib usually not prominent, lateral we rarely evident, margin entire; petioles 0.5-1 mm. long; ‘ upper and lower stipules united into a glabrous, white, ne nous, deltoid to subulate, usually lacerate-margined s¢@ i ; cyathia solitary at the nodes; peduncles slender, glabrous, na 2-4 mm. long; involucres 1.5-2 mm. diam., open-campanula : to broadly obconical, glabrous without and within except 0D ee lobes; lobes short-hairy within below, exceeding the glan iy narrowly deltoid-attenuate; glands 0.5-1 mm. long, transversé y glabrous, entire or slightly crenate; fifth gland linear, equaling sinus U-shaped, little depress as bracteoles forming one conspicuous, upwardly expan ed, m 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 203 reticulate in definite lines, opaque-white or so thin as to little obscure the brown testa.—PLATE 661 Southern San Joaquin Valley, Inyo Co., south through southern California, east to Nevada, Utah, Arizona, New Mexico, Texas and Oklahoma, south to Lower California, east to Tamaulipas (Map 23). Representative specimens seen: CaLr- FORNIA. Tulare Co.: Porterville, Sept. 20, 1921, Kelly(CA). Kings- Tulare Co.: between Tulare and Tulare Lake, 1892, Ed. Palmer 2721 (US). Kern Co.: Bakersfield, Davy 1 883 (C); Randsburg, Mohave Desert, Apr. 14, 1905, Heller (NY, Ph, US). Ventura Co.: near Frazier Borax Min oe Mt. Pinos, ‘Abrams & MeGr regor 206 (D, G, NY, US). o Co.: Shepherd Canyon, Argus Mountains, Coville & iete' 785 (G, NY). Los Angeles Co.: Santa Monica Hills, G. B. Grant 497 (US); North Fork, San Gabriel, bales 9374 (US); San Bernardino Co.: Kessler Peak, Ivanpa h Mountains, Mohave Desert, Jepson 15829 (J); Holcomb Valley, San Bernardino Mountains, altitude 7,200 feet, Ewan 4872 (G); Redlands, Kuntze 23263 (NY); Hesperia, Mohave Desert, — 552 (G, N i; Riverside Co.: Paloverde, Boe 3205 ie Lo NY, O, US). fa mperial Co.: on ay banks, Experiment Farm, Bard, Apr. 5, 1912, Dewey (US); For t Yuma, Jepson 11734 16S), Nevapa. Nye Co. Rhyolite, Heller 9674 (G, US); gravel canyon-bottom, road to Beatty, east side Grapevine Mountains, Death Valley Region, Train 676 (US). Clark Co.: Moapa, Kennedy ee ey; By i Kiernan’s, Meadow Valley Wash, Goodding 654 (NY, US); Mormon Mesa, Eastwood & Howell 6310 sis Las Vegas, Goodding 2301 (G, NY). Uvan. Washington Co.: mesa northwest of Hurricane, Maguire & ae 1436 “D; near St. George, pl 27 (1). Arizona. Mohave plain near Oatman, April, 1916, Creighton (Ph); Peach Springs April, 1893, “Wilson (G). Coconino Co.: Turkey Tanks, M acDougal 141 (G, US); north from San > Peaks, Leiberg 573 (US). Navajo Co.: Winslow, 2 miles north in desert, Stone 373 (NY); Pineisle. Hough 126 (US). Apache Co.: be- tween Chambers and Navajo, Eastwood & Howell 6893 (G). Yavapai Co.: Prescott, Aug. 7, 1876, Ed. Palmer 509 (G, US). Maricopa Co.: Hassayampa, Newlon 684 (J). Pinal Co.: sandy river bottom, Gila River, 2 miles below Sones Dam, Maguire, Richards & Moeller 10421 (G, I). Gila Co.: Roosevelt Dam Goodding 723 (G, NY, US). Greenlee ay face of conglomerate cliff above San Francisco River banks, 8. Clifton, nee’ Richards & Moeller 11788 (G, I). Graham Co. me mile west Upper Reservoir Plat, 10 miles southwest Safford, M aguire 10007 (I). Pima Co.: Santa Cruz Valley near Tucson, May 2, 1881, Pringle (G, NY, US); Tucson, Nov. 30, 1894, Toumey (G, NY 204 Rhodora [Mar US). Santa Cruz Co.: vicinity of Ruby, Mrs. A. F. Morton 8 (US). Cochise Co.: Tanner’s Canyon, Huachuca Mountains, Holzner 1573 (US); Paradise, Blumer 1691 (G, NY, US); San Bernardino Ranch, Mex. Boundary Line, Mearns 639 (NY, US). 8S. & Sept., 1914, Beals (US). Grant Co.: Mangas Springs, 18 mules northwest of Silver City, Metcalfe 722 (G, NY, US); 5 miles east of Canyon, Tharp 4427 (US). Presidio Co.: Shafter, infrequent, valleys, Hanson 552 (NY, US). Brewster Co.: along Rio Grande near Lajitas, Chisos Mountain area, Warnock 665 (US). Crockett o.: Ozona, M. E. Jones 26001 (I, W). Dallas Co.: Dallas, Bush 650 (G, NY, US). Tom Green Co.: Knickerbocker Ranch, Bee. (NY, US). ; Sutherland Springs, Aug. 1879, Ed. Palmer 1207 (G). Zavala gua south of Velderrain, Pennell 19477 (US); Alamos, atroy®: Rose, Standley & Russell 12935 (US) aedabens, Rose, Standley & Russell 15074 (NY, US). Srnazoa: Culiacan, Rose, Stand & Russell 14934 (NY, US). Curavanva: Rio Santa ag Thurber 749 (G, NY); 6 miles west of Piloncillo, Johnston 786 man 406 hihue Apr. 8-27, 1908, Ed. Palmer 16 (G, NY, US); Llano de Chilicoté, 7 miles east of Chilocote Station, Johnston 7991 (G); Santo 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 205 Tomas on railroad northwest of San Isidro, Sierra Madre Occi- dental, Pennell 18991 (US). Durango: Dura ango and vicinity, June, 1896, Ed. Palmer 296 (C, F, G); Santiago Papasquiaro, ies 1896, Ed. Palmer 42 (C, F, G, "US 5S). Coanuita: Municipio 8&3 ; to Apr 1907, Ed. Palmer 92 (US). oo , citation of additional colleen see Bull. Torr. Bot. Club 63: 446-447. 1936. The Thurber specimen here taken as type is from Torrey’s herbarium and is labeled ‘‘Huphorbia albomarginata Torr. & Gray” presumably in Torrey’s hand. This specimen is accom- panied by diagnostic drawings. This choice of type preserves the customary interpretation of this well known species. The original diagnosis would not distinguish H. Fendleri nor E. poly- carpa from E. albomarginata. However, in spite of that and the fact that some specimens of E. Fendleri were labeled as E. al- bomarginata and the description of the seeds was probably drawn from them, the majority of specimens originally referred to E. albomarginata were that, and E. Fendleri was described on the next page. One specimen of E. polycarpa was included in the concept and part of the original diagnosis may have been based upon it. Nevertheless, about three-fourths of the specimens originally referred to E. albomarginata were specifically identical with the specimen here chosen as type. Jepson’s statement, Fl. Calif. 2: 428. 1936: ‘‘type loc. ‘head- waters of the Colorado,’ that is, western Texas, Diffenderfer;” is based on the statement in the publication of EZ. albomarginata that it grew “with the preceding” (EZ. Wrightit T. & G.) which is reported from “ Headwaters of the Colorado”. This, combined with the statement on page 3 of the same volume that all the collections of plants of the expedition were made by Dr. W. L. Diffenderfer leads logically to Jepson’s bibliographical conclusion. My selection is based on actual examination of the specimens from which the original diagnosis was drawn. 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 223 EUPHORBIA SUBGENUS CHAMAESYCE IN CANADA AND THE UNITED STATES EXCLUSIVE OF SOUTHERN FLORIDA Louis CuTreR WHEELER (Continued from page 205) 32. EupHoRBIA FEenpDuERI T. & G., Rep. Expl. & Surv. Miss. R. to Pacific Ocean 2 (4): 175. 1855. Glabrous perennial from a taproot woody in age; ue several to numerous from the ae penraeit ae to rect, long, slender, internodes long; lea hbla des entire, 3-11 mm. long, Saarinen aie "Obli ique base, to lanceolate; petioles mostly ca. 1 mm. long; stipules up to 1 mm. long, distinct, narrowly linear, reba entire, rarely with a few hairs; pedunc les up to 4 mm. long, glabrous; cyathia eet at the nodes; involucres glabrous, 1.25-1.75 mm. diam., campanulate to turbinate, glabrous without, with short hairs > thin above oppo- site the gla nds; lobes deltoid, shortly-attenuate, hairy within pins equaling ‘the glands; glands Riser 1.5—-4 times as long as wide, concave or convex, up to 1 m. long; appendages white, glabrous, 0.5-1 mm. long, as side as t as gland at least at the base, obtuse-crenate to entire and narrowly deltoid, or parted into 2-4 narrow segments; fifth gland shorter than the lobes, linear; ‘ ec ee tially, ovateactt radially, angles prominent, front facets con- cave, mostly smooth, back facets slightly wrinkled, coat white, midroreseabete ee TO Vanes var. triligulata. - Fenpieri T. & G., Rep. Expl. & Surv. Miss. R. to Pact Ocean 2 (4): 175. 1855, oats vypica L. C. Wheeler, Bull. Torr. Bot. Club 63: 444. 1936. Typ: Santa Fe, New Mexico May 4—July 18, 1847, Fendler 800 (NY!; isotype G}, atypical). 224 Rhodora [JUNE See Plantae Exsiccatae Grayanae 726, cited below, for bibli- ography.—Anisophyllum Fendleri (T. & G.) Klotzsch & Garcke, Abh. Akad. Berlin, Phys. 1859: 26. 1860.—Chamaesyce Fendleri (T. & G.) Small, Fl. Se. U. S., 710, 1333. 1903. E. rupicola Scheele, Linnaea 22: 153. 1849, not Boiss., Elench., 81. 1838. Typr: “siidlich von Neubraunfels: Lindheimer. Juli.” (2). (Lindheimer 290, (M!) New Braunfels, Texas, July, 1846, designated as type collection by Blankinship, Ann. Rep. Mo. Bot. Gard. 18: 149. 1907). E. Fendlert T. & G. var. dissimilis Payson, Bot. Gaz. 60: 379. 1915. Type: dry sandy hills, Naturita, Montrose County, M)). Some of the plants unusually large-leaved and some of the ap- pendages approaching those of var. chaetoca yx. Chamaesyce Gooddingii Millsp., Field Mus. Pub. Bot. 2: 405. 1916. Typr: Lee Canyon, Spring (Charleston) Mountains, Clark County, Nevada, altitude 8,000 feet, Aug. 4, 1913, A. e PRN” Panes (F 411087!; photographs G!, W!; isorypzs C!, Gi, Pp 5 Chamaesyce Greenei (Millsp.) Rydberg sensu Rydberg, Fl. Rocky Mts., 544. 1917, and Fl. Pr. Plains, 517, fig. 350. 1932. Inyo and San Bernardino Counties, California, east to wester! Nebraska and Oklahoma, south to Sonora (2), and Texas (Mar 29). Representative specimens seen: Kansas. Ellis Co.: stony hills, A. S. Hitchcock 475 (G). Oxtanoma. Ellis Co.: neal Shattuck, Clifton 3183 (G). Woodward Co.: 8 miles southwest of Woodward, Goodman 2179 (G). Roger Mills Co.: Antelope Hills, Goodman 2612 (G). Harper Co.: near Buffalo, @._ Stevens 308 (G). Beaver Co.: 15 miles southwest of Beaver City, G. W. Stevens 350 (G). Texas. Bexar Co.: 8 miles south 0 Bulverde, Cory 6020 (G). Comal Co.: New Braunfels, Lind- young 104 tG). Wilson Co. : Sutherland Springs, 1879, Ed. Palmer of Alpine, Cory 18595 (G). Presidio Ca 154% miles southwest ? F , ow 2 of rea Cory 26319 (G). Kendall Co.: Spanish eo Bates (G). Cotorapo. Larimer Co.: Cowen 126 (G). Weld Co.' Feeney 82 (G). Mesa Co.: Grand Junction, M acbride on - Ura. Uintah Co.: 8 miles south of Ouray, Uinta Basil, . -* ys Rollins 1699 (G). Grand Co.: near Wilson Mesa, Rydberg & 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 225 Garrett 8382 (G). Emery Co.: Muddy River, 1877, Ed. Palmer 448 (G). Nevapa. Clark Co.: Las Vegas, Goodding 2282 (G). ARIZONA. Coconino Co.: Cosnino, M. E. Jones 4036 (I, 2 (}); desert near Tiba: Clute 91 (G): Williams, July 12- 1886, Bunker Pak near Flagstaff, MacDougal 209 (G, US). Seno Co.: near Montezuma Castle National Monume nt, ae Nation 2048 (NY). Navajo Co.: Kayenta, Clute 9 (G); Marsh Pass, Eastwood & t Hebel te 6616 (G). Cochise Co.: 5 miles northeast of Dragoon, Maguire 11191 (G, I). For citation . additional specimens and discussion of in tergrades see Bull. Torr. Bot. Clu 63: 445. 1936; Bull. ao Calif. Acad. 33: 106. 1934; Bull. So. Calif. Acad. Sci. 35: 128. Certain plants of ae species exhibit a rare tendency. They have occasionally three, four, or even five leaves at a node. The collection showing this tendency most strongly was one infected with rust: below Wheeler Well, rocky hillside, Juniper belt, Clark County, Nevada, May 31, 1936, I. W. Clokey 7183 (Cl, G). These plants showed as many as five leaves per whorl. Another collection made at the same place and time which was not infected with rust had only opposite leaves: Clokey 7184 (Cl). This suggested that the abnormality might have been induced by the pathogen. However another perfectly healthy collection from the same general region showed some whorled leaves: gravelly slope with Juniperus and Cercocarpus, Harris Springs Road, Spring (Charleston) Mountains, Clark Co., Nevada, July 16, 1937, Clokey 7578, Pl. Exs. Gray. 726 (G). MacDougal 209 (US) from Arizona also shows three leaves in a whorl in one rusty plant. That this variation is not confined to this region is shown by a collection which bears not only one whorl of four leaves but also one unpaired leaf. The collection bears only the data “Rio Grande, Wright 1848” (G). Previously, Bull. Torr. Bot. Club 63: 444. 1936, I made the error of combining the locality for Pope’s collection from Big Springs of the Colorado River, Texas, with New Mexico, the se ae which Fendler 800 came. 32b. —- rt T. & G. var. CHAETOCALYX Boiss. in DC. Prod. 15 (2):3 9. 1862. Typr: jesaitake sides near Frontera, El Paso County, Texas, Apr. 3, 1852, C. Wright 1847 (Ge!; photo- graphs G!, W!; 1sorype G!). Not the extreme of the variety but rather closer to typica but nevertheless acceptable in var. chaeto- calyx.—Chamaesyce chaetocalyx (Boiss.) Wooton & Standley, Contr. U. 8. Nat. Herb. 16: 144. 1916.—E. chaetocalyx (Boiss.) Tidestrom, Proc. Biol. Soc. Wash. 48: 40. 1935. Prats 666D. 226 Rhodora [JUNE Fs if ~ . 37 dots } A 3) AC A; Me LATA ’ N AP 34, dots range of Et PHORBIA SUPIN Hess , k - ACUT! ’ 36, kk , LIA . the i yYSSOPIFO 4. SERPENS in Canada & U. S., circle E. GoLONDRINA; ae ee aes "tations in U.S.; 39, E. AMMANNIOIDES in U. 8.; 40, E. wumisrrara (Virg omitted). 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 227 Arizona, New Mexico, and western Texas (Map 31). Repre- sentative specimens seen: Texas. Culberson Co.: Cory 1964 (G). Crockett Co.: 11.9 miles southwest of Leon Powell, on 3191 (G). Brewster Co.: Chisos Mountains, Mueller 8 sor? (M). For citation of additional specimens see Bull. Torr. Bot. Club 63: 445. 1936. Frontera seems to be missing on the modern maps. It is lo- cated on the Rio Grande a few miles above El Paso. It is shown on the general map, which is first, in the part entitled ‘‘Topo- graphical Maps, Profiles, and Shedehen't in Rep. Expl. & Surv. Miss. R. to Pacific Ocean. 11:—1859 32c. E. Fenpuerr T. & G. var. TRILIGULATA L. C. Wheeler, Bull. Torr. Eos Club 63: 445. 1936. Typrn: cliffs above Boquillas Canyon, Boquillas, Brewster County, pew: 2 12, 1931, Moore & Stsermaee 38444 (G!; isorypEs CA}, h!). "A very distinctive variant teased only from the aol Bebe (Map 29). Puate 666 EUPHORBIA ASTYLA Engelm. ex Boiss. in DC. Prod. 15 2): 40. 1862. Typ PE: ‘‘Valley of the Nazas River” [Coahuila Durango], Mexico, Apr. 15, 1847, A Gregg 457 (M 149805 !: fragment F!; phot tographs Gl, W!: robable isotype G!). small but others typical plant. —Chamaesyce pe (Engelm.) Millsp., Field Mus. Pub. Bot. 2: 408. 1 abrous, perennial from a root up re 6 mm. thick; stems several to numerous, 4-23 em. long, prostrate to erect, 0.6-2 mm. thick, internodes up to 2 em. long but t predominantly 3-6 mm. long; leaves sessile, 2-8 mm. Med orbicular-reniform to ovdiandnte: base oblique, often clasping, margin entire: stipules united to form a scale ca. 0.5 mm. wide and mostly only ca. 0.2 mm. high, minutely lacerate and bearing a fringe of very short hairs on the inner margin; cyathia solitary at the nodes, mostly toward or at the stem-tips where often much congested by the very short internodes; peduncles 0.5-1.4 mm. long, stout; involucres glabrous outside, densely short-hairy inside espe- cially at summit and somewhat downward beneath the glands, broadly campanulate to obconical-campanulate, 1.4-1.6 mm. in diam.; lobes narrowly triangular, equaling the glands; glands transversely oval to oblong, generally slightly depressed in the middle, 0.4-0.6 mm. long; appendages white, glabrous, 144-14 times as wide as gland, entire or crenulate; 5th gland absent or present as a mere papilla; sinus U-shaped, not depressed, densely hairy; bracteoles numerous, glabrous below, pubescent above, ca. 3 as long as the androphores, some united below and often ad- nate to the involucre beneath the glands; staminate flowers 3-6 per fascicle, 22-26 per cyathium; androphores glabrous, 1.4-1. 228 Rhodora [JUNE mm. long; gynophore glabrous, soon exserted and_ usually reflexed; ovary glabrous, 3-angled; styles 0.3-0.4 mm. long, entire, sometimes medianly creased, stigmas thickly round- capitate; capsule glabrous, subacutely 3-lobed, slightly wider below the equator, 1.9-2.6 mm. long, one carpel often sterile and shriveled; seeds sharply quadrangular, 1.5-1.7 mm. long, 0.9-1 mm. tangentially, 0.8-1 mm. radially, ovate radially, ventral facets plane to slightly convex, all traversed by irregular low rounded frequently anastamosing ridges separated by narrow grooves, angles often notched, gelatinous coat sordid to chalky white, very thick, microreticulate.—PLaTE 666A. of Hermanas, prostrate on heavy alkaline soil, Johnston 7060 (G); 4 miles west of Cuatro Cienagas, middle of saline flats 1m very alkaline soil (with Dondia, Allenrolfea, Atriplex, etc.) prostrate, succulent, brittle, Johnston 7135 (G); Sierra del Rey, Purpus 4612 (F, G, M, US). The collector’s number on the type bears some illegible suffix and the notation that it was mixed with 457a. 34. Eupnorsia Hoovert L. C. Wheeler, Proc. Biol. Soc. Wash. 53: 9. 1940. Typr: Yettem, Tulare County, California, styles 1.7-2 mm. long, entire; capsule spheroid with flattened base, roundly 3-lobed, glabrous, 1.6-1.9 mm. long; seeds ovoid- quadrangular, rotund-ovate radially, raphe slightly curved, 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 229 back eee facets covered with low irregular, smooth ridges, 1.4-1.6 mm. long, 1-1.1 mm. tangentially and radially, coat white and microreticulate.—PLATE 664B. Central Valley of California (Map 32). Other collections seen: CALIFORNIA. Tulare Co.: Yettem, R. F. Hoover 1277 (G) Ore Co.: 4 miles southeast of Vina, R. F. Hoover 8722 (G). 35. EUPHORBIA SERPYLLIFOLIA Pers., Syn. Pl. 2: 14. 1806 (as serprllefolicn. raps stems glabrous to aby gaint ease pe a to erect, ies ete leaf-blades 3-14 mm. long, ovate, ve oblong- serrulate at least toward the apex; degules distinct, ee entire or few-parted, wget yk glabrous; cyathia solitary; involucres gla- brous to villous, 0.8-1.2 mm. in diam., obconical-campanulate to cadena? glands transversely oblong, 0.2-0.5 mm. long; appendages narrow, white, glabrous, entire to crenulate or even subdentate; sinus U-shaped, saaahtlsy depressed ; staminate flowers 5-18 per cyathium; androphores glabrous, 1-1.2 mm. long; Abe ge glabrous, soon exserted and reflexed; capsule sharply 3-angled, 1.5-1.9 mm. long, wider below the equator, mostly gla- brous; seeds cael Se but often turgid, angles rounded to sharp, 1—-1.4 mm. long, broadly ovate to narrowly ovate radially, facets oe Be ® sight punctate or even rugulose, coat clay white to bro Key To VARIETIES Glabrous throughout; leaves 3-14 mm. long............... a. var. genui More or less villous; leaves 3-7 mm. Me is OU ee b. var. hirtula. 35a. E. SERPYLLIFOLIA Pers., Syn. Pl. 2: 14. 1806, var. GENU- ina Boiss. in DC. Prod. 15 (2): 43. rere Thellung in Ascherson & Graebner, Syn. Mitteleur. Fl. 7: 1917. Typs: ‘‘ Hab. in Amer. calidiore” (Ge?, not seen; Revatiay M!; photographs of fragment G!, W!). An average member of the species, stems slightly winged. Millspaugh, Pittonia 2: 82, Pl. 1, figs. 1-4. 1890, poor. oe sereigticioln (Pe) Small, Fl. Se. U.S., 712, 13338. 1903. E. inaequilatera eres sensu Engelm. in Emory, U. 8. & Mex. Bound. Surv. 2 (1): 187. 1859. Anisophyllum navomezicanim arog & Garcke, Abh. Akad. — Phys. 1859: 1860. Typr: 25th May-8rd Octbr. . Rio del Marts, sandy soil, low river bank between ayer rents Santa Fe and Poni Creek’’.! New Mexico, Fendler 7 (BI; photeerinte G! W!). ok very slightly winged; Sled slender, slightly punctately pitted. Alters is not the extreme represented by E. albicaulis.\—E. novomexicana (Klotzsch & Garcke) L. C. Wheeler, Bull. So. Calif, yer Sci. 35: 129. 1936. 1 From Fendler’s field notes (or copy?) at Gray Herbarium. 230 Rhodora [JUNE E. serpyllifolia Pers. 8 consanguinea Boiss. in DC. Prod. 15 (2): 43. 1862. Type: Nebraska, 1853-4, F. V. Hayden (Ge!; photo- graphs G!, W!). Stems slightly winged; seeds slender, punctately pitted; leaves mostly broad. While the seeds and capsules ap- proach those of E. albicaulis the leaves are broader. Jepson, Fl. W. Middle Calif., 262. 1901.—E. consanguinea, E. subserrata, and £. notata Engelm. ex Boiss., |. ¢. in synonymy.—Chamaesyce consanguinea (Engelm.) Lunell, Amer. Midl. Nat. 1: 205. 1910, with parenthetical author but without basinym; description given. Millsp., Field Mus. Pub. Bot. 2: 408. 1916, basinym given as ““Huphorbia consanguinea Engelm. Mex. Bound. 1877). that place no such name appeared. The name was first published by Boissier in synonymy; see above. E. neomexicana Greene, Bull. Calif. Acad. Sci. 2: 56. 1886. Typu: plains toward the Gila, New Mexico, Aug., 1880, £. L. Greene (Herb. Greeneanum 13713, not seen; photograph G!). Narrow-leaved, other characters not certain from picture.—4. serpyllifolia var. neomexicana (Greene) Millsp., Pittonia 2: 84, - i, figs. 16-18. 1890.—Chamaesyce neomexicana (Greene) Lunell, Amer. Midl. Nat. 1: 205. 1910 (June), with parenthetical author but without basinym; description given. Standley, Contr. U.S. Nat. Herb. 13: 199, 227. 1910 (Oct.), with basinym. E. he Pea Hochst. & Steud. sensu Greene, Bull. Calif. Acad. E. serpyllifolia var. rugulosa Engelm. ex Millsp., Pittonia 2: 85, Pl. 1, fig. 19. 1890. Typr: San Bernardino Valley, Aug., 1881, S. B. & W. F. Parish 672 (F 99060!; photographs G!, W!; 180- Nebraska, July 13, 1891, P. A. Rydber F raphs , 1891, P. A. 9 356 (NY!; photog G!, W!). Stems terete; leaves very narrow; seeds long and Cab ee ee albicaulis (Rydb.) Rydb., Bull. Torr. Bot- 6. Club 33: 145. 1 TYPE: on a railroad bank, Leeds, Benson County, North Dakota, (Mi 251952!; photographs G!, W!)- on a railroad bank, Leeds, Benson ounty, N July 27, 1909, J. Lunell 695 (Mi 251953!; photographs G!, W)): 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 231 Chamaesyce erecta Lunell, op. cit., 206. Typr: on the margin of a slough, Leeds, Benson Cou unty, North Dakota, Aug. 19, 1909, J. Lunell 696 (Mi 251955!; photographs G! W!). Erect, as is often the case with plants of this species in wet situations. ag aequata var. erecta (Lunell) Lunell, op. cit., 4: 471. 1916.—PLatE 657C. Southern British Columbia and Alberta, south to Mexico, east to western Texas, Colorado, Nebraska, Iowa, Minnesota, and northwestern Michigan (Map 15). Said to be introduced into Europe. Representative specimens seen: ALBERTA: Rosebu Creek Valley, Moodie 1199 in part (NY); Milk River, Macoun 13702" (QO). British Cotumpia: Vancouver Island, Sproat Lake, Aug. 12, 1887, Macoun (NY). MucwiGgan. Keweenaw Co.: July, 1888, O. A. Farwell 235 (NY); Phoenix, June, 1886, Farwell (Ph) ; Keweenaw Point, July 15, 1885, Farwell (NY). “Sianaire ap Brown Co.: Sleepy Eye, Sheldon 917 (Mi). Clay Co.: south of Moorhead, Ballard 2902 (Mi). Kittson Co.: North- cote, "Ballard 2682 (Mi): Kennedy, Ballard 2734 (Mi). Lincoln : Lake Benton, Sheldon 1285 (Mi). Marshall Co.: Warren, Ballard 2774 (Mi). Pipestone Co.: Pipestone, Aug. 11 | 1925, Peterson (Mi). St. Louis Co.: eran Pee 21 18 (Mi). "Iowa. onona Co.: Turin, Pammel (F). Mussourt. Jackson Co.: Courtney, Bush 11824 (M of Nortu Dakota. Benson Co.: Leeds, Lunell 639 (Mth. Cass Co.: Fargo, July, 1890, Greene (NY). Sours Daxora. Brookings Co.: near Brookings, June 26, 1894, T. A. Williams (W). Stanley Co.: Fort Pierre, June 21, 1853, Hayden (NY). Washington Co.: Sheep Mt., Visher 2359 (NY). NEBRASKA. Cheyenne Co.: south of Pater Rydberg 356 (NY). Oxianoma. Cimarron Co.: 11 miles ot of Boise City, Stratton 448 (M). Texas. sche Co.: Alpine, MacKensen 38 (F). El Paso Co.: El Paso, M. E. Jones 1342 ‘d ). Culberson Co.: Van Horn flats, hig: 1900, Eggert (M). Monrana. Lake Co.: Aug., 1924, Swingle (I). Wyomine. Platte Co.: Uva, Neon 2743 (N Y). Sweetwater Co.: Granger, Nelson 8133 (NY). Conorapo. Boulder Co.: near Boulder, Tweedy 5057 (NY). Conejos Co.: Antomito, July 19, 1898, Earle (NY). Denver Co.: Denver, along the Platte River, M. E. Jones 864 (I). Douglas Co.: Buffalo Creek, Goodman 2004 (NY). El Paso Co.: Manitou, July 27, 1888, Northrop (NY). Fremont Co.: Canyon City, 7.8. Brandegee 316 (Ph). Gunnison Co.: Parlin, B. H. Smith 81 (NY). Jefferson Co.: Golden nes July, 1879, Gaillardon (F). La Plata Co.: Durango, Baker, Earle & Trac 608 (NY). Larimer Co.: Ft. Collins, July 10, 1896, Baker (NY). Otero Co.: 15 miles northeast of La Junta, Rollins 1 874 (M, NY, W). Ouray Co.: Box Cafion west of erin Underwood & Selby Ana Co.: phar Mts., Wooton 425 (NY). Grant Co.: Fort 232 Rhodora [JUNE Bayard, Blumer 27 (NY). Lincoln Co.: near Gray, Skehan 71 (NY); White Mts., Wooton 608 (NY). Otera Co.: Clouderoft, Schulz 301 (NY). Sandoval Co.: 4 miles north of Bernalillo, Arséne 18987 (F, US). San Miguel Co.: near Pecos, Standley 505 Y). Santa Fe Co.: Santa Fe Canyon, 9 miles east of Santa Fe, Heller 3845 (NY, O). Sierra Co.: 2 miles south of Sy ROE rantiod 1298 (NY). Ipano. Ada Co.: Boise, Clark 144 : (I). Shoshone Co.: Coeur d’Alene Mts., Leiberg 1317 (NY). Fort Halleck, Heller 9275 (NY). Ormsby Co.: Eagle Valley, Baker 1274 Washoe Co.: Spanish Springs Valley, Ken- nedy 1945 (F, NY). Arizona. Mohave Co.: Peach Spring, Lemmon 89 ( Coconino Co.: Walker Lake, San Francisco Carrizo Mts., Navajo Indian Reservation, Standley 7499 (US): (G, , US). Gila Co.: Tonto Basin, Toumey 266 (US). Cochise Co.: near Fort ). Thompson 11122 (NY). Okanogan Co.: Loomiston, Elmer 602 (NY). Whitman Co.: Pullman, Elmer 200 (NY). Orecon. Curry Co: Rogue River bar at ferry, Peck 13656 (Ph). Klamath Co.: Fort Klamath, Peck 9562 (NY). Multnomah Co.: University Park, Wasco Co.: The Dalles, Thompson 11891 (NY, W). Washington Co.: Forest Grove, July 2, 1894, Lloyd (NY). CALIFORNIA. Humboldt Co.: Trinity River Valley near the South Fork, Tracy 6555 (J). Mendocino Co.: Covelo Ranger Station, Crone- miller 663 (J). Lake Co.: near Clear Lake, J. Torrey 478 (NY)- Sonoma Co.: Santa Rosa Creek east of Santa Rosa, Heller 5802 on Y, Ph). Napa Co.: Napa River near St. Helena, Jepson 5 934 (J ). Solano Co.: northeast of Elmira, Jepson 12379 (J). ce Francisco Co.: San Francisco, Bolander 186 (NY). Sa? 2 ateo Co.: Crystal Lake, Abrams 2990 (NY). Santa Clara Co.: tanford University, Baker 195 (NY). Monterey Co.: Carmel, 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 233 1905, Clemens (NY). Siskiyou Co.: Mt. Shasta, Brown 541 (NY). Shas ta Co.: near Ash Creek, Shasta Forest, Dayton 486 (NY). Butte Co.: 5 miles east of Chico, Heller 1381 5 (F, NY). Colusa Co.: Indian Creek about 7 miles north of Leesville, Bracelin 564 (I, NY). were Co.: Sacramento, M. E. Jones 28565 (1). San Joaquin Co.: Lathrop, Walker 910 (J). Alameda Co.: Berkeley, Michener & Bioletti 6189 (M, NY). Modoc Co.: Egg Lake, July 25, 1893, irae (J, NY). Lassen Co.: Beckwith Pass, Jepson 7762 (J ); Plumas Co.: Bucks Valley, sip 10643 (J). Nevada Co.: lower aay of Donner Lake, Heller 6935 oes Placer Co.: Yuba River below Cisco, Heller 13303 (NY, Ph). El Dorado Co.: near Echo Camp on Lincoln Highway, Heller pee (NY, Ph, W). Amador Co.: Ione, Braunton 1178 (NY). Calaveras Co.: ‘Dorrington, Jepson 101 12 (J). Tuolumne Co.: Deadman Creek, Jepson 6559 (J). Mariposa Co.: Yosemite Valley, Abrams 4652 (NY). Madera Co.: Fresno Big Trees, Jepson 15985 (J). Fresno Co.: Pine Ridge, Hall & Chandler r $48 (NY). Los Angeles Co.: Los Angeles, rabbis yt te oe NY); Orange Co.: New wport, May, 1908, King 43). San Bernardino Co.: Cucamonga, Abrams 2661 (M NY, Ph); Mill Creek, San Bernardino Mts., Munz 7577 (NY); 5 miles south of Barnwell, eastern Mohave Desert, Munz 13836 (W). San Diego Co.: DieEo, Spencer 964 (NY). MEXICO: Basa CaLirorNiA: Srey of La Encantada, Sierra San Pedro Martir, Wiggins & Toon 5020 (US). CxHrauanva: plains near Chihuahua, Pringle 286 (Ph, aie CoauviLa: Jimuleo, Pringle 81 (US). Durango: Durang , Apr—Nov., 1896, Ed. Palmer 899 (US). pa ei near Aén Juan del Rio, Rose, Painter & Rose 9611 Bush’s collection of this species from Missouri is not neces- sarily to be taken as evidence that the species grows there naturally. From this and several other collections of Euphorbia made by Bush at Courtney, his home town, I have been led to wonder if he were not careless with the sweepings from his her- barium and consequently found many waifs in his backyard. Euphorbia minuta Philippi appears doubtfully distinct from E. serpyllifolia. It is quite possible that the plants here called EZ. serpyllifolia var. genuina could be divided into three varieties: (1) The typical with broad leaves and seeds ; (2) slender-seeded, narrow-leaved, ranging from Nebraska to New Mexico and perhaps westward in arid regions; (3) elliptic-leaved and with numerous staminate 234 Rhodora [JUNE flowers and subdentate appendages, in the Central Valley of California. 35b. E. sERPYLLIFOLIA Pers. var. HIRTULA (Engelm.) L. ©. Wheeler, Proc. Biol. Soc. Wash. 53: 11. 1940; based on Z. Talley’s, Cuyamaca Mountains, San Diego County, Cale wild, San Jacinto Mountains, June 16, 1921, Spencer (0); Idyllwild, San Jacinto Mountains, June 26, 1920, Spencer (0); San Jacinto Mountains, Spencer 970 (G). San Diego Co.: Pine Spencer (O); Julian, Cleveland 904 (G); San Diego, 1874, Cleveland (G, M); near Julian, Abrams 3? (G). MEXICO: Lowzr CALIFoRNIA, San Vincente Rancho, July 8, 1896, T. S. Brandegee (J). _ The designation of the specimen at Gray Herbarium as tyP® Seems necessary in view of the following evidence: Two collections are cited, “Near San Diego (Cleveland); at Talley’s in the Cuy® maca Mountains, Palmer.” Engelmann, on the sheet im ! herbarium, questioned the reference of the Cleveland collectio® to E. hirtula. That means that Engelmann considered the name associated permanently with the Palmer collection. No sheet has been found at Missouri Botanical Garden which can app! priately be taken as type. The isotype at M was acquired Engelmann’s death by purchase of a private herbarium. Enge" mann’s Latin diagnosis was sent to Watson and now repose : the folder of E. hirtula, Watson translated the description in? English and published it for Engelmann. 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 235 Jepson, Fl. Calif. 2: 426. 1936, took the Cleveland collection as type but that was done without examining the actual speci- mens. Jepson has confused certain specimens of EF. supina (as E. maculata) with this species, e. g., Jepson 10,042 At US there is a collection by Ed. Palmer in 1869 bearing only the locality “Sonora”. The label is copied. In view of the curiously restricted range of this species it is fairly certain that this Beau is erroneous. the Platte River, eee County, Relea. July, 1856, H. Engelmann (M heel photographs G!, W!; ISOTYPE? G). ee RHopora 39: 496. 1937 for discussion of this choice.) An oh member of the entity. Boiss., Icon. Euph., t. 18. 1866, —Chamaesyce glyptosperma (Engelm.) Small. Fl. Se. U. 12, 1333. 1 1903. . glyptosperma Engelm. var. tenerrima Engelm. in Emory, U. . x. Bound. Surv. 2 (1): 187. 1859. Typ: pebbly ee! of the Nueces River, Texas, May 25, 1851, C. Wright 1853 (M 144673!; photographs G!, W!; isoTyPEs Gh, NY!). A plant with small leaves, a character whi ch shows no geographical correlation, SA intergrades and is probably more ecological than ae g eM es Engelm. var. pubescens Boiss. in DC. Prod. 15 (2): 4 1862. Typx: bord du Mississippi, eran Richt "178 (Ge, not ee fragment F!). A mixture of typical LZ. See : so E. Greenii Millsp., Pittonia 2: 88. 1890. Type: Beaver Can- yon, Idaho, July 30, 1889, FE. L. Greene (F 196395!, photographs G!, W)). Ent tirely typical and later abandoned by its author.— ine Greenei (Millsp.) Rydb., Fl. Rocky Mts., 344, 1063. Chamaesyce ed er ph Engelm. var. integrata eee Amer. Midl. Nat. 3: 142. 1913. Tyrer: Leeds, Benson County, North Dakota, Aug. 20, 1906, J. Lunell 693 (Mi 251964!, 2 ues Gl, W W)). Entire leaves are by no means unusual in this species and the character is too variable warrant recognition. _ Annual, glabrous; stems mostly prostrate, sometimes ascend- ing or erect, 5-33 em. me 2 fe 5 mm. thick, internodes up to ovate-oblong in the lower, 3-15 mm. long, base sig tech inequi- lateral, margin often serrulate especially at the apex and on the 236 Rhodora [JUNE . 0.3 mm. long; ca. 14 bifid; capsule glabrous, sharply 3-angled, wider below the equator, 1.4-1.7 mm. long; seeds sharply quadrangular, 1-1.3 mm. long, 0.7-0.9 mm. tangentially and radially, rae traversed by 3 to 4 rounded transverse ridges passing more 0! less through the angles, or sometimes the ridges very faint, coat specimens seen: New Brunswick: Carleton Co.: yard of the Canadian Pacific railroad, Hartland, Fernald 1981 (G). Qu=BEC: Pontiac Co.: Chichester, Vallée de VOttawa, Marie-Victori", fi Ney, Oxford Co.: Rumford, Aug. 1, 1890, Parlin (NE). NEW MPSHIRE. Cods Co.: Randolph, Pease 18138 (NE); Carroll @ 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 237 Carroll Station, poy 16591 (NE). Vermont. Franklin Co.: Swanton, Blake 3 8 (NE). Chittenden Co.: Burlington, Aug. 7, 1921, Stessanion *G). Windsor hare Mt. Tom, Woodstock, northeast of Tripoli, Sept. 19, 1915, See (G, NY). Aas Co.: Albany, House 25024 (NY). Suffolk Co.: Fishers Island, Aug. 11, 1895, Graves (NY). MrcutGan. Schooleraft Co.: near Manistique, Dodge 201 (NY). Emmet Co.: Pellston, Deam 28949A (Deam). Dickinson Co.: south of Iron Mountain, Fernald & Pease 3407 (G). Kalamazoo Co.: 6 miles west of Schooleraft, Hanes 1886 (G). INDIANA. Steuben Co.: Bass Lake, 5 miles west of Angola, Deam 55400 (Deam, G). Elkhart Co.: 3 miles northeast of Bristol, Deam 57370 (Deam, G). Newton Co.: Beaver Lake bed, 3 miles south of Lake ie Deam 56707 eam). Wisconsin. Brown Co.: Green Bay, Aug. 4, a Schuette (NY). Waupaca Co.: Lake Mendota, Mode. Aug. 2 1893, Sudworth (NY). Trempealeau Co.: Brady Peak, 1 at west of Pan ig oy Hermann 8944 (NY). La Crosse Co.: La Crosse, Pammel 620 (G). Dane Co.: Madison, Aug. 13, 1889, Trelease (M). Polk Co.: Osceola, Aug. 8, 1900, Baker (G). Pierce Co.: Prescott, Fassett 10258 (G). Inurots. Cook Co.: on ballast, West Chicago, Umbach 2318 (G). Lee Co.: Dixon, Vasey (G). Peoria Co.: Peoria, July, 1891, McDonald (G). St. Clair C Co.: St. Louis, Sept., 1886, Eggert (NY). MINNESOTA. Anoka Co.: Cooper 111 (Mi). Becker Co.: Detroit Lakes, O. A. Stevens 290 (G). = Co.: Sleepy Eye, Sheldon 5966 (Mi). Carver Co.: Chas ska, Ballard 161 (Mi). Chisago Co.: Center City, Taylor 1568 (Mi). Clay Co.: Muskoda, Ballard 8051 (Mii). hippewa Co.: Montevideo, es 28, 1893, Moyer (Mi). Clear- water Co.: near Squaw aE Moyle 850 (Deam, NY). ~~ w Co.: Brainerd, Mell 249 (M, NY). Houston Co.: Jefferson, W. Wheeler r 434 (Mi). Hennepin Co.: Minneapolis, Sheldon 1658 (Mi). Kandiyohi Co.: Spicer, Frost 347 (Mi). Kittson Co.: Humboldt, Ballard 2604 (Mi). Lincoln Co.: Lake Benton, oo 1557 ee Morrison Co.: Sp agied 888 (Mi). Otter tail July, 1886, Helene (Mi). Iowa. Boone Co.: pain Pawnal & Ball 191 (G, NY). Black Hawk Co.: Burk 904 (M). Green Co.: Rippey, July 31, 1867, Allen (G). Woodbury Co.: Sioux City, Pammel 193 (G, NY). Humboldt Co.: Dakota City, Pammel 190 AG , NY). Missovurt. Jackson Co.: Courtney, Bush 500 (NY). Nortu Daxora. wig ore ~ bar July 28, 1912, Lunell (NY). Towner Co.: Maz 1900, Ki ‘Idahl ( i). Ramsey Co.: Devil’s Lake, peer oan YMA). Oliver Gos: 238 Rhodora [JUNE Ft. Clark, upper river valley on old Indian Lodge site, 0. A. Stevens 384 (F, G). Billings Co.: Medora, Aug. 12, 1896, Brannon Broken Bow, Webber 7 (NY). Deuel Co.: sand-draws, Rydberg 197 (NY). Lancaster Co.: Lancaster, Clements 2587 (NY). 2988 (G). Kansas. Comanche Co.: west of Protection, £. 3 Palmer 41860 (NY). Ellis Co.: near Hays, Rydberg & Imler 1 250 (N Geary Co.: Ft. Riley, Gayle 576 (NY). Graham — Imler 65 (NY). Grant Co.: Ulysses, Thompson 2 be Y). Kiowa Co.: Belvidere, Sept. 5, 1898, White (NY). Mountains, Demaree 12827 (NY). Muskogee Co.: Arkansas Junction, Bush 533 (G, NY). Woods Co.: near Fairvalley, & W ; Stevens 1641 (G, M, NY). Txxas. Brown Co.: Reverchon 76 ( Crane Co.: 11 miles east of Grand Falls, : Dallas Co.: Dallas, June, 1874, Reverchon (G). El Paso Co.: El Paso, M. E. Jones 4195 (NY). Hall Co.: Estelline, Reverchon 3789 (G). Menard Co.: Menard, Cory 24700 (G). Nueces Ce hear Corpus Christi, Mar., 1894, Heller (NY). Presidio pe Marfa, Havard 16 (M). Randall Co.: Palo Duro Canyon, Ba" 1257 (NY). Reeves Co.: Verhalen, Cory 1956 (G)._ Tarrant G. low woods, Ruth 304 (NY). Taylor Co.: Abilene, Tracy 7861 (G 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 239 Tom Green Co.:3 miles east of San set ote Cory 4879 (G). Victoria Co.: Guadalupe, 105 miles southwest of San Antonio, 1879, Kd. EM As 1201 (G). Wilbarger Co.: Chillicothe, Ball 959 (NY). on Co.: Sutherland Springs, 1879, Ed. Palmer 2047 ( Motes rien! Lewis and Clark Co. : near Helena, Aug. 12, 1892, Kelsey (NY). Park Co.: Suksdorf Gulch, 9 miles west of Wilsall, pirsgts 86 (G). Yellowstone Co.: Custer, Fe) |e Blankinship 6 (M). Wyomine. Teton Co.: Spread Creek, Tweedy 368 (Nyy Crook Co.: Devil’s Towe. Little Missouri Butte, Aug. 17, 1897, Griffiths & Carter (NY). Johnson Co.: South Fork of Powder "River, Goodding 257 (G, NY). Converse Co.: Orin Junction, Nelson 2573 (G. NY). Yellowstone National Park, G. N. Jones 5285 (NY). Conorapo. Weld Co.: New Windsor, Osterhout 2369 (NY). Larimer Co.: Ft. Collins, July 9, 1896, Baker (M, NY). Boulder Co.: near Boulder, Tweedy 5058 (NY). Denver Co.: near Denver, Sept. 25, 1916, neciae (G). El Paso Co.: Manitou, F. FE. & E. S. Clements 25 (M, N Y). Fremont Co.: Canyon City, Clements 276 (NY). Pueblo Co.: Pueblo, Baker, Earle & Tracy 5 (G, NY, O). Montrose Co.: Naturita, Payson 510 (G, Ph). Ouray Co.: near Ouray, Underwood & Selby 240 (NY). Archuleta Co.: Arboles, Baker 4538 (G). Utan. Box Elder Co.: Brigham, Zundel 2 242 (NY). Cache Co.: 3 miles northeast of Logan, Maguire 208 (I). Morgan Co.: Peterson, Weber River, Devil’s Gate, Pammel & Blackwood 3911 (G, M). Salt Lake Co.: Salt Lake City, M. E. Jones 1024 (I, NY). Grand Co.: Moab, Christinsen 10039 (NY). Uintah Co.: near Jensen, Peirson 1 2599 (G). San Juan Co.: near Bluff, along San Juan River, Rydberg & Garrett 9897 (NY). NEVADA, 1872, re so (N Y). New Mexico. Rio Arriba Co.: Es pafiola on he Rio Grande above Santa Fe, Sept. 7, 1881, Engelmann (M). Gasisiuen een Miguel Co.: between Anton Chico & Las Vegas, Rose & _ 17620 (NY). Dojia Ana Co.: Dofia Ana, C. Wright 1855 (G, NY). Arizona. eee Co.: Navajo Indian Reserva- tion, north end of Carrizo Mountains, Standley 7371 (US). Navajo Co.: Holbrook, Oct. . 1897, Zuck (US). Coconino Co.: Houserock, Eastwood & Howell 6473 (G). Ipano. Nez Perces Co.: Lake Waha, Heller 3343 (NY). Oregon. Multnomah Co: Hayden Island, Thompson 3749 (M). Wasco Co.: The Dalles of the Columbia, "Aug. 2, 1880, G. Engelmann (M). Wallowa Co.: Horse Creek Canyon, Sheldon 81 14 (NY). Grant Co.: 4 miles east of John Day, Peck 18612 (NY). Wasuincron. Okanogan Co.: Palmer Lake, Loomiston, Elmer 602 (M). Chelan Co.: Peshastin, Wenatchee Valley, soe & Leiberg 472 (G, NY). Stevens Co.: Marcus, Kreager 461 (G, NY). Pierce Co.: Murray, American Lake, M uenscher 5103 (G). Walla Walla Co.: Waits- prt orner 1161 (G). Klickitat Co.: White Salmon, Suksdorf ) 240 Rhodora [JUNE This species ranges the farthest north of any species of Euphor- bia in North America. It nearly reaches 54° north latitude at Fort Saskatchewan, Alberta. On the other extreme it ranges south to the Gulf of Mexico in Texas. I was under the impression that I had seen specimens from California in some of the herbaria in that state. None of my notes confirm this and only one Californian specimen has been found in the more than three hundred collections examined and that had atypical seeds and foliage: Shasta Co., along the Sacra- mento River, but in dry ground, near Redding, Sept. 4, 1889, Greene (F). Plants nearly intermediate between EL. glyptosperm4 and E. serpyllifolia occur on the southern slopes of Mt. Shasta. The plants referred to E. glyptosperma by Jepson, Man. Fi. Pl. Calif., 599. 1925, were probably E. Abramsiana. The fact that EZ. glyptosperma does not appear in Jepson, Fl. Calif. 2:—. 1936, supports this supposition. ; There are two other collections sufficiently different to merit comment. Hastwood & Howell 6473 from Arizona has unusually narrow seeds, thus approaching HZ. Abramsiana. Suksdorf 86 from Montana has unusually large seeds. ; 37. Evpuorsia Apramsiana L. C. Wheeler, Bull. So. Calif. Acad. Sci. 33: 109. 1934. Typr: Heber, Imperial Valley, in perial County, California, June, 1904, Le 4091 (D33555!; isorypes G!, M!, Ph!). E. pediculifera Engelm. vat- Abramsiana (L. C. Wh.) Ewan in Jepson, Fl. Calif. 2: 427. 1936. hamaesyce saltonensis Millsp. in Parish, Cat. Pl. Salton Sink, 6. 1913 (preprint from Carn. Inst. Wash. Pub. 193: 110. 1914); but mostly much shorter, gradually shortening upward; leaf- the nodes but mostly con i - gested in groups of 5 to 10 on very leafy lateral branches; involucres 0.6-0.7 mm. in diam., glabrous outside, glabrous inside except on the lobes, obconical, tapering 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 241 to the peduncle; proximal lobes Core piviceii the glands, each deeply parted into 2-4 slender glabrous segments, the distal lobes exceeding the glands, mostly patted into 2 slender glabrous segments, all the lobes with a few hairs inside at the base; fifth gland equaling the glands, filiform, glabrous; sinus narrowly \- shaped, little depressed, making the fifth gland appear somewhat as one of the divisions of the Tobes; ae transversely ih, a to subcircular, 0.15—0.2 mm. long appendages mostly wider the glands, white, i ocvevene ante or slightly two-lobed ; ale oles reduced to one slender segment opposite each glan nd, adnate to the involucre below, free portion of one or two slenderly fili- form segments about equaling the androphores, with a few short hairs; staminate flowers 3-5 per cyathium; androphores glabrous, 0.7-0.9 mm. long; gynophore glabrous, exserted but often not reflexed; ovary ia rous, 3-lobed; styles bifid ca. halfway, glabrous, clavate, ca. 0.3 mm. long, rotately spreading but the tips slightly nproniliics: capsule glabrous, rotundly ellipsoid- oblong, 1.3-1.7 mm. long, subacutely praneioes seeds sharply PnP jr 1-1.4 mm. long, 0.6-0.7 mm. tangentially, ca. 0.6 mm. radially, n ear ovate to Jee bengnete radially, base truncate, goes facets slightly concave, dorsal facets plane or slightly convex, all with 4-6 irregular transverse rounded ridges slightly including the angles, coat white, microreticulate.— PLATE 660 Colorado ‘Desert, California, southern Arizona, south to northern Sinaloa (Map 1). Representative specimens seen: CALIFORNIA. Imperial rei streets of Brawley, about 15 feet about 2 feet below sea-level, Parish 8302 a oie kh ARIZONA. Yavapai Co.: Fort Ve rde, MacDougal 511 (US); Fort Verde, Mearns 155 (NY). Ma aricopa Co.: 20 miles south of Gila Bend, Harrison, Kearney & Fulton 7991 (A, US). Pinal Co.: Sacaton, Peebles, Harrison "¢ Kearney 83 (US), Peebles 10595 (US); weed in greenhouse, experiment farm, Sacaton, Peebles 5834 (A ). Pima Co.: on range reserve, Wilmot, Thornber 342 (M, among rocks, Papago rg soe Harrison & Kearney 7993 (A, US). Co.?: “Mesas,” June 23, 1881, Pringle (NY, US). MEXICO: Sonora: _Guaymas lee sandy places, Sept., 1887, Ed. Palmer 186 (G, US); Guaymas, aan gravelly ridges, Sept., 1887, Ed. Palmer 187 in part (G, US). Srnauoa: dry hills, Fu erte, Rose, Standley & Russell 13541 (US). Ae citation of additional specimens see Bull. So. Calif. Acad. 33: 110. 1934. The specimens from Fort Verde are atypical in hat they have coarser stems with coarser pubescence, and seeds approach- ing those of E. glyptosperma which is the closest relative of this species, 242 Rhodora [JUNE 38. EupHorsia theriaca sp. nov. Annua, glabra; caulibus prostratis vel suberectis; laminis foliorum ovatis, obovatis, vel ovato-orbicularibus, 3-5 mm. longis, integris; petiolis 0.7-1 mm. longis; cyathiis solitariis; involucris turbinato-campanulatis, diametro 1.3-1.4 mm.; glandulis transverse ovalibus vel oblongis, 0.6-0.8 mm. longis; appendiculis nullis vel rudimentis tantum; floribus masculis 30-36 per cyathium; stylis bifidis, 0.3-0.4 mm. longis; capsulis glabris, subacute 3-angulatis, 1.4—-1.6 mm. longs; seminibus acute quadrangularibus, 1—1.2 mm. longis, in quoque latere 2-3 rugis transversis praeditis, radialiter ovato-acutis, bas! truncatis. .3-1.4 mm. in diam., 1.7-2 mm. long, glabrous outside, with short hairs at the summit and beneath the glands inside; lobes rap fer acuml- strongly — roundly 3-lobed ; styles 0.3-0.4 mm. long, slightly m. slightly broader than long, strongly and subacutely 3-lobed; broader below the equator, base truncate; seeds t i gonal, 1-1.2 mm. long, 0.7-0.8 mm. tangentially and radially, ovate-acute radially, base truncate, facets with mostly 3, pager times 2, high rounded ridges passing but slightly through the sharp angles, coat white, microreticulate PLATE 6 : yPE: Blue Creek, Chisos Mountains, Brewster County, Texas, Aug. 1, 1931, C. H. Mueller 8080 (G!). This same number at M, NY, and US consists wholly of E. cinerascens. There was a little bit of this mixed with er collection at G. This number at F contains a few fragments ° E. theriaca, the rest is E. cinerascens. Known only from the tyP® (Map 7). 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 243 39. EUPHORBIA pee Engelm. in Emory, U. 8. & Mex Bound. Surv. 2 (1): 1859. Type: Arizona (without further locality) Nov., 1856, ei " Schott (M 46906!, fragment F!, photo- graphs F!, GI, W!). A young erect plant with large leaves.— Chamacsyce arizonica (Engelm.) Arthur, gels 11: 260. 1911. E. versicolor Greene, Bot. Gaz. 6: 184. 1881. Typx: San Fane Mowe near Clifton, Greenlee County, Arizona, Sept. 3, 1880, EZ. L. Greene (Herb. Greeneanum, Notre Dame; photograph Gi: isoTyPE NY!). About an average member of the species. —Cham ot oir versicolor (Greene) Norton, Contr. . 8. Nat. Herb. 25: 345. 1925. E. portulana 8. Watson, Proce. Amer. Acad. Arts & Sci. 24: 73. 1889. Typm: island in harbor, Guaymas, Sonora, Mexico, Oct., 1887, Ed. Palmer 321 (G!; 1soTyrE C!).—Stems unusually rigid and numerous. —Chamaesyce portulana (S. Wats.) Millsp., Field Mus. Pub. Bot. 2: 411. E. purisimana Mill, coe Calif. Acad. Sci., ser. 2, 2: 225. 1889. Typx: Purisima, Lower California, Mexico, Feb. 12, 1889, T. S. Brondegee (F 196127); photograp hs G! W!; ISOTYPE C!). Unusual in that the glands on most, but not all, of the involucres lack appendages.—Chamaesyce purisimana (Millsp.) Millsp., Field Mus. Pub. Bot. 2: 411. 1916. E. collina T. S. Brandegee, Univ. Calif. Pub. Bot. 4: 184. 1911 (not Philippi, Soma 29: 41. 1857-8). Typr: Cerro de San Ignacio, Dura Mexico, July, 1910, C. A. Purpus 4599 (C 144767!: fiaenat F!; isotypes F!, G!, MD). An entirely typical member of the species. Perennial from a inetd taproot, erect or prostrate, soepontamrae forming mats; stems up to 30 em. long, slender, with pread- Ing, mostly clavate, meroseopialy smooth hairs, Darods up to 3 cm. long, generally shorter, often very short towards the ends of the ies teat bla les reddish, 1-10 mm. long, deltoid- Ovate, ovate, ovate-oblong with oblique ‘base, 0 r the upper very small and oval-cuneate, mostly with fine spre reading hairs at least on the lower surface, margin entire, petio g, clothed as the leaves; stipules minute and often not visible lower united, — distinct: peduncles mung with fine preading airs, 1-4 m . long, mostly 1-2 mm. long; seatlaa. salary at the nodes = ‘often congested by id shortening of the upper internodes; atari long-turbinate, constricted above, ca. 1.5 mm. long, 0.8 mm. diam., with sparse short spr reading hairs without, glabrous within; lobes sabalece, equaling the glands, entire, ciliate on the margins; glands 1. 5-2 times as long as wide, mm. long, concave, red; appendages glabrous, oval, up to 1 mm. long, white but rubescent; fifth gland absent, its sinus narrowly V-shaped and depressed halfway to base of the involu- cre; bracteoles reduced to one sparsely hairy filiform appendage 244 Rhodora [JUNE adnate below to the involucre opposite each gland, two-thirds as long as the involucre or shorter; staminate flowers 5-10 (or rarely 12), mostly 6-7; androphores glabrous, equaling the glands, 1.5 mm. long; gynophore glabrous, long-exserted and reflexed at maturity; ovary three-lobed, densely pubescent, styles parted to the middle, slender, glabrous, ca. 0.6 mm. long; capsule with spreading hairs, spheroid with flattened base, ¢a. 1.5 mm. long and in diam., obtusely three-angled; seeds quadran- gular, 1-1.2 mm. long, ca. 0.6 mm. radially and tangentially, base truncate, outline narrowly ovate radially, raphe obliquely truncated by the micropylar region, facets with low, often anastomosing, ridges which slightly include the angles, white coat very thin, the brown testa showing through.—PtaTeE 661B. Colorado Desert, California, south to Lower California, east to Texas, Chihuahua, and northern Durango (Map 43). ss Pima Co.: Sierra Tucson, April 27, 1884, Pringle (NY); abo- quivari Mountains, Kearney & Peebles 10895 (US); rocky south slopes, north of Ranger Station, Baboquivari Mountains, Wiegand, Maguire, Richards & Moeller 10761 (G, 1). Santa Cruz Co.: Tumacacori, Harrison & Kearney 6021 (G, US). Greenlee Co.: disturbed soil at side of Highway 81 on Smelter Hill, south oe 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 245 My previous statement, |. c., as to type locality and type col- lection was based on confusion resulting mainly from the fact that I had not seen Engelmann’s specimens. Now it is clear that the specimen taken as type here is the one which should be and has been taken as type by Millspaugh. There is a specimen of this species purporting to have as its origin Warm Springs, Mo- have Desert, San Bernardino County, California, May, 1882, S. B. & W. F. Parish 1384 (M 144655). But this is the type collection of #. Parishiit. Another sheet of the same collection (M 144654) is EZ. Parishii. Since 144655 is the only specimen of E. arizonica claiming to have come from Mohave Desert and since it appears that a mixture may have occurred in the mount- ing, I question the source of 144655. 40. EUPHORBIA SETILOBA Engelm. ex Torrey, Rep. Expl. & Surv. Miss. R. to Pacific Ocean 5 (2): 364. 1857. Type: Fort Yuma, Bey sees Coane California, Major G. H. Thomas (N Yt; ISOTYPE M!). A biologically typical member of the species.— Chamaesyce setiloba (Engelm.) Millsp. ex Parish, Cat. PI. eon cero 6. 1913 (preprint from Carn. Inst. Wash. Pub. 193: 110. 14), E. floccosiuscula M. E. Jones, Contr. West. Bot. 15: 145. 1929. Typr: Hermosillo, Sonora, emai Oct. 27, 1926, M. E. Jones 22789 (P!; isoryPE M!). A plant with unusually long internodes, apparently from a shady situation. Annual, with microscopically smooth tapering hairs through- out; stems slender, prostrate or perhaps rarely erect, rarely forming mats, internodes up to 2 em. long, often muc shortened toward the stem-tips congesting the leaves and involucres into dense small ‘“heads’’; leaf-blades 2-7 mm. long, oblong or oblong- ovate, base slightly oblique, petioles up to 1 mm. long; stipules not apparent; peduncles up to 1.5 mm. long, clothed as the herbage, filiform; cyathia solitary at the nodes; involucres very shortly hairy without, glabrous within, long-turbinate, constricted above, ca. 1.2 mm. long, ca. 1 mm. n diam.; lobes narrowly deltoid, entire, ciliate, equaling the Foes glands red, trans- versely oblong or the ‘distal sometimes discoid, concave, 0.1—0.2 m. long; ape white, glabrous, ca. 1 mm. long al wide, parted into 3-5 narrow attenuate segments; fifth eee roster absent, its sinus depriesed haiteay to base of the involucre; bracteoles reduced to one filiform appendage adnate below to the involucre opposite each gland, with few or no hairs, sere portion only ca. 0.2 mm. long; staminate flowers 3-7 per cyathium; andr ophores glabrous, ca. 1.2 mm. long, ca. " daunlite the as. Synophore glabrous or with a bt hairs above; ovary three- 246 Rhodora [JUNE lobed, long-white-hairy, styles glabrous, 0.4-0.5 mm. long, ~~ o the base, clavate, slender; capsule long-hairy, peer ee . 1.1 mm. long, sharply angled, seeds 0.9-1 mm. long, ~ : ee radially and tangentially, quadrangular, sharply angle ‘a i with low irregular wrinkles, base truncate, ovate-acutish 1a y coat microreticulate, white but the brown testa showing through. t+ Mountain and Lone Mountains, et T562C (US); infrequent at Persimmon Gap, vo Re (US). MEXICO: Baga CALIFORNIA: Agua Verde, J. N. Bot. Club 63: 429-430. 193 Lind The specimen at M, which it would seem might well be ta as type, is too fragmentary. 41. EupHorsra vituirera Scheele, Linnaea (22: 153. i Annual, or de nitely perennial in some Mexican en stems erect, 6-45 em. tall, pilose with long white tapering | ee or rarely glabrous, epidermis papillate, papillae disappeane os age, internodes mostly 1-2 em. long, ultimate branchlets © we subcapillary; leaf-blades ovate to triangular-ovate-acute, i faleate, 3-10 mm. long, uppermost often much reduce A narrower, mostly glabrate a : ateral dermis papillate beneath, base usually strongly age ie ca. ‘ strongly amplexicaul; stipules 0.5—1 mm. long, distinct to vie united, mostly attenuate and often with erect linear ee : ie mm. ong; ate, 0.9-1 mm. in diam a ee short spreading hairs outside, pubescent at the summit, me lobes triangular, acuminate, hairy inside, about equaling 1941] Wheeler,— Euphorbia Subgenus Chamaesyce 247 glands; glands transversely elliptical to oblong, 0.3-0.5 mm. long, depressed or slightly folded in the middle; appendages 1-3 times as wide as and mostly a little longer than the glands, white, glabrous, margin entire to crenulate; fifth gland linear, 14-24 as long as the lobes; sinus small, U-shaped, little depressed; bracte- oles partly free and partly united into a radial partition adnate or ca. 24 its length to involucre, somewhat shallowly divided above, pubescent above, nearly equaling the androphores; staminate flowers 9-25 per Perse fs androphores glabrous, ca. . mm. long; gynophore glabrous, exserted and mostly Aiea vary glabrous, sharply 3-angled; styles glabrous, ca. lone. le— bifid; capsule glabrous, sharply 3-angled, ihe Hatt often produced be yond the seed into an empty carina, 1.3-1.9 mm. long, 2.6-3.2 mm. in Gan m.; see oak ovoid-qu had gular, 1-1.1 mm. long, 0.7-0.9 mm. tangentially and radially, radially ovate, subacute, back and lateral angles sharp, facets convex, smooth to faintly and broadly wrinkled, coat light brown to white, subfarinose and microreticulate. Key To VARIETIES Herbage more or less PRIOR. oS 6 ie as es a. var. typica. Hedhiags Mim w ets koe ie AS eee ce es ie ea os ae b. var. 7 E. VILLIFERA Scheele, Linnaea 22: 153. 1849, var. Feist Type: New B raunfels, Comal County, Texas, "Aug.— Sept., 1846, Pr Lindheimer 293 (?; isotype M 200497!). og collection designated as type by Blankinship, Ann. Rep. mb Gard. 18: oe 907.) A good representative of the POR some of Bnet slightly toothed.—Chamaesyce villifera (Scheele) Small, Fl. Se. U. S., 712, 1333. 190 Chamaesyce Stanfieldii Small, hc. Types: San Marcos and vicinity, Hays County, Texas, summer, 1898, S. W. Stanfield (NY!; photographs G!, W!). Leaves mostly entire, good match for isotype of EF. villifera. —E. Stanfieldii (Small) Cory, Ruopora 38: 407. 1936.—PLate 659D. Texas, south to Oaxaca, Yucatan, and Guatemala (Map 16). Representative specimens from the United States: Trxas. Hays Co.: San Marcos and vicinity, summer, 1898, Stanfield (NY); San Marcos, spring, 1897, Stanfield (N Y). Travis Co.: Mt. Bonell, Austin, Hall 550 (M, NY); Mt. Bonell, a Austin, Young 50 (G, M); Senin nes of Austin, Rose & Russell 2411 1 (NY). Comal Co. r New Braunfels, Plateau 5447 (NY); New Braunfels, 1846-1851, Lindheimer 294 (M), 808 (G, M), 508 (G, M), 1148 (G, NY), 1149 (G, NY); New Braunfels, 1850, C. Wright (M, NY). Kerr Co.: 5.8 miles southeast of Mountain Home, Cory 19310 (G); 13 34 miles southwest of Kerrville, Cory 23954 (GQ). Real Co.: 16 miles north of Bandera, Cory 8898 (G); 248 Rhodora [JUNE 16/4 miles north of Leakey, Cory 24387 (G); 14.7 miles north of Leakey, Cory 27404 (G). Brewster Co.: Blue Creek Chisos Mountains, Cory 7006 (G); Blue Creek Trail, Chisos Mountains, Cory 7329 (G); Chisos Mountains, Mueller 8081 (G, NY); im basin of Chisos Mountains, Warnock C646 (US); Glass Moun- tain, Warnock T202 (US). This species is exceedingly variable in habit and margin of the leaf. The distinction attempted by Small when he separated this species into two, one with leaves entire or essentially so and the other with leaves serrate, is useless since it would divide the species in such a way that there would be more intermediates than members of the two extremes. Furthermore, Small’s type was far from typical of the supposed entity he attempted to define. Lindheimer 293 is one of the numbers combined and distributed as Fase. HI 530. Warnock T202 is atypical in having styles - 0.7 mm. long, 33-34 staminate flowers, and a woody perennial root. lb. E. vittirera Scheele var. NUDA Engelm. ex Boiss. in DC. Prod. 15 (2): 45. 1862. Typr: Growing in the bed of a mountain torrent near New Braunfels, Comal County, Texas, Nov., 4 C. Wright (M 200505!; photographs G!, W!; isotypes G!, Ge' NY!). Known only from the type collection. This rather trifling variant is striking for being a glabrous member of an otherwise more or less pilose species. 42. E Bound. Surv. 2 (1): 188. 1859. Type: Guadalupe Pass 0” lanceolate, oblong, or obovate-oblong, but most com ‘ oblong, often faleate, base markedly inequilateral, ™ re sharply serrate on the larger leaves to subentire on some of the smaller; petioles 0.5-1 mm. long; stipules distinct, glabrous Le rarely ciliate, deltoid, attenuate, central lobe much exceeding th lateral and up to 1.7 mm. long; peduncles ca. 1 mm. long, 8% brous to rarely sparsely pilose; cyathia at the upper nodes, solitary, but often crowded toward the branch-tips by the short 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 249 upper internodes; involucre 1-1.1 mm. in diam., shortly ob- conical-campanulate, glabrous to rarely sparsely pilose outside, sparsely short-hairy at the summit inside; lobes triangular, exceeding the glands, often 2—3-parted above; glands transversely elliptical to oblong, or rarely subcircular, slightly depressed in the middle, 0.2-0.6 mm. long; appendages white, glabrous, ca. s wide as to twice as wide as the glands, rounded, subentire to crenulate; fifth gland absent to nearly as long as lobes, linear; p 41, range of EupHorpiA veRMIcULATA in Canada & U. S.; 42, E. POLYGONIFOLIA; 43, E. arizonica; 44, E. Coamarsyce in U. S.; 45, E. cAPITEL- LATA in U. S. (Maps 41 & 42 from Goode’s Series of Base Maps, North America No. 102, by permission of the University of Chicago Press. tt lines are the thousand foot contour. about equaling the androphores; staminate flowers cyathium; androphores 1-1.1 mm. long, glabrous; gynophore long-exserted and usually reflexed, glabrous or rarely very sparsely pilose; ovary 3-lobed, glabrous; styles 0.3-0.4 mm. long, bifid halfway to nearly to the base, subclavate; capsule glabrous, 2.1-2.6 mm. long, broader than long, turgidly triangular in cross- section, broader below the equator, base truncate; seeds quad- rangular, 1.5-2 mm. long, 1.1-1.5 mm. tangentially, 1.1-1.3 mm. radially, ovate to rotund-ovate radially, dorsal and lateral 250 Rhodora [JUNE angles subalate, ventral blunt, ventral facets plane to slightly concave, dorsal facets plane to slightly convex, surface essen- tially smooth, coat chalk-white to occasionally sordid.—PLATE D. Arizona, New Mexico, western Texas, Chihuahua, Coahuila, Durango, and Zacatecas (Mar 7). Representative specimens seen: ARIZONA. Mohave Co.: 4 miles east of Peach Springs, Cochise Co.: between Tombstone and Bisbee, Harrison & . Soco east of Doyle Rentfrow’s “desert claim”, Eggleston 19370 (US). Chaves Co.: 20 miles south of Rosswell, F. S. & E. S. Earle 285 US). Dofia Ana Co.: beside the white sands, Aug. 24, 1899, Wooton (US); mesa near Las Cruces, Aug. 12, 1895, Wooton (NY). E o.: dry hillside near Carlsbad Cavern, Standley 4039 : G peth Co.: Eight Mile Well, Cory 1942 (G). Culberson Co.: Van Lois Tharp 3557 (US); Walker Ranch, Cory 1941 (G). Jeff a Expedition from western Texas to El P May-Oct., 1849, C. Wright 658 (G, NY). MEXICO: Cxraua- HUA: volcanic mesa near Horcasitas, Pringle 799 (G, M, US); pass between Las Animas and Chilicote Station, J. M. 7995 (G) ; 8 miles northwest of Cruces, silty plain, Johnston Sos (G); 2 miles south of San Fernando, silty soil on plain, J ohnston 7940 (G); 3 miles north of Charco Piedra (21 miles northeast of Camargo), Johnston 7928 (G). Coanvia: 12 miles north 0 Monclova, Johnston 7196 (G); 2 miles northwest of Fronter (road from Monclova westward to beyond Cuatro Cienagas); Johnston 7183 (G); 24 miles west of Saltillo, silty benches at base of hills, Johnston 7673 (G); desert 41 miles west of Saltillo, Johnston 7691 (G); Torreon, Oct. 13-20, 1898, Ed. Palmer 501 » US). Durango: 19 miles north of Zaragoza, Shreve Sam (G); Mapimi, Oct. 21-23, 1898, Ed. Palmer 582 (G, US). ZACA- TEcAS: Conception de Oro, Aug. 11-14, 1904, Hd. Palmer a" (G, US); Hacienda de Cedros, Lloyd 209 (US); 5 miles south © — Johnston 7393 (G). Biol wy UPHORBIA INDIVisa (Engelm.) Tidestrom, Proc. Bi Soc. Wash. 48: 40, 1935.—E. ye BK. var. ? indivisa Engelm. in Emory, U. 8. & Mex. Bound. Surv. 2 (1): 187. 1859. T¥PE 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 251 Engelmann’ mounted in such a way that the relation of specimens and labels on the sheet is uncertain. Millsp., Field Mus. Pub. Bot. 2: 387. 1914, designated as type the specime “ optera Sitar come nee to several, prostrate, 6-25 cm. long, mostly 0.5-1 mm. thick, crisply pilose, the See we Sab so, — nodes up to 2.5 mm. ong, edian ca ned to but 2 mm. at the branch tips; eon ner aes peimbiarei ped faleate and strongly asymmetrical st lea tip Tange to 5-8 ] oblong or even linear-oblong, ong, mostly serrulate, mostly glabrous above, sparsely crisp-pilose ride often gla- brate; petioles crisp- pilose ca. 1 mm. long; s ipules distinct, linear-subulat ate, 1-1.4 mm long, usually Hoh with short involucre pi ir ret Nee Binh : mm. in see are exceeding the pr glands transversely oval to rane! seta folded, 0.25-0.6 mm. long, the proximal about twice as lon the distal: appendages white, rubescent, glabrous, “Pogsuedad wavy-margined, the distal fan-shaped, symmetrical, long, the proximal oblong to oblong-spatulate or pena elliptical, greatly prolonged parallel to the columella, 1.1-1.6 mm. long; fifth gland et half as long as the lobes; ‘sinus U- shaped, depressed halfway to base of involucre; bracteoles about half as long as the pen hoeli dete of 1-3 linear segments opposite each gland, iting below to the involucre, pubescent above; rlesegeels _fow 5-15 per cyathium; androphores glabrous, 1.1-1.5 m ‘oie gynophore crisply hairy above, reflexed into the sinus ea scarcely exserted; ovary 3-lobed, densely strigose; styles SS shortly united ‘below, entire or sometimes bifi 44 to base, 0.7-1.3 mm. long; capsule strigose, 1.4-1.6 mm. long, subpyriform, strongly , Slobed, base truncate; seeds quadrangu- lar, 0.9-1.2 mm. long, 0.5-0.6 m m. tangentially, 0.4-0.5 mm. radially, SS ovate-acute cadinllys base truncate, facet with 3-5 rounded transverse ridges passing slightly ius the fae nines tek by sharp deep grooves, coat white, micro- reticulate, often but little oe the pale brown testa.— PLATE 3D. Southern Arizona, southern New Mexico, western Texas, Snicket Chinen Durango, Coahuila, and San Luis Potosi 252 Rhodora [JUNE (Map 11). Representative specimens seen: ARIZONA. Pinal Co.: Oracle, Eggleston 15951 (Ph). Pima Co.: Range Reserve near Tueson, Sept. 13, 1911, Wooton (US). Santa Cruz Co.: Pata- gonia Mountains, Kearney & Peebles 10145 (US); Nogales, Harrison & Kearney 6028 (G, US). Cochise Co.: Mescal, Griffiths 1817 (NY); Bisbee, C. E. Lloyd 459 (G); Rucker Valley, 1881, Lemmon 454 (F, G). New Mexico. Sierra Co.: 2 miles south of Hillsboro, Metcalfe 1292 (G, NY). Texas. Brewster Co.: Alpine, Sept. 1, 1909, Mackensen (M); 13 miles south of Alpine, Cory 9264 (G); Limpia Canyon, Davis Mountains, Hanson 744 (G). MEXICO: Sonora: gravelly arroyo, Thurber 963 (G); Rio Magdalena on road to Cananea and Remedios, Wiggins 7078 (W). Curavanua: hills and plains near Chihuahua, Pringle 326 (G); 5 miles north of Mestefias, Johnston 7954 (G). DuRanco: 26 miles west of Mapimi, Johnston 7780 (G). COAHUILA: Soledad, Sept., 1880, Ed. Palmer 1206 (G). Sawn Luts Poros: in arenosis, Morales, Schaffner 855 (G). This entity is very closely related to E. dioeca HBK. Since the other members of the group to which E. indivisa belongs are all extra-limital, the specific lines drawn by Millspaugh, Field Mus. Pub. Bot. 2: 383-392. 1914, in his paper, “‘Euphorbia adenopter@ in North America”, are accepted, with some hesitation, until the group can be carefully studied. The collection made in Texas in 1855 by Dr. Antisell (NY), reported by Engelmann in Emory, U. 8. & Mex. Bound. Surv. 2 (1): 187. 1859, as E. dioeca HBK., is here included under F. indivisa, as are also the plants described by Coulter, Contr. U.S. Nat. Herb. 2: 389. 1894 as FE. adenoptera Bertol. Evpnorsia Tormirouia L., Sp. Pl. 1: 454. 1753. Tyee (?)——A nisophyllum thymifolium (L.) H Pl. Suce., a. 2.—“E.{uphorbia] maculata L. (var. thymifolia L.)” Griseb., Brit Isl., 53. 1859, basinym not stated but reference made to E. maculata 3. Pl. Carib. (not located ).—Chamaesy thymifolia (L.) Millsp., Field Mus. Pub. Bot. 2: 412. 1916. '; itional synonyms listed by Thellung in Ascherson Annual (or short-lived erennial?); stems rostrate, mostly 0.5-1 mm. thick, He heed ond BAbaimld oer sometimes glabrate at maturity, 15-30 em. long, internodes on the malt stems from a maximum of 2 em. long to only ca. 1 mm. long ne some of the lateral branchlets; petioles 0.5-1 mm. long; blades poy elliptic to oblong (often narrowly so) and ovate-lancee” ? ong, nearly glabrous above, sparsely tomentulose to subglabrous 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 253 beneath; stipules ca. 1 mm. long, linear-subulate, distinct or slightly united, entire or slightly parted, with erect hairs; ; eyathia subsessile, mostly in mall cymes of 2-several on shor + lateral br anchlets, reel! ea eaaly obeonical, soon distended and distorted by the base of the nactially included capsule, upwardly strigose outside, subglabrous serra lobes triangular, distal equaling the glands, proximal mu shorter, upwardly strigose; glands red, Bas rg to very toate gra se oval, the distal the more nearly circular, 0.2-0.25 mm. in diam slightly depressed in the center; : appendages on distabe beads symmetrical, margining the glands and 1 1% times as wide, to rudimentary, on proximal glands up to twice as long as glands, asymmetrical, elongated toward the sinus, glabrous, white to pink, entire to Sibceaate 5th gland absent: sinus broadly U-shaped, depressed nearly to ’ base of involucre; bracteoles few, linear, sparsely pu- bescent; deg cereee flowers 3-5 per cya athium; androphores gla- brous, 0.9-1.2 m m. long; gynophore very short t (ca .0.5 mm. long), strigose above; vila: densely hairy; styles bifid about halfway, glabrous, 0.4-0.5 m saat tips usually clavate; vag up- bic strigose, sharply 3-angled, widest near the’ base, 1.1—1.5 m. long; seeds sharply quadrangular, with low ilveiolis trans- paces vidges woh slightly including the angles, ovate to narrowly ovate radially, 0.8-0.9 mm. long, 0.45-0.55 mm. tangentially and radially. pare TE 655A Mexico, West Indies, South America, Old World tropics. This species is not known from the United States but is included here because of its close relationship with Euphorbia supina. The description given above is based on the fae specimens: Mexico: Tamaulipas: Tampico, June 3-6, 1910, Ed. Palmer 572 (M). Sinaloa: Comedero, Jan. 27, 1940, Aeets 5395 (W); Cofradia, Nov. 25, 1939, Gentry 5062 (W). The application of the name E. thymifolia here made is the traditional one. Whether the type in the Linnaean Herbarium is identical with the plants included here is uncertain. According to his notes, A. Gray considered the specimen of FE. thymzfolia in Linnaeus’ herbarium identical with the entity called E. supina in this paper. That interpretation may be correct but Gray might easily have overlooked the small differences between these two entities. The customary application will be continued until definite evidence to the contrary is submitted by someone particularly acquainted with the plants concerned. I am applying the name £. thymifolia in the sense in which it was used by Boissier in DC. Prod. 15 (2): 47. 1862. Thellung, in his discussion of his application of this name in Ascherson & 254 Rhodora [JUNE Graebner, Syn. Mitteleur. Fl. 7: 475. 1917, states: ‘“ Linné’s Herbarexemplar gehért nach Boissier (a. a. O. 1862), wie auch dasjenige des jiingeren Burmann, zu E. thymifolia in unserem Sinne.” It is not at all certain that Boissier ever saw Linnaeus’ specimen. Boissier merely stated: ‘157. E. thymifolia (Burmann fl. Ind. pag. 2, et herb.).”’ There is no definite evidence to indi- cate that the herbarium referred to is Linnaeus’. 45. EupHorBia supiIna Raf., Amer. Monthly Mag. 2: 119. 1817 (Dec.). Typr: doubtless ‘not extant. ‘‘Very common on the downs and seashores of Long-Island, north and south, — in New-Jersey, in ars &e.”’—Xamesike aN (Raf.) R Aut. Bot., 87. 1840 0.—Chamaesyce supina (Raf.) Moldenke, Au annotated and classified list of H. N. Acidenke collection num- bers from No. 1 to No. 11,277 inclusive, 135. 1939. E. littoralis Raf., simmer, Mo. Mag. 2: 119. 1817 (Dee.), not re lies HBK., Nov. Gen. et Sp. 2: 54 (quarto), i tos 7 (Apr.). Typr: doubtless not extant. s on : oo ge Ses 1819, ‘by reference to E. maculata L. sensu W. P. ‘Ba rton, Comp. Fl. Phila. 2: 185. 18 Type: sandy fields and cultivated grounds, common near New York, probably Torrey (NY!; photographs G!, W)). Elliott, Sketch Bot. So.-Car. & Goareis 2: 655. 1824; Sprengel Syst. Veg. 3: 794. 1826.—Xamesike depressa (Torr.) Raf., Aut Bot., 97. 1840. BE. thymifolia Michaux, Fl. Bor.-Amer. 2: 212. 1803. Bie Ohio and Mississippi Rivers and Illinois. Michaux (Herb. Mus Paris; photograph G!). E. thymifolia L. sensu Pursh, FI. merge Sept., 606. 1814. : _£. maculata L. sensu Torrey, "Bl State N. Y. 2: 176. 1843, and 15 (2): 46. 1862; Small in Britton & hors Ill. Fl. No. U. 8 & Can. 2: 373. 1 1897; Robinson & Fernald, Gray’s New Man. Bot. ed. 7, 547. 1908; Thellung in Ascherson & Graebner, Syn. ioe on Fl 7: 465. 1917; Jepson, Man. Fl. Pl. Calif., 600. 192 E. aac Featherman, Ann. Rep. Geol. § Louisian® p. Geol. Surv. : 2: 71. 1871. Type: Grand Isle, Louisiana, Featherman? (Low siana Sepa: University? ; ISOTYPE G!). A narrow-leaved, pe parently erect maritim Imost inde Sinan, tector € ecotype representing an a Chamaesyce Tracyi Small, Fl. Se. U. S., 713, 1333. : Type: Biloxi, Harrison County, Mississippi, joy 15, 1894 (NY! photographs G! W! ). The same ecotype as E. meganaes0s. * Date according to Barnhart, Bull. Torr. Bot. Club 29: 597. 1902. 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 255 Chamaesyce Mathewsii Small, Man. Se. Fl. 797, 1505. 1933. Type: sand dunes opposite Miami, Dade County, Florida, Nov. 26-Dec. 20, 1913, J. K. & G. K. Sma ll 4556 (NY!; photographs G!, W!). A robust maritime ecotype intergrading too generally for recognition. nnual; stems prostrate to mice sparsely villous or the young stems subtome ntose, sometim glabrate but stem-tips always vestite, 10-45 em. long; leadsblesies 4-17 mm. long, elliptic-ovate, oblong-ovate to linear-oblong, the larger with inequilateral wtp sparsely villous, often glabrate serisaialls above, serrulate to subentire; petioles 1-1.5 mm. long, 2-3-parted, villous; an solitary at the nodes but mostly Riess on aos con ngested lateral branches; involucres obconical, ca. 0.8 mm. in diam., villous; glands 0. 15-0.25 mm long, transversely elongate; fifth gland filiform, 14 as long to ca. e qualing the lobes; sinus U-shaped, depressed, 14 to ¥% to base af involucre, long-hair iry; abatidiiges narrow, white, epee Pre meena crenulate; staminate flowers 4-5 (sometimes 3-3) p athium; androphores glabrous, 0.9—1.2 mm. long; gy panphaee sean? hairy, a exserted and mostly reflexed ; ovary densely strigose; styles ca. 0.4 mm. (rarely to 0.6 mm.) | ong, 144-4 bifid, clavate, capsule sharply 3-angled. oo often partiality glabrate e, wider elow : e shite’ ca, . long; seeds quadrangular, ca. 1 mm. long, ovate radially, rea with subregular low transverse ridges pr slightly in- cluding the angles, coat whitish-brown.— PLATE 662 Quebec and Ontario; general in eastern United States, west to North Dakota and Texas: introduced in Oregon, California, and Arizona; introduced in Eur rope (Map 34). Representative specimens seen: QursBec. Wright Co.: Che lsea, Macoun 87833 (G). Pontiac Co.: Chichester, Marie-Victorin, Rolland-Germain Meilleur 44016 (G). Onrvario. Ottawa, Broad St. Station, Macoun 87834 4 (G); Amherstburgh, Macoun sige Ey Chatham, Wooler, Macoun 8 88084 (G). pee INE. Seria n Co.: Auburn, E. D. Merrill 1649 (NB). Aroostook Co.: Island Falls, “Fernald 2739 (NE). Cumberland Co.: Portland, Fernald, Long & Norton 13999 (NE, Ph); Brunswick, Aug. 23, 1910, Furbish (NE). Han- cock Co.: Schooner Head Road, Bar Ha rbor, Mount Desert Island, Sept. 18, 1896, Rand (NE). Kennebec Co.: Winslow Fernald 2736 (NE); Clinton, July 20, 1907, Bean (NE). Franklin Co.: Farmington, Sept. 2, 1904, Knowlto n (NE). Penobscot Co.: Lagrange, Fernald & Long 13997 (NE, . Piscataquis Co. 256 Rhodora [JUNE Orchard, Chamberlain 228 (NE); York, Mt. Agamenticus, Aug. 17, 1905, Chamberlain (NE). New HampsuHire. Cheshire Co.: East Swanzey, Robinson 539 (G, NE). Cods Co.: Whitefield, Pease 11616 (NE); Berlin, Moore 4329 (G). Grafton Co.: July 6, 1919, Knowlton (Ph). Windham Co.: Bellows Falls, July 18, 1896, Deane (NE). Windsor Co.: Windsor, Aug. 22, 1880, Leland (NE). Massacuuserts. Barnstable Co.: Harwich, Fernald & Long 18687 (NE, Ph); Brewster, Collins 1176 (NE); Yarmouth, Collins 981 (NE). Berkshire Co.: Stockbridge, July 20, 1919, Hoffmann (NE). Bristol Co.: New Bedford, Aug. 5, 1904, Hervey (NE). Dukes Co.: Elizabeth Islands, Fogg 2869 (NE). Essex Co.: Salem, Forbes 2315 (NE). Hampshire Co.: West Chesterfield, Robinson 489 (G). Hampden Co.: Spring- Co.: Norfolk, Ware 4138 (NE). Plymouth Co.: Hull, Rothrock (NE). Suffolk Co.: Savin Hill Beach, Boston, Aug. 28, 1853, : ; 15 ; Middlesex Co.: Killingworth, Sept. 3, 1874, Hall (NE). New Haven Co.: Naugatuck, Harger 5519 (Ph); South End, New Franklin, Oct. 1, 1910, Woodward (NE). Tolland Co.: Bolton, Weatherby 5104 (NE). Windham Co.: Central Village, Sheldon 614 (NE). New York. Albany Co.: Albany, Sept. 2, 1908 Burnham (G). Bronx Co.: Thorn’s Dock, Oct. 2, 1891, Bia (Ph). Richmond Co.: Port Richmond, Staten Island, Dowe 6735 (M). St. Lawrence Co.: Canton, Phelps 636 (G). Tomp- kins Co.: Ithaca, Bechtel 8409 (G). Washington Co.: Vaughns, north of Hudson Falls, Aug. 28, 1891, Burnham (G). New Harbor River, Catawba, Long 61184 (G). Burlington Co: Delanco, Hermann 3631 (PLB). Hae ie Co.: Lawnside, Stone 1077A (PhB). Cape May Co.: Cape May City, Aug: Z 1917, Stone (PhB). Cumberland Co.: Cumberland, Long 3894 (PhB). Essex Co.: Watchung Mt., Pennell 7459 (Ph). Middle- sex Co.: Stelton, Halsted’s American Weeds 88 (G, Mo). Ocean Co.: Lakewood, Hunnewell 8081 (G). Somerset Co.: Watchung, Moldenke 1803 (Ph). PENNSYLVANIA. Berks Co.: Fleetw00® 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 257 Leibelsperger 52 (PhB). Bucks Co.: Grenoble, Long 4601 (PhB). Chester Co.: Honey Brook, Benner 5418 (PhB). Columbia Co.: west branch Fishing Creek, 25 miles above Central, Fosberg 15968 (G). Lebanon Co.: in South Mountain, Penryn, Heller 669 (G). Lehigh Co.: near Allentown, Pretz 4842 (PhB). Mont- gomery Co.: Souderton, Long 26754 (PhB). Northampton Co.: Northampton, Aug. 17, 1923, Churchill (G). Philadelphia Co.: os Benner 1835 (G, PhB). Detaware. New Castle : Newark, Oct. 18, 1923, Otis (PhB). uae Co.: Rehoboth, pote 15, 1895, Commons (Ph). District oF Cotumsia: Brook- land, Oct. 13, 1915, Ban (G); Washington, Aug. 18, 1888, Holm (G). VIRGINIA. Amherst Co.: Monroe, Aug. 16, 1899, Pieters (US). Arlington Co.: near _ Sept. 20, 1935, Blake (W). Bedford Co.: Sept. 25, 1871, A. H. Curtiss (G). West VIRGINIA. Cabell Co.: SRP Park, Gilbert 493 (Mo, Ph). Grant Co.: Bayard, Moo 0 (G i Greenbrier Co.: White Sulphur Springs, ihunnevel o6n7 (G). Mineral Co.: Keyser, Moore ou (G). Tucker Co.: along Shavers Fork near Parsons, Greenm 408 (G). Noxva Carouina. Buncombe Co.: Biltmore, Biltmore pial 407b (G). Cartaret Co.: shore of North River, Ran- dolph 827 (G). Pender Co.: Hampstead, Randolph 989 (G). OUT Bees esi Berkeley Co.: Santee Canal, Ravenel (G). Charleston Co.: Charleston, Fernald & Long 9 "48 (G). Lancas- ter Co.: Elgin, House 2583 (M). Marion Co.: Marion, Wiegand & Mannian 1814 (G). Grorata. Whitfield Co.: Dalscn, Chicka- mauga, Harper 380 (G). Fuoripa. Bradford Co.: Hampton, Wiegand & Manning 1816 (G). Dade Co.: sand-dunes opposite Miami, J. K. & G. K. Small 4580 (NY). Duval Co.: Fredholm 5188 (G). Escambia Co.: Pensacola, 1886, ““A. W. Curtiss’’ (G). Gulf Co.: Apalachicola, Chapman (G). — Co.: Fredholm 6469 (G). Lake Co.: near Eustis, Nash 1058 (G). Lee Co.: Alva, A. S. Hitchcock 328 (G). Palm Beach Co: Palm Bea ch, Hunnewel 7365 (G). Pinellas Co.: St. Petersburg, Deam ‘279 7 (G). Mricuican. Allegan Co.: Dohmen 61 (F). St. — Co.: Sturgis, Aug., 1898, Daniels (Mo). Washtenaw Co.: n Arbor, Hermann 91165 (G ). OHIO. Portage Co.: Garretts- ville, Webb 469 (G). Scioto Co.: Moore s Lane, Friendship, 52954 Denia. Jasper Co.: 5 miles east of Wheatfield, Done 26533 (Deam). fake Co.: north side of Cedar Lake, Deam 21389 (Deam). Marion Co.: Pine, Praback 2315 (G). Morgan Co.: north of Martinsville, along. Traction Line, Deam 13935 (Deam). Noble Co.: north side of Waldron Lake on oe Township, Aug. 9, 1905, Deam (Deam).” Parke Co.: 44 mi 258 Rhodora [JUNE west of Bloomingdale, Deam 41921 (Deam). Pike Co.: 1 mile west of Cato, Deam 52655 (Deam). Porter Co.: northwest of Denham, Deam 29862 (Deam). St. Joseph Co.: South Bend, Nieuwland 11736 (M). Starke Co.: 3 miles west of North Jud- son, Deam 32212 (Deam). Sullivan Co.: 5 miles southeast of Sullivan, Deam 25691 (Deam). Tippecanoe Co.: 2 miles south of Battle Ground, Deam 49615 (Deam). Warren Co.: 7 miles west of Williamsport, Deam 37890 (Ph). Washington Co.: 6 miles west of Pekin, Deam 18980 (Deam). White Co.: bank of Tippecanoe River, 2 miles northeast of Monticello, Deam 32671 (Deam, G). Kenrucxy. Harlan Co.: near Poor Fork Post Office, Kearney 156 (G). Kenton Co.: Banklich Creek, Sept. 25, 1838 (Ph). TENNessEE. Davidson Co.: campus of Peabody Ogden? (F). Pierce Co.: Prescott, Fassett 10287 (Deam); Bay City, Fassett & Wilson 10281 (G). Iurors. Champaign Co.: Cham- paign, Gleason 1982 (G). Cook Co.: Stony Island, H. H. Smith 5979 (G, M, Ph). Hancock Co.: Augusta, Aug., 1859, Mead (@)j un Winnebago Co.: South Beloit, Wadmond 2943 (G). MINNESOTA: Chisago Co.: Taylors Falls, Sept., 1893, Ballard (Mi). Goodhue Co.: Burntside, Anderson 777 (Mi). Hennepin Co.: Fort Snelling Reservation, Rosendahl 2332 (Mi). Houston Co.: Jeffersom W. A. Wheeler 430 (Mi). Hubbard Co.: Itasca Park, Moy stone Co.: Pipestone, Aug. 11, 1925 Peterson (Mi). Scott Co-' Cleary’s Lake, — 596 (Mi) Wabasha = Book Lodge, Co.: ere: July, 1894, Fink (G). Story Co.; Ames, Sept» I Boone Co.: Columbia, Drouet 1012 (Mo). Cole Co.: Jette : use (M). Cooper Co.: Bush 15125. (Mo). unklin Co.: Campbell, Oct. 23, 1893, Bush (M). Franklin Co-' 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 259 Gray Summit, Kellogg 1845 (M). Henry Co.: July 3, 1892, ries (Mo). Jackson Co.: Courtney, Bush 7068 (G). Jefferson : 4 miles southwest of Pevely, si peccngh 506 (M). Laclede Co. : Fork of Gasconade River 1 mile t of Dry Knob, Steyer- mark 13925 (Mo). Marion Co.: Hin nibsl, Davis 1419 (M). Saint Clair Co.: 1 mile north of Collins, Oct. 4, 1930, Drouet (Mo). St. Louis Co.: Jefferson Barracks, bre 3053 (Ph). Saline Co.: below Glasgow Bridge on Highwa y 20, Drouet 1607 ( RKANSAS. Crawford Co.: Van Bur ee, Demaree 15278 (W). "Drew Co.: Monticello, Demaree 138521 (W). Garland Co.: 8 miles east of Hot Springs, Scully 284 (G). Grant Co.: Saline River, Poyen, Demaree 16331 (W). Hempstead segs Tokio, Demaree 9931 (G, M). Hot Springs Co.: Malve Demaree 16298 (Mo, W). Jefferson Co.: Pine Bluff, cere 1540 7 (W). a ee Co.: Caddo Gap, Small Creek, Demaree 9582 (M, Ph). Polk Co.: Mena, Demaree 16057 (W). Pope Co.: Russellville, Demaree 15334 (W). Prairie Co.: White River Ridges, de Va lls Bluff, Demaree 15736 (W). Pulaski Co.: Little Rock, Demaree 8202 (M). Scott Co.: Mansfield, Demaree 18180 (W). Lourstana. Boissier Co.: Alden Sa 1898, Trelease (M). Caddo Co.: Shreveport, Sept. 1847, Gregg Beli Plaquemines Co.: Port Eads, Tracy & “ted 209 (G, Ph). Bea bonne Co.: Houma, on Little sie aie Wurzlow (F }i RTH Daxora. Cass Co.: Fargo, Oct. 16, 1911, Waldron G Ph). NEBRASKA. Kearney Co.: Millen ec ‘OT, 1929, Hapeman (Ph). Nemaha Co.: Peru, Y Michels 419 (O). Otoc Co.: Nebraska City, Aug., Spe! Thornber (M). Webster Co.: Blue Hill, Bates 5025 G). Kansas. McPherson Co.: Lindsborg, June, 1884, Bodin (F). Riley Co.: Norton 471 (G). Oxianoma. Alfalfa Co.: near esa G. W. Stevens 1334.1 (G). Creek Co.: Sapulpa, Bush 536 es ). Harmon Co.: Hollis, G. W. Stevens 1054 (G). Le Flore Co.: tapp, Demaree 16041 (W). Mayes Co.: near Mazie, G. W. pp 2584 (G). Oklahoma Co.: near Oklahoma City, White 5 (G). Osage Co.: near Copan on Coon Creek, G. W. Stevens a1) (G). Ottawa Co.: near Miami, G. W. Stevens 2307 (G). Pawnee Co.: Pawnee, Aug. 31, 1933, Coffman (Mo). Woods Co.: near Alva, = W. Stevens 1603 (G). Texas. Anderson Co.: Palestine, Oct. 21, 1884, Joor (M). Ar ransas Co.: 5 miles east of 1198 Chambers Co.: 534 suites south of Anahuac, Cory 22451 (G). Fayette Co.: 34 me baa of Monument Hill, Cory 10046 ~ adres Co:: i; Cory rei (G) 260 Rhodora [June Uvalde Co.: Uvalde, 90 miles northwest of San Antonio, 1879, Ed. Palmer 1202 (G). Valverde Co.: Devils River, Orcutt 6074 : ct € € West Berkeley, July, 1891, Michener & Bioletti (J). Ventura Co.: Ventura, Abrams 4132 (G, Ph). Los Angeles Co.: Pasadena, Grant 427 (J). Riverside Co.: Santa Ana River at Chino Creek, E. thymifolia 6 5: 171. 1837. Typx: “Hab. Banks of the Mississippi and Ar- kansas” (Ph? or British Mus.?), not located. This variety may There is one glabrous collection of E. swpina: Cameron, Cameron Co., Louisiana, July 4, 1903, Tracy 8477a (G, M). It may be that authentic material of E. supina Raf. exists at Geneva. Boissier in DC. Prod. 15 (2): 47. 1862, appears to have taken the name from exsiccatae rather than from the article ™ the American Monthly Magazine, for no page is cited and Boissier does not mention the other two Euphorbiae that appeared in the same article. Thellung in Ascherson & Graebner, 5Y™ Mitteleur. Fl. 7: 471. 1917, cites “E. supina Rafin.!” which makes it appear that he has examined the Rafinesquean specimen? at Geneva. For discussion of the history of the misapplication of the nam E. maculata L. see Contr. Gray Herb. 127: 74-76. 1939. The identity of the Linnaean specimen of E. thymif olia L. has not been investigated. In view of the very close relationsh!P between the entity to which that name has been applied and B. 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 261 supina it is entirely possible that EH. thymifolia might be appli- cable to the entity here called HE. supina. The identity of HZ. swpina seems beyond question. It is the only prostrate pilose Euphorbia with opposite leaves and vestite capsules occurring in the area mentioned by ae 46. EUPHORBIA HUMISTRATA Engelm. ex Gray, Man t. No. U.S., ed. 2, 386. 1856. Typr: banks of the Mississippi, ge nile Missouri, July, 1833, Geo. 2 ae ipo 4 on 149873); photo- graphs G! W)). Boiss. in DC. Pro 2): 4 1862.—Aniso- phyllum bial sotg ‘(Engel Klotzsch & ake , Abh. Akad Berlin, Phys. 1859: 1860.—Chamaesyce humistrata (Engelm. ) Small, Fl. Se. U. S., 718: 1333. 1903. Annual; stems prostrate to ee crisply and sparsely villous, sometimes ep below, to 35 cm. long, internodes mo ostly 1-3 cm. long; leaf-blades oval to oblong-ovate on the main stems, often Aartowae and much reduced on the branchlets, 4 strongly inequilateral; petioles 1-1.5 mm. long; stipules distinct to united, ibaa rod attenuate, 1-1.5 mm. long, mostly with linear divisions , ciliate; pedunc les ca. 1 mm. long, crisply hairy to glabrate; cyathia solitary at the nodes or, more commonly, in oo clusters on short lateral branches; involucre obconical, 0.8 mm. in diam., villous outside, essentially glabrous oe lobes triangular, attenuate, slightly exceeding the glands; gl ands mostly ipeera-wre to elliptical, reap gabeiroalat 0.15-0.3 m mag pare pendages narrow, white to pink, crenulate to entire; fifth gland Wien ca. 4 as long as che lobes; sinus U- shaped, wide, depressed fully halfway to base of involuere; bracteoles minute, linear; staminate flowers 3-5 per cyathium androphores glabrous, ca. 1.3 mm. long; gynophore slightly slender; capsule strigose, sometimes Satay glabrate, sharply 3-angled: 1.3-1.4 mm. long, wider below the equator; seeds mm. radially, radially broadly ovate, facets nearly smooth, coat rown, granular.—PLa io west to Illinois and Kansas, south to Alabama, west a extreme northeastern Texas; adventive in Virgi inia (Map ns : Oo. F road, Richmond, Fernald & Long 12708 (W); waste Ht and railroad ballast, South Richmond, Fernald & Long 12709 (W). OHIO Champaign Co.: Sept. 10 , 1892, Wane tut. Ph). 262 Rhodora [JUNE Hamilton Co.: near Cincinnati, Lloyd 2485 (Ph). InpIANA: Crawford Co.: bank of Ohio River in Leavenworth, Deam 48035 (Deam). Dearborn Co.: bank of Ohio River in Aurora, Deam 41122 (Deam). Dubois Co.: Hunnely bottom, 2 miles east of Huntingburg, Deam 42635 (Deam). Gibson Co.: swamp 3 miles south of Mt. Carmel, Ill., Deam 24230 (Deam). Greene Co.: swamp I| mile north of Newberry, Deam 24101 (Deam). Harrison Co.: bank of Ohio River 4 miles southeast of Laconia, Deam 41514 (Deam). Jackson Co.: 6 miles north of Crothersville, a 1). Ba thwest_of Carlisle, Deam 51438 (Dea or Switzerland Co.: bank of Ohio River in Vevay, Deam 508 Eugene, Deam 51351 (Deam). Vigo Co.: Greenfield Bayou 9 miles west of Pimento, Deam 23989 (Deam). Warrick Co.: bank of Big Pigeon Creek 6 miles west of Boonville, Deam 24344 (Deam). Inurnors. Menard Co.: Athens, 1863, Hall (M). Peoria Co.: Illinois River bottomlands near Peoria, Aug., 1903, Me- Donald (NY). Rock Island Co.: Rock Island, Aug. 10, 1866, Engelmann (M). St. Clair Co.: Sept. 10, 1886, Eggert (G, ¥ Y). Kenrucxy: Lyon Co.: Kuttawa, Eggleston 5246 (M, NY). T. G. Lea Herb. bank of Alabama River at House Bluff, Harper 40 (G, NY, Ph). Mississtppr. Warren Co.: Vicksburg, Demaree 14111 (W). Mi sour!I. Barry Co.: White River, Sept. 22, 1896, K. K. Mackenzié (NY). Bates Co.: Osage River at bridge south of Amoret, Steyermark 20281 (M). Clark Co.: 2 miles west of Gregory Landing, Sept. 6, 1934, Drouet (G, M, Mo). Cole Co.: Osage River lock, Rickett 1134 (Mo). Cooper Co.: 10 miles southwest of Boonville & south of Lanvine, Steyermark 15910 (M, Mo). Jackson Co.: Sheffield, Bush 123824 (M, NY). Jefferson Co.: Kimmswick, Sept., 1868, Engelmann (M). Laclede Co.: Osage Fork of Gasconade River, Steyermark 13931 (M, Mo). New Madrid Co.: near La Forge, Steyermark 8843 (M). Perry Cos ES ETT PORATED Ira aS eg Re NR a EAS a a Re ee 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 263 Rivers on island, Steyermark 19716 (M). Arkansas. Chicot Co.: Eudora, near La Fourche Lake, Demaree 18575 (W). Drew Co: Monticello, Demaree 18410 (W). Lincoln Co.: Bayou Bar- tholomew, York Town, Demaree 13729 (M, NY). Prairie Co.: De Valls Bluff on White River bottoms, Demaree 15521 (NY). Pulaski Co.: along Arkansas River at Little Rock, Demaree 8328 x, NY). Lourstana. Terrebonne Co.: Houma, Aug. 21, 1912, Wurzlow (NY). Kansas. Miami Co.: Paola, Sept. 19, 1885, Oyster (NY). OxLAHoma. Nowata Co.: Lenapah, G. W. ‘Stevens 2182 (G, NY). Ottawa Co.: bar by Neosleo fier near Miami, . W. Stevens 2309 (G, M, NY). Rogers Co.: Verdigris, Bush 5387 (G, M, NY). Trxas. Bowie Co.: Texarkana, May 9, 1891, Plank (NY). This species seems to grow usually on river flats, judging by the meager habitat-data given by collectors on their labels. Some specimens are difficult to place definitely in EZ. humistrata for it approaches EF. swpina very closely. Reports of E. humi- strata from states other than those from which specimens are cited above are probably based on misidentifications of E. supina. 47. EupHoRBIA sTIcTospoRA Engelm. in Emory, U. 8. & ee Bound. Surv. 2 (1): “ie 1859. Typr: steep bank of the Paw River (‘‘fork’’), in | soil, Kansas, Sept. 8, 1847, A. Fendler 798 (M 200482!; photographs | . W)). (Type designated |b y Millspaugh, Bot. "Gaz. 26: 266. 898.) Boiss. in DC. P 15 (2): 41. 186 pe ik Seep pssst (Engelm.) Small, Fl. = U.S, 714, 1334. 1903. Aibeophithees senile Klotzsch & Garcke, Abh. Akad. Berlin, Phys. 1859: 28. 1860. Typr: Los Bafios, Mexico, C. Ehrenberg (B!; photographs G!, W!; isorypE M!). nnual; stems prostrate to ascending, hegre villous, 5-25 em. long, internodes mos tly 1-2 cm. long on the main stems; leaf- blades suborbicular, ovate, chicane to oblong-linear and obovate, 3-10 mm. long, sparsely crisply hairy and glabrate above, similarly but more densely and permanently vestite be- neath, margin sharply serrate, base oblique; petioles ca. 1 mm. long; ‘stipules triangular, often attenuate, distinct or united, sometimes divided, ca. 1 mm. long, sparsely hairy; cyathia solitary at the nodes but mostly on congested short densely leafy lateral branches; peduncles mostly 1-2 mm. long, pubescent; involucre obconical, tapering to the peduncle, 0.7-1 mm. in diam., pubescent outside, essentially glabrous inside; lobes tri- angilgiy ciliate, about equaling the glands; glands transversely oblong to elliptical or even subcircular, 0.15-0.3 mm. long; appendages narrow, glabrous, white, entire to ss ecinie « or even sharply toothed; fifth gland from a mere apiculation to a linear 264 Rhodora [June segment half as long as the lobes; sinus narrowly U-shaped, depressed ca. 14 to base of involucre; bracteoles minute, linear; staminate flowers 3-7 per cyathium; androphores glabrous, 9-1 mm. long; gynophore glabrous below, crisply hairy above, exserted and mostly reflexed; ovary 3-lobed, densely strigose; styles ca. 0.2 mm. long, expanded at the apex, entire to deeply emarginate; capsule 1.4—-1.9 mm. long, 3-angled, widest near the base, strigose, the basal angles with spreading hairs; seeds sharply quadrangular, 1.2-1.4 mm. long, 0.7-0.8 mm. tangen- tially, ca. 0.6 mm. radially, radially narrowly ovate to oblong- ovate, base truncate, apex acute, facets with shallow irregular pits to subregular low transverse ridges, coat mottled whitish brown and brown, the bottom of the pits darker colored than the ridges —PLaTE 662B. : South Dakota and Wyoming, south to Chihuahua, Coahuila, Durango, and Zacatecas, west to Arizona (Map 28). Representa- tive specimens seen: Soutu Daxora. Fall River Co.: Hot Springs, Aug. 27, 1909, Petersen (NY). Stanley Co.: Missouri Valley near Fort Pierre, June 19, 1839, Geyer in part (US). Black Hills, Aug., 1893, Thornber (M). Nepraska. Dawes Co.: Whitney, Aug., 1890, Bates (G). Webster Co.: Red Cloud, July 22, 1904, es (G). Kansas: Ellis Co.: near Hays, Rydberg & Imler 1244 Hitchcock 472a (G, M, NY). Logan Co.: on Turkey Creek, east of McCallaster, Rydberg & Imler 1153 (NY). Meade Co.: miles southeast of Guymon, Stratton 459 (M). Woods Co.: Alva G. G, NY). Texas: Brewster Co.: frequent 0” Sul Ross Campus, Alpine, Warnock T589 (US). Culberson Co.: Walker Ranch, Cory 1941 (G). El Paso Co.: El Paso, M. Jones 4195 (G, I, NY in part, O). Gaines Co.: 3.9 miles pr: of Seminole, Cory 16617 (W). Hudspeth Co.: Eight Mile Wel Cory 1946 (G). Kerr Co.: 5.8 miles southeast of Mountain Home, Cory 19317 (G). Martin Co.: Stanton, June 14, 1900, Eggert ( 95 Pecos Co.: 514 miles southeast of Fort Stockton, Cory 152 ve Reeves a Cory 1949 (G). Ward Co.: Barstow, Tra Canyon City, M. E. Jones 786 (I, M, NY). Larimer Co.: Fo Collins, Anonymous 4020 (NY). ‘Otero Co.: 15 miles northeast 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 265 of La Junta, Rollins 1876 (M, NY, W). Pueblo Co.: etary Baker, Earle & Tracy 6 (G, M, NY). Weld Co.: Sec. 18, T R.65 W, Muir 149 (G). Yuma Co.: Wray, 1907, Shantz ONY)’ New Mexico. Catron Co.: Beaverhead, Datil Forest, vee 20424 (G). Chaves Co.: south of Rosswell, Earle M). Dofia Ana Co.: Mesilla Valley, ts & Standley hore (M). Lincoln Co.: Gray, Skehan 51 (G, NY). Santa Fe Co.: Santa Fe at base of hills, iba 797 (G, M). San Miguel Co.: near oe Standley 4939 (G, M , NY). Sierra Co.: Kingston, Metcalfe 1006 (M). Arizona: Cochise Co.: Do uglas, Thornber 7111 (T); near Douglas, Harrison & Kearney. S ae (U vie aradise, Chiricahua Mts., Blumer 1690 (G, M, US); Benson, Harrison 8224 (US); Bisbee, Aug. 28-30, ee Disieg (Ph). Pim a Co.: Rose- mont, Santa Rita Mts., "Thornber 9055 (T). Santa Cruz Co.: near Nogales, Peebles, Harrison, & Kearney 5564 (US); Pata- gonia Mts., Kearney & Peebles 10140 (US). MEXICO: CuI- HUAHUA: near Chihuahua, Pringle 327 (G, US); southwest of El Paso, near Comalitos, Thurber 731 (G, M). CoanuiLa: 1 mile south of Hermanas, Johnston 7061 (G); 4 miles east of Cua- tro Cienegas, eet 7128 (G); Saltillo, Nov. 2-5, 1898, Ed. Pe 567 (G, US); Torreon, Oct. 13 3-20, 1898, Ed. Palmer 489 US); Jimulco, Pringle 80 (G, US); Ramos Arizpe poner isan. and Sacramento, Wynd & Mueller 77 (G, M, US). San Luis Porostr: near San Luis Potosi, Parry & Ed. y Seven eh (G, M). Zacatecas: 5 miles south of Majoma, Johnston 7 (GQ); near tinh del Oro, Aug. 11-14, 1904, Ed. Palme 319 (G). Durango: 26 miles west of Mapimi, J ohnston 779 (Q); Santiago Peck ats Apr. & Aug., 1896, Ed. Palmer 13 Alas , US); Tepehuanes, June 4-25, 1906, Ed. Palmer 299 48. EvpHorsra CHamasrsyce L., Sp. Pl. 1: 455. 1753. Type: Jamaica, P. Browne (Linnaean Herb., not seen; photograph F!, W!; fragment FI, the right-hand specimen wa An a ke member Niiadee if extant.) Boissier ‘in DC. ‘Prod. 15 on 47. ona, Boissier, Icon. Eu Bh 1866, poor; Fiori & Paoletti, Fl. Ital. hist; 303, as 1908; Thellung, Bull. Herb. Boiss., ser. 2, 7: 768-70. "1907; Thellung in Ascherson & Graebn go Syn. Mitteleur. Fl. 7: 476-479. 1917.—E. serpillacea Willd. Klotzsch & Garcke, Abh. Akad. Berlin, Phys. 1859: 26. 1860, va ‘ Date fide Thellung, Bull. Herb. Boiss., ser. 2, 7: 769. 1907. 266 Rhodora [JUNE synonym of Anisophyllum prostratum, and Baillon, Adansonia 1: 116. 1860-1861 as synonym of E. prostrata.—Anisophyllum Prod. 15 (2): 47. 1862 EL. tenella HBK., Nov. Gen. et Sp. 2: 53(quarto). 1817. TyPE: ad Ripas Orinoci prope Maypures et Carichana, Venezuela (Herb. Mus. Paris, not seen; photograph G!; fragment of 1soryPE from Berlin at M!). E. prostrata 8 vestita Engelm. ex Boiss. in DC. Prod. 15 (2): 47. 1862. Type: New Orleans, Louisiana, Hale (Ge!; photographs G!, W!). Scarcely more vestite than usual and of no consequence. E. stictospora Engelm. var. texensis Millsp., Bot. Gaz. Fe, 1898. Type: southern Texas, 1894, Heller 1913 (F 196483 ee Chamaesyce stictospora guadalupensis Small, Fl. Se. U. D.; be 1334. 1903.—C, stictospora var. texensis (Millsp.) Fedde in Just, Bot. Jahrb. 32 (1): 368. 1905. Chamaesyce malaca Small, Fl. Se. U. 8., 713, 1333. re Type: Kerrville, Kerr County, Texas, June 26-30, 1894, me Heller 1918 (NY!; photographs G!, W!; 1sorypes F!, G!, M’). E. malaca (Small) Little in Jeffs & Little, Univ. Okla. Biol. Surv. Pub. 2 (2): 70. 1930 (Prelim. List Ferns & Seed Pl. Okla.). : _ Boiss. in DC. Prod. 15 (2): 47. 1862, lists two additional extra limital synonyms which have not been confirmed. Annual; stems prostrate to decumbent, few to several, mostly 1-1.5 mm. thick, crisped-hairy to glabrate, at least the y ye tips vestite and often subtomentose, internodes rarely up 10 a age on main stems about 1 cm. long, on sho . — ovate, 3-11 mm. long, mostly 4-8 mm. long, margin ar . serrulate, base mostly inequilateral, upper surface mostly !° brous, lower surface finely and crisply hairy, often glabrates petioles ca. 1 mm, long; stipules with short hairs, triangular short hairs or sometimes glabrous; cyathia solitary at the no but mostly on short-noded lateral branchlets; involucres yer 0 m. in diam., obconical, tapering to the peduncle, ye Scattered crisped hairs or glabrous outside, glabrous 10S! : except on the lobes and yo equaling the glands, hairy; glands transversely oval to oblese 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 267 0.15-0.3 mm. long, usually depressed in the middle; igre white, glabrous, from about as wide as to twice as wide as the gland; sinus U-shaped, somewhat depressed, glabrous in the fower half: fifth leak filiform, 4-24 as long as the lobes; brac- teoles nearly wanting, reduced to 2-3, glabrous, filiform; stam- inate flowers 4 per cyathium; androphores glabrous, 0. 9-1.1 mm. long; gynophore hairy above or glabrous, exserted and usually eflexed; ovary roundly 3-lobed, betes white-hairy; styles bifid nearly to the base, 0.2-0.3 long, glabrous, slightly clavate; capsule sharply 3-angled, LL. 4 mm. long, wider below the equator, with crisped hairs persistent on the backs of the carpels but more nn Tess deciduous on the sides; seeds sharply quadrangular, 0.9-1 mm. long, 0.6-0.7 mm. tangentially and radially, radially ovate, pave traversed by low narrow slightly ‘irregular ridges which scarcely pass through the angles, ventral facets concave, dorsal plane to convex, coat white, microreticu- late, — little ¢ concealing the brown to gray testa.—PLATES 660C a laodans in Massachusetts and Virginia; South Eee Florida, Alabama, Louisiana, Missouri, Nebraska, Okla Texas, ‘and Arizona; introduced in the Old World (Map i) Grimes 4640 (G, NY). So Carouina. Charleston Co.: Charleston, B. L. Robinson 237 (G). Fiorma. Nassau Co Fernandina, Oct. 29, 1890, A. S. Hitchcock (M). Duval Co Jacksonville, — Curtiss oat St. Johns Co.: streets in St Augustine, Curtis 6 M, NY). Volusia Co.: Green Mound hell-midden) 5 miles south of Daytona Beach, J. K Small, J Small & DeWinkeler 10568 (NY). Monroe Co.: Gaunt Key, Der Tortugas, Lansing 2502 Gk Lee Co.: shell mounds, Marco, A. S. Hitchcock 318 (G, M, NY). Leon Co.: sidewalk on Tennessee St., Tallahassee, July 13, 1926, Harper (NY). Gulf Co.: Apalachicola, “Herb. Chapman” M, NY). Axasama. Jefferson Co.: Birmingham, in garden on 7th St. S., H. E. Wheeler 1209 (G, NY). Montgomery Co.: on slag ballast on L. & N. ee about 2 miles southwest of Montgomery, Harper 30 (G, NY, Ph). Mobile Co.: Mobile, Dukes 11 (G). Mussovrt. ae Co.: Courtney, Bush 86 30 (NY). Lovrstana. Rapides Co.: near Alexandria in on eo ground, not common, Ball 637 (G, M, NY). Orleans Co.: streets of New Orleans, Hale (M). Terrebonne Co.: Houma, (NY). rney ; Sept. 1, 1933, Hapeman (NY). Harlan Co.: Orleans, Bates 3633 NY). Oxtanoma: Kay Co.: Tonkawa, G. W. Stevens 1864 (G). Woods Co.: near Fairvalley, G. W. Stevens 1640 (G). Blaine 268 Rhodora [JUNE Co.: 2 r Canton, G. W. Stevens 833 (G, M). Rogers © Verdier, Bush 537 (G, NY). Oklahoma Co.: Mee a north Canadian River bottom, Demaree 13252 oH de 18904 ray Co.: Turner Falls, Arbuckle Mountains, ! pie (G, M) (NY). Comanche Co.: near Fort Sill, J. Clemens 1 . Ne rot Texas. Anderson Co.: Palestine, Cory yar feb: mG M). in clayey soil near peng Nat a Si peel r 40.3 ntonio, Wilkinson ; 4 ae Go: Matagorda Bay, Robbins 10 (M). Preece Harlingen, Tracy 9123 (G, M, NY). Colorado ae oe common along river, Bush 431 (G, M, NY). Coma 7 4 6 a Co.: San Marcos, E. J. Palmer 12096 (M). Hidalgo Ca Tracy 9124 (G, M, NY). Lampasas Co.: Lampasss; a 1892, Plank (NY). Llano Co.: Llano, Aug., ee to, MEd (M). Parker Co.: Weatherford, Tracy 7858 (G, M, Mo, Aug. 1 tart Red River Co.: Clarksville, Sept, 25, 1804, Plank (NY). St0t Co.: Rio Grande City, Neally 222 (US). Tarrant, Co.: (a Worth, Tracy 8167 (G, M, Mo, NY). Taylor Co.: : Ozona on & Molby 5508 (O). Valverde Co.: 48 miles south of Mi Comiatoek Road co & Duncan 3013 (M, NY). : abe mo Laredo, July 26, 1882, Letterman (M). Wilson oe sate 1196 Springs, 25 miles southeast of San Antonio, Ed. Pa Peebles G, . Arizona. Pinal Co.: 8 miles west of Flore ae y & Kearney 264 (A). Pima Co.: Tucson, Sept. 30, 1892, Merit . Santa Cruz Co.: Santa Cruz River at La Noria, 1168 (US). This species is native in the New World where it is widely distributed. It is introduced generally in the Old World. a Several additional synonyms are cited by Boissier and aa lung. All these are, of course, later than Euphorbia Chamae " and are extra-limital. Since no authentic material of any s these supposed synonyms has been available no opinion 45 their identity will be offered. follows: Small separated Chamaesyce malaca from C. prostrata as 10 “Capsules pubescent along the angles..................... 35. ©. Capeules pcooent al ofa: ately as bolo the aids’ 38. ©. ale The capsules of C. prostrata are glabrate on the sides of he: carpels. Those with slightly more persistent hairs on the st ee ae eee eee ep we 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 269 of the carpels constitute C. malaca. The character does not hold. Small described the leaves of C. malaca as ‘1-1.5 ¢ long’’, and those of C. prostrata as ‘4-6 mm. long”. The leaf- length given for C. malaca was evidently taken from a juvenile specimen of EHuphorbia humistrata collected at Verdigris, Okla- homa (Indian Territory), B. F. Bush 537 (NY) which has the leaves as long as ca. 1.5 em. But even the type of Chamaesyce malaca has leaves of mostly average size for what Small called C. prostrata. The placing in this paper of EZ. prostrata Aiton in the synonymy of E. Chamaesyce needs comment. There has been no opportuni- ty to examine any authentic material of the former. Aiton’s description is so vague as to be applicable to more than one species. As a consequence of these circumstances there is no assurance that EH. prostrata is here correctly interpreted. In ab- sence of any basic evidence, it has seemed wise to continue the customary application of the name £. prostrata. Customarily the name E. Chamaesyce is applied to an entity native in the Mediterranean region and eastward into Asia. Consequently it had not occurred to me to consider the possi- bility that the name might properly belong to a plant native in the New World. There seems to be only one author, Swartz, Obs. Bot. Pl. Ind. Occ., 196. 1791, who has applied the name to any New World plant and this was ignored as a mere casual misapplication. However, while browsing through the speci- mens at Field Museum, Chicago, in Sept., 1939, I came upon a photograph of the specimen in the Linnaean herbarium and fragments from the same specimen. The photograph shows that the sheet upon which the specimen is mounted bears the cus- tomary abbreviation used by Linnaeus to indicate specimens collected by Patrick Browne, hence the specimen was collected in Jamaica. Jackson’s Index to the Linnean Herbarium shows that the specimen was in Linnaeus’ herbarium in 1753. The photograph and fragments had been filed at Field Museum in their proper place under Ind. Oce. in the folder of E. prostrata Aiton! There is no doubt that E. Chamaesyce, as represented in the Linnaean herbarium, is identical with Z. prostrata as usually interpreted. This is not a new discovery, for, as I discovered in January, 1940, Thellung, in Ascherson & Graebner, Syn. 270 3 Rhodora [JUNE Mitteleur. Fl. 7: 477. 1917, notes in the synonymy of E. prostrata: ‘““E. Chamaesyce P. Browne in Linné’s Herbar (von Jamaica) nach B. Daydon Jackson briefl.”” Thellung then enumerates a few authors who have applied the name EH. Chamaesyce in the sense of Linnacus’ herbarium to the plant naturalized in the Mediterranean region. One of the fundamental rules to be followed in the typification of a Linnaean species is that the Linnaean specimen is to be taken as type, except in extraordinary cases, if he had it in his herbarium at the time the species in question was published and if it conforms to his diagnosis. Exceptional cases are rare an require a full and elaborate explanation to establish acceptable grounds for refusing to accept the Linnaean specimen as type The force of custom is shown by the careful and extended ex- planation deemed necessary by Merrill, Bull. Torr. Bot. Club 60: 633-638. 1933, to justify his rejection of the Linnaean spec men as type of Poa malabarica L. In order to determine whether there were any possibility that Linnaeus casually applied the name E. Chamaesyce to a plant received perhaps long after the concept had been formulated, his Hortus Cliffortianus was @* amined, since it was among the earlier references cited in the Species Plantarum. In Hort. Cliff., 198. 1737, the species appears as follows: uniformibus, ramis alternis, floribus solitariis. Several synonyms were cited, then: ‘ “12. Euphorbia inermis, foliis oppositis oblique cordatis serrulatis ” ‘Crescit in agris & vineis aridis & arenosis Siciliae, Italiae, Galliet Narbonensis & Jamaicae. Corymbus in hac nullus; folia om" uniformia, aequalia; flores ex alis ubique, albi, tetrapetalt; - a alternatim ramosissimi. Folia in Europaea magis orbiculata, Americana oblonga & saepius macula fusca in medio nolait, caulis quam arctissime terrae appressus,”’ _ Since Jamaica and the plant of America are specifically me? tioned in 1737, it was no casual chance whereby the name was applied to a New World plant in Linnaeus’ herbarium. Conse quently Euphorbia Chamaesyce must be typified by the Linnaea? specimen. This is the plant commonly known as E. prostrata Aiton, which is native of the New World but now widely int® duced in tropical regions. The numerous specimens in various herbaria which I have named E. prostrata Aiton must now called E. Chamaesyce L. | | a a a eb eh el as 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 271 Thellung, in Ascherson & Graebner, Syn. Mitteleur. Fl. 7: 455 footnote 2, 1917, stated “Das Original der E. Chamaesyce (vielleicht von Loefling aus Spanien stammend?) gehért gleichfalls zur Abart canescens [of HE. Chamaesyce] (B. Daydon Jackson briefl.)”. I do not understand what Jackson could have been taking as type unless there were two specimens in the Linnaean Herbarium. 49. EUPHORBIA LAREDANA Millsp., Pittonia 2: 88. 1890. Typr: Sandy plains, Laredo, Webb County, Texas, July 28, 1888, C. G. Pringle 2074 (F 1 197395!; photographs G!, W!; 1soTyPE NY!).— She ta cig laredana (Millsp.) Small, Fl. Se. e S., 709, 1333. 3 Annual, more or less pilose-tomentose throughout; stems several to many, 10-20 cm. long, stout (1.5 mm. diam.) toward the base, tubowiadel to 1 em. long; leaf-blades extremely oblique even to the small ee a ovate ah elliptic-oblong (distorted by the obliquity) 3 long, margin entire or the large juvenile leaves with a fay low discrete teeth, petioles less than 1 mm. long; i ea minute and obscured by the tomentum; peduncles ca. 0.6 mm. long; cyathia solitary at the nodes; in- volucres tomentose ae oe ae us within except above, short- campanulate, cuneate to the e peduncle, 1 mm. or less in diam., green-veined eat ae the lobes; fobes deltoid, equaling or rast ng, 0.2-0.3 and with long hairs; bracteoles reduced to one radial appendage opposite each gland, adnate to the evolulre for half its length, entire to one- or two-parted to the rate hairy gnc staminate flowers 3-5; androphores glabrous, 1 mm. long, i clude ; gyno- phore tomentose above, glabrous below, baci and reflexed at maturity; ovary sharply thropaheled, tomentose, styles bifid, less than 0.2 mm. long; capsule glabrate on the sides of the carpels, the backs usually persistently cep sharply three- angled, 1.3-1.5 mm. long, 1.5 mm. diam.; seeds sharply quad- rangular, 1.1-1.2 mm. long, 0.6 mm. cadinty and tangentially, ovate to narrowly ovate radially, base concave-truncate, facets concave, with subregular, round, transverse ridges including the angles, ly of ee and angles frosted white, brown be- tween.—PLaTE 662A Sandy pie in Webb, Starr and Hidalgo Counties, Texas (Map 30). Representative specimens seen: Texas. Webb Co.: Laredo, Pringle 3747 (C, F, G, NY, O, P). Starr Co.: 5 n north of Rio Grande City, Chie 1 363 (NY). Webb Co.: Lar Aug., 1899, Mackenzie 50 (NY). Hidalgo Co.: Sullivan so 272 Rhodora [JUNE Runyon 2541 (W). For citation of additional specimens see Bull. Torr. Bot. Club 63: 431. 1936. This species is very closely related to E. Chamaesyce. The following specimen, though rather intermediate, is referred to . Chamaesyce: Laredo, Webb Co. , Texas, July 26, 1882, Geo. W. Lettermann (M). ExcLupEp AnD Dovusrrut Species Anisophyllum dentatum (Michaux) Haw., Syn. Pl. Suce., 162. 1812, is member of Euphorbia su bgenus Poinsettia. thout Aplarina prostrata Raf., New Fl. N. Amer. 4: 99. 1838, proposed wi . reference to Euphorbia prostrata Aiton. Not recognizable. Said to gro in Pennsylvania. the plarina? re Raf., op. i Sasce Not recognized from description. From Louisiana ‘and Tex Club 40: peices exstipulata. (Engelm.) Rydb Bull. Torr. Bot. tion 33. 1913, is a member of Euphorbia subgenus Agaloma (including see ae gt m). n Tex iephonbia one Engelm. ex Wiese, Bot. Gaz. 24: 51. 1897, im nde Lois for E. humistrata Engelm d var. uphorbia lat ed L. 63 prostrata, var. multiflora, an simplex Raf., Med. Fi. U.S. 1: “Se 1828, nomina nuda uisiana Euphorbia’ ludoviciana Featherman, Ann. me ee Surv t Gray 2: 71. 1871, not Raf., Fl. beduvie.’ 111. Si7, An 1 sotype, a iensis Sper, of this species, shows it to he Phyllanthus se Euphorbia nen Millsp. ex -o-ate Host Index Fungi N. Amer» 462. 1929, te) 18 Peplis (L.) Scop., Fl. Carn. ed. 2, 1: 340. 1772. 5 ae Pepli (L.) Haw w., Syn. Pl. Suee. 159. 1812 ~Chamaesye Peplis (L.) P. rokhey? Consp. ide Tithym. Asiae Mediae, 15, 1933. 9: 10. Euphorbia oreo [sic] Benth. ex Frick, Euphorbia Review eedy 1936, ‘with brief ription, is evidently a garbled account of some aa species of Pups inadvertently associated with the irrelevant nel the remocarpus setigerus ome, espe to Frick in a letter, account was drawn fro newspaper ri Jorida. Xamesike >t Raf. , Aut. Bot. "98. P7840: not pee Flo imen aris according to his notes shown me by ‘dual of Tonac glauca Raf., 0 cit., 97. Sige a prostrate individu Euphorbia m maculata, Ken ae ucky, Virgi thee: "Published Jan. 18, 1938 according to Cory in letter filed at Gray Herbarium. 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 273 Xamesike gracilis (Elliott) Raf., op. cit. 98; based on Euphorbia gracilis opis Belongs to E. subgenus Agaloma (including section Tithyma- opsis Xamesike hypericifolia Raf. var. sanguinolenta Raf., 1. ¢. Probably equals Euphorbia maculata L. Pennsylvania. — hypericifolia Raf. var. parviflora Raf.,1.c. Identity uncertain. Ae micranthes Raf., 1. ec. Identity uncertain. Louisiana to Xamesike microphyla (Raf.) Raf., 1. ¢.; based on Aplarina? microphyla e x amesike prostrata (Raf.) Raf., op. cit., 99; based on A plarina prostrata af. Gandoger, Fl. Europae 20: 142. 1890 has described and given binomials under Euphorbia sect. Anisophyllum to three sup- posedly new entities from the United States (Pennsylvania, Nebraska, and Missouri). In addition he has named (as bi- nominals) and described a total of 10 segregates of 2 species native in the United States but introduced into Europe. It seems best to continue the custom of ignoring all names published in this work of Gandoger’s. The technical nomenclatural justifi- cation for this action at least in the case of Zuphorbia is that the first members of Gandoger’s binomials under this genus were the names of the various subdivisions, in the case pertinent here, Ani- sophyllum. Also, the structure of the work indicates that the new binomials were subdivisions of species which form of nomen- clature is inadmissible according to the International Rules of Botanical Nomenclature, Article 28. List oF ExsiccATAE The collectors’ numbers are printed in italics; unnumbered collections are indicated by a dash. The numbers in parentheses are those of species and varieties treated in this revision. Abrams, LeRoy. 1381 (25); 2661 Antisell, 7 = (48). (35a); 2952 (35b); 2990 (35a); 3187 Armer, Annie A. 5390 (30). (26b); 3789 (35b); 4097 (37); ed pia is G. 18987 (35a). Pie 3); 4178, 4652 (35a); §341 (3a). Arthur, J.C. 79 (13). ms, LeRoy, « and E. A. McGregor. Ashcroft, G. B. — (18). Allen, J. A. — (36, 45). — (35a, 36); 195 (35a); Allison, A. 148 (13). "43 ea), 1974 (35a). Anderson, A. P. 777 (45); 1587 (13). C. F., F. S. Earle pee cae L. E. ee (13). SS 5 5; (36) 6 (47); 608 oe Andrews, — (14, 45). Baker, H. P. — (7b). 274 ake e: a — (35a). all, C 405 (18); 637 (48); 959, 1257 6); 4950 (13); el td (30). Ballard, C. A. — (45 ys 1 1 (86); 526 (45); 527, 2604 (36); O08, 2734, 2774, 2902 (85a); 3051 (36). M.F Raioloniee. Elam. — (47). Bartlett, G. _— Bartlett, H. H. 1055. 9 (2 Sup Bartram, E pede (22a); 1793 (1); 3058 (4 assett, 13). Batchelder, C. F. (45) ates, G. Sg 298. Bates, J. M. — (82a, 47); 3633 (48); 5025 (45). Beal, Mary. —_ woe 8). Beals, Mrs. Ida —— (18, 31). Bean, R 1S Bean, R a *é. W. Bird & C, H. Knowlton. — (1 ». Bechtel, es R. 8409 (45). mire R. J., and C. P. Starr. 385 a shina H.C. 5034 (1 8). Benner, W. M. 5418 (45); 7008 (13); 7335 (45 H. F. 690 (36); 1835 (18, 20). 4084, 403b, 403d igh, 407 (45); 3838a a, 3838b, 3838c, 3838d (i); 38444 (12); 38480 oy 5895b (29); 6415 (9). ee me TAD: 528 (1); 629 Bissell, H. Linder. 21780 i; et v5P ys aoe 5 (4, 5); 2033 (14); 3158 (38); 5651 vy 7718 (12), Blanchar Blankinhi, J W. “116 (36). . A. — (36). maghte J "E, 848 (14); 849 (1); Blomquist, H. L. — (1 3). Blu — z C. 27 (35a); 1690 (47); Bodin, J. of — (7a, 45). E. &., and Molby. — Rhodora [JUNE Bracelin, Mrs. H. P. 4564 (85a). Brandegee, Katharine. de (4 + Brandegee, 22a, 35a, 35b, '39);, sé say? ag 2 13). (45); 157383, eras, 1 918 ( Butters, F. K. — (13, 15). —, 228 (45). Chandler, B. F. = (13 Cancer H. P. 7026 (31). i ee 4596, att (I): Chase, Agnes Chase, V H. — (13); 205 (1); 6% (45). pe sir V. K., & E. R. Drew. 5a). 10039 (36). Chiitieen, 1 oo (1 4, 45); ; 545 (1). Clark, June ri 144 (35a Clary, Marjorie D, 654 (26b). Clemens. — (35a). Lig (48). 1 Frais igs 2 (9); Clem E. ; 276, "2587 5 a1? 7 (80); al 3) ents Cleveland, D. —, 904 (35b). Clevenger, Cc. H. — (18). Clevenger, J ¥. — 08): Clifton, R.L. 3183 (32a). Glokey, I. W. 7183, 7184 778 (32a). 1941] Wheeler,—Euphorbia Subgenus Chamaesyce 275 Clov: 356 (5); 783 (15); Cushman, J. A. — (1 ei 1363 a; 1364 (30); 1476, 1524, Cuthbert, A. — (13); 7 1). 153 Cutler, H. C. 664 (éo): "vas (26c); clits, W. x 9 (32); ot A aa 257 (1). 2286 (4). ocks, R. 8. 3626 Coffman, Coral Fleenor — (45). Daniels, F. P. — (45). Collins, F. 8. 1, 48); 981, 1176 we Wm. — (13). 2 .N., T. H. Kearney & J. H. Kempton. 145 2a). . L. Fernald. — Collins, J. F., 14). Collom, Mrs. Rose E. — (89); 33 (25); 314 (10). Commerson. 238 ree Commons, A. — (13, 45) Congdon, J. W. 75 (35b). Conklin, J. }e ai he nue & C.D. Jacobs. 1418 Cook, H. C. (13); Cooper, H. L. 111 (36) be ory, . 608 907 (17); 1914, 1915 (18); 1919, 1921 (20); 1922 (4); 1926 (31); 1927 ) (28 O (33); 1964 (32b); 196 (43); 2273 (18); 3188 (19); 3191 26c); 26701 (19); 27315 (36); 27404 (4 oues, Elliott, & Ed. Palmer. 264 (25); 531 1 (10). ulter, J.M. — (35a) Coville, F. V ee age Coville, F. unston. 735 (31); 1008 ie), 10st (3a). Cowen, J.H 6 (32a). peti H. BR. — (81). t, Mary B. 168 (48); ~~ (31). Crone, F, P. 6 ert 45, 48); 2 (20); 2 6b, 496 (16b); 5396 Bae ee 39849 (16b); es a; 56426 Big! $0670 (12). Davi Dave John. 965 (13); 1419 (45). avy, J. B. a). 45); 6 , 24270, 24344 ee (46); 2 (13); 29566, 29862 (45); saa8s (18): 32212 (45); 32533 (14); 32671 (45); 33357 (30); Pa (45); 37495, 37598 (1 “a se06 (45); 55400, 56707, 57370 Deam s C. 2758 (12); porte sat 27 Tab 28949A (36). Deane 26a). Sean Walter. - — (1, 45). sda 36). 7625 (31); 8202 (31); (45); ¢ hes. Gs); 8328 (4 6); 8778 9582 pe 9736 (46); 15736, 6041, 16 6057, 16298, 16331, 13180 (45); 18410, 1857. ( — (13, 26a, 30, 31); , W. M. — (2). Dosa, B. A 335 (29); 376 (30). — (1, 14); 7 (13); 83 n, Victor. 31 78 (26b); $205 31). oe. D.L. — (18). 276 Eames, E. H. — (1, 13, 45); 45 (1); 54 (14). Earle, F. 8. — (35a); 290 (47); 592 20 E Sey ‘ Earle. 35 (45); 36 ‘6 is (G50) 279 pets 308 We) pate 09 eke &S.M i tec 104, 381 Eastwood, Alice. — (35b); 137 (4); O66 (26a); 273 (25); 8668 pb Eastwood, Hs & How ell. (28); 6310 (31); 6473 (36); ried (82a); 6893 (31 Eggert, eee (7b, 20, 30, 35a, 36, 45, 46, 47, 48). eston, W. W. 5 (14); 5246 ; 10274 (350). 10725 (12); 15951 (43); 15967 (25); 17352 (47); 19370 (42); 20227 (31); 20424 (47). Ehren .C. 4 Ellis, Charlotte ©. 335 (9). Imer, A E. 200 (35a); 602 (35a, > Emig, W. B H. 6265 (7b); 527 (30). Enge — Geo == (86, 46); 1139 Eagelman, Henry. — (2 ) Evermann, B. W. Ewan, 3: "sve (3 1). Farwell, Oliver A. — (35a); 235 — 8756 Peony t, N.C. 3418 (45); 10258 (36); SS 028, a5i4 (as); 18156 (1). Wilson. 1088: (45); 14 4674 (1 ). Fendler, A. 789 (9); 795 (35a); 796 (42 yi, 797 , 798 (47). Fern M. L. (1, 45); 6. (be gf 0, 1981 | (36); "66s (4); 2733 wb; 2736 (45); 2737 (14); Fernald, M. L., E. B. Bartram & B. Long. 24085 (14). Fernald, M. L., C. H. Be ag C. B. raves, B. Long & D. H. Lin der. 21756 (14). ernald, M. L., F. K. Butters & H. St. John. 15261 (1). Fernald, M. L., F. W. Hunnewell & B. Long. 9818 (1). ernald, M. L., & H. B. Jackson a4 © » Ferna 5350 (1); org iy Se Oss ‘9748 (45): 9817, 9821, 12703, 12704, 19705 Rhodora [JUNE pie 5); 12708, 12709 (46); 13994 ’ vine ie (14); 13997, 18998 (45); 17638, 18684 a Foe! oes Fernald, Bares, 13995 Bc 13999 (45) Fernald, & H John. re tee. re paste (1). Fernald, Bim Bike ong & G. Torre Fe mas M. L., & A. S. Pease. 3407 36 oe M.L., & H. St. John. 11114 B. E. — (26). RS. 8697 aie -_ i - Ferris, R. = (12); 3012 (13); TsO 3 ashe ate RB eI P. Rossbach. H. a & (2 ) b 962 a Ra “ y i M Scott & R. is Gi - ace 45 . Fisher, G. i He, (25); 74 (81); 607 ff ea mer , Jr. 2869 (45); 2870 (1); "5934 (13); 7753 At, a (1). Pate, ee es 83785 (4); 15754 nti 15908 Sas ol a 8 (45); 5499, 5517 Prehlin, 6014 (3): 6280 (12); 6348 47 (36). =. (1, 14, 45). es, ul O. Furbizh, Kate. -E. 576 (36). Gay a 1119, 1148 5 C0) 1608 mK $062, 6 395 (4 4). eyer, C. A. 493 (45). Gillespie, J ue 8418 (268); ae : ( Gillman, Henry. — (1). Gleason’ H. Allan. 1932 (45). Fy eenay H. A., & F. D. Shobe. 336 (1). aioe, K. Es art ». 1168 2 (228). oe Urb); 267 - 519 Gea), “64 (31); 708 (26a); 723 1941] (31); 759 (4); in (4); 2222 (26b); 2289 (82a); 2 1 Gl 4 (35a); 2179 20: 197 (7b); ‘sale (32a). oan Y B , 36). Graves, C. B., B. Long & D. H Linder. #1704 ( (1). Greene, E. L. — (35a, 36, 39); 266 (28). Greenman, J. M. 8 (45). Gregg, J. — (45); ie 5 (23); 4657 (88). Griffith & Schlosse. 36). io David. 181? (48); 6172 (10). Griffiths, D., & L. W. Carter. — Griffiths, David, & J. J. Thornber. Grimes, E. J. ig (13); 4640 (48). Grinnell, J. 27). ect C. A. ‘8 (1). Groth 15 (80); 55 (7a); 131 ‘(19); 208 (13). Haberer, J. V. Nett Hale, Josiah. AG A 8). Hall, Elihu. — (80, 45, 46); 547 (29); nfo tle (41a); 569 (1 9). H. P. Chandler. 345 88 (45). ae wr Hansen, Her A210 (4); 552, ss (31); Tt *e60); 744 (48). Harger, i. 8 — mae ) 5519 (4 5). 48); 30 (48); 10 (46): 380 (45); 574 (29); 736 (1); 1520 (30); $246 (1 3); 3248 (30). Harris, J. A. 4 (3e); C17205 (5). Harris, s Sidney, | — (1, 14). G. 1938 (22a); 3563 a 6617 (35a); 8142 (8); 8224 5682. a) sGer Gore 6022, 6036 (16a); 6028 (48); 6085 (4 2); 6114 (47); 6164 (9); 7998 (37); 8037 (8); fo an M4 8291 (9); 8293 (28); Harrison, oe: napeeye & H. J. Fulton. "¥991 (3 Hassler, E. 7735 (16a) Hastin, _— Hatry, O. (13). Wheeler,—Euphorbia Subgenus Chamaesyce 277 Havard, * 16 (36). Hayden, F. V. — (7b, 35a). Hayes, 5 Hayward, oe E. 246 (7b); 569 (36). Headley, F. B. 42 ae Hedgecock, G. G. Sha FE ] . A. — (18, 29, 31); 668 (13); 669 (45); 670 14); 1231 (1 & 5); 1463 (30); 1738 (19); 1804 (45); 1918 (48); 1921 (45); 1922 (13); S508; er oere Boe): 9674 (31); Bea) 8 (35a); 148 (3a); 3) {1300 (7b); 14802 (38). + a. poe (Bia); 1190 ‘ ). a gg F. J. 2248 (1); 3631 (45); 44 (3 6); 9073 (13); 9115 (45); one qd Hervey, E W — (45) ines be ). Hitche A.S. — (13 48); 3 (48 Y; 390 0G ); 3. 4 (12); 326 ea); 328 (45); 4670 (7h): 472 (47); 475 a). Hoar, nS. — Uo. Hodgdon, A. R. 677 7 (13). Hoffmann, i (13, 24, 45); 585, 617 (24). Hollister Mabel P. 81 (13). olm (ee 45). nger, J. M. Holzner, * Ba 157. a (31). F. 1277, 2583, 3722 (34). 1 (1); 1481 (13); 2583 (45); 2 835 (13); 10618 (14); idea 1997 (1); “ 3024 (36 ). 2 a): 3 (24); “6681 1 (28). asad 7 mc grit vt « T.J. 787 (45). Hoy t, Mrs. — (25). Haar, A. a ; Hughes, Mrs. Ww. R. 22 (12). Hunnewell, F. (14); 2827 (45); 6836, 6977 (14); 7365, 8081 oe C.F. — (18). R. H. 6 (47); 46 (13); 63 Ine}; 6 5 (36). Ingalls. — (5). €s, 278 Jaeger, E.C. — Brg 6b). Jepson, W. L. 7 (3a); 1374 (26b); 1650 (35b); 508 (26b); 6159 (3a); sie (35a); 6879 (27); ete (3¢e); . 5 . Q “ > ® ft w 2 ww — ce 00 10643 (35a); 11734 (31); 1237 (35a); 12475 (26a); 13924 (35b); 1892, 13 (3a); 13934 (35a); 279a (3a); 15829 (31); 15964 (35b); 15985 (35 Jermy, G. — (7a, 12) Johnson, Frank W. 113. Johnston, I. M. 105 . '; 3331, 3641 (22a); 4136, 4173 (17); 7028 (18); (33); 7061, 7123 (47) 7135 (33); 7183, 7196 (42); 721 79 i og 7991 (31); 7995 (49)? ae nes, G. N. Jones, M. E. 7, 39 ; 292 (7b 786 (47); fbi ORE pon a 3201 ( (32a); te 17 8 (85a); 4195 2 ( ;. ity (13); ays ; (56033 rated (10); 24877, 24878 O01 (31); 26013, 26015, oe W. W. C287 (9); 288 (35a). Joor, J. F. — (45) Juett, Ji << — sy: Kearney, T. H. 156 (45): “| 2108 (13) (45); 1812 (1); earney, — (26a); 9 =e (39); Flag an ( (16a); 10215 Beat, 9029 ea John H. 1845 (45 Kelly, Junea, — (31 9 si Kelsey, F. D. ~ T244 2); 449 G3); v7 Gay 1358 (26c); 1464 (28 28). Stabler, L. M. St tandley, Jeannette 1 307 (12). Standley, P. 9 (47); 5054 (35a x 1053 (i2)."7 7371 (36); 7499 (35a); 12656 287 (18); 40295 ap; ‘Stee (7b): oe (42). Standley C., & H. C. Bollma: 12243. as) Stanfield, S.W. — (7 a, 19, 29, 41a). tearns, Elmer. — (4); 245 (39). Steele, E. 8. & Mrs. 37 3). Steele, E. S., & ee: "Mary R. ao: Stevens, G. W. 308, 350 (32a); 622 Rhodora [JUNE 33 (48); 834 (80); 835 (47); eh (ab): Oks {s6); “0s (31); 1834.1 (45); 1602 (47); 1603 (45); 46); 2307 (45); 230. : (7a): 2584 (48). 2960 (7a); 3277 6). Stevens, O. A. 290, or s, W. - 16 ae ee (46). 46); 24316 (30); ae Frederick se 373 Gr Stone, Witmer. — (45, 13); 107 tratton, R. 206 (7a); 441 (20); 48 saya ); 459 (47); ae {7 rade she B on ora aig W.N. 86 (36); 489 (36); 12327 (13). Sullivant, W. (1). Svenson, H. kK 187 7 (a8) 139 (13). Swingle, D. B. — (35a). Taylor, B.C. 784, 1568 (36). Thar nA, ‘38 ie & te , 1180, 1416 an Tha aah) ¢ Gan aaey (42); 4427 (30); es Tb). Th .C.H. 2 (36); 60 ( Thompson, J. W. 3749 (36); 11122, ecbet 3 3, 36, 45, 40) Ji4 Hf 47 (42); 136 (10); 209 (12); ¥ es at (37); 5 (40; ier (18, oe 98 (31); "ST ran; ; 740 1); ~~ (43). Esta i, (po a (22s); 6902 (13); 6 Tolstead, W. f. 4 (7b); 621, 808 (36). Ed Ay Ji <7 pce 31, 48); ai (22a); has (42); 263 (26a); 35a rome C. H., & C. M. Barber 284, (9). a sae bP pee (al au 2908 (¢ 10 (29): 2913 (45); 3125 (30); ‘47 4V17 ‘By 8370 (29); 6376 ©) 1941] 7429, 7432 (13); 7435 (30); 7436 (45); ae ran 7844 (30); 1852 ; 8124 (7a i} 1 48): 8168 (7a); 8169 an 5170 (30); 84774 (45); 8953, 8987 (17); 9018 (13); 9115 Nn 6b); 2183, 9124 (48); 9125 (31); Trae & Earle. 2882 (5). y; $ M., Tracy & Lloyd. 9 ; Train, Perey. 676 (31); 1876 (4). Trelease, William. — (36, 45); 873, 874, 877, 878, 880, 1178 (138). Trueman 1), EN soa: | G. H. 66 (36). eedy, Frank. 258 (19); 268 (31); "368 (36); 5057 (35a); 5058 (36). 1917 (1); 2313 Umbach, L. M. —, (36); 2315 (45). Underwood, L. , & A. D. Selby. 156 (35a); 240 (36). — (i, 0, 26b, ies ey 1); ths (bb). Foe (8a). Visher, 8.58 2 (36); 2359 (35a); 4108 (1 3). Wadmond, S. C. 2948 (45). Waldro n, C, H. Walker, Harriet . 910 (35a). Wallac . J. — (7b, 30, 36); 121 (36). Wann, F. B. 2 Ward, L. F. (13, 30). Ware, Robt. A. 4138 (45). R. A., R. W. Woodward & E. Har; — (45). . é ge Wamock, B. 5 (26c); T (41a); C506 (ai), ig (47); Cei6 tobe)! a 657 (39); 665 (31); 985 Watson, s “8 Waugh, Weatherby, C. A. & U. te ,&J Weatherby, ©. en Coline, Webb, RJ. 469 (45); 1436 (1). Webber oH ge. (sy 1 (36); 11 heeler : 8, 14) Wheeler, [G. M.?] — (36). Wheeler, H. E. 140 (45); 1207 (13); 209 ‘48). Wheeler, J. A (13). Wheeler,—Euphorbia Subgenus Chamaesyce 283 Wheeler, C. 13380, 3652 (45); 4041 (3b waren! W. A. 836 (13); 430 (45); 434 (36 Wheeler, W. H. 296 (45). Whetzel, H. M. 12388 (14). White, 296 White, Mark. White Stephen. 1145 (45 Wiegend, K. M., B. Magu: Richards, Jr. & T. Monier woo 10778 (7). Wiegand, K. M., 1801 (29); 1814, fe ‘oo 1817 ). Wiegand, K. M., & G. B. Upton. 3748 ie Wiggins 57 (25); 4364 (22a); ( 5915 “2i); ‘5367 (22a); 6124 (10); 6125, 6311 22a, iggi “D. Demaree. 5020 6; ae ” 10761 wine Wiggins, e L., & J. W. Gillespie. 4141 Wilcox, i” E. 320 (12); 328 (35a). Wilkinson, E.H. 62 (48); 56 (18). Williams, E. F. — Williams, R.S. 160 (30 0). Williams, T. A. — (7b, 13, 30, 35a). Williamson, Cc. 8. (29, 30, Ly? 241 Williamson, ar aa J: 242 (13). Wilson, Guy. 27 (13). Wislizenus, A * 4 (7b). Wislizenus, F. 377 (13). Wolf, C B. 2300 ines 2321 (35a). lf, _ Are ior Wooton, E. O. 2s 30, “es 42, a "5 taiy: $48 @ 540 (4); 42 5, 608 (35a a); sind (12) G. Stan ey. —_— =) 3273 (47); 3617 (a. a, 41b); Wright 186 7); a6, 657 (18); 68 (42); Wat Bs & a ae a er . B. dey & C. B. 3); 77 (47); m oO oO c= — i) ~s fe") a i] Lat < ~ fo) 3 a) © a, = i=} a @ B. st E ° s = a 5 leaves. NDLERI T. & G. var. CHAETOCALYX Boiss., from Texas, Cory 1964 (G): 1, branch; 2, glands and appendages from above Fen TRILIGULATA L er, from Texas, Moore & f : D. E. rracuysperma Engelm., figs. rom ‘erry te Palmer 20 in 1869 (US 58604); figs. 6-8 from Arizona, LATE 668. A. EvpHorsia Cuamarsyce L, Photograph of Tyee from Jamaica, P. Browne (Linnaean Herb.). B. E. MacuLata L. Photograph of TYPE, source and collector unknown (Linnaean Herb.)—Photographs courtesy of Section of Photography, Field Museum of Natural History. — rneemeemaee et: A elRE C Rhodora Plate 654 A Cuphorhia polygonifolia 8 a CG. ammannioides iia m 4 Bok a SC tia Sie “Zs = aay om ae oy aN So ; S alee | fh Of .. \ Me: yn As \ co, fF x - J ANS f .. 4 y SO ees as . Q od Rr f RS fj BL oe ah > os Re x ig : A ey ¥ 4 244 rer A os e A Oy: ms WHEELER on EUPHORBIA Rhodora Plate 655 A Ci uphorbia th srt. fokia 4 Ye. = WHEELER on EUPHORBIA Rhodora Plate 656 A Cuphorbia vermiculata R C. maculata ma hyss opifolia oe ey WHEELER on EUPHORBIA Rhodora é Plate 657 A ele firta . ror f capiteledenw ay | WHEELER on EvpHoRBIA Rhodora Plate 658 A Cuphorbia B GC. Parishit /ia TERN cord: vfo & 4 GW. Dion — WHEELER on E}UPHORBIA Rhodora nce A Euphorbia angusta — GW Dillon — Vill fera Var. ty pica om 4 mm WHEELER on EupHorsia Rhodora Plate 660 A Cuph orbia theriaca ws esyce a ert N ~—e Ty i gt @ 1 ~ a WHEELER on EUPHORBIA €. glyptosperma mS D € Ahramsiana NG Rhodora Plate 661 A C uphorbia ‘g ! setiloba —e tee. os < 1/2. 12. Same, sepal « 1. . Same, petal 1. . TILLaNpsiA Pentanpi L. B. Smith(Penland & Summers 1138), branch of inflorescence >< 1/2. 15. Same, flower 2 16. TiLLanpsia Pentanpu L. B. Smith var. eeouncuratTa L. B- Smith (Daniel & Tomas 1558), branch of infloresceace « 1/2. 17. Same, flower and bract Dae 18. VRIESIA PACHYCHLAMYS (Bak. ) Mez (Jenman 2044), leaf and inflorescence < 1/5. 19. Same, sepal «& 1 20. Vrinsra spr (Bak.) Mez (Foster 111), leaf and inflo- rescence X 1 - Same, sepal ¢ a - Vriesta CLAussENIANA Mez (Claussen s. n.), inflorescence and leaf << 1/8. eo se sh eS = a) bd bed > oo nN fh oe L. B. Smiru, Studies in the « Bromeliaceae » PLATE I Fig. 1. fr) Prats II VRIESIA GLADIOLIFLORA (Wendl.) Ant. (after Antoine), leaf- blade and inflorescence >< 1/5. . Same, flower « 1/2. . VRIESIA LANCIFOLIA (Bak.) L. B. Smith (Blanchet 3458), scape and inflorescence 1/5. . Same, leaf & 1/5. : . VRIESIA VIRIDIFLORA (Regel) Wittm. ex Mez (after Antoine), leaf-blade and inflorescence < 1/5. . VRIESIA sUBSECUNDA Wittm. (Lehmann 1775), apex of leaf- blade and inflorescence >< 1/2. . Same, sepal « 1. . Same, petal < 1. - VRIESIA GUADELUPENSIS (Bak.) Mez (Duss 3321), apex of leat- blade and inflorescence < 1/5. . Same, sepal 1. : . VRIESIA LONGISCAPA Ule (Smith & Brade 2293), leaf and inflorescence < 1/5. . Same, sepal « 1. . Same, petal < 1. - VRIESIA SANGUINOLENTA Cogn. & March. (Dodge, Hunter & Steyermark 16907 ), apex of leaf-blade and lateral spike «1/5. . Same, sepal < 1, ; - Vriesia Pirriert Mez (Tonduz 12229), apex of leaf-blade ant inflorescence 1/5, - Same, sepal < 1. : - Vriusta Tonpuztana L. B. Smith (Tonduz 12349), apex oF leat-blade and inflorescence x 1/5. . Same, sepal d< 1. — 4. B. Smiru, Sludies in the « Bromeliaceae » PraTE II TT Zz ES SSS aX Puate IIT Fig. 1, Vriesta Brassicorpes (Bak.) Mez (Burchell 1898), leaf and inflorescence 1/5. 2. VRIESIA UNILATERALIS (Bak,) Mez (Burchell 3347), leaf < 1/5: 3. Same (Dusén 16663), scape and inflorescence >< 1/5. 4, Same, sepal < 1. 5. Vriesta Racinak L. B. Smith (i. B. d& R&R. Foster 270), leaf sc 1/2; 6. Same, scape and inflorescence < 1/2. : 7. Vriesia OLIGANTHA (Bak.) Mez (Glaziou 15472), plant 1/5. 8. VRIESIA AMAZONICA (Bak.) Mez (Martius s. n.), branch of inflorescence >< 1/2. 9. Vriesia PLarzMannut E. Morr. (after Morren), inflorescence xX 1/2. 10. Same, petal 1. 11. Vriesta rincens (Gris.) Harms (Wright 1518), plant X Ys 12. Same (Wright 1520a), sepal >< 1. 13. Same (Duss 3725), inflorescence < 1/5. 14, Same, sepal 1, L. B. Smiru, Studies in the « Bromeliaceae » Prate II ee FROM THE tak HERBARIUM. HARVARD UNIVERSITY CXXXVIII ‘MONOGRAPHIC STUDY OF ARABIS | IN WESTERN NORTH AMERICA . Reprinted from Ruopora, Vol. 43, July-September, 1941 CON Bipcaiat ng FROM THE GRAY HERBARIUM F HARVARD UNIVERSITY CXXXVIII MONOGRAPHIC STUDY OF ARABIS IN WESTERN NORTH AMERICA REED C. Ro.uins DartEs or IssuE dp od ONE EN RE eee 348-411 eae eae eke ate ee eet ne Paco. my Mae CONTRIBUTIONS FROM THE GRAY HERBARIUM OF HARVARD UNIVERSITY—NO. CXXXVIII A MONOGRAPHIC STUDY OF ARABIS IN WESTERN NORTH AMERICA Reep C. Rouuins In Arabis, one of the larger genera of the Cruciferae, the ex- cessive variability of certain species and the lack of sharply defined diagnostic characters throughout have given it a reputa- tion of being exceedingly complex. The speciation is especially complicated in the Cordilleran region of western North America Which is one of the principal centers of distribution for the genus. Elsewhere, Arabis is represented mostly in the north- temperate regions of the world with special areas of species- Concentration in Europe and Asia. An inadequate apprecia- tion of the importance of Arabis in North America is evident m all general systematic treatments of the Cruciferae. Even in the late pretentious work of O. E. Schulz,! the American species were not sufficiently considered in setting up the sixteen sec- tional divisions. As a result, many of our species are not properly referable to any section. erhaps one reason why American species have been largely ignored by the creators of world-wide systems of plant-classi- fieation is the lack of an inclusive, detailed study of Arabis ’ which the species and their relationships have been carefully and accurately evaluated. Recently, Hopkins? has partly met ' Engler & Harms, Pflanzenf, 17b: 542-547 (1936). * RHovors 39: 63-98, 106-148, 155-186 (1937). 290 Rhodora [Juty the need by giving us a treatment covering eastern and central North America, but the larger task of dealing with the species of the western portion of the North American Continent has not been accomplished. The central object of the present study has been the preparation of a systematic account of Arabis to fill this gap, but natural variation and the phylogenetic relation- ships of the species have also received attention. Unlike many genera of plants with endless modifications of flower-parts, the flowers of Arabis are monotonously uniform. Even the fruits and foliage are alike in some species so that in many instances trivial structures must be elevated to a position of prominence in treating the species and their relationships. Fortunately there are a number of clearly discrete species 1 Arabis which have been available as points of reference in their respective circles of affinity. These species are guide-posts to limits of variation and definitive characters and afford a gen- eral basis for interpreting species of a less clearly defined nature. The practice of constantly referring to natural, clearly defined species for aid in clarifying those of some obscurity has been such a standard procedure in the present investigation that it is expedient to list some of the species used. These are blepharophylla, A. cobrensis, A. Crandallii, A. crucisetosa, A. Cusickii, A. dispar, A. furcata, A. glaucovalvula, A. Koehler, A. Parishii, A. perennans, A. platysperma, A. rectissima, A. Shockleyi and A. suffrutescens. The taxonomic history of Arabis has been carefully worked out and presented by Hopkins. A further discussion of the subject is unnecessary. Except for one species, A. petiolaris, not in cluded by Hopkins, this study is concerned with the native species of Arabis occurring west of the one hundredth mee in North America. It is expected that the present pape? wil be used in conjunction with that of Hopkins so that wherever possible duplication of his work has been avoided. : The present work cannot be considered to be final. Arabis is too complex for that. However, as much data and expe? psa as it has been possible to assemble over a period of half-a-dozen years have been directed toward a solution of the problems 1 volved. Some species are inadequately known because the re- * Op. cit. pp. 63-66. 1941] Rollins —Monographie Study of Arabis 291 gions inhabited by them have not been botanically explored. Others should be studied genetically and cytologically to eluci- date certain problems concerning them. Still others should be cultivated in various ecological situations in order to test their presumed plasticity. These methods of approach are be- yond the possible scope of this investigation, but it is hoped they will eventually be utilized. Eighteen of the fifty-three species included in the present work have been grown continu- ously for two or three years. All flowered except A. hirsuta and A. glabra which grew vigorously for three years without pro- ducing flowering stems. The plant-cultures were grown from seed collected in the wild and have been used for observations on the ontogeny of the species and a certain amount of cytologi- cal investigation. RELATIONSHIPS WITH OTHER GENERA The question as to how far usual or well understood generic limits may be stretched without causing a complete breakdown in established concepts is particularly pertinent in the Cruciferae. In this family, the genera as often constituted are not wholly natural. This is true because of the tendency to include species which claim one or perhaps a few characteristics in common with those obviously belonging with the generic type, but which differ sufficiently to be definitely discordant. Large genera often receive many species in which the relationships are dubious and thus over a period of years may become so heterogeneous as to be almost an absurdity. It is the business of the monogra- pher to evaluate and place anomolous species. In studying Arabis, it has been my policy to check carefully the generic character of each species under consideration before placing it unequivocally in the genus. While doing so, as one might ex- pect in a genus dating from Linnaeus, several species have been found to be quite out of place. In America, the nearest related genus to Arabis is Sibara. In 1896, Greene! proposed the anagram Sibara as a generic name for a number of species with common characters previously Placed by various authors in Arabis, Cardamine, Sisymbrium and Nasturtium. These species seemingly are more closely re- * Pittonia 3: 10 (1896). 292 Rhodora [JuLy lated to Arabis than to any other genus and, indeed, they have a number of characters in common with it. Greene failed to point out reliable differences between Sibara and Arabis and as a result many subsequent writers have not recognized his genus as a valid one. Arabis is related to Sibara through such species as A. lyrata and A. arenosa which belong to the so- called section Cardaminopsis. At least two recent students of the Cruciferae, von Hayek! and O. E. Schulz,? have given generic recognition to Cardaminopsis. I do not concur with them in the opinion that Cardaminopsis is a genus distinct from Arabis, but it must be admitted that A. lyrata and its relatives are not next door to A. alpina, the type of the genus. To add species to Arabis which are only remotely related, if at all, to A. lyrata, the latter being somewhere near the outer boundary of the generic limits, simply overruns the natural definitive lines of the genus. Such a practice carried to its logical conclusion in a family like the Cruciferae would only lead to the establish- ment of another absurd Crucifera as conceived by Krause.* Sibara as a genus rests on the following characters which do not belong or belong only in part to Arabis: pectinate to pin- natifid, somewhat glaucous foliage; similar basal and cauline leaves; petiolate cauline leaves; annual or at very most biennial habit; absence of glandular tissue adjacent to the paired sta- mens; and the lack of an expanded pedicel-summit. The foliage of all the species of Sibara is highly dissected, varying from runcinate-pinnatifid to pectinate with many narrow divisions. This type of foliage is common in Cardamine and Sisymbrium, but is unknown in Arabis proper. The basal rosette is caducous or entirely absent in Sibara, but always the first formed leaves are similar to the cauline ones. In Arabis, a basal rosette 18 usually developed and in all cases the basal leaves are differ- entiated from the cauline. Such recent authors as von Hayek* and Villani,® among others, have stressed the configuration of the nectar-glands on the receptacle as being of fundamental importance in classifying the Cruciferae. It is undoubtedly 1 Fl. Steiermark 1: 477 (1908). 2 Op. cit. p. 1941] ~ Rollins,—Monographie Study of Arabis 293 true that they have overemphasized this character, since more variation in natural genera occurs than they have taken ac- count of, but when used in conjunction with other characteristics, this feature is a useful tool to aid in clarification of generic lines. In Arabis, the glandular tissue of the nectaries is well-developed and surrounds the base of the single stamens, subtends or sur- rounds the base of the paired stamens, except in one small group, and often subtends or partially surrounds the base of each petal. Sibara, on the other hand, has poorly developed nectaries. The glandular tissue subtends or rarely almost surrounds the base of the single stamens in a thin mold, but is absent or ob- solete elsewhere on the receptacle. The most noticeable differ- ence between Arabis and Sibara, aside from the foliage, is the lack of an expanded pedicel-summit in the latter genus. Arabis always has the pedicel-summit expanded in such a way as to create an enlarged support for the receptacle and nectaries. On mature fruits, a rather abrupt narrowing from pedicel-summit to replum-base is evident. In Sibara, there is practically no change in diameter or only a slight gradual expansion between pedicel and replum-base. An obscure ring of sepal-, petal-, and stamen-scars and shriveled nectaries alone mark the outer part of the receptacle after mature siliques have been developed. Very small flowers and a branching habit give Sibara a habital symmetry which is distinctive in itself. Only one or two species of Arabis have flowers even approaching the small size of the largest flowers of Sibara and only a few species have a similar habital aspect. The essential characters setting Arabis off from other genera Possessing linear siliques are its accumbent cotyledons and si- liques flattened parallel to the septum. One possible exception to the flattened siliques is found in A. glabra, but here the Siliques are often somewhat flattened and are never more than Semiterete. The nature of the siliques together with the dis- tinetive creamy-yellow petals in A. glabra are used by some botanists, especially the European, to separate this species from Arabis as Turritis glabra. There can be no doubt but that A. glabra is somewhat related to A. hirsuta which it nearly parallels in distribution, and it seems preferable to retain it in Arabis. In the related Halimolobos, the siliques are terete and the 294 Rhodora cise seeds have incumbent cotyledons. These points, together with the general habit, seem to place Halimolobos nearer to Sysim- brium than to Arabis. Several species, such as A. Whitedu and A. Hookeri, with terete siliques and incumbent cotyledons are now more naturally placed with Halimolobos. Arabis is sometimes placed near Cardamine, but it is not be- lieved that these genera are at all closely related. The peculiar elastic valves, opening with a relatively wide band of replum- tissue remaining on each side, is a distinctive feature of Carda- mine and Dentaria. It should be stressed that the stamens of Arabis are always tetradynamous. This aids in distinguishing it from genera like Thelypodium where the single and paired stamens are of equal length. RELATIONSHIPS OF THE SPECIES The systematic classifications of Arabis presented by ae Candolle,! Meyer,? Prantl,? Watson,’ v. Hayek,® N. Busch, si O. E. Schulz,” though almost successively more complicated in the chronological order given, do not satisfactorily allow for the i clusion of the North American plants without separating obvi- ously related species into different sections. The erga _ of these attempts to arrange systematically the species of Ara d into sections result from the fact that there are no clearly ve fined groups within the genus. Furthermore, many of : workers did not have sufficient material to insure an inclusive treatment. In the older works the lack of representative ae was the penalty imposed by the immaturity of the science itse 4 but the recent students have been more fortunate in this i : However, even with a quantity of material available for stu 4 the difficulty resulting from the lack of clearly defined : t genera or sections cannot be overcome. It is my conviction t al among the western American species of Arabis truly scopes subdivisions of the genus other than the species themselves af nonexistent. For this reason I have arranged these species 1 * Syst. Nat. 2: 214 (1821) and Prod. 1: 142-148 (1824). 2 Ledeb., FI. Alt. 3: 18-20 (1831). * Engler & Prantl, Pflanzenf, 3: 2 Abt. 192-193 (1891). 1895). * Fl. Sib. et Orient. Ext. 4, fam. 25: 428-475 (1926). * Engler and Harms, Pflanzenf. 17b: 2 Aufl. 542-547 (1936). 1941] Rollins—Monographie Study of Arabis 295 oxy\ ebulé gerdvline demise. SERIES 2 : 0 Fendleri—_"** issima Drum aoe S action encase He eae iene Sere Oe. ciiechaital divaricarpa sper eit\evo \ see eyo ! in nee —salbyi : hase ed kis eG / Pee ane FY ‘ ar Wellb PA\F r * Koel Hy Svind all vy a. A Twicvoghulle Fernaldians : \ MAINE re ae ee ae ee eee hatte ia US es OR es . ‘ : ‘ : “is, viqidissime . 6 Heey teh a x at or 60 - subpinnatifide al : : oT ‘, rut : a < ; \ L \ ES ails \ Ps aan area ehbivencia : ‘ at , : ere bee, On eae \ \\ aK < \ : me \ paid Mot Ae eee SERIES 3 ee ean pies fuvecta éxnciadaks Aes Adin os ~ ; x Me ote, eh } ; \ Z SS Sand \ \ \ \ oy Tricormnule =; 5 \\ : aculeolala Wedie a ere ete H oe ' ; We eee ; parish MeDenaldisne ' eo ' Bite aang ae ea ‘ ' ‘ \ PON eons ae bg 4 ‘ Vl eotetoeenn, Bot Ree ete OE SOS etc oe ble heres h \\a : eo ho es = t } ™ / Pid sativa = hi v's a ba y / Fl ayn oiesias be ' ry Z i ' 7 ! tt Maec eee alebra - } petic avis ; i SERIES 1 ist SERIES 4 ; i H Fig. 1. Relationships in ARABIS 296 : Rhodora [JuLy four more or less interrelated series of no specified taxonomic rank. It should be understood that these divisions do not cor- respond to previously published sections of Arabis, and are used primarily for the purpose of indicating probable circles of affinity. The four series are presented in chart form as Fic. 1. Solid lines between species indicate a rather close relationship, broken lines indicate a probable distant relationship. SPECIES-CRITERIA No single character, although it may appear to be very funda- mental, can be solely relied upon positively to indicate relation- ships between species or to separate interrelated groups of species in Arabis. One group is likely to have the same or a similar interplay of characters as that found in another. A simple illus- tration of the point may be taken from the cases of A. platy- sperma and A. suffrutescens, whose close relationship to each other is obvious, and of A. puberula and A. Shockleyi, another re- lated pair. The pedicel-position in one species of each pair 1s directly opposite that of the other. Thus, A. platysperma and A. Shockleyi have ascending pedicels, whereas A. suffrutescens and A. puberula have the pedicels reflexed. Taking pedicel- position as a criterion it is impossible to show that A. platy- sperma and A. suffrutescens are related but, when several char- acters are taken into account, the relationship is unmistakable. Numerous similar instances of parallel development in unrelated species and unrelated groups of species could be pointed out. This fact makes evident the necessity for great care in the evaluation of diagnostic characters. Furthermore, it 18 clear that a successful classification of Arabis must rest on a series of characters whose pattern has been carefully determined. In general, qualitative rather than quantitative differences are most reliable and should be given first importance in the characteriZ8- tion of the species. Hasit.—Most species of Arabis are strictly herbaceous, but many possess a woody caudex. In perennial species the caudex and root persist in a dormant state throughout any period inimi- cal to growth, and leaves arise from the apex of the caudex 07 caudex-branches with the return of favorable growing conditions. The caudex shows special development in such desert species as 1941] Rollins,—Monographie Study of Arabis 297 A. pulchra and A. perennans in which it is much elongated. Using desert shrubs for support, the leaves and flowering stems of these plants are often held from one to several decimeters off the ground. A number of tufted perennials have a highly branched caudex which may or may not be subterranean. In such cases many of the caudex-branches bear only leaves, while others bear both leaves and a fertile stem. Usually some rem- nants of shed leaves are to be found upon the caudex. Some- times these are of a characteristic nature, as in A. Koehleri, in which the leaf-bases resemble the peg-like pulvini found on old twigs of Picea. The stems may be either simple or branched even within a given species. Certain species show a marked constancy in pos- sessing simple fertile stems, while others constantly possess branched stems. Plants which have been grazed are usually branched even though a simple stem is the normal habit of growth. Whether the stems are branched near the base or to- ward the top markedly affects the general symmetry of the plant. Thus, the position of the branching is sometimes of importance in distinguishing between entities, but this feature is usually not constant and must be used with caution. The tap-root system is characteristic of Arabis, but that alone does not indicate that all species are either biennial or perennial as Hopkins' seems to have presumed. The plant which he treated as Arabis virginica (it is here transferred to Sibara) Was said to be biennial.2. Two lots of this plant have been grown from seed to the flowering stage in less than three months at the Harvard greenhouses. This would seem to indicate that it is usually annual in spite of its having a tap-root. All species of Arabis included in the present paper are either biennial or perennial, FoLiace.—Two sorts of leaves are found in Arabis. Those of the flowering stems are differentiated from those at the base of these stems or those terminating sterile branches of the caudex. The term “basal leaves” used throughout this paper refers to the leaves found at the base of fertile stems, those terminating caudex-branches or on the root-crown. The term “radical * Reopora 39; 67 (1937). * Ibid. p. go, 298 Rhodora [Juny leaves” has often been applied here but usually only the first year’s rosette is truly radical. Leaves subsequently produced are not near the root in many cases but, strictly speaking, are stem-leaves, though different from the leaves of the fertile stem. The term “cauline leaves” is restricted to the leaves borne on the fertile stems. The size and shape of the leaves are variable, but extremes in either may be characteristic of certain species and are use- ful in their delimitation. The leaves are entire, dentate or rarely more highly dissected. These features, though variable, are use- ful when only extremes are considered. The strongly perennial species usually have entire or nearly entire leaves. Many species have cauline leaves with auricled bases, while in others no auricles are present. This feature has been widely used as a species-criterion but it is unstable and may be relied upon only in some cases. The basal leaves are always petiolate, while the cauline leaves are only rarely so. In western America only four species with petiolate cauline leaves are known. The cauline leaves usually diminish in size from base toward apex of the flowering stem. The basal leaves are often caducous, hence they may not be present on mature plants, as in A. canadensis and A. tricornuta. However, most of the species in our area have the basal leaves present even in old plants. VestiTURE.—The trichomes of Arabis are single-celled and never glandular. They may be simple, bifureate with branches appressed (malphigiaceous) or variously branched. The term stellate has usually been applied to trichomes of the Cruciferae with several to many branches. Most of the highly branched types are actually not stellate, if by that term is meant a cel- trally attached hair with branches radiating from the middle after the manner of a star. The severally branched trichomes commonly found in Arabis are nearly all of a dendritic type, 1. with a central trunk and irregularly placed, elevated branches. These branches themselves often fork, giving rise to neW branches, thus forming a tree-like configuration. The fact that truly stellate trichomes are common in the Cruciferae and other plant families has made it desirable to distinguish the tree-like hairs by applying to them the term dendritic. 1941]. Rollins,—Monographie Study of Arabis 299 The type of trichome, whether simple, malphigiaceous, stellate, dendritic, swollen at base, terete or flattened, is of considerable importance in characterizing many species. A few species have two kinds of pubescence, but the association is usually fairly stable. The significance of the size of trichomes is not fully understood. Large differences in size may be safely used in support of other characters, but within at least a few species the trichomes vary considerably in size. An unsuccessful attempt was made to correlate trichome-size with polyploidy in A. divari- carpa. Further data are needed to determine whether poly- ploidy has the same effect on trichome-size as on the size of cells and stomata. The density and extent of the pubescence on plants of Arabis are ecologically plastic and. vary seasonally. Though often used with a great deal of assurance, the mere presence or absence of pubescence is a weak criterion. Specific observations on this point are discussed in some detail under A. Drummondi. Fiowrrs.—In general the flowers are similar throughout Arabis. They are always tetradynymous. Large differences in size may be utilized in differentiating some species, while in cthers there is considerable variation and size-differences are not reliable. Measurements to be useful comparatively must include the limits of variation in the size of any given part or organ. When accurately determined and based on a sufficient sample of the total population of a species they may be highly reliable. Unfortunately, it is only in some cases that such a Procedure can be followed by a taxonomist because of lack of material. The inner sepal-pair of the Arabis flower is never saccate at base. However, the outer pair is sometimes slightly saccate and in a few instances markedly so. The character is only limitedly usable. Petal-color varies from a deep purple to White in Arabis. A. glabra alone has a slightly yellowish petal. The nectar-glands in the Cruciferae have been shown to be diversely specialized and it has been contended that their form, Position and extent of development are of importance in the Classification of the family. Beginning with Hildebrand in 1879 and Velenovsky in 1883 the nectar-glands have received in- creased attention. They formed the chief basis for a system 300 Rhodora ae of the Cruciferae by Bayer! and were heavily Bees _— von Hayek? in his inclusive treatment of the fami e ie ta of these facts, the glandular pattern in all species o ra, apa cluded in the present paper has been carefully a i diagrammed to see if this feature might be used as an a Mee characterization of species or species-groups. ey among finding of Giinthart,? where several types were — type, in European species, only two patterns were found. : wih ‘s — which the glandular tissue surrounds the base of the s : ae men and is obsolete or entirely absent elsewhere on the - eee is confined to series one. In the rest of the species, the g 4 often tissue forms a continuous mold beneath all stamens Hh clear has short projections flanking the petal-base. ssi _ ane that the glandular pattern is of no. particular wisi an prren attempt to distinguish between the species themse ieee species. it may be used to some extent in separating groups Sta SILiQuEs.—The mature silique, as in other ne paramount significance. The shape is, in angie eptum. linear with the valves strongly flattened parallel to t fee An exception is the semi-terete capsule of A. glabra. oe the size of the silique is variable, there are ceanoenpe ong to the variation and, if accurately determined, size ii wes useful tool in characterizing certain species. The sei and & of the valves is variously developed in different setae lied limited use may be made of this feature if it is carefu ie of in conjunction with other characters. The cellular pa hole, it the septum is distinctive in a few species but, on the Ww : ‘a is so uniform and difficult of determination that little ave been made of it. The siliques of most species of gacit in glabrous but in some they are always pubescent, W a few species they may be either pubescent or glabrous. neue Silique-position is constant in most species and the ape of this feature in determining species of Arabis 1s wel pgs However, in a few instances the siliques vary considera ie their position with respect to the rachis within a given ue a Among different species the siliques vary all the way 'T Beih. Bot. Central. 18: 119-180 (1905). wie 2 i * Reprint from Bibl. Bot. 77: 1-36 (1912). 1941] Rollins —Monographie Study of Arabis 301 strictly erect position to a closely reflexed one. The siliques may be straight, curved or even strongly bow-shaped. The presence or absence of an elongated style upon the mature silique is a character which may be trusted within certain groups. An unusual development of the style is found in A. Parishii and it is only slightly less prominent in A. Johnstonii. SrEps.—The shape and size of the seeds and the absence or extent of development of wings are significant. Whether the seeds are in one or two rows in each locule has frequently been the basis for the segregation of species. Considering the point of attachment, the seeds are always in two rows, the attach- ment being at the point of replum-junction on both sides - of the locule. In certain species the seeds do not enlarge to the full width of the locule, hence the seeds themselves appear to be in two rows. In other species, the seeds entirely occupy the locule-width and are then said to be in one row. The differ- ence is not a morphological one and is more apparent than real in so far as the actual number of rows of seeds in each locule is concerned. It is clear, then, that it might be possible for the seeds to be in neither one row nor two, but half-way between. This is precisely what happens in some species. In others the character is very constant. It will be seen that the real sig- nificance is to be attached to the relationship between seed- size and silique-width. My use of the terms uniseriate and bi- Seriate refers to the position of the seed rather than to the actual number of rows as determined by seed-attachment. The cotyledons are accumbent in all the species presented in the present paper. GEOGRAPHICAL DISTRIBUTION The geographical area occupied by each species dealt with has been carefully determined and mapped. Maps showing the distribution of all species and varieties not extremely localized, except the widespread A. glabra and A. hirsuta, are presented on several pages appropriately placed in the general treatment. Many more specimens than could be cited were included in the data assembled for the maps. These geographical data have been of considerable value in determining natural boundaries of the species and varieties. Where a question as to the distinct- 302 Rhodora [Juny ness or relationship of a variety or species has arisen, a compari- son of the areas occupied by the related forms has usually been decisive. In addition to the horizontal area occupied by the species, the altitudinal zone in which they normally grow has been considered. The habitat and soil-preferences have also received attention whenever the data have been available. Arabis is indigenous to all the general habitats represented in the area, including high alpine and desert situations. No one limited area in western America is conspicuous because of the number of endemic species of Arabis contained within it. The endemic species are quite generally distributed among the Rocky Mountains, Sierra Nevada Mountains, northern Coast Ranges in California, Siskiyou Mountains, Cascade Mountains and the deserts or lesser mountain systems in the intervening region. One species, A. Hoffmannii, is exclusively insular, being found only on Santa Cruz Island. Plants of the genus extend from Alaska on the north to northern Baja California on the south. Throughout the paper, where geographical distribution has been of special importance, it has been discussed under the species concerned. Many of the accompanying maps were based on Goode Base Map no. 102; by permission of the University of Chicago Press. CyTOLoGy In addition to the usual techniques employed in Systematic Botany, some species have been examined cytologically. As 4 result of these studies it has become clear that certain species usually considered to be highly complex taxonomically are apt to contain a polyploid series. A. hirsuta in America, A. H olboel- lit, A. Fendleri and A. divaricarpa are species of this type. Far more work is necessary before the variations and complexities of these species will be fully understood, but it is now definitely known that there are polyploid plants in all of them. Results of the cytological examination of thirty-four collec- tions of Arabis are presented in the table following, which con- tains chromosome-numbers for eighteen species and varieties. All but four are reported for the first time. The counts were made from material gathered directly at the locality given OF from plants grown from seed collected at these stations. Ex- cept for one or two instances, specimens have been preserved 1941] Rollins —Monographic Study of Arabis 303 and may be consulted at the Gray Herbarium under the num- ber cited. Counts were made from pollen mother-cells, using aceto-carmine smears, or from root-tips, using a modification of Fuelgen’s technique. Unfortunately, it has been impossible to obtain sufficient data to make more than one or two general statements concerning the cytology of Arabis. Although several interesting problems have. arisen, their solution has not been attempted, but it is hoped they will be mastered during future work in the genus. On the basis of Jaretzky’s! report of the chromosome-numbers for ten species of Arabis, Manton? gave the number eight as the fundamental one for the genus. Recently, Frank H. Smith,’ after finding six pairs of chromosomes in A. dentata and seven in A. laevigata, doubted whether eight was necessarily the basic number for the genus. A glance at the table below will show that a haploid number of seven is by far the commonest among the species investigated. Undoubtedly seven is to be considered one of the fundamental chromosome-numbers of Arabis. CHROMOSOME-NUMBERS IN ARABIS Species Place of Collection Collector and N 2N Number A. alpina Harvard Botanic Garden, Rollins s. n. 8 16 : Cambridge A. alpina Gaspé Co., Quebec Knowlton s. n. 16 A. cobre ta Co., Wyomi Rollins 1656 7 A. Crandallii Gunnison Co., Colorado Rollins a“ 7 A. Crandallii u n Co., Colorado ollins 7 A. divaricarpa Uintah Co., Utah Rollins 176715 A. divaricarpa Larimer easy vo agspa Rollins 2400 8 A. Drummo Uintah Rollins 1764 7 A. Drummondi Hinsdale Go: Calocsds Rollins 1506 7 44 A. Fendleri var. typica Gunnison Co., Colorado Rollins 2098 7 14? var. typica Costilla Co., Colorado Rollins 1288 21 var. spatifolia remont Co., Colorado ollins 2069 y dae © | var. spatifolia Chaffee Co., Colorad Rollins 2077 7 var. spatifolia San Juan Co., Colorado Rollins 1511 14 re h Sahar Gunnison Co., Colorado Rollins 2090 7 . a var. pyenocarpa Windham Co., Connecticut Weatherby and Rollins s.n. 16 1 Jahrb. Wiss - Bot. 68: 1-45 (1928). * Annals Bot. 46; 516 and 543 (1932). * Am. Journ. Bot. 25: 220 (1938) 304 Rhodora [Juny CHROMOSOME-NUMBERS IN ArABIS—(Continued) Species Place of Collection Collector and N 2N Number var. pycnocarpa Daggett Co., Utah Rollins 2267 32 var. pycnocarpa Gunnison Co., Colorado Rollins 1427 16 32 A. Holboellii var. retrofracta Sublette Co., Wyoming Ownbey 1008 7 var. retrofracta Albany Co., Wyoming Rollins 1070 7 var. retrofracta Larimer Co., Colorado Rollins and Chambers 2401 7 var. retrofracta Mesa Co., Colorado Rollins 2188 var. retrofracta Uintah Co., Utah Rollins 1765 14 var. retrofracta Humboldt Co., Calif. Harris et al. ar var. pinetorum Lake Co., Colorado Rollins 1349 21 var. pinetorum Gunnison Co., Colorado Rollins 1426 14 A. laevigata Litchfield Co., Connecticut Weatherby s.n. 7 A. lignifera Uinta Co., Wyoming Rollins 2308 7? 14 A. lignifera Montrose Co., Colorado Rollins 2129 7 A. lyrata Litchfield Co., Connecticut Weatherby s.n. 8 A. microphylla var. typica Wallowa Co., Oregon Constance 1274 7 var. saximontana Big Horn Co., Wyoming Williams 3264 4 A. perennans Mesa Co., Colorado Rollins 2128 (es. A. repanda Madera Co., California Constance 2386 7 ACKNOWLEDGEMENTS Many people, too numerous to mention, have given assistance in one way or another during the course of this investigation. I am particularly indebted to Professor M. L. Fernald under whose direction the project was carried out. I appreciate the kindnesses of Mr. and Mrs. C. A. Weatherby, who have col- lected cytological material and supplied notes and photographs of type specimens of Arabis and related genera from European herbaria. Also, I wish to thank the staff-members of the Gray Herbarium, who have given unhesitatingly of their time. Pro- fessor Karl Sax has generously allowed me the use of his labora- tory and has given assistance in many ways. The efforts of Dr. Lincoln Constance in obtaining cytological and other special material are gladly acknowledged. Finally, my thanks are ¢* tended to the curators of the following herbaria for loaning ™@ terial or making facilities of the herbaria in their charge avail- able for my use: Herbarium of the National Museum of Canada, Ottawa (Can); the herbarium of Mr. I. W. Clokey, South Pasadena (Cl); Dudley Herbarium of Stanford University, 1941] Rollins —Monographie Study of Arabis 305 Palo Alto (DS) ; herbarium of the Field Museum, Chicago (F) ; herbarium of the U. S. Forest Service, Washington, D. C. (FS) ; Gray Herbarium of Harvard University, Cambridge (G); the herbarium of Dr. Eric Hultén, Lund, Sweden (L); herbarium of the Missouri Botanical Garden, St. Louis (M); herbarium of the National Arboretum, Washington, D. C. (NA); herbarium of the University of Notre Dame, Notre Dame (ND) ; herbarium of the New York Botanical Garden, New York (NY); herbar- ium of the University of Oregon, Eugene (O); herbarium of Oregon State College, Corvallis (OS); Pomona College Herbar- ium, Claremont (P); the herbarium of Mr. Frank W. Peirson, Altadena, Calif. (Peirs); herbarium of the Academy of Natural Sciences of Philadelphia, Philadelphia (Ph); the Rocky Moun- tain Herbarium of the University of Wyoming, Laramie (RM) ; herbarium of the U. S. Field Station at Sacaton, Arizona (Sac) ; the herbarium of Mr. J. W. Thompson, Seattle (T); the Inter- mountain Herbarium of the Utah Agricultural College, Logan (UAC); herbarium of the University of California, Berkeley (UC) ; herbarium of the University of California at Los Angeles, Los Angeles (UCLA); herbarium of the University of Idaho, Southern Branch, Pocatello (UIP); herbarium of the U. 8. Na- tional Museum, Washington, D. C. (US); herbarium of the University of Washington, Seattle (UW); Vegetation Type Map Herbarium, U. §. Forest Service, located at the University of California, Berkeley (VTM); herbarium of Willamette Uni- versity, Salem, Oregon (W); herbarium of the State College of Washington, Pullman (WSC). The symbols in parenthesis are those used to indicate the herbarium where the specimens cited throughout the text were seen. Collections marked (R) are in my own herbarium. Synopsis or THE Genus AraBIS L. 1n WESTERN NortTH AMERICA AraBis L, Biennial or perennial herbs, often with a ligneous base, glabrous to sparsely or densely pubescent with simple, bifureate, stellate or dendritic trichomes; caudex simple or ranched; stems terete, leafy, simple or branched; basal leaves petiolate, entire, dentate or rarely somewhat dissected, persistent or caducous; cauline leaves sessile or rarely petiolate, often auricled, approximate to remote, entire or dentate; inflorescence Tacemose, ebracteate, greatly elongating as the flowers succes- 306 Rhodora [JuLy sively expand; flowers erect to reflexed at anthesis, tetra- dynymous; sepals erect, oblong to nearly ovate, uniform, or the outer pair infrequently saccate, bounded on the sides and apex dons accumbent.—L., Sp. Pl. 2: 664 (1753); Gen. PI. ed. 5, 298 (1754); DC., Syst. 2: 213-244 (1821) and Prod. 1: 142- B 3 ( ot. (1935) ; Rollins in Res. Stud. State Coll. Wash. 4: 1-52 (1936) ; O. E. Schulz in Engler, Pflanzenf, 2 Aufl, 17b: 542 (1936) 3 ° e wt ~ Hooker, Fl. Bor.-Am. 1: 40-41 (1829); Torrey & Gray, Fl. ~ Am. 1: 78 (1838); Gray, Gen. Illustr. 1: 143, pl. 59 (1848). na, L B. auline hing petiolate, the lower dissected to the mid- ; petiole not wi : , iolari. : Ui ot winged; basal leaves 10-15 cm. rig oy petiolars. le ee eRe ead ieee ese ae Od a epee es 2S 26 See 2 O's eS 6 os do 6 68 oe 0 © 1941] Rollins,—Monographic Study of Arabis 307 B. Cauline leaves sessile, or if petiolate, then the petiole winged, entire or merely dentate; basal leaves less than 'S cmt long jocncokcis were. cae ee eee C. C. Siliques pre pedicels erect or ascend:ng............ 1D, D. Lower cauline leaves petiolate ; mre winged; basal leaves oblanceolate to bro adly spatulate, 1-3 cm sbeet Pet co pine be or send slightly — pals and: Caltoming. i560 css 5. A. repanda. D. All Sains avon sessile ; iil leaves linear hoe rg less than 8 mm. wide; ; petals definitely exceeding the VOU acs Bot ewe eee E. Leavis aed lower stems hoary with a ogo pubes- cence; pedicels eubainenk: Califomia 3 ose ec 47. A. dispar. never hoary; pedic els PIADTOUS, 4a es F. Basal leaves linear, about 2 1 mm. wide, dense ly hirsute with very coarse Keicatition persiste ent and ag ming successive rev pee cag rosettes on elongated crown; California ....... 52. A. pygmaea. F. Basal pete tga late to spatu la ote 3-8 mm. ide, gla “ pubescent with much finer trichomes ‘han 4 he above Gecaduons: not form- ing oe rosette n the crown; bw Cahfomis: ana ‘Oregon... 220.2) 3. A, ‘platysperma. C, eas and pediibels reflexed (horizontal in r suf- vulescens,. val. AOMsORTONS) 600.0 ose ss G. Leaves and stems green, pent ACRE or pubescent below only; siliques attenuate at apex; seeds un — Idaho to California and Washington........ 50. A. suffrutescens. G. Leaves and stems hoary with a minute pubescence ; siliques obtuse at apex; seeds Uieetiate: tga ges an GVAQH 65) hype ieee eases AS olewecaalote A. Seed-wing less than 1 mm. wide or seeds win ie see including wings) less than 2 mm. long; hips ‘usually less than 3 mm. wide, very rarely up to mm. wide. H. Basal leaves oborate to steed oblanceolate, obtuse and rounded at a often forming a flat rosette at base of stems, tigate. Tiler thin, ‘lade ayes as ieee as long, gl abrous or hirsute with simp r forked trichomes ex- trichomes near base; glandular tissue subtending single stamens only, or disrupted and poorly de eveloped near a stamens; outer sepals ote except in . ; Bec biseriate; siliques semi-terete; flowers pari yellow (rarely lilac); cauline leaves ample, ovate ovate-lanceolate, usually gaa widely distributed om Quebec to California. ...........--++eeeeeeeeee : ‘ Seeds pe ac "ligase tatiied parallel to a foment white to purple; cauline leaves much smaller i E pancmeirieefie in n A. hirsuta, obovate to sepals 4 r gia — _ rs purple to white.........-.-+e+seseeeees Ke Plants eat Be =e lees eo eercky oa i. 308 Rhodora [Juny L. Basal leaves 3-8 cm. long, broadly oblanceolate, Sem or rarely few-toothed ; California to ; = lit EE Ge SPORE GAR Nip MER RAP 12. A. Davidsonn. L. Basal gees less than 2.5 cm. Fig ovate to yamine dentate to rarely entire.......... M. Flow purple; pews Ficipet at apex atten CRIs ie hes aes ones 6. A. McDonaldiana. M. Flowers white; Polke nce at apex; plants of Washington and northern Oregon...... tes . furcata. K. Plants al seater. at least on ae stems and basa N. Wee urls sides. : “petals ‘9-20 mm. long; pedicels 5, peeang 12 aS cs Vaher SRS Pe O. O. Pe tals 4-8 broad, rounded at apex or retuse ; ‘lass a pubescent at least on anes s P. Plants more than 2 dm. high; siliques acute, 3.5-6.5 cm. long; style slender or obsolete; Be. regon and extreme northern RUMME G od i, aisck awe che nes Q. Q. Lowet ne appressed-pubescent; _ basal leaves not ciliate; meta of uniform, small dendritic trichom co SRE MeL ANE Q. Lower stems hirsute aT leaves siltate ubescence of lange simple, forked or dendritic trichome udex raulticipitatty eggitiaaal basa. leaves 1-4 cm. long, oa ciliate with large aris aa or rarely forked trichomes; cauline lata leaves 5-15 mm. long........-- 7. A. aculeolata. audex simple or rarely closely r 5. A. modesta. pubescent with large dendritic or forked trichomes with bulbous bases; we cauline leaves cm. sha ee i aeliae & $4 Oe P. Plants less than 2 dm. high; siliq ong, obtuse, usually with a stoutiah style; seth from Sonoma Co., Cali es. O. Petals 25-3 mm. broad, mantle truncate at apex; ici aniadly glabrous, but ay with ves ew marginal trichomes on the basa rapes Donaldiana. N. 8 white, rarely pinkish; petals less than lo: ruc Laearag a Oa or very rarely voller hirsute in A. Stems panei: aibrakiay div aricately asce poset _ A. furcata. rarely divaricately ascending; siliques ron os ong; cauline leaves auriculate ; seagoe’ comuttl — Pihhceeee Sintua sh poeta ek ae 8s _ A. hirsuta. et 9 Sader a ab ee A it; eaves ; kites wit. th large si wages or — forked trichomes to glabrous; gem ms hirsu with spreading trichomes or glabro 1941] Rollins—Monographie Study of Arabis 309 Basal leaves entire; lower ig leaves ses- sile; styles 1 mm. or more long; Montana to Utah and: Washington 3s foc. see 9. A. Nuttallir. Basal leaves iynibecinnatiad to dentate; lower cauline leaves usually petiolate; styles less n surfaces with dendritic trichomes, never og lower stems sn 2 ines septa to g —— Idaho and Washington...................- A. crucisetosa. H. Basal leaves linear to line seg alata ast (if brown, . minutely pubescent or the siliques reflexed o r erect to ae glandular tissue continuous beneat O60 8S ESE TO CS OS O58 6 Ole Che Ble te Oe be Se £6 8 ee ee 68 6 6 6 8) ee U. Styles less than 1 mm. long or the vine acai basal leaves hatrowly oblanceolate. i664... V. — uniseriate, winged, flattened, 1.5-2.5 mm. Seed-wing over 0.5 mm. wide; siliques 2.5— wide, sheavioaialy ascending ; Cali- fornia.......-. -47, A. oer Seek in a than 0.5 mm. wide; siliques —— 2 ; “eer spreading ‘at right ae 42. A. inyoensis. V. —— biseriate, eeentinily wingless plump, about d; Utah to: California .o.3 «4. 41. A. Shockleyt. U. Styles 1-8 mm. long ; ba a3 leaves reed Styles 4-8 m tage Hoses g; seeds eaeuty winged dale 12 ey long ; Calflora se: ci25 & vs 9. A. Parishiz. Sty irs Sit mm. long; seeds aera aie s 3-5 em. lo ong: California: 5.0%. 48. A. Johnstonii. it, Sildeus Taek pedicels iouely reflexed to PUM is eee. A ee sa ss she bly rg cauline leaves linear, not crowded, : Colorado to California ............-- 43. A. pulchra. W. Seade a uniseriate ; cauline leaves are to broadly anceolate, cro wded, often subpinnatifid. Siliques Sawspigs at apex; unis + anting; petals 7-1 long; Idaho to California and ‘oon OTE ir aa eee ca ce 5 . A. puberula. — i at apex; style about rami g tals 10-14 mm. long; Ca fornia cad = naa riniet wivis « 8. Cauline leaves, stems and pedicels greenish, dense pubescent to g ebtotin never hoary; basal leav of sometimes hoary-pubescent, but usually greenish, densely pubescent to never pas er lh eat ee ee X. 310 Rhodora [JuLy us ee met Sib pedicels erect to ascending, neve dive at right angles to rachis or ewadiie: - scending or m arcuate. .Y. n trichomes; lower cauline leaves variously pubes- cent or oe na ous, but not hirsute, auriculate to TT ois <4isk ae nds ns Soest wees Z. Lower fruiting pedicels 24 cm. long, glabrous; silique arcuate or at least somewhat curve ed; seeds he ate. Lower stems densely hirsute ; i meeree elow, not coriaceous; lower auline leave pubescent on both surfaces; ye 35. fi gracilipes. e sparse es- lean a Eee . Hoffmannit. Z. Lower fruiting pedicels less than 2 cm. lo pubescent or glabrous; siliques straight to arcuate, if arcuate, then the seeds a. Basal le and lower cauline leaves densely t with dendritic trich , gray, of an appearing so; siliques straight to somewhat rv arcuate ; ai usually numerous from a m oii a cake ee b. b. Basa leaves kanes to somprebianveniste, sually less th d; siliqu straight except in A. Koehleri ine Ne res eos ORE Shas iW cos c. c. Petals less than 7 mm. lon stigmas raat iiaiie: styles i up to l oa we i ociek chai d. a > fee crowded near base; perinale pubescent; lower stems finely * hewae orn, kU en ep nee meee e. nuaiee about 1 mm. wide, erect, Sine between seeds; see . wing ess to very narrowly winge _Colorado...... Reeneerrent A. Crandallit. e out eho cautine leaves hieiaasieulate os var. stylosa. Siliques blunt at apex; stigma sile, cauline leaves hig a att Nevada and Californi 18d. A. cat onl Sse depauperata. e d. Basal leaves not pannose, though o appearing so to the naked eye; cauline 1941] Rollins,—Monographiec Study of Arabis 311 leaves narrowly same remote; pedicels glabrous; lower stems hirsute to glabrous; Wyomi ing to Nevada and Weaingts (65 6 as ai, ‘A. or c. Petals 7-14 mm. long; styles about 1 m long or stigmas sessile ............... f. Basal leaves linear, rather coarsely pu- bescent, not pannose, acute; siliques penis curved; caudex-branches elon- Plants 3-9 dm. high; pedicels pubes- cent or glabrous; lower stems pu- bescent or rarely glabrous; caudex without numerous peg-like _leaf- bases; widely distributed from the Ro Sea Mountains to California Pest WURGKC Cre ae eS . A. sparsiflora. Plants any less than 3 dm. high; rous or rarely sparsely pubescent; caudex covered with numerous peg- like leaf- Bigiat: southern Oregon 30. A. Koehleri. f. Basal leaves linear-oblanceolate, minutely pubescent, pannose, obtuse; aes straight; caudex-branches not elo Style, absent, stigma sessile; pedicels siliques divaricate; Califor- bain Ry Cease es Gs ya ceeaoe es . A. inyoensis. Style about 1 mm. long; pedicels and siliques erect; Nevada ...... 32. A. Fernaldiana. b. Basal leaves oblanceolate, often broadly oe Ms 5 mm. wide; siliques somewha “ g. Stanis and pedicels agp with a al richomes; cauline leaves ovate to broadly oblong, ey approximate; California and southern Oregon...... 28. A. Breweri. WS oe acne ccc oki ndanbaeebues cease h. Siliques 1.5-3.5 mm. wide; seeds orbicular to oblong, 1-2 mm. S heoad: stems glabrous to Co eed beac ent below, never —- i ing om with sp i. Seeds gorge oblong, winged on one eon and hea stal pig ‘tiques and pedicels t; ailiques usually blun deol leaves ose wit iphiginceos hairs or cae flow: v rarely cara ae Aisecbined poses 14. A. Drummondi. 8 312 Rhodora [JuLY i. Seeds uniseriate or rarely imperfectly bi- seriate, orbicular, winged all a und ; a pedic asal lea sheet branched trichomes or glabrous; s pink to purple .........----> ]. ; ignta less than 3 dm. high; stems several Ces rgent; widely distributed ...... 13. A. Lyallii. i Plants 3-9 9 dm . high; stems ome vaya rely two or ale from a mple i gt sn and pedi ae aronscately ascending to more widely spr eading. .k. k. Siliques 15-25 mm. wi * margin trai oad. a line leaves glaucous, ovate; SI- liques g es Pca only at or nerveless; Wyom- Me PR eo cs bn vie cin 16. A. fructicosa. Basal leaves sparsely pubescent with appressed veral-branched, r ar coarse trichomes or gla brous sri ed ge gl ucous, nerved at least to the middle caudex Sebi: widely distributed. De ee oe ic hee kk ows es _ A. divaricarpa. k. Siliques 2.5-3.5 mm. wide, ma ice un- sear seeds 2-2.5 mm Sad Cal- Sy Pe ore eee 17. A. rigidissima. h. Siliques | tows Chas 15 mm. wide; seeds orbic- ular. me 1 mm. broad; stems sparsely insu ase; Wyoming to — : and Brit ish ec. ows 5h: » microphylla. Y. Basal leaves na outs linear, densely Seats "with large simple or lerksd trichomes; lower cauline leaves ‘hae, sh ar, ion aaidlate- Idaho to Re ee ee ew, ee ee ee ee 99. A. Cusickit. fe at right angles to rachis to strictly . . ee Se ee 8 Fe eR ee Oe ee eee 6 ey SOE 6 2 e 6 68) oo woe Oe Oe hirea wi mple or for ee hairs to gl 18; pedicels glabrous, gently curving downward ex- c ie in A. rectissima where ye are oe (eeessees 2 eee ee ee ONS ee Oe oe eee eee Cee oO oe never n. Plants 2 m. “sy ce to ‘several, — eacionece io biseri lower cauline inate ce ed to slightly. overlapping or mote; pedicels 1-2 em. lon 1941] Rollins,—Monographie Study of Arabis 313 Siliques 2-4 cm. long; cauline leaves small, mostly remote; basal leaves hirsute with simple trichomes to glabrous; Utah and Nevaeh sodcisus ciuk vu Heb is 8655+ <8 at. A. pendulina. Siliques 4-6 cm. long; cauline leaves fairly ample, lower imbricated; basal leaves hir- sute on the blade-surfaces with forked . A. Fendleri. RE ere a pee rare . pendulina. rae sesrob “uniseiaie: ic slightly tapering Siliques’ dnaaaele, pendulous, acute to obtuse; at least some of the basal leaves densely hirsute; Wyoming, Daloende and 20. A. demissa. Siliques almost oblique to widely pen 6 lous, acuminate; basal leaves glabrous, thin; Colorado ‘(r = Re eee 9. A. orylobula. m. Siliques strictly reflexed, appressed pedicels geniculate; ba ats le es ie birees) to slightly broader; California and Oregon 22. A. rectissima. 1. Basal leaves densely pubescent with fine to cee dendri trichomes, blades always pubescent, never merely hirsute with simple or forked hairs; pedicels pubescent to een spreading at right angles to strictly reflexed ..........-.+2.++++5 p. Cauline leaves ey Ses mag naa petiolate ; filaments e stamens straight, erect; petals at r. es spatulate . eli ccohatc never linea a pitino g less —_ 0. aaa wide; aver gi panno: rigid ; siliques widely reel a to sgrenn re- 8. cae bale © Se o.8 2 bk eae we eee ee, «a8 88S ee s. Pedicels —. mm. long; sili iques rts 0 strictly reflexed; siliques straight and spreading at right angles or arcuate. .u. 314 Rhodora [JuLy u. a ie (-3) dm. high, cegancoert numerous, filiform; caulin fever few, small and remote. Cauline leaves linear-lanceolate, 1 1-2 ish Ces bi eis 31. ry Rete digs Cauline leaves oblo hirsu aes he spreading trichomes; nen bs os et _ A. Gunnisoniana. u. Plants 39 ek: high, rarely caespito ose ; stems one to several, fairly stout; acs 8 stems densely ap- ssed-pubescent at least below; usually curved; hea My very narrow (less than 02 mm. wide); Wyoming pee to Arizona and Nevada .36. A. lignifera. v. At least the outer basal “Jeaves below, rarely appress epiiihact tgant — — seni stems one Outer baad] leon i eoaaily oblance- edicels slender, em. lo 6-9 mm. long, 1.5-2.5 mm. wide; ne ce to Mexico wi Cali- Oe SS 94, A. perennans. Outer be sal leaves narrowly ob- — acute; pedicels stout, pressed-pubescent in > cali- fornica; petals 8-12 mm. lon mm. wide; sridiely hie aes ra 26. A. sparsiflora- t. — fruiting pedicels definitely i ending to strictly reflexed, straight, not widely spreading with tips de- scending; siliques mostly straight, pen- dulous to strictly peel again the rachis ee eee 6 eee 1941] Rollins,—Monographie Study of Arabis 315 w. Pedicels strictly reflexed, appressed to the rachis; widely distributed. ots a Holboelliz. w. Pedicels strongly descending, but n the: Tachis. 237254. x. Stems hirsute with spreading tiie below; siliques straight or SONS ae ews . Holboellit. x. gig finely appressed-pubescent be- w; siliques often slightly prt and usually widely pendulous; Wy- oming to Arizona and Nevada. "36. A. lignifera. r. Basal reagent linear, minutely pubescent, usually ose; seed-wing abou wide; Pauline ate ings , weal: aha es pendulous on wi spre ading 5a i pedicels; Woomine ss Nevada and 0. A. cobrensis. er lo Pr showy, Sen Fes at t angles; siliques densely pubescent except in var. gracilis; to California and Mexico...... 43. A. pulchra. p. Cauline leaves petiolate; filaments of single sta- mens arising at right ‘angles to receptacle and curved upward; petals barely exceeding the : BOOGIE: ATNOE iis aoa sk bate eo VAs 44. A. tricornuta. 1. A. cuaBra (L.) Bernhardi. Biennial or rarely perennial; stems one or few from a tap-root, simple or rarely branche d eile deeply divided, ee entir ire, coars rsely pubescent with forked or dendritic ichowiee rarely almost glabrous, 6-15 cm long, 1-3 em. wide, apap gna hirsute; cauline leaves lanceolate to ovate, entire or the lower toothed, = a Sagittate, glabrous or the oink eit sely pubese ent, 5c long, 14 em Pils ser oblong, non-saccate, obtuse, yellowish to rarely purp lish, 3-5 mm. long; petals Lease to na y spatulate, slendanclaret *yelfowieh- white very le pur- plish, 5-7 mm. long; glandular tissue subtending 5 seg stamens, absent or nearly obsolete near paired stamens; auauen pripned erect, semi-terete to slightly more ed glabrou us, 4— long, slightly more than 1 ae h ide DMosthsn ner rved = the middle baste of the stem simple or rarely hala rigger .la. Var. typica. errs: of the stem several-branched, app pressed. . b. Var. furcatipilis. Var. typica Hopkins in Ruopora 39: 106 (1937). A. sine (L.) Bernh., Syst. Verz. Erf. 195 (1800). Twurritis glabra 316 Rhodora [Juuy L., Sp. Pl. 2: 666 (1753). Arabis perfoliata Lam., Dict. 1: 219 (1793). Turritis macrocarpa Nuttall in T. & G., Fl. N. Am. 1: 78 (1838). Arabis macrocarpa (Nutt.) Torrey in Bot. Mex. Boundary pt. 1: 32 (1858). Turritis glabra, var. lilacina O. E Schulz in Fedde, Repert. Sp. Nov. 33: 191 (1933).—Quebec to North Carolina, California and British Columbia. Only the leading synonyms have been given above. For full synonymy, references, and citations of specimens see Hopkins, 1. c. . Var. FuRcATIPILIS Hopkins. Stems pubescent below with appressed, several-branched trichomes. Pubescence smaller than in var. typica—Ruopora 39: 109 (1937).—Utah and California. Uran: Logan City Camp, Logan Canyon, Cache 0., Maguire 3437 (G, Typ). CAaLirorNiA: without locality, T. Bridges 15 (G); Santa Lucia Mts., Monterey Co., March, 1898, Plaskett 53 (G). A. glabra is abundant and widely distributed in the mountains of our area where it is usually found in habitats ranging from moist shady places to open dry disturbed soils. The total dis- tribution is very wide in temperate Asia and Europe as wel as in North America. This species is often removed from Arabis and placed in the genus Turritis, but the characters used to make the separation are not convincing. The distine- tive characters supposedly include yellowish petals as opposed to white or purple ones, semi-terete instead of definitely flattened siliques and biseriate instead of uniseriate seeds. As shown above, the seed-position in the silique may be untenable even as a species-criterion in Arabis, therefore it is of even less value as a generic character. Furthermore, a number of un- doubted species of Arabis have biseriate seeds. The siliques of A. glabra are only semi-terete and in many cases they are flattened as much as those of A. hirsuta. As for petal-color, A. glabra does not have exclusively yellowish petals. Some- times the petals are pink or even purplish. Other minor charac- ters such as the angular nature of the silique or the complete nervation of the valves are sometimes pointed to as being dis- tinctive, but these are variable in A. glabra and cannot be satis- factorily used to bolster generic segregation. : Two varieties of A. glabra have been distinguished in America. O. E. Schulz in 1933 described Turritis glabra, var. lilacina,' basing it upon the lilac color of the petals and sepal-margins. * Fedde, Repert. Sp. Nov. 33: 191 (1933). 1941] Rollins,—Monographie Study of Arabis 317 An isotype of this variety does not show the characters de- scribed by Schulz, but specimens from widely separated sta- tions (Wyoming, Rollins no. 991; Idaho, Rollins & Constance no. 1109 and Utah, Garrett no. 2236) have purplish to llac- colored buds and newly expanded petals. The supposedly dis- tinctive color fades to the usual straw-yellow as the flower fully expands. I do not consider plants with such dubious ephemeral characters to be worthy of any nomenclatorial recognition. In the case of Hopkins’s var. furcatipilis,! a more fundamental char- acter, that of pubescence-type, is used as a distinguishing charac- ter. The distribution of this variety, including as it does only isolated stations in northeastern Utah and western California, is peculiar. Perhaps the variety is not a natural one, but until more information is available on its distribution this cannot be determined. 2. A. HirsuTA (L.) Scop. Biennial or perennial; stems wets: one to several from a simple or branching caudex, sim branched above, hirsute with coarse, spreading, simple or ade trichomes (strigose with appressed malphigiaceous trichomes in var, agate often glabrous above, 2-7 dm. high; basal leaves oblong to oblanceolate or ageaiy spatulate, short- petioled, sbi to rarely acutish, entire, dentat repand, hirsute on both surfaces with ‘seat simple or forked aehasbies or rarely almost glabrous, 2-8 em. long, em. wide; cauline leaves lanceolate to nelehie tives or nearly spatulate, acute or obtuse, sessile, auriculate, entire to coarsely dentate, hirsute on both surfaces or the upper glabrous, 1-5 (—7) em. long 0.52. . Wide, approximate to remote; pedicels erect to teisceete ascending, glabrous or rarely sparsely hirsute, 0.5-1.5 em. long; sepals oblong, glabrous or rarely with a few trichomes, 2.5—4.5 mm. long, about 1 mm. wide; petals white to rarely Spear h, oblong to spatulate, 3-9 mm. ‘long; siliques erect, strict t o di- varicately ascending, glabrous, nerved below or searly. the entire length, 3-6 cm. long, 1-2 mm. wide; style 0.5-1 mm. long, * Rxopora 39: 109 (1937). 318 Rhodora ‘ei Key To THE AMERICAN VARIETIES OF A. HIRSUTA a. Flowers small; petals 3-5 mm. long, white to yellowish- white; cauline leaves approximate to subremote, siliques strictly erect; outer sepals moderately saccate........ : b. Stem-pubescence spreading, mostly simple....... 2a. Var. pycnocarpa. b. Stem-pubescence appressed, mostly malphigiaceous 2b. Var. adpressipilis. (Pursh) T. & G., Fl. N. Am. 1: 80 (1838)?