( enit amt ANNALS OF THE MISSOURI BOTANICAL GARDEN b^. NECS "T E48 Annals of the Missouri Botanical Garden Volume XX 1933 With Twenty-three Plates and Forty-three Figures Published quarterly at 8 West King Street, Lancaster, Pa., Kid the Board of Trustees of the Missouri Botanical Garden, St. Louis, Mo Entered as second-class matter at the post-office at Lancaster, Pennsylvania, under the Act of March 3, 1879. 549006 Annals of the Missouri Botanical Garden A Quarterly Journal containing Scientific Contributions from the Missouri Botanical Garden and the Graduate Labora- tory of the Henry Shaw School of Botany of Washington Uni- versity in affiliation with the Missouri Botanical Garden. Information The Annals of the Missouri Botanical Garden appears four times during the calendar year: February, April, September, and November. Four numbers constitute a volume. Subscription Price — — — $6.00 per volume Single Numbers - — — — 1.50 each Contents of previous issues of the Annals of the Missouri Botanical Garden are listed in the Agricultural Index, published by the H. W. Wilson Co. STAFF OF THE MISSOURI BOTANICAL GARDEN GEORGE T. Moore. Assistant to the Director, KATHERINE H. LEIGE. HERMANN VON SCHRENE, CARROLL W. DoDGE, Pathologist. Mycologist Jesse M. GREENMAN, RoBERT E. WOODSON, JR., Curator of de Herbarium. Research Assistant. EnNEsT S. REYNOLDs, NEeLL C. H Physiologist. Con Ea dn "Editor of Publications. BOARD OF TRUSTEES OF THE MISSOURI BOTANICAL GARDEN President, GEORGE C. Hrrcncock. Vice-President, SAMUEL C. Davis. Second Vice-President, DANIEL K, CATLIN. L. RAY CARTER. ALBERT T. PERKINS. Tuomas S. Marrirr. EUGENE PETTUS. GEORGE T. Moore. ETHAN A. H. SHEPLEY. FRED G. ZEIBIG. EX-OFFICIO MEMBERS: GEORGE R. THR BERNARD F. Dic Chancellor oe n: University. Mayor of the e p St. Louis, WILLIAM SCARLETT ALBERT KUNTZ Bishop of the Diodi of Missouri. President of The Academy of Sci- ence of St. Louis. Davip C. Topp President of the Board of Education of St. Louis. GEORGE F. HEFFERNAN, Secretary. TABLE OF CONTENTS A New Yellow Nymphaea from Tropical Africa George H. Pring and Robert E. Woodson, Jr ee om m n Contribution to the Lichen Flora of North Lo 0 oe oo kh Wes ee OR ale Veli J. P. B. Räsänen Astranthium and Related Genera....Esther L. Larsen Ferns and Fern Allies of Missouri........... M. Hlizabeth Pinkerton $9 wx € & «o9 WLW E Row Ee wow $e «9 Blastomycosis: Report of a Case, with a Study of an Etiologie Factor and a Classification of aue Organism. ........... cane eee Morris Moore A Monograph of the American Species of the Genus Halenis...........-22 ee Caroline K. Allen The Sensitivity of Orchid Seedlings to Nutri- tional Ions :.......4.66..5543 eee F. L. Wynd A Revision of the North American Species of Di 5.0 MMRMCPRENNRUNENEENRO (s son Lily M. Perry Nutrient Solutions for Orchids. ........ F. Lyle Wynd The Foliose and Fruticose Lichens of Costa BE E Sees es on eS ERN Carroll W. Dodge A Neutral (?) Strain of Mucor sphaerosporus from Missouri.......... Sa Morris Moore A Study of Hndomyces capsulatus Rewbridge, Dodge and Ayers: A Causative Agent of Fatal Cerebrospinal Meningitis....... Morris Moore Sources of Carbohydrate for Germination and Growth of Orchid Seedlings........ F. Lyle Wynd.. 79-118 119-222 223—231 239-362 363-372 373—407 469 471—568 569-581 Super Optimal and Thermal Death Tempera- tures of the Cotton Plant as Affected by Variations in Relative Humidity.......... .... Dorothy Megowen Berkley and Earl E. Berkley Studies in the Apocynaceae. IV. The Amer- ican Genera of Echitoideae. . Robert E. Woodson, Jr. Report of a Botanical Expedition into the Mountains of Western Texas............. nee Julian A. Steyermark and John Adam Moore General Index to Volume XX................... PAGE 583-604 605—790 791-8006 807—817 Annals of the Missouri Botanical Garden Vor. 20 FEBRUARY, 1933 No. 1 A NEW YELLOW NYMPHAEA FROM TROPICAL AFRICA! GEORGE H. PRING Superintendent, Missouri Botanical Garden ROBERT E. WOODSON, Jr. Research Assistant, Missouri Botanical Garden Instructor in Botany, Henry Shaw School of Botany of Washington University Nymphaea ($ Brachyceras) Burttii Pring & Woodson, spec. nov., foliis magnis longe-petiolatis orbiculare-sagittatis margine plus minusve conspicue undulato-sinuatis apice late obtusis vel rotundatis sinu profundo auriculis leviter divergentibus obtusis 25-35 cm. longis subcoriaceis utrinque viridibus vel saepius purpureo-maculatis glaberrimis vel umbilico paulo papillato subtus nervis manifestis sed vix prominentibus; floribus speci- osissimis primulino-flavis 15-18 cm. diametro metientibus, sepalis ovato-lanceolatis acutiusculis 5-8 em. longis 2-3 cm. latis dilute viridibus immaculatis laevibus, petalis ca. 20-23 anguste ellipticis apice acuminatis vel anguste acutis medio versus grada- tim angustatis plerisque 5-nervatis exterioribus quam sepalis paulo brevioribus ca. 4.5-7.0 em. longis 1.0-2.5 cm. latis, stami- nibus 190-200 linearibus connectivo manifeste elongato basi paulo ampliato exterioribus ca. 5 cm. longis dilute cadmio-flavis, carpellis 28-30 stylo linearo profunde inclinato stigmate concavo; fructu minore 3.5—4.0 cm. diametro metiente, seminibus ovoideo- oblongoideis apice minutissime apiculatis longitudine leviter punc- ticulato-striatis griseo-brunneis maximo ca. 0.075 cm. minimo ca. 0.05 cm. diametro metientibus.—Cult. Missouri Botanical Garden, Aug., 1930, G. H. Pring s. n. (Herb. Mo. Bot. Garden, TYPE). Leaves large, long-petiolate, orbicular-sagittate, margin more or less conspicuously undulate-sinuate, apex broadly obtuse or rotund, sinus relatively deep and narrow, auricles obtuse, slightly 1 Issued April 29, 1933. ANN. Mo. Bor. Garb., Vor. 20, 1933. (1) [Vor. 20 2 ANNALS OF THE MISSOURI BOTANICAL GARDEN divergent, 25-35 cm. long, subcoriaceous, either surface green, or occasionally somewhat purplish-maculate, particularly when young, glabrous, or the umbilicus slightly papillate, the nerves manifest but not prominent beneath; flowers showy, primrose- yellow, very fragrant, 15-18 cm. in diameter; sepals ovate- lanceolate, acute, 5-8 cm. long, 2-3 cm. broad, pale green, without purple spots, smooth; petals about 20-23, narrowly elliptic, apex Fig. 1. Nymphaea Burttii Pring & Woodson: S., sepals; P., petals; St., stamens; S.L., Au ee leaves; F. L., floating leaves. acuminate to narrowly acute, the base gradually narrowed from about the middle, 5-nerved, the exterior somewhat shorter than the sepals, usually 4.5-7.0 em. long, 1.0-2.5 em. broad; stamens 190-200, linear, the exterior about 5 cm. long, cadmium-yellow, the connective manifestly elongate, somewhat broadened at the base; carpels 28-30, the style linear, sharply inclined, the stigma concave; fruit relatively small, about 3.5—4.0 cm. in diameter; seeds ovoid-oblongoid, minutely apiculate, rather inconspicuously 1933] PRING AND WOODSON—A NEW YELLOW NYMPHAEA 3 puncticulate-striate longitudinally, about 0.075 cm. long, 0.05 cm. broad, grayish-brown.—Cultivated at the Missouri Botanical Garden, Aug., 1930, G. H. Pring s. n. (Herb. Mo. Bot. Garden, TYPE). Only two yellow tropical Nymphaeas have previously been described from Africa: N. sulphurea Gilg and N. Stuhlmannit Schwfth. & Gilg. The former differs from N. Burttii chiefly in the smaller flowers (4-7 cm. in diam.) and leaves (4.5-5.5 cm. long), which are almost exactly orbicular-cordate in outline, purplish-maculate sepals, and more elongate tubers. N. Stuhl- mannii is easily distinguishable from N. Burttit by the somewhat smaller flowers (10-15 cm. in diam.), with broader, obovate, obtuse or rounded petals, and smaller (21-25 cm. long), entire leaves, which are broadly ovate-cordate in outline, with broad, rounded, regular lobes. Furthermore, the venation of the leaves of N. Stuhlmannii is extremely verrucose beneath. The type specimen of N. Stuhlmannii has not been available to the writers for personal examination, but has been compared with & duplicate specimen of N. Burttit by Dr. H. Melchior, of the Botanical Museum at Berlin-Dahlem, who kindly affirmed the distinction of either species. A photograph of the type specimen of N. Stuhlmannit (Stuhlmann 410 in Hb. Berol.), generously provided by Dr. L. Diels, Director of the Botanical Garden and Museum at Berlin-Dahlem, has been deposited in the herbarium of the Missouri Botanical Garden. After a search of over ten years, the director of the Missouri Botanical Garden obtained in September, 1929, a seed-pod of what was presumed to be Nymphaea Stuhlmannii through the personal efforts of Mr. B. D. Burtt, Esq., botanist for the Tsetse Research Bureau, Kondoa, Tanganyika Territory. An excerpt from Mr. Burtt’s notes is quoted. “The seed was collected from plants growing in a seasonal rain-pond in the Sambala ‘Mbuga’ seasonal swamp, the plants having spectacular yellow flowers 8 inches in diameter and sweet scented. The flowers float on the surface of the water and on examination were found to contain dead bees (Apis melifera [?]) that were imprisoned by the anthers over the stigmatic surface of the flowers. The plants were collected on May 19, 1929, the seed from the same locality on July 15, 1929. Other plants were collected March 16, 1929, from a seasonal rain pond near Salia, Kondoa Distr. I have observed the plant in seasonal rain ponds at Magungila, Wembare Steppe in 1928, also near Lilbilin, Massai Land, in 1927." [Vor. 20, 1933] 4 ANNALS OF THE MISSOURI BOTANICAL GARDEN The single, small seed-pod, in somewhat immature condition, was received at the Garden on September 19, 1929. It was im- mediately cleaned, and the many seeds planted in the greenhouse water-lily tanks. Three weeks later a single seed germinated from the lot, producing its first primrose-yellow flower on June 17, 1930. On July 30, the plant was removed from its pot and plant- ed in the pond out of doors where it continued to bloom until frost. 'The successful introduction of a yellow tropical water-lily has opened an entirely new field for the hybridist. For many years the available colors in the Brachyceras group have been limited to blue and pink, and more recently to white through the intro- duction from the Missouri Botanical Garden of the hybrid “Mrs. G. H. Pring.” During the past three seasons 250 pol- linations have been made with the pollen of N. Burttii, resulting in many variable forms of commercial interest. The first season (1930) was devoted to selfing the yellow-flowered species, and many fertile seed-pods resulted. The following year twenty specimens were grown from seed and planted in the ponds out of doors. No variation in the color of the flowers was observed, but a variable character was evident in the peduncle and petioles, some being pure green, while about an equal number were brownish. An unusual character, which appears to be dominant in all the plants, is a peculiar twisting of the peduncle, the day before the flower opens, in such a manner as to submerge the bud. The following morning, however, the peduncle straightens, holding the open flower erect in a natural position. This action has also been observed in several hybrids. As a propagator, Nymphaea Burttii is by far the poorest of any grown at the Garden. Of the fifty propagating tubers secured from pot plants during the past two seasons, only one has pro- duced growth up to the present, despite the fact that they have been in the heated propagating tanks for a period of six months. Since it is an extremely poor propagator from tubers and extreme heat is necessary for growth, it is very doubtful if it will find a place in garden pools. On the other hand, the hybrids derived from it, particularly the “Saint Louis," are both readily propa- gated from tubers and suitable for cooler water. [Vor. 20, 1933] ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 1 Leaf, bud, and flower of Nymphaea Burttii Pring & Woodson, X 2$. del. A. A. Heinze ANN. Mo. Bor. GARD., Vor. 20, 1933 PLATE d) i WOODSON—A NEW YELLOW NYMPHAEA PRING 4 CONTRIBUTION TO THE LICHEN FLORA OF NORTH AMERICA! VELI J. P. B. RASANEN Instructor in the Agricultural School, Kurkijoki, Finnland The lichens here described have been collected in the northern section of North America. Of these specimens, 18 are Alaskan, 60 from eastern Canada (New Brunswick, except one from Quebec), approximately 150 from western Canada (British Columbia and Alberta), and a few from the state of Wyoming. The Alaskan lichens, 15 species in all, are largely the usual arctic ground lichens. The New Brunswick lichens, which com- prise 46 species, come from around Dalhousie and were collected in the year 1930 by the agronomist Tapio Reijonen. Usnea longissima was collected near Franklin, Quebec, by Dr. Viljo Kujala in 1931 during his scientific journey. On this same journey, he collected in western Canada (120 species, varieties, and forms) and in Wyoming (6 species). The latter specimens come from the sandstone in a cactus desert. The lichens from western Canada were collected by Dr. Viljo Kujala in various types of forest in British Columbia in the following localities: Jackman, Longworth, Aleza Lake, Prince George, Six Mile Lake (Trout Lake), Hazleton, Dorreen, Co- wichan Lake (Vancouver Island), Kamloops, Fish Lake (near Kamloops), Blue River, Valemount, Golden, Field, and in Al- berta in the vicinity of Lake Louise. The desert lichens come principally from the vicinity of Kamloops in British Columbia and from Wainwright in Alberta. Although the same species have been encountered in several different localities, the total number of species, varieties, and forms is 171 with 5 entirely new species, as follows: Usnea Kujalae, Placodium xanthostigmoideum, Nephromium canadense, Diplo- schistes canadensis, and Sphaerophorus Tuckermaniv. The list further includes eleven new varieties and forms, as follows: Usnea comosa var. stuppea, Alectoria sarmentosa var. gigantea, A. Fremontu f. perfertilis, Cetraria nigricans var. pallida, ! Duplicates of most of the specimens have been deposited in the herbarium of the Missouri Botanical Garden. Issued April 29, 1933. ANN. Mo. Bor. Garp., Vor. 20, 1933. (7) [Vor. 20 8 ANNALS OF THE MISSOURI BOTANICAL GARDEN C. tenuifolia var. reticulata, C. tenuifolia var. pseudoislandica f. septentrionalis, C. juniperina var. canadensis, C. juniperina var. crispata, Parmelia elegantula var. americana, Cyanisticta Hookeri var. septentrionalis, M ycoblastus sanguinarius var. Dodgeanus. Species new to North America or otherwise noteworthy include the following: Usnea lapponica, U. dasypoga, U. caucasica, U. prostrata, U. rugulosa, U. similis, Alectoria Fremontii subsp. olivacea, A. jubata var. Vrangiana, Ramalina Roesleri, Ochrolechia upsaliensis, Pertusaria xanthostoma, Physcia pyrithrocardia, P. muscigena f. squarrosa, P. endophoenicea, Peltigera Maurizüi, P. canina var. suomensis, P. lepidophora, P. Nylanderi, Cetraria Merrillii, Parmelia Delavayi, Squamaria alphoplaca, Nephromium subparile, Lobaria oregana, Diploschistes bryophiloides. The lichens collected by Dr. Kujala are principally epiphytic and soil-lichens from a variety of forest types, although a few of the soil lichens come from fields or desert. It might be inter- esting to note here that in regions as widely separated and isolated as the North American prairies and the south Russian and Asiatic steppes and deserts there appear many identical, or, at least similar, lichen forms living on the ground. Such steppe lichens are: Parmelia vagans (appearing also in south Russia and Asia), Physcia muscigena (many forms, a few also growing in the mountains), Acarospora Schleicheri (appears in the Asiatic, North- African, and Russian steppes, in addition to the European Alps and Pyrenees),! Psora decipiens (also in deserts and mountains), Diploschistes canadensis (the similar Russian D. scruposus var. terrestris), Ochrolechia upsaliensis (sporadically found outside of the steppes), Cladonia cariosa var. cribrosa (particularly on other mineral substrates, slightly calciphile). As the lichens in this collection were not collected by a lichen- ologist, the lichen flora of the various localities is very poorly represented and poor in species. They are, however, of particular interest to the European investigator on account of the many similar forms. The typical European Cetraria juniperina and C. islandica do not appear at all amongst the American lichen specimens (the latter also being absent from the ''Cetraria islandica" specimens collected in Tierra del Fuego by Professor Auer's expedition in 1929). ! Magnusson, Monogr. Acarospora. p. 395, 1929. 1933] RASANEN—CONTRIBUTION TO LICHEN FLORA OF AMERICA 9 As the lichens from the various localities have not been suffi- ciently collected, conclusions as to the relationship between the American and European species are still hypothetical, particularly as the present systematic revision of lichens necessitates many corrections of the earlier determinations. We can, however, state that as far as our present knowledge extends a great number of identical lichen forms appear both in North America and Eurasia, although the quantitative relationships between the different species differ considerably on the two sides of the Atlantic. I wish to acknowledge with thanks Dr. C. W. Dodge, My- cologist to the Missouri Botanical Garden, who has kindly trans- lated my manuscript from German to English, and T. Hidén, of Helsinski, Finnland, who has gone over the latin text of this work. UsNEA DASYPOGA (Ach. Röhl. New Brunswick: Dalhousie; British Columbia: Prince George and Six Mile Lake, on Picea. USNEA DASYPOGA Var. SUBSCABRATA Vain. British Columbia: Hazleton, on Betula and Pinus. Sterile. SNEA CAUCASICA Vain. Thallus prostratus, glauco-viridis, medulla sublaxa, K—.—British Columbia: Six Mile Lake, on Picea. Fertile. USNEA PROSTRATA (Vain.) Ras. British Columbia: Aleza Lake, on Picea. Sterile. UsNEA RUGULOSA Vain. British Columbia: Hazleton, on trunks of Pinus Murray- ana. Sterile USNEA LONGISSIMA Ach. Quebec: north shore of the St. Lawrence Bay, Franklin, especially on branches of Picea in a mixed forest; British Columbia: Vancouver, Catillano Canyon, on Thuja in a forest with abundant ferns. Sterile. Usnea comosa (Ach.) Vain. Thallus erectus, parce sorediosus, soredia isidiosa; medulla crebra, K —.—New Brunswick: Dalhousie, on pue British Columbia: Hazleton, Xn on branches of Pinus Murrayana. Steri SNEA COMOSA var. stuppea Räs., var. nov. Thallus erectus aut suberectus, brevior fruticulosus, 5-7 cm. longus, laevigatus vel leviter verrucosus, sorediosus, pallido-stramineus; soredia maculiformia, demum parce isidiosa. Medulla laxa, stuppea, K—.—British Columbia: Hazleton, iba pin on small, half-dried Pinus, Picea, and Betula, on a sandy heath. Steri Usnea simitis Motyka U. subcomosa Vain. in Kgl. Danske Vidensk. Selsk. Skrifter, Nat. og Math. Afd. VIII. 6: 392 (110) 1924. With the above species, occasional in Hazleton. Sterile UsNEA LAPPONICA Vain. With the above species, Hazleton, on Pinus. Sterile. UsNEA HIRTA (L.) Motyka. pae Columbia: Aleza Lake, occasional on Picea canadensis and P. Engelmannii. i UsNEA Kujalae! Räs., sp. nov. Thallus curtus, idum fruticulosus, 3-5 cm. longus, bene ramosus, lacteo-stramineus vel albo-viridis, sorediosus. Rami laevigati, sine verruculis nervisque. Medulla laxissima, K—. borate rara, terminalia, 2-3 mm. lata, albido-straminea, fibrillosa, pinni: Sporae 9-10 X 5.3-7 u, ovoideo- ! In honor of Dr. Viljo Kujala. [Vor. 20 10 ANNALS OF THE MISSOURI BOTANICAL GARDEN ellipsoideae, 8 : nae, monostichae. Asci cylindrici. Hymenium 80 u crassum, I+, caerulescens. Subsimilis Usneae sorediiferae Motyka, sed thallus albidus et parce sorediosus.— British Columbia: Hazleton, abundant on branches of Pinus Murray- ana. Rarely fertile. ALECTORIA SARMENTOSA Ach. Thallus elongatus, pendulus vel prostratus, K = K(Cl) + leviter roseus, demum ferrugineus, I= leviter caerulescens. Sporae normaliter binae vel 4:nae, fuscae.—British Columbia: Aleza Lake, frequent on branches of Abies lasiocarpa, Picea canadensis, and P. Engelmannii; Longworth, frequent on branches of Thuja; Hazleton, rarely on Pinus. Fertile. ALECTORIA SARMENTOSA var. gigantea Riis., var. nov. Thallus elongatus, pendu- lus vel prostratus, I em. longus, albidovetrami neus, parce pseudocyphellatus, esorediatus, K=, K(Cl)=, Cl, I= leviter caerulescens. Apothecia vulgaria, plana vel concava, atro-fuece vel livido-fusca, nuda vel leviter pruinosa a normaliter 3:nae, 25-31 X 14.5-20 y, ellipsoideae, fusco-nigrae. Hymenium 130 pu crassum, I + UN —British Columbia: Longworth, frequent on Thuja; Dorreen, on Pseudots marron LAETA (Tayl) Linds. A. japonica Tuck., A. osteina Nyl., A. lata DR. (Arkiv f. Bot. 20": 24. 1926). Thallus singes vel prostratus, osteo-ochroleucus, apicem versus concolor, K= passim obsolete intensive fulvescens, K(Cl)-r roseus.—British Columbia: Hazleton, occasional on the ground; Alaska, occasional on the ground. Sterile. ALECTORIA OCHROLEUCA (Ehrh.) N Thallus pro parte ochroleucus, majore parte viridi-niger, K= vel basim versus obsolete + dilute fulvescens, K(Cl) =. Alaska, on the ground. Sterile. ALECTORIA IMPLEXA (Hoffm.) Nyl. f. FUscibuLA Arn. Thallus pallido-fuscescens, K= lutescens, sorediosus. Soredia albida.—British Columbia: Hazleton, on branches of Pinus, Aleza Lake, on branches of Abies and Picea i y E c on trunks of Abies; British Columbia: Kamloops, Fish Lake, Golden, and Aleza Lake, very frequent on branches and trunks of Pseudotsuga and Pinus Murrayana. Also on conifers near Hazleton. Sterile. — JUBATA var. VRANGIANA (Gyeln.) Ras. ngiana Gyeln. (Magyar Bot. Lap. 31: 46. 1932) Thallus 1 parce sorediosus, vel fere esorediatus, olivaceo-viridis vel fusco-viridis, K- oon .— British Columbia: Hazleton, on conifers, occasional. Sterile. Aisoni A Fremonti Tuck. f. perfertilis Räs., forma nov. Thallus prostratus, E RAO meti laevis, subnitidus, esorediatus, K=, Cl=, K(Cl)=. Apothecia vulgaria numerosissima, lateralia, convexa, flavida. Sporae 5-8 X 5 y, ovoideae, 8: nae, hyalinae. Hymenium I+ caerulescens.— British Colmibis: Golden, on the trunks and branches of Pinus Murrayana. ALECTORIA FREMONTI subsp. oLrvACEA Ris. Thallus olivaceo-castaneus, prostratus, laevis, nitidus, esorediatus, K=, Cl=, K(Cl)=. Apothecia rarissima, minora, flavida.—British Columbia: Golden, on iion Mies iE and Pseudotsuga, Kamloops, Fish Lake, on Pinus Murrayana. Ste The sorediose form, named by Du Rietz (Arkiv f. Bot. 20": 8. 1926) subsp. Eriks- sonit, was not seen among the North American specimens. It would appear more correct, as Du Rietz (l. c.) thinks, not to regard the sorediose form as the main species. My subspecies olivacea (Rüsünen, Medd. Soc. Fauna Flora Fennica 43: 4. 1933] RASANEN—CONTRIBUTION TO LICHEN FLORA OF AMERICA 11 1916) is much shinier than the North American bid and has very rare apothecia, so that we have here three different Alectoria t A THRAUSTA (Ach.) Nyl. British Cui: Hazleton, on branches of posi Prince George and Six Mile Lake, on Picea. : ALINA phan nox Nyl. British Columbia: Hazleton, on branches of Pinus os rare. Ster RAMALINA ROESLERI ae Nyl. New Brunswick: Dalhousie, on branches of Picea, Sorbus, and on a wooden wall. Sterile. RAMALINA CALICARIS (L.) Fr. New Brunswick: Dalhousie, on trunks of Populus. Fertile. LETHARIA VULPINA (L.) Vain. Thallus sorediosus. Medulla I+ caerulescens.— British Columbia: Kamloops, Fish Lake, and Field (subalpine), on branches and trunks of Pseudotsuga. Sterile. LETHARIA VULPINA f. INCOMPTA Ach. Thallus sorediosus et bene isidiosus, me- dulla I+ eaerulescens.—Found with the bci m near Fish Lake, Golden, and Jackman, B. C., on Pinus Murrayana. Ste CORNICULARIA DIVERGENS Ach. Alaska b on the ground. Sterile. CETRARIA MERRILLII D C. californica var. Tuckermanii Howe Thallus 0.5-2 cm. longus, rigidus, erectus, divaricato-ramosus, opacus, olivaceo- niger. Rami angulato-nervosi, foveolati. Apothecia vulgaria, 0.5-3 mm. lata, plana vel convexa, terminalia, atra vel fusco-nigra, nuda, subnitida, subciliata. Sporae 8:nae, 6-8 X 4-5.5 y, ovoideae, incoloratae.—British Columbia: Golden, x ocu d Fish Lake, and Hazleton, on branches of Pinus Murrayana and other conifer; ertile. ARIA NIGRICANS Nyl. var. pallida Rüs., var. nov. I" superne pallidus vel E a ia, quove albidus, basim versus san eo-lentus; medulla I+ caerulescens. Laciniae 1-2 mm. latae, subcanaliculatae, AME ap ope dicho- tome ramosae.—Alaska, on the ground with Alectoria ochroleuca and Cornicularia divergens. Sterile. CETRARIA TENUIFOLIA (L.) Vain. Thallus canaliculatus, laevigatus, medulla K —.—Alaska, rarely with the previous species. Sterile. CETRARIA TENUIFOLIA var. reticulata Rüs. var. nov. Thallus subtus foveato- reticulatus, castaneo-fuscus, in marginibus dense isidioso-dentatus. ^ Laciniae angustae, SPSE EE UE A crispatae, marginem versus pseudocyphellatae. Medulla K—.—British Columbia: Kamloops, on stones, Valemount and Aleza Lake, on the ground. Sterile. CETRARIA TENUIFOLIA Var. PSEUDOISLANDICA Ris. (Ann. Bot. Soc. Zool.-bot. Fenn. Vanamo 2: 15. 1932) f. septentrionalis Räs., forma nov. Medulla I—. Thallus similis C. islandicae, sed I—, et margo laciniarum fere sine spinis vel ciliis. Color pallido-castaneus.—British Columbia: Golden, on trunks and branches of dwarf shrubs near the ground. Sterile. The true Cetraria islandica (L.) Ach., which gives the reaction ‘‘Medulla I lutes- cens, demum ferrugineo-rubescens," has not been seen by me from either North or South America. TRARIA NIVALIS (L.) Ach. British Columbia: Kamloops, on the ground and rocks in semi-desert; Alaska, on the ground. Sterile. CETRARIA CUCULLATA (Bell) Ach. British Columbia: rage on the ground with Arctostaphylos Uva-ursi; Alaska (alpine), on the ground. Steril [Vor. 20 12 ANNALS OF THE MISSOURI BOTANICAL GARDEN CETRARIA CAPERATA B. Vain. C. pinastri (Scop.) R British Columbia: eal a small specimen with other Cetraria species on branches of Pinus Murrayana, Aleza Lake, a chiefly geoplese lichen (Rüsünen, Uber Flechtenstand u. Fl. Veget. im Westl. Nordfinnl., p. 84. 1927), with Parmelia ambigua and P. pallescens on the base of Abies lastocarpa. Sterile. CETRARIA JUNIPERINA (L.) Fr. var. viRIDIS (Schweinitz) Riis. C. viridis gprs apud Halsey, Ann. Lyceum Nat. Hist. New York 1: 16. 1824; C. juniperina var. virescens Tuck. Thallus ae rigidus; laciniae obtusae, obscuro-virides; K=, Cl=, I=. Apothecia liri ed AEN 3-8 mm. lata, fusco-nigra, subnitida, margo sub- crenata, tenu Sporae 8 : nae, ovoida 6-8 X 3.5-6.5 u, incoloratae.—British Columbia: Golden, common on branches of Pinus Murrayana. Fertile. CETRARIA JUNIPERINA var. canadensis Räs., var. nov. Thallus erectus vel sub- erectus, subrigidus; laciniae obtusae, puni Pelih vel interdum viridi-flavae, super reticulato-nervosae, intus flavae. Apothecia terminalia, minora, 1-3 mm. lata, castaneo-fusca, nitida; margo lacerato-crenata. Sporae 6.5-8 X 4-5.2 u, ovoideae, 8:nae, hyalinae.—British Columbia: Kamloops, Fish Lake, Aleza Lake, and Hazleton, very frequent on twigs of conifers. C. juniperina (the main species) has apothecia chiefly in the middle of the thallus and the lobes are very divided. Also, the color of this species is not so bright a golden yellow as in the variety canadensis, and the surface is furthermore much smoother in the species. TRARIA JUNIPERINA var. crispata Riis., var. nov. Sicut var. canadensis thallus dense lacerato-crispatus vel mibpaoiliatas: Apothecia rara. Est Ripe cum C. juniperina var. terrestris Schaer. (= C. Tilesii Ach. in Zahlbr. Cat. Lich. Univ. 6: 341. 1930).— British Columbia: Golden, rarely on branches of Pinus Murrayana with var. viridis. CETRARIA ALEURITES (Ach.) Th. Fr. New Brunswick: Dalhousie, on trunks of Abies and on wood. Sterile. CETRARIA CHRYSANTHA Tuck. Thallus stramineus, lacunoso-nervosus, REN subtus fusco-niger, K+ lutescens, Cl=, K(Cl)-r roseus, I=.—Alaska (alpine), on the ground. Sterile. CETRARIA LAURON Ach. var. han DR. Thallus subtus fusco-niger, non isidiosus, I =.—British Columbia: Kamloops, on stones in semi desert. Sterile. CETRARIA NoRVEGICA (Lynge) DR. Thallus isidiosus, I -.—With the above species in Kamloops, on stones. Sterile. CETRARIA GLAUCA (L.) Ach. Thallus IF caerulescens.—With the above species in Kamloops, on stones; Prince George, Six Mile c" abundant on Picea; Aleza Lake and Vancouver, on branches of Picea. CETRARIA GLAUCA V&r. STENOPHYLLA Tuck. (Syn. North Amer. Lich. 1: 36. 1882). Thallus I F caerulescens, anguste laciniatus, glaucus, subtus niger.—British Columbia: Vancouver Is, Lake Cowichan, occasional on the lower branches of Picea in the forest. Sterile. CETRARIA SCUTATA (Wulf.) Poetsch C. chlorophylla (Willd.) Dalla Torre et Sarnth British Columbia: Kamloops, Fish Lake, Golden, Hazleton, Aleza Lzke, on the branches of Pinus Murrayana and Abies; Vancouver Is., Lake Cowichan, occasional on Thuja gigantea. Sterile. 1933] RASANEN—CONTRIBUTION TO LICHEN FLORA OF AMERICA 13 NEPHROMOPSIS PLATYPHYLLA (Tuck.) Herre Cetraria er Lab Tuck. Thallus 1-3 cm. latus, irregulariter laciniatus, super opacus, obscure fuscus, tuberculosus, s pad pallide fuscus, reticulato-nervosus. Apothecia marginalia, minora, 1-2 mm. lata, plana, fusca, tenuiter marginata; margo crenata. Sporae sphaeroideae vel ‘eibiellipsoidene, 8:nae, incoloratae, 4-6.5 u. Hymenium ca. 50 p crassum, I+ caerulescens.—British Columbia: Golden, on twigs of Pinus Murray- ana. NEPHROMOPSIS CILIARIS (Ach.) Hue. British Columbia: Aleza Lake, a small specimen on Pinus Murrayana. Steri e. PARMELIA PALLESCENS (Neck.) Ras P. hyperopta New Brunswick: Dalhousie, on trunk of Abies; E rs: Columbia: Aleza Lake, on base of Abies lasiocarpa as & geoplese lichen. Ste RMELIA AMBIGUA (Wulf.) Ach. New aga Dalhousie, on wood; British Columbia: Aleza Lake, on base of Abies lasiocarpa and Pinus Murrayana. Sterile PARMELIA PHYSODES (L.) Ach. f. LABROSA Ach. New Brunswick: xr mdi on trunk of Abies; British Columbia: Aleza Lake, on trunks of Picea, Kamloops Pinus Murrayana. rile. PARMELIA PHYSODES f. viTTATA Mereshk. British Columbia: Aleza Lake and Hazleton, on branches of Pinus Murrayana. Sterile PARMELIA LOPHYREA Ach. Thallus cinereus, — Ir K+ eee foe = roseus. Apothecia vulgaria, subtus ampullacea, demum Sporae sphaeroideae, 2.5-5 u, 8 : nae, incoloratae. Hymenium 50 u eat T. caerulescens.— British Columbia: Golden, on branches of Pinus Murrayana; Aleza Lake, on branches of Picea PARMELIA DELAVAYI (Hue) Nyl. Thallus parvus, esorediatus, obscure cinereus. Apothecia vulgaria, demum 5 mm. lata, subtus non ampullacea, discus fuscus, urceolatus vel deplanatus. Sporae 5-7 X 4-5 u, 8:nae, ovoideae. Hymenium ca. 55 u crassum, I+ caerulescens.—British Columbia: Golden, on Pinus Murray- ana. PARMELIA ENTEROMORPHA Ach. Thallus esorediatus, ventricoso-inflatus, cavus, K= lutescens, K(Cl) =.—British Columbia: Vancouver Is, Lake Cowichan, on trunk of Thuja gigantea. Sterile. PARMELIA VITTATA (Ach.) Róhl. British Columbia: Prince George, Six Mile Lake, on trunk of Betula?. Sterile. PARMELIA SULCATA Tayl. New Brunswick: Dalhousie, on trunks of Betula, Sorbus, Picea, Abies, and on wood; British Columbia: Prince George, Six Mile Lake, on Betula; Hazleton, on Betula; Golden, on Pinus Murrayana and Pseudotsuga; Kamloops on arid ground. Sterile. PARMELIA VAGANS Nyl. P. molliuscula Tuck. (Syn. North Amer. Lich. 1: 64. 1882). Thallus dichotome laciniatus, sine sorediis et isidiis, stramineo-virens, K F fulves- cens.—British Columbia: Kamloops, on the ground between grasses and shrubs. Sterile. The lichen is a typical desert lichen and appears also in such environment in Russia and Asia. PARMELIA SUBAURIFERA Nyl. New Brunswick: Dalhousie, on the trunk of Sorbus; British Columbia: Hazleton, on the trunk of a Populus. Steri PARMELIA PAPULOSA (Schaer.) Vain. P. exasperatula Nyl [Vor. 20 14 ANNALS OF THE MISSOURI BOTANICAL GARDEN New Brunswick: Dalhousie, on the trunk of Sorbus. Sterile. PARMELIA ELEGANTULA (Zahlbr.) Rüs. var. americana "e var. nov. Thallus opacus, obscure olivaceo-fuscescens, isidiosus, K=, Cl=, K(Cl)=. Isidia sub- cylindrica, curta, concoloria. Apothecia desunt.—British Sema Kamloops, on a Pseudotsuga in semi-desert PARMELIA PUBESCENS (L.) Vain. var. RETICULATA Cromb. British Columbia: Kamloops, on a stone in semi-desert. Steri THAMNOLIA VERMICULARIS (Sw.) Schaer. : ddl (regio alpina), on the ground between tufts of Cornicularia divergens and Cladonia rangiferina SIPHULA cERATITES (Wahlbg.) Fr. British Columbia: Doreen, on stones in moist mountain forest. Sterile. STEREOCAULON TOMENTOSUM Fr. British Columbia: Dorreen, on branches of Tsuga; Alberta: Lake Louise, on mossy earth; ay m alpina), between Cornic- ularia divergens and Cetraria chrysantha. Rarely fer LECANIA DIMERA (Nyl.) Th. Fr. British ci diae - Aleza Lake, very common on the trunks of Populus. Fertile. SQUAMARIA ALPHOPLACA (Wahlbg.) Dub. "Thallus orbicularis, radiatus, griseus, K+ partim rubescens. Medulla I—. Apothecia ca. 1 mm. lata, nigra, nuda vel tenuiter pruinosa. Sporae 9, 6.5 m, 8 : nae.—U. S. A.: Cody, Wyoming, on sand- stone in a cactus desert. ASPICILIA CALCAREA T Mudd. With the previous on sandstone in cactus desert, in Cody. Ferti LECANORA COILOCARPA A (Ach. ) Nyl. New Brunswick: Dalhousie, on a rail fence. Fertile. LECANORA UMBRINA (Ehrh.) Róhl. New Brunswick: Dalhousie, on a stone in a field. Fertile. LECANORA 8YMMICTA Ach. "Thallus areolato-verruculosus, mox totus sorediosus, virescens, K—, Cl4- aurantiaco-rubescens. Apothecia livida, mox convexa, fere immarginata.—New Brunswick: Dalhousie, on a wood rail fence mictera (Nyl.) Rüs.—in the summer of 1931, on a wooden fence in Petsamo, Finnland. Except for this I have not seen this central European and North American species. LECANORA SUBINTRICATA Nyl. British Columbia: Kamloops, Fish Lake, on trunks and branches of Pinus Murrayana. OCHROLECHIA UPSALIENSIS (L.) Mass. Thallus K-, Cl-, K(Cl)-. Epi- thecium K(Cl)—. Apothecia 1-2 mm. lata, planiuscula vel urceolata, pruinosa. Margo crassa, subrugosa.—British Columbia: on rotting vegetable remains in somewhat shaded positions in the semi-desert. PHLYCTIS ARGENA (Ach.) Flot. New Brunswick: Dalhousie, on trunk of Abies sp. Sterile. PERTUSARIA PERTUSA (L.) Tuck. New Brunswick: Dalhousie, together with previous species on Abies. Fertile. PERTUSARIA XANTHOSTOMA (Smrft.) Fr. Thallus verruculoso-inaequalis, K —, K(Cl)—. Apothecia 0.5 mm. lata, punctiformia, pallido-lutescentia. Excipulum K+, sanguineo-rubescens. Sporae 4:nae, ellipsoideae, incoloratae, 58-66 X 26-40 u. Hymenium I+ caerulescens.—New Brunswick: Dalhousie, on a Thuja trunk. PrERTUSARIA MULTIPUNCTA (Turn.) Nyl. British Columbia: Aleza Lake, on the trunk of an Abies. Fertile. PERTUSARIA FAGINEA UT CURES I IU stranthium ce. ae E a nd unequally lacerated crown; ray-flowers in 2 or Achaeto 88. bins d Mas ud subterete, or distinctly 4—5-angled.... Aphanostephus KEY To SPECIES OF ASTRANTHIUM A. Stems erect or suberect from a distinctly ligneous base. s. Leaves glalgmus. 056054 605s RR SEMEL Aces y es 1. A. mimum aa. Leaves more or less pubescent. b. Stems usually unbranched above. €. Leaves oblong-elliptic. .... Eat DEI eis ss 2. A. xylopodum cc. Leaves oblanceolate to lineal spe. a renas roie 3. A. orthopodum bb. Stems usually branched above. 6 Lenves dentate.. 5c er EN E oe ATTA e s 4. A, mexicanum | Leaves ntie, 66s 1o" Re Ar aie estas 4a. Var. chihuahuense AA. Seas erect or procumbent from an annual or slightly ligneous base. a. Plants with long slender runners....................-. 5. A. xanthocomoides aa. Plants without long slender runners. bk Aches glabrous.: Loo EIER QU ess 6. A. purpurascens bb. Achenes pubescent m glochidiate-tipped hairs. c. Leaves not rosulat d. Plants not nane UT a seeoose 7. A. integrifolium dd. Plants conspicuously branched.................. 7a. Var. ciliatum 66. Leaves TORAO. oea reesi ee des ees 7b. Var. rosulatum TAXONOMY ASTRANTHIUM Astranthium Nutt. Trans. Am. Phil. Soc. N. S. 7: 312. 1841; Benth. & Hook. Gen. Pl. 2: 265. 1873, as to synonymy. Bellis Michx. and Am. authors, Fl. Bor. Am. 2: 131. 1803, not of Linnaeus; Hook. Bot. Mag. 52: pl. 3465. 1835; DC. Prodr. 5: 304. 1836; Raf. New Fl. Am. 2: 24. 1836; Torr. & [Vor. 20 30 ANNALS OF THE MISSOURI BOTANICAL GARDEN Gray, Fl. N. Am. 2:189. 1842; Gray in Hemsl. Biol. Cent.-Am. Bot. 2: 118. 1881; Gray, Syn. Fl. N. Am. 12: 163. 1884, and ed. 2, 163. 1886 and 1888; Britton & Brown, Ill. Fl. 3: 350. 1898, and ed. 2, 3: 402. 1913; Britton, Manual, 943. 1901, and ed. 2, 1905; Rob. & Fern. in Gray, Manual, ed. 7, 799. 1908; Small, Fl. Southeastern U. S. 1202. 1903, and ed. 2, 1913. Herbaceous, caulescent, glabrate, or pubescent annuals or perennials. Leaves alternate, sessile or petioled, linear-lanceo- late to obovate-spathulate, entire or dentate. Involucre 2-3- seriate, imbricated, appressed, lanceolate bracts usually with lacerately ciliate, membranaceous margins. Heads few- to many- flowered. Ray-flowers pistillate, rays 2-3-dentate. Disk-flowers tubular, perfect, corolla 5-lobed. Style-branches linear-lanceo- late, acute, papillose at the tip, stigmatic surfaces confined to the lowermost margins of the style-branches and extending one third of their length. Pappus entirely lacking or an inconspicu- ous ring-like crown. Achenes obovate, compressed, narrowed at the base, pubescent with glochidiate or emarginate hairs, rarely glabrous. Type species: Astranthium integrifolium (Michx.) Nutt. Trans. Am. Phil. Soc. N. S. 7: 312. 1841, which was based on Bellis integrifolia Michx. Fl. Bor. Am. 2: 131. 1803, ''ad ripas rivu- lorum et in collibus umbrosis Tennassée.”’ 1. Astranthium mimum (Blake) Larsen, n. comb. Bellis mima Blake, Contr. U. 8. Nat. Herb. 22: 594. 1924. Herbaceous perennial, 38-50 cm. high; stems several, simple, erect, monocephalous, greenish, striate, glabrous or sparsely pubescent with spreading hairs; basal leaves few, obovate- oblong, apiculate, narrowed at the base into a slender petiole, 15-20 em. long, 2-3 em. wide, somewhat membranous in the dried state; stem-leaves linear-lanceolate, hirsute-ciliate, lower 5-6 cm. long, 4-6 mm. wide, gradually reduced above, the uppermost 1 cm. long, 1 mm. wide; peduncles terminal, mono- cephalous, enlarged just below the head; involucre 2-2.5 cm. in diameter; bracts 2-seriate, equal, 7 mm. high, linear, acute, sparsely pilose, ciliate; ray-flowers about 60, white, fertile, 3- dentate, about 12 mm. long; disk-flowers yellow, fertile, pappus 1933] LARSEN—ASTRANTHIUM AND RELATED GENERA 31 none; achenes of ray and disk similar, oblong, compressed, glabrous. Distribution: State of Durango, Mexico. Known only from type specimen. Specimens examined: Duranco: from Sierra Madre, 30 miles north of Guanacevi, alt. 2440-2745 m., Aug. 18, 1898, Nelson 4786 (US, No. 332830, TYPE). 2. Astranthium xylopodum Larsen, n. name. Keerlia mexicana Gray, Proc. Am. Acad. 22: 422. 1887, not Bellis mexicana Gray, Smiths. Contr. [Pl. Wright. pt. 1] 3: 93. 1852. Perennial from a woody base; stems several, simple, or slightly branched near the top, striate, hirsute with spreading pubescence; peduncles 10-14 cm. long, naked or bearing 1-2 small linear- lanceolate bracts near the inflorescence, hirsute-pubescent; leaves oblong-elliptic, evenly pubescent with appressed hirsute hairs, margins strongly ciliate; involucre 1-2 cm. in diameter; bracts linear-lanceolate, pubescent, margins lacerately ciliate, somewhat membranaceous; ray-flowers fertile, about 20; pappus lacking; achenes 4-nerved, glabrous. Distribution: known only from the State of Jalisco, Mexico. Specimens examined: JALISCO: shaded hillsides near Guadalajara, June 27-July 14, 1893, Pringle 4412 (M, G); Rio Blanco, July 1886, Edward Palmer 146 (G, TYPE). 3. Astranthium orthopodum (Rob. & Fern.) Larsen, n. comb. Bellis orthopoda Rob. & Fern. Proc. Am. Acad. 30: 117. 1894. Perennial from a short, thick rootstalk; stems several, de- cumbent or suberect, 10-20 cm. high, simple or somewhat branched, appressed-pubescent, monocephalous; leaves thick, entire, appressed-pubescent, basal leaves oblong-spathulate, 3—4 cm. long, 3-5 mm. broad, stem-leaves linear-lanceolate, erect, gradually reduced toward the inflorescence; peduncles 3-4.5 cm. long, appressed-pubescent; involucre 10-12 mm. in diameter; bracts 2-seriate, linear-lanceolate, acute, purple-margined, sparse- ly and somewhat appressed-pubescent, 4-5 mm. long; ray- flowers about 30, 1 cm. long; achenes sparsely pubescent with glochidiate or straight and slightly emarginate hairs; pappus an inconspicuous whitish ring-like crown. Distribution: western Chihuahua, southward along mountains in the state of Durango, Mexico. [Vor. 20 32 ANNALS OF THE MISSOURI BOTANICAL GARDEN Specimens examined: Cuimuanva: Guachachie, June 25, 1892, Hartman 523 (G, TYPE); vicinity of Madera, alt. 2250 m., May 27-June 3, 1908, Edward Palmer 287 ( DvnaNao: City of Durango and vicinity, Apr.-Nov. 1896, Edward Palmer 163 (M); Otinapa, July 25-Aug. 5, 1906, Edward Palmer 425 (M). 4. Astranthium mexicanum (Gray) Larsen, n. comb. Bellis mexicana Gray, Smiths. Contr. [Pl. Wright. pt. 1] 3: 93. 1852; Gray in Hemsl. Biol. Cent.-Am. Bot. 2: 118. 1881. Perennial from a branched ligneous base; stems 15-60 cm. high, several, branched above, striate, pubescent with spreading hairs; leaves appressed-pubescent, those of the stem sessile, entire, or saliently dentate, oblong-linear, gradually reduced toward the inflorescence, basal leaves ellipsoid-spathulate, crenate-dentate, 5-20 cm. long, .8-2.5 em. broad; involucre .8- 1.4 em. in diameter; bracts 2-3-seriate, cinereous-pubescent, linear-lanceolate, purple-tipped; ray-flowers white, numerous, .5-1.0 em. long; pappus an inconspicuous ring-like crown or almost obsolete; achenes compressed, pubescent. Distribution: mountains of south-central Mexico. Specimens examin HripALao: Sierra de Pachuca, alt. 3076 m., Aug. 26, 1902, Pringle 9857 (M). Mexico: Apr. 26, 1849, Gregg 701 (M, even); along brooks, Ixtaccihuatl, March- July 1903, Purpus 159 (M); moist open woods and meadows about timber line, Ixtaccihuatl, Oct. 1905, Purpus 1575 (M); “above timber line,” Popocatepetl, Oct. 1908, Purpus 3640 (M). FEpERAL District: Cima, Aug. 24, 1910, Orcutt 3768 (M); Serrania de Ajusco, alt. 3076 m., Aug. 8, 1896, Pringle 6442 (M). More tos: Toro, alt. 3015 m., Aug. 5, el Fisher (M, No. 914805). Oaxaca: Sierra de San Felipe, alt. 3076 m., June 28, 1894, pee 4719 (M); Sierra de San Felipe, alt. 3076 m., Sept. 1894, ym 261 (M). 4a. Var. chihuahuense Larsen, n. var.” Plate 3. Resembles the species in habit but differs from it in that it is much less pubescent; the leaf margins are essentially entire, ciliate; the pappus is more conspicuous than that of the species and consists of a narrow fluted crown; the achenes are glabrous, or glabrate. ? Astranthium mexicanum (Gray) Larsen var. chihuahuense Larsen, n. var., planta herbacea perennis, 6-7 dm. alta, supra ramosa; marginibus foliorum integris vel subintegris; pappo coroniforme, margine irregulariter sinuato-dentato; achaeniis glabris vel glabratis—Mexico: canyons of Sierra Madre, State of Chihuahua, Oct. 4, 1888, Pringle 2015 (M, No. 122921, TYPE). 1933] LARSEN—ASTRANTHIUM AND RELATED GENERA 33 Distribution: State of Chihuahua, Mexico, known only from the type. Specimens examined: CurmvuanHva: canyons of Sierra Madre, Oct. 4, 1888, Pringle 2015 (M, No. 122921, TYPE). 5. Astranthium xanthocomoides (Less.) Larsen, n. comb. Brachycome xanthocomoides Less. Syn. 192. 1832, nomen sub- nudum; Schlecht. in Linnaea 9: 265. 1835. Brachycome xeranthemoides Steud. Nomencl. Bot. ed. 2, 220. 1840, in part (typographical error). Keerlia linearifolia DC. Prodr. 5: 310. 1836, in part as to synon. Bellis xanthocomoides (Less.) Gray in Hemsl. Biol. Cent.-Am. Bot. 2:118. 1881 Annual; stems several, striate, procumbent or ascending, branching, leafy throughout, producing slender runners; leaves spathulate, apiculate, sparsely pubescent, gradually reduced toward the peduncles, 1.5-2.5 cm. long, .3—.8 cm. broad; pe- duncles 1-6 cm. long, densely pubescent at the base of the invo- lucre; involucre .8-1.2 cm. in diameter; bracts 2-seriate, lanceo- late, acute, pubescent along the lower portion of the main axis, membranaceous-margined; ray-flowers white, .5-.8 cm. long, about 16; pappus practically lacking; achenes compressed, pubescent, becoming more or less glabrate. Distribution: xe meadows and open woods of east-central Mexico. Specimens exam Vera Cruz: prope la Jaya (La Hoya), June 29, Schiede 206 (M, TYPE). HripALGO: meadows of Sierra de Pachuca, alt. 2760 m., July 17-28, 1898, Pringle 6888 (M); alpine meadows, Sierra de Pachuca, alt. 2023 m., Aug. 26, 1902, Pringle 9858 (M); moist meadows and open woods, Pachuca, July 1905, Purpus 1844 (M). 6. Astranthium purpurascens (Rob.) Larsen, n. comb. Bellis purpurascens Rob. Proc. Am. Acad. 27: 172. 1892. Perennial; roots of numerous fibers; stems erect, branched, 4— 45 cm. high, striate, pubescent with spreading hairs; leaves apiculate, pubescent with spreading hairs, margins Bewteshat ciliate; lower leaves oblong-ovate, 2.5—4.0 cm. long, 1.5-6 cm. broad; upper leaves linear to linear-lanceolate, gradually reduced, 1 cm. or less in length at the base of the peduncles; peduncles 3-8 cm. long; involucre .8-1.0 cm. in diameter; involucral bracts about 14-16, lanceolate, acute, scarious-margined, only slightly [Vor. 20 34 ANNALS OF THE MISSOURI BOTANICAL GARDEN ciliate near the apex, pubescent along the main axis with a few upwardly appressed hairs; rays 10-15, whitish-purple, .5—8 cm. long; pappus lacking; achenes glabrous, golden-brown at matu- rity. This species is closely allied to A. integrifolium (Michx.) Nutt. ; the chief difference is to be found in achenial characters. Astran- thium integrifolium has achenes which are pubescent with glo- chidiate-tipped hairs whereas A. purpurascens (Rob.) Larsen has glabrous achenes. Distribution: known -— from the type locality and Chiapas, Mexico. Specimens examine San Luis Porosi: shaded grassy slopes, barranca of Las Canoas, Aug. 18, 1891, Pringle 3819 (M, coTYPE). Curapas: without definite locality, coll. of 1864—70, Ghiesbreght 548 (M). 7. Astranthium integrifolium (Michx.) Nutt. Trans. Am. Phil. Soc. N. 8. 7: 312. 1841. Bellis integrifolia Michx. Fl. Bor. Am. 2: 131. 1803; Hook. Bot. Mag. 52: pl. 3455. 1835, in part, exclusive of Brachycome zanthocomoides Less.; DC. Prodr. 5: 304. 1836; Raf. New FI. Am. 2: 24. 1836; Torr. & Gray, Fl. N. Am. 2: 189. 1842; Gray, Syn. Fl. N. Am. 1?: 163. 1884, and ed. 2, 1886 and 1888; Britton & Brown, Ill. Fl. 3: 350. 1898, and ed. 2, 3: 402. 1913; Britton, Manual, 943. 1901, and ed. 2, 1905; Rob. & Fern. in Gray, Manual, ed. 7, 799. 1908; Small, Fl. Southeastern U. S. 1202. 1903, and ed. 2, 1913. Bellis nutans Raf. New Fl. Am. 2: 23. 1830. Bellis parviflora Raf. Ibid. Eclipta integrifolia Spreng. Syst. Veg. 3: 602. 1826. Annual, openly branched from near the base, 8-45 cm. high, sparsely pubescent with spreading or subappressed hairs, fre- quently conspicuously pubescent near the base, leafy throughout; lower leaves oblong-spathulate, 2.5-4.0 cm. long, .8-1.4 cm. broad; stem-leaves oblong-ovate, sparsely pubescent, margins ciliate, 3.5 em. long, 1.5 em. broad, gradually reduced toward the base of the peduncle; peduncles 4-9 em. long, densely pubescent at the base of the involucre; involucre 2-seriate, about .6-1.2 cm. in diameter, bracts lanceolate, acuminate, acute, membranous margins only slightly lacerate, sparsely pubescent along the main 1933] LARSEN—ASTRANTHIUM AND RELATED GENERA 35 axis with a few spreading hairs; ray-flowers fertile, 16-22, pur- plish-blue; achenes compressed, pubescent with glochidiate- tipped hairs. Distribution: sandy soil, south-central Kentucky to northwestern Georgia, west- ward to southeastern Kansas and eastern Oklahoma. Specimens examin GroraIa: dry ra Ringold Road, Chickamauga, May 27, 1911, Churchill (M, No. 839094). Mississippi: Tchula, Holmes Co., Apr. 18, 1927, ves is Anderson 1555 (M). Kentucky: Bowling Green, May 6, 1892, Price (M, No. 122882 TENNESSEE: West Nashville, May 26-27, 1909, 2 4422 (M); moist field, Joelton, Davidson Co., July 16, 1922, Svenson 118 (G); copses around Nashville, May-June, Gattinger 1297 (M); rocky open hillsides, near Erin, Houston Co., May 24, 1920, E. J. Palmer 17623 (M, G Missouri: gravelly barrens, Noel, May 10, 1915, Bush 7534 (M, G); gravelly barrens, Noel, May 10, 1915, Bush 7534 A (M); rocky open ground, near Jane, MacDonald Co., May 23, 1931, E. J. Palmer 39297 (M); rocky Mà r bald knobs, along creek, near Oasis, Taney Co., June 3, 1931, E. J. Palmer 8 M). ARKANSAS: river bottoms near Fayettotille, June 1835, nci 129 [607] M); alluvial soil, waste places, cultivated fields, Fayetteville, May, Harvey 45 (M Fayetteville, 1880, Harvey (M, No. 122880); Hot Springs, May, 1879, Soulard (M, No. 122885); sandy soil, Oik, Marion Co., June 15, 1914, E. J. Palmer 5992 (M). Kansas: rocky soil, Chautauqua Co., Mar 7, 1897, Hitchcock 1056 (M). OKLAHOMA: in woods and BREL. common near Tishomingo, Johnston Co., Apr. 27, 1916, Houghton 3549 (G); clay washes, near Ardmore, Carter Co., Apr. 17, 1913, Stevens 77 (M); Arbuckle Mts., Crusher, May 12, 1916, Emig 606 (M); Catoosa, May 8, 1895, Bush 897 (M); Ind. Terr., 1877, Butler (M, No. 122886); upland prairies, sandy soil, near Howe, Le Flore Co., May 25, 1931, E. J. Palmer 39343 (M). 7a. Var. ciliatum (Raf.) Larsen, n. comb. Pl. 4, fig. 1. Bellis ciliata Raf. New. Fl. Am. 2: 24. 1836. Bellis integrifolia Gray, Smiths. Contr. [Pl. Wright. pt. 2] 5:78. 1853, in part. Stems several, diffusely branched from near the base, 1-3 dm. high, terminated by long slender peduncles; leaves oblong- spathulate to linear-oblong, reduced to small bract-like leaves at the base of the peduncles, .8-3.0 cm. long, .2-.8 cm. broad, sparsely pubescent with subappressed hairs, margins ciliate; peduncles 4-6 cm. long; involucre .5-.8 cm. in diameter; in- volucral bracts lance-elliptic, acute, the narrow Tina scous margins lacerately ciliate; ray-flowers about 9-15; achenes pubes- cent with glochidiate-tipped hairs. Distribution: from southeastern Oklahoma and southwestern Arkansas to south- central Texas, west to the Pecos [Vor. 20 36 ANNALS OF THE MISSOURI BOTANICAL GARDEN Specimens examined: ARKANSAS: moist open ground, Fayetteville, Washington Co., July 6, 1915, E. J. Palmer 8176 (M); Coal Bank, May 14, 1895, Bush 895 (M); Redfork, May 14, 1895, Bush 896 (M). OKLAHOMA: open woods near Idabel, McCurtain Co., May 18, 1916, Houghton 3645 G). Texas: Apr.-May, 1844, Lindheimer 251 (M); “Pine’s Island," May 5, 1903, Rever- chon (M, No. 122889); Reverchon 440 in part (M, No. 122890); clay barrens, Bryan, Brazos Co., May 27, 1915, E. J. Palmer 7785 (M); Pecos and Limpio, June 1851-2, Wright 1176 in part (M, No. 123093); dry banks, Hempstead, Apr. 24, 1872, Hall 3806 (M). 7b. Var. rosulatum Larsen, n. var.? Pl. 4, fig. 2. Bellis integrifolia Gray, Smiths. Contr. [Pl. Wright. pt. 2] 5: 78. 1853, in part. Stems several, sparsely branched near the base, pubescent with spreading hairs, .5-1.5 dm. high; leaves sparsely pubescent with subvillose hairs; basal leaves numerous, forming a rosette, oblong- spathulate, 2-7 em. long, .5-1 cm. broad; stem-leaves oblong- spathulate to linear-lanceolate, gradually reduced to the base of the peduncle; peduncle 1.5-5 cm. long; involucre about .7 em. in diameter; bracts lance-elliptic, acute, the narrow membranaceous margins somewhat lacerately ciliate near the apex; ray-flowers about 16; achenes densely pubescent with glochidiate-tipped hairs. Distribution: — Oklahoma and eastern 'Texas west to Presidio County. — exa OKLAHOMA: open sepas near Mannsville, Johnston Co., May 16, 1916, Griffith 3455 (M, G); pnis cu Gap, Apr. 16, 1877, Butler 63 (M); Sapulpa, Apr. 28, 1895, Bush 927 (M); * Arkansas," May 20, 1895, Bush 929 (M). Texas: sandy prairies, Columbia, Brazoria Co., March 25, 1914, E. J. Palmer 5028 (M); Vietoria, Vietoria Co., March 5, 1916, E. J. Palmer 9070 (M); sandy open ground, Larissa, Cherokee Oó., Apr. 7, 1916, E. J. Palmer 9388 (M); sandy prairies, Matagorda, Matagorda Co., March 5, 1914, E. J. Palmer 4856 (M, No. 753171, TYPE); on plains, Handley, Apr. 15, 1913, Ruth 74 (M, G); along railroads near Houston, Apr. 23, 1899, Eggert (M, Nos. 122893 & 122895); Terrell, Apr. 6, 1903, Reverchon 4006 (M); prairies near Victoria, Apr. 7 & 10, 1900, Eggert (M, No. 122899); Pecos and Limpio, June 1851-2, Wright 1176 in part (M, No. 123091); near San Felipe, 233 Astranthium integrifolium (Michx.) Nutt. var. rosulatum Larsen, n. var., caule simplice vel basi ramoso, 0.5-1.5 dm. alto, subvilloso; foliis basalibus rosulatis, oblongo-spathulatis, 2-7 cm. longis, usque ad 1 cm. latis, integris, ad apicem rotun- datis vel obtusis, sparse pubescentibus; foliis caulinis oblongo-spathulatis vel lineari- —— integris, superne sensim minoribus; pedunculis 1-6 em. longis.— TExas: sandy prairies, Matagorda, Matagorda County, March 5, 1914, E. J. Palmer 4856 (M, 753171, TYPE). 1933] LARSEN——ASTRANTHIUM AND RELATED GENERA 37 Apr. 1839, Lindheimer (M, No. 122904); along river, Columbia, Apr. 19, 1899, Bush 195 (M); on prairie, Chenango Junction, Apr. 18, 1900, Bush 60 (M); Reverchon 440 in part (M, No. 122898); open ground, along railway grade near Marfa, Presidio Co., June 18, 1926, E. J. Palmer 31036 (M). KEERLIA Keerlia Gray, Bost. Jour. Nat. Hist. [Pl. Lindh. pt. 2] 6: 221, 222. 1850, in part as to K. bellidifolia and K. effusa, not DC.; Smiths. Contr. [Pl. Wright. pt. 1] 3: 92. 1852, not DC.; Syn. Fl. N. Am. 1?: 164. 1884, and ed. 2, 164. 1886 and 1888. Bourdonia Greene, Erythea 1: 207. 1893. Herbaceous, caulescent, pubescent annuals or perennials. Leaves alternate, sessile or petioled, oblong to obovate-spathulate, entire. Involucre of imbricated lanceolate bracts, with mem- branaceous margins. Heads few-flowered. Ray-flowers in a single series, pistillate, fertile. Disk-flowers tubular, five-lobed, frequently sterile. Branches of the style lanceolate, acutish, hairy toward the tip. Pappus a thickened ring-like or slightly lacerated crown. Achenes compressed, glabrate or hirsute- pubescent. Type species: K. bellidifolia Gray & Engelm. Proc. Am. Acad. 1: 47. 1848, which was based on No. 415 of Lindheimer, collected above Guadeloupe, Texas, 1845-1846. KEY TO THE SPECIES Stems branched from the base... .....ccds ccc dnioe cccscece 8. K. Bigs od pa Stems simple to the inflorescence................0 cee cece ceeneues 9. K. effusa 8. Keerlia bellidifolia Gray and Engelm. Proc. Am. Acad. 1: 47. 1848; Gray, Bost. Jour. Nat. Hist. [Pl. Lindh. pt. 2] 6: 220. 1850; Smiths. Contr. [Pl. Wright. pt. 1] 3: 92. 1852; Syn. FI. N. o 1:: 164. 1884, and ed. 2,164. 1886 and 1888. Bourdonia bellidifolia (Gray & Engelm.) Greene, Erythea 1: 207. 1893. Annual; the caudex giving rise to much-branched stems 9—30 em. high, leafy throughout, striate, pubescent with multicellular spreading hairs; leaves thin, sparsely pubescent, apiculate, lower leaves obovate-spathulate, uppermost somewhat linear, gradually reduced toward the inflorescence, 1—4 cm. long, .4-1.0 em. broad, narrowed below to a subpetiolate base; heads small; involucre [Vor. 20 38 ANNALS OF THE MISSOURI BOTANICAL GARDEN .3-.5 em. in diameter, bracts 2-3-seriate, glabrous, linear-lanceo- late, acute, with membranaceous margins; ray-flowers 4-15, blue, fertile; disk-flowers 15-20, frequently sterile, style branches shorter than those of the ray-flowers; pappus a thickened ring- like crown; achenes nerved, compressed, glabrate. Distribution: southern Texas; doubtless also in adjacent Mexico. Texas: along Nueces Bay, Nueces Co., alt. 6 m., March 12, 1894, Heller 1436 (M); Rock Springs, Apr. 17, 1930, Marcus E. Jones 26575 (M); above Guada- lupe, 1845-46, Lindheimer 415 (M, co-rvrE); near New Braunfels, May 1848, Lindheimer 628 (M); Comanche Spring, New Braunfels, Apr. 1850, Lindheimer 932 (M); rocky calcareous ground, Del Rio, Valverde Co., March 26, 1917, E. J. Palmer 11384 (M); dry limestone hills, Concan, Uvalde Co., Apr. 13, 1917, E. J. Palmer 11550 (M); vicinity of San Antonio, Apr. 1919, von Schrenk (M, No. 88063); near Austin, March 17, 1908, York 410 (M). 9. Keerlia effusa Gray, Bost. Jour. Nat. Hist. [Pl. Lindh. pt. 2]6:222. 1850; Smiths. Contr. [Pl. Wright. pt. 1] 3: 93. 1852; Syn. Fl. N. Am. 1?: 165. 1884, and ed. 2, 165. 1886 and 1888. Bourdonia effusa (Gray) Greene, Erythea 1: 207. 1893. Perennial, 15-60 cm. high; stem simple below, diffusely branched above into an open glabrous panicle; stem and leaves hispid- pubescent; leaves thin, lowermost leaves obovate-spathulate, 3-9 cm. long, .5-1.3 cm. broad, narrowed at the base into a petiole, those of the stem oblong with a broad, sessile, somewhat amplexicaul base, gradually reduced toward the inflorescence; heads numerous, small; involucre turbinate, 2-3 cm. in diameter; involucral bracts glabrous, 3-4-seriate, linear-lanceolate, acute, membranaceous-margined; ray-flowers 4-7, white, fertile; disk- flowers 6-9, sterile; pappus consisting of a thickened irregularly lacerate crown; achenes compressed, hispid-pubescent. Distribution: south-central Texas. Specimens examined: Texas: Berlandier 499 (M); White Oak Creek, Gillespie Co., Jermy 815 (M); shady banks of the Upper Guadalupe, 50 miles above New Braunfels, 1847-48, Lindheimer 629 (M, co-ryre); dry hillside thickets, Leakey, Edwards Co., June 10, 1916, E. J. Palmer 10165 (M); dry hillside thickets, Spanish Pass, Kendall Co., Sept. 28, 1916, E. J. Palmer 10839 (M); San Geronimo Creek, June 1884, Reverchon 1534 (M); Comanche Spring, New Braunfels, 1849, Lindheimer 933 (M). 1933] LARSEN—ASTRANTHIUM AND RELATED GENERA 39 List or ExsiCCATAE The distribution numbers are dii in italics. The number in parenthesis is the species number used in this revisio Berlandier, J. L. 499 Bush, B. F. 897, HA 754A (7); 896, 896 (7a); 60, - xdi 7b). Churehill, J J. R. -— ji Eggert, H. — (7b). Eggleston, W. M 4422 (7). Emig, W. H. 606 (7). Engelmann, " p [607] (7). Fisher, G. L. — (4 Gattinger, A. 1297 (7). Griffith, F. 3455 (7b). Hall, Elihu. 306 (7a). me C. V. 623 (3). Harvey, F. L. —, 45 (7). Heller, A. A. 1436 M. Hitchcock, A. S. 1055 Houghton, H. W. 3549 na 3645 (Ta). Jermy, Gustav. 815 (9). Jones, Marcus E. 26575 (8). Lindheimer, F. 251 (7a); 628, 932 (8); T 933 (9). Nelson, E. W. 4786 (1). Orcutt, C. R. bos (4). —, (7b); 416, Palmer, E. J. 5992, 17623, 39297, 39343, 39483 (7); 7785, 8176 (Ta); 4855, 5088, 9070, 9388, 31036 (Tb); 11384, 11550 (8); 10165, 10839 (9). Palmer, Edward, 146 (2); 163, 287, 425 (3). Price, Sadie F. — Pringle, C. G. 2); 4719, 6442, 9857 nt 2016 AS 6888, 9858 (5); 3819 Purpus, A. 159, 1575, 8640 (4); 1344 Reventhaty J. —, 440 in t (7a); — (7b); 4006, 440 in part (7b); ; 1584 (9). Ruth, Albert. ?4 (7b). Schiede, G. 206 (5). Smith, Charles L. 261 (4). Svenson, H. K. 118 (7). von Schrenk, H. (8). Woodson, R. E. Jr., and Anderson, E. 8. 1666 (7). e Charles. 1176 in part (7a); 1176 in rt (7b Yo H. H. 410 (8). [Vor. 20, 1933] ANNALS OF THE MISSOURI BOTANICAL GARDEN PLATE 2 Astranthium — (Michx.) Nutt. from Eggleston 4422 in the Missouri Botanical Garden Herbar 918135 Involucre d receptacle x 5. Style-branches of the disk-flower, X 10. Style-branch greatly enlarged. Bellis perennis L. from Pring, in the Missouri Botanical Garden Herbarium No. Involucre and receptacle, X 5. Ray-flower, X 10. Disk-flower, X 10. Stamen, X 10. Style-branches of the disk-flower, X 10. Style-branch greatly enlarged. ANN. Mo. Bor. Garb., Vor. 20, 1933 PLAYE 2 Sys he . ce e. LA nm ^ "js "^s o y ae 9 Tu 7, E31 Nu NÉ T^ Pal 4 | gu | d X9 ff i” + : / / / j IH pH LARSEN—ASTRANTHIUM AND RELATED GENERA [Vor. 20, 1933] 42 ANNALS OF THE MISSOURI BOTANICAL GARDEN PLATE 3 Astranthium mexicanum (Gray) Larsen var. chihuahuense Larsen. From the type specimen, Pringle 2015, in the Missouri Botanical Garden Herbarium. ANN. Mo. Bor. GARD., Vor. 20, 1933 ^ PLATE 3 \ CEE Uu «Ira? uate, a4 4° Yen ate 3 C. G. PRINGLE, PLANTA MEXICAN Æ. 1858, TATE CP CEIRTANA | a 2015 Aphantostephus Arkansanus, Gray. Canyons of Seren Waites 4, fhteter LARSEN—ASTRANTHIUM AND RELATED GENERA [Vor. 20, 1933] 44 ANNALS OF THE MISSOURI BOTANICAL GARDEN PLATE 4 Fig. 1. Astranthium integrifolium (Michx.) Nutt. var. ciliatum (Raf.) Larsen. From Lindheimer 251, in the Missouri Botanical Garden Herbarium. Fig. 2. Astranthium integrifolium (Michx.) Nutt. var. rosulatum Larsen. From the type specimen, Palmer 4855, in the Missouri Botanical Garden Herbarium. PLATE 4 ) D 0, 193: 2 ANN. Mo, Bor. GARD., Vor. d Au on fie yer Pacman imd Sadun 1marta iia A gat, b ge tats * Pa re tan ^x * * Flora Texana exsiceata. $251. Foe 4) Jag V. Lidicinse, 1541 LARSEN—ASTRANTHIUM AND RELATED GENERA FERNS AND FERN ALLIES OF MISSOURI! M. ELIZABETH PINKERTON Assistant in Botany, Henry Shaw School of Botany of Washington University This paper is based primarily upon material in the herbarium of the Missouri Botanical Garden, the University of Missouri, and the private collections of Mr. John H. Kellogg and of the author. After each specific name reference is given to the original publication; this is followed by reference to the two standard manuals and to an illustrated monograph covering the fern-flora of the region concerned, where the species or variety is treated under the same or a different name. These are: Gray, ‘New Manual of Botany,’ ed. 7, 1908; Britton and Brown, ‘Illustrated Flora of the Northern States and Canada,’ ed. 1, 1896, and ed. 2, 1913; and Eaton, ‘Ferns of North America,’ vol. 1, 1879 (plates 1-45) and vol. 2, 1880 (plates 46-81). Time of fruiting, the general distribution for North America, and the habitat for Missouri in particular precede the specific description. The names of the principal collectors whose material was ex- amined are listed, and definite citations made when the plant is rare. Key To FAMILIES 1. Plants with short vertical stems or rootstocks. .............. eee 2 1. Plants with horizontally spreading stems or rootstocks...............-00+065 4 2. Leaves onion-like, producing sporangia at their bases............... Isoetaceae 2. Leaves neither onion-like nor producing sporangia at their bases............... 3. Sporangia borne on under sides of leaves................ sese. Polypodiaceae 3. Sporangia borne terminally in special fruiting clusters.......... Ophioglossaceae 4. Sterile leaves large, usually compound or variously divided.................. 5 4. Sterile leaves small, often acale-lik®, 5.2.02. ccc c ccc ce cece ecb oeseseberseues 6 5. Fertile se leaf-like or pod-like; if the latter, sporangia covered with Inque sla ei P MPA r ei ice ara 8 (Sie Polypodiaceae 5. Fertile segments 1 not leaf-like; sporangia naked................ Osmundaceae 6. Plants: truly quaie......... REA EREAMERA REA e rrr RR Salviniaceae 6. Plants terrestria " aba do dre ot 0 ono 6g c Cor oU Coo yp cL Sin oe 7 T: Stems jointed; sporangia clustered underneath the scales of terminal cone- (ee E T ET Equisetaceae 7. Stems not jointed; sporangia borne on upper surfaces of leaves............... ! This article is a portion only of a thesis submitted to the Board of Graduate Studies of livor aed Ho dide in partial fulfillment for the requirements of the degree of master of sci Issued April 29, 1933. ANN. Mo. Bor. GARD., Vor. 20, 1933. (45) [Vor. 20 46 ANNALS OF THE MISSOURI BOTANICAL GARDEN 8. Bpores all of to mine Mid. i... eoe error Lycopodiaceae 8. Spores of two kinds (larger macrospores and much smaller microspores)... . . edith eS ON ERA DONORS ET A Wee E Rhode E Selaginellaceae OPHIOGLOSSACEAE 1. Veins free; sporangia separate in compound inflorescences.......... Botrychium 1. Veins reticulate; sporangia disposed in a solitary spike.......... Ophioglossum Fig. 1. Key map to distribution chart (fig. 2) showing location of counties in Missouri. OPHIOGLOSSUM 1. Sterile segments obtuse, usually solitary; middle areoles long and narrow, outer hexagonal and containing a few included veinlets...... 1. O. vulgatum 1. Sterile segments cuspidate, 2-5; areoles wide and containing many anasto- mosing TARNEN 5 isi i443 EAR RAAOAOALARXo EE EAE 0455058785 2. O. Engelmanni 1. Ophioglossum vulgatum L. Sp. Pl.2:1062. 1753. Gray, p. 47; Britton & Brown, ed. 1 and 2, 1: 2. ATCHISON HOLT ES 2 a A e o mn a el H Bash a obey ¥f zü "EE: Q = 5 * 5 j E Eod t 5 san un 2 = z a <= E» z o E PIRE INIM FEE EERTE FEN TPAR: Saan agied aaps guanako deb 533 3,3 LOEEERE TIFE ETET EFEFED E ERE FFEEREREE TIFE gas Bee ec Gre Sa x -| EERERCEE RET EDELLLLLERELPLELLLLEEEEHEEEEEETEEPEEUEEUSEEEEEEEEEE FEE E FH E 8298S PEELE EP Er FEFEFE TEF l Adiantum a TF ae die te ef, + ET + +t + T ure + g ripe TECHS dot t + ++ ttti + Eb ot + + seco? T H * 3 Asplenium” Bradley! + + + . + : " iL nts + y + i" + ebenoides 6. ae pinnatifidum + 4 ( Dp Ge! platyneuro he a Aha BE + * toto +++ ++ do Bog, + MuR. esiliens ar + 4+ a + t ++ + ] Trichomanes Tt A - F 44 + + het ) Athyrium acrostichoides el + + $ + angus ex lium + T T + + ¥ + iz + $ + + + $ RWE angust EHE d. * + + 1 eue phia iA + + + ++ +++ + + T tpar + + trycb um ob mee A 2^ + + FA + , diss x aA; s temifolius t ES pU rginian ys le i Bl i + Piece Mop ++ MR = NE n d 4 +4 ++ + ff OR rAlpophyline es eo + + + + +++4+4+4+ GF ms Mette: - x rmm + + +++ + ) Gheilanthes alabamensis Pieces T 3 4 Lt t AE t ml i + + LAUS + + : lanosa + b+ + + ttt + Js: oae $ FEEF 4 5 oystopt eris bulbifera + +44 ++ + + + fe + T fe ark + + ide + i fragili ttt. B, ET ++ tto +44 ++ + woe rS Ax + + +++ + Xx j Donnøtaoatsa” puneti lobule + 26 Ea uis etum EFFEEF "m ++ + TRIN + T" 27 incid 1 + + =- + a + 28 praealtum Aat" d ++ +4 ++ ee grits Ms + 29 isoetes Butleri + 30 rae pits ge ; i Š E melan poda : 32 Lrgopoaiin luein ++ + + 33 broken iiber £ ve ^ 34 Ni io A A $5 yothetee na dealbata Emi T + + + + + ++ E+ + > 4 36 Ophioglossum mentenek mao he + F + + F 4- 37 ep huir + = 38 Osmunda dene ttt + + B $9 Claytonian ps T ++ + 40 galis spectabilis t + + + t t tt + — 41 Pellsos atropurpur Tt tkt + err 6s ee + je t —++ d 42 labella Eq. Rt Ne d + + tE + + + HH 43 Polypodium Le rei Ttt e + + i ortt ++ +++ iia 4 rginianum +++ + - + 45 Polys blend ror on PE H--oc-FEFATLFLGMA + + + + +t + +t+tttt ++4+ + Ta ab +44 + 46 Pteretis nodulosa + ; 47 Pteridium letiusculum NN. + It e t 48 Len X 3elaginella rupestris EA EER kp ; s r FT. i. + 4 [ z 50 Thelypteris Da seRoppu M : 4- Vu 51 arginalis Bas + * ENT ae + t + + a $2.4, n palustris. ubescens TU + : + 59. sp nulosa 2 : 54 in t di 3 55 Wo MEA dede sn -tCkRR TEM + + ^ P bees. + a + — x 56 woodwardia areolata WAR E ; Fig. 2. Chart showing distribution of posi by counties: — signifies reported only; +, specimens seen. vo t+ + + + Me ++ + ++ +4 tt + 1933] PINKERTON—FERNS OF MISSOURI 47 Rootstock short, oblique or erect; common stalk half or more above ground, constituting 1/3 to 2/5 the length of the plant; sterile segments sessile, ovate, 1-5 inches long, 0.5-2 inches broad; basal veins about 9; sporophyll 0.75-2 inches long, apicu- late. May to August. Distribution: Nova Scotia to momi bentii to Missouri and Texas. South- eastern Missouri—swamps or moist me Specimens examined: Palmer 14716, 1 ges 1 4748, 14766. 2. Ophioglossum Engelmanni Prantl in Ber. Deut. Bot. Ges. 1: 351. 1883. Gray, p. 47; Britton & Brown, ed. 1 and 2, 1: 2. Rootstock cylindrical; common stalk mostly below ground, sheathed by persistent leaf-bases; sterile segments sessile, ellip- tic to ovate, 1-3.5 inches long, 0.5-2 inches broad; basal veins 13 or more, transverse ones oblique; sporophyll 0.5-1 inch long, apiculate. March to October. Distribution: South Carolina to Florida, westward to Arizona. Limestone glades, especially where there is a thin layer of fine rich and damp soil. Specimens examined: Palmer, Kellogg, Letterman, Engelmann, Bush, Pinkerton, Eggert, Greenman. BOTRYCHIUM . Bud wholly enclosed; cells of epidermis straight; sterile blade petiolate and ever over 7 inches broad; common stalk mostly underground............ _ 1. Bud exposed along one side at base of rootstock; cells of epidermis flexuous; sterile blade sessile, 2-16 inches broad and nearly as long; common stalk above ground, half the length of plant................. ec cece teen eens 4 2. Ultimate segments deeply laciniate................. 2. B. obliquum var. dissectum 2. Ultimate segments serrulate-dentate. ......... ccc cece cece cece nnn 3 3. Plant coarse; common stalk very shave underground; frond E ee aaiae 655 os 6 5 osc s o's 0 0 0 ohh eee 1o 11 RS 1. B. obliquum 3. Plant slender; common stalk up to 2 inches above ground; frond coriaceous on drying. iesnas sear cine t EERERRES 8. B. obliquum var. Weit 4. Ultimate segments T ripe sporangia mtraw-colored, openi but slightly in dehiscence.............. 5. B. virginianum var. i A 4. Ultimate segments acute; pa sporangia brown, opening widely in dehiscence. TC TUMOR CES a Re cee 4. B. virginianum 1. Botrychium obliquum Mühl. in Willd. Sp. Pl. 5: 63. 1810. Gray, p. 48; Britton & Brown, ed. 1, 1: 3, ed. 2, 1: 5; Eaton, pl. 20, fig. 2. B. ternatum Sw. in Schrad. Jour. für die Bot. 2 (1800): 111. 1801 B. ternatum var. obliquum D. C. Eaton, Ferns N. Am. 1: 149. 1879 [Vor. 20 48 ANNALS OF THE MISSOURI BOTANICAL GARDEN Plant 5-15 inches tall; bud pilose; sterile blade long-stalked, 2-5 inches broad, tripinnatifid or tripinnate below; ultimate segments obliquely ovate or oblong-lanceolate, acutish terminal ones elongate; sporophyll long-stalked, usually stout, several- pinnate. August to October. Distribution: Maine to Alabama, westward to Michigan and Texas. Moist woods, neutral soil. Specimens examined: Palmer, Bush, Kellogg, Letterman, Eggert, Trelease. 2. Botrychium obliquum var. dissectum Prantl in K. Bot. Gard. Berlin, Jahrb. 3: 342. 1884. Gray, p. 49; Britton & Brown, ed. 1, 1:3, ed. 2, 1: 5; Eaton, pl. 20, fig. 1. B. dissectum Spreng. Anleit. 3: 172. 1804. B. ternatum var. dissectum D. C. Eaton, Ferns N. Am. 1: 150. 1879. Character of rootstock, fruiting stalks, and texture of the plant all similar to the species; frond subternately divided, basal divisions unequally and broadly deltoid, decompound; the upper and secondary pinnae deltoid-lanceolate, pinnate, with laciniate or deeply cut pinnules; the ultimate divisions divergent, narrow, and incised. Distribution: Maine to Florida, westward to Illinois and Missouri. Rich moist woods, deeply sha Specimens examined: Eggert, Letterman, Pinkerton. 3. Botrychium obliquum var. tenuifolium (Underw.) Gilbert in Fern Bull. 11:99. 1903. Gray, p. 49. Pl. 5, fig. 2. Plant small, up to 12 inches high; common stalk up to 2 inches high, not all underground, slender; blade ternate with few divi- sions, 5-10 em. broad and about 5 em. long; leaves coriaceous; ultimate segments broad, about 1.5 cm. long, ovate, acutish, conspicuously serrulate. September and October. Distribution: Virginia to Arkansas. Southern Missouri—swampy land. Specimens examined: Dunklin Co., Bush 134, and coll. of Sept. 28 and Oct. 28, 1897, T'release; Butler Co., Bush 3110. 4. Botrychium virginianum (L.) Sw. in Schrad. Jour. für die Bot. 2 (1800): 111. 1801. Gray, p. 49; Britton & Brown, ed. 1, 1: 4, ed. 2, 1: 6; Eaton, pl. 33. Common stalk slender, above ground 1/2-2/3 the length of 1933] PINKERTON—FERNS OF MISSOURI 49 plant; blade sessile or nearly so, membranaceous, ultimate segments toothed; sporophyll long-stalked, bi-tripinnate. June and July. Distribution: New Brunswick to Alabama, westward to Idaho and Texas. Rich mois woods. mens examined: Davis, Engelmann, Letterman, Palmer, Bush, Pinkerton, Eggert, "Kellogg, Trelease, Larsen, Daniels, Emig, Blankenship. 5. Botrychium virginianum var. intermedium Butters in Rho- dora 19: 207. 1917. Ultimate segments of the sterile frond spatulate, penultimate ones ovate, not crowded; segments of the fertile frond opening wide in dehiscence; sporangia straw-colored, up to 0.8 mm. long. Specimens examined: Monteer, Shannon Co., Bush 4724; Whiteside, Lincoln Co., Davis, coll. of 1910. The writer feels that the species is so generally variable that this variety is not very distinct. POLYPODIACEAE I. Indum lacking. o.oo si eu i serva ng DD rie rx Re GO ee i à 2 1. Indusia present, at least in carly stages, ee cog oka ss os) shen cose tenons 4 2. Fruit dots near margin, becoming panne somewhat protected by over- lapping leaf margins; lower surfaces of leaves covered with a white powder. "PUE rn ee Gaye otholaena 2. Fruit dots separate, era absolutely bores COWPOUPNFP Iur 3 3. Fronds linear, once-plunatid. ......... DERI Polypodium 3. Fronds triangular, more Ph once- pinnata e. qua sns ss enda ria Thelypteris 4. Fronds dimorphic E- fronds not MG eu aes + > she vi a eee R PA 5 4; Fronds monomorgnhie « . se- «5.6. s.s-ccdeualege en SO Bees 2 5.6 5s 6 na AE eens 6 5. Sterile fronds pin innatiid; veins anastomosed; fertile segments bipinnate.Onoclea 5. Sterile fronds bipinnatifid; veins free and unbranched; fertile segments pinnate .eiscooesiocosv-XARRPTTRIBRSRENE vies C++ dys eo Pteretis 6. Indusa MARINA: . «6.6 ns ness 0 ears ne Eade ARES er ERR Fen ke 6. Indusia not marginal. ...... a ETRAS IU. Las cha Sees ens 11 7. Indusia not formed of revolute margins, but cup-shaped and opening ter- Pcie | socero co ede eR IS EE cen ennstaedtia 7. Indusia formed in part at least by revolute margins. .......... eese 8 8. Indusia definitely interrupted, occurring in is little rounded flaps... DUMMIES EE ur ddinnium 8. Indusia continuous or only slightly broken. .......... leen 9. Indusia double... ... 60620. vs oss ss sen eee. esos es) E Pteridium 9. Indusia single... cscs cap REA ees RR 10. Blades with a few large, relatively smooth segments.............. s. Pellaea 10. Blades with many small, usually tomentose or hairy segments. ..... m I1. Bori elongated... eode eorr oie 8 RRS Exod Rn ERA rrr IL Bori rounded... 50s eas vs cdo + 05605 eee eer RD eee ES. [Vor. 20 50 ANNALS OF THE MISSOURI BOTANICAL GARDEN 12. Sori dispersed irregularly on lateral veins... 13 JE vM MM Fri Feria. U E HDADUE ORA CR CES RO Camptosorus IM TE DNE Re T ee en en we ee ee yor eS 14 a ee X Aa ACE ee: eee ee A Asplenium 14. Sori mostly curved over the end of veins..................0000 000s Athyrium rr ene as Pe ree ee ree ta ae re es 16 15. Indusia attached at base or at one side... 2.1.6... 17 16. Indusia orbicular, attached centrally.............00....0000 0000. Polystichum 16. Indusia reniform, attached at sinus... eese Thelypteris 17. Plants never glandular, but scaly on stipes and rootstocks; segments more or - hein indusia tapering, attached at one side and becoming obscure M MAN LrrruaeQureetkgtrrCOra a REAREXSAEXA4 xS ORIREERM PES Cystopteris 17. Plants ver d also scaly; segments obtuse; indusia entire when young, splitting at top into several stellate segments.................... Wo ADIANTUM 1. Main rachis unbranched..............0. 00.000 cee eee eee 1. A. Capillus-Veneris 1. Main rachis dichotomously branched with 4-5 pinnae on each side. .2. A. pedatum 1. Adiantum Capillus-Veneris L. Sp. Pl. 2: 1096. 1753. Gray, p. 36; Britton & Brown, ed. 1, 1: 27, ed. 2, 2: 31; Eaton, pl. 87. Venus hair. Rootstock creeping, chaffy; stipe very slender, 3-12 inches long, black or brownish, slightly scaly; frond ovate-lanceolate, 6-12 inches long, 4-12 inches wide at base; pinnules wedge- obovate or rhomboid, long-stalked, glabrous, membranaceous, margins variously incised, veinlets flabellately forking from base; fruit dots lunate or transversely oblong. June to August. Distribution: New Jersey to Florida, westward to South Dakota and California. Moist rocky places, ravines, wet limestone cliffs. Specimens examined: Kellogg, Daniels, Bush, Palmer, Trelease, Emig, Shepard. 2. Adiantum pedatum L. Sp. Pl. 2: 1095. 1753. Gray, p. 35; Britton & Brown, ed. 1, 1: 27, ed. 2, 2: 31; Eaton, pl. 18. Maidenhair. Rootstock long, creeping, chaffy; stipe 9-18 inches long, shining, dark brown to black, slightly scaly at base, once-forked at summit, each division bearing on one side only several pinnate divisions (occasionally tri-forked); blade reniform-orbicular, 8-18 inches broad, membranaceous, glabrous; pinnules short- stalked, oblong, triangular or end ones fan-shaped; lower margin entire, all veins proceeding from it, upper margin lobed. June to August. 1933] PINKERTON—FERNS OF MISSOURI 51 Distribution: Nova Scotia and Quebec to Georgia, westward to Alaska and California. Moist rich woods. Specimens examined: Bush, Davis, Merge Palmer, Trelease, Pinkerton, Woodson, Kellogg, Broadhead, Emig, Macken: ASPLENIUM 1. Blades pinnatifid, or only lower segments pinnate; apices long-attenuate........ 2 1. Blades 1-3-pinnate; apices not long-attenuate. ......... selle 3 2. Blades membranaceous; lower midribs black and shining on under side; apices crenate, sometimes proliferous..................000005 8. A. ebenoides 2. p E ei xs green, herbaceous; prolongations sinuous- BL NOt OOM NOUN es hx i cance EEEN Xt E uo 4. nnatifidum 3. Blades Pak only; pee of regular shape; margins not deeply Fahey CCC HBC H HCCC ETO HHS SPO HHH CTH FE HHC OCHO HH EHH HOH HHC CHE HHH HH ETHEH HHH HAE OEE Sousa os kk ic kN WEEE E 64 nd + err OU Oe RR TL 4. Pinnae mostly roundish, not auriculate...................04. 7. A. Trichomanes 4. Pinnae oblong or lanceolate, auriculate.... 2.2.0... ec ce ee cece ee ees 5 5. E and rachises shiny black, slender; pinnae mostly opposite; sori few and om 000 than COBE. DEEP DES ^ . A. resiliens 5. Stives pe rachises shiny "reddish brown, coarse; pinnae m sitéciinte; RH numerous HORT COSlAO, . oos or rn RENS . A. platyneuron 6. Stipes and rachises green throughout; ultimate segments ew cuneate; mendo inbrisie....... oce ono tae eee ere De xs A. cryptolepis 6. Stipes and rachises shining chestnut-brown; lower pinnae bd into obtuse sumentes; margins cronate. .. s. S. QUITE s erede 1. A. Bradleyi 1. Asplenium Bradleyi D. C. Eaton, Bull. Torr. Bot. Club 4: 11. 1873. Gray, p. 39; Britton & Brown, ed. 1, 1: 26, ed. 2, 1: 30; Eaton, pl. 51, figs, 4-8. Rootstock short, covered with narrow acuminate blackish- fuscous scales; finds 4-7 inches high, oblong-lanceolate; stipes 2-3.5 inches long, dark chestnut and shining, tufted, slender; rachises brown, or green above; pinnae numerous, lower ones no larger than the middle ones, obtuse or acutish, toothed, in the largest fronds pinnatifid into oblong lobes which are toothed at apices; sori short, borne near midveins, becoming confluent; indusia membranaceous, persistent. July to September. Distribution: New York to Georgia, westward to Missouri and Arkansas. On sandstone or chert outcrops—comparatively rare and local Specimens examined: Palmer, Bush, Trelease, Mackenzie, Greene, Pinkerton, Van Dugen, Shepard 2. Asplenium cryptolepis (L.) Fernald in Rhodora 30: 37. [Vor. 20 52 ANNALS OF THE MISSOURI BOTANICAL GARDEN 1928. Gray, p. 39; Britton & Brown, ed. 1, 1: 25, ed. 2, 1: 29; Eaton, pl. 15, fig. 1. Asplenium Ruta-muraria L. Sp. Pl. 2: 1081. 1753. Rootstock short, creeping, entangled, tufted; fronds 1-6 inches tall; stipes and rachises entirely green or slightly brown at base; blades deltoid-ovate, smooth, subcoriaceous, bi-tri- pinnate; ultimate segments few, stalked, 3-14 mm. long, narrowly cuneate to roundish obovate; margins deeply fimbriate; veins flabellate, no midveins; sori few (2-4 per pinna), oblong, covering whole segment when mature; indusia delicate with ciliated margin. July to September. Distribution: Vermont to Georgia, westward to Illinois and Missouri. Shaded limestone cliffs—scarce. Specimens examined: T'release, Palmer, Bush, Russell. 3. Asplenium ebenoides R. R. Scott, Berkeley in Roy. Hort. Soc. Jour. N. S. 1: 137. 1866. Gray, p. 38; Britton & Brown, ed. 1, 1: 23, ed. 2, 1: 26; Eaton, pl. 4, fig. 2. Rootstock short, creeping, chaffy; fronds up to a foot high; stipes tufted, 4-9 inches long, young ones reddish brown, older ones black, shining, slender; lower rachises dark and shining underneath; blades firm-membranaceous, triangular-lanceolate, variable, 3-12 inches long, 1-3 inches at base, tapering to a long- acuminate apex which may become proliferous, lowest divisions distinct, shorter; sori numerous throughout, mostly single and opening obliquely upwards. August and September. Distribution: Vermont to Alabama, westward to Missouri. Limestone—rare. Specimens examined: T'release, Russell. 4. Asplenium pinnatifidum Nutt. Gen. 2: 251. 1818. Gray, p. 38; Britton & Brown, ed. 1, 1: 22, ed. 2, 1: 27; Eaton, pl. 8, . 2. Rootstock short, creeping, branched, chaffy; fronds 6-9 inches high; stipes brownish near base and green above, clustered; blades 2-5 inches long, subcoriaceous, herbaceous, lanceolate- acuminate from broad and sub-hastate base, pinnatifid; the basal pinnae sometimes long-attenuate, lower lobes of pinnae roundish-ovate, margin crenate, upper pinnae gradually smaller and more adnate to winged rachises; sori straight, many, be- 1933] PINKERTON—FERNS OF MISSOURI 53 coming confluent with age, mostly solitary, occurring also on the slender prolongation. July to October. Distribution: Massachusetts to Alabama, westward to Missouri; recorded from Georgia and Arkansas. On La Motte sandstone only, in shaded crevices of cliffs. Specimens examined: T'release, Engelmann, Russell, Pinkerton. 5. Asplenium platyneuron? (L.) Oakes in D. C. Eaton, Ferns N. Am. 1: 24. 1879. Gray, p. 39; Britton & Brown, ed. 1, 1: 23, ed. 2, 1: 27; Eaton, pl. 4, fig. 1. Ebony spleenwort. Asplenium ebeneum Ait. Hort. Kew. 3: 462. 1789. Fronds 4-20 inches high, fertile ones tall and upright, sterile ones short and spreading; stipes and rachises reddish brown, shining, rather thick; blades linear-oblanceolate, tapering at base, once-pinnate; pinnae 20—40 pairs, lanceolate, 0.5-1.5 inches long, alternate, sessile, auricled on upper or both sides of base, and more or less overlapping rachis; sori 8-15 in number, nearer midveins than margins, becoming confluent. July to September. Distribution: Vermont to Alabama, westward to Texas and Oklahoma; recorded from Ontario and Colorado. Rocky open woods, preferring alkaline soil. Specimens examined: Davis, Kellogg, Engelmann, Eggert, Trelease, Bush, Palmer, Daniels, Pinkerton, Woodson, Russell, Mann, Dewart, Emig, Krause, Meek. 6. Asplenium resiliens Kunze, Linnaea 18: 331. 1844. Gray, p. 39; Britton & Brown, ed. 1, 1: 23, ed. 2, 1:27; Eaton, pl. 36, figs. 5&6 Asplenium parvulum Mart. & Gal. in Mém. Acad. Brux. 15: 60. 1842. Rootstocks with black scales; fronds 4-12.5 inches long; stipes and rachises black and shining, slender; blades normally linear-oblanceolate, pinnate; pinnae 4-12 mm. long, mostly opposite, nearly sessile, upper edges auricled and on lower pinnae both edges auricled, deflexed; blades widest in middle; margins mostly entire or slightly crenate, tendency to incurve slightly; fruit dots nearer outer margins than midribs, nearly parallel to the midribs, oblong, few, sometimes becoming confluent. June to October. 2 Var. incisum (E. C. Howe) Robinson in Rhodora 10: 29. 1908, has very brittle stipes and the pinnae deeply pinnatifid. This appears to be merely an ecological variation OL. 20 54 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: Massachusetts to Florida, westward to New Mexico. Limestone cliffs. Specimens examined: T'release, Bush, Palmer, Russell, Pinkerton, Drouet. 7. Asplenium Trichomanes L. Sp. Pl. 2: 1080. 1753. Gray, p. 39; Britton & Brown, ed. 1, 1: 24, ed. 2, 1: 28; Eaton, pl. 36, fig. 1. Maidenhair spleenwort. Rootstocks nearly erect, inconspicuously chaffy with narrow black scales; fronds 3-6 inches high; stipes slender, densely tufted, brownish-purple, polished, rachis similar to tip; blades once-pinnate, linear; pinnae 3-7 mm. long, herbaceous, mostly opposite, roundish, crenate margins, obliquely wedge-truncate at base, attached by narrow points; sori medial or nearer the midveins than margins, 3-6 pairs on outer sides of veins, be- coming confluent; indusia membranaceous. July to September. Distribution: Ontario to Alabama, westward to British Columbia and California; widely distributed but local. Sandstone rocks where plenty of water is available. Specimens examined: Kellogg, Letterman, Eggert, Engelmann, Palmer, Russell, Pinkerton, Trelease, Morrison, Rickett, Mackenzie, Blankenshi ip. ATHYRIUM 1. Fronds simply pinnate...............0000 s s sss esses 1. A. angustifolium 1; Fronds more NB DEIN. Loose Ayo eese ea ovo Pa ETSI 2 2. Fronds deeply bipinnatifid, margins lightly serrate-crenate. ... . » A. acrostichoides 2. Fronds usually tripinnatifid, margins deeply and irregularly incised........... 3. Rhizomes creeping, not densely covered with persistent feres fronds widest near base; indusia with glandular cilia; spores nigrescent, wrinkled. E nh oor eee eRe eT Rad da qoa Ro OE d 3. A. asplenioides 3. Rhizomes horizontal, completely concealed by thick fleshy bases of old fronds; fronds widest near middle; indusia toothed or short-ciliate, never glandular; spores yellow, slightly papillate.................. 4. A. angustum 1. Athyrium angustifolium (Michx.) Milde in Bot. Zeit. 48: 376. 1886. Gray, p. 39; Britton & Brown, ed. 1, 1: 24, ed. 2, 1: 28; Eaton, pl. 56, fig. 1. Narrow-leaved spleenwort. Asplenium pycnocarpon Spreng. Anleit. 3: 112. 1804. Asplenium angustifolium Michx. Fl. Bor. Am. 2: 265. 1803. Athyrium pycnocarpon Tidestrom, Elys. Marianum, p. 36. 1906. Rootstocks stout, creeping, with many long, branched rootlets; stipes green except for brown base; fronds 1-2.5 feet long, mem- branaceous, herbaceous, pinnate; pinnae 2-5 inches long, 20-30 pairs, short-stalked, linear-oblong, attenuate, margins slightly 1933] PINKERTON—FERNS OF MISSOURI 55 wavy; fertile pinnae near top, narrower and shorter; sori 20-30 pairs, linear, slightly curved, lying along outer of bifurcated veins; indusia firm, convex, concealed by strongly confluent sori at maturity. August and September. Distribution: Quebec to Georgia, westward to Michigan and Missouri; recorded from Kansas and Minnesota. Moist woods and shaded ravines, reported occasion- ally on sandy soil. ecimens examined: Davis, Kellogg, Palmer, Eggert, Trelease, Bush, Pinkerton, Daniels, Emig, Glatfelter. 2. Athyrium acrostichoides (Sw.) Diels in Engl. & Prantl, Nat. Pfl. 14: 223. 1899. Gray, p. 39; Britton & Brown, ed. 1, 1: 26, ed. 2, 1: 80; Eaton, pl. 50. Silvery spleenwort. Asplenium acrostichoides Sw. in Schrad. Jour. fiir die Bot. 2: (1800): 54. 1801. Asplenium thelypteroides Michx. Fl. Bor. Am. 2: 265. 1803. Athyrium thelypteroides (Michx.) Desv. in Mem. Soc. Linn. Paris [Prodr. p. 266] 6: 266. 1827. Rootstocks creeping, horizontal; stipes 8-16 inches long, straw-colored, herbaceous, with a few scales on lower portion; blades lanceolate to ovate-oblong, 1-3 feet long and 6-12 inches broad, narrowed to base, deeply bipinnatifid; ultimate segments distinct, obtuse; margins slightly serrate-crenate; sori 3-6 pairs per segment, arranged more or less evenly along lateral veins, mostly straight, oblong, some double. August to October. Distribution: New Brunswick to Georgia, westward to Missouri; recorded north- ward to Minnesota. Rich moist woods, or moist sandy soil. Specimens examined: Palmer, Eggert, Davis, Letterman, Broadhead. 3. Athyrium asplenioides? (Michx.) Desv. in Mem. Soc. Linn. Paris [Prodr. p. 266] 6: 266. 1827. Gray, p. 40; Britton & Brown, ed. 1, 1: 26, ed. 2, 1: 30; Butters in Rhodora 19: 169. 1917. Asplenium Athyrium Spreng. Anleit. 3: 113. 1804. Athyrium Filix-foemina (L.) Roth in Rómer's Arch. f. Bot. 21:106. 1799. Rhizomes horizontally creeping, partially covered by short persistent leaf bases, the whole structure 1-1.5 cm. in diameter, *'lhis species and A. angustum are very difficult to distinguish. It is often necessary to have the whole plant, fruiting and not too mature, to be absolutely certain. I have taken the character of the spore as my ultimate criterion. [Vor. 20 56 ANNALS OF THE MISSOURI BOTANICAL GARDEN with conspicuous projections of new growths before fronds of the current season; stipes long, about equal to the deltoid lanceolate fronds; young erowtha covered with scales soon deciduous, these small and light tan-colored, cell walls thin and inconspicuous; fronds bipinnate to tripinnatifid, second pair of pinnae commonly the longest, the lowest only slightly shorter; pinnae narrower at base; pinnules variously incised but apex more or less obtuse due to the venation, since two veins usually end on the same level; sori longer and narrower than in A. angustum, on the anterior side of anterior vein of each lobe of pinnule, and some- times on the lower veins, of typical athyroid type, rarely double; young indusia ciliate with multicellular glandular-tipped hairs nearly disappearing at maturity, leaving quite even margins; sporangia stalk frequently supplied with a glandular hair; spores furnished with a somewhat nigrescent, wrinkled exospore. July to October. Distribution: pe stent to Florida, westward to Missouri and Texas. Shaded rich woods or cliffs, sandy so Specimens examined: Kellogg, Trelease, Palmer, Bush, Soulard, Blankenship. 4. Athyrium angustum* (Willd.) Presl in Rel. Haenk. 1: 39. 1825. Gray, p. 40; Britton & Brown, ed. 1, 1: 26, ed. 2, 1: 30; Butters in Rhodora 19: 169. 1917. Asplenium Michauzii Spreng. Syst. 4: 88. 1827. Asplenium Filix-foemina (L.) Roth in Rómer's Arch. f. Bot. 2': 106. 1799. Rhizomes horizontal and somewhat oblique, condensed and completely covered by thick fleshy bases, 2-5 cm. in diameter; stipes up to half as long as the frond, a moderate number of scales persistent, 1 x 1.5 mm., dark opaque with thick darker cell walls and narrow cells; fronds bi-tripinnatifid, middle of fronds widest and lower pinnae much shorter and often deflexed, often large forms and polymorphic; pinnae not narrower at base; pinnules and segments with acute apices, due to one vein ending considerably beyond any of its neighbors; fertile fronds consider- ! Butters describes two varieties, namely: var. elatius and var. rubellum, distin- guished chiefly by the dimorphism and thicker texture of the former (a sun form and the monomorphism and thinner texture of the latter (a shade form). He claims that the latter has a more northerly distribution; but in Missouri the two seem to be purely ecological variations and scarcely worth varietal rank. 1933] PINKERTON—FERNS OF MISSOURI 57 ably narrower and more acute than sterile ones; sori short and wide, of typical athyroid type; indusia never glandular but persistently short-ciliate; sporangia stalk rarely bearing glandular hair; spores bright yellow and slightly papillate, no exospore and not wrinkled. June to October. Distribution: Labrador to Pennsylvania, westward to South Dakota and Missouri. Rich woods. Specimens examined: Bush, Davis, Palmer, Engelmann, Kellogg, Eggert, Russell. CAMPTOSORUS 1. Camptosorus rhizophyllus (L.) Link, Hort. Berol. 2: 69. 1833. Gray, p. 40; Britton & Brown, ed. 1, 1: 21, ed. 2, 1: 26; Eaton, pl. 8, fig. 1. Walking fern. Rootstocks short, creeping; stipes grouped in tufts, spreading, green, fleshy, 1-6 inches long; blades evergreen, subcoriaceous, 4-12 inches long, base auriculate, cordate or hastate, apex attenuate and filiform, rooting at tips (or from auricles); sori numerous, straight or slightly curved, single and on inside of and parallel to veins, near midribs, on both sides of outer veins and becoming confluent at exterior tips; indusia membranaceous. All year. Distribution: Ontario to Alabama, westward to oe and Oklahoma. Lime- stone rocks, usually associated with Entodon (a m Specimens examined: Bush, Davis, Engelmann, aie. Kellogg, Daniels, Palmer, Pinkerton, Dewart, Russell, Rickett, Weller. CHEILANTHES I Fronds relatively amootih. i... 4o 444 ae scan eee 1. C. alabamensis 1. Fronds hairy or tomentose "ie 2. Plants small, 2-6 inches tall, matted; mature stipes nearly glabrous. A s 2. Plants taller, 4-20 inches, not matted; stipes hirsute.............. bibe s» e**?289€829wó9À€9€v90466$o.0e*9»9e0e999259226»0698&€9484€9"9»2929252299 1. Cheilanthes alabamensis (Buckl. Kunze in Linnaea 20: 4. 1847. Gray, p. 36; Britton & Brown, ed. 1, 1: 30, ed. 2, 1: 34; Eaton, pl. 57, fig. 7. Rootstocks creeping, short, slender, with dark ferruginous scales; stipes black, wiry, slightly villous at base due to fine rusty scales 3-7 mm. long; blades scabrous to smooth, 2-10 inches long, lanceolate, bipinnate; pinnae acuminate, lower ones shorter than those above; pinnules often auriculate; indusia broad, pale but firm, frequently broken by incision of pinnules. August to October. [Vor. 20 58 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: Florida, westward to Missouri and Arizona. Limestone cliffs. Specimens examined: Kellogg, Palmer, Bush. 2. Cheilanthes Feei Moore, Ind. Fil. Gen. 38. 1857. Gray, p. 36; Britton & Brown, ed. 2, 1: 34; Eaton, pl. 6, fig. 1. Cheilanthes lanuginosa Nutt. in Hk. Sp. Fil. 2: 99. 1852. Rootstocks short, clothed with narrow scales of black centers and thin brown edges; fronds 2-6 inches tall; stipes densely tufted, black or brown, originally woolly, becoming glabrous when mature; blades bi-tripinnate, slightly tomentose above and woolly below ; pinnules divided into minute rounded segments, densely crowded; indusia herbaceous, continuous. July to Oc- tober. Distribution: Wisconsin to Texas, westward to Nevada and Arizona. Limestone cliffs in dry and exposed localities. Specimens examined: Davis, Kellogg, Engelmann, Trelease, Bush, Palmer, Daniels, Pinkerton, Morrison, U phof. 3. Cheilanthes lanosa (Michx.) Watt in Jour. Bot. 12: 48. 1874. Gray, p. 36; Britton & Brown, ed. 1, 1: 31, ed. 2, 1: 34; Eaton, pl. 2, fig. 2. Cheilanthes vestita Sw. Syn. Fil. 128. 1806. Rootstocks short, creeping, with pale brown scales; fronds 4-16 inches long; stipes wiry, dark brown, hirsute; blades her- baceous, bipinnate; pinnae hirsute and somewhat glandular; in- dusia inconspicuous, discontinuous. July to September. Distribution: New York to Alabama, westward to Oklahoma; recorded from — and Texas. Sandstone rocks, usually in dry and exposed places Specimens examined: Kellogg, Eggert, Letterman, Palmer, Broadhead, Pinkerion, Muller, Glatfelter, Mackenzie, Swallow, Emig, Blankenship, Link CYSTOPTERIS 1. Pinnae short-stalked on rachises; pinnules at least narrowed at points of attachment; segments ovate, acute, usually variously incised; indusia ne err rr er CR e i cee 1. C. fragilis 1. Pinnae sessile on rachises; pinnules oblong, obtuse, regularly toothed; indusia round or perona E AEE E E EE N E ID LE RA 2 Xo T 2. Fronds T oo EEEE "T 2. C. bulbifera 2. Fronds not estes NILUS Ead RR 8. C. bulbifera var. horizontalis 1. Cystopteris fragilis’ (L.) Bernh. in Schrad. Neues Jour. 5 An exceedingly variable species. Hybridization with C. bulbifera might account for some of the aberration. Forms bearing bulblets have been included under this group, as that character does not seem to be a constant feature for C. bulbifera alone. 1933] PINKERTON—FERNS OF MISSOURI 59 Bot. 1*: 27. 1806. Gray, p. 43; Britton & Brown, ed. Leas MA ed. 2, 1: 15; Eaton, pl. 53, fig. 1 Filiz Pots Under. Nat. Tesis ed. 6, 119. 1900. Rootstocks elongated, often 4-5 inches long, or shorter and condensed, slender, but covered with persistent leaf-bases, chaffy at apex, scales delicate, ovate, acuminate, ferruginous; stipes in a dense cluster, slender, brittle, 4-6 inches long; blades 6-8 inches long and about half as wide, thin and membranaceous, ovate-lanceolate; basal pinnae commonly narrower than the second and third pairs, apparently bipinnate but segments usually connected by narrowly winged midribs, segments round- ish-oval to ovate to rhomboid-ovate to ovate-lanceolate, toothed, dentate or irregularly laciniate; veinlets pinnately arranged on midveins, lower ones forked; sori small, roundish, seated on middle of veins nearest midrib; indusia delicate, rounded, ovate, or occasionally with narrow beak-like points, concealed by mature sporangia; lower pinnae often sterile. May to July. Distribution: Cosmopolitan. Rocky soil, moist woods, preferring alkaline soil. ecimens examined: Davis, Jnosiptana: Trelease, Bush, Palmer, Pinkerton, Eggert, Kellogg, Woodson, Harrison, Mann, Daniels, Blankenship, Duncan, Williams. 2. Cystopteris bulbifera (L.) Bernh. in Schrad. Neues Jour. Bot. 1*: 10. 1806. Gray, p. 43; Britton & Brown, ed. 1, 1: 12, ed. 2, 1: 15; Eaton, pl. 58, fig. 13. Filix bulbifera Underw. Nat. Ferns, ed. 6, 119. 1900. Rootstocks seldom over one inch long, chaffy at apex and covered with persistent leaf-bases; stipes slender, rather brittle, 6-10 inches long; blades submembranaceous but of a brittle rigidity, triangular-attenuate, 1—4 feet long, 3-5 inches broad at base; pinnae attenuated upward, bipinnate at base, upper pinnules attached by winged rachises; pinnae numerous (up to 40 pairs), oblong; pinnules oblong, obtuse, pinnately lobed; sori numerous, all pinnae fertile, arranged in rows along each side of midveins of pinnules, placed on the lowest superior veinlet of each group near its middle and so near the midvein; indusia truncate and fragile, covered by mature sporangia; bulblets some- times present on the under side of the frond attached near the base of or on the pinnae. July to August. Distribution: Newfoundland to Georgia, westward to Michigan and Arkansas; [Vor. 20 60 ANNALS OF THE MISSOURI BOTANICAL GARDEN recorded pm Utah and Arizona. Rocky soil, preferring limestone, in moist shady situation nel examined: Bush, Davis, Kellogg, Trelease, Palmer, Pinkerton, Daniels, Letterman 3. Cystopteris bulbifera var. horizontalis Lawson in Bot. Soc. Edinb. Trans. 8:40. 1866. Pl. 5, fig. 1. Fronds triangular-lanceolate, broad at base, not more than three or four times longer than broad; pinnae horizontal; lowest pinnules often quite broad with irregularly cut lobes and bearing numerous medium-sized sori along the lateral veins, sometimes almost tripinnate, or pinnules irregularly lobed, merely ovate- lanceolate. May to October. Distribution: southern Missouri and northern Arkansas. Damp limestone bluffs. Specimens examined: Ilasco, Ralls Co., Davis 2663; Sulphur Springs, Jefferson Co., coll. of Oct. 23, 1898, T'release; Terre Bleue Cr., Ste. Genevieve Co., coll. of Aug. 29, 30, 1898, T'release; Lesterville, Reynolds Co., coll. of June 5, 1929, Kellogg; Tecumseh, Ozark Co., Palmer 32896. DENNSTAEDTIA 1. Dennstaedtia punctilobula (Michx.) Moore, Ind. Fil. Gen. 97. 1857. Gray, p. 45; Britton & Brown, ed. 1, 1: 12, ed. 2, 1: 14; Eaton, pl. 44. Dicksonia pilosiuscula Willd. Enum. 1076. 1809. Dicksonia punctilobula Hk. Sp. Fil. 1: 79. 1846. Rootstocks extensively creeping, slender, scaleless but finely hairy at tips, irregularly branching with many long slender root- lets; stipes rather stout, light brown, chaffless, slightly puberulent ; blades 1-3 feet long, ovate-lanceolate, acuminate, delicately herbaceous, hairy and minutely glandular, tripinnatifid; pinnae numerous, lanceolate, pointed, second pair a little longer than first; pinnules adnate to rachis and usually decurrent on it, rhomboid-ovate, pinnatifid into oblong and obtuse cut-toothed lobes; sori minute, on upper margins of the lobes of the pinnules; indusia cup-like, delicate. August. Distribution: New Brunswick to Georgia, westward to Illinois and Missouri. Crevices of La Motte sandstone, rare. Specimens examined: T'release, Russell, Eggert, Pinkerton. NOTHOLAENA 1. Notholaena dealbata (Pursh) Kunze in Amer. Jour. Sci. II. 6: 82. 1848, as Nothochlaena. Gray, p. 35; Britton & Brown, ed. 1, 1: 32, ed. 2, 1: 35; Eaton, pl. 9, fig. 2 1933] PINKERTON—FERNS OF MISSOURI 61 Cheilanthes dealbata Pursh, Fl. Am. Sept. 2: 671. 1814. Nothochlaena pulchella Kunze in Bot. Zeit. 1: 633. 1848. Pellaea dealbata (Pursh) Prantl in Engler's Bot. Jahrb. 3: 417. 1882. Notholaena nivea var. dealbata Davenp. in Cat. Davenp. Herb. Suppl. 44. 1883. Rootstocks short, creeping, with narrow brown chaffy scales; stipes 1—4 inches long, tufted, wiry, slender, copper-brown, as are the rachises; blades 2-4 inches long, broadly deltoid-ovate, 4—5- pinnate, all but ultimate segments alternate, those sometimes opposite; segments obovate-oval and entire or several-lobed, 1-2 mm. broad, upper surfaces pale green, coriaceous, lower white and powdery, giving a silvery appearance; sporangia seated on upper portions of the veins; no indusia, but a protection afforded by the slightly turned-back margins. June to September. Distribution: Missouri and Kansas to central Texas; recorded from Nebraska. Dry calcareous rocks. Specimens examined: Daniels, Bush, Palmer, Blankenship. ONOCLEA 1. Onoclea sensibilis L. Sp. Pl. 2: 1062. 1753. Gray, p. 45; Britton & Brown, ed. 1, 1: 9, ed. 2, 1: 11; Eaton, pl. 72, fig. 1. Sensitive fern. Rootstocks slender, creeping, rooting freely and often forking; stipes coarse, straw-like, hollow, flattened, light brown when dry; sterile blades triangular or triovate, midribs winged, widening toward the apex, sinuses rounded; lowest segments broadly lanceolate, herbaceous, sensitive to frost; veins conspicuous, re- ticulate; margins variously rounded, lobed, toothed, serrate or acute; fertile fronds 12-18 inches long, pinnate, contracted; each segment a pouch filled with several sporangia; delicate hood-like indusia. August to November. Distribution: Newfoundland to Florida, westward to Kansas. Swamps or damp rich soil. Specimens examined: Bush, Davis, Eggert, Palmer, Morrison, M uller, Daniels, Blankenship, Mackenzie, Williams. PELLAEA 1. Pinnae dichotomously branched at apex........ 3. P. atropurpurea var. cristata L. Pinnae not branched at apex......«4 4 4WANAAE ee nese a error 2 [Vor. 20 62 ANNALS OF THE MISSOURI BOTANICAL GARDEN 2. Stipes M: reddish-brown; pinnae usually membranaceous, pale green, ore or s^ pec at both corners and tending to i at bases into i Lo ee Os ba ks pad. new a kaw aden wad ees . P. glabella 2. Stipes scabrous, Sale scale sis. pinnae coriaceous, blue-green, ‘knee And seldom PVN. 6555s 5.60504 ocn iw ook eh 2. P. atropurpurea 1. Pellaea glabella Mett., Kuhn in Linnaea 36: 87. 1869. Gray, p. 37; Britton & Brown, ed. 2, 1: 33. Pellaea atropurpurea var. Bushii Mackenzie, Flora Jackson County, Mo., p. 5. 1902 Stipes and rachises brownish-red, smooth or but slightly hairy; fronds simply pinnate above, the lower ternate or rarely quinate (the entire pinnule seems to be breaking at the base—a possible tendency toward compound pinnules) ; pinnae membranaceous to coriaceous but always pale green. April to October. Distribution: Ontario and Vermont to Pennsylvania, westward to South Dakota and northern Arkansas. Exposed high places on limestone cliffs Specimens examined: Davis, Eggert, Kellogg, Bush, Palmer, Pinkerton, Trelease. 2. Pellaea atropurpurea (L.) Link, Fil. Sp. in Hort. Berol. 59. 1841. Gray, p. 37; Britton & Brown, ed. 1, 1: 29, ed. 2, 1: 33; Eaton, pl. 54, fig. 4. Rootstoek short, densely covered with rusty scales about 2 mm. in length; fronds 4-12 inches long, pinnate or below bipinnate, coriaceous; fertile segments linear, more or less pointed at apex; sterile segments approaching oval, shortly stalked; veins obscure; continuous indusia of reflexed margins. June to September. Distribution: Connecticut to Florida, westward to South Dakota and Texas. Pinkerton, Kellogg, Daniels. 3. Pellaea atropurpurea var. cristata Trel. in Rept. Mo. Bot. Gard. 12: 77. 1901. Gray, p. 37. Pinnae dichotomously forked. Distribution: known only from Eureka, Missouri. Limestone. Specimens examined: Eureka, 1899, Pauls. POLYPODIUM . Blades smooth, green... sess 1. P. virginianum : Blades densely hun a | ERETETITIETTCTT TT TT TOIT S.P. elige 1933] PINKERTON—FERNS OF MISSOURI 63 1. Polypodium virginianum L. Sp. Pl. 2: 1085. 1753. Gray, p. 34; Britton & Brown, ed. 1, 1: 32, ed. 2, 1: 36; Eaton, pl. 31, fig. 1. Common polypody. Polypodium vulgare L. Sp. Pl. 2: 1085. 1753. Rootstocks close to surface of soil, covered with chaffy, red- brown scales; stipes smooth, herbaceous, light green; blades ovate- oblong or narrowly oblong, subcoriaceous or chartaceous, ever- green, simple, deeply pinnatifid, smooth; segments linear-oblong, obtuse or slightly acute, crenulate and serrate, sinuses rounded, alternate, margins obscurely dentate; sorilarge, naked. July. pde Newfoundland to Florida, westward to Minnesota and Arkansas. sandstone or sandy soil. [OE HA examined: Eggert, Russell, Engelmann, Trelease, Pinkerton, Letterman, Greene 2. Polypodium polypodioides (L.) Hitche. in Rept. Mo. Bot. Gard. 4:156. 1893. Gray, p. 34; Britton & Brown, ed. 1, 1:33, ed. 2, 1: 36; Eaton, pl. 26, fig. 2. Gray polypody. Polypodium incanum Sw. Fl. Ind. Occ. 3: 1645. 1806. Rootstocks woody, covered with small dark brown scales; stipes 1-4 inches long, slender, bearing peltate ovate scales with dark brown centers; blades oblong-lanceolate, pinnate; segments ob- long, obtuse, entire, sessile, separated by rounded sinuses, alter- nate, upper surfaces smooth or with few scales, lower densely scaly; fruit dots small and naked. July to September. Distribution: Massachusetts to Florida, westward to Missouri and Texas. Usually an epiphyte on swamp trees, particularly Taxodium distichum; also on sandstone rocks. Specimens examined: Eggert, Trelease, Bush, Pinkerton, Rickett, Kellogg, Mac- kenzie, Meek. POLYSTICHUM 1. Margins serrulate; fertile pinnae contracted; sori confluent... . 1. P. acrostichoides 1. Margins deeply toothed or pinnatifid; fertile pinnae scarcely contracted; sori not confluent and appearing only on tips of lower pinnae.............. a seti cule YR e ir e onus hi ERE 2. P. acrostichoides var. incisum 1. Polystichum acrostichoides (Michx.) Schott, Gen. Fil. 17. 1834. Gray, p. 40; Britton & Brown, ed. 1, 1: 14, ed. 2, 1: 16; Eaton, pl. 34. Christmas fern. Aspidium acrostichoides Sw. Syn. Fil. 44. 1806. Dryopteris acrostichoides Kuntze, Rev. Gen. Pl.2:812. 1891. [Vor. 20 64 ANNALS OF THE MISSOURI BOTANICAL GARDEN Rootstocks stout, creeping, with persistent leaf-bases; stipes densely tufted, with chaff of large golden-brown scales, 5-7 mm. wide; blades lanceolate, pinnate, 1-2 feet long, 3-5 inches wide, rigid, evergreen, subcoriaceous; pinnae numerous, 1-3 inches long, oblong-lanceolate, short-stalked, upwardly falcate or lowest slightly deflexed, apex acutish, upper sides auriculate; margins serrulate to incised with incurved bristle-pointed teeth; veins free, branching three to four times; upper pinnae of fertile fronds more or less contracted and heavily soriferous; sori terminal on lower veinlets in 2-4 rows, becoming confluent with age; indusia round, indurated, not glandular, persistent. July to August. Distribution: Maine to Florida, westward to Michigan and Texas; recorded from Nova Scotia, Shady hillsides of ravines, in rich soil which is interspersed with on. Specimens examined: Bush, — Eggert, Trelease, Palmer, Pinkerton, Emig, Kellogg, Daniels, Thomas, Mackenzie 2. Polystichum acrostichoides var. incisum Gray, Man. Bot. ed. 1, 632. 1848. Gray, p. 40; Britton & Brown, ed. 1, 1: 14, ed. 2, 1: 16. Polystichum acrostichoides var. Schweinitzii (Beck) Small in Bull. Torr. Bot. Club 20: 464. i Aspidium schweinitzii Beck, Bot. North. & Mid. States, ed. 1, 449. Aspidium acrostichoides var. incisum D. C. Eaton, Ferns N. Am. 1: 258. 1879. Segments few and distant, large, irregularly incised; upper pinnae covered by confluent sori, lower ones fertile at tips only, sori large; veins numerous, frequently overlapping to form ir- regular areoles. Distribution: This seems to be nearly as common as the type in southern Missouri and northern Arkansas; but since there appears to be a series of intermediate forms between the two I have not attempted to separate the individual ranges. PTERETIS 1. Pteretis nodulosa (Michx.) Nieuwl. in Am. Midl. Nat. 4: 334. 1916. Gray, p. 45; Britton & Brown, ed. 1, 1: 9, ed. 2, 1: 11; Eaton, pl. 73. Ostrich fern. Osmunda Struthiopteris L. Sp. Pl. 2: 1066. 1753. Onoclea Struthiopteris Hoffm. Deutsch. Fl. 2:11. 1795. 1933] PINKERTON—FERNS OF MISSOURI 65 Struthiopteris germanica Willd. Enum. 1071. 1809. Matteuccia Struthiopteris (L.) Todaro in Syn. Pl. Acot. Vasc. Sicilia, p. 30. 1866. Matteuccia nodulosa (Michx.) Fernald in Rhodora 17: 164. 1915. Rootstocks stout and ascending, with slender underground stolons; sterile blades 2-7 feet high, 6-15 inches broad, short- stalked, broadly oblanceolate, abruptly short-acuminate, gradual- ly narrowed below middle, lower pinnae reduced; pinnae narrow, deeply pinnatifid; segments oblong, obtuse, entire; veins simple, fertile blades 1-7.5 feet high, with pod-like brown pinnae, in- cluded sori confluent. July. Distribution: Newfoundland to New York, westward to Michigan and Missouri; recorded from Virginia. Alluvial soil. Specimens examined: Livonia, Putnam Co., Bush 7780, 7780A, 7780B. PTERIDIUM 1. l hel ted, seldom redivided. . .2. P. latiusculum var. Paa 1. Pinnules oblong-lanceolate, at least a few vinnatifid ida d NN RE . P. latiusculum 1. Pteridium latiusculum (Desv.) Hieron. Wissenschaftl. Ergeb. d. Schwed. Rhodesia-Kongo-Exp. 1911-12, Heft 1, p. 7. 1914; Maxon in Am. Fern Jour. 9: 43. 1919. Gray, p. 36; Britton & Brown, ed. 1, 1: 28, ed. 2, 1: 32; Eaton, pl. 35. Pteris aquilina L. Sp. Pl. 2: 1075. 1753. Pteridium aquilinum Kuhn in Decken, Reisen in Ost-Afrika 33:11. 1879. Rootstocks cord-like, creeping, blackish, deeply buried; stipes solitary, erect, naked, 1-3 feet high, swollen at base, brownish; blades 2-4 feet long, 1-3 feet broad, triangular-ovate, rigidly subcoriaceous, subternate; the long-stalked basal pinnae and middle ones bipinnate, those above lobed or simple; segments oblong-lanceolate, under surfaces glabrous or pubescent; veins close-packed, free; indusia continuous around the edge of the pinnules, double. July to September. Distribution: Cosmopolitan, open woods, preferring sandy soil. Specimens examined: Kellogg, Eggert, Palmer, Bush. 2. Pteridium latiusculum var. pseudocaudatum Maxon in Am. Fern Jour. 9: 44. 1919. Gray, p. 36; Britton & Brown, ed. 2, 1: 32. [Vor. 20 66 ANNALS OF THE MISSOURI BOTANICAL GARDEN The variety occurs in this region but does not seem to have a distribution different from that of the species. It is distinguished by its very narrow elongated pinnules. THELYPTERIS Dryopteris Adans. Polystichum Roth Aspidium Swartz Nephrodium Richard Lastrea Bory Phegopteris (Presl) Fée L INQUE DNE s 5: 65:5.65 6050000 Sra eR OSA EX SEA 5. T. heragonoptera b 30 MEE LS EE 05055 0 ek EEEEREVEYREAAVCWXTEPS ECEERSERERESDIRARI 2 2. Blades bipinnatifid or bipinnate; segments not spinulose.................... 3 2. Blades tripinnatifid or tripinnate; segments spinulose....................... 4 3. Sori medial, small; blades membranaceous, not evergreen................. COSERKRLEREXENAATCRERARWARDEREKENRE SREZSATLATAA 1. T. palustris var. pubescens 3. Sori near margins, large; blades subcoriaceous, evergreen...... 2. T. marginalis 4. Indusia glandless; pinnae decidedly oblique to rachises, scales of stipe pale UNI Hoe d Eee eee Tey: avesECNICO EI I 8. T. spinulosa 4. Indusia glandular; pinnae more or less at right angles to rachises; scales of stipe brown with a dark center.............. 4. T. spinulosa var. intermedia 1. Thelypteris palustris var. pubescens’ (Lawson) Fernald in Rhodora 31: 34. 1929. Gray, p. 41; Britton & Brown, ed. 1, 1: 15, ed. 2, 1: 18; Eaton, pl. 30. Marsh shield-fern. Thelypteris Thelypteris Nieuwl. in Am. Midl. Nat. 1: 226. 1910 Rootstocks slender, creeping, blackish, nearly naked; stipes as long or longer than the blades, blackish at base, sparingly chaffy; blades oblong-lanceolate, pinnate, 1-3 feet long, 4-6 inches wide, membranaceous, scarcely narrowed at base, short- acuminate; pinnae 20-30 pairs, alternate, short-stalked, approxi- mately at right angles to stalks, linear-lanceolate, broadest at base, deeply pinnatifid; segments oblong-obtuse, mostly entire; veins pinnate, usually once-forked; fertile fronds usually on longer stalks and of narrower segments than the former; sori nearly medial, crowded; indusia glabrous. August. * Eaton mentions an unusual form—‘the lower two or three pairs are usually but little shorter than those above them; but fronds are occasionally found in which they are conspicuously reduced." Such & form is one from Iron Lake, Iron Co., Kellogg 1634. 1933] : PINKERTON—FERNS OF MISSOURI 07 Distribution: New Brunswick to Florida, westward to Texas. Wet woods. Specimens examined: Eggert, Kellogg, Bush, Pinkerton. 2. Thelypteris marginalis (L.) Nieuwl. in Am. Midl. Nat. 1: 226. 1910. Gray, p. 42; Britton & Brown, ed. 1, 1: 17, ed. 2, 1: 20; Eaton, pl. 55. Evergreen wood fern. Rootstocks stout, ascending, covered with long chaffy brown scales; stipes several inches to a foot long, light tan, somewhat chaffy; blades 6-30 inches long, evergreen, subcoriaceous, ovate- lanceolate, scarcely narrowed at base, deeply bipinnatifid; pinnae numerous, practically sessile, lanceolate, acuminate, slightly broader above the base; pinnules adnate to narrowly winged sec- ondary rachis, oblong to oblong-lanceolate, faintly crenately- toothed; veins free, forked or pinnately branched; sori large, near margins of segments; indusia hard, orbicular-reniform, glabrous, dark brown. July to August. Distribution: Nova Scotia to Georgia, westward to Kansas and Oklahoma. Sand- Stone ledges where it is moderately moist. Specimens examined: Letterman, Engelmann, Eggert, Kellogg, Palmer, Pinkerton, Muller, Rickett, Mackenzie, Broadhead, Link, Blankenship, Trelease. 3. Thelypteris spinulosa (Retz.) Nieuwl. in Am. Midl. Nat. 1: 226. 1910. Gray, p. 43; Britton & Brown, ed. 1, 1: 18, ed. 2, 1: 21; Eaton, pl. 68. Spinulose shield-fern. Rootstocks stout, creeping, chaffy; stipes 4-14 inches long, chaffy ; blades 0.5-3 feet long, ovate-lanceolate to oblong, acumi- nate, bi-tripinnate, firmly membranaceous; primary pinnae short- stalked, lower pairs triangular-ovate to triangular-lanceolate, re- maining pinnae gradually narrower in outline; secondary rachises narrowly wing-margined; pinnules oblong, subacute, incised with spinulose-serrate lobes; sori small, sub-marginal, terminal on vein- lets; indusia flat, round-reniform, glandless. Distribution: Labrador to Virginia, westward to Idaho. Moist woods, alluvial soil. Specimens examined: Neeleyville, Butler Co., coll. of Oct. 30, 1899, Russell. 4. Thelypteris spinulosa var. intermedia (Retz.) Nieuwl. in Am. Midl. Nat. 2: 278. 1912. Gray, p. 43; Britton & Brown, ed. 2, 1: 22. Similar to the species except for the glandular indusia and right-angled relation of pinnae to rachis. [Vor. 20 68 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: In Missouri—more northern, sandy so Specimens examined: Pickle Springs, Ste. ced Co., Pinkerton 1; and coll. of May 21, 1930, Kellogg. 5. Thelypteris hexagonoptera (Michx.) Weatherby in Rhodora 21: 179. 1919. Gray, p. 35; Britton & Brown, ed. 1, 1: 19, ed. 2, 1: 23; Eaton, pl. 65. Broad beech-fern. Phegopteris hexagonoptera (Michx.) Fée, Gen. Fil. 243. Rootstocks elongated, slender, creeping, chaffy with gold scales; stipes 8-20 inches long, sender: greenish, or straw-colored; blades triangular, 7-12 inches long, 7-15 inches wide, thinly herbaceous, deep green, slightly hairy or granular, bipinnatifid; pinnae adnate to winged rachis, the lowermost ones broadest and largely ovate to ovate-lanceolate, and others lanceolate; seg- ments usually bluntly acuminate and crenate; veins pinnate and free, branched or not; sori borne over whole frond, small, near end of veins and so near margins; no indusia. August. a New Brunswick to Delaware, westward to Oklahoma. Moist woods and rav "iens examined: Davis, Eggert, Palmer, Kellogg, Bush, Pinkerton, Rickett, Link. WOODSIA 1. Woodsia obtusa (Spreng.) Torr. Cat. Pl. in Geol. Rept. N. Y. 195. 1840. Gray, p. 44; Britton & Brown, ed. 1, 1: 11, ed. 2, 1: 14; Eaton, pl. 71. Rootstocks short, creeping, chaffy with narrow light brown scales; stipes 2-6 inches long, green with darkish base in living plants and drying to a bright brownish straw color; fronds 8-15 inches long, broadly lanceolate, narrower at base than in middle, abruptly terminating at tip, membranaceo-herbaceous, minutely glandular, deep blue-green in color, nearly bipinnate; pinnae remote, short-stalked, obtuse, triangular, ovate to oblong, mostly opposite; segments oblong, obtuse, crenately toothed; sori sub- terminal on veins, nearer margin than midveins; young indusia subglobose, splitting into several irregular lobes which extend out beyond the sporangia, difficult to detect. September. Distribution: Vermont to Alabama, westward to Wisconsin and Texas; recorded from Nova Scotia. Moist calcareous or acid soil. Specimens examined: Davis, Kellogg, Palmer, Pinkerton. 1933] PINKERTON—FERNS OF MISSOURI 69 WOODWARDIA 1. Woodwardia areolata (L.) Moore, Ind. Fil. Gen. 45. 1857. Gray, p. 38; Britton & Brown, ed. 1, 1: 20, ed. 2, 1: 25; Eaton, pl. 22, fig. 2. Lorinseria areolata (L.) Presl in Epim. Bot. 72. 1849. Acrostichum areolatum L. Sp. Pl. 2: 1069. 1753. Woodwardia angustifolia J. E. Smith in Mem. Acad. Turin Se Ap Rootstocks creeping, several to 12 inches long, often branched, less than 1/4 inch thick, with some scales near apex; stipes dark at base, paler above, bearing a few scales, dimorphic; sterile fronds 9-10 inches long, oblong-ovate, pinnate, bright green above, paler below; rachis winged from tip to just below base of blades; sinuses rounded, segments acute, 1—4 inches long, 0.5-1 inch wide, finely serrate, membranaceous; veins finely reticulated, with a longitudinal row of narrow areoles along each side of midribs and midveins, and several rows of hexagonal aeroles and free veins running outwards to serrated edges; fertile fronds taller, with a darker stalk; segments and wing of rachis much narrowed; one row of areoles on each side of midribs, each covered by a brown involucre attached to the outer enclosing veins and open along midrib; sporangia also from enclosing veinlets; sterile fronds appearing in May and fertile ones later. August to October. Distribution: Massachusetts to Florida, westward to Missouri and Texas; re- corded from Maine. Swamps and moist soil. Specimens examined: Poplar Bluff, Butler Co., July, 1898, Eby. EQUISETACEAE EQUISETUM 1. Stems annual; dir&orphie......... Vet sees 1. E. arvense 1. Stems perennial; monomorphic. o. -graer hae eesi e nn heh 2 2. Sheaths cylindrical, green, turning gray, with black lines at bases and tops, vue and undilated, splitting with age; ridges almost smooth; often very plant... roD Oe rr rn 2. E. praealtum 2. Sheathe ae eae, green with narrow black limbs, elongate, not splitting; ridges with one row of tubercules; medium-sized plants....... . E. laevigatum 1. Equisetum arvense L. Sp. Pl. 2: 1061. 1753. Gray, p. 52; Britton & Brown, ed. 1, 1: 36, ed. 2, 1:39. [Vor. 20 70 ANNALS OF THE MISSOURI BOTANICAL GARDEN Rootstock tuberiferous, felted with brown wool, extensively creeping; fronds dimorphic; sterile ones annual, prostrate or erect, green, rather slender, 12-24 inches high, 6-19 furrowed, with scattered stomata; sheaths whitish, tipped with about 12 acuminate, brown, separate teeth; branches whorled, simple or compound, not drooping, the 3-4 angled sheaths of branches consisting usually of 4 teeth, or often 3, rarely 5, long and acumin- ate; fertile fronds annual, appearing in early spring before sterile ones, usually unbranched, succulent, and withering after spores are ripe, 4-10 inches high, light brown, sheaths conspicuous, long, flaring and pointed, of 8-12 teeth; spikes not apiculate; variable. May. Distribution: Greenland to Alabama, westward to Alaska and California. Sandy soil, in waste places, along streams, etc. Specimens examined: ^ Kellogg, Bush, Daniels, Palmer, Letterman, Eggert, Trelease, Daniels, Mackenzie 2. Equisetum praealtum’ Raf. Fl. Ludovic. 13. 1817. Gray, p. 53; Britton & Brown, ed. 1, 1: 38, ed. 2, 1: 41. Equisetum hyemale var. robustum A. A. Eaton in Fern Bull. 11: 74. i Hippochaete prealta (Raf.) Farwell in Mem. N. Y. Bot. Gard. 6: 467. 1916. Equisetum robustum A. Br., Engelm. in Amer. Jour. Sci. 46 88. 1844 Fronds perennial, evergreen, 3-11 feet high, erect; stems rough, 20-48 ridges bearing silica in single rows; sheaths cylindrical, short-appressed, not dilated or only slightly when young,* at first green but soon turning black or gray with black bands above and below, splitting with age; sheath segments normally tri- carinate; teeth dark and caducous; cones pointed. May. T Schaffner (Am. Fern Jour. 13: 33-41. 1923), says: "Although P E. ipsas usually bears cones on shoots of the season. Shoots ster firs year may bear cones the second, both terminal and on lateral branches. Branching is rare the first season unless the shoot is injured, but the second year branching i is common even on uninjured shoots. E. praealtum is an exceedingly le species, some forms recognized probably being genetie and some ecological, but none of these forms passes out of the specific limits as usually drawn. Some are short and robust; some tall and massive; some very slender." * Young shoots are often very difficult to distinguish from E. laevigatum. 1933] PINKERTON—FERNS OF MISSOURI 71 Distribution: Quebec to Georgia, westward to British Columbia and New Mexico. Wet sandy places. Specimens examined: Davis, Engelmann, Bush, Palmer, Pinkerton, Eggert, Trelease, Daniels, Throuse, Demetrio. 3. Equisetum laevigatum A. Br., Engelm. in Amer. Jour. Sci. 46:87. 1844. Gray, p. 53; Britton & Brown, ed. 1, 1: 38, ed. ATIS Fronds perennial, erect, mostly simple, pale green, 1-5 feet high, 14-30 ridged, almost smooth, with stomata in two rows on each side of depressions; sheaths funnel-shaped, elongated, and green with usually a narrow black band at thetop; white-margined teeth soon deciduous; cones pointed. May to June. Distribution: New York to North Carolina, westward to Washington and Cali- fornia. Along streams, especially in san il. Specimens examined: Engelmann, Bush, Daniels, Eggert, Palmer. OSMUNDACEAE OSMUNDA 1. Sterile fronds truly bipinnate; pinnules stalked and bend separated....... P EA Nen Sah 4a SER FP PUR be n ASS . O. regalis var. spectabilis i, Sterile fronds bipintattd. ....2. «4.56 don nena ou Eee BU EN 2 2. Fronds dimorphic; sterile pinnae with tufts of m hairs at base; apices of fronds and pinnae tapering; veins inconspicuous............ . O. cinnamomea 2. Fronds monomorphic but fertile part of fioul ia is die middle several pairs of pinnae; no tufts of hair at base of pinnae; apices of fronds and pinnae abruptly narrowed and scarcely acute; veins dark-colored, conspicuous. dala bac E EG e riu d ME M Ae A E 3. O. Claytoniana 1. Osmunda regalis var. spectabilis Fernald in Rhodora 32: 72. 1930. Gray, p. 46; Britton & Brown, ed. 1, 1:5, ed. 2, 1: 7. Rootstocks stout, creeping, covered with persistent leaf-bases; fronds 2-6 feet high, stipes never chaffy, green or yellow, rounded on back, flattened on front; blades ovate-oblong, bipinnate; pinnae iios opposite, ovate; pinnules unequal, 6-12 pairs plus the terminal one, ES Corno short-petioled, distant, oval- oblong or oblong-lanceolate, vidoe ones often auriculate on lower side, margins crenulate-serrate, apex obtuse or subacute; apical pinnae fertile, bipinnate, ultimate divisions thread-like, containing no chlorophyll, entirely covered with sporangia. May to July. [Vor. 20 72 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: Newfoundland to Florida, westward to Saskatchewan and Missis- sippi. Lowlands, swamps, marshes, and wet woods. Specimens examined: Eggert, Palmer, Kellogg, Bush, Pinkerton, Engelmann, Trelease, Mackenzie. 2. Osmunda cinnamomea L. Sp. Pl. 2: 1066. 1753. Gray, p. 47; Britton & Brown, ed. 1, 1: 5, ed. 2, 1: 7. Rootstocks creeping, massive, bearing circular clusters of sterile leaves with one or more fertile ones within; fronds dimor- phic, sterile ones 1 foot or more high, oblong-lanceolate, acu- minate, tapering, deeply pinnatifid; pinnae oblong-lanceolate, acute, tapering, tomentose tuft at base of each pinna; pinnules obtuse, subcoriaceous, green; veins inconspicuous, veinlets once- forked near midvein; margins entire or obscurely crenulate; fertile fronds about as tall as the sterile, bipinnate, and covered with cinnamon-colored sporangia, arising early in the spring pre- ceding the sterile ones. May an Distribution: Newfoundland, westward to Minnesota and New Mexico. Sand- Specimens examined: Eggert, T'release, Kellogg, Pinkerton, Russell. 3. Osmunda Claytoniana L. Sp. Pl. 2: 1066. 1758. Gray, p 46; Britton & Brown, ed. 1, 1: 6, ed. 2, 1: 8. Rootstocks creeping, stout, with imbricated leaf-bases; stipes several inches to 2 feet long, woolly when young but never chaffy ; fertile fronds taller than the sterile and in the midst of the crown formation, oblong-lanceolate, 1—4 feet long; lowest pinnae about half as long as middle ones, acute and often rounded; pinnae barely acute, never acuminate, short-stalked, lütioeclste from a broad base; pinnules close; 2—6 pairs of fertile pinnae near middle of frond, shorter than sterile pinnae and deflexed in a mature speci- men, closely bipinnate, woolly, covered with bivalvular reticulated sporangia. May to July. mto Adr Newfoundland to North Carolina, westward to Minnesota and ri. mps and moist woods, moist sandstone ledges. el thence Bush, Eggert, Davis, Palmer. LYCOPODIACEAE LYCOPODIUM 1. Sporophylls segregated into slender cones; habit of plants fan-like......... els Hise EGET ELS Ae heed ve hee oo ae 3. L. complanatum var. flabelliforme 1933] PINKERTON—FERNS OF MISSOURI 73 1. Sporophylls not differing from vegetative leaves; habit of plants rope-like...... 2 2. All the leaves broadest above the middle; margins jagged...... 1. L. lucidulum 2. Shorter leaves broadest at base; margins entire or slightly denticulate...... OPE PUDRS M RODA c E 2. L. lucidulum var. porophilum 1. Lycopodium lucidulum Michx. Fl. Bor. Am. 2: 284. 1803. Gray, p. 55; Britton & Brown, ed. 1, 1: 40, ed. 2, 1: 44. Stems assurgent from decumbent persistent bases giving rise to a few vertical stems; leaves dark green and shining, widespread or becoming deflexed, acute, broadest above middle, erose- denticulate, arranged in alternating series of long and short members, the latter often entire and usually bearing the sporangia; gemmiferous. August to October. Distribution: Newfoundland to Delaware, westward to Alaska and Washington. On sandstone only, usually associated with Sphagnum. Specimens examined: Kellogg, Eggert, Pinkerton. 2. Lycopodium lucidulum var. porophilum (Lloyd & Underw.) Clute, Fern Allies, p. 3. 1905. Gray, p. 55; Britton & Brown, ed. 2, 1: 44. Lycopodium porophilum Lloyd & Underw. in Bull. Torr. Bot. Club 27:150. 1900. Essentially like L. lucidulum except that the shorter leaves are broadest at the base and the margins are nearly smooth, and are not deflexed. Distribution: in the same places as the species. Sandstone Specimens examined: Terre Bleue Cr., Ste. Genevieve Co., coll. of Aug. 29, 1898, Trelease, and Pinkerton 81. 3. Lycopodium complanatum var. flabelliforme Fernald in Rhodora 3: 280. 1901. Gray, p. 57; Britton & Brown, ed. 1, 1: 43, ed. 2, 1: 48. Rhizomes slender, creeping, with numerous erect stems which branch irregularly, giving rise to a flattened fan-shaped vege- tative structure, about a foot high with 4-ranked imbricated scale-leaves, those of the two lateral rows broad, with spreading tips, of the upper row narrow and incurved, and of the lower row minute deltoid-cuspidate; peduncles long, dichotomously branched at tips and bearing a number of slender cones about an inch long. August and September. [Vor. 20 74 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: — to West Virginia, westward to Alaska and Idaho. Open pine woods on sandy so Specimens examined: "Pickle Springs, Ste. Genevieve Co., Kellogg 8718. SELAGINELLACEAE SELAGINELLA 1. Plants bearing ill-defined strobili; leaves dimorphic, 4-ranked....... 1. S. apoda 1. Plants bearing distinct strobili; leaves of one kind, spirally arranged.2. S. rupestris l. Selaginella apoda (L.) Fernald in Rhodora 17: 68. 1915. Gray, p. 58; Britton & Brown, ed. 1, 1: 45, ed. 2, 1: 49. Selaginella apus (L.) Spring in Mart. Fl. Bras. 12:119. 1840. Stems prostrate and creeping, 1-4 inches long, pale green, delicate; leaves of two kinds, four-ranked and spreading, the smaller pointed and appressed to the stem; no distinct cones; fertile leaves near tip of branches, those containing macrospores conspicuously bulged. July to September. Distribution: Massachusetts to Florida, westward to Michigan and Louisiana. Moist shaded places, among grasses. Specimens examined: Eggert, Bush, Palmer, Mackenzie. 2. Selaginella rupestris (L.) Spring in Mart. Fl. Bras. 12: 118. 1840. Gray, p. 57; Britton & Brown, ed. 1, 1: 44, ed. 2, 1: 49. Stems densely tufted, bearing occasional sterile runners; all leaves alike, narrow, appressed, and imbricated, bristle-tipped, gray-green; strobili four-sided. August to October. Distribution: Quebec to Alabama, westward to Minnesota and Oklahoma. In dry exposed places where there is a little soil, sandstone, or chert. Specimens examined: Eggert, Russell, Bush, Palmer, Pinkerton, Greene, Broadhead, VanIngen, Shepard. ISOETACEAE ISOETES 1. Megaspores reticulate; sporangia unmarked................. 1. I. Engelmanni . Megaspores tuberculate; sporangia marked in some w 2 . Megaspores less than 480 in diameter; sporangia ae BE with brown spots. Fleck 92 UR RD PAR RUE RR ERE RUE OR ala aaa Galeua 4s Gee EORR 2. I. melanopoda 2. Megaspores more than 480 y in diameter; sporangia marked with brown lines. ee E re a ere er re ES 8. I. Butleri t2 m= 1. Isoetes Engelmanni A. Br. in Flora 29: 178. 1846. Gray, p. 61; Britton & Brown, ed. 1, 1: 48, ed. 2, 1: 53. Corms 2-lobed; leaves 15-60, 13-50 cm. long, light green; sto- 1933] PINKERTON—FERNS OF MISSOURI 75 mata numerous; peripheral strands variable in number or none; sporangia oblong, unspotted, with narrow velum; megaspores white, 400-570 y in diameter, distinctly marked with honey- comb network of narrow ridges; microspores 21-30 y, seldom 33 y, in length, smooth to minutely roughened. Distribution: eastern border to rom valley. Near ponds. Specimens examined: Engelma 2. Isoetes melanopoda Gay and Dur. in Bull. Soc. Bot. Fr. 11:102. 1864. Gray, p. 61; Britton & Brown, ed. 1, 1: 48, ed. 21:54. Corms 2-lobed; leaves 15-60, 15—40 cm. long, slender, erect, firm, bright green, usually black and shining at base, with Mat pale membranaceous border, little (2-3 em.) extended above spo- rangium level; stomata present; peripheral strands 4—6 cardinal, plus as many as 14 accessory groups; ligule subulate, triangular; sporangia oblong, 0.5-3 cm. long, marked by numerous brown spots; velum variable, from very narrow to covering half of sporangium; megaspores 280—440 y. in diameter, marked with low tubercules, frequently confluent into short low wrinkles; micro- spores frequently ashy-gray, 20-30 u long, fine-spinulose. Distribution: Illinois to Texas. Wet prairies. Specimens examined: Pfeiffer, Bush, Palmer. 3. Isoetes Butleri Engelm. in Bot. Gaz. 3: 1. 1878. Gray, p. 61; Britton & Brown, ed. 1, 1: 48, ed. 2, 1: 54. Corms 2-lobed; leaves 8-30, 8-15 cm. long, more slender and rigid than J. melanopoda, tapering to apex; stomata numerous; peripheral strands usually 4, sometimes more in number; ligule elongated, cordate at base; sporangium oblong, 6-7 mm. long, marked with brown lines; velum very narrow; megaspores vari- able, commonly 480-650 u in diameter, marked with numerous tubercules, usually distinct, occasionally confluent; microspores 27-37 y long, papillose. Distribution: Tennessee, westward to Kansas and Oklahoma. Limestone barrens. Specimens examined: Eggert, Bush, Palmer. SALVINIACEAE AZOLLA 1. Azolla caroliniana Willd. Sp. Pl. 5: 541. 1810. Gray, p 50; Britton & Brown, ed. 1, 1: 35, ed. 2, 1: 38. [Vor. 20 76 ANNALS OF THE MISSOURI BOTANICAL GARDEN Plants floating on surface of water, often covering large areas, deltoid or triangular-ovate in outline, 6—25 mm. broad, pinnately branched; lobes ovate, lower lobe reddish, upper greenish with a reddish border; megaspores minutely granulate with three acces- sory corpuscles; masses of microspores armed with rigid septate processes. Distribution: Lake Ontario to Florida, westward to Washington and California. On surface of still waters. Specimens examined: Eggert, Engelmann, T'release, Bush, Mackenzie. INDEX TO SPECIES Acrostichum areolatum............ Adiantum CCC eC cece ere se eeenernesees ? Names in italics are synonyms. asplenioides................... 55 —— ERETT x4 doit 55 pycnocarpom... cce 54 irterethi DET OTT ER Fat 55 D — Pc 75 earolinianga...............Lses 75 Botrycbium.............. sees 47 OÉSasclum. 6 5.6 60 65.5 E ark a 48 OU 5 israr a TII 47 var. dissectum.............. 48 var. tenuifolium............. 48 ternatum... 6... ccc ee eee ee 47 var. dissectum............04. 48 var. a cease eevee TEM 47 bp colpo oeer ike sin anaes 48 var. peas a A REE 49 CDU. reda icine 57 rhizophyllus.................. 57 Cheilanthes..............2...... 57 alabamensis................... 57 dealbata.. 2.0.0... ce cee eee 61 b i v5 45-3 oe eee 58 IF c RE 58 lanuginosa... 0.0.06... 58 Vesttia. wb we cede nh 58 es o e s PEETTTTOURT TTC TET. 58 bulbifera. oo... 0.000.000.0000... 59 var. horizontalis............. 60 His... ETA eee eee 58 Dennstaedtia................... 60 punctilobula.................. 60 EURO 5k iiss eR REFER 60 gilostuscula ........... esee. 60 1933 PINKERTON—FERNS OF MISSOURI 77 punciilobula. ...... e cae eee 60. PD.» teen ERES 61 y penc cor mots or 66 AUODUPDUINA.,.... Feed cane oo 62 ocrosticoMel . Vers 6 val. DUME LS io CONSU 62 acq DA Re eae tess Mews 6 Var. Gh.) oe ie) 62 EWR Lolo aS Fascists codes 69 2s os taeda ane E 61 voies FOE o e eek 70 gabela: sos cv Xs oe Phan 62 BV EUER. eerte I PROG. ica CR eR Ee 66 WRI PAPERS RO EE TETTE EE T 70 hezagonoptera. ooo ss Bc esee 68 Mg, RTT CET TT eee E E E 70 einem DEDERIS aia 62 XR UMEN EAE LL rex 59 TM cec ee a ae eer anne 59 eae TOWN IU eee €3 BEL slicers tas 2 rb 59 hay Tp umet 63 Hippochacie POG ooe ra aea 70 ee Ae Sey ee 63 Tr c crc 74 Posen, 204 BORNE 63 push BEocerua ipu s M e E RE 75 acrostichoides. . ....1.... 2... 63 Ingelmannic cs ose 74 var. méisum. anru Bik 6.0035 64 verei NEC 75 var. Schweinitsli. ods oos 64 SEEN eso vies SAR E 66: . Polystichum.. c. en E 66 anaes aa: Peer Perey Terre 600 Preeti ccc Ss 64 DOPU ea ean a a 72 ici e VETUS E T. 64 bce flabelliforme. ..... 78 Piendium..... o CBE 65 lucidulum..................... 73 aquilinum. .... ves nee nn 65 var. porophilum............. 73 latiusculum. 2:5 Es 65 porophilum.......... lessen. 73 var. pseudocaudatum......... 65 o MINER TERT TT TOT eee ree 65 NEN cocos odit eee eee 65 lire us PORE Ee "or 7 Birulhiopteris. ............ LL... 00. "dE Iesse rex 74 EDO. Leo eod nese R 66 uis TOM E Pe: ee ene 74 BODDOlBBUA. o erae rex 60 oc MEME VETERES 74 so dio lí MEME ERU NI, 60 Struthiopteris germanica........... 65 nivea dealbata. ........... Lulu. OMNEM 0o. s aer a ae 66 Nothochlaena pulchella............ 6l hexagonoptera TONER SUA 68 UBND ech ewes a be aes 61 a, TET T 67 os pl MEME NEMINEM dead 61 palustris pubescens............ 66 gg en Tasse ees 64 a Pee RA 67 CDI PRETI TEES et 46 var. ^ NU MES 67 PERDERE, (Coe d xe d Core 47 AA TAn a P Sees TE 66 FUSRLUIM. ou adien aaa cedit (AN T R E T rus 68 MUNN cios REESE T 71 | PME INE MT. n 68 MUDMBODN. ric o ON ada 4 xh r3. o AMEMMMERÉ 69 Olaytoniana o Lond Eo an we ees 72 TO ....... dero DS 69 regalis spectabilis. ............. 71 ae, mee” pe 69 Struthioptreris.................. 64 [Vor. 20, 1933] 78 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 5 Fig. 1. Cystopteris bulbifera var. horizontalis Lawson. From E. J. Palmer No. 32896, in the Herbarium of the Missouri Botanical Garden. Fig. 2. Botrychium obliquum var. tenuifolium (Underw.) Gilbert. From Dr. William — coll. of October 28, 1897, in the Herbarium of the Missouri Bo- tanical Garden INNOSSIN AO SNUAA—NOLUAMNId Ann. Mo. Bor. GARD., Vor. 20, 1933 PLATE 5 wan: Min ade. dai] PG y —— mue e corns tng c " v 3 » z A 3 ^F » f EN o ^ " a Mw 2l e = © wow. Se f oc g $ yg I c z - R9 w os ekz Pu s yer ee? C Er ih i a ba * n 2 $ Us m is - tA “SS BLASTOMYCOSIS: REPORT OF A CASE, WITH A STUDY OF AN ETIOLOGIC FACTOR AND A CLASSIFICATION OF THE ORGANISM MORRIS MOORE Rufus J. Lackland Research Fellow in the Henry Shaw School of Botany of Washington University INTRODUCTION It is the purpose of this paper to describe briefly the disease known clinically as blastomycosis, and to try to clarify the rec- ognition of the organism involved. The literature of the field is at present too extensive for an entire review, and since numer- ous workers have already given excellent discussions on the clinical aspects of the infection, as to its gross pathology, micro- scopic histo-pathology or cellular reactions, and the biological or rather immunological phenomena, a review would be un- necessary here. However numerous such papers may be, there is still much work to be done on the subject. The author has attempted to clear up, at least in his own mind, several undecided points in the disease: first, the establishment of the etiological agent of blastomycosis; second, the determina- tion of the exact classification of the organism. In the past, and even at the present, medical men have grouped under one general heading all organisms which were responsible for the same clinical condition. Good as this system may be for general diagnosis, much difficulty is encountered because of the fact that physicians are inclined to devote very little time to a study of the organism, thus rendering any therapeutic measures, if avail- able, indefinite, inasmuch as several of the fungi present varying degrees of pathogenicity and require different therapeutie meas- ures. Thus we find that numerous species of the genera Saccharo- myces, Monilia, Cryptococcus, Endomyces, Sporotrichum, and others have, at one time or another, been considered etiological agents of blastomycosis. A review of the history will illustrate these facts. HisTORY Years before the first case was definitely described as blastomy- cosis, investigators had performed a certain amount of work on ANN. Mo. Garp., Vor. 20, 1933 (79) [Vor. 20 80 ANNALS OF THE MISSOURI BOTANICAL GARDEN fungi involved in cases of infection and had established these organisms as etiological factors, particularly the yeast and yeast- like groups. Chronologically, the list is quite long, but it is worthy of note. According to Hufschmitt, Sartory, Sartory, and Meyer (731), we find that in 1845 Remak, and in 1853 Robin, in his ‘Histoire Naturelle des Végétaux Parasites de l'Homme et des Animaux,’ discovered the normal existence of the yeast Cryptococcus guttu- latus in the rabbit intestine. A few years previously, Hannover (cited in Buschke and Joseph, '28) had found yeast in the urine of diabetie patients. Investigations then tended to turn to the parasitism of these organisms in animals, with the result that Bernard during the course of his work on fermentations attempted the first animal experiment by injecting beer yeast into these subjects. Following this work, Popoff, Grohe, Roussy, and several others showed the pathogenic actions of the yeasts on mammals, and Rivolta in 1873 in his ‘Parasiti Vegetali’ dem- onstrated for the first time a yeast infection in a horse. In the meantime, Metchnikoff and Weismann showed the parasitism of the Saccharomycetaceae in the lower animals. In 1892 Wernicke showed the first mycosis in man and named it ** maladie protozoique de la peau." The following year Troisier and Achalme (793) definitely established the relation between yeasts and man. In the meantime, several workers attempted to show the destructive ability of these organisms on the animal tissues. Popoff in his work had used dogs as his subjects, but impure cultures. Raum ('91) inoculated animals with large amounts of yeasts, and a rise in their temperature, shortness of breath, and death resulted. Neumayer (91) fed animals with cultures and also inoculated them subepidermally. His feeding developed a gastro-enteritis which he believed due to fermentation, since the skin inoculations were of no value. The yeasts of these workers were probably of the non-pathogenic types, for L. Rabinowitsch (96), a few years later, showed fifty various yeast-like organisms with seven pathogenic for animals. Nesczadimenko (’99) made peritoneal injections of yeasts in a physiological saline solution into rats, mice, guinea-pigs, and dogs, with death ensuing from eight to twelve days. He concluded, however, that these organisms were not so deadly, although causing this mortality. 1933] MOORE—BLASTOMYCOSIS 81 The first actual case of blastomycosis, so-called, was reported by Gilchrist at the June, 1894, session of the American Dermato- logical Association. His paper resulted from the finding of peculiar yeast-like bodies in the diseased tissue of a patient. The doctor attending the patient had given the diagnosis as a typical case of chronic scrofuloderma. Several months after Gilchrist’s report, Busse (’94) brought to light the extraordinary case to which he later gave the name “‘Saccharomycosis hominis." The patient was a woman thirty-one years of age who had suffered from a localized subperiosteal inflammation of the left tibia. An examination of the abscess, which opened spontaneously, re- vealed ‘‘numerous doubly contoured, very refractive, roundish and ovoid bodies,” and these were found to be situated both intracellularly and extracellularly in the pus and abscess wall. These organisms when isolated in pure culture and then inocu- lated experimentally in animals proved to be what were later known to be blastomycetes. The patient later developed super- ficial ulcers on the face, subperiosteal swellings on the right ulna and the left sixth rib near the axillary line, with death ensuing. Busse cultured the yeast from the ulnar swellings and from the bottom of the ulcers. Approximately two years later, the first case reported by Gilchrist was published in detail in the Johns Hopkins Hospital Reports of 1896. In the meantime, however, several others had noticed similar cases among guinea-pigs, horses, mice, and other lower animals (Sanfelice, 795, 796, 96a, Roncali, '95, Corselli and Frisco, '95, Tokishige, '96, and others). In 1896, Curtis isolated a fungus similar to that described by the former writers from a myxomatous tumor of theleg. In the same year, Gilchrist, in conjunction with Stokes, published a short paper on a second case of blastomycosis, and this was published in detail two years later (Gilchrist and Stokes, ’98). Simoni (’97), working on the diseased tonsils of patients, found budding yeast-like cells in twenty tonsils. Maffuci and Sirleo (98) examined numerous tumors and found budding cells in a great number of tissues. Many other reports followed, as that of Hyde, Hektoen, and Bevan (’99) with a supplement by Hek- toen (’99) later in the year, Hessler (’99) with a case report, and or. 20 82 ANNALS OF THE MISSOURI BOTANICAL GARDEN several during the same year and 1900. In the following year appeared the elaborate work of Ricketts (01), with a study of the organism from a case of systemic blastomycosis by Otis and Evans (03). Eisendrath and Ormsby (705) described a systemic infection, and Irons and Graham (706) reported a severe gener- alized systemic disorder. Hektoen (’07) gave a comprehensive review of the literature, and from that time on the medical journals have published too great a number of cases of infection due to yeast-like organisms, under the heading of blastomycosis, to render a complete survey of literature a matter for a paper of this length. ETIOLOGY AND CLINICAL MANIFESTATIONS The disease known clinically as blastomycosis is very likely due to a plurality of organisms and not species of the same genus as indicated by previous writers. This is clearly evident as seen by the great number of papers published and the cases reported, involving such fungi as Saccharomyces, Oidium, Monilia, En- domyces, Cryptococcus, Coccidioides, and even such a form as Sporotrichum. The clinicians have referred to the category of blastomycosis any clinico-pathological condition which may be due to yeast-like or budding fungi. It must be understood, therefore, that the term as here used refers only to the clinical aspect of the condition. In America, clinicians and medical men, more especially medical mycologists, are inclined to class as the cause of blas- tomycosis only that organism which was described originally by Gilehrist and Stokes in 1894 and in this view the author is greatly in accord. On the other hand, European workers consider only that organism which was reported by Busse (94) and so elaborated on by Buschke ('98). However, by reason of priority, the organism of the former workers should hold the position so designated and be established as such. Further remarks on the Gilchrist organ- ism will be found in the discussion. Blastomycosis presents numerous clinical manifestations and in this respect it is protean, being found in practically every organ of the human body either in biopsy or autopsy material. No immunity towards the invading organism is established by any 1933] MOORE —BLASTOMYCOSIS 83 of the anatomical structures. Clinically, the condition presents lesions which are alike both for the cutaneous type of the disease, that is that group of infections which may be found occurring superficially, or for the systemic type of the disorder, occurring in the lungs, bones, meninges, liver, or other viscera. This division is based on the work of Jacobson and his associates (32), who further separate the cutaneous type into that of primary in character, as occurring in the epidermic layers or the cutis, as shown by Hagiwara (22) and Hashimoto (722), whose organisms, although not of the Gilchrist type yet coming under the general heading of blastomycosis as known clinically, also the work of Grschebin (727) ; and secondary, due to an infection of the deeper tissues, internal viscera, or bony structures, as shown by Irons and Graham (706), Ryerson (709) for bones, and many others. The primary form or the cutis infection Jacobson further designates as presenting one of three varied appearances: papulo- ulcerative; verrucous or papillomatous; gummatous. The papulo-ulcerative type Jacobson designates as being initial lesions which are papulo-pustular in character and of epidermic origin (shown by Hessler, '99, Hektoen, ’99, Ricketts, '01, Engelhardt, '24, and Fabry, ’27, '28). These lesions rupture in the course of time and empty out the purulent exudate on the surface of the skin, with the probable ultimate formation of crusts. The process may be proliferative and involve a great area of the immediate vicinity. The lesions usually show a violaceous border with the involution of the peripheral surfaces and perhaps consequent scarring and atrophy. The verrucous or papillomatous type (Froilano De Mello and Rodrigues, '29) is characterized as being nodular or papular in character and present on a normal or deep-red, infiltrated skin. Several of the lesions may coalesce to form papillomatous patches which resemble verrucous tuberculosis. These lesions may break up into healing areas which upon drying present irregular scars. The characteristic color as noted above is found here too, as well as the sloping periphery. The gummatous type develops from the subcutaneous layers of the tissue of the deeper portion of the cutis in the form of small, [Vor. 20 84 ANNALS OF THE MISSOURI BOTANICAL GARDEN slightly elevated, somewhat tender, reddish, deep-seated, soft nodules situated on the characteristic violaceous-red surface of the skin. There is a diffusion of the color with subsequent establishment of new nodules in the vicinity. The nodules en- large, become soft and gummatous, and then break down to form masses of ulcerative, proliferative materials bordered as in the other two types, and contain numerous abscesses. The secondary cutaneous form consists chiefly of variously formed ulcers which give off a purulent or sanguino-purulent discharge from a soft, granulating floor. Some may develop crusts with raised edges, while some may assume hyperplastic functions with papillomatous characteristics, and usually there may be a metastatic action on the part of the ulcers represented by the formation of new lesions which are surrounded by a dark red or purplish zone. Healing may be spontaneous with indurated scars, as noted in some cases, or infection may persist but may finally succumb to treatment with iodides as was noted in the case reported here. A study of the ulcers formed in blastomycosis shows them to originate in abscesses which from a clinical point of view can be divided into the superficial and the deep types. Secondary, cutaneous, superficial ulcers arise usually in the subcutaneous tissues as nodules of varying size as shown by Stober (14), Engelhardt (24), Ferguson (’28), and Montgomery and Ormsby (08). These ulcers usually enlarge, rupture, and spread the material over the surface of the skin, setting up new foci, or in some cases they have been found to dry up and disappear. The deep type of secondary cutaneous blastomycosis (Grsche- bin, '27, ’28), characterized by smaller number and deep- seatedness, is by far the most serious of the two, involving de- structive processes of the bone, muscle, and deep tissues and organs. It rarely shows any inflammatory reaction, but can be distinguished by the purulent character of the abscesses as con- trasted with the mucoid or mucopurulent nature of the super- ficial abscesses. The above types represent the typical forms occurring in a clinic. However, Weidman and Douglas (’21) reported the occurrence of a sarcoma-like tumor on the leg of a patient, which 1933] MOORE—BLASTOMYCOSIS 85 looked like lupus vulgaris and yielded blastomycetoid bodies on histological sectioning. Then, about six years later, Cleland (27) reported a case with the formation of a myxomatous-looking tumor mass which also showed typical cells on sectioning. These, however, are rare and until more cases are reported cannot be placed in the definite clinical types. Under the heading of cutaneous blastomycosis, Castellani has also established principal types of blastomycosis of the cutis from a clinical point of view. 1. Blastomycosis verrucosa (Synonym: Blastomycosis, Gilchrist t s Etiology: Cryptococcus dermatitidis Gilchrist and Stokes, 1896 (Synonym: Cryp- tococcus gilchristi Vuillemin). Castellani here creates the genus Blastomycoides. 2. Blastomycosis ulcerativa profunda sen mutilans (Synonym: Blastomycosis, Wernicke-Ophüls type or Blastomycosis coccidioides type). Etiology: Coccidioides immitis Rixford and Gilchrist, 1896. 3. Blastomycosis purulenta profunda (Synonym: Blastomycosis, Busse type; Blastomycosis subcutanea purulenta). Etiology: Cryptococcus hominis Vuillemin, 1901 (probably covers many species). 4. Blastomycosis glutealis (Synonym: Blastomycosis, Kartulis type Etiology: mycological investigations iis dis completed. The fungi seem to belong to the genera M onilia and Cryptoco astomycetica Pil eee (Folliculitis decalvans cryptococcica, pro parte, Castellani type). (Synonyms: Furunculosis cryptococcica; Pseudo- furunculosis blastomycetica; Furunculosis mycetica; Folliculitis decalvans sacchar- omycetica J shape decalvans moniliaca). Etiology: yeast-like fungi either of genus Cryptococcus or Monilia (No asci or ascospores ects to Castellani). In addition to the above types, Castellani adds the following, although they are in no way connected with the blastomycosis organism: Blastomycosis epidermica; Intertrigo blastomycetica; Dermatitis blastomycetica interdigitalis. For clinical purposes in diagnosing skin infections such a classification is good, but for correct determination of the etio- logical agent, it is essential that each type of involvement be named with the infective agent designated as such. For example, if the organism be a saccharomycete, the disease should be called saccharomycosis; if a monilia, then moniliomycosis; if an endomy- cete, then endomycosis. Thus, when the organism is correctly Fr the amount of time necessary to determine the right sort of curative measurefor that type of infection may be taken, and the amount of time necessary for healing reduced. ó [Vor. 20 86 ANNALS OF THE MISSOURI BOTANICAL GARDEN SYSTEMIC BLASTOMYCOSIS As mentioned previously, the disease is protean in its clinical manifestations, with the result that practically every organ in the body has been shown to be infected either in the living, by various measures, such as X-ray, or in autopsy material. No vital organ is immune, and this in itself is sufficient to cause a careful physician to give more attention to therapeutic measures. This universal infectivity of the agent was especially shown by such writers as Otis and Evans ('03), Eisendrath and Ormsby (05), Le Count and Myers (’05), Irons and Graham (06), Montgomery and Ormsby ('08), Wade and Bell (16), Garr (725), Panja (25), Toepel (29), and Maner and Hammack (30). There is, however, a difference in frequency with which the vari- ous organs show their susceptibility. 'The portion of the anatomy that shows the greatest amount of infection is the skin, either primary or secondary, having about 95 per cent of all cases recorded. This phase of blastomycosis has received the greatest amount of attention principally because it is so prevalent, but also because it is usually a manifestation of a metastasis from the deeper organs, and this helps bring forth the diagnosis of blastomycosis. The pulmonary system, including the lungs and bronchi, constitutes the second most frequent and the most common sys- temic form, being present to the extent of approximately 95 percent in systemie infection in available autopsy records, and about 35 per cent in primary cases. This was shown by such menas Stober (14), Wade and Bell (16), Wade (18), Dennis (18), Miller (27), Medlar (27), and Mazza and Niño (28). The disease probably is primary in the bronchi and from there spreads to thelungs. If secondary in the lungs, as in systemic disorders, the process may be slow and chronic; if primary, however, the spread may be rapid and fatal. The kidneys are next in frequency of infection. The genito- urinary involvement is usually secondary by way of metastatie foci through the blood. The disease in these organs manifests itself in the form of nephritis, showing casts and albumen in the urine. The culturability of the organism from samples, however, 1933] MOORE—BLASTOMYCOSIS 87 cannot be demonstrated unless the kidneys are accompanied by infected bladder or prostates. The spleen follows next in order, but this organ is usually easily susceptible so that a great amount of infection is to be expected in any systemic disorder of this sort. The complication of the bones and joints seems next in the amount of infection. This form of the syndrome is very common in systemic disorders, and may even be a primary infection if the diagnosis in the patient here described was correct seven years previous to his entry at the Barnard Skin and Cancer Hospital. This type of the disease may manifest itself in the form of arthritis, osteitis, osteomyelitis, or periostitis, according to Ryerson ('08-'09) and Stober (’14). The process usually results in a suppuration, formation of sequestra, and abscesses which break down intervening cell walls and coalesce, causing great damage. The liver appears to be a rather frequent subject to the in- fection, coming next in the order of frequency. This is to be expected in systemic disorders where the blood plays an impor- tant part. Metastases through the blood vessels are fairly com- mon, and yeast cells are easily transported to the main organs in this manner. The pleura too are susceptible to a great extent, and here the proximity to the lungs is a great factor in their thfedtion. The lymph glands follow along rather closely, as shown by Wanamaker (728), especially for the cervical lymph glands. Cerebrospinal involvement, including the brain, meninges, spinal cord, and skull bones, occurs fairly often as a secondary metastatic process in generalized systemic blastomycosis, ac- cording to studies made (J. T. Moore, ’20, Freeman and Weid- man, '23, Greenfield, '24, Wilhelmj, '25, and Gáspár, '29), being found in at least 12 per cent of the cases. When the disease is secondary to a general systemic infection, there may be osteo- myelitis of the skull bones with destructive processes. The diagnosis rests not upon any clinical entities which may be present, because the inflammatory reaction simulates many other condi- tions, but upon the actual laboratory finding of the organism either in the spinal fluid or in sections of the diseased brain tissue. [Vor. 20 88 ANNALS OF THE MISSOURI BOTANICAL GARDEN Wilhelmj (725) states that in those cases where there is no patho- logical symptomatology or clinical manifestation on any other part of the body, and when the meninges are infected during the primary stages of the invasion, death may occur without the initial appearance of general metastatic foci, and such a condition he calls primary cerebrospinal blastomycosis. The spine may be involved in the process in a suppurating condition (Parker, ’23), but this condition is relatively rare. Jacobson lists the vertebrae as being next to the brain in susceptibility to attack. This condition has been noted on several occasions. Roentgenographic studies usually reveal an infection of bodies of the vertebrae, and lamina and posterior ligaments may show an involvement which simulates greatly tuberculous Pott’s disease. Prostatic infection in blastomycosis is often noticed (Parmenter and Simpson, '19). Usually it is associated with a genito-urinary complication and involves the urinary bladder (Rhamy, ’26). In these cases acute urinary urgency and pyuria are well-defined symptoms. The heart lesions in blastomycosis are shown first in the pericardium and then in the myocardium in the form of an in- flammatory reaction (Hurley, 716). Pancreatic involvement follows in frequency. Infection in the peritoneum is the next most common. Jacob- son reports finding the disease in the abdominal viscera in the following decreasing order of frequency: kidneys, spleen, liver, lymph glands, pancreas, peritoneum, adrenals, and gastro- intestinal tract. These organs, as pointed out previously, become involved usually through metastasis by way of the blood stream or by direct transmission from tissue to tissue. In this manner, testicular blastomycosis is usually developed. Blanchard, Swartz and Binot (’03), as early as 1908, noted an intraperitoneal in- volvement. The eyes may also be involved. The infection here is very painful and often dangerous, leading to blindness with perhaps a complication of the nervous system and the brain. McKee (26) and Ferguson (’28) noted cases of the eye which were secondary infections due to a metastasis from the pulmonary apparatus. 1933] MOORE—BLASTOMYCOSIS 89 Laryngeal and tracheal blastomycosis are rare infections. Jacobson records four in America and one in Europe. All the patients were adult males working either with the soil or its products, one being a clerk in a general store (Dennis,’ 18, Downing, '18, C. Jackson, '26, and New, '28). The larynx showed a “chronic inflammatory mucosa with a grayish, minutely nodular surface in some portion of the lesion, with a few minute, isolated, yellowish nodules." ‘There was often a reddish, raw portion of the larynx due to the ease with which the superficial layers came off with coughing. The process resembled very closely tuberculosis. Involvement of the tongue is perhaps a rare occurrence, but cases are not reported in great numbers purely because sputa smears usually show a variety of yeasts and thus no definite etiological significance is attached to those obtained. The first case was reported by Copelli (13), a second one by New (17) from the Mayo Clinie. Since that time, however, Mazza and Canevari (29) reported a case from Argentina, and Niño (29) reported an infection of the lower lip with the involvement of the tongue. Such an infection usually hinders respiration, inasmuch as a tumor-like growth, as evidenced by Copelli and New, de- veloped which enlarged in the back portion of the tongue and practically filled the entire larynx. SYMPTOMATOLOGY Blastomycosis when of the primary cutaneous type presents no clinieal symptomatology of discomfort or pain except for that expressed because of the lesions. When of the systemic type, however, the condition is very different. There are numerous clinical factors to make a picture which might easily be confused with a number of diseases. According to numerous investigators, the onset of the disease varies with the person and amount of infection. It may be intense and acute, leading to death in a short time, or insidious and mild, with a prolonged chronic condition, death finally occurring as a result of a secondary complication. There is a typieal set of symptoms once the disease is well established. This consists of typical malaise, recurrent chills, loss of weight, [Vor. 20 90 ANNALS OF THE MISSOURI BOTANICAL GARDEN as evidenced in the present case, loss of strength leading to general emaciation, night sweats, although morning sweats may be present too, irregular fever, pain in affected parts, and a rapid pulse. The disease, as noted before, may be primary in the skin with subsequent spread systemically or it may be systemic with the formation in the later stages of nodular growths on the skin. Unless the patient is well taken care of, systemic infection re- sults, leading ultimately to death. DIFFERENTIAL DIAGNOSIS The final diagnosis of blastomycosis rests on the finding of the organism either culturally in a lesion of the patient or, if that be unavailable as in systemic disorders, the identification of the fungus in biopsy or autopsy material. As pointed out previously, the disease is protean in character, with the result that a careful examination must be made to diagnose it blastomycosis, comparing it with the several well- known clinico-pathological entities which it may simulate. The most noteworthy of these complicating diseases are as follows: (a) The dermic lesions described previously, developing necrotic and papillomatous growths or ulcers, resemble very closely epitheliomas, differing only in the rapidity of evolution and the absence of deep induration, verrucas, tuberculosis in its various forms, and syphilis. Its resemblance to sporotrichosis (Lewis, 17) has often been noted, but it differs in being less sluggish. It differs from syphilis only by the softness of the lesions, the red- dish-blue ring around the lesion, and by a negative Wassermann, with, of course, the presence of the organism in the blastomycosis infection; (b) The systemic infections of blastomycosis must be distinguished through laboratory methods from a great many complications, particularly coccidioidal granuloma described in a previous paper by the author (M. Moore, ’32). The organism of coccidioidal granuloma, Coccidioides immitis, reproduces by en- dosporulation, and the blastomycosis organism, through budding. Furthermore, the lesions in the so-called ‘‘California disease" are more rapid in evolution than in blastomycosis; (c) Infection of the glands which is quite rare in blastomycosis often suggests 1933] MOORE—BLASTOMYCOSIS 91 lymphatic leukemia, Hodgkin’s disease, and possibly lympho- sarcoma; (d) abre i testinal lesions may often resemble typhoid and in some cases the isolation of the organism is necessary to rule out this disease; (e) Osseous infection quite often resembles tuberculosis, petticularly as reported by Ryerson (708-09); (f) Pulmonary blastomycosis very often presents the same clinical, histological, and pathological pictures as tuberculosis, as noted by Medlar (27) and Miller (27); (g) Infections of the brain often confuse the pathologist or clinician with its similarity to torula (Freeman and Weidman, ’23), epidemic meningitis, and even tumors of the spinal cord; (h) Sutejew, Utenkow, and Zeitlin (29) find that the use of bromides and iodides evidently causes an allergy which in its reactions presents lesions similar to those caused by the infective agent of blastomycosis and is often confused with it, the latter differing in their more rapid evo- lution. It would seem, therefore, that the recognition of blastomycosis is not a very easy matter. The really important fact concerned with this work is to find the organism, which usually requires laboratory technique, and to verify its pathogenicity by animal inoculation in order to comply with Koch’s postulates. Predisposing factors.—The infective agent shows no particular preference as to sex, although more cases have been males than females, of the industrial classes, chiefly the workers of the soil and its products, a fact well exemplified in the present case. There is no discrimination as to race or color, all peoples being affected in like fashion. Stober ('14) correlates the type and amount of infection with the habitat and environment of the patient. As far as is known, all ages are susceptible, a similar condition existing for many other fungi which tend to become systemic, such as coccidioidal granuloma. Treatment.—The successful therapeutic measures in blastomy- cosis are few and limited. The best treatment or cure for the disease rests of course on the skill of the physician in detecting it at an early stage before it has seized a definite foothold, when it can be kept from becoming systemic. If the disease is definitely located in a particular section, surgery may be used to eliminate it, as suggested by several authors (Wade and Bell, ’16, R. H. [Vor. 20 92 ANNALS OF THE MISSOURI BOTANICAL GARDEN Jackson, '26). Cautery has also been used with beneficial results. Hedge ('28) has employed carbon dioxide snow in freezing cutaneous lesions with measurable success. X-ray treatment has also been used frequently, being combined with the administration of iodides. On the whole, primary cutaneous lesions yield fairly readily to iodides and even the application of crystal violet and gentian violet, although their use if at all successful is empirical because dye therapy does not rest on any scientific basis (Spring, '29). In systemic infections, however, one has to contend with complications. The extent of the infection cannot be determined easily. The only cases to recover from an infection of this sort are those to which special care and attention coupled with a change to a clean, pure atmosphere had been made, with the administration of large amounts of saturated iodide solution, either potassium or sodium, although the former has been used more extensively. The dyes mentioned above have proven worthless to the systemic type of disorder. Several workers have advocated copper sulphate, but many others have found this to be useless. Roentgenotherapy has, as yet, no really therapeutic importance. Stober has applied immuno-therapy in the form of a vaccine of suspended blastomycotic cells, heated to 110° C., but no definite results can be shown. It would seem, therefore, that therapeusis is greatly in need of investigation. Immunological reactions.—Immunology in blastomycosis has not reached any definite point as yet. Agglutinins have been reported by some, and negative results by others. Precipitins have also had the same reaction, as well as complement fixation. On the whole, results are indefinite and a good deal more in- formation is needed. The main difficulty seems to be that the toxins of the blastomycosis fungus are difficult to demonstrate. So far, immunology is an open book with only attempts proving nothing written on its pages, and it is to be hoped that more work can be done along this line for the benefit of those who may be inflicted with this syndrome-complex. Mortality.—The number of deaths resulting from blastomycosis is a factor worthy of note, since systemic disorders due to the 1933] MOORE—BLASTOMYCOSIS 93 typical blastomycotic organism usually prove fatal. This is so because systemic infections are rarely diagnosed as such until there have been cutaneous outbreaks, with the result that thera- peutic measures are given too late for beneficial results. It is difficult to quote figures because there are many cases occurring which have rapid recovery and the physician does not report them. Moreover, true blastomycosis is difficult to diagnose un- less smears and cultures are made from the abscesses. In many instances, several attempts are necessary before any fungous growth is obtained, and unless the investigator is well trained in mycological technique and in the recognition of such forms typical of the Gilchrist organism the application of Koch’s postulates would be essential, particularly with mice and guinea-pigs. REPORT OF CASE Clinical History—M. H. L. (Hospital No. 50095), a white, widowed male, farmer by occupation, 43 years of age entered the Barnard Free Skin and Cancer Hospital, at Saint Louis, Missouri, April 5, 1932, with an ulcerating, proliferating infection of the left hand and forearm, which the patient claims to have had for 5 years. Family History.—Father died as a result of high blood pressure, hypertension, at age of 56 years. Mother dead due to throat trouble, at age of 30 years. Three brothers and one sister dead due to an infancy cause. Patient has 7 children, all living and well. Wife dead due to double pneumonia at age of 38 years. No history of diabetes, nephritis, syphilis, or cancer in fami Past History.—Past health has been good; usual obildliood diseases. He had pneumonia at 15 years of age, bronchial pneumonia at 18 years of age, and influenza in 1918. He has had no operations other than those to be mentioned in the present illness. Right leg injured 7 years ago Personal History.—Patient denied any venereal disease. Present Illness.—Patient stated that 7 years previously, a limb had fallen on his right lower leg, causing a knot to form. The physician who examined his leg thought he had a periostitis and opened the lesion on the leg. About 2 years later, or 5 years previous to entering the hospital, a mossy, verrucous-like lesion similar to that on the leg appeared on the back of the left hand. Both the lesions on the right lower leg and on the hand continued to spread. The patient went to the Vanderbilt Univer- sity Hospital, at Nashville, Tennessee, about a year later, at which time the eruption had spread over most of the anterior and lateral sides of the right lower leg with two lesions of the right thigh above the right knee and with involvement with most of the back of the left hand and part of the forearm. He remained in the Vanderbilt Hospital for about 3 months, during which time the lesions on the right leg and arm were curetted and treated with X-ray. At the time of his discharge, the lesions on his right leg had healed, as had most of the hand except that portion near the thumb and the wrist. Another currettement was performed, but the lesion per- sisted. He received two X-ray treatments each on the leg, arm, and hand. Six years previous to entering the Barnard, he had lesions on the right elbow region; [Vor. 20 94 ANNALS OF THE MISSOURI BOTANICAL GARDEN these lesions healed. About 4 to 6 years before his entry he had lost some weight, but at the present time complains of no recent loss of weight. He gives no histo of hemaptysis, night sweats, fever, or frequent colds. About 10 days previous to entry, a reddish lump developed on the flexor surface of the left lower arm near the elbow. This lump is not very tender or painful. He has been applying potash on the ulcerated area of the left hand Physical Examination.—Patient is a thin, coóperative, moderately active, white male 43 years of age, who has had the condition herein to be described as blastomy- cosis of the left arm and hand for approximately 6 years and who had the same con- dition on the leg for 3 years until it was curetted and X-rayed 4 years previously. The right leg and left arm show scarring which will be described later. Head.—Normal size and shape Eyes.—Eyes react to light and accommodation. Reactions normal. Ears.—No discharge, — normal. Nose.—Septum intact, no ulce rales —Slight redness. Mouth.—Several teeth mi Neck.—No stiffness or rigidity, tonsillar and cervical — pee Thorax.—Thorax hairy, thin, and symmetrical; expansion fair. Lungs.—Breath sounds a little harsh and rough over both | E principally the right and right apex. Voice sounds normal, but louder on the right than on the left. No persistent or moist ráles or ráles heard after coughing. Expansion and resonance normal. Cardiac.—Cardiac sounds a little slow and distant, but of a normal quality. No enlargement or pathological murmur. Blood pressure, 106 systolic and 66 diastolic. Abdomen.—No masses or tenderness. Inguinal glands a little enlarged. Genitalia.—Normal male; no discharge or penile sore Reflexes.—Superficial—present and active. Babinski negative. Deep—present and active Extremities.—Left arm and hand—The lower third of the left forearm, back of the hand and extending 2 or 3 cm. on to the palm, is involved in an atrophic process sharply defined superiorly and inferiorly with some scaling, no telangiectasis. On the external surface of the forearm extending on to adjoining parts of the hand and thumb is an ulcerative process which has been covered with a black crust. Ul- cerated area of the left thumb and the flexor surface of the left forearm consists of warty-like and cone-like, multiple abscesses with some elevation of the borders. There is a deep scar on the right elbow region. The anterior, medial, and lateral aspects of the right lower leg are covered with a thin, smooth scar about 12 to 14 inches long and covering two-thirds of the right lower leg. There are two scars just above the right knee, about 4 inches, and 8 X 4 inches in diameter, respectively. On the flexor surface of the left forearm, just below the elbow, is an abscess which is red in color, oval in shape, semifluctuant, practically non-tender and not hot. Extending up from and about the abscess is a chin of firm lymph nodes. The lymph nodes above the left elbow and epitrochlear region are a little enlarged. Laboratory Findings.—Urine negative, being pale straw in color and clear. Blood tests showed 4,600,000 red blood cells and 10,200 white blood cells, with 84 per cent hemoglobin. The differential blood count showed 26 lymphocytes, 8 large mononu- clears and transitionals, 156 polynuclear neutrophiles, 2 polymorphonuclear eosino- philes, and 2 basophiles. Wassermann negative. Smears from the left hand and 1933] MOORE—BLASTOMYCOSIS 95 arm showed budding forms of yeast cells on April 7, 1932. Blood, urine, and pus from left hand and lower arm and pus withdrawn from abscess of left arm were inoculated on glucose-glycerine agar on April 8. Spinal puncture on April 12 showed a clear fluid with a normal pressure. Glucose-glycerine agar and blood agar were inoculated with spinal fluid, with no growth occurring April 9, 1932.—The patient was started on potassium iodide with dosage up till signs of intoxication, and an ointment saturated with gentian violet was applied locally. This treatment was followed with 10 per cent sodium iodide intravenously for several days when patient showed an improvement April 16, 1932.—X-ray of left forearm and wrist, right femur, and right leg showed no abnormal bone changes. The hilar and bronchial structures of both lungs showed considerable thickening, inflammatory in character, but the parenchymatous portion of both lungs appeared free from any active pathology 7, 1932.—Patient showed & very marked alone AM with still some evidence of trouble around the left thumb. Clinical Diagnosis.—Blastomycosis of left hand and a The agar inoculated from the left arm developed a ania which went through the three stages typical of the organism of blastomycosis: the yeast-like growth; the prickly type of growth; and the cottony type of growth. This culture was used for the studies carried out in this paper. ANIMAL INOCULATION A suspension of a 10-day-old culture of the organism in sterile saline was inoculated in a mouse, intratesticularly. An orchitis developed and the mouse showed typical malaise, emaciation, loss of appetite, loss of weight, rise in temperature, with death ensuing in three weeks. The internal viscera showed numerous, pin-point lesions at autopsy, which when squeezed exuded a muco-purulent material from which the yeast cells were isolated. This was in accordance with Koch’s postulates. TECHNIQUE In order to ascertain the morphological characteristics of the fungus, the organism was suspended in hanging-drop cultures, allowed to grow, and observations made from time to time. For finer detail and structure, several transfers were made to slides on which had been placed a drop of a mixture of glycerine (Merck C. P.) and a 1 per cent solution of crystal violet. The fungus was allowed to remain for a period of one-half to one hour to allow for a clearing of the material and sufficient staining to render satisfactory results. This method proved adequate for the work here described. Another method was used also, whereby material was fixed on a slide smeared with albumin and then stained with [Vor. 20 96 ANNALS OF THE MISSOURI BOTANICAL GARDEN methylene blue and eosine. The first method, however, was sufficient. DESCRIPTION A study of the organism in lesions reveals a yeast-like growth of spherical or ovoid, budding and branching cells with no evidence of any filamentous hyphae. These cells, measuring approxi- mately 7-12 y in diameter and sometimes as much as 20 y in length, may occur singly, in groups of two's, three's or four's, as individual colonies. On closer examination microscopically, these yeast-like cells (pl. 6, figs. 1-9) show a heavy, reticulated, granular, and in many cases, vacuolated, protoplasm, and a very definite nucleus with emanating streams of cytoplasm. In many cells the nucleus is barely distinguishable, being a mass of cyto- plasmie structures, but further study after several subcultures shows up this part of the cell very frequently. Many of the yeasts of this group may show a double-contoured, highly refractile membrane, and this feature is of little diagnostic use unless dem- onstrated in tissue sections, as by Rewbridge, Dodge, and Ayers (29) for Endomyces capsulatus, and by Moore (McBride and Thompson, '33) for E. capsulatus var. isabellinus. Thus Wade (16) ascribes to the fungus structure in tissues: (1) an inner delicate capsula vera; and (2) an outer applied capsula sclerotica. In any case, the capsule is lost on repeated cultivation on arti- ficial media. With agar as a substrate, the yeast-like cells elongate (pl. 6, figs. 10-12, 14-17) and on acid media become thin-walled and long with a diameter of 2-214 u. These hyphae become inter- twined and are composed chiefly of isodiametric cells. On neutral or slightly alkaline media, with protein as the chief source of nitrogen, the hyphae are thicker and shorter, with a diameter of 3-4 u. This condition is especially true of media with an excess of carbohydrate as found in glycerine agar. Budding cells are numerous on slightly acid media, being about 9 u in diameter (pl. 6, figs. 26 and 28). The hyphae, at first clear, become granular, and at various points along the sides develop numerous small, knob-like pro- jections of the limiting membrane, which enlarge, round out or become pyriform and sessile, measure approximately 5 y in 1933] MOORE—BLASTOMYCOSIS 97 diameter, and occur usually near a septum. These are known as conidia and may remain attached, at times to small stems or sterigmata, or may become free and develop in the media by budding (pl. 6, figs. 24, 26, 28-29). Racquet mycelium (pl. 6, figs. 23-24), a phenomenon char- acteristic of the fungi of this group and especially of the Tri- chophytons and various other Ascomycetes, is common here, having the swollen portion 5-6 y in diameter and the narrow section 3-314 y in diameter. Chlamydospores may be found arising in the hyphae or terminal as hypnospores, varying in size and shape from round cells 7-81% y in diameter, to elongated, widened cells 5144-714 u wide and 12-15 y long (pl. 6, figs. 22, 25, 27—29, 31, 32, and 37), or they may arise as sessile cells from the hyphae (pl. 6, fig. 36). Round, thick cells with a coarse, granular cytoplasm are particularly evident on cornmeal agar (pl. 6, figs. 26 and 28). When first examined in a hanging-drop preparation, one may see oil droplets on the hyphae which are strongly suggestive of endospores. These disappear, however, when the organism is stained. In some cases, as in pl. 6, fig. 25, a hollow sphere or vacuole surrounded by a hyaline, gelatinous substance may be present in the filament. The development in tissue, as has been noted, is chiefly by budding or gemmation. ‘The process begins with the projection of the inner layer of the cell or endosporium, following Hektoen, which pushes the transparent zone and outer membrane in front of it. The bud becomes enlarged and surrounded by the same walls as surround the mother cell, and division takes place by the presence of a cross-wall which is formed by the pinching in of the cell (pl. 6, figs. 1-3, 6-7). On artificial media, on the other hand, proliferation of the fungus is brought about through sexual reproduction which is heterogamous. Two terminal cells may fuse (pl. 6, fig. 19) or two hyphae growing side by side may send out lateral cells which copulate (pl. 6, fig. 18). In either case, a spherical ascus results which may be terminal on a long filament or lateral on a short peduncle (pl. 6, fig. 33) and has a thick capsule (pl. 6, figs. 21 and 33), sometimes surrounded by a sheath as in pl. 6, fig. 35. The latter case, however, is rare and was [Vor. 20 98 ANNALS OF THE MISSOURI BOTANICAL GARDEN observed only three times. There are formed 8 spherical to ovoid, smooth spores which are hyaline when young and held in a gelatinous substance and when older become chamois-colored and granular, and have at maturity a diameter of 2-3 y, varying on different media (pl. 6, figs. 38 and 33). This organism therefore agrees with that described by Gilchrist and Stokes as Blastomyces dermatitidis, but further observation has here been made on the sexual development. CULTURAL DESCRIPTIONS The culture in this case was growing on glucose-glycerine agar, being an inoculum from an abscess on the left arm. All cultures used in the cultural determinations were taken from the above tube and grown at room temperature, approximately 22° C. Having the two stages so common with yeast-like organisms and characteristic of several members of the Endomycetales, it was thought desirable to transfer the fungus on a wide variety of media and pH range. Possessed with saccharomycetous prop- erties, on the one hand, and filamentous fungous affinities, on the other, the above method of culturing proved to be satis- factory in this work. The following media used are arranged in the order of decreas- ing hydrogen-ion concentration. Raulin's Solution (pH 4.1).—Culture shows a thin, smooth suspension of yeast-like cells, budding and branching, varying in size from 414-514 u x 7-815 u, with several showing a change to filamentous formation. Richards’ Solution Agar (Media consisting of Richards’ solution with the addition of 1.5 per cent agar. pH 4.4).—Growth sparse, of fine filaments. Colony 3V$ cm. in diameter at end of 30 days. Culture shows long hyphae projecting from edge of growth, 214- 3 u in diameter, with numerous budding cells. Filaments branching, with cross-walls, numerous chlamydospores, and swellings. Characteristics of the group present. Color of colony isabella to cinnamon, strongly suggestive of chamois, due to the spores and asci which are in abundance. Czapek's Agar (pH 4.4).—Color of colony white, becoming chamois with age. Growth very sparse and cottony, with much 1933] MOORE—BLASTOMYCOSIS 99 of the mycelium submerged in the agar. Colony 4 cm. in diam- eter at end of 24 days. Hyphae long and thin, 112-2 y in diameter, with swellings approximately 4 x 12 y, several thick- walled chlamydospores 714 x 14 y, and numerous terminal hypnospores. Several 8-spored asci seen, as well as many conidia. Malt Extract Agar (pH 5.8).—Growth slow and cultural characteristics insufficiently different to be taken into discussion. Sabouraud's Agar (pH 5.6).— Growth rapid, profuse, obtaining a diameter of 714 cm. at end of 30 days. Culturally the colony simulates very much that of Microsporon audouini of Ota and Langeron in the presence of several radiating ridges from the round center, the inoculum, and the several concentric rings of growth of decreasing abundance, just outside the ridges. Color of colony white when young and becoming the characteristic chamois when older. Like M. audouini, it has racquet mycelium, chains of round cells on a hypha measuring 314 y in diameter, numerous chlamydospores 8 x 12 y, terminal hypnospores 5 x 11 u, and many conidia, characters found also in the Tricho- phytons and peculiar to Zndomyces capsulatus. Unlike M. audouini, however, this organism reproduces by the formation of asci which are numerous here, measuring from 10 to 13 y iu diameter, containing 8 spores. Sabouraud's Broth (Sabouraud's medium minus the agar. pH 5.6).—Culture consists of submerged mycelium of large flakes, each measuring approximately 2 cm. in diameter at end of 24 days. Mycelium floating on surface, dry, and chamois colored, with white region, presumably the young hyphal elements. In general, growth is good. Microscopically, the culture shows long, narrow hyphae 214 y in diameter, branching and inter- twining. Submerged mycelium shows almost no swellings, chlamydospores, terminal hypnospores, nor thick-walled cells, as compared with the great number found in that on the surface: The several that are present, however, show a great reduction in size and form from the exposed, the measurements of which are similar to those on agar. Potato-dextrose Agar (pH 5.6).—Growth profuse and cottony, covering the surface of the agar completely. Diameter of colony [Vor. 20 100 ANNALS OF THE MISSOURI BOTANICAL GARDEN 71% cm. after 24 days. Color cinnamon, with colony showing concentrie circles of color alternating with white, and the cinna- mon very pronounced, due perhaps to the medium constituents. Hyphae 3 y. in diameter, with numerous, thick-walled cells 744 u in diameter, budding cells, swellings 5 x 12 u, and chlamydospores varying in size from 4-7 x 9-14 u. Asci numerous, measuring approximately 13 y in diameter. Corn-meal Agar (product of Digestive Ferments Co. pH 6.0).— Growth poor, colony being 114 cm. in diameter at end of 24 days. Color white. Growth around inoculum loose and cottony. Hyphae short, thick-walled, 214 » in diameter, with numerous budding cells approximately 7 » in diameter. Chlamydospores numerous, 7 x 13 y, terminal hypnospores several, 5 x 12 u; a few asci seen, 11 y in diameter. Conidia abundant, 5 y in diameter. June-beetle Agar (medium consisting of a 4 per cent extract of June beetles, Lachnosterna fusca, plus 1.5 per cent agar, sterilized at 20 pounds pressure for 20 minutes, with a final pH 6.1).— Growth of loose, flat, cottony mycelium, forming concentric circles of decreasing abundance until a ring of fine filaments surrounds the culture. Colony 514 cm. in diameter at end of 24 days. Hyphae 2-21% y in diameter, with many conidia 5 y in diameter. Asci 12-13 y in diameter, thick-walled, enclosed in a sheath. Abundance of racquet mycelium. June-beetle Dextrose Agar (above medium plus 2 per cent dextrose). —Growth fair, attaining a diameter of 3 cm. at end of 30 days. Colony bright chamois in color, cerebriform, and cottony. Many conidia, 415-5 y in diameter. Hyphae 215-3 y in diameter and fairly short. Chlamydospores 8 x 16 y and numerous, as well as terminal hypnosp 5x12u. Asciround, 12-14 y in diameter. Lactose Agar (product of Digestive Ferments Co., lactose broth plus 1.5 per cent agar. pH 6.8).—Growth good, reaching a diam- eter of 6 cm. at end of 24 days. Colony chamois-color, profuse and cottony, with a region of very fine mycelium surrounding it. Hyphae 31⁄4 y. in diameter with numerous conidia 5 y. in diameter, budding off. Many thick-walled resting cells 7 » in diameter. Characteristic racquet mycelium, chlamydospores, terminal hypnospores, with properties similar to those on Sabouraud’s agar. 1933] MOORE—BLASTOMYCOSIS 101 Lactose Broth (product of Digestive Ferments Co. pH 6.8).— Growth good, large white flakes being formed in the solution which later become intertwined, forming a mat of mycelium 7 y in diameter. Hyphae slightly reduced, 3 u in diameter, 114 y in the younger filaments. Preponderance of budding cells 7 u in di- ameter, with thick-walled chlamydospores, asci, and terminal hyp- nospores, but reduced in size as compared with the growth on agar. Eosine-methylene-blue Agar (agar used as one of a routine, prod- uct of Digestive Ferments Co. pH 7.0).—Growth good, with a diameter of 514 cm. at end of 24 days. Culture compact, due to the hyphae having absorbed the stain from the substrate and turning the mycelium pink. Colony appears powdery with age. Hyphae characteristic, with swellings, 3 » in diameter. Many conidia, 5 x in diameter, and hypnospores with several chlamy- dospores. Glycerine Agar (nutrient agar as prepared by the Digestive Fer- ments Co. plus 6 per cent glycerine, Merck C. P. pH 7.1).—Growth fair, having a diameter of 5 cm. at end of 24 days. Culture shows a crinkled, moist region of budding yeast-like cells and a dry filamentous, cottony, chamois-colored region which has changed to the mycelial form characteristic on agar. Filamentous hy- phae 314 y in diameter, characteristic swellings being present which are slightly larger than those found on lactose agar. Racquet mycelium also present. Nutrient Agar (product of Digestive Ferments Co. pH 7.2).— Growth rapid, covering a region 7 cm. in diameter at end of 30 days. Colony filamentous, cottony, brown, with concentric rings of growth, the outermost being white. Hyphae 214 u in diam- eter. Growth similar to that on Sabouraud’s agar micro- scopically, with numerous conidia 5 y in diameter and asci 13 y in diameter. Nutrient Broth (pH 7.2).—Culture forms a mat of intertwining mycelium of long hyphae 2-214 y in diameter, with swellings, asci, and chlamydospores. Very few conidia. Terminal hyp- nospores several, but reduced in size, 4 x 9 y. Endo's Agar (product of Digestive Ferments Co. pH 7.5).— Growth fair, colony having a diameter of 314 cm. at end of 30 [Vor. 20 102 ANNALS OF THE MISSOURI BOTANICAL GARDEN days. Culture shows radiating ridges from center of inoculum, with growth becoming flat due to the stain in the medium which is absorbed by the hyphae, as in the case of the eosine-methylene- blue agar, giving the mycelium a pink color. Microscopically, the hyphae have a diameter of 24% y. Numerous conidia 5 y in diameter. Culture otherwise similar to that on eosine- methylene-blue agar. Gelatine (nutrient agar plus 1.5 per cent gelatine).—Slow lique- faction beginning after 30 days. Culturally, the fungus is very characteristic of the organism of blastomycosis in that it passes through the three typical stages: the moist, yeast-like stage with a flat growth; the prickly culture with the colonies simulating greatly small burrs (coremia); and the final, cottony growth present on agar after extended growth. DISCUSSION As stated in the introduction of this paper, it would seem that the syndrome-complex, commonly known as blastomycosis, has an innumerable list of etiological factors, each causing a condition so much like the other that clinicians have grouped them under one head. However, should one encounter any of these in a clinic one would find that therapeutic measures are so vastly different, varying with the organism, that a direct and accurate knowledge of the causative agent in each particular patient is absolutely essential. In the past, medical men, not particularly trained in mycologi- eal taxonomy, were inclined to class together all fungi presenting ascomycetous characters under one name, Blastomyces. So great is the confusion to-day that it is necessary to pick out these pathogenic fungi and classify each one separately. The organism isolated in the first case was termed Blastomyces dermatitidis by Gilchrist in 1894 because of its budding properties in the lesion. In a case of dermatitis reported by Gilchrist and Stokes (’96) the organism, which was evidently of the type termed Blastomyces, was called an Oidium. In a following paper (Gil- christ and Stokes, '98), it was made known that the organism described in the previous paper was called an Oidium because it did not ferment glucose, saccharose, or lactose, and although 1933] MOORE—BLASTOMYCOSIS 103 developing by gemmation or budding in the tissues, human and animal, developed mycelia with the formation of conidia upon artificial media. Ricketts (01) made an extensive study of the organism, distinguishing it from several of the yeasts but failing to consider several of the yeast-like fungi, and proposed definitely the name Oidium for the genus of the Gilchrist fungus. After this work, several terms were applied to the disease. Busse (’94) described his case a short time after Gilchrist reported his and he named the organism Saccharomyces hominis. Vuillemin in a later publication assigned the organism to the genus Cryptococcus and called it C. gilchristi. However, he failed to make a careful study of the organism culturally on artificial media, and a classification which places a great emphasis on the yeast-like appearance of the fungus in lesions is not exactly justifiable. Brumpt (’27) places the organism in the genus Mycoderma, calling it M. dermatitis. This terminology, however, is synony- mous with Oidium, and in that case is likewise useless. For a great number of years, no great work was done to establish definitely the position of Gilchrist’s organism, and the name Blastomyces as created by him still held sway. The term pre- sents a lot of difficulties. In the first place, the Blastomycetes, according to Buschke, are that group which develops through budding, provided à mycelium is formed on agar, while to the group of Blastomycetes, as Naegeli names budding, would belong the genera Endomyces, Saccharomyces, Cryptococcus, Monilia, and Oidium. Now the question arises as to what the actual meaning of the word blastomycete is. According to Vuillemin (’01), it does not designate a natural group, a botanical family based on genealogical affinities. There is in existence a genus Blastomyces, but these organisms are not budding fungi in the sense of Buschke. They are filamentous fungi whose spore-bearing elements, whether terminal, lateral, or intercalary, can be isolated by disarticulation, following Costantin and Rolland (quoted by Vuillemin). Frank (quoted by Vuillemin) established the Blastomycetes as an order to include such fungi as the beer yeasts whose elements are iso- lated by budding and not by disarticulation. In this respect, by virtue of the law of priority, it would seem that the name Blas- tomycetes, as designated by Frank, should remain. However, [Vor. 20 104 ANNALS OF THE MISSOURI BOTANICAL GARDEN if by general agreement the name of a genus could replace that of an order, then, according to the rules of nomenclature, the genus of Costantin and Rolland is legal, and the name as designated here is not legitimate by reason of the lack of distinct characters which have no generic value. Castellani recently proposed a new classification of yeast-like or budding fungi based on the presence or absence of ascospores, which includes families of both the Ascomycetes and Fungi Im- perfecti. 1. Saccharomycetaceae: budding cells, asci, and ascospores, but no mycelium in culture. 2. Endomycetaceae: budding cells, asci, and ascospores, with mycelium in culture. 3. Cryptococcaceae: budding cells (blastospores), no asci and no mycelium in culture. 4. Oosporaceae: budding cells, no asci, but mycelium in culture. In addition to this family classification, he created a new genus which he calls Blastomycoides, to which he assigns three species, and places it in the family Oosporaceae: 1. Blastomycoides der- matitidis, synonym Blastomyces dermatitidis Gilchrist and Stokes; 2. Blastomycoides immitis, synonym Coccidioides immitis Rixford and Gilchrist; 3. Blastomycoides tulanensis Castellani. He de- fines the genus Blastomycoides as: ‘‘Oosporaceae appearing in the lesions as large roundish cells from eight to twenty microns in diameter, or larger, with the protoplasm containing a number of well-marked granules or spherules, and with a membrane showing a well-defined double contour; in dextrose agar cultures a large amount of mycelium is present." He bases further differentiation of the three species on their cultural differences when grown on mannitol, lactose, glucose, and galactose agar. The second species that he names, Blastomycoides immitis, has already been discussed and classified by the author in a previous paper (M. Moore, ’32). The author has made no pretext of studying the third species, so that nothing can be said about that. The first species, however, Blastomycoides dermatitidis, is alto- gether misplaced, simply because there are asci present in the mycelium in culture. This of course would refer the genus to the family Endomycetaceae, in which group the writer definitely establishes the organism. 1933] MOORE—BLASTOMYCOSIS 105 Observations on the growth, development, reproduction, and further evolution of the fungus show that there are budding cells in the lesions, mycelium formed on agar with an intermediate stage showing the change from the yeast-like to the filamentous form. In accordance with this, Mellon (724, ’26, ’26a) has recorded the fact that asci do occur particularly ‘‘in the so-called secondary colonies of the cultures which also contained ‘dau- ernzellen’ and pigmented oidia." The author wishes to affirm Mellon’s findings as to the presence of asci, but suggests that these structures are present in the third stage, whereas the second step would consist of the intermediary forms which have an appearance very much like greatly enlarged oidia. Furthermore, it would seem that Mellon’s description was indefinite, inasmuch as he refers to an ascus as an ascospore, and oil droplets are suggested by him as being chromatin indicators and forerunners of the future spores. Such factors as these are very important in the taxonomy of this type of fungi and should not be dealt with so promiscu- ously. Furthermore, Mellon has not paid much attention to the fact that no matter how old the lesion may be the blastomycosis organism found there does not change from its yeast-like, budding growth until it has been transferred to artificial media, where the change is an adaptation to the mode of life it must lead; in other words, the change from active parasitism to one of saprophytism. It is to be understood, however, that a change such as suggested here will not necessarily reduce its viability, at least for the time being. A study of the evolution of the organism has repeatedly shown, in several hanging-drop cultures, that reproduction is hetero- gamous, as given in the description, with the final formation in the series of a large eight-spored ascus. Mellon in his papers consistently shows a four-spored ascus. It would seem, therefore, that he either has an organism unlike the one here described for the blastomycosis parasite, or else he has taken for granted as spores the four oil droplets which may and often have been found to occur on a mature eight-spored ascus, as was evidenced by the author on another ascomycete, Endomyces capsulatus var. isabellinus Moore, which was described in a case in another paper (McBride and Thompson, ’33) and also in Endomyces [Vor. 20 106 ANNALS OF THE MISSOURI BOTANICAL GARDEN capsulatus Rewbridge, Dodge and Ayers (29). Furthermore, it is quite possible that Mellon has observed the ascus just previous to the division of the nuclei, in the formation of the eight-spored ascus. "This latter statement is only a conjecture on the part of the writer, but in any case the cultural descriptions do not agree with those given by the early investigators and with which the organism here described does agree. In view of such criteria, it would seem that the organism formerly described as Blastomyces dermatitidis Gilchrist 1894, is not strictly a member of that genus, the name of which, on account of its etymological derivation, is essentially a misnomer. Because of its morphological characteristics, Blastomyces derma- titidis does not present those affinities entirely but simply as one phase of its life cycle. However, particularly because of its asco- mycetous attributes, it should belong to the class Ascomycetes, order Endomycetales, family Endomycetaceae, and because of its similarity in morphology and reproduction (perfect stage) to that of Endomyces capsulatus and its variety, it should belong to the genus Endomyces. If taxonomic position in this family and genus be dependent on the number of spores in the ascus, it would seem, according to Whitman ('13), that this organism should belong in the genus Oleina. However, the genus Endomy- ces contains a number of pathogenic species with eight-spored asci, whereas Oleina has no pathogenic species, and until a classifica- tion better than the one now in existence be established, dermati- tidis should be placed with Endommyces. From the above statement it would appear that the organism should now be known as: Endomyces dermatitidis (Gilchrist 1894), M. Moore, n. comb. Mycelium in lesions of budding yeast-like cells 7-12 y in diam- eter and sometimes as much as 20 y in length, occurring singly, in groups of two's, three's, or four's. Growth on agar of isodia- metric cells 2-214 y in diameter on acid media and 3-4 y in diam- eter on slightly alkaline media. Hyphae septate, with conidia pyriform or round, pedunculate or sessile, 5 v in diameter. Racquet mycelium present, 5-6 y in diameter at swollen portion and 3-31 y in diameter at narrow portion. Chlamydospores terminal or lateral or intercalary, 514—714 x 12-15 u, or sometimes 1933] MOORE—BLASTOMYCOSIS 107 round, 7 y. in diameter. Copulation heterogamous, asci spherical, 8-13 y in diameter, with 8 spherical to ovoid, smooth, hyaline to light chamois-colored spores 2-3 y in diameter, at maturity. Colony white in color, becoming cinnamon to brown with age. SUMMARY 1. The history of blastomycosis is given, with a review of the early work on yeast-like, fungous pathogenicity, and a report of the first case published. 2. The etiology and clinical manifestations represent a number of conditions due to several yeast or yeast-like organisms: Sac- charomyces, Oidium, Monilia, Endomyces, Cryptococcus, and Coccidioides, which have been placed in one category to constitute the agents responsible for the syndrome-complex, blastomycosis. 3. The disease is shown to simulate several conditions, in which cases the diagnosis must be arrived at through the isolation of the organism and the application of Koch’s postulates. 4. Immunological reactions and therapeutic measures are as yet indefinite as to specific results, although beneficial results have been reported by the use of iodides. 5. A case of blastomycosis of the arm and hand is reported. 6. There is a description of the organism, culturally and mor- phologically, showing its relationship to the class Ascomycetes. 7. The fungus is definitely established as Endomyces dermati- tidis of the family Endomycetaceae. ACKNOWLEDGMENTS The author wishes to express his sincere gratitude to the following: Dr. Carroll W. Dodge, Professor of Botany in the Henry Shaw School of Botany of Washington University, for his interest, criticisms, and helpful suggestions; Dr. George T. Moore, Director of the Missouri Botanical Garden, for the courtesies extended; Dr. Martin F. Engman, Dermatologist to the Wash- ington University Hospitals, for the use of the data with respect to the case reported; and Miss Nell C. Horner, librarian of the Missouri Botanical Garden, for her assistance. [Vor. 20 108 ANNALS OF THE MISSOURI BOTANICAL GARDEN BIBLIOGRAPHY Agostini, A. (31). On Blastomycoides lanuginosus Castellani. Jour. Trop. Med. & Hyg. 34: 287-288. 1931. Basgal, W. (’ “i Contribuição ao estudo das blastomycoses pulmonares. Doctor- ate thesis in medicine. Rio de Janeiro, 1931 Bassoe, P. (06). Report of a case of disseminated blastomycosis of the lungs, lumbar vertebrae and subcutaneous tissues. Chicago Path. Soc., Trans. 6: 380. Benedek, T. (28). Bemerkungen zum Zuchtungsverfahren des Schizosaccharo- myces hominis Benedek, 1927. I Mitteilung. Die Primarkultur. Derm. Wochenschr. 87: 1203-1214. 1928. and R. Frühwald (28). Clinical picture, mycology and serum diagnosis of Ni huie ose 2 cases. Ibid., 1566-1577. 1928. Blanchard, R., E. Swartz, et J. Binot (703). Sur une blastomycose intrapéritonéale. Arch. de Parasitol. 7: 489-507. 1903. Bigot, A., et H. Velu (25). Isolement rapide de Cryptococcus mirandei en culture pure. Soc. Path. Exot., Bull. 18: 127-129. 1925. — — ——, —————, (25a). Étude biologique de Cryptococcus mirandei agent de la bidatit des voies lacrymales de l'àne. Ibid., 231-235. 1925. —————, ——— ——, (325b). Contribution à l'étude des blastomycoses animales, Rev. Path. RA et Hyg. Gén. 25: 280, 281, 283. 1925. Borzone, R. A. (29). Un caso de blastomicosis en Santa Fé y ensaya de revisión de las blastomicosis americanas. Soc. Scient. Santa Fé, An. 1: 58-62. 3 Bowen, J., and S. B. Wolbach ('06). A case of blastomycosis; the results of culture and animal experiments. Jour. Med. Res. 10: 167-177. i Brown, P. K., and W. T. Cummins (15). I. A differential study of coccidioidal granuloma and blastomycosis. II. Report of two additional cases of cocci- dioidal disease. Arch. Int. Med. 15: 608-627. 1915. Brumpt, E. (27). Précis de gie. pp. 1213, 13883. Masson et Cie. Paris, 1927 Burkhead, C. E. (22). Oidiomycosis, including one case of coccidioidal granuloma and one of cutaneous blastomycosis. Kan. Med. Soc., Jour. 22: 101. 1922. Buschke, A. (98). Ueber Hautblastomykose. Deutsch. Derm. Gesell., Verhandl. 6: 181-222. 1898. , und A. Joseph (28). Blastomykose (Ascomykose). " T EMI: Hand- buch der Haut- und Geschlechtskrankheiten 11: 825-925. Busse, O. (’94). Ueber parasitaire Zelleinschlusse und ihre emi Centralbl. f. Bakt. Orig. 16: 175-180. 1894. —————, (95). Ueber Saccharomycosis hominis. Virch. Arch. 140: 23-46. 1895. Castellani, A. (25). Observations on some diseases of Central America. (Blas- tomycosis in man in Central America.) Jour. Trop. Med. & Hyg. 28: 1-14. 1925. , 25a). Notes on three new yeast-like organisms and a new bacillus, with remarks on the clinical conditions from which they have been isolated; furun- culosis SANG macroglossia blastomycetica, stomatitis eryptococco- bacillaris. Ibid., 1925. —, (28). Fungi did vmi, diseases. Am. Med. Assoc. Chicago, 1928. —— — ——, (28a). Notes on blastomycosis; its etiology and clinical varieties. Roy. Soc. Med., Sect. Trop. Dis. & Parasitol., Proc. 21: 447-462. 1928. 1933] MOORE—BLASTOMYCOSIS 109 Castellani, A., (28b). Blastomycosis and some other conditions due to yeast-like fungi (buddiuE fungi). Am. Jour. Trop. Med. 8: 379-422. 1928. Pi e Mannitol agar in the differentiation of the fungi of type Blastomy- c. Exp. Biol. & Med., Proc. 26: 544. 1929. Cile TE (28). Material zur experimentellen Blastomykose des Kaninchens. Derm. Wochenschr. 87: 1649. 1928. Chiari, H (30). Zur Pathologie und Histologie der generalisierten Torulose (Blastomykose). Arch. f. Derm. u. Syph. 162: 422—441. 1930. Chyurlia, N. (26). Notes on a case of bronchony ede. Jour. Trop. Med. & Hyg. 29: 145-146. 1920. Cleary, J. * (04). A case of generalized blastomycosis. Chicago Path. Soc., Trans. 6: 105-113. 1904 Cleland, J. d (27). A case of systemic blastomycosis with the formation of a yxomatous-looking tumor-like mass. Med. Jour. Australia 14: 337-340. Cole, W. H. (24). Systemic blastomycosis. Ann. Surg. 80: 124-134. 1924. Copelli, M. (13). A case of blastomycosis. Jour. Cut. Dis. 31: 51-52. 1913. Corselli, G., und B. Frisco (795). Pathogene Blastomyceten beim Menschen. Beiträge zur Aetiologie der bosartigen Geschwülste. Centralbl. f. Bakt. 18: 368-373. Coupal, J. F. (24). Diagnosis and treatment of d disease entities. Report of six cases of blastomycosis. Internat. Clin. 4: 1-14. 1924. Curtis, F. (96). Contribution à l'étude de n e humaine. Inst. Past., Ann. 10: 449—468. 1896. Davis, B. F. (11). The ER IL reactions of oidiomycosis (blastomycosis) in the guinea-pig. Jour. Inf. Dis. 8: 190. 1911. —— — ——, (22). Blastomycosis: lid pathology and therapeusis. Minn. Med. 5:311-315. 1922 Davis, C. N. (^06). n case of blastomycetic dermatitis. Jour. Cut. & Vener. Dis. Dennis, F. L. 18). Blastomycosis of the upper respiratory tract with a report of a case primary in the larynx. Ann. Otol., Rhin. & Laryng. 27: 571. 1918. Desjardins, A. U. (25). Roentgenotherapy and diathermy in blastomycosis. Am. Jour. Roentgenol. 14: 14-16. 1925. eee G. B., and R. R. Elworthy (’25). A case of blastomycetie dermatitis (Gilchrist). Roy. Soc. Med., Proc. 19: 4-10. 1925. Downing, E. D. (18). A case of blastomycosis with laryngeal involvement. Am. Med. Assoc., Jour. 70: 85-86. 8. D pea D. N. . and O. S. Ormsby ('05). A case of systemic blastomycosis in utum. Ibid. 45: 1045. 1905. Engslhardt, W. (24). Ein Beitrag zur Aetiologie oberflachlicher Hautblastomyko- sen und Hautsoormykosen. Arch. f. Derm. u. Syphil. 146: 313-322. 1924. Evans, N. (03). A clinical report of a case of blastomycosis of the skin from acci- dental inoculation. Am. Med. Assoc., Jour. 40: 1772-1775. 1903. (09). Coccidioidal granuloma ond blastomycosis in the central nervous DN Jour. Inf. Dis. 6: 523-526. Fabry, J. (25). Superficial erosive meis Derm. Wochenschr. 81: 1071- 1075. 1925 , (27). Über akneformige blastomycosis cutis. Ibid. 84: 824-827. 1927. [Vor. 20 110 ANNALS OF THE MISSOURI BOTANICAL GARDEN Ferguson, A. S. (28). Blastomycosis of eye and face secondary to lung infection. Brit. Med. Jour. 1: 442-443. 1928. da Fonseca, 0 . (22). Sobre as agentes das blastomycoses europeas. Cyclosexuadoe possicao systematico do levedo de Hudelo. Brasil-Med. 36: 101-102. 1922. ——, (28). Ensaya de revisión de las blastomicosis sudamericanos. Inst. Clin. Quirurg, " Hen 4: 469-502. 1928 ——— , . de Aréa Lefo (’28), Dermatite blastomycosique. Soc. Biol., Gok: ‘aL 08: 622-623. 1928. Fontaine, B. W., M. Haase, and v - Mitchell (09). Systemic blastomycosis. rch. Int. Med. 4: 101-117. Warne J. B. C. (13). Contribution : l'étude des exoascées pathogènes. Thèse de Bordeaux, 100 pp. 19 — = (00). On 2 pathogenic action of blastomycetes. Jour. Path. & Bact. 6: 37-63. 1900. org W., and F. D. Weidman (23). Cystic blastomycosis of cerebral gray matter caused by igi oe Stoddard and Cutler. Arch. Neurol. & Psychiat. 9: 589-603 Froilano de Mello, et A. ere (29). Sur un cas de blastomycose à placards multiples végétants verruqueux ou pustulo-uleérés. Soc. Path. Exot., Bull. 22: 142-147. ! Garr, C. C. (25). Systemic blastomycosis. Surg. Gyn. & Obs. 41: 490-492. 1925. Gáspár, I. (29). Blastomycotic meningo-encephalitis. Arch. Neurol. & Psychiat. 22:475-486. 1929. Gilchrist, T. C. (96). A case of blastomycetic dermatitis in man. Johns Hopkins Hosp., Repts. 1: 269-283. 1896. — ———, (02). Blastomycetie dermatitis in the negro. Brit. Med. Jour. 2: 1321- 1328. 1902. " W. R. Stokes ('96). The presence of an Oidium in the tissues of a case of pseudo-lupus vulgaris. Johns Hopkins Hosp., Bull. 7: 129-133. 1896. ————, ————, x^ A case of pseudo-lupus-vulgaris caused by a blastomyces. Jour. Exp. Med. N Y. 3: 53-78. 98. Graves, M. L. (22). Breieitie blastomycosis. Am. Jour. Trop. Med. 2: 123-132. Greenfield J. G. (24). Blastomyeosis of nervous system. Med. Sci. 10: 267-273. 192 Eee 8. (27). Ein Fall von tiefer puu Blastomykosis der Haut (Busse- chke). Derm. Wochenschr. 85: 1049-1055. 1927. , (28). Deep primary blastomycosis of the skin. Urol. & Cutan. Rev. 32: 453-457. 1928. d L. N. Maschkilleisson (26). Beitrage zur Lehre von der pathologi- schen Anatomie der Gilchristchen Hautblastomykose. Derm. Wochenschr. 82: 811-818. 1926. Haase, M., E. R. Hall, and C. H. Marshall Soi Local blastomycosis, report of a case. p . Med. Assoc., Jour. 79: 820-82 1922. Hagiwara, S. (22). Uber Dikara ien Jap. Zeitschr. Derm. Urol. 22: 941-980. 1922. Hamburger, W. W. ('07). A comparative study of four strains of organism isolated rom four cases of generalized blastomycosis. Jour. Inf. Dis. 4: 201-209. Hamilton, C. M. (26). Blastomycosis. South. Med. Jour. 19: 431-435. 1926. 1933] MOORE—BLASTOMYCOSIS 111 Harter, A. (09). De la blastomycose humaine. Thèse Fac. Méd. Nancy 8: 222. 1909. Hashimoto, T. (22). Uber Blastomycosis cutis. Jap. Zeitschr. Derm. Urol. 22: 1-34. 1922. Hedge, H. M. (28). The use of carbon dioxide snow in treating blastomycosis. Am. Med. Assoc., Jour. 90: 1367-1369. 1928. Hektoen, L. (99). The organism in a case of blastomycetic dermatitis. Jour. Exp. Med. 4: 261-278. 1899. , (07). Systemic blastomycosis and coccidioidal granuloma. Am. Med. ssoc., Jour. 49: 1071-1077. à Herrick, J. B. (07). Generalized blastomycosis. bid. 328. 1907. Hessler, R. (99). Blastomycetic dermatitis. bid. 32: 760. 1899. Hicks, J. A. B., and F. R. Chopping (24). Case of perionychia due to a blastomyces. Lancet 206!: 128. . Hill, H. P., and E. C. Dickson (14). Report of a case of systemic blastomycosis. Calif. State Jour. Med. 12: 120. 1914. Howes, W. B., and P. F. Morse (21). Report of two cases of blastomycosis. Boston Med. & Surg. Jour. 185: 315-317. 1921. Hudelo, Rubens-Duval, et Laederich (06). Etude d'un cas de blastomycose à foyers multiples. Soc. Méd. Hóp. Paris, Bull. et Mém. 23: 723-734. 1906. Hufschmitt, G., A. Sartory, R. Sartory, et J. Meyer ('31). Un cas de blastomycose cutanée à foyers multiples. Ann. Dermatol. 7: 850-876. 1931. Hurley, T. D. (16). Heart lesion in blastomycosis. Jour. Med. Res. 33: 499-502. 1916. Hyde, J. N., L. Hektoen, and A. D. Bevan (’99). A contribution to the study of blastomycetic dermatitis. Brit. Jour. Derm. 11: 261-276. 1899. Irons, E. E., and E. A. Graham (’06). Report of a case with miliary and ulcerative blastomycosis of the lungs. Miliary blastomycosis of the spleen and multiple superficial and deep abscesses. Jour. Inf. Dis. 3: 666-682. 1906. Mig C. (26). Blastomycosis of the larynx. Arch. Otolaryng. 3: 99-107. P ed R. H. (26). Surgical treatment of certain massive blastomycetic skin lesions. Am. Jour. Surg. 1: 185-187. 26. Jacobson, H. P., J. F. Schamberg, and H. Morrow (’32). Fungous diseases. clinico-mycological text. pp. 149-181. Charles C. Thomas Co., Eu Illinois. 1932. Jeaume, G., et M. Dekester (25). Isolement de l'agent pathogène de la blastomy- cose des voies lacrymales. Soc. Path. Exot., Bull. 18: 124—127. 1925. Jona, G. (97). Die Schutzmittel des Crpeuitinis gegen Blastomyceten. Centralbl. f. Bakt. 21: 147-150. 1897. LeCount, E. R., and J. Myers (05). Systemic blastomycosis. Final report of the case described by Eisendrath and Ormsby in 1900. Jour. Inf. Dis. 4: 187-200. 1905 Legendre, J. (727). A propos de la dermatite blastomycosique chéloidienne. Soc. Path. Exot., Bull. 20: 323. 1927. Lewis, D. 17). Blastomycosis and sporotrichosis. Surg. Clinics, Chicago 1: 1125. 1917. MacLeod, J. M. H. (330). Some skin affections due to yeast-like fungi. Brit. Med. Jour. 1930: 1119-1123. 1930. [Vor. 20 112 ANNALS OF THE MISSOURI BOTANICAL GARDEN Maffuci, A., und L. Sirleo ('98). Ueber die vpn ala Infektionserreger bei bósartigen Tumoren. Zeitschr. f. Hyg. 27: 1-30. 1898 Maner, G. D., and R. W. Hammack (30. Systemic blastemisooels. Calif. & West. Med. 32: 87-90. 1930. Massey, A. Y. (16). Blastomycosis (?) in Central Africa. Jour. Trop. Med. & H 916. Mazza, S., y F. Niño (28). Notas sobre blastomicosis de las vias respiratorias. Reunión Soc. Argentina Patol. Reg. Norte en Santiago del Estero 4: 545—548. 1928. Quintana, y V. Bernasconi (30). Blastomicosis grave SAA por m n. sp. Reunión Soc. Argentina Patol. Reg. Norte en Salta 6 ig 214. — —, . Canevari (29). Uleeras blastomicosicas de la lengua. Reunión Soc. j diio, Patol. Reg. Norte 5: 226-230. 1929. — — ——, F. L. Niño, y P. Nicolini (29). Blastomicosis de la mucosa labiogeniana. Ibid. 231-239. 1929. ———— —, — — ——, y A. Egües (29). Perionixis blastomicética por Monilia (n. sp.). Ibid. 284.288. 1929. , L. Stabile de Nucci, y E. J. Canal Feijóo (Santiago del Estero) ('29). Blastomicosis cutánea de forma lenta por criptococa (n. sp.). Ibid. 293-308. 1929. y B. Palamedi (32). Caso mortal de blastomicosis cutáneo mucosa. "em d Soc. Argentina Patol. Reg. Norte en Tucumán 7: 424—467. 1932. McKee, S. H. (26). Blastomycosis of the cornea, with review of reported cases of blastomycosis of the eye. Internat. Clin. 3: 50-57 6. Meckel, M. (27). Weitere Mitteilungen über erosive Blastomykosen. Derm, Wochenschr. 84: 817-824. 1927. "— E. M. (27). Pulmonary blastomycosis; its similarity to tuberculosis. m. Jour. Path. 3: 305-314. 1927. us R. R. (24). Observations on an re stage for the parasites of blas- tomycosis eh Exp. Biol. & Med., c. 22: 69. 1924. , (26). Studies in microbic indi. "VL The infective and taxonomic significance of a newly described ascospore stage for the fungi of blastomycosis. Jour. Bact. 11: 229-252. 1926. , (26a). Studies in mierobie heredity. VII. Observations on the genetic origin of the several types of fungi found in the lesions of blastomycosis hominis. Ibid. 419-482. 1926. Michelson, I. D. (28). reser ge P i and bacteriologic study. Am. Med. Assoc., Jour. 91: 1871-1876. 192 Miller, $ E. (’ 28). Yeast-cell formation in man. U.S. Navy Med. Bull. 23: 229- 192 Miller, W. S. m The reticulum of the lung: Its similarity in blastomycosis to that in tuberculosis. Am. Jour. Path. 3: 315-320. 1927. Montel, R., and R. Pons (26). Dermatite blastomycosique chéloidienne. Soc. Path. Exot., Bull. 19: 876-880. 1926. Montgomery, F. H. (03). A case of obese blastomycosis followed by laryngeal and eee "qiia Death; autopsy. Jour. Cut. Dis. 21: 19-22. 1903. y (08). "iiA blastomycosis: Its etiologie, pathologie and clinical "ie aine as y established by a critical survey and summary of twenty- 1933] MOORE—BLASTOMYCOSIS 113 two cases, seven previously unpublished. The relation of blastomycosis to coccidioidal granuloma. Arch. Int. Med. 2: 1-41. 1908. Montpellier, J., et A. Catanei (’26). Blastomycose de l'avant-bras chez une femme indigéne d'Alger. Soc. Path. Exot., Bull. 19: 586-592. 1926. Moore, J. T. (20). Blastomycosis. Report of a case dying from abscess of brain. Surg., Gyn. & Obs. 31: 590-594. 1920. Moore, M. (732). Coccidioidal granuloma: A classification of the causative agent, Coccidioides immitis. Mo. Bot. Gard., Ann. 19: may 1932. [—_——_——], MacBryde, C. M., and E. J. Thompso (33). Meningitis and dermatitis caused by a new variety of ae Ne: (endomycete). Arch. of Derm. & Syphil. 27: 49-69. 1933. Morris, R. T. (13). A case of systemic blastomycosis. Am. Med. Assoc., Jour. 61: 2043-2044. 1913 Nesezadimenko, A. (99). Zur Pathogenese der Blastomyceten. Centralbl. f. Bakt. 25: 55-58. 1899. Neumayer, J. (’91). Untersuchungen über die Wirkungen der verschiedenen Hefearten, welche bei der Bereitung weingustiger Getrünke vorkommen auf den thierischen und menschlichen Organismus. Arch. f. Hyg. 12: 1-60. 1891. New, G. B. (17). Blastomyoeosis of the tongue. Am. Med. P» iy Jour. 68: 186. 1917. , (28). Blastomycosis of the larynx. Ann. Otol, Rhin. & Laryng. 37: 240-250. 1928. Nieberle, N. (27). Blastomycosis of skin in pig. Virch. Arch. f. path. Anat. 263: 1927 Niño, F. L. (’29). Uleeración blastomicética cutáneomucosa del labio inferior (Consideraciones acerca de su diagnóstica etiológico). Reunión Soc. Argentina Patol. tick Norte 5: 213-225. 1929. —— ———, (29a). Onixis y perionixis de origen blastomicósico (Estudio clinico y fidológitoi: Ibid. 270-282. 1929. , (30). Blastomicosis humano generalizada por Criptococo (n. sp.). Reunión Soc. Patol. Reg. Norte en Salta 6: 117-167. 1930 ——————, y J. Fernandez (7290). Nueva observacion de perionixis per Monilia iie Reunión Soc. Patol. Reg. Norte 5: 282-283. 1929. —, Palant (30). Nuevas observaciones de onixis y peri- onixis s de origen SaL icoaiec. Reunión Soc. Argentina Patol. Reg. Norte en Salta 6: 35-99. 1930. Ormsby, 2 S. (21). Blastomycosis. A practical treatise on diseases of the skin. Lea & Febiger, Philadelphia & New York. 1921. and H. M. Miller (03). Report of a case of systemic blastomycosis with multiple cutaneous and subcutaneous lesions. Jour. Cut. Dis. 21: 121-130. 03 1903. Ota, M. (24). Essai de classification des blastomycétes pathogènes. Ann. Para- sitol. 2: 34-61. 1924. Otis, F. J., and N. Evans ('03). Morphology and biology of the parasite from a case of systemic blastomycosis. Am. Med. Assoc., Jour. 41: 1075-1082. 1903. Panja, G. ((25). A case of generalized blastomycosis. Ind. Med. Gaz. 60: 475-476. 1925. Parker, C. A. (23). Actinomycosis and blastomycosis of the spine. Jour. Bone & Joint Surg. 5: 759-777. 1923. [Vor. 20 114 ANNALS OF THE MISSOURI BOTANICAL GARDEN Parmenter, F. J., and B. T. Simpson (19). A case of blastomycosis involving the prostate and Lame vesicles. Jour. Urol. 3: 449. 1919. Rabinowitsch, L. ('96). erra n evi pathogene Hefearten. Zeitschr. f. Hyg. u. Infektionskrank. 21: Mii 4 (91). Zur mac a en T e der Sprosspilze. bid. 10: 1-50. Sad T A. (26). Systemic blastomycosis. Neb. Med. Jour. 11: 257-260. 1926. Rewbridge, A. G., C. W. Dodge, and T. T. Ayers (29). A case of meningitis due to Sindontront capsulatus (new species). Am. Jour. Path. 5: 349-364. 1929. "—— B. W. (26). Blastomycosis of the bladder. Am. Med. Assoc., Jour. 87: 06. 1920. s W. (’ a Beiträge zur Hefepilzerkrankung. Derm. Wochenschr. 87: 931-940. Ricketts, H. : (7 Oidiomycosis (blastomycosis) of the skin and its fungi. Jour. Med. Res. 6: 374—547. 1 , (Ol A new mould fungus as the cause of three cases of blastomycosis idiren of the skin. Boston Soc. Med. Sci., Jour. 5: 453-459. 1901. Roncali, D. B. (95). Die Blastomyceten s den P E Centralbl. f. Bakt. 18: 432-434. 1895. Ryerson, E. W. ('08-'09). Blastomycosis: Report of two cases resembling bone tubereulosis. Am. Jour. Orthoped. Surg. 6: 79-83. 1908-1909. — E. 8., and D. C. Smith (27). The effect of gentian-violet on the organ- blasioaireaia infection. Arch. Derm. & Syph. 16: 153-155. 1927. Sanfelice, F. (95). Ueber einen neuen pathogenen Blastomyceten, welcher inner- halb der Gewebe unter Bildung kalkartig aussehender Massen degeneriert. Centralbl. f. Bakt. 18: 521-526. 95. , (96). Ueber die Loggen nd Wirkung der Blastomyceten. I. Abhand- lung. Zeitschr. f. Hyg. 21: 32-58. 1896. —— — ——, (96a). Ibid. II. Aor eta Ibid. 390—420. 1896. epum d C. R. (29). Two cases of blastomycosis cutis. Acta Dermato- Venereol. 10: 83-94. 1929. Simoni, A. de (97). Ueber das Vorkommen von Blastomyceten in der Hyper- trophischen Tonsille. Centralbl. f. Bakt. 22: 120-122. 1897. Smith, D. C., H. C. Turner, and E. S. Sanderson (28). Systemic blastomycosis with a report of a fatal case. Brit. Jour. Derm. 40: 344—359. 1928. Speroni, D., J. Llambias, S. E. Parodi, y. F. L. Niño ((29). Blastomicosis humano susetelisado por criptococo (n. sp.) Estudio parasitológico, anátomopato- gica, clinico y experimental. Reunión Soc. Argentina Patol. Reg. Norte 5: 9. Spring, D. (29). Comparison of seven strains of organisms causing blastomycosis in man. Jour. Inf. Dis. 44: 169-185. 1929. Stearn, E. W., and A. E. Stearn (29). Comparative inhibiting effect of gentian violet hod mercurochrome on the growth of certain fungi. Jour. Lab. & Clin. Med. 14: 1057-1060. 1929. Stober, A. M. (14). Systemic blastomycosis. Arch. Int. Med. 13: 509-556. 1914. Stovall, W. D., and H. P. Greeley (28). Bronchomycosis. Report of eighteen cases of primary infection in the lung. Am. Med. Assoc., Jour. 91: 1346-1351. 1928. Sugden, F. (23). Case of blastomycosis. Brit. Med. Jour. 2: 63. 1923. 1933] MOORE—BLASTOMYCOSIS 115 Sutejew, G., M. Utenkow, and A. Zeitlin (’29). Beitrag zur Ätiologie, Röntgen- diagnose und Róntgentherapie der Blastomykose. Fortschr. Geb. Rontgenstr. 11: 475-483. 1929. T., F. E. (28). Cutaneous moniliases. Trop. Med. & Hyg. Jour. 31: 37-38. 1928. Toepel, T. (29). Systemic blastomycosis. Am. Med. Assoc., Jour. 93: 32. 1929. Tokishige, H. (96). Ueber pathogene Blastomyceten. Central, f. Bakt. 19: Troisier, E., et P. Achalme (’93). Sur une angine parasitaire causée par une levure et cliniquement semblable au muguet. Arch. Méd. Expér. 5: 29-37. 1893. Urbach, E., und F. Zach (’30). Generalisierte Torulose (Europaeische Blastomy- kose). Eine klinisch-botanisch Studie. Arch. f. Derm. u. Syphil. 162: 401-421. 1930. Vuillemin, P. (01). Les blastomycétes pathogènes. Rev. Gén. des Sci. 12: 732- , (10). Matériaux pour une classification rationelle des Fungi Imperfecti. Compt. Rend. Acad. Paris 150: 882. 1910. Wade, H. W. (16). A variation of gemmation of Blastomyces dermatitidis in the tissue lesion. Jour. Inf. Dis. 18: 618-629. 1916. ae ig | Portal of entry in experimental chronic pulmonary (systemic) mec Philipp. Jour. Sci. 13: 271. 1 —— ————, and G. S. Bell (16). A critical coniare of systemic blastomycosis. Arch. Int. Med. 18: 103. Walker, J. W., and F. H. H aiomay (02). Further report of a previously re ported case of blastomycosis of the skin: Systemic infection with Msoi death; autopsy. Am. Med. Assoc., Jour. 38: 867-871. 1902. Wanamaker, T. (28). A case of TEOS of the cervical lymph gland. Am. Laryng., Rhin. & Otol. Soc., Trans. 34: 450—452. 1928. Weidman, F. D., and H. R. Douglas (21). Blastomycetoid bodies in a sarcoma- like tumor of the leg. Arch. of Derm. & Syphil. 3: 743-752. 21. Wernicke, R. (’92). Uber einen Protozoenbefund bei Mycosis fungoides (?). Centralbl. f. Bakt. 12: 859-861. 92. Whitman, R. C. (13). A contribution to the botany of the organism of blastomy- cosis. Jour. Inf. Dis. 13: 85-94. 1913. Wilhelmj, "i M. (25). The primary meningeal form of systemic blastomycosis. Am r. Med. Sci. 169: 712-721. 1925. Wohl, M. G. (23). Fungous dinbi of man in the State of Nebraska; sporotri- chosis; blastomycosis; actinomycosis. Am. Med. Assoc., Jour. 81: 647-653. 23. Yakimoff, W. L., and W. J. Wassilewsky (25). Au sujet de la blastomycose. Soc. Path. Exot., Bull. 18: 130-132. 1925. Zoon, J. J. (30). Blastomycosis cutis durch Monilia floccoi mit positiver Blutkultur. Derm. Wochenschr. 58: 356-367. 1930. [Vor. 20, 1933] 116 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 6 All drawings made with camera lucida at a magnification of 800. Figs. 1-9. Yeast-like cells Figs. 1, 6, 7. On Raulin’s solution. Figs. 2-5, 9. On Sabouraud's agar. Fig. 8. On glycerine agar Figs. 10-12, 14-17. Yeast-like cells showing a change to mycelial formation. Figs. 10, 12. On potato-dextrose agar. Figs. 11, 14. On glycerine agar. Figs. 15-17. On Sabouraud’s agar. Fig. 13. Germinating spores on Richards’ solution agar. Fig. 18. Heterogamous copulation of lateral cells on Sabouraud's agar. Fig. 19. Heterogamous copulation of terminal cells on Richards’ solution agar. Fig. 20. Copulating branch on corn-meal agar Fig. 21. Maturing ascus on Richards’ solution agar. Fig. 22. Terminal hypnospore on Sabouraud’s agar. Fig. 23. Racquet mycelium on June-beetle agar. Fig. 24. Mycelium showing conidia on Richards’ solution agar. Fig. 25. AN showing round terminal chlamydospores and swollen hypha on nutrient agar Figs. 26, - Myoslium showing conidia, oidia-like cells, and resting cells on corn-meal ag Fig. 27. Chesed on Czapek’s a Fig. 29. Mycelium showing conidia on oue agar. Fig. 30. Racquet formation on Czapek's agar Fig. 31. Terminal chlamydospore on — ager. Fig. 32. Terminal hypnospore on Endo’ Fig. 33. Maturing lateral ascus on cedi agar Fig. 34. Racquet mycelium on June-beetle dextrose agar. Fig. 35. Ascus — with a third sheath in proximity to a round resting cell, on potato-dextrose a Lateral na on — agar. Fig. 37. Resting cell on Sabouraud's a Fig. 38. Mature ascus on sor oca agar. Ann. Mo. Bor. Ganp., Vor. 20, 1933 PLATE 6 MOORE—BLASTOMYCOSIS [Vor. 20, 1933] 118 ANNALS OF THE MISSOURI BOTANICAL GARDEN EXPLANATION OF PLATE PLATE 7 Fig. 1. Photograph of hand of patient on day of entry, April 8, 1932, showing lesion involving portion of thum ig. 2. Photograph taken on Apel 8, 1932, showing abscess on flexor surface of left a arm. g. 3. Photograph showing marked improvement after treatment with sodium iodide intravenously. Photograph showing almost complete healing. ANN. Mo. Bor. Ganp., Vor. 20, 1933 9017 50099 PLATE 7 ee F3312? D50095 MOORE—BLASTOM Y COSIS A MONOGRAPH OF THE AMERICAN SPECIES OF THE GENUS HALENIA! CAROLINE K. ALLEN Formerly Missouri Botanical Garden Special Fellow in Botany, Henry Shaw School of Botany of Washington University INTRODUCTION Frequent attempts to determine recent collections of Halenia, especially from Central and South America, have revealed the need for a comprehensive taxonomic treatment of the American species of the genus. Incidental determination of isolated species in herbaria, which furnish scanty material, has led to error and the duplication of species has resulted. Few of these fragmentary treatments are provided with adequate descriptions, and still fewer are accompanied by illustrations. The present paper is a monographie study of the American species of Halenia. The first portion is devoted to the North and Central American representatives of the genus and the second to those of South America. The writer at this time wishes to express her appreciation to Dr. George T. Moore, Director of the Missouri Botanical Garden, where this work has been carried on, for the privilege of using the facilities afforded by the herbarium, library, and laboratory. It is with pleasure, indeed, that she also acknowledges the courte- ous assistance and helpful criticism offered by Dr. J. M. Green- man, under whose personal supervision the study was begun and completed. For the loan of specimens for study the writer is especially indebted to the curators of the herbaria of the following institu- tions: Gray Herbarium of Harvard University, New York Botanical Garden, Philadelphia Academy of Natural Sciences, United States National Herbarium, Field Museum of Natural History, the Dudley Herbarium of Stanford University, Brooklyn Botanic Garden, Iowa Agricultural College, Natural History Mu- 1 An investigation carried out at the Missouri Botanical Garden in the Graduate Laboratory of the Henry Shaw School of Botany of Washington University, and submitted as a thesis in partial fulfillment of the requirements for the degree of doctor of philosophy in the Henry Shaw School of Botany of Washington University. Issued April 29, 1933. Ann. Mo. Bor. Ganp., Vor. 20, 1933. (119) [Vor. 20 120 ANNALS OF THE MISSOURI BOTANICAL GARDEN seum of Vienna, Botanical Museum of Stockholm, the Delessert Herbarium at Geneva, Royal Botanie Gardens, Kew, and the British Museum of Natural History. Particularly does the writer wish to express her gratitude to Mr. R. I. Cratty, of the Iowa State College, for the loan of specimens from the Parry Herbarium deposited there; to the curator of the Herbarium of the Botanical Garden of Madrid, for supplying a photograph of the type of Swertia cucullata; and to the curator of the Herbarium of the Jardin des Plantes, Paris, for photographs of the types of Humboldt, Bonpland and Kunth. In connection with visits to various herbaria, the author desires to acknowledge the courtesy and kindly assistance of Dr. Ernst Gilg and Dr. Robert Pilger, of the Botanical Garden at Berlin-Dahlem; Dr. Walter Robyns of the Botanical Garden, Brussels; and Mr. Spencer Savage, in charge of the Linnaean Herbarium, of the Linnaean Society of London. "Thanks are due also to Dr. George E. Nichols, Director of the Marsh Botanical Garden, Yale University, for his kindness in procuring seeds of Halenia; to Miss Nell C. Horner, Librarian of the Missouri Botanical Garden, and Dr. John H. Barnhart, of the New York Botanical Garden, for assistance in bibliography; to Dr. Roland V. La Garde, of the Missouri Botanical Garden, for preparation of photographs; and to the George F. Cram Com- pany, for permission to use the copyright outline maps of North and South America. HISTORY OF THE GENUS Linnaeus in the ‘Amoenitates Academicae," which appeared in 1751, published short descriptions of two genera of the Genti- anaceae, namely, Swertia, consisting of five species, and Gentiana, of twenty-three. Under Swertia Linnaeus listed the spurred gentian with the following description. “4. Swertia corollis quadrifidis quadricornibus. Amoen. acad. 2. p. 344. Habitat in Sibiria, Gmelin; Canada, Kalm.” This was apparently the only species of spurred gentian he had ! Linnaeus, Amoen. Acad. 2: 344. 1751. 1933] ALLEN—THE GENUS HALENIA 121 ever seen, and, it being closely related to the Swertia he knew, he placed it in that group, giving it the specific name corniculata or "^ horn-tipped," as distinct from the others. Gmelin, in his ‘Flora Sibirica,? published in 1769, referred to this as being synonymous with his genus T'etragonanthus, which he had described or mentioned in a previous book or manuscript, and which he evidently based upon specimens collected by G. H. Stellar. 'This work contains a brief description, but a good illustration of Swertia. On account of the presence of spurs on the corolla, Bórck- hausen,’ in 1796, segregated Swertia corniculata from the Linnaean genus Swertia and called it Halenia after Jonas Halen. Although the latter had included a short description of it in a previously published dissertation on Kamtchatka plants, Bórckhausen must be considered the author of the genus. He cited as a synonym Swertia corniculata Linnaeus, but changed the binomial name to Halenia sibirica. The name Swertia, however, persisted for some time in litera- ture, the generic descriptions becoming more elaborate and detailed with each publication. Ruiz and Pavon described and illustrated Swertia umbellata from Peru‘ in 1802. Michaux included the genus in his ‘Flora Boreali-Americana'5 appearing the following year, and Humboldt, Bonpland and Kunth,’ described six new species from Mexico and South America in 1818. A few species, together with a new variety, were published in ‘Linnaea’ by Schlectendal and Chamisso? in 1830. But it was not until Grisebach began his extensive study of the Gentian- aceae that any attempt was made to bring together the species of the world. As a result, when his ‘Observationes’! appeared in 1836, followed in 1839 by the 'Genera et Species Gentian- earum,'? many of the existing names fell to synonymy. Grisebach divided Halenia into two main divisions: the first, 2 Gmelin, Fl. Sib. 4: 114. 1769. 3 Bórckhausen in Roemer, Arkiv für Botanik 11: 25. 1796. * Ruiz & Pavon, Fl. Peruv. 3:21. pl. 242. 1802. 5 Michaux, Fl. Bor.-Am. 1: 97. 1803. $ Humboldt, Bonpland & Kunth, Nov. Gen. & Sp. Pl. 3: 174. 1818. 1 Schlectendal & Chamisso in Linnaea 5: 122. $ Grisebach, Obs. Gent. 36. ? Grisebach, Gen. & Sp. Gent. 322-398, 1839. [Vor. 20 ev. ANNALS OF THE MISSOURI BOTANICAL GARDEN with spurs ascending and spreading; the second, with spurs pendulous to ineurved. He devoted careful attention to the descriptions, synonymy, affinities, differentiating characters, and habitat of each species mentioned. The collector and type were given in each case. He also separated from the Swertia of Humboldt, Bonpland and Kunth two species, brevicornis and parviflora, and founded on them a new genus Hxadenus distin- guished by the presence of pits at the base of the corolla, instead of spurs, and by central placentation. This abolished the Swertia of Humboldt, Bonpland, and Kunth, which was based on Swertia corniculata Linnaeus, and left the non-spurred Swertia originally described by Linnaeus a genus entirely distinct from our present Halenia. Hooker's *Flora,'!? published in 1840, contained good illustra- tions of Halenia deflexa with the varieties Brentoniana and heterantha. Nearly a decade elapsed before any significant study was done on the genus as a whole. Bentham," in 1839-1840, in describing Hartweg’s plants from Mexico added two new species, multiflora and decumbens. Martens and Galeotti in 1844 in their *'Enu- meration of Mexican plants collected by Galeotti"!? described, along with other new Halenia species, two new species of Exa- denus. Walpers’ ' Repertorium '? contains reference to Exadenus, but Weddell in 18591 merged the two genera. He considered the group as a whole to consist of two main subdivisions, the first being that group with spurs, and the second the spurless species. ‘The former he subdivided into section l—ovary uni- locular; section 2—ovary bilocular, and the corolla having small spurs. Here he placed the two original species of Hzxadenus Grisebach which were based on the species of Swertia Humboldt, Bonpland & Kunth, namely, parviflora and brevicornis, but neglected to transfer the species described by Martens and Galeotti. The second subdivision, without spurs, he also divided into two sections, on the presence of a uni- or bilocular ovary. 10 Hooker, Fl. Bor.-Am. 2: 67. pl. 155-6. 1840. 11 Bentham, Pl. Hartw. 24. 1839; 67. 1840. ? Martens & Galeotti in Bull. Acad. Brux. 11!: 370. 1844. 18 Walpers, Rep. Bot. Syst. 6: 508. 1846-47. ^ Weddell, Chlor. And. 2: 74. 1859. 1933] ALLEN—THE GENUS HALENIA 123 His argument for combining the two genera was based on the fact that the two characters, the bilocular ovary and the short spurs do not always coincide. In this work, he described five new South American species based on collections of Triana, Funck & Schlim, Goudot, Purdie, etc. At the time of the publication of Bentham and Hooker's ‘Genera Plantarum,’ 1876, there were about twenty-five rec- ognized species of Halenia from the whole world. Hemsley, in ‘Biologia Centrali-Americana,'5* which appeared in 1882, referred the Exadenus Martens and Galeotti to Halenia. Gilg in Engler and Prantl’s ‘Natiirlichen Pflanzenfamilien '? recognized the validity of the generic name Halenia and included a careful description and good illustrations. Sessé & Mocifio's 'Flora* contains a description of Swertia cucullata which has escaped the notice of succeeding monog- raphers, and has not been transferred to Halenia. A photo- graph of the type specimen kindly furnished by Dr. E. Balguerias, Curator of the Botanical Garden at Madrid, reveals the fact that it is without question Halenia brevicornis Griseb., though it is rather difficult to determine whether or not it is the species proper or one of its many forms. From this time until the monographic treatment of South American species was published by Gilg!? in 1916, several species and varieties of Mexican and Central American Halenia were published. Halenia Rothrockit Gray,?° now Halenia recurva, in 1876 was based on plants collected in Arizona by Dr. J. T. Rothrock. G. Don, in 1838, in the ‘General History of the Dichlamydeous Plants’?! included a description of the existing species of Halenia, making several new combinations from the species of Humboldt, Bonpland and Kunth.? Some of his species were taken from the manuscript of D. Don, thus appearing in % Bentham & Hooker, Gen. Pl. 2: 817. 1876. 16 Hemsley, Biol. Cent.-Am. Bot. 2: 351. 1882. 17 Gilg in Engler & Prantl, Nat. Pflanzenfam. 42: 89. 1895. 18 Sessé & Mocifio, Flora Mexicana, 73. 1894. 19 Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, pp. 93-122. 1916. 20 Gray in Proc. Am. Acad. 11: 84. 1876. 21 Don, G. Gen. Hist. 4: 177. 1838. 2 Humboldt, Bonpland & Kunth, Nov. Gen. & Sp. Pl. 3: 174. 1818. [Vor. 20 124 ANNALS OF THE MISSOURI BOTANICAL GARDEN publication for the first time. Kuntze,” in 1891, revived the generic name T'eiragonanthus Gmelin. He has been followed by a few later taxonomists, among whom are Britton, Small,” Rydberg,” and others. However, Halenia was placed on the list of nomina conservanda, and thus must be regarded as the correct name in accordance with the International Rules of Botanical Nomenclature. Halenia crassiuscula Robinson and Seaton?’ ap- peared in 1893. Britton,?* in 1894, reduced H. Brentoniana Grisebach to a variety of Tetragonanthus deflexus, while in 1899 Fernald?* placed H. heterantha Grisebach under Halenia deflexa var. heterantha. Other new species and new combinations ap- peared in rapid succession, due to the extensive collecting done during the period from 1890 to about 1920, by Conzatti, Pringle, Purpus, and others in Mexico, and Weberbauer in South America. Among these newly published species were: Halenia candida Ramirez,*° 1895; H. chlorantha Greenman,* 1905; H. bella, H. caespitosa Gilg, 1906; H. Conzattii Greenman,* 1912; H. guate- malensis Loesener, and H. plantaginea var. latifolia Loesener* (now H. guatemalensis var. latifolia (Loesener) Allen) 1913. The type species up to this time had been called H. sibirica, the name given it by Börckhausen. However, according to the rules of priority, sibirica should give way to the older name corniculata used by Linnaeus. Accordingly, Druce% in 1914 revived corniculata which is at present the accepted name for the type species of Halenia. Britton and Brown in the 'Illus- trated Flora, reduced Grisebach’s species Brentoniana and heterantha to varieties of deflexa. * Kuntze, Rev. Gen. Pl. 2: 431. 1891. ^ Britton, Manual, 734. 1901. *5 Small, Fl. Southeastern U. S. 931. 1913. * Rydberg, Fl. Rocky Mts. 666. 1922. *7 Robinson & Seaton in Proc. Am. Acad. 28: 113. 1893. *8 Britton in Mem. Torrey Bot. Club 5: 261. 1894. ?? Fernald in Rhodora 1: 37. 3 Ramirez in Inform. Secret. Foment. Mexic. "ed Mont. Ajusco). 34. 1895. 3l Greenman in Proc. Am. Acad. 41: 240. 1 2 Gilg in Fedde, Rep. Spec. Nov. 2: 52. D. % Greenman in Publ. Field Mus. Bot. 2: 335. 1912. ^ Loesener in Verh. Bot. Ver. Brandenb. 55: 182. 1913. % Druce in Rept. Bot. Exch. Cl. Brit. Isles 3: 419. 1914. * Britton & Brown, Ill. Fl. 3:15. 1913. 1933] ALLEN—THE GENUS HALENIA 125 Gilg,?" in 1916, monographed the South American representa- tives of the genus, adding several new species. At that time, the herbarium at Berlin contained as complete a series of South American plants as were to be found any where, but, in some cases at least, the material was too inadequate to determine the limits of variation of certain of the species proposed. Hence, some of the species recognized, studied in the light of subsequent collec- tions, have fallen into synonymy. Gilg created three main divisions: A. Nectaria parva vel obsoleta, rarius extrinsecus breviter semigloboso-prominentia. Folia manifeste carnoso-sub- coriacea. Nectaria haud calcariformia, sed extrinsecus ad basin corollae alte globoso- vel coniformi-prominentia. Folia semper tenuiter herbacea. C. Nectaria extrinsecus calcaria manifeste evoluta formantia. Division A contains the largest known Halenia, native of South America only, which in inflorescence, flower, and leaf habit, shows affinities with the western species of Swertia and Frasera as well. Plants in division A are very distinct and occur in Colombia and Venezuela. These Gilg apparently considered the most primitive. In division B we find the brevicornis-parviflora complex. Gilg considered H. brevicornis (HBK.) Don a valid species known only from South America, and parviflora a native of Mexico; he fur- ther described a new species erythraeoides from Venezuela which agrees in every respect with the parviflora type from Mexico. Division C is subdivided according to the length of spurs; here again, confusion has occurred, as is apt to be the case when herbarium material is scanty and field work impossible. From a limited experience in the field with a North American species, the author can state with conviction that it is possible for two plants belonging to the same species, growing side by side, to vary not only in form but in size as well. It has been found also that the spur character is inconstant. Axillary flowers and those blooming late in the season frequently possess shorter spurs or no spurs at all. If this character is so variable in one species, it 3 Gilg in Engl. Bot Jahrb. 54: Beibl. 118, p. 98. 1916. w [Vor. 20 126 ANNALS OF THE MISSOURI BOTANICAL GARDEN is reasonable to suppose that the same situation may obtain in other species. Only close attention to ecological detail and a wealth of material not yet available from South America can enable the monographer to delimit species from these little- known regions even with a fair degree of accuracy. Briquet,?* in 1931, described several new species which will be treated below. Gross MORPHOLOGY Habit.—The American species of the genus Halenia are gla- brous, strictly herbaceous, or somewhat ligneous annuals or per- ennials. They may be of caespitose habit, as is illustrated by H. caespitosa, or coarse, fleshy, foliose plants with single stems, 5 or more dm. high, as in H. hygrophila and related species, or scapose with the basal leaves disposed in a rosette, as in H. plantaginea, or slender, graceful, simple or branching plants 1.5-6 dm. high, as in H. brevicornis and its varieties. As a whole, the genus is not colorful, the flowers being various shades of yellow and yellow- green, except in the northernmost species, which has purple owers. Hoots.— The root system of the North American species is fibrous, with a persistent, slender tap-root, frequently more or less woody in texture, though typically that of an annual or perennial in cross-section. In the South American species the root is, for the most part, ligneous and thick. Stems.—The stems are simple or branched, usually erect, though they may be decumbent, as in Halenia decumbens and H. Weddelliana. They are of two types, mostly foliose as in H. deflexa and Schiedeana, or scapose as in H. plantaginea. The stems often continue underground for a short distance, sending out erect, flowering stems at irregular intervals. These may be angular or terete; if the former they are often slightly winged, due to the decurrence of the leaves, and usually faintly striate. Leaves.—The leaves are opposite and decussate, or infrequently whorled as in H. verticillata, either entirely cauline or disposed in a rosette, sessile or petiolate. The petioles exhibit the same characteristics as the stem, the decurrence of the calyx-lobes * Briquet in Candollea 4: 317. 1931. 1933] ALLEN—THE GENUS HALENIA 127 frequently being apparent. Where both petiolate and sessile leaves occur on the same plant, the basal leaves have petioles often equal to or longer than the blade, and the petioles become decreasingly shorter toward the summit of the stem, the upper- most leaves being sessile. The leaves are entire, 1-, 3- or 5- nerved, and range from ovate to obovate or spatulate, or lanceolate to linear, varying from .5 to 8 or 10 cm. in length and up to 6 em. in width. The leaves of the North American species are, on the whole, thin and herbaceous, whereas in the South American species they are frequently coarse, fleshy, or coriaceous. Inflorescence.—The inflorescence consists of terminal or axillary cymes of varying density. The flowers occur on short pedicels at the tip of a stem or branch in a compact cluster or head, as in H. brevicornis, or they may be borne loosely on long pedicels at the tip of a branch or at a node, as in H. Schiedeana. In the Central American and the more primitive of the South American species, the inflorescence is usually a spicate, racemose, or more or less umbellate cyme. A pair of small, linear, foliaceous bracts is usually present at the base of the inflorescence, often bearing tiny, undeveloped buds in their axils. Frequently the bracts approximate the leaves in size and structure, and in H. involucrata they form an involucre almost entirely enveloping the inflores- cence. Calyx.—The calyx is persistent, foliaceous, 4-parted, and the segments united only at the extreme base and arranged in pairs; the inner pair represents an inner cycle, the second an outer cycle. The segments may be caudate as in the type species, H. corni- culata, or spatulate, obovate, elliptic, lanceolate-linear, or linear, with intergradations, with obtuse, acute, acuminate, mucronate or apiculate tips. The tips very rarely are reflexed, as in H. Schiedeana. The length of the calyx varies from one-third to nearly equal that of the corolla. The calyx-segments are from 1- to 3-nerved, often reticulately veined at the tip, and both surfaces are often papillate. Squamellae are usually found on the inner surface of the calyx-segments at their base (see fig. 1). They vary in size, shape, number, and position on the lobes, and [Vor. 20 128 ANNALS OF THE MISSOURI BOTANICAL GARDEN are usually distinguishable only under the dissecting lens after the herbarium specimens have been boiled. Squamellae occur in many of the Contorti both on the calyx and the corolla-lobes. Engler and Prantl? have referred to similar structures in several genera of the Gentianaceae as discs. util Fig. 1. Types of squamellae found on the calyx in the genus Halenia. Corolla.—The corolla is marcescent, campanulate, 4-lobed, white, yellow, yellow-green, green, or purple, ranging in length from 5 mm. to nearly 3 cm. The lobes are triangularly ovate or obovate, acute, acuminate, apiculate, mucronate or obtuse, often auriculate, with an entire, erose, or crisped margin, and fre- quently papillate on both surfaces. The veining of the dex- trorsely convolute lobes may or may not be reticulate. The corolla-tube varies from one-fourth to three-fourths the length of the entire corolla. At or near the base of the tube, opposite each lobe, is a tubercle which may be merely a slight swelling in the lower portion of the tube, as in H. brevicornis, or a definite *? Gilg in Engler & Prantl, Nat. Pflanzenfam. 47: 89. 1895. 1933] ALLEN—THE GENUS HALENIA 129 spur longer than the corolla, as in H. guatemalensis (pls. 8-11). It frequently happens that the corolla lacks spurs entirely in the axillary flowers or in those occurring late in the season, as, for example, in H. deflexa and other species. Not only in Halenia is this situation apparent, but it is recalled that in other normally spurred plants, for example, Linaria canadensis, spurs are fre- quently absent. "Therefore it has seemed advisable to discon- tinue heterantha as a variety of H. deflexa, since the occurrence of these spurless forms is more or less frequent within the genus. Gilg noted that in certain species of Halenia, for example, H. brevicornis, etc., the flowers on the main stalk are normally large and wide open, while below, on the same stem, they are definitely smaller, apparently not opening at all or else very slightly. The structure is similar, with the exception of the absence of spurs. The ovary in these abnormal flowers contains fewer seeds, and the capsule is much reduced in size. This would indicate, according to Gilg, a gradual reduction of chasmogamous flowers which depend on insect pollination, to more or less cleistogamous flowers. The shape of the spurs varies exceedingly, from slender to very thick and broad, or from spreading and ascending to pendulous and incurved. Intergrading forms are frequent. The spurs are frequently conspicuously veined and apparently glandular. Stamens.—The stamens are equal in number to, and are borne alternate with, the lobes of the corolla at or near the summit of the tube. The filaments are adnate to the corolla, and the anthers are often enfolded in the bud by the margins of the corolla-lobe. The anthers are versatile, deltoid, ovate or oblong, often mucronate and papillate. The filaments are usually linear, but they may become more or less dilated and, rarely, papillate. The pollen grains are constant for the genus, being more or less tetrahedrally spherical, and having three pores. Pistil.—The pistil is sessile. The stigma is usually sessile and cleft to expose the two inner stigmatic surfaces; the lobes may be truncate or ovate. The ovary is bicarpellate, the margins of each carpel being infolded and becoming the region of attachment for the numerous ovules. Fruit.—The fruit is a flattened, lanceolate to lanceolate- [Vor. 20 130 ANNALS OF THE MISSOURI BOTANICAL GARDEN obovate capsule, frequently subfalcate, and usually exserted. It is unilocular at maturity, though in the young state it appears to be more or less two-celled. The fruit dehisces septicidally along the inner surface of each locule tip. Seed.—' The seeds vary in shape, being globose to ovoid or elliptic, often flattened. The surface is reticulate in the majority of the South American species and in the brevicornis complex, except for var. latifolia. The remaining species show the surface of the seeds to be minutely granular instead of reticulate. They vary in size from 0.5 to 1 mm. in diameter, and in color from dull greenish-brown and yellow-brown to dark, shiny brown, the latter usually being typical of those with reticulate surface. The age of the plant and the conditions attending its collection no doubt have their influence on the color, size, and, to some extent, the texture of the seed-coat. FLoRAL ANATOMY The major portion of the anatomical investigation of the Gentianaceae, particularly the Menyanthoideae, has been con- cerned with the stem and leaf structure. The most complete anatomical data is found in Gilg’s treatment of the Gentianaceae in ‘Die Natürlichen Pflanzenfamilien.^? Solereder’s ‘Systematic Anatomy of the Dicotyledons'* gives very little additional in- formation. Since that time more attention has been given to floral morphology as a separate study. Stolt,? in 1921, made an exhaustive cytological survey of the flowers of several genera, among which was included Halenia elliptica, an Asiatic species having affinities with Halenia deflexa of North America. Inci- dentally, Stolt inserted a diagram of the transverse section through the ovary, which indicates clearly the vascular system of that portion of the flower. In order to make the present monograph as complete as possible, anatomical study was undertaken. Fresh flowers of Halenia deflexa, a species with both spurred and spurless forms, were obtained by the author in Vermont and preserved in 70 per cent 40 Gilg in Engler & Prantl, Nat. Pflanzenfam. 4?: 50. 1895. “| Solereder, Systematic Anatomy of the Dicotyledons 1: 548-550. 1908. * Stolt in K. Svensk. Vet.-Akad. Handl. 61"; 1-56. 1921. 1933] ALLEN—THE GENUS HALENIA 131 alcohol. The pickled material was dehydrated and embedded in paraffin following the butyl alcohol method outlined by Zirkle, sectioned at 10 u, and stained with crystal violet and erythrosin. The accompanying drawings (pl. 12) were made with the aid of a “Promi” microscopic drawing and projecting appara- tus. The xylem has been cross-hatched in order to differentiate it from the other vascular elements. Transverse sections of the spurred form of Halenia deflexa show that the vascular system of the pedicel is an amphiphloic siphonostele (pl. 12, fig. 1). Approaching the receptacle, the stele enlarges and assumes a rhombic form (fig. 2). The de- currence of the outer lobes of the calyx is apparent. Shortly thereafter (fig. 3), the midribs (a) of the two outer calyx-lobes leave the receptacular stele (r). They migrate outward, and from either, two lateral traces (a’) are given off (fig. 4). At this point, four lacunae (z) appear, prior to the severing of the calyx from the receptacle. In the succeeding illustration (fig. 5) these lacunae (z) merge into two crescent-shaped fissures, and the midribs of the two inner calycine lobes (b) leave the stele. In fig. 6 the calyx-tube is entirely free from the receptacle, and in the axil of each potential lobe are visible 4-8 minute emergences or squamellae (e). The stele has again assumed a more or less rhombic shape, but from the 4 angles, traces (c) depart centri- fugally, soon resolving into the midrib (n) and two laterals (c’) destined to supply each of the 4 corolla-lobes (fig. 7). The disruption of the residual vascular cylinder is continued, and the 4 staminal traces (f) are fully differentiated. At this in- terval the corolla-tube is virtually free from the receptacle. In the succeeding illustration (fig. 8) lacunae appear at m, the ovary at this point appearing bilocular. "The residual stele con- sists of two roughly semi-circular masses with a concentration of lignified elements (l) at either end. The calyx-lobes are free, and sections near the tip of the reflexed spurs are found, the origin of which is to be described. In the following illustration (fig. 9) the origin of the spurs is apparent, and a cleft (y) is visible, indicating the sinus. Simul- taneously, 4 protrusions occur on the inner surface of the corolla, 55 Zirkle in Science, N. S. 71: 103. 1930. [Vor. 20 132 ANNALS OF THE MISSOURI BOTANICAL GARDEN below the sinus, preparatory to the severing of the staminal filaments (f). The ventral traces (1) and dorsal (k) are dis- cernible. The ovules (0) are evident, showing their position in the axils of the placentae (p). In the final stage, taken from a section through the tip of the flower (fig. 10), the differentiation of the corolla-lobes and the isolation of the staminal filaments (f) are complete. The pla- centae (p) have diminished in size. The spurless form shows a similar vascular system except for the absence of spurs, the presence of fewer ovules, and finally, a more pronounced dorsal trace (k), after the cessation of ovular production. GEOGRAPHICAL DISTRIBUTION The accompanying maps show three centers of distribution of the species of Halenia in America. The first (figs. 2, 4) extends from Labrador and Newfoundland, south to New York and west to British Columbia and Montana. The one species and its variety found in this area grow in moist or dry situations in calcareous, slaty, or alluvial soil, in open woods or fields, on stream banks or along the sea-shore, usually in the shade. The habit varies with the habitat. This distribution follows closely the northern region of glaciation, and coincides with the usual distribution areas of herbaceous species common in that territory. This same species has been collected three times in the State of Mexico, but has not been reported from the intervening region, a fact that might suggest a previously more continuous distribu- tion of the species from the northern Rockies along the mountain ranges to the Mexican Sierras. The second area comprises a region extending from the Chiri- cahua Mountains of New Mexico and Arizona, southward to Costa Rica in Central America. Here are a few wide-spread mountain species; but for the most part, they are endemics occurring in volcanic areas (fig. 3). The third large center of distribution is the northern part of South America, where the genus is represented by a relatively large number of endemic species (fig. 5). All of the North and Central American species, with the Ld ee d l3 Fig. 2. Map of southern Canada and northern United States, showing the geographical distribution of the species of Halenia: 1 = H. defleza; 2 = H. deflexa var. Brentoniana. 14. H. brevicornis Fig. 3. Map of southern United States, Mexico, and Central America showing the geographical distribution of Halenia. 16. . var. multiflora 4. H. crassiuscula 5. H. Palmeri. 3. H. recurva 24. H. Shannonit 20. var. ovata 13. H. decumbens 8. H. plantaginea 26. H. rhyacophila 23. f. compacta 21. var. Tuerckheimii 25. H. guatemalensis 9. f. grandiflora 28. var. procumbens 29. H. caleoides 30. var. latifolia 22. H. platyphylla 27. var. macropoda 6. H. Conzaitii 12. H. nudicaulis 10. H. Pringlei 11. H. Schiedeana 1933] ALLEN—THE GENUS HALENIA P A o^ CRAM'S 85x 11 Outline Map SOUTH AMERICA -— convmant THE GEORGE F CRAM COMPANY D Ü INOIAMAPOL! en, E PN x, , meats A — DLE Fig. 4 s * 2 2. ^c t e L 5 ) APT 7 \ p jJ | $ H LES s $, , H | ty os AM Z A i p "d | hag fs (eos i T / " . ] t| Ai Ya / i a a 2 / me š í A $2 .] ^ x I AOA STT | ETE d ann PETER IE i FERES d ü EB `~ 7 [3 i | Te s s s z Lj = * mn w o2 (Maps reproduced Fig. 5 Maps of North and South America, showing the principal areas of distribution of Halenia. by permission of the George F. Cram Company). exception of Halenia alata and H. brevicornis and its varieties, are spurred forms which show a definite relationship to one another. Those from South America are of two types: spurred type, similar to the North American species (excluding brevicornis and [Vor. 20 134 ANNALS OF THE MISSOURI BOTANICAL GARDEN alata) and showing affinities with them; non-spurred type, with tubercles instead of spurs and quite different in habit from the brevicornis and alata. Halenia brevicornis and its varieties form a connecting link between the second and third areas, as well as a morphological link, so to speak, between the primitive Swertiella and the more advanced Haleniastrum. The varieties of brevi- cornis do not extend further south than Guatemala, though the species proper is found in South America. Here again, as in Mexico and Central America, exist endemies which show a development almost parallel with that which has occurred north of the equator. An example of this is clearly shown by Halenia decumbens, from the mountains of Mexico, which bears a striking resemblance to H. Weddelliana, a species from Ecuador, Colombia, and Peru. SYSTEMATIC POSITION The Gentianaceae, as pointed out by Gilg,“ shows closer affinity to the Loganiaceae than to any other family in the Contorti, yet the differences are so well marked that the two families are never confused. Grisebach separates the Gentianaceae into two subfamilies, the Gentianoideae and the Menyanthoideae. In the first subfamily the leaves are always simple, entire, sessile, and never al- ternate. The aestivation of the corolla is never valvate. The second subfamily, the Menyanthoideae, has alternate and mostly petiolate, sometimes trifoliate, leaves, and the aestivation is induplicate-valvate. Halenia belongs to the first division. The structures called squamellae, described early in the text, are never found in the Menyanthoideae but may occur in other genera of the Gentianaceae, at the base of the corolla or calyx- lobes. The nearest relative of Halenia is Swertia. The most primitive forms of Halenia, particularly the South American species, are often confused with Swertia. The primitive members of Halenia, instead of the definitely spurred corolla typical of the majority of the species of the genus, possess small, knob-like protuberances or nectaries which upon casual examination might pass unnoticed “ Gilg in Engler & Prantl, Nat. Pflanzenfam. 42: 50. 1895. 1933] ALLEN—THE GENUS HALENIA 135 or be taken for the nectaries which are structurally distinct and characteristic of Swertia. More detailed study reveals the fact that the depressions or spurs of Halenia are without the marginal fringe which is always conspicuous in Swertia (fig. 6). Halenia never has a corona, but in Swertia it is frequently present. The corolla-lobes of Halenia are dextrorsely convolute in the bud, whereas the reverse is the case in Swertia. Anatomical study discloses other generic differences. a b C d Fig. 6. Swertia perennis L.: a, bud; b, interior surface of petal. Halenia brevicornis (HBK.) Don: c, bud; d, interior surface of petal. Within the genus itself two rather distinct sections are ap- parent. The more primitive, which was mentioned above as being near Swertia and possibly originating from it, may be called tentatively Swertiella. It contains those species from South America which were considered by Gilg as being most primitive. Halenia alata of Mexico also belongs to this section. The Swer- tiellae may be characterized by the absence of spurs or the pres- ence of small tubercles or prominences at the base of the corolla, which are not visible usually from the outside. They are coarse, fleshy, foliose plants, and for the most part with dense inflores- cences. The second section, Haleniastrum, is more advanced and is readily recognized by the presence of spurs of varying [Vor. 20 136 ANNALS OF THE MISSOURI BOTANICAL GARDEN length and shape. Both herbaceous and more or less woody types comprise this section. An interesting discussion of the phylogeny of the Gentianaceae was presented by T. H. Huxley* before the Linnaean Society of London, April 7, 1887. Confining his study to the structure of the corolla, he separated the family into two groups; the first division Permelitae was segregated mainly on the presence of à series of nectarial cells on the inner surface of the cup. The Permelitae were again divided into four groups with various BREVIGORNIS — DASYANTRA xe du HÉRZOGII mig WERERBAUERI mapa — T anarTLA ELATA TERRIEN.. Tien €— umir M 98 EA ps LE CURANDA sso KILLIPII HYPOTHETICAL ANCESTOR Fig. 7. Chart showing the probable phylogenetic relationship of species of Halenia. modifications of floral structure. The first, Actinanthe, which he considered the most primitive and the least differentiated, con- tained among other elements, Exadenus. The latter was in- cluded as a transitional stage on account of its rudimentary spurs, leading to the second division, the Keratanthe, containing only one genus Halenia, which represents an extreme modification of the Exadenus type. The third related division, the Lophanthe, 5 Huxley, T. H. Jour. Linn. Soc. Bot. 24: 101. 1887. 1933] ALLEN—THE GENUS HALENIA 137 depicts another outgrowth of the Actinanthe type, but with filamentous appendages or fimbriae. Swertia and Frasera fall into this category. Text-figure 7 shows the probable relationship of the various species of Halenia, according to the author's interpretation. For the sake of convenience, in the taxonomic treatment of the genus, the North and South American species are taken up separately in each section. ABBREVIATIONS In the citations of specimens examined the following abbrevia- tions have been used to denote the various herbaria from which specimens were used for study. ANSP = Herbarium of the Academy of Natural Sciences of Philadelphia. B n. G = Botanical Garden, Berlin BB = Brooklyn Botanical Garden. BM - British Museum of Natural History, London. B = Herbarium of the Botanical Garden, B ls. C — Herbarium of the En of Chicago, deposited in the Field Museum of Natural Histo CAS = Herbarium of the California Academy of Sciences. D = Dudley Herbarium of ^r burns Jr. University. DH = Delessert Herbarium of Gen F = Herbarium of the Field Musei of Natural History. G = Gray Herbarium of Harvard University. HP = Herbarium of H. Pittier, Director de Museo Comercial, Caracas, Venezuela. HJP - Herbarium of the Jardin des Plantes, Paris IAC = Herbarium of the Iowa Agricultural Collage. K = Herbarium of the Royal Botanic Gardens, Kew L = Linnaean Herbarium, Linnaean Society of cm London. M = Herbarium of the Missouri Botanical Mera MU = Herbarium of the University of Missour NY = Herbarium of the New York Botanical rte P — Parry Herbarium deposited at the Iowa Rosipulturel College. S = Herbarium of the Botanical Museum, Stockholm. SM = Herbarium of the State Museum, Albany, New York. UC = Herbarium of the Botanical Mri of the University of Copenhagen. US — United States National Herbariu i = Herbarium of the Natural History ‘Museum, Vienna. TAXONOMY Halenia Bórckh. in Roemer, Arkiv für Botanik 1!: 25. 1796; Ruiz & Pavon, Fl. Peruv. 3: 21, pl. 242, fig. 1. 1802; Endl. Gen. Pl. 601. 1836-40; Grisebach, Obs. Gent. 36. 1836; G. Don, Gen. [Vor. 20 138 ANNALS OF THE MISSOURI BOTANICAL GARDEN Hist. 4: 177. 1838; Grisebach, Gen. & Sp. Gent. 322. 1839; Dietrich, Syn. Pl. 2: 918. 1840; Hooker, Fl. Bor.-Am. 2: 67, pl. 155-6. 1840; Grisebach in DC. Prodr. 9: 128. 1845; Grisebach in Linnaea 22: 45. 1849; Weddell, Chlor. And. 2: 74. 1859; Benth. & Hooker, Gen. Pl. 2: 817. 1876; Hemsl. Biol. Cent.-Am. Bot. 2:351. 1882; Baillon, Hist. Pl. 10:142. 1891; Gilg in Engler & Prantl, Nat. Pflanzenfam. 42: 89. 1895; Conzatti, Fl. Syn. Mexico, 174. 1897; Rouy, Ill. Pl. Eur. 17: pl. 412. 1902; Gilg in Fedde, Rep. Spec. Nov. 2:52. 1906; Robinson in Gray's Manual, ed. 7. 659. 1908; Britton & Brown, Ill. Fl. 3: 15, fig. 3365. 1913; Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p. 93. 1916; Johnson, Tax. Fl. Pl. 488. fig. 340. 1931; Briquet in Candollea 4: 317. 1931; Rydberg, Fl. Cent. N. Am. 636. 1982. Tetragonanthus Gmelin, Fl. Sib. 4: 114, pl. 53. 1769; Kuntze, Rev. Gen. Pl. 2: 431. 1891; Britton, Manual, 734. 1901; Small, Fl. Southeastern U. S. 931. 1913; Rydberg, Fl. Rocky Mts. 666. 1922. Swertia Linnaeus, Amoen. Acad. 2: 344. 1751; Jussieu, Gen. Pl. 158. 1791; Michaux, Fl. Bor.-Am. 1: 97. 1803; Humboldt, Bonpland & Kunth, Nov. Gen. & Sp. Pl. 3: 174. 1818; Kunth, Syn. Pl. 2: 266. 1823; Schlectendal & Chamisso in Linnaea 5: 122. 1830. Ceratia Persoon, Syn. Pl. 1: 287. 1805; Hedw. Gen. 181. 1806. Exadenus Grisebach, Obs. Gent. 36. 1836; Gen. & Sp. Gent. 322. 1839; in DC. Prodr. 9: 128. 1845. Glabrous caulescent annual, biennial, or perennial herbs. Root fibrous, varying from typically herbaceous to extremely ligneous. Leaves membranaceous or fleshy, opposite, entire, sessile or petiolate, usually 3-5-veined, veins frequently sub- merged in fleshy type. Inflorescence a terminal or axillary, subumbellate, or rarely racemose or spicate cyme. Calyx 4- parted, foliaceous, linear, lanceolate, ovate or spatulate, usually papillate, often bearing small squamellae at the base of each lobe. Corolla 4-parted, white, yellow, green, or purple, marcescent, campanulate, the tube of varying length; the lobes dextrorsely convolute, elliptic to ovate, obtuse to acute or apiculate, entire, crisped or erose, frequently auriculate and papillate. Stamens 1933] ALLEN—THE GENUS HALENIA 139 4, included, adnate to the corolla-tube at varying heights, and alternate with the corolla-lobes; filaments linear, occasionally somewhat dilated; anthers 2-celled, ovate, oblong or subtriangu- lar, versatile. Carpels 2, sessile, the edges being infolded to form a parietal placenta bearing many ovules; stigma sessile, composed of two oblong or ovate lobes, the inner surfaces of which are stigmatic. Fruit a compressed capsule, lanceolate to ovate, often subfalcate, septicidially dehiscent from the tip. Seeds globose or slightly flattened, brown or greenish tan, granular or reticulate. Type species: H. corniculata (L.) Druce in Rept. Bot. Exch. Cl. Brit. Isles 3: 419. 1914. SYNOPSIS OF THE SECTIONS OF THE GENUS Plants usually coarse with fleshy leaves, rarely slender with thin, herbaceous leaves; stem usually leafy; spurs absent, or present as very small inconspicu- ous ida frequently obscured by calyx; distribution chiefly Sout esie eer errr eT CORE D 1. Swertiella Planta usually slender, with thin, herbaceous leaves; om leafy or unio spurs present; distribution North and South America.......... 2. Haleniastrum SECTION 1. SWERTIELLA KEY TO NORTH AMERICAN SPECIES AND VARIETIES 1. Leaves mostly radical; stem more or less BOBRDOSÓE ID sis d ns seruo OS 1. H. alata 1. Leaves mostly cauline; stem not scapose. 2. Leaves ovate, less than 1.2 cm. long......... 2f. H. brevicornis var. ovata 2. Leaves not ovate, longer than 1.2 cm. 3. Pedicels filiform, slender, elongate; habit decidedly spreading....... VE OP RO Bik "ke win fe RA n GE g. H. brevicornis var. Tuerckheimii 3. Pedicels stouter than filiform, —— 4. Corolla without distinct spur 5. Inflorescence strict................ 2c. H. brevicornis var. micranthella 5. Inflorescence compact. 6. Leaves linear, slender... 60 Lese cecil 2. H. brevicornis 6. Leaves ovate to lanceolate, coarse....2a. H. brevicornis var. latifolia 4. Corolla with small, but distinct spurs. 5. Spurs thick, conical, more or less pendulous................... n a BS Rs. 6 96% 4 so X o oc 2b. H. brevicornis var. multiflora 5. Spurs blunt, spreading, hog .2e. H. brevicornis var. chihuahuensis 5. Spurs slender, divergent. ............. 9d. H. brevicornis var. divergens 1. H. alata (Mart. & Gal.) Hemsl. Biol. Cent.-Am. Bot. 2: 351. 1882. Exadenus alatus Mart. & Gal. Bull. Acad. Brux. 11!: 372. 1844; Walper's Rep. Bot. Syst. 6: 508. 1846-7. [Vor. 20 140 ANNALS OF THE MISSOURI BOTANICAL GARDEN Tetragonanthus alatus Kuntze, Rev. Gen. Pl. 2: 431. 1891. Small perennial with 1-2 simple erect slightly winged stems, about 0.5-1 dm. high; numerous radical leaves, up to 3.5 em. long and .4 em. broad, crowded, attenuate into long petioles, oblanceolate, obtuse, 3-nerved; lower cauline leaves almost twice the length of the radical, subsessile; upper linear, obtuse, sessile; inflorescence consisting of a few (4-6) terminal flowers on 4- winged pedicels; calyx foliaceous, slightly shorter than corolla, segments oblong, 3-nerved, papillate, acute; corolla yellow, subrotate, up to .6 em. long, tube about one-half as long as the entire corolla; corolla-lobes ovate, obtuse; filaments linear; anthers ovate; capsule broadly ovate; seeds yellow-brown, globose, granular. Distribution: rocky forests of southern Mexico. Specimens examined: Vera Cruz: in forests and on trachytic rocks on Mt. Orizaba, alt. 2250-3000 m., June—Oct. 1840, Galeotti 7221 (BG, B TYPE, DH, K, V); Mt. Orizaba, alt. 3000-3125 m., Aug. 1838, Linden 934 (DH, K). 2. H. brevicornis (HBK.) G. Don, Gen. Hist. 4: 177. 1838; Wedd. Chlor. And. 2: 77. 1859; Hemsl. Biol. Cent.-Am. Bot. 2: 352. 1882; Gilg in Engler & Prantl, Nat. Pflanzenfam. 4?: 89. 1895. Swertia brevicornis HBK. Nov. Gen. & Sp. Pl. 3: 174. 1818. S. parviflora HBK. Nov. Gen. & Sp. Pl. 3: 174. 1818. S. parviflora var. « angustifolia Sch. & Cham. in Linnaea 5: 122. 1830. S. cucullata Sessé & Mociño, Fl. Mex. 79. 1894. Halenia parviflora G. Don, Gen. Hist. 4: 177. 1838; Wedd Chlor. And. 2: 77. 1859; Hemsl. Biol. Cent.-Am. Bot. 2: 352. 1882. H. paucifolia Hemsl. Biol. Cent.-Am. Bot. 2: 352. 1882. H. erythraeoides Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p. 105. 1916. Exadenus parviflorus Griseb. Gen. & Sp. Gent. 322. 1839. E. brevicornis Griseb. l. c. 323. E. paucifolius Mart. & Gal. Bull. Acad. Brux. 11!: 372. 1844. Tetragonanthus paucifolius Kuntze, Rev. Gen. Pl. 2: 431. 1891. T. parviflorus Kuntze, l. c. T. brevicornis Kuntze, l. c. 1933] ALLEN—THE GENUS HALENIA 141 Annual, 1.5-4.5 dm. high; root slender; stems erect, slender, usually simple below and branched above, sometimes branched from the base, slightly angled, and striate; leaves sessile, sub- connate, lower linear to lanceolate, up to 3 cm. long, .4 cm. broad, usually prominently uninerviate, upper shorter, narrower, up to 2 em. long, linear; inflorescence compact cymose clusters, pedicels erect, 1.8 em. or less in length; calyx-segments lanceolate, approximately one-half to two-thirds the length of the corolla, usually 3-nerved; corolla .4—.8 cm. long, tube one-half or slightly more the length of the entire corolla; corolla-lobes broadly to narrowly ovate, acute, more or less auriculate; spurs merely small depressions near the base of the corolla-tube, usually not visible to the naked eye; stamens about .2 cm. long, attached usually near the middle of the tube; filaments linear, anthers usually deltoid; ovary lanceolate, about .5 cm. long; seeds minute, ovoid to subglobose, brownish, finely reticulate. Distribution: isa in mountains, from Mexico, southward to Peru. Specimens exami Mexico: San Luis Porosi: Oct. 1879, Schaffner 421 (BG, BM, F, NY, UC, US); region of San Luis Potosi, alt. 1500-2000 m., 1878, Parry & Palmer 600 (ANSP, BM, F, G, IAC, K, M, NY, US) VERA Cruz: Mt. Orizaba, alt. 1875 m., 1841-43, Liebmann 10775 (UC); same locality and date, alt. 2500 m., Liebmann 10778 (UC); same locality, 1840, Galeotti 7219 (B TYPE of Exadenus paucifolius, DH); “in dumetis prope Jalapam," 1830, Schiede & Deppe 247a (BG TYPE of Swertia parviflora var. « angustifolia, BM, M) GUANAJUATO: near city Guanajuato, alt. 1650 m., Humboldt & Bonpland (BG TYPE not seen, HJP, M photo.). PUEBLA: vicinity of Puebla, ravines, Hacienda Alamos, route Vera Cruz, alt. 2170 m., Nov. 5, 1907, Arsène 2098 (G, M, US); mountains, Esperanza, Aug. 1907, Purpus 2697 (BG, BM, CAS, F, M, US); Chinantha, alt. 1750-2000 m., May 1841, Liebmann 10776 NY, UC). Mexico: La Cima, alt. 2500 m., | alt. 1500 m., Aug. 1904, Kuntze 23783 (NY); Volean de Toluca, Heller 401 tee 1864-70, Ghiesbreght 137 He K, NY); Ghiesbreght 618 (G, K, M). XICO WITHOUT LOCALITY: Tate (DH); Oct. 8, 1897, Berlandier 1207 ex 13 (DH); Berlandier 1207 (V); Pavon (DH); Wawra 425 (V). CENTRAL AMERICA: CosrA Rica: *in monte Reventado,” alt. 2250 m., coll. of 1847, Oersted 10779 (UC) NicanAGUA: El Viejo, Oersted (M). 2a. H. brevicornis var. latifolia (Sch. & Cham.) Allen, n. comb. [Vo 20 142 ANNALS OF THE MISSOURI BOTANICAL GARDEN Swertia parviflora var. 8 latifolia Sch. & Cham. in Linnaea 5: 122. Ezadenus parviflorus var. 6 latifolius Griseb. Gen. & Sp. Gent. 322. 1839; in DC. Prodr. 9: 128. 1845 Halenia parsiftera var. latifolia Herrsl. Biol. Cent.-Am. Bot. 2:351. 1882. A more sturdy plant than the species, attaining a height of 6 dm.; leaves larger than species, 1.5-4 em. long, .3-1.5 cm. broad, acute or obtuse, lowermost smaller, ovate, rotund, and with longer petioles; inflorescence frequently less compact than that of species and with longer pedicels; calyx-segments lanceolate, foliaceous, one-half to two-thirds the length of the corolla, papillate, 3-nerved; corolla .6-.8 em. long, tube about equalling the ovate, frequently papillate, auriculate lobes; spurs short, consisting of very slight angular protrusions froi the extreme base of the corolla-tube, giving the corolla a square appearance. Distribution: mountains of Mexico. Vera Cruz: “in dumetis prope Jalapam,” 1830, Schiede & Deppe 247b (BG TYPE of Swertia parviflora var. 8 latifolia Sch. & Cham., BM, M, S, V); same locality, coll. of 1833, Beyrich (S); region of Orizaba, Oct. 20, 1866, Powpesü 3126 (BG, B, K, 8, UC, US). PuxBLA: Manzanilla, alt. 2250 m., Nov. 24, 1908, — E 703 (B, US); Barrancas, Hacienda Alamos, route Vera Oris, alt. 2170 m., Dec. 10, 1907, Arsène (M, US); Boca del Monte, alt. 2300 m., Nov. 16-19, 1907, Anim fa More tos: Lecima, Sierra de Ajusco, Aug. 18, 1896, Harehberger 13? (ANSP, G, US ). TLaxcaLa: Chiautempan, alt. 2250 m., Nov. 10, 1908, Arsène 1711 (B, US). Mexico: Ajusco Mountains, 1905, Luton & Lemmon (CAS); Valley of Mexico, ate Le Ixtaccihuatl, Jan. 1903, Purpus (CAS); due de Agua, Nov. 1 Purpus 1762 (CAS, F, G, M, US); Sierra de las Cruces, alt. 2375 m. , Sept. 12, 1904, Pringle 13120 (BG, F, G, K, 8, UC, US); Eslava, Salazar (US); near Salazar, Sept. 14, 1903, Rose & fee 7026 (US); Désierto Viejo, near Mexico, Sept. 6, 1865, Bourgeau 799 (BG, B, DH, G, K, NY, 8, UC, US). MICHOACAN: vicinity of Morelia, alt. 2500 m., Oct. 26, 1911, Arsène 5610 (M, US). Mexico witHout Locauity: Berlandier 1207 (DH); 1832, Alaman (DH). 2b. H. brevicornis var. multiflora (Benth.) Allen, n. comb. Halenia multiflora Benth. Pl. Hartw. 24. 1839 Plant more sturdy than the species, frequently branched above, rarely branched at the base, 4.5 dm. or less high; leaves up to 1.3-2.5 em. long, .2-.9 em. broad, obtuse or acute, ovate to 1933] ALLEN—THE GENUS HALENIA 143 narrowly lanceolate, frequently faintly 3-nerved, midvein always prominent; inflorescence usually densely flowered, though com- pactness of arrangement varies; mature corolla with small thick rounded pendant spurs at its base. Distribution: in «oi ERN of Mexico. Specimens examin San Luis Porosi: in posta baie of San Miquelito, Aug. 1877, Schaffner 88 (G); Alvarez, Sept. 28-Oct. 3, 1902, Palmer 160 (CAS, F, G, M, NY, UC, US); Parry & Palmer 600b (IAC, M). ZACATECAS: on the Sierra de los Morones, near Plateado, Sept. 1, 1897, Rose 2732 US). Janisco: ‘‘in pinetis Bolaños,” 1839, Hartweg 210 (BM, DH, G, K, NY, V); Sierra minis: west of Bolafios, Sept. 15-17, 1897, Rose 2962 (US); banks of ravines near Guadalajara, alt. 1250 m., Oct. 21, 1903, Pringle 11636 (BG, F, G, K, US); hills near Guadalajara, Oct. 14, 1889, Pringle 2735 (C, IAC); Rio Blanco, Oct. 1886, Palmer 680 (G, NY, US); same locality and date, Palmer 683 (G). GUERRERO: between Ajusinapa and Petatlan, alt. 1250-1750 m., Dec. 14, 1894, Nelson 2126 (US). Oaxaca: Sierra de San Felipe, alt. 2000 m., Oct. 13, 1894, Smith 665a (M, NY, US); same locality, alt. 1800 m., Aug. 15, 1898, Conzatti & Gonzalez 878 (G); San Pedro Nolasco, alt. 1875 m., Oct. 1840, Galeotti 1490 (B, DH) 2c. H. brevicornis var. micranthella (Briq.) Allen, n. comb. Halenia micranthella Briq. in Candollea 4: 320. 1931. Plant 1.5-6 dm. high; leaves linear-lanceolate, often obtuse, 3-nerved, the lower long-petiolate, the upper linear, sessile; inflorescence usually less compact, but more strict, than in the species, and the stem and pedicels more erect; the nodes of the inflorescence approximately equidistant, giving the appearance of a narrow raceme; corolla campanulate, but more narrowed at the base than in be species; calyx-segments usually one-half the length of the corolla; corolla-lobes broadly ovate, acuminate, auriculate; spurs reduced to minute depressions, frequently not visible to the naked eye; anthers usually broadly ovate; filaments varying, usually linear, rarely dilated. Distribution: mountains of Mexico. Specimens examined: HipaLGo: El Chico, near Pachuca, Sept. 1905, Purpus 1761 (CAS, F, G, M, NY, US); wet meadows, Sierra de Pachuca, alt. 2450 m., Aug. 13, 1898, Pringle 6964 (ANSP, BG, B, CAS, DH, F, G, IAC, K, M, NY, S, US, V); same locality, Sept. 8, 1899, Pringle 7943 (BG, F, G, K, M, NY). Mexico: Sierra de a o Nov. 9, 1903, Pringle 11842 (BG, F, G, K, S, UC, US); near Ozumba, alt. 2000 m., Nov. 3, 1902, Pringle 11329 (G, UC, US). Souta Mexico wiht Locauity: July 1841, Liebmann 10777 (UC). [Vor. 20 144 ANNALS OF THE MISSOURI BOTANICAL GARDEN Briquet based his new species on Pringle 6964. This is cited in the original publication as 1964, but this is merely a typo- graphical error. The Pringle specimens are about 12 cm. high. The additional material cited above undoubtedly belongs to the same species based on Pringle 6964, but for the most part it consists of plants over 15 cm. in height. It is possible that there are two distinct plants under Purpus 1761, but since the variation in the brevicornis complex, as a whole, is so pronounced, these differences have been considered as variations typical of the variety. Hence, all sheets of Purpus 1761 have been determined as var. micranthella. 2d. H. brevicornis var. divergens“ Allen, n. var. Similar to var. multiflora but with more slender spurs which diverge, making the corolla broader at the base than at the tip. Distribution: central Mexico. Specimens ex t Mıcnoacan: Loma Sta. Maria, vicinity of Morelia, alt. 2000 m., Sept. 4, 1910, Arsène 55 (F); vicinity of Morelia, near La Huerta, alt. 1950 m., Sept. 1, 1910, Arsène (M TvPrE, US); Loma Sta. Maria, alt. 2050 m., Sept. 19, 1910, Arsène (M, US); same locality, alt. 1950 m., Sept. 4, 1910, Arsène 5957 (K, M, S, US); same locality, Oct. 28, 1910, Arsène 5864 (K, M, US) Vera Cruz: Orizaba, 1853, Müller (NY). Mexico wrrnovT Locauity: 1858, Sumichrast (DH). 2e. H. brevicornis var. chihuahuensis‘’ Allen, n. var Similar to var. multiflora but with lower leaves always elliptic- ovate, about 1 cm. long, increasing in length and acuteness and decreasing in width as they approach the summit of the stem; inflorescence more loosely arranged, and spurs more blunt but not quite so pronounced as in multiflora; calyx-segments three- fourths the length of the corolla, and usually narrower. ** H. brevicornis var. divergens Allen, var. sse intei a var. multiflora calcari- bus tenuioribus divergentibus latioribus basi quam summo.—Mi1cHoACAN: vicinity of Morelia, near La Huerta, alt. 1950 m., Sept. 1, 1910, Arsène (M TYPE, *' H. brevicornis var. chihuahuensis Allen, var. nov.—Differt a var. multifor folis inferioribus semper elliptico-ovatis ca. 1 cm. longis, prope summum caulis longitudine augmentibus et latitudine deminuentibus; inflorescentia laxiora; Ld bus obtusis; calycis segmentibus 34 corollae longitudini adaequantibus, plerumque angustioribus.—CmrHuAHUA: pine plains, base of the Sierra Madre, Sept. 26, 1888, Pringle 1664 (BG, BM, B, CAS, DH, M yer, S, V). 1933] THE GENUS HALENIA 145 ALLEN Distribution: mountains of Chihuahua, Mexico. Specimens examine CHIHUAHUA: pine plains, base of the Sierra Madre, Oct. 4, 1887, Pringle 1830 (ANSP, CAS, F, G, K, NY, US); same locality, Sept. 26, 1888, Pringle 1664 (BG, BM, B, CAS, DH, M ryps, S, V); mesa, west of Hop Valley, Sierra Madre Mountains, alt. 1750 m., Sept. 17, 1903, Jones (BM, M, US); southwestern Chihuahua, Aug.- Nov., 1885, Palmer 408 (ANSP, G, IAC, K, NY, US) 2f. H. brevicornis var. ovata‘? Allen, n. var. Plant 5-6 dm. high; stems erect, simple below, bearing short floriferous branches only at tip; leaves about 12 pairs, shortly petiolate, frequently bearing buds in the axis; lower leaves broadly ovate, usually with a prominent midvein, reticulate, less than .5 cm. long, .3—4 em. broad, abruptly acuminate; middle cauline leaves ovate, 1-1.2 cm. long, .7-.8 cm. broad, acuminate; upper leaves up to 1.5 em. long, lanceolate; inflores- cence and flowers similar to multiflora type, but spurs more slender, slightly incurved, and divaricate. Distribution: known only from type locality. Specimens examined: Mexico: Tepic, Jan. 5-Feb. 6, 1892 Palmer (US TYPE). 2g. H. brevicornis var. Tuerckheimii (Briq.) Allen, n. comb. Halenia Tuerkheimii Briq. in Candollea 4: 317. 1931 (dedi- cated to H. von Tuerckheim). Slender graceful stem, loosely branching, up to 6 dm. high; leaves lanceolate to lincar-lanedelata or elliptic, about 1 cm. long, .5 em. broad, lower cauline leaves becoming more acute, faintly 3-nerved, midvein prominent; middle cauline leaves 3 cm. or less long, .5 em. broad; inflorescence in loose, terminal or axillary, few-flowered cymes, each flower borne on a long slender attenuate pedicel; first flowers with very small, though definitely formed, spurs, at the base of the corolla-tube; later and usually axillary owers of more slender habit and without spurs. 48H. brevicornis var. ovata Allen, var. nov.—Planta 5-6 dm. "T caulibus erectis, infra simplicibus modo summo ramos breves floriferos gerentibus; foliis ca. 12 geminis, breve petiolatis, saepe in axibus gemmas gerentibus; foliis inferioribus late ovatis, medio-nervo plerumque prominenti, reticulatis, minusquam .5 em. longis, 3—.4 em. latis, abrupte Ead foliis mediis caulinis 1-1.2 cm. longis, .7—.8 em. latis, ovatis, acuminatis; foliis superioribus usque ad 1.5 cm. longis, lanceolatis; inflorescentia et floribus multiflorae similibus sed calcaribus tenuioribus, parvulum incurvatis divaricatisque.—MExico: Tepic, Jan. 5-Feb. 6, 1892, Palmer (US TYPE). [Vor. 20 146 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: known only from type locality. Specimens examine GUATEMALA: “Alta Verapes, Fichtenwülder bei San Joaquin," alt. 1000 m., Dec. 1907, von T'uerckheim 2041 (F, DH, G, NY, US, V). From an examination of the types of Halenia brevicornis, H. parviflora, H. multiflora, and Exadenus paucifolius, etc., it ap- pears that this group presents a complex, all members of which show variation in habit and spurs, which can not be considered specific differences but differences of degree. Careful perusal of the specimens available discloses the fact that nearly every locality produces some variation in the species. Thus, Chihuahua gives rise to a definitely spurred form, while Guatemalan material presents an entirely different aspect habitally. Ecological ex- periments, as well as field work, would doubtless prove highly valuable as a supplement to the taxonomic treatment of this complex. At present there is no distinction which warrants the retention of these as distinct entities. To draw a line between these forms mentioned and treat them specifically would com- plicate further an already confused situation. Therefore it has seemed advisable to draw attention to these differences, or variations from the specific form, by relegating them to the status of variety of form, with the note that this is merely an arbitrary disposal, that intergradation is existent and that en- vironment is in a large measure responsible for the variation found within the species. KEY TO THE SOUTH AMERICAN SPECIES 1. Leaves always thin, herbaceous; stem and branches more or less slender. 2. Plant slender; flowers in enthenls less than 1 em. long 3. Stem ueuadly branched above, sparingly leafy; baal rosette wee boxed Spy ea a vidi Verc D XO Eier RrrDDCR Ce - breskoornte 3. Stem not branched above, leafy; leaves adpressed; a. ern DUSenb.s o ioo direc Ra der yon DC rs aub DAS 3. H. adpressa 2. Plant coarse; flowers in anthesis usually more than 1 em. long. 3. Nodes 8 or less; leaves not apiculate...................... . H. macrantha 3. Nodes 12 and more; leaves apiculate..................... 5. H. Karstenii 1. Leaves more or less fleshy; stem and branches coars 2. Upper leaves of inflorescence subinvolucrate and buds rs — H 2. Upper leaves of inflorescence not subinvolucrate. 3. Rosette usually present; stem-leaves linear-lanceolate. 4. Plant usually less than 30 em. high; leaves not more than 5 em. long. 5. Calyx-lobes minutely hirtellous. 1933] ALLEN—THE GENUS HALENIA 147 6. Plant usually less than 20 cm. high, or if more than 20 em. high, H. owers in axils of all caulineleaves................ 7. H. inaequalis 6. Plant 10-30 em. high; flowers in axils of upper leaves only.. GREG RS 6b UE POTETE sys 5 CRORE TREE 8. H. . viridis 5. Calyx-lobes not hirtellous. 6. Stem-leaves less than 3 pairs, linear................ 9. H. Schultzei 6. Stem-leaves more than 3 ud lanceolate to HART ERR 7. Cauline leaves less than 2 em. long............ 10. H. gentianoides 7. Cauline leaves more than 2. Ca, ONE. ees 11. H. stellarioides 4. Plant usually more than 30 cm. high; leaves more than 5 cm. long. 5. Inflorescence verticillate; leaves ternate............ 12. H. verticillata 5. Inflorescence not verticillate; leaves not ternate. Nodes more gn Dialed ate i NOE ME 18. H. foliosa 6. Nodes less than 1 7. Root-stalk aia conspicuously larger than stem. 8. Flowers more than 1.5 cm. long, 8-10 mm. broad........ BR ss datdxpe we Pas uA V aaa eT ae 14. H. dasyantha 8. Flowers less than 1.3 cm. long, 5-7 mm. broad...... 16. H. elata 7. Root-stalk not swollen, scarcely larger than stem 8. Plant less than 30 em. high.................... 16. H. Tolimae 8. Plant more than 30 cm. high 9. porcine elongate, spicate; niim > con- em ER TE RET TEE E ee . H. hygrophila 9. Inforseec spreading; calyx-venation Be bua eT D 18. H. parallela 3. Rosette absent; pron es ovate to ovate-lanceolate........ 19. H. major 2. Halenia brevicornis (HBK.) G. Don, Gen. Hist. 4: 177. 1838 For synonymy and description, see North American species, page 140 of this work. Specimens inge. SOUTH AME VIRA Trujillo & Mérida, alt. 1300-4900 m., 1842, Linden 456 (DH). CoroMnzia: Bogota, Dec. 19, 1853, Holton 19 (464) (NY); Dept. Cundinamarca, - moist grassy loam, southwest of Sibate, alt. 2600-2800 m., Oct. 13-15, 1917, Pennell 2434 (M, NY, US); Dept. Antioquia, La Sierra, Medellin, alt. 2000 m., Jan. 4, 1931, Archer 1348 (M); Dept. El Cauca, open banks near Rio Cauca, Coconuco to Popayan, alt. 2000-2500 m., June 18, 1922, Killip 6889 (US); field, near Rio San Andreas, Calaguala, Coconuco, alt. 2500-2800 m., June 18, 1922, Pennell 7154 (ANSP, G, NY, US); field, north of Coconuco, alt. 2300-2500 m., June 11, 1922, Pennell & Killip 6480 (ANSP). Ecvapon: near Quito, coll. of 1859, Jameson (DH, G, NY, UC, V). Peru: steep grassy slopes, Mito, alt. 3000 m., April 8-18, 1923, Macbride 3431 (F, M); 1840, Mathews 3133 (DH Attention should be Salted to the fact that the Swertia brevi- cornis of Humboldt, Bonpland & Kunth, which Gilg has relegated [Vor. 20 148 ANNALS OF THE MISSOURI BOTANICAL GARDEN to synonymy, is a very short much-branched specimen, differing markedly in habit from the S. parviflora HBK. A close examina- tion of the type of the former reveals the fact that the terminal branch has been broken off. Subsequent lateral offshoots give an entirely different appearance habitally. ‘This situation is apparent also in the specimens collected by Oersted in Central America, which Gilg in 1915 labeled H. parviflora. 3. H. adpressa*? Allen, n. sp. Perennial, .5-2.5 dm. high; root coarse; stem usually solitary, erect, slender; basal leaves in dense rosette, oblanceolate, 1-1.5 em. long, .3 em. broad; cauline leaves 5-6 pairs, sessile, lanceolate, 1-1.5 em. long, .2-.3 em. broad, acute; inflorescence a several- flowered cyme, pedicels 2 cm. long; calyx-lobes lanceolate, .4—.6 em. long, .15 em. broad, obsoletely 3-nerved, midvein prominent; corolla .7—.9 em. long, yellow, tube slightly less than one-half the length of the entire corolla; lobes ovate, auriculate, papillate, apiculate; spurs thick or slender, pendulous, more or less diver- gent, giving the flower a triangular appearance just preceding anthesis; stamens approximately .4 em. long; filaments linear, anthers oval; capsule lanceolate; seeds ovoid. Distribution: known only from Colombia. Specimens examine CoroMBia: Dept. Bentesder, Páramo de las Vegas, alt. 3700-3800 m., Dec. 20—21, 1926, Killip & Smith 15679 (M cian NY, US). Species very similar to H. brevicorn 4. H. macrantha Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p. 105. 1916 Perennial herb, 5-6 cm. high, with root curved-erect, ligneous, 49H. adpressa Allen, sp. nov.—Herba perennis, usque ad 2.5 dm. alta; radice crassa; caule plerumque solitario, erecto, tenui; foliis basalibus in rosula densa, oblanceolatis, 1.-1.5 cm. longis, .3 cm. latis; foliis caulinis 5-6 geminis, sessilibus, lanceolatis, 1-1.5 cm. longis, .2-.3 cm. latis, acutis; inflorescentia multiflorifera cyma, pedicellis 2 cm. longis; calycis lobis lanceolatis, .4—.6 cm. longis, .15 cm. latis, ob- solete 3-nerviis, medio-nervo prominenti; corolla .7—.9 em. longa, flava, tubo 14 totae corollae longitudini parum subaequanti; corollae lobis ovatis, auriculatis, papil- latis, apiculatis; calcaribus crassis vel tenuibus, gibbis, pendulis, plus minusve divaricatis; staminibus ca. .4 cm. longis; filamentis linearibus, antheris ovalibus; capsula lanceolata; seminibus ovoideis.—CoroMnBiA: Dept. Santander, Páramo de las Vegas, alt. 3700-3800 m., Dec. 20-21, 1926, Killip & Smith 15679 (M TYPE, NY, US). 1933] ALLEN—THE GENUS HALENIA 149 rather elongate, covered with the blackish remains of marcescent leaves; flowering stem single, erect, internodes 4-5 cm. long, with no rosette leaves before anthesis; leaves herbaceous, acute, 3-nerved veins sunken above, prominent below; lower cauline leaves oblanceolate, gradually narrowed into a long but broad petiole, 6-7 cm. long, .5-1.5 cm. broad; upper leaves lanceolate to ovate-lanceolate, broadly sessile, about 3 cm. long; inflores- cence an apical 3-flowered cyme, solitary in axils of the upper- most euphylloid leaves; pedicels of apical flowers 3 cm. long, those of laterals 2 cm.; Locos oblanceolate, 1-1.3 cm. long, 3-nerved, acute; Boro about 1.5 em. long, green, tube about one-third the length of the entire corolla; lobes ovate to broadly ovate, very acute, somewhat apiculate; spurs large, globose, conical, calluses about .3 cm. long at the base of the tube. Distribution: Colombia No specimens exicoloed, but description compiled from original publication. (TYPE, Kalbreyer 702, BG). 5. H. Karstenii Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, 105. 1916. Biennial or perennial herb up to 5 dm. high; root branched; stems simple, covered with leaf-bases for 7-8 cm. at intervals of less than .4 cm., internodes up to 4—5 cm. long; leaves in pairs at the nodes, sessile, lanceolate, narrowed at base, 2-5.5 cm. long, up to .9 em. broad, 3-nerved, veins excurrent, forming a mucro; inflorescence 4—12 flowers in terminal and axillary cymes, pedicels erect, up to 4 cm. long; calyx-lobes lanceolate, .6—.8 cm. long, .2-.25 em. broad, acuminate, 3-nerved, papillate; corolla about 1 em. long, tube not quite one-half the length of the entire corolla; lobes ovate, subrotund, erose at apex; spurs midway up the corolla tube, small, scarcely noticeable, glandular convex depressions in the corolla; stamens about .3 cm. long, at the ori- fice of the tube; filaments linear, anthers oval, acute; capsule 15-18 cm. long, ovate; seeds ovoid, light brown, wrinkled. Distribution: páramos of Bogota. ned: Specimens examine CoLoMBia: Páramo de Bogota, Karsten (V TYPE). 6. H. subinvolucrata Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p. 99. 1916. [Vor. 20 150 ANNALS OF THE MISSOURI BOTANICAL GARDEN Perennial herb, about 2 dm. high; root multifibrous; subter- ranean stem vertical, short, thick, flowering stem erect, loosely leafy; leaves thickly herbaceous or somewhat fleshy, 3-nerved, veins sub-parallel, deeply sunken above, prominent below; basal leaves linear-lanceolate, gradually narrowed at the base into narrow petiole, 4-5 cm. long, .4-.5 em. broad; lower cauline leaves similar to basal leaves; upper cauline leaves broadly sessile, lanceolate or lanceolate-ovate to ovate, 2-4 cm. long, .6-1 em. broad; inflorescence terminal or axillary, almost sessile 3-flowered cymes, peduncles .3—.6 em. long; floral leaves sheathing inflorescence and longer than flowers; calyx-lobes oblong-ovate, about 1 em. long, up to .3 em. broad, 5-nerved, minutely hirtellous, acute; corolla about 1 em. long, tube about one-third or less the length of the entire corolla; lobes oblong-ovate, .7 cm. long, .3—.35 em. broad, subrotund, erose at apex; spurs scarcely prom- inent blackish maculations at the base of the tube; stamens about .2 cm. long; anthers ovate-oblong; filaments linear. Distribution: Venezuela. Specimens examined VENEZUELA: high mountains of Trujillo and Mérida, alt. 1300-4900 m., 1842, Linden 437 (DH TYPE). 7. H. inaequalis Wedd. Chlor. And. 2: 78. pl. 53 C. 1859; Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p. 99. 1916. Herb 1.5-2(-3) dm. high; stem aditafy. erect, loosely leafy, with internodes about 4-5 cm. long, and inflorescence comprising the upper third of the stem, or stem short, leafy, with internodes equidistant, 1-2 em. long, arising in axil of every leaf; radical leaves petiolate, lanceolate to oblanceolate, 2-3 em. long; upper cauline leaves sessile, lanceolate to elliptic, minutely hirtellous; flowers disposed in small loose pedunculate cymules; peduncles erect or slightly nodding, 1-2 em. long; calyx-lobes lanceolate or oblong-lanceolate, acute, margin minutely hirtellous; corolla scarcely 1 cm. long, exceeding the calyx by one-fourth its length; lobes ovate, acute; spurs inconspicuous. Distribution: known nir from Venezuela. Specimens examine VENEZUELA: "PN de Timotes, Mérida, alt. 3800 m., Sept. 4, 1921, Jahn 558 (HP, US); same locality, alt. 3600 m., Jan. 21, 1922, Jue 839 (HP, US); Mérida, alt. 3000 m., 1846, Funck & Schlim 901 (DH TIN). 1933] ALLEN—THE GENUS HALENIA 151 Funck & Schlim 1148 is cited by Weddell in the original publication of the species, but Gilg has placed this number under H. viridis. 8. H. viridis E Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p. 100. 1916. erba viridis Griseb. in Linnaea 22: 43. 1849; Wedd. Chlor. And. 2:62. 1859. H. inaequalis Wedd. Chlor. And. 2: 78. 1859; Gilg, l. c. Perennial herb, 1-3 dm. high; stem solitary, erect, thick, simple, internodes 4—5 cm. long; basal leaves in rosette, more or less coriaceous, sessile, lanceolate, subequal, up to 6 cm. long, .4 cm. broad, minutely hirtellous, 3-nerved; cauline leaves 3-4 pairs, approximately 5 cm. long; inflorescence a narrow racemi- form cyme arising in the middle of the stem, with simple 3-5- flowered cymules; pedicels erect before anthesis, cernuous after, about 2.5 cm. or less long; calyx-lobes oblong-lanceolate, minutely hirtellous, about .4-.5 cm. long, nerved; corolla about 1 cm. long, greenish, tube about one-third the length of the entire co- rolla; lobes elliptic-lanceolate, rather obtuse, somewhat erose at apex; spurs inconspicuous; stamens .5 cm. long; anthers ovate; capsule oblong-lanceolate. Distribution: known wy from Venezuela. Specimens examine VENEZUELA: Sierra Nevada: Mérida, alt. 3300 m., 1846, Funck & Schlim 1148 (DH TYPE); Laguna Mucuy, Cabeceras del Saisay, Mérida, 4200-4300 m., April 19, 1930, Gehriger 92a (HP). 9. H. Schultzei* Gilg, n. sp. Caespitose mat-like perennial, about 3 dm. high; root tough, fibrous; stems one or more, erect, simple, slender, internodes 5—7 50H. Schultzei Gilg, sp. nov.—Herba perennis, caespitosa, ca. 3 dm. alta; radice lenta, fibrata; caulibus 1—multis, simplicibus, tenuibus, internodiis 5-7 cm. folis basalibus in rosula densa, in petiolis longis tenuibus attenuatis, ad basin dilatatis, lineari-lanceolatis, ca. 5 cm. longis, .5 cm. latis, nerviis, acutis; foliis caulinis plerumque 2 geminis, sessilibus, linearibus, plerumque ca. 1.5 cm. longis, nerviis, acutis; inflorescentia pauco-florifera cymis terminalibus lateralibusque, pedicellis tenuibus, 1-1.5 cm. longis, ad apicem recurvatis; calycis lobis lanceolatis, ca longis, acutis; corolla latissime campanulata, ca. 1.3 cm. longa, flava; corollae lobis ovatis, obtusis marginibus erosis vel crispis; calcaribus inconspicuis.—(TYPE, Schultze 1304, BG). [Vor. 20 152 ANNALS OF THE MISSOURI BOTANICAL GARDEN em. long; basal leaves in dense rosette, narrowed into long slender petioles, dilated at base, linear-lanceolate, about 5 cm. long, .5 em. broad, nerved, acute; stem-leaves usually 2 pairs, sessile, linear, usually about 1.5 em. long, nerved, acute; inflores- cence few-flowered in terminal and lateral cymes, pedicels slender, 1-1.5 em. long, recurved at tip; calyx-lobes lanceolate, about .6 em. long, acute; corolla very broadly campanulate, about 1.3 cm. long, yellow; corolla-lobes ovate, rather obtuse, erose or crisped margins; spurs not discernible. Distribution: Colombia ecimens examined, but description compiled from photograph. (TYPE, Schultze 1304, BG). 10. H. gentianoides Wedd. Chlor. And. 2: 78. pl. 53B. 1859. Perennial herb, up to 3 dm. high; root more or less thick, ligneous; stems sterile or flowering, long, loosely leafy; basal leaves in a dense rosette, petiolate; cauline leaves 2-6 pairs, subsessile, lanceolate, acute, 3-nerved; inflorescence lateral or terminal, small racemiform cymes, peduncles more or less resu- pinate; calyx-lobes oblong, up to .7 cm. long, papillate, acuminate, prominently 3-nerved; corolla about 1 cm. long, tube approxi- mately one-third the length of the entire corolla; lobes ovate, rotund, crenulate-erose; spurs reduced to inconspicuous glandular areas not visible to the naked eye; stamens approximately .4 cm. long; filaments linear, anthers ovate, acuminate; capsule lanceo- late. Distribution: páramos of Colombia. p mens examined: Corownia: Páramo de Bogota, Karsten (V); same locality, Goudot (BG TYPE, M photo, V); same locality, T'riana 1964 (DH, V); same locality, Jan. 17, 1854, Holton 467 (DH, G); Guasca, 1919, Ariste-Joseph A423 (US). 11. H. stellarioides Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p. 100. 1916. Perennial? herb, up to 3 dm. high; flowering stem erect, loosely leafy, internodes 4-7 cm. long; basal leaves lacking; cauline leaves sessile, lanceolate to linear-lanceolate, 2.5-3.5 cm. long, gradually decreasing toward the summit, .4—.5 cm. broad, acute, 3-nerved; inflorescence lateral or terminal, 5-7-flowered cymes, pedicels more or less erect, up to 1.5 em. long; calyx-lobes lance- 1933] ALLEN—THE GENUS HALENIA 153 olate or oblong-lanceolate, about .8 em. long, .25 cm. broad, acute, 3-nerved, or nerves obsolete; corolla about 1 cm. long, tube about one-third the length of the entire corolla; lobes ovate- oblong, upper margin subcrenulate, acute; spurs semi-globose callous prominences at the base of the tube. Distribution: páramos of Colombia. No specimens examined, but description compiled from original publication and photograph. (TYPE, Lehmann 3080, BG). 12. H. verticillata Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p.103. 1916. Annual herb, slender, up to 1 m. high; root fibrous; stem .5 cm. thick, becoming black, covered with the remains of early leaves, densely leafy at the middle or just below the middle of the stem, in- ternodes up to 8 em. long; basal leaves in dense whorls for a varying distance up the stem, connate, sessile, linear-lanceolate, 10 cm. or less long, .8 cm. broad, acuminate, 5-nerved fleshy; cauline leaves in 7-8 whorls of 3, lanceolate (extreme upper subt ovate-lanceolate), 5-nerved, 5-6 cm. long, gradually decreasing in length toward the summit, with corresponding increase in width; flowers numerous, disposed in 3-5 verticillate, axillary and terminal cymose clusters; pedicels erect, up to 3.5 cm. long; calyx-lobes oblong-lanceolate, papillate on under surface of veins and entire upper tip, up to 1.0 em. long, .2-.35 cm. broad, acu- minate to acute, 5-nerved, veins parallel, becoming confluent at tip; corolla apple-green or yellowish-green, up to 1.7 cm. long, tube about one-third the length of the entire corolla; lobes broadly ovate, subrotund, erose and papillate at tip; spurs sub- globose glandular prominences at the base of the corolla. Distribution: wet rS about voleanos, Colombia. Specimens examine CoroMnia: ‘“‘Cauca am Vulkan," Sotará, 3500 m., Lehmann 6190 (BG TYPE, F); Dept. of El Cauca, Mt. Pan de Azucar, alt. 3500-3700 m., June 16, 1922, Pennell 7052 (ANSP, G, NY, S, US). 13. H. foliosa Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p. 101. 1916. Biennial herb (?), up to 8 dm. high; root stout (?); flowering stems l-many, erect, simple, few or no leaves at base, lower internodes 1-1.5 cm. long, upper 5-7 cm. long; numerous pairs [Vor. 20 154 ANNALS OF THE MISSOURI BOTANICAL GARDEN of lower cauline leaves, thickly herbaceous, gradually narrowed toward the base but dilated again at base, lanceolate, 3-5 cm. long, about 1 em. broad, acute, 3-5 nerved; upper cauline leaves herbaceous, gradually narrowed toward the base and broadly sessile, ovate-elliptie, 3-5 em. long, about 1 em. broad, acute to very acute, 5-nerved, veins parallel, sunken above, prominent below; terminal and axillary, loosely arranged, many-flowered elongate cymes, comprising a thyrsoid inflorescence 10 cm. long, pedicels about 2.5 em. long; small upper leaves shorter than the inflorescence; calyx-lobes ovate-oblong, .7—.8 cm. long, .3 em. broad, acute or very acute, 3-nerved; corolla about 1.3 cm. long, tube about one-fourth the length of the entire corolla; spurs semi-globose protuberances at the base of the tube. Distribution: plrunos of Colómbia. Specimens examine CoroMnia: Dept. of ‘Bolivar-Antioquia, Páramo de Chaquiro, alt. 3000-3200 m., Feb. 23, 1918, Pennell 4268 (NY, US). (TYPE not seen, Stübel 276, BG, M photo). 14. H. dasyantha Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p.103. 1916. Perennial herb, (1.5—-)3—7 dm. high; root thick ;stem erect, simple, more than .5 em. thick, fleshy, brown in dried specimens, faintly striate, covered with remains of marcescent leaves, 2—5 internodes up to 10 em. long; basal leaves in a dense rosette, broadly elliptic, up to 10 em. long, nearly 2 cm. broad, acute, 5-nerved; 3-5 pairs of cauline leaves, sessile, dilated at in| oblong-lanceolate, 4—10 em. long, their length gradually decreasing toward the summit, 1-2 em. broad, acute, 5-nerved; inflorescence axillary and ter- minal, many-flowered racemose cymes, usually 10 em. long; pedicels erect, for most part up to 3.5 cm. long; calyx-lobes ovate to ovate-lanceolate, papillate, up to 1 cm. long, .4-.5 cm. broad, acute to abruptly acuminate, 3-nerved; squamellae frequently scale-like lobed bodies; corolla 1.5-2 cm. long, pale greenish- yellow, tube nearly equal in length to the entire corolla; lobes ovate, subrotund, erose at apex, somewhat papillate; spurs small subglobose prominences at the base of the corolla-lobes, almost obscured by the calyx; stamens approximately .4 cm. long, attached at the summit of the tube; filaments linear, anthers ovate-oval; stigma reflexed; capsule up to 2 cm. long; seeds elliptical, reticulate. 1933] ALLEN—THE GENUS HALENIA 155 Distribution: moist grassy páramos, or dry open woods, Colombia. Specimens examined: CoroMBia: Dept. of Caldas, Páramo del Quindio, alt. 3700-4200 m., Aug. 15-20, 1922, Pennell & Hazen 9997 (ANSP, NY, US); Dept. of Tolima, Páramo de Ruiz, alt. 3500-3800 m., Dec. 16-17, 1917, Pennell 3001 (NY, M, US); Dept. of Cauca, Páramo de Buena Vista, Huila group, Central Cordillera, alt. 3000-3600 m., Jan 1906, Pittier 1111 (US). (TYPE not seen, Lehmann 2065, BG, M photo.). This species is very similar to H. elata but is, on the whole, a larger and coarser plant. Since the material is scanty and the geographical location different, H. dasyantha has for the present been maintained as a distinct species. 15. H. elata Wedd. Chlor. And. 2:78. 1859. Perennial, up to 5 dm. high; stem thick, coarse, erect, loosely leafy; numerous basal leaves in dense rosette, broadly lanceolate, attenuate into petiole, dilated at base, 7-10 cm. long, approxi- mately 1 cm. broad, 5-nerved, acuminate; cauline leaves more or less sessile, shorter, lanceolate, acute; inflorescence numerous loose terminal and lateral cymes, peduncles up to 3 cm. long; calyx-lobes oblong-lanceolate to ovate-lanceolate, .7—.8 cm. long, 3-nerved, papillate, acute; corolla hardly one-fourth more than the length of the calyx, tube slightly more than one-half the length of the entire corolla; lobes ovate, obtuse; spurs inconspicu- ous tubercules at the extreme base of the corolla; stamens .6 em. long; anthers narrowly ovate, filaments linear. Distribution: in ie vas Nevada of Santa Marta, Venezuela. Specimens exami VENEZUELA: MN Nevada de Santa Marta, Caracas, 1844, Funck 4156 (DH TYPE). 16. H. Tolimae Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p. 101. 1916. Perennial herb, 2.5-3 dm. high; root thick; root-stalk thick, densely covered with remains of marcescent leaves; one to few flowering stems, thick, erect, simple for the most part, very narrowly winged, internodes 3-6 cm. long; basal leaves arranged in rosette, slightly narrowed into long broad petioles, dilated at the base, lanceolate, 5-8 cm. long, up to .5-.6 em. broad, 3-5- nerved, acute; stem-leaves 2-3 at a node, sessile, broadly lanceo- late, 4-6 cm. long, .6—7 cm. broad, their size gradually de- creasing toward the summit, 3-5 subparallel veins, sunken above, [Vor. 20 156 ANNALS OF THE MISSOURI BOTANICAL GARDEN prominent below, reticulate; inflorescence few-flowered (5-7) cymes, axillary and terminal; pedicels more or less resupinate, up to 3 em. long, the central one longer; calyx-lobes ovate to ovate- oblong, papillate, .7—.9 cm. long, .25—.35 em. broad, subacuminate, 3-nerved; corolla up to 1.3 em. long, light greenish-yellow, tube slightly less than one-third the length of the entire corolla; lobes ovate, erose and papillate at the tip; spurs glandular subglobose protuberances at the base of the tube, obscured by the calyx; stamens approximately .4 cm. long; filaments linear, anthers ovate; capsule lanceolate, apiculate, 2 cm. long; seed ovoid- elliptical, very minutely reticulate, pale tan. Distribution: grassy páramos of Colombia. Specimens examined: CoroMnBia: Dept. of Caldas, Páramo del Quindio, alt. 4100-4400 m., Aug. 15-20, 1922, Pennell & Hazen 9841 (ANSP, G, NY, US); (typ not seen, Stübel 228, BG, M photo.); bare loam slopes below snow, same locality and date, alt. 4300-4500 m., Pennell & Hazen 9894 (ANSP, NY, S, US). This last-cited specimen has leaves broader than those of the type, but is similar otherwise. H. Tolimae appears to be closely related to H. elata. 17. H. hygrophila Gilg in Engl. Bot. Jahrb. 54: Beibl. 118, p.102. 1916. Biennial herb, about 6 dm. high; root ligneous; subterranean root-stalk covered with darkened leaf-bases; flowering stem fleshy, erect, simple, 1 cm. thick, faintly winged, internodes up to 8 em. long; 30-40 basal leaves in a dense rosette, linear, up to 12 em. long, .5-.9 em. broad, acute, 5-nerved; cauline leaves yellow-green, 3-4 pairs, the lower lanceolate, the upper sub- tending the inflorescences, ovate-lanceolate, sessile, 5-7 «cm. long, gradually decreasing in size toward the summit, approxi- mately .8-1.2 em. broad, 7-nerved, parallel veins confluent at tip; inflorescence terminal and axillary in upper stem-leaves, in dense many-flowered cymes, giving spicate appearance, pedicels erect, up to 3 cm. long; calyx-lobes ovate-oblong to ovate, papillate, .9-1.0 cm. long, narrowly acute, 5-nerved, reticulate; corolla up to 1.5 em. long, yellowish-green, tube approximately one-third the length of the entire corolla; lobes ovate, erose at apex; spurs angular sac-like prominences at the extreme base of the corolla lobes, obscured by calyx; stamens approximately .7 em. long, attached at the orifice of the tube; filaments linear, 1933] ALLEN—THE GENUS HALENIA 157 anthers broadly oblong, apiculate; capsule lanceolate; seeds elliptic, pale tan, reticulate. etnias páramos of Andes, Colombia. Speci d: diens : Páramo of Guanacas, Central Andes of Popayan, 3000-3600 m., Lehmann 7990 (F, à. BG TYPE, M photo, US) 18. H. parallela* Allen, n. sp. Perennial herb, up to 6 dm. high; root-stalk thick; flowering stem single, erect, simple, or branched above, striate, lower inter- nodes 1-1.5 em. long, upper 4-6 cm. long; numerous basal leaves in rosette, almost sheathing, lanceolate, 3-5-nerved, about 9 em. long, 1 em. broad; cauline leaves about twelve pairs, more remote toward the apex, sessile, lanceolate to elliptic, 4-6 cm. long, gradually decreasing toward the summit, approximately 1 cm. broad, 3-5 subparallel veins sunken above, prominent below; inflorescence 1 or more many-flowered axillary and terminal cymes, pedicels more or less erect, up to 3.5 cm. long, the lateral somewhat shorter than the terminal; calyx-lobes ovate, papillate, .7-.9 cm. long, .4 em. broad, attenuately acute, prominently 5—7-nerved, veins subparallel; corolla about 1.3 em. long, probably greenish?, tube about one-fourth the length of the entire corolla; lobes broadly ovate, slightly papillate, crisped toward the tip, abruptly acuminate; spurs large globose protuberances at the base of the tube, obscured by calyx; stamens nearly .5 cm. long, attached at the orifice of the tube; filaments linear, anthers oblong; capsule lanceolate. 5 H. parallela Allen, sp. nov.—Herba perennis, usque ad 6 dm. alta; radice crassa; aule florifero solitario, erecto, simplice vel supra ramoso, striato; internodiis enn 1-1.5 em. longis, superioribus yet: m. longis; foliis iuo multis, in rosula densa, fere Serin bur, lanceolatis, P vi iis, ca. 9 em. longis, 1 cm. latis; foliis caulinis ca. duodecem geminis, ad apicem remotioribus, seii. lanceo- latis vel ellipticis, 4-6 cm. longis, sensim sursum decrescentibus, ca. 1 cm. latis, 3-5-nerviis, nervis subparallelis, supra immersis, infra prae infor- escentia cymis 1- I apri axillaribus et terminalibus; pedicellis plus minusve erectis, usque ad 3.5 cm. longis, lateralibus terminalibus brevioribus; calycis lobis ovatis, popillats, 4-. 9 cm. longis, .4 cm. latis, attenuate acutis, pro- minente 5-7-nerviis, nervis subparallelis; corolla ca. 1.3 cm. longa, forte viridula, tubo ca. 14 totae corollae longitudini adaequanti; lobis late ovatis, parum papillatis, ad apicem crispis abrupte acuminatis; calcaribus magnis globosis gibbis ad basin tubi calyce obscuratis; staminibus ca. .5 cm. longis, tubi summo adjunctis; filamentis linearibus, antheris oblongis; capsula lanceolata.—V ENEZUELA: Páramo de La Negra, Mérida, Dec. 1927, Gutzwiller 32 (HP TYPE, US) [Vor. 20 158 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: Venezuela. Specimens examin VENEZUELA: Páramo de La Negra, Mérida, Dec. 1927, Gutzwiller 32 (HP TYPE, US). Species near H. foliosa and H. hygrophila. 19. H. major Wedd. Chlor. And. 2: 79. 1859. Annual probably, up to 6 (?) dm. high; (root not seen); stem erect, simple below, branched above, loosely leafy, very narrowly winged and striate; leaves sessile, subconnate, broadly elliptic, 2-7 em. long, .5-2 em. broad, somewhat abruptly acuminate, 3-5-nerved; inflorescence terminal and axillary 3-6-flowered sub- umbelliform cymes, pedicels 4 cm. long, usually erect, frequently nodding at apex; calyx-lobes oblanceolate to subspatulate, up to .8 em. long, .25 em. wide, 3-nerved, midvein very prominent ; corolla 1.0-1.3 em. long, tube almost one-half the length of the corolla; lobes ovate, erose at apex; spurs small upcurved conical protrusions almost at the orifice of the tube; stamens approxi- mately .5 em. long, at the orifice of the tube; filaments linear; anthers not seen; capsule ovate, 1.3-1.6 cm. long; seeds oval, brown-black, wrinkled. Distribution: shrub zone, mountain bases, Colombia. d: BIA: Dept. of El Cauca, Mt. Pan de Azucar, alt. 3300-3600 m., June 16, 1922, Pennell 7084 (ANSP, NY, US); Pag of Cundinamarca, Sibate, alt. 2700-2800 m., Oct. 13-15, 1917, Pennell 2438 (M, US); Dept. of Caldas, Cerro Tatama, alt. 3400-3700 m., Sept. 8-10, 1922, slg 10575 (US). (TYPE not seen, Goudot, HJP). SECTION 2. HALENIASTRUM KEY TO NORTH AMERICAN SPECIES AND VARIETIES l. m spreading to ascending. ant less than 2.5 dm. high. 3. Biennial; flowers white; distribution Mexico............ 20. H. crassiuscula 3. Annual; flowers purple; distribution Canada. 23a. H. deflexa var. Brentoniana 2. Plant more than 2.5 dm. high. 3. Spurs, if present, about !4 the length of the corolla......... 21. H. Pringlet 3. Spurs less than 1% the length of the corolla. 4. Annuals; leaves mostly cauline. 5. Stem-leaves linear. ............ ccc cece cc ccc eeeecees 22. H. recurva 5. Stem-leaves lanceolate to ovate..........00... 0000 cece 23. H. deflexa 4. Perennials; leaves mostly radical. . Strict, many-flowered, spike-like inflorescence. D DEM RE Lieber eres q9EXECODCKERK ER" wens 24. H. rhyacophila 6. Stem procumbent.............. 24a. H. rhyacophila var. procumbens 1933] ALLEN—THE GENUS HALENIA 159 5. Loose, broad, few-flowered inflorescence............... lesus. YogeAet "PTS REPE A 24b. H. rhyacophila var. macropoda 1. he pendulous to incurved. . Leaves not apiculate or very rarely so; ender not mucronulate. 3. Basal rosette absent; leaves mostly cau 4. Habit erect; leaves less than 3 cm. (he D.C NE I oo ees ses ec tee ee CES 25. H. Palmeri 5. Leaves lanceolate or ovate. 6. Calyx-lobes mostly obtuse, appressed........... 26. H. Conzattit 6. Calyx-lobes acute, reflexed.............. suus. 27. H. Schiedeana 4. Habit prostrate; leaves 5-12 cm. long.................. 28. H. caleoides 3. Basal leaves present, cauline few o 4. Spurs incurved, 14 or less than 4 the linen of the corolla. 5. Sterile branches present; leaves numerous. ......... 29. H. platyphylla 5. Sterile branches absent; leaves less than 15. 6. Flowers 1.5 cm. or less lon 7. Flowers less than 1 em. long; spurs rudimentary. .30. H. nudicaulis 7. Flowers more than 1 cm. long; spurs 14-14 the length of the oo ee ik os oe haa a dod ae ee 31. H. plantaginea 6. Flowers more than 1.5 cm. long...31a. H. plantaginea f. grandiflora 4. Spurs spreading, about l4 the length of the corolla. 5. Stems always erect. Gea ves lineaments coos Se ie lene oU 32. H. Shannonii 0. Leaves elipüeal...........Leeees. 82a. H. Shannonii f. compacta 5. Stems more or less deceumbent.................ss. 33. H. decumbens 2. Leaves conspicuously apiculate; calyx-segments mucronulate. 3. Plant more than 2 dm. hi 5 Ou Bul ee Oe eed OE . H. guatemalensis 3. Plant less than 2 dm. high........... 34a. H. guatemalensis var. latifolia 20. H. crassiuscula Robinson & Seaton in Proc. Am. Acad. 28:113. 1899. Small caespitose biennial of dense habit, slightly fleshy; stems erect, 0.4-1.0 dm. high, narrowly winged, much branched; radical leaves broadly oblanceolate to elliptic, 2 cm. long, obtuse, attenuate into long petiole nearly equalling the blade, 3-nerved; eauline leaves 1-3 pairs, narrowly oblanceolate to oblong, nar- rowed at the base; inflorescence dense compact umbellate cyme; flowers terminal or axillary, pedicellate, after anthesis slightly nodding, not at all resupinate; calyx-segments lanceolate to oblanceolate, .45-.6 cm. long, obtuse, papillate, 3-nerved; corolla white, up to 1.5 cm. long, about .5 em. broad at base, tube .35-.4 em. long; lobes oblong-elliptic, acute; spurs .4 cm. long, arising slightly below the midpoint of the tube, slender, spreading, and curved upwards; stamens .25 cm. long, anthers ovate; filaments linear; capsule lanceolate, frequently subfalcate, acute, 1.4 cm. long; seeds globose, light yellow-brown, granular. [Vor. 20 160 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: bare alpine summits, Mexico. Specimens examine Mexico: Nevado de Toluca, bare alpine summits, alt. 3500 m., Sept. 2, 1892, Pringle 4229 (ANSP, BG, BM, B, CAS, C, DH, F, G TYm, IAC, K, M, NY, S, US, V); Ixtaccihuatl, wet meadows, alt. 3000-3250 m., March-July 1903, Purpus 318 (CAS, M, US); Popocatepetl, Aug. 7-8, 1901, Rose & Hay 5999 (US). 21. H. Pringlei Robinson & Seaton in Proc. Am. Acad. 28: 113. 1893. Halenia candida Ramirez in Inform. Secret. Foment. Mexico (Excurs. Mont. Ajusco), 34. 1895; Estud. Hist. Nat. 102. 1904. Biennial, of glaucous aspect; root thick, tough, ligneous; stem usually solitary, occasionally caespitose, simple or nearly so, scape-like, slender, erect, 1-2.5 dm. high; leaves less than 3 cm. long, radical leaves elliptic to narrowly oblanceolate, faintly 3- nerved, attenuate below into slender petioles, usually as long as the leaf-blade and persistent; cauline leaves 1-2 pairs, sessile, short, sublinear, 1.5-3 cm. long, about .3 em. broad; flowers disposed in terminal, or occasionally lateral, few-flowered um- belliform cymes, pedicels less than 2 cm. long, usually about .8 em.; calyx-lobes oblong-spatulate, .3—.5 cm. long, acuminate, 3- nerved; corolla white, .8-1.5 cm. long, spurless in the majority of cases; when spurs present, slender, spreading, and curved- assending, 1.5-1.8 em. from tip to tip, with prominent veins and glandular tips; corolla-tube up to .2 cm. long; corolla-lobes elliptic, acuminate; stamens .2-.35 em. long; anthers narrowly ovate, minutely papillate, filaments linear, slightly uncinate; capsule lanceolate, acute, exserted; seeds subglobose, yellow- brown, granular. Distribution: springy meadows of central and south Mexico. Specimens examine Mexico: springy alpina meadows, Sierra de las Cruces, alt. 2450 m., Aug. 28, 1904, Pringle 13121 (BG, C, UC, G, K, US); same locality, Aug. 23, 1892, Pringle 4209 (ANSP, C, CAS, DH, F, G TYm, IAC, K, M, NY, 8, US, V); same locality, June 1895, €—— 908 (US). X1CO: without locality, 1920-21, Reiche 36 (BG). The habit of this species 1s very similar to that of H. nudicaulis. The Pringle specimens cited above were collected in August, and the plants are smaller and grow less luxuriantly than the single specimen collected in June by Altamirano. The specimens collected later in the season very rarely possess spurs, while the 1933] ALLEN—THE GENUS HALENIA 161 earlier plants show a distinctly spurred corolla. This condition is shown in other species to a somewhat less extent, and is in all probability traceable to variation in environmental conditions. 22. H. recurva (Sm.) Allen, n. comb. Swertia recurva Smith in Rees, Cyclopedia 34: sub Swertia. 1819. Halenia Rothrockit Gray in Proc. Am. Acad. 11: 84. 1876; Rothrock, Rept. Wheeler Exped. 195, pl. 21. 1878; Hemsl. Biol. Cent.-Am. Bot. 2: 353. 1882. Tetragonanthus Rothrockii Heller, Cat. N. Am. Pl. 6. 1898, and ed. 2, 16. : Annual, 2.5-5 dm. high; stem simple, often branched above; basal leaves less than 3.5 cm. long, .6 cm. broad, elliptic-lanceolate to spatulate; cauline leaves remote, lance-linear, 1.5—4 cm. long, about .35 em. broad, obscurely 3-nerved, midrib prominent below; inflorescence a loosely flowered subumbellate cyme; flowers on slender pedicels, .5-3 cm. long, often in sevens; calyx lobes lanceolate, elongate-acute, up to .6 cm. long, uninerviate, papillate; corolla bright-yellow, about 1-1.2 cm. long, tube less than one-half the length of the entire corolla; corolla-lobes ovate, subacuminate, delicately veined, papillate; spurs curved, hori- zontal or ascending, up to 1.6 cm. from tip to tip; anthers broadly oblong, mucronate, papillate; filaments slightly obovate; capsule ovate-lanceolate; seeds yellow-brown, subglobose-ovoid, granular. Distribution: southern United States and Mexico. Specimens examined: UNITED STATES: Arizona: Mt. Graham, alt. 2250 m., Sept. 1874, Rothrock 733 (ANSP, F, IAC, M, US); same locality, Aug. 1874, Rothrock (G, NY); Chiricahua Mts., Sept. 22. 1931, Jones 28603 (M); Barfoot Park, Chiricahua Mts., alt. 2000-2050 m., rolling andesitic, recently lumbered pine land, Sept. 8, 1906, Blumer 1859 (BG, D, G, K, ^ US, V); same locality, alt. 2480 m., Sept. 22-23, 1914, Eggleston 10774 (US); Apache Pass, Chiricahua Mts., Sept. 1881, Lemmen & Lain mon (CAS); Hermitage, Chiricahua Mts., Sept. 1881, Lemmon & Lemmon (CAS); Rucker Valley, Chiricahua Mts., B 1881, collector unknown 1874 (CAS, DH, F, SM); White Mts., Aug. 1878, fw (F); summit of White Mts. (Springerville-Fort Apache Road), Apache Reservation, alt. 2270-2880 m., Aug. 29, 1919, Eggleston 15781 (F); Riverside Ranger Station, Greer, Apache Forest, alt. 2700 m., Aug. 24, 1920, Eggleston 17137 (NY, US); grassy flats near Brinkley’s Ranch, White Mts., Aug. 5, 1915, Ellis 20 (US); Riggs Flat, Pinaleno Mts., alt. 2000 m., Sept. 23, 1917, Shreve 5373 (G); Columbia Trail, Pinaleno Mts., alt. 2500 m., Sept. 13, 1914, Shreve 4312 (CAS, US). [Vor. 20 162 ANNALS OF THE MISSOURI BOTANICAL GARDEN New Mexico: Mogollon Mts., on or near the west fork of the Gila River, Socorro Co., alt. 2125 m., Aug. 14, 1903, Metcalfe 501 (M, NY); swampy ground, divide of Mogollon Mts., Sept. 7, 1881, Rusby 264 (ANSP, BM, CAS, F, K, M, NY, US); same locality, Aug. 1881, Rusby (IAC, NY). M . EXICO: Cuinuanvua: Mt. Mohinora, Sept. 1, 1898, Nelson 4868 (G, US); cool slopes, Sierra Madre, alt. 1750-2375 m., Sept. 27, 1888, Pringle 1663 (BG, BM, B, CAS, DH, M, NY, 8, V); same locality, Sept. 24, 1887, Pringle 1329 (ANSP, BG, C, G, K, NY, US); Meadow Valley, Sierra Madre Mts., alt. 1750 m., Sept. 17, 1903, Jones (S); Sierra Madres, near Colonia Garcia, alt. 2000 m., ent, 6, 1899, Townsend & Barber 309 (BB, F, DH, G, M, NY, US); Escalon, Mutis (L). CoanviLa: Sierra Madre, 40 m. south of Saltillo, July 1880, Palmer 839 (ANSP, G, K, US). DunaNao: Barranca, below Sandia Station, alt. 1625 m., Oct. 13, 1905, Pringle 13588 (G, 8, UC, US); Sierra de Candela, alt. 3000 m., Aug. 27, 1903, Endlich 53 ). JALISCO: Sierra de Tequila, alt. 2000 m., July 5, 1893, Pringle 5465 (G). The name Swertia recurva Smith was given to the specimen collected by Mutis and sent to Linnaeus, now preserved in the herbarium of the Linnaean Society of London. The description published in Rees’ ‘Cyclopedia’ was inadequate, and later the specific name recurva was placed under deflexa, the well-known northern species. Since few of the succeeding monographers ever saw the original Swertia recurva Sm., it is not strange that the error persisted. Over fifty years later, Gray described Halenia Rothrockii as a new species. si A S Y ‘Al J5) EN FAN i e s po) ALLEN—THE GENUS HALENIA THE SENSITIVITY OF ORCHID SEEDLINGS TO NUTRITIONAL IONS F. L. WYND Assistant in the Henry Shaw School of Botany of Washington University I. INTRODUCTION The germination and growth of orchid seedlings on artificial media have received considerable attention from plant physiolo- gists. Almost invariably, however, their investigations have been concerned with the carbohydrate metabolism to the ex- clusion of the inorganic nutritional relationships. The purpose of the present paper is to report the results of a study on the sensi- tivity of germinating orchid seeds and young seedlings to different nutritional ions. The few available data on inorganic nutrition are scattered in the papers dealing primarily with other phases of orchid development on artificial media. Burgeff (09) pointed out that orchid seeds germinated more favorably when nitrogen was available as ammonium sulphate than as a nitrate salt. This fact led Knudson (22) to modify Pfeffer's solution, and this is the solution frequently mentioned in orchid literature as ‘‘Knudson’s Solution B." Clement (724, ’24a) has said that the nutrition of different species of Odonto- glossum varies widely, but he did not elucidate the matter further. Ballion and Ballion (’24) made a similar statement, that in their work with Cattleya, Miltonia, and Odontoglossum the nutritional conditions varied according to the species, but here again no specific information was given. La Garde (29) found that his solution L gave about 10 per cent better growth than did Knud- son's solution B, and attributed this superiority to the slight increase of nitrogen and the large increase of potassium and phosphorus in his solution. Smith (32) used Knudson's solution B, but found that doubling or trebling the amount of ammonium sulphate gave better growth and a deeper color to his seedlings. In so far as we have been able to ascertain, the above brief review is complete as to published data concerning the inorganic nutrition of orchid seedlings. ANN. Mo. Bor. GARD., Vor. 20, 1933. (223) [Vor. 20 224 ANNALS OF THE MISSOURI BOTANICAL GARDEN II. METHODS Livingston ('19) has clearly defined four possible criteria for the comparison of different nutrient solutions, as follows:— (1) The volume-molecular proportion of any one salt. (2) The osmotic proportion of any one salt. (3) The total volume-molecular concentration. (4) The total osmotic concentration. As it is impossible to vary all the features of the composition of a series of solutions regularly at the same time, it is necessary to select arbitrarily one characteristic as the basis for comparison. Since it is probable that the solution must act upon plants pri- marily in an osmotic way, we have followed Livingston and many others in using the total osmotie concentration as the basis for comparative studies. The triangular representation of a series of solutions of three components varying in any of the above ways was first introduced in physical chemistry by Schreinemakers (093) and again by Bancroft (02). Schreiner and Skinner (710) first applied this method of investigating nutrition in their work on fertilizers. Since 1910 it has been used by various investigators to study the nutrition of several agriculturally important plants: wheat, by Shive (15), McCall (16), Livingston and Tottingham ('18), Van Alstine (19), Meier and Halstead (21), and by Tottingham and Rankin (22); buckwheat, by Shive (715, '17) and by Shive and Martin (718, ’18a); potato, by Johnston ('24) and by Martin and Shive (720); celery, by Poole and Fant (22); cranberry, by Addoms and Mounce (731); peach seedlings, by Davidson (728). After investigating the work of the above authors, it appeared to us that the triangular system of arranging a varying series of nutrient solutions provided the most satisfactory approach to the problem of the reaction of orchid seedlings to their inorganic substrate. Seeds for the investigation were furnished from the green- houses of the Missouri Botanical Garden through the courtesy of Dr. George T. Moore, the Director. Flowers of Cattleya Trianae Linden and Rchb. f. were pollinated November 17, 1930, and the largest and apparently the best pod was picked May 3, 1932, after a development period of almost eighteen months. It has 1933] WYND—ORCHID SEEDLINGS AND NUTRITIONAL IONS 225 been the author’s experience that in general the capsules first to mature furnish seeds of lower vitality than those maturing later; hence the one requiring the longest time to mature was used to supply seeds for the present work. The nutrient solutions investigated were those of Type I and Type IV, as described by Livingston (19). By the use of only three salts it is possible to arrange solutions containing the six major nutritional ions in six different combinations. Of these six possible combinations, only two will have all their ions added in unlike combinations. The compositions of the solutions used are given in tables 1 and rr. In both types the total osmotic TABLE I TYPE I—KH:;PO:, Ca(NO??, MgSO: Molecular proportion Volume-molecular concentration Culture No. KH;PO, Ca(NOs)2 MgSO* KH;PO* Ca(NO3); MgSO* R1 S1 1 1 6 0027 .0027 0161 S2 1 2 5 0025 .0049 0123 S3 1 3 4 0024 0071 0094 S4 1 4 3 0022 0089 0067 S5 1 5 2 0022 0108 0043 S6 1 6 1 0020 0122 0020 R2 S1 2 1 5 .0053 .0027 0132 S2 2 2 4 .0049 .0049 0099 S3 2 3 3 .0047 .0071 0071 S4 2 4 2 .0045 .0090 0045 S5 2 5 1 .0041 0104 0021 R3 Sl 3 1 4 0076 .0025 0101 S2 3 2 3 0072 . 0048 0072 S3 3 3 2 0068 . 0068 0045 S4 3 4 1 0065 .0086 0021 R4 S1 4 1 3 0099 0025 0074 S2 4 2 0094 .0047 0047 S3 1 90 .0068 0022 R5 $81 5 1 2 0123 .0024 0049 S2 5 2 1 0118 .0047 0023 R6 S1 6 1 1 0145 0024 0024 [Vor. 20 226 ANNALS OF THE MISSOURI BOTANICAL GARDEN TABLE II TYPE IV—K380,, Ca(HiPOQ9s Mg(NO3):* Molecular proportions Volume-molecular proportions Culture xs K;SO, |Ca(H;POj,*|Mg(NOss|| K;SO, | Ca(H;PO,.| Mg(NOs)2 Ri S1 1 1 6 .0018 .0018 .0108 82 1 2 5 .0018 .0036 .0092 S3 1 3 4 .0019 .0056 .0075 R2 S1 2 1 5 0037 0018 0091 82 2 2 4 0037 0037 0074 S3 2 3 3 0037 0056 0056 R3 S1 3 1 4 0056 .0019 0075 82 3 2 3 0056 0037 0056 83 3 3 2 0056 0056 0038 R4 S1 4 1 3 0076 0019 0057 S2 4 2 2 0077 0039 0039 83 4 3 1 0078 0059 0019 R5 Sl 5 1 2 0097 0019 0039 82 5 2 1 0098 0039 0019 R6 $1 6 1 1 .0116 .0019 .0019 * The solutions containing the highest proportion of Ca(H;PO,); could not be pre- pared, because sterilization at the temperature and pH used caused precipitation. concentration was one atmosphere. Mallinckrodt salts of high- est reagent quality were used. One-hundred-ce. portions of the nutrient solutions were placed in 200-cc. Erlenmeyer flasks, 1.75 per cent of Merck's powdered “Reagent” agar added, and steri- lized by autoclaving at 20 Ibs. pressure for 20 minutes. Follow- ing the work of La Garde (’29), who found that maltose was the best source of carbohydrate, a 2 per cent concentration of this sugar was used in all cultures. A study of the hydrolysis of this sugar under our conditions of sterilization showed that about 10 per cent was hydrolyzed to glucose. In addition to the six major nutrient ions there were added one-half part per million of manganese as manganese sulphate, and one-half part per million of boron as sodium borate. A ferric phosphate suspension was prepared as described by Living- 1933] WYND— ORCHID SEEDLINGS AND NUTRITIONAL IONS 227 ston (’19) and 1 cc. added to each liter of nutrient solution. This gave ferric phosphate in the concentration of about 3 mgms. per liter. Owing to the sensitivity of orchids to the acidity of the medium, each solution was so adjusted that the pH after sterilization was 4.8 to 5.0. Tables mr and rv indicate the pH relationships be- fore and after sterilization. The seeds were sterilized by shaking TABLE III THE ACIDITY RELATIONSHIPS OF TYPE I SOLUTIONS* A B C Sol. eps Fe, a Mn, ee pH after 1.75% agar and 2% maltose | py as made up | PH adjusted before | a aged and auto- autoclaving claved R1 Sl 5.74 3.93 4.9 82 5.10 3.91 5.0 S3 4.93 3.90 5.0 S4 4.90 3.90 5.0 S5 4.83 3.90 5.0 S6 5.08 3.90 5.1 R2 S1 4.86 3.93 4.8 82 4.75 3.88 5.0 S3 4.73 3.93 4.9 S4 4.73 4.10 5.0 S5 83 4.10 5.0 R3 S1 5.26 4.22 5.0 82 4.86 4.24 5.0 53 4.83 4.20 5.0 S4 76 4.20 5.0 R4 81 4 4.20 5.0 S2 4.81 4.22 5.0 53 4.24 4.9 R5 S1 4.86 4 4.9 S2 4.83 4.22 4.9 R6 S1 4.80 4.29 4.9 *'The pH determinations indicated in columns A and B were obtained by the quinhydrone electrode, while the results in Column C were obtained colorimetrically by the use of brom-cresol-green as an indicator. [Vor. 20 228 ANNALS OF THE MISSOURI BOTANICAL GARDEN TABLE IV ACIDITY RELATIONSHIPS OF TYPE IV SOLUTIONS DETERMINATIONS MADE AS INDICATED IN TABLE III A B C Sol. plus Fe, B, Mn y, 2n, . pH after 1.75% agar and 2% maltose pH as made up pH adjusted. before |" | (ded and auto- autoclaving claved R1 S1 4.3 4.1 5.0 82 3.71 4.18 4.9 S3 3.5 4 4.9 R2 S1 3.89 5.0 S2 3.67 4.20 4.9 53 3.59 4.29 4.9 R3 S1 4.05 3.90 4.9 S2 3.73 4.22 5.0 S3 3.64 4.30 4.8 R4 S1 3.91 3.91 5.0 S2 3.69 4.25 4.9 S3 .54 32 4.9 R5 Sl 4.34 3.81 4.9 82 3.78 4.15 5.0 R6 S1 4.27 3.90 5.0 20 minutes in à small vial of calcium hypochlorite solution pre- pared as described by Wilson (15). They were then inoculated directly into the flasks of nutrient agar by a platinum wire in- oculating needle. The flasks were incubated at 20-25? C. in a specially shaded greenhouse compartment. During the warmer summer days the greenhouse temperature could not be held below 30°, and on rare occasions 35°, but the short duration of these high temperatures apparently was without any harmful effect. III. RESULTS Type I.—The growth period of these cultures extended from April 15 until November 12, 1932. At the end of this time, the 1933] WYND— ORCHID SEEDLINGS AND NUTRITIONAL IONS 229 diameter of the seedlings was measured by a low-power microscope equipped with a calibrated eye-piece micrometer according to the method of Quednow (730). The flasks had been very lightly inoculated so as to eliminate crowding of the growing seedlings. The figures in table v indicate the diameter in microns, each TABLE V GROWTH OF SEEDLINGS IN TYPE I SOLUTION. THE FIGURES EPRESENT DIAMETERS IN MICRONS Solution Series 1 Series 2 R1 S1 915 910 82 1100 1130 S3 970 1020 S4 1015 1010 S5 925 940 S6 810 830 R2 S1 (Contaminated) (Contaminated) S2 925 S3 855 845 S4 600 740 S5 740 690 R3 S1 805 715 82 660 740 S3 690 725 S4 585 445 R4 S1 860 690 S2 620 620 S3 620 670 R5 S1 635 640 S2 475 585 R6 Sl 680 615 based on an average of 25 individuals. Figure 1 shows graphi- cally the comparative development on the different media used. Figures 3 and 4 indicate the areas of the triangles corresponding to greatest, medium, and least growth. The circles are drawn to scale and represent comparatively the magnitude of the seedlings of the respective cultures. [Vor. 20 230 ANNALS OF THE MISSOURI BOTANICAL GARDEN Type IV.—The growth period of these cultures lasted from April 28 until November 20, 1932. Table vr presents the data for the growth on these solutions. As before, the diameter is indicated in microns and is the average of 25 determinations. Figure 2 shows graphically the comparative development. ha Ri Rive Ri Ss + e amaaan ene e ee A a R 5s —————— n fare aan m atas s ttai B MM n t Ra S2 LTD Rid ——————— BTE Ra Si ———————— MM —Ó—M Mo M ne esti ars a RR eae Re 53 Ri $e RE A RR RR t s e Se A anae f: $6 LM 5$1 np —»—————— xd 15$ A oem an Dd 355 CES OR urnam ars ce aa $i LS Ty $ Ry 9*1 Ed 3 LE RA $2 Re 3! Re $5 w3: F 2 Res Wis a b Fig. 1. a: diagrammatic representation of the relative diameters of seedlings germinated on Type I solution. Series 1; b: relative diameters of seedlings germinated on Type I solution. Series 2. SI $55! $551 Ris RISI Ysi Rasi PE rs nena ee ae ae a ee A WSI MSI LIETI M $2 $3$2 RIS Rass 1252 2$2 f4$32 $52 RSS Rode Red RES! $353 R233 R233 RAS3 T $3 Riss Riss a Fig. a: diagrammatic — of the relative diameters of seedlings germinated on Type IV solution. Series 1; b: relative diameters of seedlings germi- nated on Type IV solution. Series 2. Figures 5 and 6 show the areas of the triangles producing greatest, medium, and least growth, the circles being drawn as before to indicate the average diameters of the seedlings of the respective cultures. 1933] WYND—ORCHID SEEDLINGS AND NUTRITIONAL IONS 231 TABLE VI GROWTH OF SEEDLINGS IN TYPE IV SOLUTION. THE FIGURES REPRESENT DIAMETERS IN MICRONS Solution Series 1 Series 2 R1 S1 920 990 S 825 825 S3 550 595 R2 Sl 955 950 S2 740 740 S3 635 665 R3 S1 965 965 S2 765 770 S3 750 690 R4 S1 1010 950 S2 730 730 S3 570 630 R5 S1 975 1035 S2 715 715 R6 S1 675 705 IV. Discussion Examination of figs. 3 and 4 shows perfect agreement between the duplicate series of Type I in the areas of greatest growth. These areas are characterized by having very low concentrations of KH;PO, The five best cultures in row 1 have only one-eighth of the total osmotic concentration due to this salt. The ratios of calcium nitrate to magnesium sulphate are seen to vary be- tween wide limits, .1667 to 2.500, with no significant effect on growth. The areas of medium and least growth do not show such perfect agreement in the duplicate series, but a comparison shows that the areas of medium growth correspond in general to those having the medium KH;PO, concentration, and the areas of least growth correspond to those having the highest proportion of this salt. This relation is particularly clearly shown in fig. 4. From these data one may be led to the conclusion that germinating orchid seeds and young seedlings are comparatively sensitive to [Vor. 20 232 ANNALS OF THE MISSOURI BOTANICAL GARDEN varying concentrations of KH;PO,, but relatively insensitive to wide variations in Ca(NO;); and MgSQ,. Since the ions were present as salts, it is impossible to determine which ion is responsible for the result. It is possible to add the six ions in but one other combination of three salts so that each anion will be linked with a different cation. This combination is represented by the Type-IV triangle. It is unfortunately impossible to prepare the complete Type-IV series because at the pH used (5.0) the solutions having the higher Ca(H;PO;), concentrations precipitated on sterilization due to the decom- position of this compound. (EP v. JA QU K Ha Po, Fig. 3. Type I, series 1, showing Fig. 4. Type I, series 2, showing areas of high, medium, and low yield. areas of high, medium, and low yield. Culture R2 S1 omitted because of Culture R2 $1 omitted because of Aspergillus contamination. Aspergillus contamination. Growth on the possible solutions of this type shows a remark- able result. The areas of greatest growth perfectly agree and are characterized by solutions having the minimal proportion of the total osmotic concentration due to Ca(H;PO). Again, the ratios of the other salts, KSO, to Mg (NOs). vary between wide limits, .1667 to 2.500, with very little effect on growth. The areas of medium growth correspond to those of medium concen- tration of this salt, and those of least growth correspond to the areas of high concentration of this salt. This comparison of the growth on solutions of Type I with that on Type IV indicates that in both cases greatest growth was 1933] WYND—ORCHID SEEDLINGS AND NUTRITIONAL IONS 233 associated with the lowest proportions of the phosphate salt, irrespective of which ion it accompanied, and that the relative proportions of all other ions were of comparatively little effect. The fact that growth was inversely proportional to the concen- tration of the H:PO, ion seems to be related significantly to the results of Gregory (28). This author studied the effect of the different ions on growth by an ingenious mathematical treatment of the results published previously by Johnston (’24) concerning the growth of potato plants in three-salt solutions. In studying duplicate series, Gregory found that in each the fraction of the total growth produced by the H;PO, ion was a negative quantity. Fig. 5. Type IV, series 1, showing Fig. 6. Type IV, series 2, showing areas of high, medium, and low yield. areas of high, medium, and low yield. The cultures of high concentration The cultures of high concentration Ca(H;PO,) omitted (see text). Ca(H2PO,)2 omitted. In general, he also found that the growth produced by other negative ions was much less than that produced by the positive ions. If we indicate the concentration of each ion in a “molar” sense and compare the composition of the best solution of Type I and Type IV in table vi, an interesting analogy is apparent. Despite the great apparent variation in composition of these two solutions, we see that in both the concentrations of the negative ions are of the same order of magnitude, while those of the positive ions vary greatly. It seems therefore probable that growth was [Vor. 20 234 ANNALS OF THE MISSOURI BOTANICAL GARDEN being conditioned in both cases by the negative rather than the positive ions. TABLE VII CONCENTRATION OF IONS IN “MOLS" OF THE BEST SOLUTION OF TYPE I AND TYPE IV Type I Type IV H3PO,7 .0025 .0038 NO; . 0098 .0078 SO, - 0123 0097 K* 0025 0194 Ca* * 0049 0019 Mg* * 0123 0039 These results need not be interpreted as contradictory to those of Smith (732), who found that better growth was obtained in solutions having larger amounts of ammonium sulphate added to Knudson's solution B. Smith added nitrogen in the ammonium ion, while we used the nitrate ion. It is entirely possible that orchid seedlings might react favorably to an increase of nitrogen as the ammonium ion, while growth would not improve with increases in the form of nitrate. This is indeed made very prob- able by the work of Burgeff ('09), who found better growth with ammonium than with nitrate salts. We are at a loss to relate these results to those of La Garde (29) who states, in regard to superiority of his solution L over that of Knudson’s solution B, as follows:—'' The seedlings ap- peared darker in color and had progressed relatively further in development. This effect might be ascribed to the larger dose of phosphorus and potassium.” Experiments now in progress, which will be reported later, show that La Garde’s solution L is indeed a very favorable orchid medium, but it seems probable that we may not ascribe this quality to the increase of phosphorus and potassium per se. This study further emphasizes the suggestion made by some authors that the Ca:Mg ratio is not of such importance as it was once thought. At least the limits of variation used in our experiments were not sufficiently great to produce any significant effect. 1933] WYND—ORCHID SEEDLINGS AND NUTRITIONAL IONS 235 V. SUMMARY Seeds of Cattleya Trianae Linden and Rchb. f. were germinated on three-salt solutions of Type I and Type IV, with the following results :— (1) The seedlings showed best development on media having low proportions of the total osmotic concentration due to the phosphate salt, irrespective of whether it was present as the potassium of the calcium compound. (2) The proportion of all other ions within the limits used appeared to have relatively little effect on germination and growth. (3) The negative ions appeared of greater importance in con- ditioning orchid seedling development than the positive ions. (4) The best growth in Type I solutions was found in culture R1 S2 having the composition :— OT ee aurea ad 0025 M à po eM 0049 M Mg... oe sro xr Pe ee rae 0123 M (b) The best growth in Type IV solutions occurred in culture R5 S1 containing :— BBG. cen sons 0s Xx cs chee .0097 M Ca(H;PO4s. . eee cea lior ey .0019 M Mg(NO3s. ...... eere rni rt rrr 0039 M VI. ACKNOWLEDGEMENTS The author wishes to thank Dr. George T. Moore, Director of the Missouri Botanical Garden, for furnishing seeds for this investigation; and also to express his appreciation to Elinor Al- berts Linder, formerly Orchidologist to the Missouri Botanical Garden, for helpful suggestions while employed at this institu- tion; and also to Dr. E. S. Reynolds, under whose supervision the work was carried out. [Vor. 20 236 ANNALS OF THE MISSOURI BOTANICAL GARDEN BIBLIOGRAPHY Schreinemakers, F. A. H. ('93). Konzentrierung oder Verdünnung einer Lösung bei nter Temperatur. Zeit. Phys. Chem, 11: 81-85. 1893. Bancroft, A. D. (02). Synthetic analysis of solid phases. Jour. Phys. Chem. 6: 178- 1902 Burgeff, H. (09). Die Wurzelpilze der Orchideen; ihre Kultur und ihr Leben in flanze. Jena, 1909. Schreiner, O., and Skinner, J. J. (10). Ratio of DA nitrate, and potassium in absorption and growth. Bot. Gaz. 50: 1-30. 1910. Shive, J. W. (15). A three salt nutrient solution *, plants. Am. Jour. Bot. 2: 157-160. 1915. Wilson, J. K. (15). Calcium hypochlorite as a seed sterilizer. Ibid. 420-427. 1915. McCall, A. G. (16). Physiological balance of nutrient solution for plants in sand cultures. Soil Sci. 2: 207-253. 1916. Shive, J. W. (17). A study of physiological er 2 buckwheat in three-salt solution. N. J. Agr. Exp. Sta. Bull. 319: 917. — — ——, and Martin, W. H. (18). A decre Ped of salt requirements for pe: “and for mature buckwheat plants in solution cultures. Jour. Agr. Res. : 151-175. 1918. — — ——, (18a). A comparison of salt requirements for young and for mature buckwheat plants in water culture and sand culture. Am. Jour. Bot. 5: 186-191. 1918. Livingston, B. E., and Tottingham, W. E. (18). A new three-salt nutrient solution for plant cultures. Ibid. 337-346. 1918. , (19). A plan for comparative research on the salt requirements of repre- sentative agricultural plants. Baltimore, 1919. Van Alstine, E. (19). The relation of salt proportion to the growth of wheat in sand cultures. N. J. Agr. Exp. Sta. 40: 366-374. 1919. Martin, W. H., and Shive, J. W. (’20). A study of the salt requirements of the potato. N. J. Agr. Exp. Sta. Rept. 41: 409-412. 1920 Meier, H, F. A., and Halstead, C. E. (21). Hydrogen-ion eiticntitwiibt relations in a ihreoquit solution. Soil Sci. 11: 325-352. 1921. Tottingham, W. E., and Rankin, E. J. (22). Nutrient solutions for wheat. Am. Jour. Bot. 9: 270-276. 1922. Knudson, L. (22). Non-symbiotie germination of orchid seeds, Bot. Gaz. 73: 1-25. 1922. Poole, R. F., and Fant, G. W. (22). Further study of the relation of various fertilisers mixtures to the growth of celery in muck soil. N. J. Agr. Exp. Sta. Rept. 43: 395-399. Clement, E. (24). Germination : E TU and other seed without fungal aid. Orchid Rev. 32: 233-2 — ———, (248). The apa odia of orchid seeds. Ibid. 359- 366. 4. Ballion, M., and Ballion, G. (24). The non-symbiotic germination of orchid seeds in Belgium. Jbid. 305-309. 1924. Johnston, E. S. (24). Growth of potato plants in sand cultures treated with the “six types" of nutrient solutions. Md. Agr. Exp. Sta. Bull. 270: 53-86. 1924. Davidson, O. W. ('28). An application of the triangle system in determining a nutrient solution suitable for research with the peach in sand culture. Am. Soc. Hort. Sci., Proc. 25: 354-358. 1928. 1933] WYND—ORCHID SEEDLINGS AND NUTRITIONAL IONS 237 Gregory, F. G. (’28). The differential effect of the ions of three-salt solutions on the growth of potato plants in sand cultures. Roy. Soc. Lond., Proc. B. 102: 311 1928. La Garda; R . 29). Non-symbiotic germination of orchids. Ann. Mo. Bot. Gard. Quednow, K. G. (30). Beiträge zur Frage der Aufnahme gelóster Kohlenstoffver- bindungen durch Orchideen und andere Pflanzen. Bot. Archiv. 30:51-108. 1930. Addoms, Ruth M., and Mounee, F. C. (31). Notes on the nutrient requirement and the histology of the cranberry (Vaccinium macrocarpon Ait.) with special reference to mycorrhiza. Plant Physiol. 6: 653-669. 1931. Smith, F. E. V. (32). Raising orchid seedlings asymbiotically under tropical conditions. Gard. Chron. 91: 9-11. Annals of the Missouri Botanical Garden Vor. 20 APRIL, 1933 No. 2 A REVISION OF THE NORTH AMERICAN SPECIES OF VERBENA! LILY M. PERRY Instructor in Botany, University of Georgia, Athens, Georgia Formerly Rufus J. Lackland Research Fellow in Botany in the Henry Shaw School of Botany of Washington University INTRODUCTION Verbena, the type-genus of the family Verbenaceae, is pre- eminently American. Owing to inherent variability and varia- tions caused by adaptation to local conditions of growth, the group is arather complex one. Moreover, since the last comprehensive monograph of the genus was published so long ago that it contains only some twenty of the hundred or more described North American species and varieties, considerable material has been incorrectly named. It appeared, therefore, that a critical examination of the group might be helpful in untangling the existing confusion of identities. The following paper is an at- tempt to present the results of a study of the North American species, together with a brief review of the historical, morpho- logical, and phylogenetic aspects of the species under considera- tion. The materials studied are indicated in the text by abbreviations for the herbaria to which they belong: the Academy of Natural Sciences of Philadelphia (ANSP); Field Museum of Natural History (F); Gray Herbarium of Harvard University (G); the Royal Botanic Gardens, Kew (K); the Missouri Botanical 1 An investigation carried out at the Missouri Botanical Garden in the Graduate Laboratory of the Henry Shaw School of Botany of Washington University, and submitted as a thesis in partial fulfillment of the requirements for the degree of doctor of philosophy in the Henry Shaw School of Botany of Washington University. Issued July 10, 1933. Ann. Mo. Bor. Garb., Vor. 20, 1933. (239) [Vor. 20 240 ANNALS OF THE MISSOURI BOTANICAL GARDEN Garden (MBG); New York Botanical Garden (NY); Pomona College (P); United States National Herbarium (US); photo- graphs of types and authentic specimens have been received from the Museum of Natural History, Paris (Par.); the Herbarium of the Botanical Garden and Museum, Berlin-Dahlem (Bot. Mus. Berl.-Dahl.), and the Herbarium of the Botanical Garden, Madrid (Herb. Bot. Gard. Madrid). For the photographs and the use of the loans, as well as for the many courtesies received while visiting different herbaria, the writer is truly grateful and hereby acknowledges her indebtedness to all who have given assistance in this way. She expresses her hearty thanks and cordial appreciation to Dr. George T. Moore, Director of the Missouri Botanical Garden, for the use of the extensive privileges of the excellent library and the herbarium. She is also indebted to Miss Nell Horner, Librarian, for any suggestions in bibliographical matters. To Dr. J. M. Greenman, who has directed and supervised the work, giving ready aid and helpful suggestions at all times, the writer expresses her particular gratitude for the encouragement and personal interest shown throughout this study. Taxonomic HISTORY The derivation of the word Verbena is obscure. The name has been handed down to us as representing certain herbs used in sacred rites by the ancients. Although apparently one of the well-known plants in botanical history, Verbena was described comprehensively for the first time by Tournefort,! whose descrip- tion formed the basis of the synopsis of the genus as accepted by Linnaeus.’ In the ‘Species Plantarum,’ Linnaeus? records fourteen species of which eight have been removed since to other genera. The latter fact is sufficient to indicate that the Linnaean concept was too inclusive. It was held, however, until 1806, when Jussieu,‘ in an article on ''Observationes sur la famille des plantes 1 Tournefort, Institutiones Rei Herbariae, 200, pl. 94. 1700. 2 Linnaeus, Gen. Pl. 12. 1754. 3 Linnaeus, Sp. Pl. 18-21. 1753. * Jussieu, Ann. Mus. Paris 7: 71-73. 1806. 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 241 Verbénacées," pointed out that the genus was a composite one containing several units. Amongst these he showed that Verbena differed from the others in having four stamens and four nutlets. This definition was accepted by Endlicher* under the heading '" VERBENAE VERAE," and it has continued to the present time. Although the genus is predominantly American, the earlier species were described largely from plants grown in various European gardens. ‘This was unfortunate, since such types are often extinct or inaccessible or sometimes difficult to interpret on account of probable fusion or other modifications appearing in the species after some few years of cultivation. The first treatment of the North American species, as such, is that of Michaux.® He recognized eight species, all from the region of the middle and southeastern United States. From time to time new species were collected by pioneer explorers. Kunth’ de- scribed eight from the Mexican region, Bentham: added three more, and Martens and Galeotti® reported eleven, of which six were new. ‘The ablest and most inclusive work on the genus as a whole is that of Schauer.!° In this he accepted seventy-one species and listed fourteen more as insufficiently known; of these, twenty-seven are indigenous to North America. Appearing about the same time as Schauer’s work is that of Walpers,1 whose review might be considered as an assembling of the litera- ture of the genus rather than the results of a special monographie study; nevertheless, it is comprehensive and contains a few new species. Subsequently no publication of the North American species ?n toto has appeared, although the works of Gray” and Small! are very helpful in so far as regional floras are concerned. The last two publications contain the only keys to the group under consideration. The type-species of the genus is V. officinalis L. 5 Endlicher, Gen. Pl. 633 m 3085). 1838. 6 Michaux, Fl. Bor.-Am. 2: 18-15. 1803. 1 HBK. Nov. Gen. et n 2: 272-271. 1818. 8 Bentham, Pl. Hartw. 21. 39. ? Martens & Galeotti, Bull. Acad. Brux. 11?: 320-324. 1844. 10 Schauer in DC. Prodr. 11: 535-556. 1847. 1! Walpers, Rep. 4: 13-33. 1844-48; 6: 686-687. 1846-47. ? Gray, Syn. Fl. N. Am. 2!: 335-338. 1878. 13 Small, Fl. Southeast. U. S. ed. 1, 1008-1011. 1903. [Vor. 20 242 ANNALS OF THE MISSOURI BOTANICAL GARDEN Gross MORPHOLOGY Roots.—Some species are annual with a fibrous root system branching from a slender primary root; others are perennial with a larger tap-root. ‘They maintain themselves from year to year by crowns from which clustered new shoots rise near the base of the old ones. A few species show a tendency to root at the nodes. Probably, for the most part, the main purpose of these secondary roots is to aid in obtaining nourishment for the plant, as instance, V. canadensis, a hardy plant with stems rooting only at the lower nodes. Nevertheless, in V. teucriifolia, they may be a possible means of propagation, as definite roots are found at more or less remote nodes; occasionally, both roots and fruiting spike are developed from the same node. In many cases, it is difficult to tell whether a species is an annual or only a short-lived perennial. The question of duration depends in some measure on climatic factors. Stems.—' l he stems are slender, terete, or, more commonly, quadrangular in cross-section with sharp or obtuse angles, and chiefly pubescent. Apparently all are herbaceous, the perennials dying down to the ground in winter and reappearing again in the spring. ‘The shorter forms branch freely with decumbent- ascending or prostrate branches; the taller are usually erect with branches loosely ascending-spreading. All the lower branches are opposite, but the upper may be somewhat irregularly placed. Leaves.—In the majority of the species of the section Glandu- laria the leaves are deeply cleft, whereas those of the Verbenaca are in large part merely dentate or incised. As a whole they are membranaceous, varying greatly in size and somewhat in thick- ness. In a few species they are strictly sessile; in most, however, they are either subsessile or short-petioled and tend to be strongly veined beneath. All have more or less short, sheathing, and connate bases. In general the leaves are opposite, but at times they are subalternate or irregularly disposed on the upper part of the stem; they are rarely ternate, and this arrangement appears to be an individual expression rather than a specific or even a varietal character. Pubescence.—The trichomes are simple and unicellular, varying in quantity, length, rigidity, coarseness, and direction. None 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 243 of the pubescence is truly hispid, but in some species the hairs are slightly stiffer than in others; this indument, as exemplified in the calyx of V. bipinnatifida, is designated as ‘‘hispid-hirsute.”’ In V. prostrata and V. Gooddingii the pubescence is character- istically soft-villous, but in the majority of species the general tendency is to approach a hispidulous-hirsute to strigillose condition. In the former instance, the trichomes, especially along the stem and on the inflorescence, are of different lengths and more or less spreading; whereas in the latter, as in V. tam- pensis, they are short, approximately equal, and appressed. Often the hairs of the upper surface of the leaf spring from minute bulbous bases; these, according to Solereder," are calcified or silicified hairs, and the trichomes with somewhat pustulate bases, in V. carolina and occasionally in other species, are cystolith- hairs. In any case, Solereder did not find cystoliths dissociated from the hairs. A viscid condition is rarely present, but glandular hairs are widely distributed throughout the genus. Inflorescence.—The inflorescence is terminal at the apices of stems or branches. The flowers are subtended by bracts and are borne in solitary or somewhat panicled or cymosely arranged spikes. Although these vary considerably, it is fairly easy to determine the section to which a plant belongs by the type of its spike. In the majority of the indigenous North American species of the Verbenaca, the spikes are elongated and tend to be irregularly placed, as in Verbena carolina, so that they appear more or less panicled. In V. sphaerocarpa and V. litoralis, however, they are commonly arranged in simple or compound cymes. A spike may be either very slender with the flowers somewhat remote, or dense with imbricated flowers. In general, the elongated and comparatively narrow spike with small and inconspicuous flowers is a distinctive feature of the Verbenaca native to our continent. On the other hand, the spikes characteristic of the section Glandularia are shorter and relatively broad with larger and more showy flowers. In the earlier stages of anthesis these are in fascicle-like clusters; later, when the rhachis elongates, the in- florescence becomes spicate. 14 Solereder, Syst. Anat. Dicot. ed. 2, 1: 631-633. 1908. [Vor. 20 244 ANNALS OF THE MISSOURI BOTANICAL GARDEN Calyx.—The calyx is tubular, 5-lobed or subtruncate, 5- nerved and persistent. The nerves protrude somewhat beyond the margin of the lobes, forming calyx-teeth; these vary more or less specifically and occasionally intraspecifically. The nerves are unequal in length and the posterior one, along which the calyx ruptures to release the schizocarp, is always the shortest. In the Glandularia, the calyx is at least twice as long as the schizocarp and at maturity the teeth and lobes are more or less contorted. In contrast, the calyx of the Verbenaca is generally shorter and may be either connivent, concealing the schizocarp as in V. scabra, or, as in such species as V. urticifolia and V. officinalis, open, disclosing the apices of the nutlets. With the developing ovary, the calyx enlarges a little at the base; usually it is somewhat angled, but in V. tumidula it is round at the base and tends to be inflated. Although only relative in value, the pubescence of the calyx is probably of more worth in specific determination than the other previously mentioned characters. Corolla.—' The corolla is salverform and various in color. The tube is either straight or incurved, as long as or much longer than the calyx, and uniform or slightly enlarged at the throat. The degree of the exsertion of the corolla-tube is apparently a conveni- ent and obvious character, but one to be handled with care; first, it is somewhat variable, and second, when the corolla falls after anthesis, it does not drop at once but may be some little time working its way out of the calyx-tube; hence, the length of the exserted portion appears much greater than it really is. The limb is lobed and may be either conspicuous (6-12 mm. broad) or inconspicuous (2 mm. broad). The shape and the direction of the lobes may be of value in field identification but are not at all satisfactory in work on herbarium material. The throat inside is bearded with straight minutely roughened retrorse hairs and at the orifice with moniliform ones; outside and beyond the calyx it is either glabrous or finely pubescent. The inner hairiness is more prominent in the section Glandularia. Stamens.—The stamens are didynamous, often very short- stalked, and in the North American species, inserted on the upper middle half of the corolla-tube, but not exserted. In V. bipin- natifida and all the related species of its section, the anterior pair 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 245 is very near the orifice of the throat and usually bears a gland- like appendage on the connective. The stamens are too much alike to be useful in specific differentiation. Pistil.—The ovary is superior, bicarpellary, and entire or very shallowly lobed at the apex. The style varies in length with the length of the corolla-tube, bearing the mature stigma approxi- mately in the region of the lower anthers. The typical form is slender and bilobed, the posterior lobe being sterile and laterally compressed, the anterior stigmatic and broad. The relation of these lobes to each other is reasonably constant throughout each section of the genus. In V. scabra, however, the broad stigmatic surface appears to be subtended on either side by a sterile style- lobe. Another interesting variation of the style is produced by V. quadrangulata; here, the base is enlarged or thickened into a somewhat hemispherical-angulate stylopodium and persists as an integral portion of the nutlets. Fruit.—The fruit of Verbena is a schizocarp enveloped by the persistent calyx. As previously stated, the calyx splits along the posterior nerve to release the schizocarp attached to a more or less round-quadrangular gynobase. For the most part, four nutlets or pyrenae develop, two from each biovulate carpel. Generally the nutlets are either subcylindrie or subtrigonous, but in V. quadrangulata they are beaked cylindric, and in V tumidula subovoid. The dorsal surface is more or less convex with the lateral surfaces scarcely definite, rather appearing as continuations of the dorsal portion and adjoining the commis- sural face or faces. This is not always the case, since in V. tumidula the lower part of the nutlet seems inflated and the lateral is somewhat ventricose and smooth. The surface of the nutlet varies from essentially smooth, as in V. carolina and its allies, to striate or sulcate, and is oftenest raised-reticulate or reticulate-scrobiculate on the upper part. Occasionally, the character or scope of these depressions is of specific value, as in V. gracilis; but usually a certain type of nutlet is characteristic of a group of species rather than a single one. The commissure may be more or less muricate or practically smooth. In the subcylindric pyrenae of the section Glandularia, the commissural face is commonly convex and does not extend [Vor. 20 246 ANNALS OF THE MISSOURI BOTANICAL GARDEN to the tip of the nutlet; whereas in the trigonous nutlets of the Verbenaca, the commissure ordinarily has two faces meeting at right angles, and is as long as the nutlet. The gynobase varies with the size and shape of the pyrenae which it bears. In V. carolina and V. Ehrenbergiana, it is nearly flat and almost orbicular, but in species with larger nutlets, as in V. simplex and V. neomexicana, it is slightly deeper and in large part somewhat quadrangular. It shows its greatest develop- ment in the section Glandularia where the pyrenae are largest and often somewhat broadened at the base. Perhaps the gyno- base should be used as a character complementary to that of the nutlets, but its definite value is rather elusive. The more important morphological characters, which have been used in the delineation of the genus and its sections, are illustrated with a legend (pl. 13). Hybrids.—It is a well-accepted fact that the cultivated races of Verbena have arisen through the more or less fusing of species attractive for horticultural purposes. Granting this, without field observation and knowledge gained through experimental work, the writer does not propose to discuss the question of hybridization in Verbena. Nevertheless, it is necessary to point out that there is a considerable number of specimens amongst the collections of this genus from the Middle States, which show variation in several directions or a combination of the characters of two or more species; it is the prevailing opinion that these plants are hybrids. RELATIONSHIPS AND DISTRIBUTION The genus Verbena is a member of the tribe Verbeneae accord- ing to Engler and Gilg,5 or of the Huverbeneae according to Briquet in Engler and Prantl.^ It is distinguished from its immediate relatives by the four dry and hardened nutlets. In this paper fifty-one species are recognized, four naturalized and forty-seven indigenous. With the exception of a few border-line species, it is relatively easy to separate the genus into two fairly distinct sections in which the majority of the members are so 5 Engler & Gilg, Syllab. der Pflanzenf. eds. ; AJ 10, 339. 1924. Engler & Prantl, Nat. Pflanzenf. 4?*: 146. 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 247 different that it would seem as if they had followed divergent trends for a long period of time. In attempting to present any genus in à natural order, it is highly desirable to know what species are primitive. However, without an intimate knowledge of the more diversified South American species, it seems prefer- able to state the evidence derived from the study of the North American group and leave the conclusions vague rather than more or less arbitrary. The section Verbenaca is probably the older of the two. It has the wider distribution, ranging from Canada to Panama as well as from the east to the west coast of United States and Mex- ico. Also, it is larger, containing thirty-two species. A number of these are annual, the remainder are perennial, but it is often hard to distinguish the members by this criterion. Although admittedly varying and apparently intergrading, the species as a whole are more stable and more readily recognized than those of the section Glandularia. The flowers are much smaller and for the most part are produced in greater abundance; the nutlets also are abundant and in some cases minute. The combination of these characters leads one to conclude that this section is the more primitive one. Within the section, some species appear to be more closely related than others. V. officinalis, V. menthaefolia, V. Halei, and V. riparia form a natural group possessing the same general habit, type of pubescence and of nutlets. These are species of more or less adjacent regions except V. officinalis, which has been introduced from Europe. Again, V. Ehrenbergiana, V. carolina, V. longifolia, and V. recta are very closely related, with V. urticifolia and V. scabra only a little farther removed. The last five species differ from the first mentioned in their more or less upright habit and less divided leaves, also in their smaller and somewhat smoother nutlets. It would appear as if all had arisen very close to the ancestral form and diverged only a little in the process of adaptation to environment. V. hastata and V. simplex are quite distinctive in habit, but the abundance of intermediate forms occurring between these and the two species, V. stricta and V. bracteata, would seem to indicate a close relationship amongst these units. V. Orcuttiana, although [Vor. 20 248 ANNALS OF THE MISSOURI BOTANICAL GARDEN somewhat isolated, appears to have its nearest affinity in V. simplex. A third closely related group contains V. canescens, V. neo- mexicana, V. plicata, and V. gracilis; the common characters here are those of leaf-incision, pubescence, and general features of inflorescence. V. perennis resembles this group in all respects except in its quite aberrant pubescence of short antrorse and hispidulous trichomes; in this character it simulates more closely the first-mentioned group. The following are more or less isolated species of limited distribution and with fairly static characters: V. xutha, V. prostrata, and V. robusta. V. carnea is probably a relic of some ancient form; the distinctive character of the schizocarp gives no clue to its affinities, but rather emphasizes the anomaly of the species. The section Glandularia seems to be the more nascent. The species are perennial, developing from a crown, and have a more or less sprawling to upright habit. The characters are highly variable and some species are difficult to distinguish, apparently passing into each other. Particularly is this true of V. ambrosi- folia, V. Wrightii, and V. ciliata. Some of the specimens from northern Mexico have been most perplexing; while it is perfectly evident that they belong to this group, none of them is typical. V. bipinnatifida is a closely related and wide-ranging species; by comparison, its characters are fairly clear-cut. The remaining species of the section are easily associated in groups or are anomalous. V. canadensis, V. maritima, and V. tampensis are very much alike either in habit or in floral structure. V. tam- pensis, however, is one of the species of this group where the glands may or may not be present on the anthers, and may represent the development of some transitional stage between the two sections. V. elegans of Mexico is very closely parallel to V. canadensis of the southeastern and central United States. V. delticola seems to belong in this association also; yet it is interest- ing to note that its fruit shows a tendency to develop a beak, and this may be significant of some little affinity with V. quad- rangulata, a species so aberrant that it has been treated as a separate genus. V. Gooddingi V. tumidula, and V. pumila 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 249 might very well be related to any of the previously mentioned species of the section. V. setacea resembles V. Gooddingw var. nepetifolia in habit, but it has such long calyx-teeth that one is a little hesitant about expressing any ideas of its relationship. V. teucriifolia and V. ciliata appear to intergrade in Mexico. The former, as well as V. pumila, shows a tendency toward reduced corollas and perhaps parallel development. V. lilacina and V. amoena, are without close relatives. Considering the genus as a whole, some species are in a state of flux showing many atypical forms and covering fairly large geographie areas; others are just as widespread and yet are characterized by more static characters; and again, others are endemie showing a very limited distribution and appearing as outlying members of the group. The species probably vary from youth to age in the order mentioned. As previously stated, the genus Verbena is fairly widely distributed in North America; its range extends from southern Canada to Panama and from the east coast to the west both in United States and in Mexico. It is found also in the West Indies and on islands in the Pacific off the coasts of California and Mexico. It occurs in great abundance in Texas, New Mexico, and Arizona, south along the Sierra Madre and throughout the eastern part of Mexico. It is also fairly well distributed in southern Mexico, but only a few species are reported from Central America. The region of greatest specific concentration for North America is in the state of Texas, where twenty out of the forty-seven indigenous North American species are rep- resented. Amongst these species are members of both sections, some with very definitely clear-cut characters and more or less limited distribution, others with fluctuating characters and more widely distributed. By way of contrast, let us consider the outlying members of the genus which have stable characters and are isolated or in- habit very limited areas—V. lilacina, V. sphaerocarpa, V. Or- cuttiana, V. setacea, V. macrodonta, V. amoena, and V. subuligera; or again species such as V. carnea and V. quadrangulata, which are singular but have a larger geographic range. All these appear to be relies of a more ancient distribution. Undoubtedly [Vor. 20 250 ANNALS OF THE MISSOURI BOTANICAL GARDEN V. lilacina and V. sphaerocarpa, endemics of Cedros and Socorro Islands respectively, developed as a result of geographic isolation. The same may be true of the Lower Californian entities, V. macrodonta, V. setacea, and V. Orcuttiana. Perhaps V. amoena from Lecheria, a locality some little distance north of the city of Mexico, and V. subuligera, from the Sierra Madre west of Durango, owe both their restriction and their differentiation to the presence of mountainous barriers. However, the same conclusion is hardly valid for either V. carnea, ranging from North Carolina to Florida and west to Texas, or V. quadrangulata, of southern Texas and Tamaulipas; yet they are probably relic endemics which have survived in favoring areas. If this be true, here is another bit of interesting evidence supporting the idea that the Verbenaca is the older of the two sections. Of the nine species enumerated above as being of more ancient derivation, four, V. setacea, V. lilacina, V. amoena, and V. quadrangulata, have been placed in the section Glandularia; yet, in the flowers examined, the anthers are definitely unappendaged. To be sure, this disposition of the species is not in accord with the previous classifications and may not be acceptable; nevertheless, it seems logical to believe there must be transitional forms in the develop- ment and evolution of species; and in this case, the major portion of the characters of the species in question are those of the section Glandularia. Two other species which have developed a little farther along this same line are V. tampensis and V. pumila; gland-like appendages may or may not be present on the anthers; if present, they are usually very small. It may be of interest to note that, in agreement with Jordan’s law of distribution, the species associated in groups usually occupy adjacent regions. Now as to origin.—It is possible that the progenitors of Verbena were very widespread at the beginning of the Pleistocene and with the glacial advance migrated southward; later, when more favorable climatic conditions developed they spread northward, establishing themselves in the present regions of distribution. Another possibility is that the centre of distribution is in South America and migration has been northward. However this may be, V. litoralis is the only known species indigenous to both 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 251 countries and does not seem to be very closely related to the remaining North American species. Without a detailed in- vestigation of the South American group, it seems preferable to reserve assumptions in the matter. Economic VALUE Verbena is of very little value economically. It was introduced into cultivation in European gardens relatively early in the his- tory of botanical explorations. Although not so popular as it was once, on account of its vigor and hardiness, it is still used in many gardens to impart bright colors to the landscape. TAXONOMY Verbena [Tourn.] Linn. Gen. Pl. 12. 1754. Obletia Rozier, Jour. Phys. 1: 367, pl. 2. 1773. Glandularia Gmel. Syst. Veg. 920. 1791. Billardiera Moench, Meth. 369. 1794. Shuttleworthia Meisn. Pl. Vasc. Gen. 290. 1839, and Comm. 198. 1839. Uwarowia Bunge, Acad. St. Petersb. Bull. Sci. 7: 278. 1840. Calyx tubular, 5-toothed, 5-nerved, at maturity often slightly enlarged at the base, persistent. Corolla minute to conspicuous, hypocraterimorphous, subbilabiate, 5-lobed; tube erect or in- eurved, uniform or slightly enlarged above, equalling the calyx- lobes or surpassing them; lobes imbricate in the bud, broadly oblong, obtuse or retuse, unequal. Stamens 4, didynamous, included; anthers ovate, unappendaged or the connective of the upper pair expanding extrorsely into a clavate and gland-like appendage; filaments usually short, attached above the base of the corolla-tube. Ovary superior, entire at apex or very shortly 4-lobed, bicarpellary, 4-loculed; ovules in pairs, one to each locule, anatropous; style terminal, mostly bilobed, the anterior lobe stigmatose and the posterior usually sterile. Schizocarp included in the persistent calyx, dry, hardened or brittle, separat- ing into 4 homomorphous nutlets. Seed erect; embryo straight; endosperm none.—Chiefly American. Herbs or shrubs, erect or decumbent to prostrate, pubescent or occasionally glabrous. Leaves opposite, rarely ternate or verticilate, or the upper [Vor. 20 252 ANNALS OF THE MISSOURI BOTANICAL GARDEN alternate (irregular), variously dentate, incised or dissected, not often entire. Spikes terminal, rarely axillary, peduncled or sessile, densely flowered or elongate and slender with flowers remote, solitary, cymose or paniculate. Flowers hermaphroditic, zygomorphic, sessile and bracteate. SECTION 1. VERBENACA Schauer 1. VERBENACA Schauer in DC. Prodr. 11: 536. 1847. Sterile style-lobe adjacent to the stigmatie surface or sub- lateral but usually not protruding beyond it; ovary entire at the apex with the style attached at the distal end, or, if at all lobed, style inserted in the indefinite depression between the lobes; schizocarp commonly not constricted along the lines of cleavage; anthers unappendaged. Annuals or herbaceous perennials with prostrate, decumbent-ascending or erect stems. Flowers mostly not showy. Calyx rarely more than twice as long as the schizo- carp and not contorted beyond it. Species 1-32. Ser. I. PACHYSTACHYAE Schauer. Heads or spikes, at least in anthesis, crowded and short, not at any time greatly elongated or open, generally disposed in compound cymes.—Chiefly introduced South American species. KEY TO THE SPECIES A. Leaves semiamplexicaul and subcordat B. Inflorescence glandular; bracts ditioni longer than the — corolla-tube 2-3 times as long as the calyx................0... rigida B. Inflorescence not glandular; bracts barely equalling or only dd. ex- ceeding the calyx; corolla-tube scarcely twice as long as the calyx. . Vuarasrumzrag. T S T E S RI eL EROR EVUY. bonarlóits A. Leaves not semiamplexicaul nor subcordate, tapering into a cuneate- attenuate subsessile or petiolar base. C. Spikes short (3-5 mm. long), very dense and appressed-pubescent. D. Fruiting-calyx ascending; schizocarp longer than broad, raised- reticulate above, striate toward the base............ 3. V. brasiliensis D. Fruiting-calyx spreading; schizocarp as broad as long, practically feat: ad: PIE ESPE E ke oe T RD E 4. V. sphaerocarpa C. Spikes 5-10 mm. long, dense at the apex, somewhat open bulón & nd ROAS Pe ii cf ie A te ese o PENNE UN 5. V. litoralis 1. V. rigida Spreng. in L. Syst., Cur. Post. 42: 230. 1827. V. venosa Gill. & Hook. Bot. Misc. 1: 167. 1830. 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 253 V. bonariensis var. rigida O. Kuntze, Rev. Gen. 3?: 255. 1898; Briq. Ann. Conserv. & Jard. Bot. Genéve 7-8: 291. 1904. Stems 2-6 dm. tall, sharply 4-angled in cross-section, scabrous- pubescent; leaves oblong to oblong-lanceolate or narrowly obovate, sessile, more or less semiamplexicaul, subcordate, sharply and coarsely serrate, midrib and veins impressed above but prominently reticulated beneath, scabrous and more or less hispidulous on both surfaces, margins of older leaves somewhat revolute; spikes usually short, dense and cylindrical, disposed in subternate cymes with the laterals pedunculate; bracts lanceolate, acuminate-subulate, usually closely imbricated and longer than the calyx, glandular-pubescent, ciliate; calyx 4 mm. long, glan- dular-pubescent, lobes acute with short mucronate-subulate tips; corolla-tube twice (-thrice) as long as the calyx, pubescent without; corolla-limb 5-7 mm. broad; stamens inserted on the lower middle of the corolla; nutlets trigonous, slightly broader at the base than at the apex, scarcely 2 mm. long, raised-reticulate on the upper half, striate toward the base; commissural faces muricately scabrous. Distribution: indigenous to South America; introduced in the West Indies and southern United States. Specimens examined: Norts Carona: roadside, Williamston, Martin Co., 4 July 1922, Randolph 688 (G). Grorata: uncultivated ground, in perennial spreading patches, 26 May 1928, Reade (NY); vicinity of Thomson, McDuffie Co., 1 Sept. 1907, Bartlett 1122 (P). ALABAMA: Mobile, Mohr (US); Mobile, June 1919, Graves 525 (MBG, US). LourisiANA: Mandeville, 15 Aug. 1912, Pennell 4204 (NY); Covington, Aug. 1919, Arsène (US); Catalpa, 24 Aug. 1912, Pennell 4332 (NY); Baton Rouge, 28 March 1904, Billings 49 (G); roadsides, Plaquemines Co., July 1883, Langlois (NY). Texas: Orange, 1914, Young 662 (P, US); Huntsville, 24 May 1917, E. J. Palmer 12038 (MBG); Houston, 1872, Hall 433 (G, MBG, NY, P, US); Houston, 1913, Fisher 460, 625 (US); Houston, 25 Sept. 1915, Thurow (US); Houston, 22 May 1917, E. J. Palmer 12001 (MBG); 16 km. southeast of Houston, 12 April 1925, Small & Wherry 11813 (NY). Mexico: Vera Cruz: Zacuapan, May 1913, Purpus 6413 (F, G, MBG, NY, US). West INDIES: BERMUDA: St. George's, 13 July 1905, Moore 2984 (G); near Devil's Hole, 13 June 1905, Harshberger (ANSP, G, MBG); near Hamilton, 31 Aug.-21 Sept. 1905, Brown & Britton 153 (ANSP); Tucker’s Town, 3 May 1912, Collins 45 (G); Tucker's Town, 8 Aug. 1913, Collins 269 (G, NY); fields, Montrose, Sept. 1913, Brown, Britton & Wortley 1645 (ANSP, NY); St. David's Island, 22 May-2 June 1919, Brown 693 (ANSP, NY). [Vor. 20 254 ANNALS OF THE MISSOURI BOTANICAL GARDEN Cuna: near Farallones, Oriente, 5 Aug. 1913, Leon 3910 (NY). Jamaica: road to Salt Hill, St. Andrew, 6 May 1915, Harris 11969 (G, MBG, NY); Blue Mountains, St. Andrew, 20 March 1916, Perkins 1024 (G); near Fairfield, 3-7 Sept. 1908, Britton 3176 (NY). GuADELOvPE: without locality, 1893, Duss 3470 (NY). MARTINIQUE: without locality, Sept. 1901, Duss 4697 (NY). This South American species is commonly cultivated. It has escaped and established itself in various places. Kuntze and Briquet have regarded it as a variety of V. bonariensis, but lacking a considerable amount of material for examination, the writer prefers, at present, to maintain it as a specific entity. 2. V. bonariensis L. Sp. Pl. 20. 1753. V. elongata Salisb. Prodr. 71. 1796. ? V. quadrangularis Vellozo, Fl. Flum. 16. 1825; 1: pl. 39. 1827. Stems 1 m. more or less in height, 4-angled in cross-section, somewhat scabrous-pubescent; leaves decussately opposite, lance- olate, subauriculate, semiamplexicaul, acutely serrate, entire toward the base, rugose and hirtellous above, spreading-pubescent beneath, particularly along the prominently reticulated veins; spikes compact, mostly short, commonly sessile and crowded in dense fasciculate cymes; bracts lanceolate-acuminate, barely equalling or slightly exceeding the calyx, pubescent; calyx 3 mm. long, pubescent, lobes acute with short subulate tips; corolla-tube scarcely twice the length of the calyx, pubescent without; corolla- limb inconspicuous; nutlets 2 mm. long, trigonous, chiefly striate, slightly raised-reticulate at apex; commissural faces scarcely reaching the tip of the nutlet, muricate-scabrous. Distribution: native of South America; introduced into southern United States and the West Indies. Specimens examined: Sourn CanoLINA: roadsides near Charleston, Curtiss 1963 in part (F, G, MBG, NY); Charleston Neck, 28 Sept. 1853, Gibbes (NY, US); Charleston, 30 April 1912, Robinson 127 (G); about 6 km. south of Charleston, 9 Nov. 1929, Moldenke 148 (MBG, NY). Georata: along canal bank, Augusta, Cuthbert 267, 358 (NY). LovistaNa: vicinity of Covington, 1920, Arsène 11859, 12534 (F, US). ARKANSAS: roadsides near Forrest City, 17 Oct. 1925, E. J. Palner 29303 (G, MBG, NY). TExas: Huntington, 15 Sept. 1923, Tharp 2658 (US). West INDIES BERMUDA: waste ground, 29 March-3 May 1909, Marble 787 (ANSP, NY); Pem- 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 255 broke Swamp, 21 June 1905, pica (ANSP, G, MBG, NY); roadside, Pem- broke, 31 Aug.-20 Sept. 1905, Brown & Britton 98 (ANSP, G, NY, P); Hamilton, 12 July 1905, Moore 2946 (G); E: Hamilton 16 Jan. 1912, Robinson 94 (G); Tucker's Town, Collins 58 (G), 266 (G, N JAMAICA: Blue Mountains, 14 ie 1890, Hitchcock (MBG); Blue Mountains, Mt. Hybla, 5 Apr. 1916, Perkins 102 (G); St. Helen's Gap, 7 April 1909, Taylor 4244 (NY); St. Helen's Gap, St. Andrew, 4 March 1920, Mazon & Killip 569 (G, NY); Cinchona, 26 July 1903, Nichols 168 (G, MBG, NY); Cinchona, 22 Dec. 1905, Harris 9132 (NY); Arntully, 25 Aug. 1927, Orcutt 2736 (G, MBG). This is another introduced South American species evidently related to V. rigida, but it is a larger and coarser plant with smaller flowers as well as less harsh and somewhat viscid pubes- cence. 3. V. brasiliensis Vellozo, Fl. Flum. 17. 1825; 1:pl. 40. 1827. V. litoralis var. brasiliensis Briq. Ann. Conserv. & Jard. Bot. Genéve 7-8: 202. 19 Stems 1 m. more or less in height, 4-angled in cross-section, practically glabrous below, somewhat scabrous-pubescent above, slightly contracted at the nodes; leaves decussate, lanceolate, tapering into a POE iterate: subsessile or petiolar base, sharply or incised serrate, strigillose and somewhat pustulate above with veins impressed, sparsely pubescent beneath; spikes compact, mostly short and strict, usually sessile in open cymes; bracts scarcely as long as the calyx, lanceolate, subulate at apex, ciliate; calyx approximately 3 mm. long, somewhat appressed- pubescent, lobes acute with short subulate tips; corolla-tube a little longer than the calyx, pubescent without; corolla-limb inconspicuous; nutlets trigonous, about 2 mm. long, strongly striate, raised-reticulate at apex; commissural faces scarcely reaching the tip of the nutlet, muricate-scabrous. Distribution: Glide America; introduced into southern United States. Specimens examined: 2 ea ballast dumps at Wilmington, 2 July 1897, Small 5707 (G, MBG, NY); dry sandy soil, open woods about 3 km. south of Wilmington, 25 July 1922, Biidolph 1004 (G). Groraia: savanna, 27 May 1927, Korthof (US); Brunswick, 12 May 1930, Mol- denke 1184 (MBG, NY). FLoripa: waste place, Chipley, 24 May 1930, Blanton 6598 (MBG, US); waste ground, Pensacola, 24 July 1899, Curtiss 6490 (G, MBG, NY, US); ballast ground, Pensacola, 30 April 1903, Tracy 8706 (F, G, MBG, NY, US). ALABAMA: ballast ground, estuary of Mobile River, 4 July 1893, Mohr (US); Port [Vor. 20 256 ANNALS OF THE MISSOURI BOTANICAL GARDEN Eads, 22 Aug. 1900, Lloyd & Tracy 20 (F, G, MBG, NY, US); north bank, Bayou Terrebonne, Houma, 29 May 1914, Wurzlow (NY). A species very similar in habit to V. litoralis but readily distinguished by the difference in inflorescence. The spikes of V. brasiliensis are short, compact, sessile, and regularly arranged in open cymes; whereas those of V. litoralis are longer, compact or somewhat elongate, peduncled, and arranged in more or less paniculate cymes. 4. V. sphaerocarpa Perry," n. sp. Stems erect, square in cross-section, branched above, glabrous or sparsely scabrous-pubescent; leaves oblong, tapering at the base into a short petiole, 3-6(-10) em. long, acute, sharply serrate, the upper often entire, scabrous-pubescent above with trichomes minutely bulbous at the bases, more or less short- strigillose on both surfaces, prominently veined beneath; spikes short and dense, subsessile or short-pedunculate, cymosely arranged; bracts ovate-lanceolate, shorter than the calyx, acute, pubescent; calyx scarcely 2 mm. long, connivent over the schizo- carp, appressed-pubescent, teeth very short, acutish; corolla- tube protruding slightly beyond the calyx; corolla-limb et 1.5-2 mm. broad; schizocarp 1 mm. long, approximately 1 in diameter; nutlets very faintly striate or essentially ans commissural faces muriculately scabrous. Distribution: known only from Socorro Island. = ek tage examined: x1co: CoLIMA: Socorro Island: March-June 1897, Anthony 380 (G, MBG But US); 27 May-3 July 1903, Barkelew 231 (G, MBG, P, US); 9 May 1925, Solis 70 (US). This endemic from Socorro Island is most nearly related to V. 1 V. sphaerocarpa Perry, sp. nov., annua vel perennis (basi ignota); caule erecto quadrangulari glabro vel sparse scabro-pubescente; foliis oblongis basi in brevem petiolum attenuatis 3-6(-10) em. longis acutis, infimis argute serratis, superioribus saepe integris, omnibus scabro-pubescentibus supra subtusque plus minusve breviter strigosis et reticulatis; spicis brevibus et compactis subsessilibus vel breviter pe- dunculatis in cymas dispositis; bracteis ovato-lanceolatis acutis pubescentibus calyce brevioribus; calyce vix 2 mm. longo connivente adpresso-pubescente; calycis dentibus brevissimis acutiusculis; corollae tubo paulo exserto; corollae limbo circiter 1.5-2 mm. lato; schizocarpio 1 mm. alto circiter 1 mm. lato; coccis obsolete striato- reticulatis.—Collected on Socorro Island, Mexico, March-June 1897, Anthony 380 (MBG), TYPE 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 257 litoralis and has a similar habit, but is easily separated on the distinctive characters of the inflorescence. The spikes are shorter and denser, the flowers are smaller and so crowded that the lower ones appear to be inserted at right angles to the rhachis of the spike; moreover, the schizocarp is fully as broad as or even broader than long, an unusual trait not found elsewhere in the North American species of Verbena. 5. V. litoralis HBK. Nov. Gen. et Sp. 2: 276, pl. 137. 1818. V. caracasana HBK. I. c. 275. V. bonariensis var. littoralis Hook. Bot. Misc. 1: 166. 1830. V. affinis Mart. & Gal. Bull. Acad. Brux. 11?: 322. 1844. V. littoralis « pycnostachya Schauer in DC. Prodr. 11: 542. 1847, excluding V. brasiliensis Vell. Fl. Flum. 1: pl. 40. V. littoralis 8 leptostachya Schauer in DC. Prodr. l. c. V. Hanseni Greene, Pittonia 3: 308. 1898. V. litoralis var. caracasana Briq. Ann. Conserv. & Jard. Bot. Genéve 7-8: 292. 1904 Stems approaching 1 m. in height, square in cross-section, somewhat fastigiately branched above, glabrous or sparsely strigillose, slightly contracted at the nodes; leaves lanceolate or oblong, tapering into a very short petiole or a subsessile base, 3-10 cm. long, decussate, more or less sharply and coarsely serrate, sparsely strigillose on both surfaces, scabrous and some- what rugose above, prominently veined beneath; spikes terminal, fairly dense, pedunculate, cymosely arranged or tending to be panicled, often elongate; bracts ovate-lanceolate, acuminate, sub- equalling or somewhat shorter than the calyx, glabrate; calyx 2-2.5 mm. long, finely strigillose, subtruncate, teeth minute, subulate; corolla-tube variable in length, always somewhat longer than the calyx; corolla-limb inconspicuous, 2.5-3 mm. broad; nutlets trigonous, hardly 2 mm. long, striate, somewhat reticulate at the apex; commissural faces about as long as the nutlet, muricate-scabrous. Distribution: Mexico, Central and South America; introduced into the United tates. Specimens examined: Lourstana: north bank, Bayou Black, Houma, 5 and 9 May 1914, Wurzlow (NY). CALIFORNIA: Clinton, Amador Co., June 1896, Hansen 2025 (MBG). [Vor. 20 258 ANNALS OF THE MISSOURI BOTANICAL GARDEN Mexico: San Luis Porosr: on mountains around the city, San Luis Potosi, 1876, Schaffner 718 (G); Alvarez, 28 Sept.—3 Oct. 1902, Palmer 141 in part (G). Nayarit: Tepic, 5 Jan.-6 Feb. 1882, Palmer 2019 (US), 2014a (G, NY). Jasco: fields, Guadalajara, 11 Aug. 1902, Pringle 11093 (F, G, MBG, NY, US). HipALGOo: near Ixmiquilpan, 1905, Rose, Painter & Rose 9077, 9151 (US) Vera Cruz: Orizaba, 27 July 1891, Seaton 27 (F, G, NY, US); Huatusco, April 1857, Mohr (US); Fortin, Feb. 1883, Kerber 311 (US); Santa Rosa, 13 Aug. 1926, Fisher 168 (US). TABASCO: Mayito, 10 April 1889, Rovirosa (ANSP, NY, US). PuEBLA: Pahuatlan, 14 June 1913, Salazar (US); near Puebla, 20 Oct. 1908, Arsène (US). MicHoacan: low valley, Zinapecuaro, 2 May 1849, Gregg 764 (MBG); Quinceo, 11 Nov. 1909, Arsène (US); Morelia, Aug.-April 1840, Galeotti 781 (K), TYPE col- lection of V. affinis; Morelia, Coronilla, 8 Aug. 1909, Arséne (US). Oaxaca: wet meadows, Sierra de San Felipe, 11 Sept. 1894, Pringle 4877 (ANSP, G, MBG, NY, P, US); Sierra de San Felipe, 6 Oct. 1894, C. L. Smith 222 (MBG, US); San Jacinto, 25 Sept. 1895, L. C. Smith 794 (G). CENTRAL AMERICA: GUATEMALA: uneultivated places, Oct. 1885, Bernoulli 128 (NY); without locality, 1892, Heyde 610 (US); waste places near railway bridge, 5 June 1909, Deam 6180 (G, US); vicinity of Los Amates, Dept. Izabal, 24 May 1922, Standley 24404 (US); Coban, Dept. Alta Verapaz, May 1886, von T'uerckheim 904 (ANSP, F, G, NY, US); near the Finca Sepacuite, Dept. Alta Verapaz, 25 March 1902, Cook & Griggs 148 (US); Canjutz, Dept. San Marcos, 1 Sept. 1922, Salas 32 (US); Guatemala, 1923, Ruano 332, 403 (US); Guatemala City, 1912, Aguirre 4 (US); near Guatemala, July 1921, T'onduz 627 (US); Chiapas, Dept. Santa Rosa, Dec. 1892, Heyde & Lux 4370 (G, NY, US); Santa Rosa, May 1892, Heyde & Lux 3019 in part (G). Honvuras: vicinity of Siguatepeque, Dept. Comayagua, Standley 56082 (F, US). SAN BALVADOR: vicinity of Ateos, Dept. La Libertad, 17 April 1922, Standley 23326 (G, US); Volean de San Vicente, Dept. San Vicente, 7-8 March 1922, Standley 21486 (G, NY, US); Cerro de San Jacinto, near San Salvador, 8 Feb. 1922, Standley 20617 (G, US); San Jacinto, 5 April 1905, Velasco 8848 (G, US); Volean de San Salvador, 7 April 1922, Standley 22842 (US); vicinity of San Salvador, 1905, Renson 291 (NY, US); San Salvador, 1922, Calderon 729 (G, NY, US), 925 (G, US) Nicaraaua: Casa Colorada and vicinity, south of Managua, 27 June 1923, Mazon, Harvey & Valentine 73856 (NY, US); Las Nubes and vicinity, south of Managua, 28 June 1923, Maxon, Harvey & Valentine 7478 (US). Costa Rica: without locality, 19 June 1874, Kuntze 2109 (NY, US); without locality, April 1910, Worthen (MBG); Los Ayotes, near Tilaran, Prov. Guanacaste, 21 Jan. 1926, Standley & Valerio 45430 (US); Aguacate, Nov. 1846, Oersted 11522 (US); vicinity of San Jose, Feb. 1924, Standley 33293 (US); vicinity of La Palma, on the road to La Hondura, 17-18 July 1923, Maron & Harvey 7951 (US); Cerro de Piedra Blanca, above Escasu, 31 Jan. 1924, Standley 32652 (US); near Cartago, 1845-8, Oersted 11324 (US); Cartago, Oct. 1887, Cooper 5890 (F, G, MBG, NY, US). PANAMA: fields along Rio Caldera above El Boquete, 5 Feb. 1918, Killip 3510 (US). This is a widely distributed species with very distinctive habit and somewhat variable inflorescence. In some specimens the 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 259 spikes appear to remain compact, in others they tend to elongate. Schauer used this difference to separate the forms pycnostachya and leptostachya, although he frankly admits the difficulty of distinguishing the two owing to the intermediate phases. V. affinis is characterized by à somewhat coarser floral pubescence; this, however, seems to be a variable feature and, as such, does not appear to merit more than passing mention. Ser. II. LEPTOSTACHYAE Schauer. Spikes slender and open or compact at anthesis, greatly elongating in fruit, solitary or in simple cymes or panicled.—North America. KEY TO THE SPECIES A. Spikes panicled at the apices of stems and LE subtended chiefly by inconspicuous bracts; floral bracts not prom B. Spikes very slender, elongated and graceful, ined with remote fruits. Leaves 1-2-pinnatifid or 3—5-cleft or deeply incised. D. Corolla-limb not less than 3 mm. broad; nutlets 1.5 mm. or more E. diccre a little longer than broad. F. eise) usually densely glandular and somewhat viscid- WAIDOMOORNG ss.. ederin ke 6 oe ROB rre ene . V. officinalis E. Schizocarp about twice as long as broad. G. Leaf-blades diverse in outline, the basal incised-dentate, the middle stem-leaves 1-2-pinnatifid, the upper sparingly hot pd or entire; bracts one-half as long as the calyces at ts n NPEREETE SEPIUS Pon a ae hoe os 8. V. Halei G. Leaf-blades jud in outline; bracts about as long as the onkpoes at snthesia. «205 RR esa rore 9. V. riparia D. Corolla-limb eldest more than 1 mm. broad; nutlets 1.5 mm. or len Joni. ia cap cesses e Eee eee 10. V. Ehrenbergiana C. Leaves serrate or shallowly incised. . Leaves subsessile or rider with attenuate base I. Stems mum leaves mostly crenate-serrate; Nee ulou hispidulous... 7... 2.22. ov EO OS LUE . V. carolina I. Stems inis leaves acutely serrate; calyx pe glabrous TOME. ES E E N . longifolia H. ly petioled with round MG Mon x aid et im spreading; calyx-lob gmatic surf subtended by two sterile ike 1-3- MP MROCDNENUT Ve 14. V. scabra J. Fruiting-calyx ascending; calyx-lobes not connivent; stigmatic e subtended by one sterile style-lobe........ 18. V. urticifolia B. Spikes thicker or densely flowered, usually with contiguous fruits. [Vor. 20 260 ANNALS OF THE MISSOURI BOTANICAL GARDEN K. Spikes subsessile or very short-peduncled, forming a "— in- florescence, not strict; corolla-limb 2 mm. broad.............. 2. V. recta K. Spikes peduncled, usually not indictum strict; corolla-limb à 5 M e ere rer yee PES eT Se TT TRY TIT eT Ors Free 16. V. hastata A. Spikes solitary or in 3's at the apices of stem and branches, or res and subtended by leafy bracts at the base; floral bracts prominent or no L. Schizocarp readily separating into 4 nutlets at maturity. . Leaves serrate-dentate or shallowly incised, or predominantly entire. N. Leaves serrate or edic or shallowly incised; spikes with more or less contiguous fr O. Plants coarse, more or less d hirsute-pubescent or hirsute- villous; leaves —— to ovate-orbicular; spikes stout at anthesis, 7-10 m P. Leaves aai + to ovate-acuminate, short-petiolate; corolla-limb 5-6 mm s Q. Plants em A ineei spikes flexuous, elongated and open in fruit; bracts not exceeding the fruiting calyx; nutlets 2 mm. dis shallowly scrobiculate above, sulcate toward the base... isses ese 21. V. macrodonta Q. Plants lucis oO spike strict, dense in fruit; bracts surpassing the fruiting calyx; nutlets 2.5 mm. long, retic- ulate-scrobiculate above, striate toward the base........ NEUE SERGE ENG T ECRAPRERRUREVERPELEAXCPDUUR 20. V. MacDougalii P. Leaves ovate-orbicular, sessile; corolla-limb 8-9 mm. broad.. MEET ToT RET Te Pe eee eae TIT ee NT 19. V. stricta O. Plants more slender, hirtellous or — strigillose with short trichomes, canescent or not; leaves linear to narrowly elliptical or spathulate; spikes at an nibus 5-6 mm. broad. R. Leaves strigillose; inflorescence not p—— kia seit eave ta ani wh Wa Oke prie. 7. V. engin R. — riiin inflorescence glandular-pubescent. 8. ; bracts lanceo- nione: sinh corolla-limb 3—4 mm. broad. . .18. V. Orcuttiana 8. Leaves hirtellous and canescent; rhachis — ie bracts ovate-acuminate; eoeolia-limb 6-8 mm V Ed MERE E sw acea les 26b. V. neomexicana var. ‘Meridia N. Leaves predominantly entire (mostly ^ie the lower ones w few salient teeth); spikes with more or less remote fruits. . 27. y. perennis M. Leaves deeply incised-dentate or pinnatifid or 3-cleft. T. Spikes not essentially bracteose; flowers readily seen; corolla-limb U. Leaf-blades, at least the lower ones, oblong-ovate or obtusely elliptie-ovate, not narrowly elongated, usually 3-cleft with the segments incised-dentate. V. Leaves petioled; pubescence various, but not rns hirsute; spikes usually stout, if slender not greatly elong W. Leaves not plicate, venation not noticeably int near the margin; spikes stout at anthesis, elongating or not in ruit. 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 261 X. Plants pilose to hirsute-villous; leaves not scabrous above; spikes elongating in fruit; bracts not WE the s i f MEER TE m . V. prostrata X. Plants sparsely hirsute; leaves scabrous E spikes occasionally elongating in fruit; bracts slightly sur- passing the calyx. 36 cos cc chases Eas oS 23. V. robusta W. Leaves more or less strongly plicate; venation a aed d whitish near the margin; spikes not stout at anthesis. . jis dint ae Capes as late a see eee ee 25. y. aplioala V. Leaves subsessile or at most very short-petioled; pubescence coarsely hirsute; spikes slender and greatly elongated. . 24. V. zutha U. Leaf-blades, at least the lower ones, oblong-lanceolate to spathu- late, narrowly elongated, usually incised-pinnatifid or incised- dentate (excl. V. plicata). Y. Plants coarse with a low somewhat compact habit, canescent- hirsute; leaves subpinnatifid, contracted into a broadly margined semiamplexicaul or petiola Z. Inflorescence somewhat glandular-hirsute; bracts lanceolate. ba dos dbi db das o id vb Mea pd ED Sie . canescens Z. Inflorescence sparsely, if at all, glandular and densely hermes bracts ovate, abruptly acuminate...................... eer ee ee Per re err a. V. canescens var. Roemeriana Y. Plants more slender with a taller E open habit, hirsute to canescent-hirtellous; leaves .pinnately cleft to incised- entate, with a narrowly margined petiolar base. AA. Corolla-limb 4 mm. broad; commissural faces extending to the tip of the nulle... ovo err cas eer RERO ICE ET Ced 26. V. neomezicana AA. Corolla-limb 6-10 mm. broad; commissural faces scarcely extending to the tip of the nutlet. BB. Plant more or less hirsute; leaves MUS cleft; bracts lanceolate. . "Tr 6a. V. neomexicana var. zylopoda BB. Plants canescent-hirtellous; leaves Rena or somewhat m eeply incised; bracts ovate................. 26b. V. neomexicana var. hirtella T. Spikes usually bracteose with somewhat foliaceous bracts at th ani of the spike; flowers inconspicuous; corolla-limb 2.5-3 m road (larger in V. plicata). CC. md delicate; leaves incised-pinnatifid to pinnately cleft, not pli- cate; inflorescence glandular-hirtellous; nutlets 1.5-2 mm. long, scrobiculate practically to the base...................4. 28. V. gracilis CC. Plants coarse; leaves incised-dentate to subpinnatifid or 3-cleft, pli- cate or an inflorescence hirsute or pun nutlets 2-2.5 mm. long, not scrobiculate to the bas DD. Leaves more or less strongly plicate, UN whitish- veined near the margin... .. r or ZU Pre LEVE D 25. V. plicata DD. oie ue plicate nor conspicuously whitish-veined near the EE. Plants j EA leaves spreading-hirsute; nutlets raised- reticulate above, striate below FF. Leaves with a subpetiolar or scialinxplixicadl base; bracts [Vor. 20 262 ANNALS OF THE MISSOURI BOTANICAL GARDEN ovate (a little longer than the oe abruptly acumin- ate, ascending.............. 29a. anescens var. Roemeriana FF. Leaves narrowed into a margined Seide: bracts linear- lanceolate (much longer than the flowers), usually re- 81. 8 8. EE E E TE V. bracteata EE. Plants hirsute-pubescent; leaves appressed-hirsute; nutlets only faintly reticulate or essentially smooth. ..... 80. V. subuligera L. Schizocarp tardily separating into 4 nutlets at maturity.......... 32. V. carnea 6. V. officinalis L. Sp. Pl. 20. 1753. V. spuria L. l.c. V. domingensis Urb. Symb. Ant. 5: 484. 1908. Stems ascending or erect, branched, glabrous or nearly so; leaves 2-7 cm. long, strigillose on both surfaces; basal and lower stem-leaves more or less ovate, narrowed below into a petiole, 1-2-pinnatifid or 3—5-cleft with parts incised; upper leaves similar but smaller and less divided; spikes paniculately disposed or in 3's or solitary, slender and elongate; bracts usually about half as long as the calyx; calyx 2-2.5 mm. long, glandular-pubescent, subtruneate, teeth minute; corolla-tube a little longer than the calyx; corolla-limb about 4 mm. broad, segments more or less rounded; nutlets trigonous, barely 2 mm. long, strongly striate, slightly reticulate above; commissural faces muricate. Distribution: Europe; introduced in waste places in eastern North America and West Indies. Specimens examined: MassaAcuusETTS: Rowley, Oakes (G). RuopkE IsnAND: Warwick Neck, 1848, Thurber (G). New York: Brooklyn, 13 Sept. 1879, von Schrenk (MBG). New Jersey: ballast, Camden, 30 Aug. 1874, Parker (G); roadside, Longacoming, Camden Co., 23 July 1867, Parker (G); Cold Spring, Cape May Co., 30 Aug. 1917, Gershoy 583 (G). PENNSYLVANIA: Philadelphia, 1844, Lea (MBG); Lancaster, 21 Aug. 1861, Porter (G); banks of Susquehanna, Harrisburg, Oct. 1852, Porter (G); York Furnace, York Co., 28 June 1899, MacElwee 873 (MBG). DEgrLAwARE: Wilmington, 1845, T'atnall (G). District OF CoruMsBia: vicinity of Washington, 23 June 1878, Ward (MBG). VinaiNIA: Parksley, 11 Sept. 1902, Norton (MBG); Bedford Co., 30 June 1870, Curtiss (MBG); Marion, 1892, Small (MBG). ORTH CAROLINA: without data, Curtis (MBG); waste grounds, Moyock, 1 July 1922, Randolph 587 (G); dry sandy bank about 3 km. west of Plymouth, 4 July 1922, Randolph 645 (G); Clarkton, 23 June 1897, Biltmore Herbarium 4762 (G, NY); Cranberry Forge, July 1895, F. Wislizenus 1214 (MBG). Sourn CAROLINA: streets in Graniteville, 22 May 1899, Eggert (MBG). Georata: Rome, Chapman (MBG). 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 263 Fronipa: without data, Rugel 121 (F, MBG). ALABAMA: Collinsville, 29 July 1897, Eggert (MBG); Attalla, 9 July 1898, Eggert (MBG). LourstANA: Port Eads, 22 Aug. 1900, Tracy & Lloyd (G, MBG, NY). TENNESSEE: along the banks of Doe River, Carter Co., 16-17 July 1891, Small & Heller 484 (G, MBG); Knoxville, July 1893, Ruth (MBG); Hollow Rock, Carroll Co., 5 Aug. 1897, m (F, MBG). West INDIE BERMUDA: Hamilton Parish, 11 July 1905, Moore 2939a (G); Flatts, 31 Aug.-20 Sept. 1905, Brown & Britton 28 (ANSP, G, NY); Flatts, 3 Aug. 1915, Collins 267 (G, NY); Bailey's Bay, 10 Feb.-9 March 1908, Brown 492 (ANSP, G, NY); St. George’s, 18 Jan. 1912, Robinson 113 (G). Santo Dominao: Angostura del Rio Yaqui, 8 May 1887, Eggers 1828 (NY), TYPE collection of V. domingensis; Culo de Maco, Prov. Azua, Aug. 1912, Fuertes 1856 (NY); Loma Rosilla, Prov. de la Vega, July 1912, Fuertes 1771 (NY); vicinity of Mission, Fonds Varettes, 12 April-4 May 1920, Leonard 3939 (G, NY). Cusa: without data, Wright 3658 (G); Vecindad de Vento, May 1906, Baker 2591 (NY, P); Playa de Marianao, 29 March 1911, Britton & Cowell 103826 (NY); near Playa de Marianao, Leon (NY), Leon & Edmunds (NY); cultivated field, Campo Florido, Prov. Havana, 13 March 1905, Curtiss 677 (ANSP, G, MBG, NY). On the whole, the specimens from Santo Domingo and Cuba differ from the typical V. officinalis in their slender and more elongate habit; the inflorescence is scarcely as glandular, the flowers are smaller, and the nutlets often do not exceed 1.5 mm. in length. Nevertheless, the Cuban specimens vary greatly in size, and Curtiss 677 is hardis separable from typical V. officinalis. Since many of the specimens are rather poor, it appears probable that they may very well represent an impoverished condition. Urban himself was somewhat uncertain of the status of his species as he appended the following note in a later publication: “ An re vera a formis V. officinalis L. separanda?' 7. V. menthaefolia Benth. Pl. Hartw. 21. 1839. V. setosa Mart. & Gal. Bull. Acad. Brux. 11?: 321. 1844. Stems decumbent or ascending, branched, sparsely and minutely hispidulous; leaves ovate, tapering at base into a margined petiole, 3-6 cm. long, deeply cleft or subincised with divisions remotely serrate-dentate, strigillose on both surfaces, somewhat pustulate above; spikes panicled, slender, elongate, compact only at the apex; bracts ovate-lanceolate, acuminate, variable in length, usually shorter than the calyx, ciliate and sparsely strigillose; calyx 2.5-3 mm. long, strigillose, sparsely (if at all) glandular, teeth minute; corolla-tube only slightly longer [Vor. 20 264 ANNALS OF THE MISSOURI BOTANICAL GARDEN than the calyx; corolla-limb about 6 mm. broad, segments more or less truncate; nutlets trigonous with convex back, 2-2.5 mm. long, striate, raised-reticulate above; commissural faces muricate. Distribution: Arizona to southern California, northwest Mexico to Oaxaca. Specimens examined: ARIZONA: north of Yuma, 26 April 1906, Jones (P). CaLrFOonNIA: San Diego, May 1852, Thurber 555 (G), June 1875, Palmer 308 (MBG), 5 May 1903, Abrams 3406 (G, MBG, NY, P), 6 July 1915, Macbride & Payson 781 (G),17 April 1918, Carlson (G), in canyons, Aug. 1918, Spencer 971 (G, P), waysides, 31 Oct. 1919, Spencer 1414 (G, P); Otay Creek, San Diego Co., 27 March 1923, Peirin 8379 (P). Mexico LOWER CALIFORNIA: Salton River Crossing, 27 April 1894, Schoenfeldt 2915 (G, NY, US); sandy roadside, Tia Juana, 1 Feb. 1920, Bartram (ANSP); ranch, 46 km. southwest of Tia Juana, 13 April 1925, Jones (P); base of Cucopa Mountains, 6 April 1905, MacDougal 168 (NY). Sonora: vicinity of Alamos, 16 March 1910, Rose, Standley & Russell 12934 (US); vicinity of Hermosillo, 6 March 1910, Rose, Standley & Russell 12451 (US); vicinity of Navojoa, 21 March 1910, Rose, Standley & Russell 13130 (US). SINALOA: collection of 1921, Ortega 4215 (US); vicinity of Fuerte, 25 March 1910, Rose, Standley & Russell 13447 (US); vicinity of San Blas, 24 March 1910, Rose, Standley & Russell 13422 (NY, US); Topolobampo, 15-25 Sept. 1897, Palmer 268 (US); near Plomosas, 18 July 1897, Rose 1763 (US). Curnuanva: valley of the San Pedro, Ortiz, 11 April 1887, Pringle 1599 (MBG). CoaHUILA: near Saltillo, 7 May 1848, Gregg 11 (MBG); near Saltillo, 14 July 1848, Gregg 265 (G, MBG); vicinity of Saltillo, May 1898, Palmer 191 (G, MBG, NY, US); vicinity of Buena Vista, 24 July 1848, Gregg 276 (MBG); valley of Parras, 11 April 1847, Gregg 406 (MBG, NY); San Lorenzo de Laguna, about 120 km. southwest of Parras, 1-10 May 1880, Palmer 1042 (ANSP, G, US). Duranco: near El Salto, 12 July 1898, Nelson 4577 (MBG, US); bottom-lands and ravines, Durango, April-Nov. 1896, Palmer 153 (F, G, MBG, NY, US), 356 (G, MBG, NY, US). San Luis Porosi: Alvarez, 28 Feb.-3 Oct. 1902, Palmer 141 in part (F, G, MBG, NY, US); San Luis Potosi, 1878, Parry & Palmer 717 (G). AGUASCALIENTES: near Aguas Calientes, 10 Oct. 1903, Rose & Painter 7799 US). GUANAJUATO: margin of stream, Sirena Mountain, 1894, Duges (G); Leon, 1839, Hartweg 175 (K Tyre, NY QUERETARO: vicinity of Queretaro, 1912, Basile 99 (US); Queretaro, 1910-13, Arsène & Agniel 10242 (F, G, MBG, US); Queretaro, July 1914, Arsène 9998 (US). Vera Cruz: Maltrata, 20 Aug. 1891, Seaton 7 (F, G, US); Huatusco, 1857, Mohr S). HipaLao: Sierra de Pachuca, 20-24 July 1905, Rose, Painter & Rose 8753 (G, NY, US); Nopala, Aug. 1913, Salazar (US); Moran, 1840, Galeotti 778 (K), TYPE collec- tion of V. setosa. Mexico: near Tlalnepantla, 6 July 1905, ps Painter & Rose 8382 (G, US); Lomas de Santa Fe, July 1928, Lyonnet (US); V ley of Mexico, 1865-66, Bourgeau 860 (US), 547 (G); Valley of Mexico, 27 June Ag Pringle 8534 (ANSP, F, G, MBG, 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 265 NY, P, US); mountains between Toluca and Mexico City, 28 June 1910, Rusby 181 (NY). More tos: La Cascada, 29 May 1901, Pringle 9529 in part (F, US). MicHoACcAN: Morelia-Rincon, Arsène 2798 (G); east of Maravatio, 30 April 1844, Gregg 823 (MBG). Oaxaca: mountains, San Juan del Estado, 18 June 1894, L. C. Smith 27 (G); Sierra de San Felipe, 20 Aug. 1894, Pringle 5715 (G); between Coixtlahuaca and Tamazulapam, 12 Nov. 1894, Nelson 1943 (US); between Las Sedas and Salome, 30 Aug. 1921, Conzatti 4207 (US). In the specimens cited from California south including Sinaloa, the inflorescence is more densely strigillose than in the collections from the southern part of Mexico, the calyces are about 1 mm. longer, with teeth strongly unequal and the subtending bracts often as long as the calyces. Although this apparently indigenous species has been known generally as V. officinalis, it has somewhat harsher pubescence and is scarcely, if at all, glandular. The fruiting calyx tends to be connivent, concealing the apex of the schizocarp rather than open and disclosing it. Perhaps these are differences only of degree and may be merely variations of V. officinalis; nevertheless, for the present it seems preferable to retain the name V. menthaefolia for the American representative. 8. V. Halei Small, Bull. Torr. Bot. Club 25: 617. 1898. V. leucanthemifolia Greene, Pittonia 5: 135. 1903. Stems usually several from a woody base, erect, ascendingly branched, glabrous or strigillose above; leaves 3-10 cm. long, strigillose on both surfaces, diverse in outline—the basal and lower stem-leaves oblong to ovate, tapering into a petiole ap- proximately as long as the blade, irregularly dentate or incised; the middle stem-leaves 1—2-pinnatifid with shorter petioles; the upper sparingly dentate or entire; spikes paniculately disposed, slender and elongate; bracts about one-half as long as the calyx, appressed, ciliate; fruiting calyx 3-3.5 mm. long, strigillose, subtruncate, the broad nerves terminating in unequal subulate teeth; corolla-tube scarcely longer than the calyx; corolla-limb 6—7 mm. broad, segments retuse; nutlets trigonous, approximately 2.5 mm. long, usually prominently striate, raised-reticulate at apex; commissural faces muricate. Distribution: Alabama to Texas, Mexico ALABAMA: without data, Buckley (MBG); p 11 May 1904, Deweys (G). [Vor. 20 266 ANNALS OF THE MISSOURI BOTANICAL GARDEN MississIPPI: cultivated grounds near coast, May 1859, Hilgard (MBG); Ocean Springs, 14 Sept. 1891, Seymour 50 (G); Ocean Springs, June 1892, Skehan 109 (G, MBG), 46 (G); Biloxi, 9 June 1900, Tracy (NY); Natchez, June 1898, Shimek (MBG). LourstaNna: without data, Riddell 1268 (NY); without data, Hale (G), 245 (NY); sandy open ground, Natchitoches, 24 April 1915, E. J. Palmer 7369 (MBG); Alex- andria, 20 Aug. 1847, Gregg (MBG); vicinity of Covington, 1920, Arsène 11831, 12242 (US); Gretna, opposite New Orleans, 5 May 1899, Ball 344 (G, MBG, NY); Madisonville, 14 May 1888, Joor (MBG); along Calcasieu River near Lake Charles, 11 April 1925, Small & Wherry 11774 (NY); Cameron, 4 July 1903, Tracy 8709 (F, G, MBG, NY); vicinity of Cameron, 4 Dec. 1910, McAtee 1953 (US). OKLAHOMA: south of Dougherty, Murray Co., 1 May 1926, Stratton 62 (MBG). Texas: Texarkana, Letterman (MBG); railroad near Polk, 13 June 1898, Eggert (MBG); between Iowa Park and Electra, Wichita Co., 20 Aug. 1921, Ferris & Duncan 3337 (MBG); near Longview, 19 April 1899, Biyori (MBG); Dallas, Reverchon 42 (G), 732 (MBG); Dallas, Jones (P); North Dallas, 27 June 1899, Eggert (MBG); Brazos, July 1843, Lindheimer 155 (77) (MBG); Tarrant Co., Ruth 108 (G, MBG, NY, US); Van Zandt, April 1929, Ezell 5699 (US); prairies near Granbury, Hood Co., 4 May 1900, Eggert (MBG); Abilene, 19 May 1902, Tracy 7996 (F, G, MBG, NY); near Comanche, 8 May 1900, Eggert (MBG); Waco, 1904, Pace 22 (MBG); Gurley, 20 April 1907, Howell 362 (US); San Augustine, 19 April 1916, E. J. Palmer 9485 (MBG); College Station, Brazos Co., 28 April 1917, E. J. Palmer 11715 (MBG); Mill Creek, April 1839, Lindheimer (MBG); near Houston, March 1842, Lindheimer 155 (106) (MBG); prairies, Houston, 1872, Hall 432 (MBG, P); Houston, 21 April 1899, Eggert (MBG); Houston, 17 April 1900, Bush 28 (MBG, US); Galveston Island, May 1843, Lindheimer (MBG); Galveston Island, 22 Sept. 1901, T'racy 7533 (F, G, MBG, NY); Alvin, 22 April 1918, Young (P); Columbia, 9 April 1899, Bush 84 (G, MBG); Columbia, 20 April 1900, Bush 77 (G, MBG, NY, US); sandy prairies, Columbia, 29 March 1914, E. J. Palmer 5044 (MBG); Austin, 1 May 1915, Young 77 (MBG); Austin, 7 April 1922, Tharp 2818 (US); flood plains of Colorado River, near Austin, 2 April 1929, Armer 5385 (US); Crab Apple Creek, Gillespie Co., Jermy 203 (MBG); Kerrville, 4 June 1916, E. J. Palmer 10037 (MBG); New Braunfels, April 1851, Lindheimer 1076 (537) (G, MBG, NY); near Bracken, 13 July 1903, Groth 75 (G, NY); Bexar to Austin, April 1828, Berlandier 322, 1592 (G); San Antonio, 1900, Wilkinson (MBG); San Antonio, 1911, Clemens & Clemens 969, 970 (P); San An- tonio, April 1922, Schulz 766 (US); Sutherland Springs, Wilson Co., Aug. 1879, Palmer 1043 (G); near Sabinal, 21 April 1925, Small & Wherry 11998 (NY); Del Rio, 20 April 1930, Jones 26229 in part (P); Uvalde, 11 May 1918, E. J. Palmer 13564 (MBG); Uvalde, 28 April 1928, E. J. Palmer 33604 (MBG, NY); Millett, 11 May 1897, Trelease (MBG); Cuero, 22 March 1907, Howell 313 (US); prairies near Goliad, 8 April 1900, Eggert (G, MBG); Refugio, 8 March 1916, E. J. Palmer 9111 (MBG); Corpus Christi, 5-12 March 1894, Heller 1419 (G, MBG, NY, US); Corpus Christi, 1 May 1913, Orcutt 5867 (MBG); Kingsville, 25 March 1920, High 52 (MBG); near Laredo, Aug. 1899, Mackenzie 96 (MBG); near Laredo, 6 April 1901, Eggert (MBG); Rio Hondo, Cameron Co., Sept. 1913, Chandler (G, MBG); Sierra Blanca, 11 April 1930, Jones 26229 in part (MBG). Mexico: TAMAULIPAS: Matamoros, Berlandier 1511, 3016 (G, MBG); vicinity of Tampico, 1-31 Jan. 1910, Palmer 78 (F, G, MBG, NY, US). A species closely related to V. menthaefolia and V. officinalis, 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 267 but readily distinguished by the diverse outline of the leaves and the somewhat more slender achenes. 9. V. riparia Raf. ex Small & Heller, Mem. Torr. Bot. Club 3:12. 1892. ? V. hastata var. B oblongifolia Nutt. Gen. 2: 40. 1818. V. urticifolia var. riparia Britton, Mem. Torr. Bot. Club 5: 276. 1894. Stems 6-15 dm. tall, erect, sparsely pubescent or glabrate, widely branched; leaves oblong to ovate, 4-12 em. long, petioled, pinnatifid or nearly tripartite toward the base, sparsely strigillose on both surfaces, venation prominent beneath; spikes paniculately disposed, slender, elongate; bracts lanceolate-ovate, as long as the calyx at anthesis, acuminate; calyx 3 mm. long, minutely glandular- pubescent, subtruncate, teeth minute; corolla-tube only slightly longer than the calyx, puberulent without; corolla-limb 3.5 mm. broad; segments more or less rounded, the middle posterior one emarginate; nutlets oblong, 2-2.5 mm. long. Distribution: New Jersey and Virginia to North Carolina (acc. to Small). Specimens examined: VirGINIA: Marion, peden Co., 1 July 1892, Small (P); Marion, Smyth Co., 6 July 1892, Small (G, M NoRTH CAROLINA: near E Caldwell Co., 3 July 1891, Small & Heller (F); near falls of Yadkin, Stanley Co., 18 Aug. 1891, Small & Heller (F). The material at hand is too scanty and too immature to give many clues to the probable relationship of this species. It would seem as nearly related to V. officinalis as to either V. hastata or V. urticifolia. It is characterized by pinnatifid or tripartite leaves, minutely glandular-pubescent inflorescence, and fruit about twice as long as thick. 10. V. Ehrenbergiana Schauer in DC. Prodr. 11: 548. 1847. Stem erect, branched, hirsute; leaves trifid with lateral lobes small, somewhat ovate with cuneate base narrowed into a mar- gined petiole, attenuate at apex, 4-8(-10) cm. long, coarsely serrate-dentate, strigillose above, hirtellous below, also somewhat paler and prominently veined; spikes paniculately disposed, very slender, remotely flowered; bracts ovate, about one-half as long as the calyx or shorter, subulate, ciliate; fruiting calyx [Vor. 20 268 ANNALS OF THE MISSOURI BOTANICAL GARDEN about 1.5 mm. long, strigillose, lobes very short, obtuse, mucro- nate; corolla inconspieuous, limb about 1 mm. broad; nutlets tigonone with convex back, 1 mm. long; commissural faces meeting sharply at right angles, almost smooth. Distribution: Mexico. Specimens examined: Mkxico: Coanuita: Saltillo, July 1880, Palmer 2037 (G). NvuEvo — moist places near Monterey, July 1888, Pringle 1948 (ANSP, F, G, MBG, NY, AN Luis ix oTOSI: Tancanhuitz, Feb. 1888, Seler 722 (G, US); Rio Verde, 17 Nov. 1910, Orcutt 5423 (MBG); Bagre, Minas de San Rafael, July 1911, Purpus 5451 (F, G, MBG, NY, US). Hipauco: Cazadero, April 1841, Liebmann 11335 (US). Vera Cruz: Wartenberg, near Tantoyuca, 1858, Ervendberg 153 (ANSP, G). PuEBrA: Pahuatlan, 12 July 1913, Salazar (US). Mexico: near Los Reyes, Ehrenberg 713 (Bot. Mus. Berl.-Dahl. TYPE, MBG phot.). In general habit this species somewhat approaches V. carolina. It differs, however, in its tripartite leaves, smaller flowers, and fruits. The eollsotion from Saltillo is much more densely hirsute on all parts, but apparently is conspecific. 11. V. carolina L. Syst. ed. 10, 852. 1759; Sp. Pl. ed. 2, 29. 1762. caroliniana Willd. Sp. Pl. 1: 119. 1798. polystachya HBK. Nov. Gen. et Sp. 2: 274. 1818. biserrata HBK. Ll. c. 275. veronicaefolia HBK. L. c. hirsuta Mart. & Gal. in Bull. Acad. Brux. 11?: 321. 1844. . mollis Mart. & Gal. l. c. 323. . paucifolia Mart. & Gal. l. c. 324, as V. pauciflora M. & G. in Walp. Rep. 6: 687. 1846-47. V. caroliniana forma or var. polystachya (Kunth) Loes. in Fedde, Rep. Sp. Nov. Veg. 9: 362. 1911. Stems erect or ascending, usually solitary, branched, hirsute- hispid; leaves lanceolate-oblong or somewhat elliptical, 3-8(-12) em. long, obtusish or acute, subsessile or narrowed into a short petiole, coarsely crenate-serrate, strigose and somewhat pustulate above, hirsute-strigose especially along the midrib and veins beneath; spikes paniculately disposed, slender, open in fruit; bracts ovate, varying in length, usually about one-half as long as Must Bis i Big Bia Me 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 269 the calyx or shorter, acuminate-subulate, somewhat spreading, ciliate; fruiting calyx approximately 2 mm. long, with the obtuse ROAS lobes connivent, hispidulous; corolla-tube scarcely exserted; corolla-limb inconspicuous, about 2 mm. broad; mature schizocarp inclosed by calyx; nutlets trigonous with convex back, hardly 1.5 mm. long, faintly striate; commissural faces meeting at right angles, almost smooth. Distribution: Arizona and Mexico to Salvador, Central America. Specimens examined: Arizona: Wilgus Ranch, Chiricahua Mountains, 12 Oct. 1907, Blumer 1783 (G, MBG, NY, US); Cave Creek, Chiricahua Mountains, 2 Sept. 1929, Harrison & ` Kearney 6144 (US); Santa Rita Mountains, 15 July 1881, Pringle (G); Fort Huachuca, July 1893, Wilcox (NY); Huachuca Mountains, 4 Sept. 1903, J pes E US); Ramsay Canyon, Huachuca Mountains, 30 Sept. 1929, Jones 25006 (G, N Nevapa: Carson Valley, Aug. 1872, Lemmon Mexico: without data, Graham (G); collection of 1844, ‘Galeotti 795A (G, US). Lower CALIFORNIA: Sierra de San Francisquito, 1 Oct. 1899, Brandegee (NY). Sonora: Santa Cruz, 22 Oct. 1893, Mearns 2627 (US); Cochuto, 1 Oct. 1890, Hartman 94 (G, NY, US); La Cruz de los Canadas, 3 Nov. 1890, Lloyd 448 (G). SiNALOA: near Colomas, foothills of Sierra Madre, 15 July 1897, Rose 1677 (US); Santa Lucia, 1919, Dehesa 1551 (US). Curavanva: Norogachi, Aug.-Nov. 1885, Palmer 364 (ANSP, G, US); San Diego Canyon, Sierra Madre Mountains, 16 Sept. 1903, Jones (P); Guayanopa Canyon, Sierra Madre Mountains, 23 Sept. 1903, Jones (P). DvnaNGo: rich bottom-lands, vicinity of Durango, Apr.-Nov. 1896, Palmer 339 (F, G, MBG, NY, US). Nayarit: west of Tepic, 31 May 1849, Gregg 1001 (MBG); Tepic, 5 Jan.-6 Feb. 1892, Palmer 2057 (US). Jauisco: Cuidad Guzman, 14 Oct. 1921, Kempton & Collins (US). GvANAJUATO: Guanajuato, 1880, Duges (G). Hipataco: Real del Monte, 15 Sept. 1910, Clokey 1865 n d ; near Moran, Humboldt & Bonpland (Par. TYPE of V. veronicaefolia, MBG p Vera Cruz: Sierra Templada in general, June-Oct. 1840, ze | 735 (K), TYPE PI of V. hirsuta; Orizaba, 1853, Mueller (NY); Orizaba, 1855, Mueller 766, 87 (NY); Orizaba, 1855-7, Botteri 180 (F, G, MBG, US); Orizaba, 1857, Mohr (US); ata Schiede 88 (NY); Jalapa, Jan. 1894, C. L. Smith 1737 (G). PuEBLA: region of Atoyac, near Puebla, 15 July 1909, Nicholas (US); Tezuitlan, 9 June 1910, Orcutt 4045 (F). Mexico: near city of Mexico, Humboldt & Bonpland (Par. TYPE of V. biserrata, MBG phot.); valley of Mexico to Tizapan, 19 Aug. 1865-66, Bourgeau 119 (G, US); Chalco region, 4 Oct. 1921, Kempton & Collins (US); near Tlalpam, 1905, Rose, Painter & Rose 8496 (US); Tlalpam, 9 Aug. 1910, Orcutt 3488 (F, US); Eslava, 15 June 1901, Pringle 9312 (F, G, MBG, NY, US); near Toluca, 18 Sept. 1889, Pringle 2813 (F, G More os: fields, La Cascada, 29 May 1901, Pringle 9529 in part (G, MBG, NY); Toro, 5 Aug. 1924, Fisher (F, MBG); Tepoztlan, autumn 1926, R. Redfield 4 (US). Micuoacan: Morelia, Coronilla, 8 Aug. 1909, Arsène 3000 (US); Quinceo, near [Vor. 20 270 ANNALS OF THE MISSOURI BOTANICAL GARDEN Morelia, 11 Nov. 1909, Arsène (US); on slope of voleanie mountain, Xorullo, Hum- boldt & Bonpland (Par. TYPE of V. polystachya, MBG phot.). Corma: Colima, 9 Jan.-6 Feb. 1891, Palmer 1156 (G, NY, US). Oaxaca: mountains near Oaxaca, June-Oet. 1840, Galeotti 737 (K), TYPE collec- tion of V. mollis; Sierra de San Felipe, 12 Sept. 1894, Pringle 4892 (ANSP, G, MBG, NY, US); Sierra de San Felipe, 2 Oct. 1894, C. L. Smith 224 (MBG, NY, US); vicinity of La Parada, 19 Aug. 1894, Nelson 1021 (US); Cuyamecalco, 5 Sept. 1895, L. C. Smith 634 (G). CENTRAL AMERICA: GUATEMALA: uncultivated places, Oct. 1865, Bernoulli 127 (N Y); without locality, Sept. 1927, Morales 786 (US); without locality, 1892, Heyde 120, 530, 477 (US); San Siguan, Dept. Quiche, April 1892, Heyde & Lua 3018 (ANSP, MBG, US); Coban, Dept. Alta Verapaz, Nov. 1902, von Tuerckheim 8442 (F, G, NY, US); Coban, Dec. 1906, von Tuerckheim IT 651 (F, MBG); Tactic, Dept. Alta Verapaz, May 1886, von Tuerckheim 913 (G, US); vicinity of Los Amates, Dept. Izabal, 24 May 1922, Standley 24443 (MBG, US); Proderos, Dept. Guatemala, July 1921, Tonduz 628 (US); Chilloui, Dept. Guatemala, 20 June 1921, Rojos 44 (US); Volean Pacaya, Dept. Amatitlan, July 1892, Shannon 3638 (US); San Vincente, Pacaya, Dept. Amatitlan, 5 May 1921, Tonduz 488 (US); San Lucas, Dept. Solola, 16 Feb. 1906, Kellerman 5825 (US); San Pedro, 1913, Tejada 72 (US); Tejutla, 1913, Tejada 5? (US); Santa Rosa, Dept. Santa Rosa, May 1892, Heyde & Lux 3019 (G in part, NY, US); vicinity of Siguatepeque, Dept. Comayagua, 14-27 Feb. 1928, Standley 55944 (F, US). SALVADOR: vicinity of Apastepeque, Dept. San Vincente, 4 March 1922, Standley 21331 (G, MBG, US); vicinity of San Salvador, Renson 175 (NY, US); Plazuela de Aculhaca, San Salvador, Dec. 1906, Velasco 8999 (US); Amatepeque Hill, near San Salvador, 2 Feb. 1907, Pittier 1909 (US); San Salvador, 1922, Calderon 794 (G, US); vicinity of San Salvador, 30 March-24 April 1922, Standley 22414 (G, NY, US); vicinity of Santa Tecla, Dept. La Libertad, 10 April 1922, Standley 23083 (G, US) Finca San Nicolas, 1923, Choussy 36 (US). This species is probably most nearly related to V. urticifolia and V. scabra, but is easily set apart from them by its subsessile or very short-petiolate leaves; moreover, their areas of distribu- tion are practically distinct. The collections cited appear to be conspecific, notwithstanding the fact that they vary greatly in size, serration of leaves, and length of floral bracts. The speci- mens from the northern part of the range are slightly coarser, with leaves more remotely crenate-serrate, yet they match very closely Pringle 4892, from the Sierra de San Felipe, Oaxaca. In the latter, however, the leaves are more densely pubescent on the under surface. The specimens from Guatemala and Sal- vador have somewhat crowded spikes and are, as a whole, smaller in every way and more densely hispidulous. However, since most of the collections were made from December to Febru- ary, it seems as if this more compact habit might be a seasonal rather than a geographic variation. . , 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 271 A specimen labelled *'Inter Tallahassee et St. Marks, Florida, legit Rugel, April-June 1843" appears in the herbarium of the Missouri Botanical Garden. As this is the only representative of V. carolina seen from Florida, and as it is so far from the natural range of the species, it would appear to be an error in labelling. Another collection worthy of mention is Nelson 752. It is almost glabrous, and the lobes of the corolla are emarginate, a rather unusual feature in this series. The lack of pubescence suggests V. longifolia, but unfortunately none of the available material of the species is in sufficiently good condition to reveal the character of the corolla. Although the general habit of the collection is that of V. carolina, it 1s rather doubtfully con- specific. 12. V. recta HBK. Nov. Gen. et Sp. 2: 277. 1818. V. caroliniana forma or var. recta (Kunth) Loes. in Fedde, Rep. Sp. Nov. Veg. 9: 362. 1911. Stem erect, branched, hispidulous, reddish; leaves lanceolate- oblong to ovate with cuneate base narrowed into a short margined petiole, 4-8 em. long, acute, serrate-dentate, reticulately veined, strigillose on both salare, minutely pustulate above; spikes numerous, congested at the apices of stems and ee short- cylindrical, usually dense but not strict; bracts a little shorter than the calyx, ovate-lanceolate, acuminate, hispidulous; fruiting calyx about 2 mm. long, hispidulous, lobes short, obtuse, mu- cronate, commonly connivent; corolla-tube scarcely exserted; corolla-limb inconspicuous; nutlets trigonous with convex back, 1.5-2 mm. long, faintly striate; commissural faces meeting at right angles, practically smooth. Distribution: Mexico. Specimens examined: MExico: Hipauao: Real del Monte, 12 July 1913, Salazar (US); between Pachuca and Cerro Ventoso, Humboldt & Bonpland 4066 (PAR. TYPE, MBG phot.). Pursta: Huitzuilzilapam, 9 June 1910, Orcutt 3950 (F). Mexico: Cima Station, 30 Aug. 1905, Pringle 13597 (G, US); Toluca, Berlandier 1222 (US). ORELOS: Tres Marias, 4 July 1901, Rose & Hay 5310 (US); Toro, 5 Aug. 1924, Fisher 320 (MBG, US). Oaxaca: vicinity of Cerro San Felipe, 1894, Nelson 1105 (US); Sierra de San Felipe, 9 Aug. 1894, Pringle 4769 (ANSP, F, G, MBG, NY, US). [Vor. 20 272 ANNALS OF THE MISSOURI BOTANICAL GARDEN Verbena recta, generally determined in the material at hand as “aff. V. polystachya HBK.," is easily distinguished from this closely related species by the short hispidulous pubescence, the commonly broader leaves, and the compact cylindrical congested spikes. The floral characters of both are similar, although generally the fruiting calyx of V. carolina is more acutely ovoid and the subtending bract is much sharper at the apex. 13. V. longifolia Mart. & Gal. Bull. Acad. Brux. 11?: 323. 1844 Stems erect, tall, glabrous, obtusely 4-angled; leaves lanceolate to elongate-elliptie, 10-12 cm. long, the upper somewhat smaller, decussate, short-petiolate, acutely serrate from below the middle to the apex, appressed-pubescent or very short-strigillose on both surfaces, venation prominent beneath; spikes paniculately disposed, elongate, slender, open in fruit, glabrous; bracts ovate, about one-half as long as the calyx, acute-acuminate, minutely ciliate; calyx 2 mm. long, with the obtuse lobes connivent over the schizocarp, practically glabrous; corolla-tube scarcely pro- truding beyond the calyx; corolla-limb inconspicuous; nutlets trigonous, hardly 1.5 mm. long, smooth or faintly striate; com- missural faces smooth. Distribution: southern Mexico. Cami ens examined: Mex! VERA ‘Cavs: Colipa, March 1841, Liebmann 11318 (US); near Chila, April 1888, Seler 724 (G, US). PukBLA: in vicinity of San Luis Tultitlanapa, July 1908, Purpus 3406 (F, G, MBG, NY, US). Morevos: San Anton, near Cuernavaca, 14 Oct. 1904, Seler 4194 (G, US); Cuer- navaca, 28 Oct. 1905, Seler 4347 (G, US). Oaxaca: Tehuantepec, 19 April 1910, Orcutt 3321 (F, MBG, US). A rather singular species combining the foliar characters of V. litoralis with the inflorescence characters of V. carolina. 14. V. scabra Vahl, Eclog. Am. 2: 2. 1798, not V. scabra Marnock, Floricult. Mag. 5: 87, pl. 54, fig. 2. 1840-41. Stem 1 m. more or less tall, erect, solitary, simple or branched, hispidulous; leaves ovate to elongate-ovate, 3-10(-13) em. long, 2.5-5 em. broad, petiolate, serrate-dentate, acute or obtusish, 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 273 scabrous and commonly strigillose above, less scabrous and some- what paler beneath, also hispidulous along the veins; spikes paniculately disposed, slender, pedunculate, copiously and closely flowered; bracts ovate-acuminate, about half as long as the calyx, hispidulous; fruiting calyx 2.5-3 mm. long, hispidulous, ovoid, with the somewhat unequal lobes acutely connivent, diverging from the rhachis of the spike 45° or more; corolla-tube scarcely, if at all, longer than the calyx; corolla-limb about 2 mm. wide, lobes obtuse; anthers glandless; stigmatic surface midway between two almost equal obtusish sterile lobes; nutlets trigonous, 1-1.3 mm. long, faintly striate, reticulate above; commissural faces extending to the top of the schizocarp, meeting sharply at right angles, muriculate. Distribution: North Carolina, Florida west to California, south into northwestern Mexico; West Indies. Specimens examined: NonTH CAROLINA: moist rich soil, open swamp land, Wilmington, 25 July 1922, Randolph 1012 (G). LORIDA: without locality, 1842-49, Rugel 156 (F, MBG); common in everglades, 27 Aug. 1925, O'Neill (MBG); vicinity of Mayport and Jacksonville, 1870—76, Keeler (NY); near Jacksonville, 25 June-13 July 1894, Curtiss 5111 (MBG, NY); Lake City, 21 July 1893, Quaintance (MBG); vicinity of Eustis, Lake Co., 1-15 July 1894, Nash 1248 (G, MBG, NY); low ground, Orange Co., 22 July 1902, Fredholm 5416 (G, MBG, P, US); open low ground, near Lake Okeechobee, 20 May 1925, E. J. Palmer 27462 (MBG); east shore of Lake Okeechobee, 11-25 Nov. 1913, Small & Small 4337 (NY); east shore of Observation Island, 11-25 Nov. 1913, Small & Small 4403 (NY); hammocks along Taylor Creek, 11-25 Nov. 1913, Small & Small 4341 (NY); Fort Lauderdale, 19-25 Nov. 1903, Small & Carter 1072 (F, NY); moist sandy soil, Hollywood, Broward Co., 10 Feb. 1930, Moldenke 591, 599a (MBG, NY); everglades west of Miami, 3 Nov.-7 Dec. 1912, Small 4020 (NY); hammocks near Miami River, 26 Nov.-20 Dec. 1913, Small & Small 4520 (NY); Biscayne Bay-Indian River, Palmer 397 (G, MBG); Myers, Lee Co., July-Aug. 1900, Hitchcock 269 (MBG); sandy ditch, south of Estero, Lee Co., 14 April 1930, Moldenke 972 (MBG, NY); Sanibel Island, 18 May 1901, Tracy (NY); Manatee, 8 May 1900, Tracy 6652 (G, MBG, NY); Hernando Co., June-July 1898, Hitchcock (MBG); near Tallahassee, Berg (NY); shore of Dog Island Sound, near Lanark, 26 July 1920, Harper 242 (NY, US); streets of Apa[lachicola], Chapman (MBG). Mississipri: Scranton, Jackson Co., 5-6 Aug. 1896, Pollard 1191 (G, MBG, NY); Ocean Springs, 13 Aug. 1895, Skehan (MBG). LouistANA: South Pass, 20 Aug. 1900, Tracy & Lloyd 22 (G, MBG); New Orleans, Riddell (G); New Orleans, Drummond (NY); New Orleans, 21 Oct. 1885, Joor (MBG) Texas: without locality, July 1851, Lindheimer 618 (MBG); Beaumont, Jefferson Co., 11 Sept. 1916, E. J. Palmer 10692 (MBG); Beaumont, Sept. 1904, Kirn 2139 (P); Houston, July 1842, Lindheimer (MBG); Comanche Springs, New Braunfels, [Vor 20 274 ANNALS OF THE MISSOURI BOTANICAL GARDEN etc., Aug. 1851, Lindheimer 1077 (G, MBG, NY); Leon River, Oct. 1850, Bigelow (G); San I and vicinity, 1898, Stanfield (NY); New Braunfels, 1847, Lind- heimer (MBG). Arizona: north of Rice, Apache Reservation, 4 Oct. 1927, Harrison 4897 (US); Santa Cruz River, Tucson, 11 May 1881, Pringle (G); banks of Santa Cruz River, near Tucson, 18 July 1884, Pringle (ANSP, F, G, NY, P, US) CaLrFORNIA: without locality, Wallace (G); Cienega, on Santa Monica R. R., yon 6 (G); Los Angeles, 1892, Davidson (G); near Pasadena, 17 Dec. 1892, Mc- Clatchie (N Y); San Bernardino, Oct. 1891, Parish (MBG), 5338 (NY); San Bernar- dino Valley, Parish & Parish 1043 (F, MBG); San Bernardino Valley, 25 July 1909, Parish 7149 (P); 3 km. south of San Bernardino, 2 Sept. 1924, Johnston (P); Upland, July 1917, Johnston (P); 8 km. southeast of Chino, Munz & Johnston 11289 (P); Cocomozo Mountains, Aug. 1881, Parish & Parish 11143 (G). Mexico: Lower CALIFORNIA: San Tomas, 15 July 1885, Orcutt 1802 (MBG). CoanuiLA: Soledad, a section of low mountains with few oaks, 40 km. southwest from Monoclova, 9-19 Sept. 1880, Palmer 1040 (ANSP, US); San Lorenzo de Laguna, 120 km. southwest of Parras, May 1880, Palmer 1040 (G) West INDIES: Cuna: without locality, 1865, Wright 3659 (G); Santa Rosalia, 2 Aug. 1895, Combs 389 (G, MBG, NY); valley of San Juan, vicinity of Matanzas, 14 March 1903, Britton, Britton & Shafer 292 (NY); Havana, 15 Oct. 1908, Leon 685 (NY); Havana, 21 April 1919, Leon & Edmunds 8719 (NY); Rio Arimao, 22 March 1910, Britton & Wilson 5771 (NY); Isle of Pines, 8 March 1916, Britton, Wilson & Leon 15256 (NY); Pinar del Rio, 7 April 1924, Krig 3187 (NY). Jamaica: Port Antonio, Dec. 1890, Hitchcock (MBG); Port Antonio, Aug. 1910, Lang 102 (ANSP); a near Black River, 12 Sept. 1907, Harris 9987 (NY); road- side, Hopeton, 13-22 Sept. 1907, Britton 1527 (NY); Och Rios and vicinity, 4 April 1908, Britton & Hollick E (NY); banks of Cabaritta River, Meylersfield, 16 Dec. 1914, Harris 11808 (G, MBG, NY). Harti: Prov. Barahona, Aug. 1910, Fuertes 391 (NY); Prov. La Vega, July 1912, Fuertes 1758 (NY); Miragoane and vicinity, July 1927, Eyerdam 201, 432 (G). Porto Rico: without locality, Read (ANSP); Bayamon, 22 March 1885, Sintenis 1074 (MBG, NY); Aybonito, 3 Nov. 1885, Sintenis 2010 (ANSP, P); Vega Baja, 5 Nov. 1913, Stevens & Hess 4260 (NY); El Teudal, Coamo River, 13 Feb. 1922, Britton, Britton & Brown 6019 (NY); border of Mary Lake, near Luquillo, 3 April 1922, Britton, Britton & Brown 70387 (NY). Brrmupa: Smith's Parish, 12 July 1905, Moore 2947 (G, NY); Hamilton Parish, 9 July 1905, Moore 2874 (G, NY); cultivated land north of Hamilton Parish, 31 ug.-20 Sept. 1905, Brown & Britton 878 (ANSP, G, NY); Washington Sound, Sept. 1913, Brown & Britton 1681 (ANSP, NY); roadside, 3 Aug. 1913, Collins 268 (G, NY Verbena scabra, often confused with V. urticifolia, is easily separable by the very scabrous upper surface of the leaf and the rather conspicuous divergence of the fruit from the rhachis of the spike. A quite distinctive feature is the position of the stigmatie surface apparently between two almost equal sterile 1933] PERRY— NORTH AMERICAN SPECIES OF VERBENA 275 style-lobes. This condition is somewhat approached in V. carolina, but in the other species examined the second sterile lobe has never been so definitely developed. 15. V. urticifolia L. Sp. Pl. 20. 1753. V. diffusa Poir. in Lam. Encyc. 8: 550. 1808. V. diffusa Desf. ex Spreng. in L. Syst. 2: 748. 1825. V. urticifolia var. simplex Farwell, Papers Mich. Acad. Sci. 3: 108. 4. Erect herb 5-15 dm. tall; stem solitary, simple or more often branching from near the base, hirtellous to almost glabrous; leaves 8-20 cm. long, petiolate, blades broadly lanceolate to oblong-ovate, with rounded base decurrent on the petiole, short- acuminate or acute, coarsely and somewhat doubly crenate- serrate, hirtellous or glabrate on both surfaces, often minutely pustulate above; spikes paniculately disposed, slender, peduncu- late, more or less sparsely flowered; bracts ovate-acuminate, very short, ciliate; calyx 2 mm. long, pubescent particularly along the nerves, lobes obtuse, not connivent, teeth short, subulate, subequal; corolla-tube scarcely exserted; corolla-limb about 2 mm. wide, lobes obtuse; mature schizocarp exposed at the distal end; nutlets trigonous with a convex back, about 2 mm. long, faintly striate; commissural faces meeting at right angles, almost smooth. Distribution: Quebec and Ontario, Maine to Nebraska and southward. pecimens examined: UEBEC: vicinity of Longueuil, Aug. 1916, Victorin 3124 (MBG); Cape Tour- mente, 40 km. below Quebec, 12 Aug. 1922, Victorin 15750 (G). Ontario: Casselman, 24 Aug. 1891, Scott (G); Niagara Falls, 1830, ? Torrey 154 (NY). Maine: North Berwick, 25 Sept. 1897, Fernald & Parlin 928 (G). New HampsuireE: Walpole, 28 July 1901, Williams (G); Ashuelot, 1 Aug. 1898, Robinson 559 (G). Vermont: East Middlebury, 14 July 1908, Williams (G); Rutland, 31 Aug. 1899, Eggleston 1531 (G); Manchester, 21 July 1898, Day 147 (G); Brattleboro, 2 Aug. 898, Robinson 134 (G). ASSACHUSETTS: Lowell, 28 July 1927, Beattie (P); slopes above Connecticut River, Holyoke, 7 Sept. 1926, Seymour 601 (MBG, NY); Waverley, 10 Aug. 1878, Lane (G); Belmont, 27 Sept. 1891, Deane (G); Milton, 3 Aug. 1887, Kennedy (G); South Framingham, 25 July 1890, Sturtevant (MBG); Dedham, 22 Aug. 1897, Williams (G); Adams, 24 Aug. 1901, Day ?9 (G); Manchester, 18 Aug. 1895, Williams (G); Monson, Aug. 1897, Morris (MBG); Nonquit, 1888, Sturtevant (G). [Vor. 20 276 ANNALS OF THE MISSOURI BOTANICAL GARDEN Ruope IsrnAND: dry roadside banks and thickets northeast of Great Salt Pond, Block Island, 20 Aug. 1913, Fernald, Hunnewell & Long 10262 (G); Providence, July 1844, Thurber (G); Providence, 20 July 1892, J. F. Collins (G). Connecticut: in damp woods, Wethersfield, Wright (G); Hartford meadow, 3 Aug. 1882, Wright (G); New Haven, 1858, Eaton (MBG). New York: open pasture, Canton, 18 July 1914, Phelps 800 (G); Bolton Landing, Lake George, 11 Aug. 1906, Seler 4560 (G); Vaughns, north of Hudson Falls, 8 Aug. 1912, Burnham (G); near station, Ithaca, 4 Aug. 1915, Hames & McDaniels 4870 (G); Fall Creek, Ithaca, 12 July 1916, Munz 616 (P); Six Mile Creek, Ithaca, 8 July 1878, Trelease (MBG); Fleischmanns, Delaware Co., 30 July 1892, von Schrenk (MBG); Port Chester, 3 Sept. 1886, Stabler (G). New Jersey: Green Pond, Morris Co., 21 Sept. 1886, Britton (NY); Denmark Pond, 7 Aug. 1910, Mackenzie 4746 (G, MBG); Watchung, Somerset Co., 7 July 1930, Moldenke 1330 (MBG, NY); South Amboy, July 1892, Halsted 175 (G, NY). PENNSYLVANIA: about Tunkhannock, 18 July 1929, Osterhout (P); Meadville, Aug. 1893, Curtis (P); Harrisburg, 20 July 1888, Small (G); Lancaster, 29 Aug. 1900, Heller (G); woodlands, Angora, 9 July 1899, MacElwee 882 (MBG); near Philadel- phia, 29 July 1871, Redfield (MBG); near Philadelphia, 1890, Greenman 1376 (G, MBG). ELAWARE: marsh along Delaware River, south of Newcastle, 2 Aug. 1923, Tidestrom 11572 (G). ManvLAND: College Park, Symons (MBG); Chesapeake City, 11 July 1923, Tidestrom 11413 (G) District or COLUMBIA: gravelly and cindery artificial soil, northeast Washington, 24 July 1893, Boettcher 226 (G, MBG); Anacostia, 19 July 1893, Boettcher 176 (G, MBG); Brookland, 19 July 1912, Holm (MBG); near Long Bridge, 11 July 1891, Blanchard (MBG) Vrana: Chisels Run, west of Williamsburg, 22 June 1921, Grimes 3794 (NY); Reed Creek, Wythe Co., 23 July 1892, Small (MBG, NY); Marion, Smythe Co., 20 July 1892, Small (MBG Nortu CAROLINA: Onobuim Station, 13 Sept. 1885, Britton (NY). Sourn CAROLINA: near Anderson, 8 July 1919, Davis 9117, 8458 (MBG); Santee Canal, Ravenel (G); Charleston, 19 Aug. 1859, Gibbes (NY). FrLomipa: Apalachicola, Chapman (MBG); near Tallahassee, May 1843, Rugel (MBG). ALABAMA: Tensaw, 22 Aug. 1904, Tracy 8037 (G, MBG, NY). Mississipri: Woodville, 7 Sept. 1887, Joor (MBG); Taylorville, 2 Aug. 1903, Tracy (NY). LourstaNa: without data, Hale (G); Natchitoches, 14 June 1915, E. J. Palmer 8000 (MBG); waste pps Alexandria, 5 June 1899, Ball 556 (G, MBG); Lockport, 23 Oct. 1919, Guidroz 3 Ouro: Shaker Pond, near RE 8 July 1896, Greenman 1380 (MBG); Oberlin, June 1895, Hicks (MBG); Oxford, 6 June 1910, Overholis (MBG); Cincinnati, 18 Oct. 1930, Stephenson (MBG) West Virama: Tygart River valley, above Huttonsville, 23 Sept. 1904, Green- man 244 (G); banks of Blackwater River, Hendricks, 10 Sept. 1904, Greenman 246 (G); low ground west of Salt Pond, 28 June 1922, laa c 474 (G). Inp1ANA: East Chicago, 10 Aug. 1910, Lansing 28 Kentucky: Stamping Ground, 1 July 1930, vw 308 (MBG); Rockdale, July 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 277 1928, Runyon 1260 (US); Nicholasville, 18 July 1923, McFarland 92 (MBG); Bowling Green, Aug. 1896, Price (MBG); Calvert City, 16-19 June 1909, Eggleston 4841 (NY). TENNESSEE: borders of woods, Knoxville, July 1898, Ruth 833 (MBG), 788 (NY). Wisconsin: St. Croix Falls, July 1899, Baker (P); Green Bay, 7 July 1878, Schuette (NY InumNors: Bowmanville, Cook Co., 7 July 1896, Chase (MBG); Chicago, 19 Aug. 1907, Greenman 1877 (MBG); Starved Rock, June-Sept. 1921, Thone 50 (MBG); Urbana, 6 July 1899, Gleason 718 (G); Olney, 17 July 1927, Ridgway 2831 (MBG); near Alton, 24 July 1927, Bucholz (MBG); Freeburg, Aug. 1917, Hertel (MBG). MINNESOTA: open grounds, Collegeville, Stearns Co., 29 July 1912, Chandonnet (MBG); near Minneapolis, July 1891, Aiton 8467 (P); meadow, Northfield, 1912, Goldsmith 119 (NY); Houston Co., July 1912, Freiberg (MBG). Iowa: Black Hawk Co., 22 July 1929, Burk 591 (MBG); Ames, Hitchcock (MBG); Ames, Sept. 1873, Bessey (G); Ames, Sept. 1909, Campbell 67 (G, MBG); Bentons- port, July 1920, Graves 1691 (MBG); Decatur Co., 13 July 1903, Anderson (MBG). Missounr: Hannibal, Marion Co., Davis 73, 127 4a, 3458, 8591, 4450, 4565 (MBG); Ilaseo, Ralls Co., 13 Aug. 1915, Davis 3852 (MBG); Aberdeen, 11 Sept. 1911, Davis 950 (MBG, NY); St. Louis, Aug. 1838, Riehl 135 (MBG, NY); St. Louis, 11 Aug. 1894, Glatfelter (MBG); St. Louis, Sept. 1899, Baker (P); Creve Coeur Lake, 24 Aug. 1930, Kellogg 16277 (MBG); Fern Glen, St. Louis Co., 14 July 1906, Johnson (MBG); Allenton, Letterman (MBG); Allenton, 27 July 1884, Kellogg (MBG); Kimmswick, 15 July 1885, F. Wislizenus 282 (MBG); Victoria, 6 July 1890, Hitchcock (MBG); Shepard Mountain, near Arcadia, 22 July 1915, Greenman 3766 (MBG); Poplar Bluff, 14 Aug. 1892, Dewart 15 (MBG); Jefferson City, July 1866, Knause (MBG); Brumley to Bagnell, 21 Sept. 1897, T'release 719 (MBG); Springfield, 31 July 1892, Dewart 18 (MBG); Willard, 9 July 1919, Blankinship (P); Swan, 24 Sept. 1899, Bush 452 (MBG); Jackson Co., 7 Aug. 1893, Bush (MBG); bottoms, Cass Co., 24 July 1864, Broadhead (MBG); Webb City, 4 July 1902, E. J. Palmer 198 (MBG); gravel bars, Butler Creek, Noel, 9 Sept. 1913, E. J. Palmer 4229 (MBG ARKANSAS: waste places, near Corning, 21 Aug. 1896, Eggert (MBG); near McNab, 5 Oct. 1923, Greenman 4412 (MBG); Fulton, 18 Sept. 1900, Bush 884 (MBG). H Dakora: Brookings Co., 1903, Hia (MBG); Mitchell, July 1903, Hoffstetter (MBG); sand hills near La Creek P. O., Bennett Co., 14 Aug. 1911, Visher 2263 (NY); Vermilion, 11 Sept. 1911, Visher (MBG P NEBRASKA: Dickson’s Bluffs, on Missouri River, 12 July 1853-4, Hayden (MBG); near Plummer Ford, Dismal River, 8 Aug. 1893, Rydberg 1716 (NY); Lincoln, Aug. 1899, Hedgcock (MBG); Franklin, 1893, Laybourne (MBG). Kansas: Riley Co., 16 Aug. 1892, Waugh (MBG); low woods, Riley Co., 12 July 1895, Norton 389 (G, MBG, NY); Manhattan, Sept. 1893, Norton (MBG); vicinity of Arkansas City, south of Arkansas River, 2 July 1929, Rydberg & Imler 482 (NY). OKLAHOMA: near Alva, 11 July 1913, Stevens 1673 (G, MBG, NY); near Tonkawa, 4 Aug. 1913, Stevens 1821 (G); 16 km. south of Stillwater, 14 July 1927, Stratton 158 MBG); near Cleo, 19 July 1913, Stevens 1742 (G); vicinity of Fort Sill, 3 July 1916, Clemens 11749 (MBG); Davis, 10 July 1916, Emig 716 (MBG); near Grant, 4 June 1916, Houghton 4015 (G); “Arkansas,” 23 July 1894, Bush 432 (MBG). Texas: Tarrant Co., 10 Aug. 1926, Killiam 6933 (US); Dallas, Aug. 1876, Rever- chon (G), ?34 (MBG); Houston, May 1842, Lindheimer 195 (MBG); Columbia, 13 Oet. 1900, Bush 1482 (MBG) Verbena urticifolia is readily distinguished by petiolate and [Vor. 20 278 ANNALS OF THE MISSOURI BOTANICAL GARDEN rarely scabrous leaves, remotely flowered slender spikes, and unconnivent fruiting calyces. On account of its close resemblance to V. scabra the two were confused, and V. urticifolia formerly was believed to be a native of tropieal America. It would now appear to be indigenous in the Middle Eastern States, establishing itself as à weed in various places. 16. V. hastata L. Sp. Pl. 20. 1753. V. pinnatifida Lam. Tab. Encyc. 1: 57. 1791, as V. hastata y Poir in Lam. Encyc. 8: 546. 1808. V. paniculata Lam. Tab. Encyc. 1:57. 1791. V. paniculata var. pinnatifida Schauer in DC. Prodr. 11: 546. 1847. V. hastata var. pinnatifida Gray, Syn. Fl. N. Am. 2!: 330. 1878; Britton, Mem. Torr. Bot. Club 5: 276. 1894. V. hastala var. paniculata Farwell, Ann. Rept. Comm. Parks and Boul. Detroit 11: 82. 1900. V. hastata f. rosea Cheney, Rhodora 4: 245. 1902. V. hastata var. paniculata f. rosea Farwell, Papers Mich. Acad. Sci. 2:37. 1923. Stems 4-15 dm. tall, branched above, rough-pubescent with short antrorse hairs; leaves lanceolate, oblong-lanceolate or occasionally ovate-lanceolate, 5-15(-18) em. long, gradually acuminate, petiolate, coarsely or incised serrate, often hastately 3-lobed at base, veins impressed above, rough-pubescent on both surfaces; spikes strict, usually numerous in a panicle, pedunculate, compact; bracts lanceolate-subulate, commonly a little shorter than the calyx; calyx 2.5-3 mm. long, pubescent, lobes acute with short subulate tips, more or less connivent; corolla-tube somewhat longer than the calyx, pubescent without; corolla-limb 3-4.5 mm. broad; nutlets about 2 mm. long, nearly smooth or very faintly striate; commissural faces muriculate or almost smooth. LN Nova Seotia to British Columbia, south throughout the United Sta sail Scotta: bank of Five-mile River, Hants Co., 19 July 1920, Pease & Long 22350 (G). New Brunswick: Nepisiguit, 30 July 1873, Fowler (MBG); alluvium of Nash- waak River, Nashwaak, 31 July 1922, Fernald & Pease 25247 (G Quesec: La Trappe, 5 Aug. 1926, Louis-Marie 141 (G); swamp, ‘Ascot, Sherbrooke 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 279 Co., 20 July 1923, koh (G); sand-plains, northwest of Three Rivers, 1 Aug. 1923, Chenin & Knowlton (G). Ontario: Ashdod, ss July ee Fowler (MBG); Kingston, Sept. 1897, Fowler (G); Aylmer, 5 Aug. 1912, Fisher (MBG). Marne: gravelly shore of St. John River, St. Francis, 5 Aug. 1893, Fernald (G); Foxcroft, 18 July 1895, Fernald 296 (G, MBG); North Berwick, 28 Aug. 1891, Parlin (G); Eliot, 27 Aug. 1895, Williams (G). New Hampsuire: bank of “Connt” [Connecticut] River, 1894, Jesup (G); Suc- cess, Coos Co., 24 Aug. 1908, Moore 4316 (G); roadside, Walpole, 28 July 1901, Williams (G); roadside, East Jaffrey, 10 July 1897, Robinson 196 (G). Vermont: Weybridge, 8 July 1908, Williams (G); fields, Brandon, 7 July 1921, Dutton (MBG); Mechanicsville, 6 Aug. 1906, J. W. Anderson (G); Manchester, 6 July 1898, Day 149 (G); Brattleboro, 2 Aug. 1898, Robinson 153 (G). MassACHUSETTS: by Merrimac River, Lowell, 8 Aug. 1927, Beattie (P); Melrose, 17 July 1876, Morong (MBG); Waverly, 10 Aug. 1878, Lane (G); South Framingham, 11 Aug. 1888, Sturtevant (MBG); Jamaica Plain, Faxon (G); Wigwam Pond, Dedham, 6 Oct. 1901, Williams (G): Nine Mile Pond, Cape Cod, 4 Sept. 1898, Greenman 398 (G, MBG); shore of Wequawket Pond, Centreville, 6 Sept. 1896, Williams (G); Centreville, 20 Aug. 1902, Cheney (G); Nonquit, 29 Aug. 1888, Sturtevant (MBG); Fisher’s Pond, West Tisbury, Martha’s Vineyard, 4 Sept. 1917, Seymour 1324 (G); Monson, July 1897, Morris (MBG); near Water Shop Pond, Springfield, 5 Aug. 1924, Seymour 511 (MBG); South Worthington, Hampshire Co., 17 Aug. 1912, Robinson RHODE IsrAND: oec Olney (G); Providence, 16 July 1892, Collins (G); dense wet thickets at borders of sphagnous swamps southwest of Harbor Pond, Block Island, 19 Aug. 1913, Fernald & Long 10263 (G). Connecticut: waste ground near river, Stafford, 9 July 1922, Weatherby 5076 (MBG); Southington, 10 Aug. 1897, Bissell (MBG); Southington, 31 July 1898, Andrews (G); Oxford, 11 Aug. 1888, Harger (G); New Haven, 1858, Eaton (MBG). New York: Granville, 22 Aug. 1924, Drushel 6351 (MBG); Port Chester, 3 Sept. 1886, Stabler (G); Willets Neck, Long Island, 8 Aug. 1853, Hexamer & Maier (G); Peconic, 15 Aug. 1895, von Schrenk (MBG); Fleischmanns, Delaware Co., 15 Jul 1892, von Schrenk (MBG); Dryden, 3 Sept. 1915, Dean & Thomas 4874 (G); low flats north of the PRÉ op 16 July 1915, MacDaniels 4873 (G); near Bool’s Brook, Ithaca, Munz 617 (P); Keene Valley, Essex Co., 1 Aug. 1891, von Schrenk (MBG); Canton, 17 E 1914, Phelps 799 (G). New JEnsEY: Fort Lee, 30 July 1921, Rydberg (P); Plainfield, 9 Aug. 1877, T're- lease (MBG); Watehühg: 11 July 1930, Moldenke 1339 (MBG); New Market, July 1892, Kelsey 173 (G). PENNSYLVANIA: banks of Susquehanna River at Perdix, 4 Aug. 1926, Heller 14222 (MBG); near Harrisburg, 13 July 1888, Small (G); near Philadelphia, 1889, Greenman 1273 (G); shores of Schuylkill River, Fairmount Park, 22 July 1871, Redfield (MBG). DELAWARE: along Delaware River, south of Newcastle, 2 Aug. 1923, Tidestrom 11551 (G). Maryan: below Havre de iex Park, 2 Aug. 1902, Shull 156 (MBG); Ellicott City, 3 Aug. 1916, Arséne 672 (M District or COLUMBIA: near ed Bridge, Washington, 11 July 1891, Blanchard (MBG). Virainia: Langley, 10 Aug. 1901, W. Palmer (US); South Fork, Holston River, at St. Clair’s Bottom, Smyth Co., 30 July 1892, Small (G, MBG). [Vor. 20 280 ANNALS OF THE MISSOURI BOTANICAL GARDEN Fromipa: Apalachicola, Chapman (MBG). Onto: dry places, July, Riehl (MBG); alluvial soil, Windham Township, Portage Co., 12 Aug. 1924, Webb 5441 (G); Shaker Pond near Cleveland, 8 July 1896, Green- man 1379 (G, MBG); near Cleveland, Greenman 1377, 1882 (MBG). West Virainia: near Tygart Junction, 24 Sept. 1914, Moore 2569 (G, P); Marlin- ton, 29 July 1930, Berkley 1804 (MBG); Woods Co., 7 Aug. 1897, Pollock (MBG). MicniaAN: meadow, Turin, 9 July 1901, Barlow (G); thickets along West Maple River, west of Pellston, 7 Aug. 1917, Gates 10708 (MBG); near Trenton, 13 Aug. ES Chandler (US); — College, 1890, Baker (P); South Haven, 4 Sept. 1911, Lansing 3340 (G); B n Harbor, 4 July 1911, Lansing 3241 (G); New Buffalo, 20 July 1911, rani (G DIANA: open woods, East lines: 10 Aug. 1910, Lansing 2805 (G). TENNESSEE: Hollow Rock, 5 Aug. 1897, Eggert (MBG); along railroads, MeNairy Co., July 1893, Bain 444 (G). Wisconsin: along Wolf River, near Keshena, 13 Sept. 1925, E. J. Palmer 28738 (MBG); St. Croix Falls, Polk Co., July 1899, Baker (P); near Beef Slough, Buffalo Co., 13 Aug. 1926, Fassett & Hotchkiss 3322 (G); near Mirror Lake, Sauk Co., July 1903, Eggert (MBG). ors: meadow, State Street, Chicago, 4 Aug. 1907, Greenman 1953 (MBG); old beaches, Lake Chicago, Chicago, 5 Sept. 1909, Steele 106 (MBG); waste ground, 39th and Stoney Island, Chicago, 12 July 1913, H. H. Smith 5654 (G); west of Kan- kakee, 3 Aug. 1912, Sherff 1649 (MBG); waste land, Champaign, 31 July 1900, Gleason 1945 (G); Beardstown, July 1842, Geyer (MBG); Mississippi bottoms, Shepherd, Pike Co., 29 Aug. 1914, Davis 3565 (MBG); banks of the Mississippi, opposite St. Louis, July 1842, Engelmann (M BG). Minnesora: shore, De Soto Lake, Itaska Park, Becker Co., 13 Aug. 1929, Grant 3073 (G, MBG); fields, Collegeville, 29 July 1912, Chandonnet (MBG); Rockville, July 1896, Campbell (P); Swan Lake, Minneapolis, July 1892, Ballard (G). Iowa: Ames, Sept. 1873, Bessey (G); Ames, Sept. 1909, Pammel 78 (MBG); ledges, Boone Co., 1898, Pammel 1806 (MBG); Grinnell, Aug. 1875, Jones (P); Grinnell, Aug. 1877, Jones (G, MBG); Decatur Co., 13 July 1903, Anderson (MBG). Missoumi: Medill, 24 Aug. 1920, Bush 9165 (MBG, P); Palmyra, 5 Oct. "h Davis 8578, 8590 (MBG); Oakwood, Davis 3192, 3817 (MBG); St. Louis Co., Aug. 1877, Eggert (MBG, P); Creve Coeur Lake, 14 Aug. 1927, Kellogg 1131 dios St. PM July 1839, Lindheimer (MBG); banks of Chouteau's Pond, St. Louis, Aug. 1, Engelmann (MBG); south banks of Chouteau's Pond, St. Louis, July 1842, Races 337 (MBG); banks of River des Peres, near St. Louis, Sept. 1841, Engel- mann (MBG); St. Louis, Aug. 1861, Engelmann (G); Allenton, 18 June 1896, Kellogg (MBG); Courtney, 31 Aug. 1904, Bush 2214 (MBG); Webb City, E. J. Palmer 197, 198, 1085 (MBG); Neck City, 4 Aug. 1916, E. J. Palmer 10568 (MBG); Joplin, 10 July 1897, T'release 716 (MBG). Manirosa: low ground, Muskeg Island, Lake Winnipeg, 11 Aug. 1884, Macoun ie TH Daxkora: peninsula of Lake Ibsen, 1 Aug. 1899, Lunell (G); Grand Forks, 10 pes 1893, Brannon 128 (MBG); edge of slough, Fairmount, 22 July 1912, Berg- man 2335 (MBG). Sours Dakora: Brookings, 1 Aug. 1893, Thornber (MBG); Oakwood Lakes, Brookings Co., 27 July 1903, Johnson (MBG); Iroquois, 9 Aug. 1894, Thornber (G); Forestburg, 18 July 1910, Visher 4439 (MBG). NEBRASKA: Dickson’s Bluffs on the Missouri, 12 July 1853, Hayden (MBG); 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 281 prairies, Middle Loup River, near Thedford, 8 Sept. 1893, Rydberg 1515 (G); St. Paul, 24 July 1909, Bates (G); Lincoln, 9 July 1885, Webber (MBG); waste places, Nuckolls Co., Aug. 1899, Hedgcock (MBG). KANSAS: Olathe, Aug. 1892, Hitchcock (MBG); wet places, Riley Co., 16 July 1895, Norton 390 (G, MBG); banks of Joy Creek, about 8 km. from Osborne City, 11 July 1894, Shear 202 (G); Reno Co., 24 July 1899, White 174 (MBG). OKLAHOMA: near Lamont, 2 Aug. 1913, Stevens 1791 (G, MBG); Doby Springs, Harper Co., 20 Aug. 1927, Stratton ved (MBG); elose to edge of Lake Ivanhoe, near M 11 Oct. 1913, Stevens 2901 Texas: 1.6 km. north of Canadian = on Amarillo-Dalhart Road, Oldham Co., 27 Aug. 1921, Ferris & Duncan 8513 (MBG) SASKATCHEWAN: without locality, 1857-8, Bourgeau (G). Montana: Hot Springs, Flathead Lake, 8 Sept. 1908, Jones 8963 (P). WvoniNa: Goose Creek, 10 < 1896, A. Nelson 2258 (G, MBG); Sheridan, 24 July a A. Nelson 8439 (MBG OLORADO: Wray, 11-13 Aug. 1919, Eggleston 15551 (P); Timnath, 22 July 1901, or (P); Fort Collins, 7 July 1897, Crandall 2014 (G); along Platte River, Denver, 16 Aug. 1910, Eastwood 41 (G, MBG); foothills near Golden, 1 Aug. 1878, Jones 523 (P); Brighton, Sept.-Oct. 1908, E. L. Johnston 507 (MBG); Georgetown, 24 July 1878, Jones (P); Canon City, 1872, Brandegee (MBG). New Mexico: valley of the Mimbres, 1851, Wright 1498 (ANSP, G). Arizona: Flagstaff, 1891, MacDougal 566 (NY); Hassayampa Creek, Fort Whipple, 6 June 1865, Cowes & Palmer 279 (MBG). Ivano: Boise, 18 Aug. 1911, Clark 253 (G, MBG); Falk’s Store, Canyon Co., 28 June 1910, Macbride 304 (MBG). Uran, Salt Lake City, 18 June 1878, Jones 487 (P); near Salt Lake City, 14 July 1902, Pammel & Blackwood 3638 (G); American Fork Canyon, 3 Aug. 1880, Jones 1487 (P). Nevapa: Panaca, 5 Sept. 1912, Jones (P). British CoLUMBIA: Alberni, Vancouver Island, July 1915, Carter (G). WasniNGTON: Meyers Falls, 20 Aug. 1902, Kreager 469 (G); Alma, Okanogan Co., July 1897, Elmer 537 (MBG, P); foot of Priest Rapids, 17 July 1903, Cotton 1396 (G); Yakima region, 1882, Brandegee (MBG); Walla Walla, July 1898, Savage, Cameron & Lenocker (MBG); Prosser, 14 July 1929, Gotfredson 90 (P). OnEGOoN: near Pendleton, 11 Sept. 1896, Leiberg 2630 (G); Multnomah Co., July 1877, T. Howell 292 (G); Mount Scott, Multnomah Co., 15 Aug. 1902, Sheldon 11167 (G, MBG, P); low ground, Salem, 30 July 1917, J. C. Nelson 1804 (G). CALIFORNIA: near San Joaquin River, Louis Park, Stockton, 26 Sept. 1927, Stan- ford ?21 (P). This rather distinct species shows a tendency to blend with several members of the Verbenaca in the region of the Middle States. Some botanists have regarded the variability in the compactness and the profusion of the inflorescence, as well as the tendency of the leaves to develop hastate lobes near the base, as being of nomenclatorial value. Unfortunately in the rather complete series of specimens at hand, the writer has not found any combination of characters varying greatly in one direction; [Vor. 20 282 ANNALS OF THE MISSOURI BOTANICAL GARDEN and, at the present time, sees no reason to maintain other than the specific category. 17. V. simplex Lehm. Ind. Sem. Hort. Hamb. 17. 1825; in Nova Acta K. Acad. Leop. (Pugill. Pl. 1: 37) 14: 824. 1828; Linnaea 3: Litt.-Ber. 10. 1828. V. angustifolia Michx. Fl. Bor.-Am. 2: 14. 1803, not V. angustifolia Mill. Gard. Dict. no. 15. 1768. V. rugosa Muhl. ex Willd. Enum. Hort. Berol. 633. 1809, not V. rugosa Mill. Gard. Dict. no. 18. 1768, nor V. rugosa D. Don in Sweet, Brit. Fl. Gard. II. 4: pl. 318. 1838. Stems chiefly erect, branched above, branches ascending, sparsely pubescent or strigillose; leaves lanceolate or spathulate, tapering into a subsessile base, 3-10 em. long, subacute or ob- tusish, serrate or serrulate, reticulately rugose above and oc- casionally scabrous, somewhat prominently veined beneath, glabrate or sparsely strigillose on both surfaces; spikes solitary at the apices of stems and branches, short- pedunoled, strict, usually somewhat crowded; bracts lanosolateanbulate, commonly shorter than the calyx, glabrate or glabrous; fruiting calyx 4 (-5) mm. long, sparsely pubescent, lobes acuminate; corolla-tube scarcely longer than the calyx, with scattering hairs outside the throat; corolla-limb about 6 mm. broad; nutlets trigonous, 2.5 mm. long, raised-reticulate above, striate toward base; com- missural faces fully as long as the nutlets, muriculate-scabrous. Distribution: Ontario and Vermont, south to Florida, west to Oklahoma and Nebraska. Specimens examined: ONTARIO: rocky ground, Belleville, 20 June 1876, Macoun 1305 (G). Vermonr: without locality, 5 July 1903, Blanchard 26, 60 (G), 161 (NY). Massacuuserts: dry hillside, Sheffield, 27 Aug. 1902, Hoffman (G). Connecticut: railroad cut, Fair Haven, 18 June 1886, Harger (G); New Haven, Eaton (G); New Haven, 26 June 1884, Safford 216 (US). New York: sandy field north of Tripoli, 25 July 1920, Burnham (G); Pine Plains, 1875, Hoyeradt (G). Jersey: Scotch Plains, 13 Aug. 1877, Trelease (MBG); Camden Co., 22 June 1871, Parker (G); Atco, 22 June 1871, Redfield 6430 (MBG); Sicklerville, 21 June 1894, Brinton (G); Oakland, 25 July 1909, Mackenzie 4213 (MBG) PENNSYLVANIA: without data, Muehlenberg (Bot. Mus. Berl.-Dahl. TYPE of V. rugosa, MBG phot.); Philadelphia, Nuttall (G); Whiteland, Chester Co., 26 June 1910, Bartram 1021 (G); vicinity of Conewago, 28 May 1889, Small (US); York Furnace, 28 June 1899, MacElwee 760 (MBG, NY); Mercersburg, Green (G) 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 283 Maryann: Great Falls, 6 Aug. 1905, House 1863 (MBG); Hurlock, 29 May 1919, C. P. Smith 3186 (G). District OF CoLuMB1A: Washington, 26 May 1889, Sudworth (G); Joy City, 19 June 1891, Blanchard (MBG). Vrana: Bluemont, 25 May 1905, House 875 (US); open woods near Churchview, 5 June 1921, Leonard & Killip 534 (US); north of Williamsburg, 17 June 1921, Grimes 8788 (NY); Bedford Co., June 1873, Curtiss (MBG); Chilhowie, 4 Aug. 1892, Small (F, NY); about Cumberland Gap, Lee Co., 27 July 1892, Small (F, G, MBG, US); Blue Ridge, 22 July 1891, Seymour 48 (G), 49 (MBG). NonTH CAROLINA: near Chapel Hill, Ashe (US); Morgantown, 1872, Ruger (US); near Biltmore, 17 June 1897, Biltmore Herbarium 4759b (G, NY, US); Madison Co., 27 July 1880, J. D. Smith (US). Sours CAROLINA: Abbeville District, July 1855, Hexamer & Maier (G). Grorata: Dalton, 9 Aug. 1900, Harper 385 (NY, US); Chickamauga Creek, near Ringgold, 6-12 Aug. 1895, Small (F, NY); between Taylor Ridge and Lafayette, 3 July 1900, P. Wilson 161 (NY, US). Fronipa: without locality, 1842-49, Rugel 127 (F, MBG, US). AMA: Cullman, 29 May 1892, Mohr (US); Blount Springs, 5 May 1898, Baker (P); Blountsville, 6 May 1898, Baker (US); Choctaw Co., 13-15 Oct. 1896, Schuchert (US). Mississippi: Agricultural College, Oktibbeha Co., 11-17 Aug. 1896, Pollard 1312 (F, G, MBG, NY, P, US). Onro: on limestone soil, Castalia, 27 July 1894, Moseley Er Sandusky, 28 July 1894, Moseley (F, US); Oxford, 16 June 1910, Overholts (MBG West VIRGINIA: Shenandoah Junction, June 1891, Millepaugh 865 (NY); along Roanoke River, south of Roanoke, 2 June 1891, Small & Heller 482 (G); roadside, south of Williamsburg, 13 June 1921, Grimes 3716 (G); Sweet Springs, 14 Sept. 1903, Steele 314 (G, MBG, NY, US). InpIANA: near boundary line to Galien, St. Joseph Co., 15 June 1911, Nieuwland 2685 (MBG, US); about 3 km. west of Goldsmith, 9 July 1913, ae 13619 (MBG); about 1.6 km. west of Palmyra, 22 June 1916, Deam 2035? (G, U Kentucky: without data, Short (NY); Stamping Ground, 29 ae 1930, Singer 178 (US); Smithfield, 30 May 1909, Eggleston (NY); Nicholasville, 8 June 1923, McFarland 109 (MBG); Mammoth Cave, Edmonson Co., May 1899, Palmer (NY); Bowling Green, 21 May 1899, Price (MBG); near Franklin, 22 May 1926, Anderson & Woodson 47 (MBG); Kuttawa, 27 Sept.-9 Oct. 1909, Eggleston 5237 (MBG, NY). ENNESSEE: Carter Co., 8 July 1880, J. D. Smith (US); Knoxville, Ruth (F, P), 731 (US), 740 (NY), 765 (MBG); roadside at base of Chilhowee Mountains, Curtiss 1955 (F, G, MBG, NY); Chattanooga, 25 Aug. 1876, Engelmann (MBG); West Nash- ville, 26-27 May 1909, Eggleston 4430 (MBG, NY, US); Kingston Spring, 5 Aug. 1897, Eggert (MBG); Hollow Rock, Carroll Co., 14 Aug. 1897, Bilimore Herbarium 4759 (G, MBG, NY, US). ue Baraboo, 1861, Hale (G, MBG). Inuinors: Stony Island, June 1911, Greenman 3640 (MBG); Stony Island, 25 June 1914, H. H. Smith 5923 (G, MBG); near 91st Street, Chicago, 28 July 1907, Greenman 1981 (MBG); South Chisage, 18 July 1913, H. H. Smith 5670 (G); Rock Island, 17 Aug. 1866, Engelmann (MBG); East St. Louis, 9 July 1898, Norton (MBG); bluffs near Prairie du Pont, 20 June 1876, Eggert (MBG); Red Bud, 3 June 1888, Pammel (MBG); Shawneetown, 26 May 1919, E. J. Palmer 15262 (MBG); Mound City, June 1859, Vasey (G). [Vor. 20 284 ANNALS OF THE MISSOURI BOTANICAL GARDEN Iowa: without locality, 17 July 1875, Arthur 24 (MBG); Cedar Falls, 12 July 1926, Pammel 272 ys Iowa City, Hitchcock (MBG); Columbus Junction, 27 Aug. 1899, Pammel 1699 (MBG). Missounr: Hannibal, 27 June 1917, Davis 3588 (MBG); Ethel, 11 June 1915, Bush 7586a (MBG); Silex, 29 May 1915, Davis 4552 (MBG); prairie, St. Louis, May 1833, Engelmann 336 (MBG); banks of Mississippi above St. Louis, Oct. 1841, —— (M BG); St. Louis, Sept. 1841, Engelmann (G, US); mineral region south- of St. Louis, June 1845, King (MBG); Pacific, 13 June 1897, Trelease 713 (MBG); Pacific, 3 June 1918, Greenman 4125 (MBG); Gray Summit, 16 June 1927, Kellogg 1180 (MBG); Allenton, 6 July 1911, Letterman (MBG); near Sulphur Springs, May 1927, Steyermark 292 (MBG); Crystal City, 14 May 1887, Trelease (MBG); Washington, 25 June 1888, Pammel (MBG); Victoria, 8 July 1890, Hitchcock (MBG); Jefferson Co., 20 June 1876, Eggert (MBG); near Hillsboro, June, Riehl 456 (MBG); Hillsboro, 24 May 1885, F. Wislizenus 280 (MBG); near St. Genevieve, 27 Ma 1928, Greenman 4578 (MBG); Jerome, 24 May 1914, Kellogg 496 (MBG); Sheffield, 17 June 1915, Bush 7647 (G, MBG, US); Westport, 17 May 1896, Bush 914 (US); Independence, 1 June 1895, Tindall (MBG); south of Cedar Gap, 22 May-3 June 1911, Lansing 2976 (F, G); about 2 km. west of Mansfield, 5-12 June 1911, Lansing 135 (F, G); Springfield, 31 July 1892, Dewart 42 (MBG); Willard, 24 July 1919, Blankinship (P); limestone barrens, Ash Grove, 24 Aug. 1912, Standley 9336 (US); Carterville, 8 July 1910, E. J. Palmer 2981 (MBG); Webb City, E. J. Palmer 200, 3319 (MBG); Swan, 24 Sept. 1899, Bush 450 (MBG); Galena, Stone Co., 27 Ma 1914, E. J. Palmer 5769 (MBG); Gainesville, 26 June 1928, E. J. Palmer 34766 (MBG); Eagle Rock, 24 June nv Bush 206 (MBG, US); Noel, McDonald Co., 9 Sept. 1913, E. J. Palmer 4232 (MBG ARKANSAS: Beaver Station, Eureka Springs, Glatfeller (MBG). NEBRASKA: Wahoo, June 1890, Rydberg 154 (NY). ANSAS: Lawrence, Stevens (US); between Sereni var Hopes 20 June 1929, Rydberg & Imler 120 (MBG); sterile soil, Anderson Co. , Hitchcock 791 (G, MBG, NY, US); Mound City, 18 July 1887, Kellerman pe OKLAHOMA: near Miami, 26 Aug. 1913, Stevens 2299 (G, ved near Copan, 15 Aug. 1913, Stevens 2080 (G); Foyil, 5 Aug. 1894, Bush 433 (MBG); on ereek bank, near Pawhuska, 9 Aug. 1913, Stevens 1980 (G); Caddo, 20 range 1891, Sheldon 49 (US); near Idabel, 18 May 1916, Houghton 3622, 3646 (G, MBG). This species appears to be very closely related to V. hastata. In its typical form, it is readily distinguished by its elongate spikes and narrow leaves. Often, however, intermediate forms occur between the two, making it somewhat difficult to find clear lines of demarcation between them. 18. V. Orcuttiana Perry,'? n. sp. 5 V. Orcuttiana Perry, spec. nov., herbacea verisimiliter perennis; caulibus erectis quadrangularibus sparse trábesbentibus ramosis; foliis lanceolato-ellipticis spathulatisve in petiolum alatum attenuatis 4-6 em. longis grosse serratis lineato- rugosis adpresso-pubescentibus supra subtusque reticulatis dense patento-hirtellis; spicis pedunculatis compactis elongatisque; bracteis lanceolato-acuminatis sparse ciliatis nervo medio et margine decurrentibus calyce brevioribus; calyce 4 mm. longo 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 285 Stems several from a common base, 4-angled in cross-section, glabrous or very sparsely hirtellous, branching; leaves lanceolate- elliptical to spathulate, tapering into a margined petiole (1-2 em. long), 4-6 em. long, decussate, coarsely serrate, rugose and appressed-pubescent above, more densely spreading-pubescent or hirtellous beneath and prominently veined; spikes pedunculate, solitary or somewhat panicled, strict, mostly dense, elongate; rhachis more or less angulate; bracts lanceolate-acuminate, shorter than the calyx, midrib and margins more or less decurrent along the rhachis, sparsely ciliate; calyx 4 mm. long, appressed- pubescent and finely glandular, teeth short-subulate or acuminate, more or less connivent above the schizocarp; corolla-tube about as long as the calyx; corolla-limb 3-4 mm. broad; nutlets tri- gonous, 2 mm. long, raised-reticulate at the apex, striate toward the base; commissural faces extending to the tip of the nutlet, muriculate-scabrous. Distribution: Lower California. Specimens examined: Mexico: Lower CALIFORNIA: Pinery, 27 July 1883, Orcutt (US); Hanson's Ranch, 29 July 1883, Orcutt (G, NY, US); table-lands, Hanson’s Ranch, 30 July 1883, Orcutt 909 G TYPE); mountains, northern Lower California, 8 July 1885, Orcutt (US); Sierra Juarez, 12 July 1924, Gallegos 2342 (US) This species, which has been passing as V. litoralis, is much like V. neomexicana var. xylopoda in the finely glandular spike and the angle of insertion of the flowers. It differs, however, in the shorter nutlets, the smaller corollas, and the type of the pubescence. In V. Orcuttiana, the trichomes are short and somewhat hirtellous and the pubescence of the inflorescence is closely appressed. In gross habit, it is scarcely to be distin- guished from V. simplex, but the latter has somewhat harsher pubescence and larger non-glandular flowers. 19. V. stricta Vent. Hort. Cels. 53, pl. 53. 1800. V. Alopecurus Cav. Descr. 68. 1802. adpresso-pubescente tenuiter glanduloso; calycis dentibus breviter subulatis vel acuminatis; corollae tubo vix exserto; corollae limbo 3-4 mm. lato; coccis subtrigonis 2 mm. longis dorso striatis apice elevato-reticulatis; commissura muriculata.—Col- lected on table-lands, Hanson's Ranch, Lower California, 30 July 1883, Orcutt 909 (G), TYPE. [Vor. 20 286 ANNALS OF THE MISSOURI BOTANICAL GARDEN V. rigens Michx. Fl. Bor.-Am. 2:14. 1803. cuneifolia Raf. Med. Repos. N. Y. II. 5: 360. 1808. stricta B ? mollis Torr. Ann. Lyc. N. Y. 2:234. 1827. mollis Raf. Atl. Jour. 146. 1832. stricta f. roseiflora Benke, Rhodora 34: 10. 1932. stricta f. albiflora Wadmond, Rhodora 34: 19. 1932. Stems 3-12 dm. tall, subterete, simple or branched above, rather densely hirsute; leaves ovate or suborbicular, 6-10 cm. long, sessile or nearly so, sharply and mostly biserrate, thickish, hirsute and rugose above, densely hirsute-villous and prominently veined beneath; spikes solitary or several, short-pedunculate, thick, usually quite dense both in flower and in fruit; bracts lanceolate-subulate, approximately as long as the calyx, hirsute, ciliate; calyx 4-5 mm. long, densely hirsute, lobes acuminate; corolla protruding slightly beyond the calyx, pubescent without; corolla-limb 8-9 mm. broad; nutlets trigonous, 2.5 mm. long, raised-reticulate above, strongly striate below ; commissural faces reaching tip of nutlet, muriculate or almost smooth. Distribution: eastern and central United States, from Pennsylvania westward through the Rocky Mountains. Probably introduced into other localities. Specimens examined: (Herb. Bot. Gard. Madrid TYPE of V. Alopecurus, MBG phot.). Massacuuserts: Fall River, 2 Sept. 1908, Sanford (G Connecticut: waste ground, Nangatuck, 19 July 1908, Blewitt 14 (G); Bridgeport, 27 Aug. 1892, Hames (G). New York: pasture south of Pulpit Rock, Ithaca, 2 Aug. 1919, A. J. Hames 12797 (Q). New Jersey: ballast, New Durham, 15 July 1893, Van Sickle (US). PENNSYLVANIA: along Lincoln Highway at Gap, Lancaster Co., Urban (G); vicinity of Conewago, Sept. 1892, Small (NY); Conewago, Sept. 1892, Heller 638 (US). hin ec Da Onto: Dayton, Short (MBG). MicniGAN: Constantine, 2 July 1923, Fisher 19 (MBG); Pine Lake, Charlevoix Co., 18 Aug. 1917, Ehlers 642 (P). InpIANA: East Chicago, 10 Aug. 1910, Lansing 2810 (G); Lake Maximkuckee, 9 Aug. 1889, Evermann 970 (US); sand pit northeast of Winona Lake, 2 Aug. 1897, Deam (MBG); west of Palmyra, 22 June 1916, Deam 20357a (G KENTUCKY: waste places, Hickman, 14 Aug. 1897, Biltmore Herbarium 3653a (NY); Calvert City, Marshall Co., 16-19 aee 1909, Eggleston 4837 (NY); Wickliffe, 16 Aug. 1923, McFarland & Anderson 253 (MBG). TENNESSEE: Hickmann, Gattinger (US); Henderson, June 1892, Bain 328 (NY). Wisconsin: Mirror Lake, 15 July 1903, Eggert (MBG); Waupaca, 1907, Garesche (MBG); Trempeleau, 1861, Hale (G); Oregon road, Madison, 29 July 1889, T'release (MBG). 1933] PERRY— NORTH AMERICAN SPECIES OF VERBENA 287 InnrNors: Stony Island, Chicago, H. H. Smith 5645, 6028 (G, MBG), 5952 (G); Romeo, 26 July 1897, Umbach (MBG); Starved Rock, La Salle Co., June-Sept. 1921, Thone 88 (MBG); Oquawka, 1872, Patterson (G, NY); Bloomington, July 1886, Robinson (G); Carthage Township, 2 Sept. 1916, Gates 9994 (MBG); Decatur, 19 June 1896, Gleason 377 (G); Athens, Aug. 1863, Hall (MBG); East Hannibal, 13 June 1913, Davis 16 (MBG); Shepherd, 25 June 1915, Davis 6370 (MBG); Mississippi River, bluffs north of Alton, 4 Aug. 1910, Sherff (G); Bonpas Township, 15 July 1925, Ridgway 2431 (MBG); E. St. Louis, 28 July 1900, Eggert (MBG); Grand Tower, 22 Aug. 1900, Gleason (G). Minnesota: Perham, Ottertail Co., 8 Aug. 1912, Chandonnet (MBG); St. Anthony, 7 July 1888, Schuette (G, NY); Willmar, July 1892, Frost (G); Morton, July 1890, MacMillan (P); Lake City, 28 July 1883, Manning (G); Winona Co., Aug. 1901, Holzinger (NY); Houston Co., July 1912, Freiberg (MBG). owa: Fayette Co., 12 July-4 Sept. 1904, Fink 251 (US); Hawkeye, Aug. 1896, Gardner 588 (NY); Cerro Gordo Co., 14 Aug. 1899, J'ones petal: Ames, 29 Aug. 1896, Pammel 85 (G, MBG, NY); Grinnell, Aug. 1877, Jones (NY, P); Mount Pleasant, 7 July 1898, Ball 1585 (MBG); rn ig Aug. 1920, a 1947 (MBG); Ben- tonsport, July 1920, Graves 1994 (M Missouri: suburbs of Hannibal, pies 1202, 1613, 2791, 2960, 2962, 3227, 3589, 4461, 4487 (all MBG); Chain of Rocks, Aug. 1915, Beckwith 48 (MBG); St. Louis Co., 15 July 1872, Redfield 522 (MBG); St. Louis, Aug. 1838, Riehl 196 (MBG, NY); St. Louis, Aug. 1841, Engelmann (G); Allenton, Letterman (MBG); Williamsville, Wayne Co., 27 June 1914, E. J. Palmer 6107 (MBG); Jerome, 16 June 1914, Kellogg 498 (MBG); Warsaw to Linn Creek, Camden Co., July-Aug. 1913, Emig 107 (MBG); Cole Camp, Benton Co., 12 July 1897, Trelease 717 (MBG); railroad embankments, Mansfield, 5-12 June 1911, Lansing 3164 (G); Randolph, 17 July 1898, Mackenzie 261 (MBG); Independence, 26 June 1895, Bush 475 (MBG, NY); Willard, 9 July 1919, Blankinship (P); Carterville, 8 July 1910, E. J. Palmer 2980 (MBG); Webb City, 3 Aug. 1902, E. J. Palmer 199 (MBG); Turkey Creek, Joplin, 10 July 1897, Trelease 718 (MBG); Swan, 25 Sept. 1899, Bush 670 (MBG). ARKANSAS: waste places, Harvey 1958 (G, NY); Big Lake, 20-22 June 1911, McAtee 2053 (NY, US); Jonesboro, Craighead Co., 4 July 1927, Demaree 3553 (MBG); Fayetteville, Harvey 61 (MBG); Fort Smith, 1853-4, Bigelow (US). Sovru Daxora: Big Stone Lake, Roberts Co., July 1922, Over 14386 (US); Wind- sor Township, Brookings Co., 27 July 1903, Johnson (MBG); Forestburg, 3 July 1910, Visher 4450 (MBG); 16 km. south of Interior, 29 June 1929, E. J. Palmer 37627 (G, MBG); near Fort Meade, Black Hills, 19 June 1887, Forwood 299 (US); canyons, Lead, 9 Aug. 1913, Carr 118 (G, MBG, NY, US); Rapid City, 22 July 1912, Visher 1507 = Hot Springs, 3 Aug. 1892, Rydberg 982 (US). EBR : Lincoln, 27 June 1885, Webber (MBG); E bel City, 26 Aug. 1926, Heller 14290 (MBG); South Fork of Platte, July 1 H. Engelmann (G, MBG); Alma, 21 June 1897, Pammel (MBG); Anselmo, 8 m 1889, Webber (US); Broken Bow, 7 July 1897, Pammel (MBG); Callaway, 27 June 1901, Bates (G); Dismal River, south of Thedford, 27 June 1893, Rydberg 1422 (G, NY, US); North Platte, July 1896, Plank (NY); Ogallala, 16 June 1925, Jones (P). Kansas: Pottawatomie Co., 1895, Hitchcock 972 (US); prairie, Riley Co., 26 July 1895, Norton 391 (G, MBG, NY, US); Concordia, 24 July 1929, Benke 5164 (G, NY), forma roseiflora; hill 8 km. from Osborne, 11 July 1894, Shear 191 (G); Florence, 28-30 July 1903, Griffiths 5050 (US); vicinity of Caney, 29 June 1929, Rydberg & Imler 433 (MBG, NY); Syracuse, 28 July 1893, C. H. Thompson 154 (G, MBG, NY, US). [Vor. 20 288 ANNALS OF THE MISSOURI BOTANICAL GARDEN OKLAHOMA: Verdigris, 2 Aug. 1894, Bush 435 (MBG); 12 km. west of Pawnee, 27 July 1927, Stratton 249 (MBG); near Tonkawa, 5 Aug. 1913, Stevens 1865 (G); Doby Springs, 20 Aug. 1927, Stratton 408 (MBG); near Shattuck, 11 Oct. 1913, Stevens 2930 (G); Norman, 10 Oct. 1914, Emig 365 (MBG); near Granite, 17 June um Stevens 1084 (G); PEC Mountains near Davis, 23 June 1917, Emig 787 MBG); vicinity of Fort Sill, 16 Aug. 1916, Clemens 11748 (MBG); near Cache, 25 eg 1913, Stevens 135414 (G, MBG); near Grant, 2 June 1916, Houghton 4000 (G, MBG, NY). Texas: Lipscomb, 1 July 1903, Howell 64 (US); prairies, near Canadian, 11 Aug. 1900, Eggert (MBG); Dallas, Reverchon (G, NY), 735 (MBG). ONTANA: bottoms, Crow Agency, 14 July 1901, Blankinship (G). Wyomina: Whalen Canyon, 16 July 1894, A. Nelson 538 (G); Hartville, 20 July 1894, A. Nelson 505 (G, MBG). Cotorapo: Wray, 1-4 July 1919, Eggleston 15224 (MBG); Denver, 2 Sept. 1910, Eastwood 90 (G, MBG, US). New Mexico: creek bottom, Santa Fe, 1847, Fendler 597 in a (ANSP, G). WasurNGTON: Meyers Falls, 20 Aug. 1902, Kreager 475 (G, N Y, US). A very distinct species somewhat incapable of sharp delimita- tion on account of the tendency to hybridize with neighboring species. Normally it is recognized by its stout compact spike, imbricated flowers, and ovate-orbicular sessile leaves. 20. V. MacDougalii Heller, Bull. Torr. Bot. Club 26: 588. 1899. V. MacDougalii mut. rosella Cockerell, Am. Nat. 36: 809. 1902 Stems 3-8 dm. tall, stout, obtusely four-angled, simple or occasionally branched, cinereous-green, hirsute-pubescent ; leaves oblong-elliptical or elongate-ovate, 6-10 em. long, short-petiolate or narrowed into a subpetiolar base, coarsely and irregularly serrate-dentate, hirtellous, rugose and minutely pustulate above, densely pilose-pubescent and prominently veined beneath; spikes solitary or sometimes several, short-pedunculate, thick, comparatively dense both in flower and in fruit; bracts lanceolate- subulate, for the most part noticeably longer than the calyx, pubescent, ciliate; calyx 4-5 mm. long, rather densely pubescent, lobes very obtuse, terminating in short subulate teeth; corolla- tube scarcely protruding beyond the calyx; corolla- limb 6 mm. broad; nutlets trigonous with convex back, 2.5 mm. long, raised- tetioulate toward the distal end, strongly or faintly striate below; commissural faces reaching tip of nutlet, muriculate or almost smooth. 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 289 Distribution: ae Wyoming to New Mexico and Arizona. Specimens examined: WYOMING: Platte! Canyon, Laramie Co., 2 July 1901, A. Nelson 8354 (MBG). Cotoravo: Palmer Lake, 22 July 1895, Osterhout (NY); La Veta, 14 July 1896, Shear 3577 (NY); Cucharas Valley, near La Veta, 20 July 1900, Vreeland 636 (NY); Stonewall, July 1912, Beckwith 188 (NY); Arboles, 10 July 1899, Baker 565 (F, G, MBG, NY, P, US). New Mexico: without locality, 1847, Fendler 597 in part (MBG); Vermejo Park, Colfax Co., 31 Aug. 1913, Wooton (NY); vicinity of Ute Park, Colfax Co., 2 Sept. 1916, Standley 14224 (NY); foot of Baldy Mountain, near Elizabethtown, Oct. 1898, St. John (US), mut. rosella; near Sierra Grande Union, 18 June 1911, Standley 6065 (US); Jicarilla Apache Reservation, near Dulce, 20 Aug. 1911, Standley 8239 (US); southeast of Cuba, 22 Aug. 1915, Read 19 (US); Sandia Mountains, 15 July 1914, Ellis 258 (MBG, US); Santa Fe Co., 10 July 1847, Edwards (NY); Santa Fe Co., 1889, Brandegee (MBG); Santa Fe Co., 20 July 1898, Greene 77 (NY); Santa Fe Canyon, 3 Oct. 1913, Rose, Fitch & Parkhurst 17717 (US); Winsar’s Ranch, Pecos River National Forest, 6 July 1908, Standley 4223 (F, G, MBG, NY, US); near Pecos, 15 Aug. 1908, Standley 4927 (F, G, MBG, NY, US); west of Las Vegas, St. John (P); Las Vegas, 26 June 1893, Mulford 39 (MBG); vicinity of Las Vegas, Anect 57 (G), 211 (NY); mountains west of Grant’s Station, 2 Aug. 1892, Wooton (NY, US); White Mountains, Lincoln Co., 26 July 1897, Wooton 208 (MBG, NY, P, US); Cloudcroft, 1912, Stearns 342, 358 (US); Cloudcroft, Aug. 1920, Schulz 255 (NY); Mescalero Reservation, Sacramento Mountains, 21 July 1905, Wooton (US). Arizona: Clark’s Valley, 1 Aug. 1883, Rusby (NY, US); South Fork, Little Colorado River, Apache Forest, 23 Aug. 1920, Eggleston 17106 (NY); 16 km. south of Rowe’s Point on the Grand Canyon, 5 Nov. 1899, Ward 11 (US); Flagstaff, 29 Aug. 1884, Jones (P); vicinity of Flagstaff, 8 July 1898, aire a 249 (ANSP, F, G, NY TYPE); im pines, Flagstaff, 16 Aug. 1922, Hanson A148 (F, MBG). Uram: Mammoth Creek, near head of Sevier River, 10 Sept. 1894, Jones 6026 (MBG, NY, P. This southwestern representative of V. stricta is readily distin- guished by the short-petiolate elongated leaves, the compact spikes, and the floral bracts surpassing the calyx. 21. V. macrodonta Perry,!? n. sp. PI. 14. 19 V. macrodonta Perry, spec. nov., herbacea (basi ignota) verisimiliter perennis; caule 1-1.5 m. alto erecto hirsuto-hispidulo ramoso; foliis elongato-ovatis basi cuneata in brevem petiolum alatum attenuatis 10-14 em. longis biserratis vel mu- cronulato-denticulatis supra scabro-hirsutis et inconspicue pustulatis subtus reticu- latis hirsutis; spicis paniculatis glanduloso-hirsutis basi foliolatis compactis deinde elongatis basi laxisque; bracteis lineari-lanceolatis subulatis ciliatis calyci subae- quantibus; calyce 5 mm. longo glanduloso aliquantulum viscido-pubescente; calycis dentibus 1.5 mm. longis subulatis; corollae tubo paulo exserto extus glabro vel uberulo; corollae limbo 5-6 mm. lato; coccis subtrigonis 2 mm. longis dorso sulcatis superiore parte tenuiter Birrae commissura muriculata.—Collected on the road from Miraflores to San Bernardo Ranch in Sierra La Laguna, Lower California, about 750 m. alt., 20 Jan. 1906, Nelson & Goldman 7425 (MBG), TYPE. [Vor. 20 290 ANNALS OF THE MISSOURI BOTANICAL GARDEN Stem 1-1.5 m. tall, erect, branching, hirsute-hispidulous; leaves elongate-ovate with cuneate base narrowed into a short (1-2 em. long) margined petiole, 10-14 cm. long, coarsely and sharply biserrate-dentate with mucronate teeth, rugose with veins impressed and scabrous-hirsute above with minutely pustulate hairs, prominently veined and hirsute beneath; spikes panicled, subtended by leafy bracts, dense before anthesis, becoming elongated and open in fruit, glandular-hirsute; floral bracts lanceolate-linear, approximately equalling the fruiting calyx, subulate, ciliate; calyx 5 mm. long, glandular, somewhat viscid-pubescent, teeth 1.5 mm. long, subulate; corolla-tube protruding a little beyond the calyx, glabrous or puberulent without; corolla-limb 5-6 mm. broad; nutlets trigonous, 2 mm. long, shallowly scrobiculate on the upper half, tending to be sulcate toward the base; commissural faces muriculate. Distribution: Lower California. Specimens examined: Mexico: Lower CALIFORNIA: road from Miraflores to San Bernardo Ranch in Sierra La Laguna, about 750 m. alt., 20 Jan. 1906, Nelson & Goldman 7425 (MBG TYPE, US). Verbena macrodonta is a coarse plant with large thickish leaves and open inflorescence, in a measure similar to V. MacDougalit, but differing in its less strict habit, somewhat remote fruits, more glandular calyces, and shorter plumper nutlets. 22. V. prostrata R. Br. in Ait. Hort. Kew. ed. 2,4:41. 1812. V. lasiostachys Link, Enum. Hort. Berol. 2: 122. 1822. Stems at first erect or ascending, at length spreading and diffusely branched, sparsely villous; leaves ovate or oblong-ovate with cuneate base tapering into a margined petiole, 5-10 cm. long, commonly 3-cleft, coarsely and irregularly serrate-dentate with apiculate teeth, or incised, veins impressed above, more or less conspicuous below, both surfaces pilose to sparsely villous; spikes solitary or more often loosely paniculate, dense before anthesis, becoming elongated and open in fruit; bracts lanceolate- subulate, not longer than the calyx, villous and at times finely glandular; calyx 4-5 mm. long, villous or glandular-hirsute, subtruncate, the distal end connivent above the schizocarp, teeth subulate; corolla-tube a little longer than the calyx, very 1933] PERRY—NORTH AMERICAN SPECIES OF YERBENA 291 sparsely pubescent outside or apparently glabrous; corolla-limb 3-5 mm. broad; nutlets oblong-trigonous with convex back, 2-2.5 mm. long, slightly thicker at the distal end, raised-reticulate above, striate below, often with striae fading out toward the base; commissural faces more or less muricate, approximately reaching the tip of the nutlet. Distribution: ae and California. Specimens examin OnEGON: dry um along railroad, about 1.5 km. north of Comstock, 18 June 1919, J. C. Nelson 2669 (G); banks of Umpqua at Roseberg, 2 June 1928, Thompson 4412 (MBG); Glendale, 19 June 1902, Jones (P); Pleasant Creek, near Wimer, 26 July 1892, Hammond 322 (MBG, NY, US); Grant’s Pass, 24 June 1884, T. Howell 174 (US); Grant’s Pass, 3 July 1887, T. Howell (F), 1249 (MBG); gravel bar about 5 km. above mouth of Rogue River, 8 July 1919, Peck 8702 (G, MBG, NY); in clearing about hotel, Agness, 22 June 1917, J. C. Nelson 1429 (G); Dryden, 16 June 1904, Piper 6160 (US); near Medford, Aug. 1922, Epling 5445 (MBG); south of Ashland, 19 May 1898, Applegate 2228 (US). CALIFORNIA: near Yreka, Siskiyou, 19 June 1876, Greene 860 (F, G, MBG, NY); Yreka, 2 July 1910, Butler 1621 (MBG, P, US); Mount Shasta, 13-27 July 1892, Palmer 2529a (US); Soda Springs, Nevada Co., 30 July 1881, Jones 2598 (P); Round Valley, Mendocino Co., 20 May-20 June 1898, Chestnut (US); near Clear Lake, 1865, Torrey 417 (G, NY); Clear Lake, Lake Co., 12 July 1929, Blankinship (MBG); southern slope of Mount Sanhedrin, Lake Co., 19 July 1902, Heller 5919 (G, MBG, NY, P, US); west of Windsor, near Russian River, Sonoma Co., 27 June 1902, Heller 5785 (F, G, MBG, NY, P, US); Oak Knoll, 5 May 1901, Braunton 378 (US); Irish- town, Amador Co., June 1893, Hansen 964 (MBG); West Point Bridge, Calaveras Co., 7 July 1896, Hansen 1823 (US); Stanford University, Santa Clara Co., April 1898, Abrams (P); San Jose, 14 July 1899, Pammel 187 (MBG); Santa Cruz, 21 June 1881, Jones (G, P); in pine woods, Pacific Grove, 25 May 1903, Heller 6778 (F, G, MBG, NY, P, US); Carmel Bay, Sept. 1902, Elmer 4045 (G, P); Monterey, Hartweg 1924 (G, NY); Monterey, Aug. 1917, Parish 11590 (P); Santa Lucia Mountains, May 1898, Plaskett 142 (NY, US); San Antonio River, 1880, Vasey (US); Wood’s Creek, Fresno Co., 19 June 1910, Clemens (P); San Luis Obispo, 26 June 1876, Palmer 342 (MBG, NY, US); San Luis Obispo Co., 19 June 1887, Summers (MBG, P); Santa Barbara Co., 1865, Torrey 416 (G, NY); Santa Barbara, May 1902, Elmer 3846 (MBG, P, US); seep, base of ocean bluffs, 32 km. northwest of Santa Barbara, 26 March 1925, Munz 9295 (P); Sulphur Mountain Spring, Ventura Co., 1-2 June 1908, Abrams & McGregor 6 (G, NY, US); Ventura, 13 April 1923, Kendall (P); San Buenaventura, March 1861, Brewer 229 (US); Los Angeles, 1860-62, Brewer 31 (G, US); Los Angeles, 29 May 1891, Fritchey 29 (MBG); near Mesmer, 10 June 1917, Johnston 1820 (P, US); near Pasadena, June 1893, Haynes (P); San Antonio Canyon, 28 June 1917, Johnston 1608 (P); edge of woods, Palomar Mountain, 11 Aug. 1918, Spencer 994 (G, P); wayside, Mesa Grande, 2 June 1919, Spencer 1164 (G); Spencer Valley, near Julian, San Diego Co., 20 June 1903, Abrams 3787 (F, G, MBG, NY, P); bottom of canyon south of “Lemon Tank," west coast, San Clemente Island, 10 April 1923, Munz 6734 (P). Although, from the original publication, one would naturally [Vor. 20 202 ANNALS OF THE MISSOURI BOTANICAL GARDEN infer that V. prostrata is Aiton’s species, it is in all probability Robert Brown's. Aiton, in his acknowledgments (postscript to the fifth volume of Hortus Kewensis), mentions the new matter added by his friend Robert Brown, some without reference to his name; more tangible evidence is furnished by Schauer, who, in his monograph, indicates the specimen at Kew as V. prostrata R. Br The species is comparatively easy to distinguish by its de- cumbent habit, soft villous pubescence, and elongated spikes. It closely resembles V. robusta, which is much harsher and of limited distribution. The following specimens differ from the species in having scarcely glandular inflorescence and fruiting calyx only 2.5- 3 mm. long shortly conniving beyond the fruit: Catirornia: Three Rivers, Tulare Co., 9 July 1904, Culbertson 4210 (MBG, P); San Bernardino, Sept. 1886, Parish & Parish 969 (F, US); near San Bernardino, May 1894, Parish 2819 (MBG, P); Inglewood, Los Angeles Co., 31 May 1902, Abrams 2481 (P); San Diego Co., 1875, Palmer 309 (G, MBG); southern California, 1876, Parry & Lemmon 842 (NY). 23. V. robusta Greene, Pittonia 3: 309. 1898. Stems 6-9 dm. high, erect, paniculately branched above, glabrate or sparsely hirsute; leaves ovate or oblong-ovate with cuneate base tapering into a margined petiole, 4—7(-10) cm. long, usually 3-cleft, irregularly serrate-dentate with apiculate teeth, or incised, rugose and scabrous-pubescent above, less harshly pubescent beneath with veins prominently reticulate; spikes often crowded, subsessile, usually dense but occasionally elongated; bracts lanceolate-subulate, more or less exceeding the calyx in length, densely glandular-hirtellous; calyx 4 mm. long, densely glandular-hirsute, lobes obtusish, terminating in very unequal acuminate-subulate teeth; corolla-tube a little longer than the calyx, puberulent without; corolla-limb 3-4 mm. broad; nutlets oblong-trigonous, 2-2.5 mm. long, raised-reticulate above with striae fading out toward the base; commissural faces more or less muricate, approximately extending to the tip of the nutlet. Distribution: California and Lower California. Specimens examined: 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 203 CALIFORNIA: vicinity of Ione, Dec. 1904, Braunton 1268 (MBG, NY); Tiburon, Marin Co., 26 July 1900, Eastwood (G); valley back of Berkeley, 23 Oct. 1880, Engelmann (MBG); West Berkeley, May 1887, Greene (US); salt marshes, Oakland, Sept. 1886, Congdon (G); Temescal, Alameda Co., Aug. 1891, Michener & Bioletti 123 (G); Crystal Springs Lake, San Mateo Co., July 1903, Elmer 4950 (MBG, P, US); San Luis Obispo, 1876, Palmer 34114 (F, G, NY, US); west Sherman Canyon, 18 June 1901, Braunton 99 (US); Valdez Bay, Santa Cruz Island, 5 Sept. 1927, Jones (P); sandy soil near coast, Santa Catalina Island, 19 July 1915, Macbride & Payson 850 (G); Santa Catalina Island, 20-25 July 1917, Eastwood 6500 (US); dry stream- bed, pebbly beach canyon, Santa Catalina Island, 13 May 1928, Dunkle 1955 (P); Avalon, Santa Catalina Island, Sept. 1896, Trask (MBG); Avalon, May 1897, Trask (F, US); Avalon, Aug. 1901, Trask (NY); Avalon, 28 April 1914, Carlson (US); Avalon, 13 June 1915, Carlson (G, MBG); Jamuel Valley, San Diego Co., 26 June 1875, Palmer (G), 310 (F, MBG, NY) Mexico: Lower CALIFORNIA: near Rancho Salina, foot of Guatay Grade, about 6 km. south of Rio Guadalupe, 11 Sept. 1929, Wiggins & Gillespie 3977 (MBG); Santa Tomas, 15 July 1885, Orcutt 1301 (MBG, US). Verbena robusta has been much confused with the nearly related V. prostrata, but is quite readily distinguished from the latter by the brighter green color of the herbage, the scabrous upper surface of the leaves, and the usually dense spikes. More- over, the mature calyx lacks the marked tendency toward sub- connivent lobes, a characteristic of V. prostrata. The schizo- carps of the two are very much alike. The collection Orcutt 1301 has a greatly elongated inflorescence and the leaves are not particularly scabrous. The specimens Braunton 99 and 1263 show unusually long floral bracts. These are probably atypical phases of the species or possibly hybrids. 24. V. xutha Lehm. Del. Sem. Hort. Hamb. 7, 8. 1834; Linnaea 10: Litt.-Ber. 115. 1835-6. V. strigosa Hook. Comp. Bot. Mag. 1: 176. 1836, not V. strigosa Cham. Linnaea 7: 256. 1832. V. Lucaeana Walpers, Rep. 4: 23. 1844-48. Stems upright, 6-10 dm. tall, branched, hirsute-hispid; leaves oblong or broadly ovate in outline, 5-8(-12) cm. long, incised- pinnatifid or more often trifid with segments coarsely dentate, the lateral much smaller than the middle segment and close to the sessile base of the blade, strigose above, canescent and spread- ing-hirsute below, trichomes particularly prominent along the somewhat paler midrib and veins; spikes elongate, somewhat strict, not compact except at anthek: short-pedunculate; bracts [Vor. 20 204 ANNALS OF THE MISSOURI BOTANICAL GARDEN lanceolate-subulate, subequalling or commonly a little shorter than the calyx, strigillose, ciliate; calyx 3-4 mm. long, strigose- hirsute, lobes acuminate-subulate; corolla-tube approximately as long as the calyx, the throat pubescent without; corolla-limb 5-8 mm. broad; nutlets 2 mm. long, raised-reticulate on the upper half, faintly striate below, commissural faces reaching the apex of the nutlet, muricate or muriculately scabrous. Distribution: Alabama to Texas. Specimens examined: ALABAMA: Navy Cove, Aug. 1889, Mohr (US). Louisiana: without data, Hale (G, MBG), Drummond (G); vicinity of Alexandria, 9 June 1899, Ball 605 (MBG, US); Chopin, 6 May 1915, E. J. Palmer 7556 (MBG); St. Martinville, 30 May 1893, Langlois (MBG, US); Pointe a la Hache, 25 June 1884, Langlois 123 (US); New Orleans, 1832, Drummond 253 bis (K TYPE of V. wider yrs Schriever, 8 June 1917, Munz 1607 (P); Cameron, 4 July 1903, Tracy 8708 ( , MBG, NY, US). RKANBAS: near Homan, 10 June 1898, Eggert (MBG); Texarkana, Aug. 1881, Letterman (MBG, US). Texas: Gladewater, 18 June 1901, Reverchon 2582 (MBG, NY); near Longview, 7 June 1899, Eggert (MBG); Corsicana, 3 Oct. 1900, Reverchon 2118 (MBG, US); Huntsville, 1920, Warner 43 (US); sandy land, Brazos Co., 1899, Ness (G); Brazos, July 1899, Lindheimer (MBG); dry banks of Brazos River, 23 June 1917, Munz 1470 (P); College Station, 24 July 1899, Reverchon (MBG); black land prairie, Montgomery Co., 18-21 July 1900, Dixon 473 (F, G, NY); Burton, 26 May 1872, Hall 434 (F, G, ; , P, US); Austin, 12 July 1920, Tharp 667, 668 (US); Austin, Aug. 1921, Schulz 679, 701 (US); San Marcos and vicinity, 1898, Stanfield (N Y); Industry, 1844, Lindheimer 145 (MBG); Cypress City, May 1877, J. Ball (MBG); near Houston, April 1842, Lindheimer 154 (G, MBG); swamps, Houston, Aug. 1904, Kuntze 23811 (NY); Lotus, about 16 km. west of Houston, 8 Aug. 1921, Ferris & Duncan 8268 (MBG); Strand, Jefferson Co., 9 April 1924, Tharp 3166 (MBG); Galveston, 6 June 1920, Fisher 212 (US); Columbis, 5 Oct. 1900, Bush 1275 (MBG); Columbia, 23 Sept. 1901, Bush 899 (MBG); Hallettsville, 9 Aug. 1912, Fisher 122 (US); Ottine, 30 Aug. 1926, Bogusch 1235 in part (P); Sequin, 17 June 1903, Groth 18? (F, G, NY, US); Cave Lake, Jackson Co., 30 June 1915, Drushel 2842 (MBG) This species is easily distinguished by its erect habit, coarse pubescence, elongated open spike, and sessile commonly trifid leaves. It has often been confused with V. canescens var. Roemeriana, but the latter is a smaller and more compact plant. 25. V. plicata Greene, Pittonia 5: 135. 1903. Coarse herb with stems decumbent to ascending, branched, hirtellous; lower leaf-blades elliptic-ovate, narrowed into a margined petiole of approximately the same length, 1-3(-4) em. long, broadly obtuse, plicate, coarsely incised-dentate, often 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 205 3-lobed, more or less canescent, rugose and somewhat appressed- hirsute above, hirsute beneath and prominently marked (particu- larly near the margin) with whitish veins; upper leaves similar but smaller, often appearing spathulate; spikes terminal on stems and branches, ordinarily not compact; bracts ovate- lanceolate, usually exceeding the calyx, at times barely equalling it, acute, hirsute, midrib often noticeable; calyx 3.5-4 mm. long, more or less glandular-hirsute, lobes very obtuse or subtruncate, terminating abruptly in subulate teeth; corolla-tube scarcely longer than the calyx; corolla-limb 4-6 mm. broad, anterior lobe retuse; nutlets cylindric-trigonous, 2-2.5 mm. long, shallowly scrobiculate above, changing to indefinitely striate toward the base; commissural faces finely muricate-scabrous to practically smooth, not reaching the tip of the nutlet. Distribution: Texas, New Mexico, Arizona, and northern Mexico. Specimens examined: Texas: without data, Berlandier 2506 in part, 644 (MBG); without locality, Sept. 1881, Havard (US); southwestern Texas, Reverchon 118 (G); western Texas, 1851-2, Wright 1496 (G, MBG, NY); sandy roadside, Chillicothe, 28 Sept.-3 Oct. 1906, Ball 1171 (US); sands, Estelline, 25 May 1904, Reverchon 4314 (MBG); sandy waste, Garza Co., 6 June 1925, Ruth 1289 (US); Post, 22 May 1925, Wooton (US); Lubbock, 24 April 1930, Demaree 7539 (US); moist open ground along creeks, Sweetwater, Nolan Co., 27 May 1918, E. J. Palmer 18730 (MBG); north of Colorado, Mitchell Co., June 1900, Eggert (G, MBG); prairie north of Stanton, Martin Co., 13 June 1900, Eggert (MBG); rocky and sandy soils, Comanche, 8 Aug. 1877, Reverchon 834 in part (MBG); near Comanche, 10 May 1900, Eggert (G, MBG); rocky prairies, rown Co., 10 Aug. 1877, Reverchon 737 in part (MBG); Brown Co., April 1882, Reverchon 737 (US); San Angelo, 19 May 1908, Reverchon 1953 (MBG); Barstow, 14 April 1902, Tracy & Earle 30 (F, G, MBG, NY), Tyre collection; Barstow, 15 April- 3 May 1902, Tracy & Earle 41 (NY); Oxona, 13 April 1930, Jones 26221 (P); Davis Mountains, 5 Aug. 1918, Young (P); cliffs back of Fort Davis, Davis Mountains, 9-12 July 1921, Ferris & Duncan 2726 (MBG, NY, P); about 5 km. east of Study Butte, Brewster Co., 30 June 1931, Moore & Steyermark 5299 (MBG); Alpine, 7 June 1926, E. J. Palmer 30523 (MBG); near Boquillas, 17 April 1919, Hanson 619 (MBG, US); Austin, 12 May 1872, Hall 429 (F, G, NY); gravel bars of Blanco River, Blanco, 5 April 1918, E. J. Palmer 13282 (MBG); near Feodora, Terrell Co., 26 April 1928, E. J. Palmer 33537 (MBG, NY); Devils River, Valverde Co., May 1913, Orcutt 6235 (MBG); Corpus Christi Bay, Dec. 1879, Palmer 2038 (G); Laredo, Aug. 1879, Palmer 2040 (G); Laredo, Berlandier 1485 (= 225) in part (G); Laredo, 21 March 1903, Reverchon 3904 (G, MBG, US); Laredo, 1913, Orcutt 5555, 5717 (MBG); near Pharr, Hidalgo Co., 6 April 1931, McKelvey 1756 (G). New Mexico: plains, Carrizoza, 8-19 May 1902, Earle 606 (NY); in valley near Gray, 26 July 1900, Earle 427 (NY, US); neighborhood of San Miguel, 12 Aug. 1847, Fendler 594 (G, MBG). ARIZONA: desert prairie, north of Tucson, 24 April 1913, Greenman & Greenman 28 [Vor. 20 296 ANNALS OF THE MISSOURI BOTANICAL GARDEN (MBG); Tucson, 2 May 1892, T'oumey 806 (US); slopes west of Tucson, 30 Dec. 1919, Bartram 256 (ANSP); Cienega, near Pantano, 14 June 1881, Pringle (F, G, MBG); valley near Camp Lowell, 8 June 1882, Pringle (ANSP, F, NY); near Fort Lowell, 15 Sept. 1900, Griffiths 1595 (NY). Mexico: Crimuanva: Colonia Juarez, Sierra Madre Mountains, 11 Sept. 1903, Jones (P). Coanuita: Sierra Mojada Mountains, 19 April 1892, Jones 372 (P, US). In habit V. plicata is somewhat similar to V. neomexicana, but in inflorescence it strongly resembles V. canescens var. Roemeriana. Its distinetive character is foliar and is most easily seen in the basal and the lower stem-leaves. The leaf is obviously petioled, often 3-lobed, and very shallowly incised-dentate; moreover, the veins beneath are whitish and particularly prominent near the margin owing to the plication of the leaf and the apparent broadening of the veins in this region. 26. V. neomexicana (Gray) Small, Fl. Southeast. U. S. ed. 1, 1010. 1903, and ed. 2, 1913. V. canescens var. neomexicana Gray, Syn. Fl. N. Am. 2!: 337. 1878. V. officinalis var. hirsuta Torr. Bot. Mex. Bound. 2:128. 1859. Plant slender; stems upright, branched, hirsute; leaves 1-5 cm. long, pinnately cleft or almost parted, segments incised or coarsely toothed, rugose, somewhat scabrous and finely pustulate above, the veins more or less prominent beneath, hirsute on both sur- faces; spikes solitary or tending to be panicled, usually short- peduncled, hirsute; bracts lanceolate-acuminate, commonly not longer than the calyx; calyx about 3 mm. long, hirsute-pubescent and very slightly glandular, teeth short and subulate; corolla- tube scarcely longer than the calyx; corolla-limb approximately 4 mm. broad; nutlets trigonous with convex back, 2 mm. long, very shallowly reticulate-scrobiculate on the upper half, longi- tudinally striate below; commissural faces extending to the tip of the nutlets. Distribution: ie and New Mexico. Specimens examin Texas: Fort Davis, 13 Sept. 1918, Young 1703 (US). New Mexico: White Mountains, Lincoln Co., 12 Aug. 1897, Wooton 646 (NY); Ruidoso Creek, Lincoln Co., 3 July 1895, Wooton (NY, US); Kingston, Sierra Co., 6 July 1904, Metcalfe 955 (US); borders of thickets, near Coppermines, 1851, Wright 149? (G tyre, MBG, NY, US); Pinos Altos Mountains, 1880, Greene (F, MBG); 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 207 bottom of Tierra Blanca Canyon, Gila Forest, 18 Aug. 1916, Chapline 609 (NY); G S. Ranch, vicinity of Silver City, 27 Aug.-12 Sept. 1911, Holzinger (US); Mogollon Mountains, on or near West Fork of Gila River, Socorro Co., 28 Au 1903, Metcalfe 612 (G, MBG, NY, P, US The collections of Young 1708, Wooton 646, and Wooton (collection of July 3, 1895) are by no means typical of the species, but perhaps are better placed here than elsewhere. This species appears to be very closely related to V. canescens and V. gracilis. It differs from both, however, in the upright habit and the nutlets. In both V. canescens and V. gracilis the commissural face does not extend to the tip of the nutlet and the striae on the dorsal surface are less conspicuous. 26a. Var. xylopoda Perry,” n. var. Stem somewhat coarser; pubescence shorter, denser and more glandular; calyx 4 mm. long, glandular-hirsute, teeth acuminate; corolla-limb 6-10 mm. broad. Distribution: ia eT California, and northern Mexico. ecimens ex d: ARIZONA: Clarkdale, 17 Sept. 1921, W. W. Jones 344, 346 (G); Rio Verde, 8 Sept. 1865, Cowes & Palmer 571 (MBG); Skull Valley, 28 April 1903, Jones (MBG, P, US); near Oracle, 20 April 1930, Harrison & Kearney 6689 (NY, P); Chiricahua Moun- tains, Cochise Co., 1 May 1894, W. W. Price (P); Paradise, Chiricahua Mountains, 16 Sept. 1907, Blumer 2170 (F, US); Chiricahua Mine, 21 Oct. 1907, Blumer 1804 (F, G, MBG, NY, US); Warren, Cochise Co., 20 May 1915, Carlson (US); Bisbee, 3 Oct. 1892, Mearns 1013 (US); near Fort Huachuca, 1882, Lemmon 2857 (G); Fort Huachuca, Aug. 1892, Wilcor (NY); Huachuca Mountains, 3 Sept. 1903, Jones (P); Ash Canyon, Huachuca Mountains, 6 Aug. 1909, Goodding 334 (G, NY); Huachuca Mountains, 3 Sept. 1928, Harrison & Kearney 5796 (US); Santa Catalina Mountains, 27 July 1917, Munz 1149 (P); Sabino Canyon, Santa Catalina Mountains, 21 April 1922, Hanson A1130 (MBG TYPE); about 13 km. south of Vail, 31 Aug. 1903, Jones (P); foothills of the Santa Rita Mountains, 11 May 1884, Pringle (ANSP, F, G, US); Santa Rita Forest Reserve, 8-13 Sept. 1902, Griffiths 3431 (US); Santa Rita Range Reserve, 12 May 1912, Wooton (US); slopes about Calabasas, 21 April 1908, Tides- trom 872 (US); Baboquivari Mountains, 28 March 1927, Peebles, Harrison & Kearne 8790 (US); Baboquivari Mountains, 26 Sept. 1927, Harrison 4778 (US); Baboquivari Mountains, 12 April 1928, Gilman 1120 (P). CALIFORNIA: without locality, 1876, Palmer (G), 33914 (US). Mexico: Lower CALIFORNIA: Big Canyon of Tantillas Mountains, 10 Sept. 1875, Palmer (G), sis (F, MBG). 20 Var. xylopoda Perry, var. nov., typicam simulans sed indumentum brevius densiusque; calyce 4 mm. longo glanduloso-hirsuto; calycis dentibus acuminatis; corollae limbo 6-10 mm. lato.—Collected on rocky slopes, Sabino Canyon, Santa Catalina Mountains, Arizona, 21 April 1922, Hanson A1130 (MBG), TYPE. or. 20 298 ANNALS OF THE MISSOURI BOTANICAL GARDEN Sonora: Guadalupe Canyon, 27 Aug. 1893, Merton 2042 (US); '"Niggerhead Mountains, near monument no. 82," Aug. 1893, Mearns 1887, 1918 (US); Fronteras, June 1851, Thurber 446 (G, NY). CurIHuAnvA: Saint Eulalia Hills, 30 July 1885, Wilkinson (US); San Diego, 26 April 1891, Hartman 608 (G); rocky hills, near Chihuahua, 16 April 1885, Pringle 270 (ANSP, F, G, US); vicinity of Chihuahua, 8-27 April 1908, Palmer 52 (F, G, MBG, NY, US); “between Sacramento and Chihuahua," 24 Aug. 1846, Wislizenus 150 (MBG); in the Sierra Madre, 21 June-29 July 1899, Nelson 6161 (G, US); near Colonia Garcia, 29 July 1899, T'ownsend & Barber 192 (F, G, MBG, NY, P, US); between Colonia Garcia and Pratt's Ranch, below Pacheco, 22-24 Aug. 1899, Nelson 6271 (G, US). This variety differs from the species in the denser and more glandular pubescence and the larger corolla. The nutlets also are slightly longer, with the reticulations somewhat deeper than in the species, and the commissural faces hardly extend to the tip of the nutlets. The specimens Nelson 6161, 6271 and Town- send & Barber 192 closely resemble the above variety in in- florescence, but are more like V. neomexicana in the long and somewhat sparsely hirsute indument on the lower part of the stem; Pringle 270 and Wilkinson approach variety hirtella. 26b. Var. hirtella Perry," n. var. Plants densely canescent-hirtellous; leaves more or less shal- lowly incised; bracts usually broadly ovate-acuminate; corolla- limb about 8 mm. broad. Distribution: Texas, New Mexico, and Coahuila, Mexico. Specimens examine Texas: rocky ii between Van Horn Wells and one 2 July 1852, Parry, Bigelow, Wright & Schott (NY, US); Sivermore Peak, Davis Mountains, 9-12 July 1921, Ferris & Duncan 2607 (MBG, NY); sand bars of Sieb Davis Mountains, 11 July 1926, E. J. Palmer 30791 (MBG); Valentine, 28 April 1930, Jones 2622/ in part (P); near Shafter, Presidio Co., 26 April 1931, McKelvey 2046 (G); Pinto Canyon, near Ruidosa, 13 April 1919, Hanson 645 (G, NY, US); Chisos Mountains, 12 Aug. 1915, Young 112 (MBG); foothills of Chisos Mountains, 22 May 1928, E. J. Palmer 84065 (MBG "tyre, NY); gravelly mu north side of Chisos Mountains, 27 June 1931, Moore & tavernas 8277 (MBG). New Mexico: Socorro Mountains, 11 July 1897, Herrick 715 (US); low moun- tains west of San Antonio, 14 April 1908, Wooton 3852 (US). Mexico: CoanuiLa: Sierra de Parras, Purpus 1094 (F, G, MBG, NY, P). The pubescence of this variety is much finer and shorter than ?! Var. hirtella Perry, var. nov., planta dense hirtello-canescens; foliis plus minusve breviter incisis; bracteis fere late ovato-acuminatis; corollae limbo circiter 8 m lato.—Collected on ns foothills of the Chisos Mountains, Texas, 22 May 1928, E. J. Palmer 84065 (MBG), TYPE. 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 209 in the above. The leaves are not so deeply incised and often tend to be elongated. The leaves of Purpus 1094 are so narrow and shallowly incised that it appears superficially like V. perennis; nevertheless, the character of the pubescence allies it with this variety. 27. V. perennis Wooton, Bull. Torr. Bot. Club 25: 262. 1898. Stems several from a woody base, divaricately ascending- erect, more or less strictly branched, glabrate or often finely glandular and slightly hispidulous with short stiff antrorse hairs; leaves predominantly linear, 1-4 cm. long, entire or pinnately few-lobed, erect-ascending, sparsely hispidulous, mar- gins revolute; spikes terminal, pedunculate, slender-filiform, elongate; T ovate, 1.5-3 mm. long, acute, hispidulous, ciliate; calyx 4-5 mm. long, pubescence more abundant along the nerves, lobes subequal, short, acute; corolla-tube slightly longer than the calyx, pubescent; corolla-limb 5-7 mm. broad, lobes repand; fruit more or less remote; schizocarp 3 mm. long, strongly constricted along the lines of cleavage; nutlets subcylindric, reticulate-scrobiculate except at base; commissural faces smooth or slightly scabrous and not extending to the tip of the nutlet. Distribution: Texas and New Mexico. Specimens examined: Texas: Glass Mountains, 25 Aug. 1925, Tharp 3682 (US); Loyola, 5 Nov. 1902, Seler (G); Guadalupe Mountains, 1882, Havard 197 (G); Guadalupe Mountains, 13 Aug. 1916, Young (MBG); rough grassy slopes, McKittrick Canyon, Guadalupe Mountains, 23 July 1931, Moore & Steyermark 3611 (MBG); Guadalupe Mountains, 8 Aug. 1931, Clarke 4250 (MBG). New Mexico: Guadalupe, Oct. 1881, Havard (US); Queen, Aug. 1909, Wooton (MBG, US); Queen, 12-20 Aug. 1924, Standley 40686 (US); Otero Co., 17 Aug. 1899, Standley (US); White Mountains, Lincoln Co., 21 July 1897, Wooton 187 (G, MBG, NY, P), type collection; Ruidoso Creek, White Mountains, 5 Aug. 1901, Wooton (MBG); north of El Capitan Mountains, Lincoln Co., 31 Aug. 1900, Earle 387 (MBG, NY, P, US); Capitan, 8-19 May 1902, Earle 619 (NY); Gray, Lincoln Co., 6 June 1898, Skehan 20 (F, G, MBG, NY, P, US); Berendo Creek, 13 May 1905, Metcalfe 1568 (F, G, MBG, NY, US). The relationship of this species is somewhat anomalous. The lobing of the leaves and the character of the nutlets seem to ally it with V. canescens and its relatives; whereas the pubescence and the predominance of practically entire linear-oblong leaves recall V. simplex. It could scarcely be confused with either, [Vor. 20 300 ANNALS OF THE MISSOURI BOTANICAL GARDEN since the character combination of an open spike, very narrow leaves, and sparsely short-strigillose hairs is not found elsewhere in the group under consideration. 28. V. gracilis Desf. Cat. Hort. Paris, ed. 3, 393. 1829. V. remota Benth. Pl. Hartw. 21. 1839. V. arizonica Gray, Proc. Am. Acad. 19: 95. 1883. Low diffusely branched herb; branches decumbent to ascending, more or less glandular, canescent, hirsute; leaves ovate, cuneate at base and narrowed into a margined petiole, 1-3 cm. long, incised-pinnatifid to pinnately cleft, segments oblong, subincised, acutish, midrib and veins impressed, sometimes glandular, hirsute particularly on the lower surface; spike terminal, slender, more or less sparsely flowered, sessile, the lowermost flowers often appearing in the axils of the upper leaves; bracts linear, attenuate, gradually reduced in length toward the tip of the spike, usually much longer than the calyx; calyx 3 mm. long, more or less glandular, hirtellous, subtruncate, teeth minute and subulate; corolla inconspicuous, the tube somewhat longer than the calyx, the limb about 2-3 mm. broad; nutlets 1.5-2 mm. long, finely scrobiculate from the apex practically to the base; commissural faces muricately scabrous. Distribution: Arizona, Utah, and Mexico. Specimens examined: (Par. Tyre, MBG phot.). Described from specimen cultivated in the Botanical Garden of Paris. Arizona: Gardiner's Spring, 25 June 1882, Pringle (ANSP, F, G, NY); Gulching Ground, Pine Canyon, Chiricahua Mountains, 5 July 1907, Blumer 1612 (F, G, MBG, NY, US); San Pedro River, Mexican Boundary Line, 12 Oct. 1892, Mearns 1116 (US); Tanner’s Cafion, near Fort Huachuca, 1882, Lemmon (G); Fort Huachuca, 1890, Patzky (US); Fort Huachuca, Wilcox (NY, US), 29 (US). Uran: mesa east of Monticello, 25 July 1911, Rydberg & Garrett 9201 (NY). Mexico: Cuinuanua: waste places, Chihuahua, 26 May 1885, Pringle 54 (G); near Chihua- hua, 7 May 1887, Pringle (F, MBG); vicinity of Chihuahua, 1-21 May 1908, Palmer 200 (US). DvnaNao: city of Durango and vicinity, April-Nov. 1896, Palmer 135 (F, G, MBG, NY, US), 911 (US); city of Durango, 1 Aug. 1898, Nelson 4593 (MBG, US); Otinapa, 25 July-5 Aug. 1906, Palmer 456 (G, US). AGUASCALIENTES: (Zacatecas according to the Kew specimen): Hartweg 174 (G, K tyre of V. remota, NY). San Luis Porost: region of San Luis Potosi, 1878, Parry & Palmer 722 (G, MBG, US); valley of San Luis Potosi, 1876, Schaffner 720 (ANSP, G, MBG). 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 301 QUERETARO: Queretaro, 2100 m. alt., July 1904, Kuntze 23444 (NY). Hipauco: Nopala, 1 Aug. 1913, Salazar (US); bare hills above Pachuca, about 2500 m. alt., 23 July 1898, Pringle 7590 (F, MBG, P); dry calcareous mesas, near Metepec Station, about 2500 m. alt., 22 June 1904, Pringle 13159 (F, G, US). Mexico: vicinity of Mexico, ? Berlandier 578 (MBG, US); valley of Mexico, Pedrigal, about 2200 m. alt., 1 Sept. 1896, Pringle 6539 (ANSP, F, G, MBG, NY, US); Santa Fe, 26 June 1865-6, Bourgeau 361 (G); near Tlalpam, 15 July 1901, Rose & Hay 5488 (US); near Tlalpam, July 1905, Rose, Painter & Rose 8495 (G, US); lava beds, Tizapan, about 2250 m. alt., 24 Aug. 1900, Pringle 9135 (G, MBG, NY). Oaxaca: near Mitla, June 1888, Seler 22 (G). V. gracilis is probably a relative of V. canescens and V. neo- mexicana. It is easily recognized by its slender habit, usually long-attenuate bracts, tiny flowers, and essentially scrobiculate nutlets. The specimen from Utah is atypical, but for practical purposes seems better referred here. 29. V. canescens HBK. Nov. Gen. et Sp. 2: 274, pl. 136. 1818. Low coarse herb; stems several and branched, decumbent to ascending, canescent, hirsute; leaves oblong-lanceolate, 1-5 em. long, acute, remotely incised-dentate or subpinnatifid, contracted into a margined entire semiamplexicaul or petiolar base, rugose and hirtellous above, the trichomes often with minute bulbous bases, canescent-hirsute and somewhat conspicuously veined beneath, margins revolute; spikes sessile or short-peduncled, mostly solitary, loose-flowered or compact, glandular-hirsute; bracts lanceolate, acuminate, variable in length, often exceeding the calyx, hirsute, ciliate; calyx about 3 mm. long, subtruncate, with very short inconspicuous teeth; corolla-tube slightly longer than the calyx; corolla-limb 4-6 mm. broad, lobes retuse; nutlets subcylindrical, 2 mm. long, raised-reticulate above, longitudinally striate at least half way from base to apex; commissural faces muricate-scabrous, not reaching the tip of the nutlets. Distribution: Nevada and Mexico. Specimens examined: Nevapa: Caliente, 9 May 1892, Jones 554 (P). Mexico: CoanuirLa: Saltillo and vicinity, 15-30 April 1898, Palmer 25 (F, G, MBG, NY, US); Saltillo, 1200 m. alt., 22 Aug. 1913, Adole 22 [Arsène 10626] (F, G, MBG, NY, US); valley near Saltillo, 1848-49, Gregg 99, 246 (MBG); canyon and elevated portion of Sierra Madre, south of Saltillo, 25 July-1 Aug. 1880, Palmer 1047 (ANSP, G, US); [Vor. 20 302 ANNALS OF THE MISSOURI BOTANICAL GARDEN near Buenavista, 22 ry "e^ Gregg (NY), 740 (MBG); near San Pablo, 29 April 1847, Gregg (NY), 645 (MBG). ZACATECAS: near Conception del Oro, 11-14 Aug. 1904, phe 803 (F, G, MBG, NY, US); Zacatecas, 1800-2100 m. alt., Aug. 1903, Purpus 467 (US). AGUASCALIENTES: stony places, rim Calientes, Hartweg 177 (G, NY); near Aguas Calientes, 20 Aug. 1901, Rose & Hay 6212 (NY, US); near Ape Calientes, 9 Oct. 1903, Rose & Painter 7741 (G, NY, US). San Luis Porosi: Angostura, June 1911, Purpus 6515 (F, G, NY, US); in sands around San Luis Potosi, 1876, Schaffner 719 (ANSP, G); San Luis Potosi, 1878, Rose, Painter & Rose 8971, 9162 (US). EBLA: Cerro Guadalupe, 20 June 1909, Nicolas (US); margin of Atoyac, 20 June 1910, Nicolas (US); barren hills about Esperanza, 2660 m. alt., 17 Aug. 1905, Pittier bed (US VER UZ: Ceiba, M Mueller 1215 (NY); Esperanza, 2400 m. alt., 14 Aug. 1891, Boson 822 (F, G, U : Cerro de r Ae 16 Oct. 1921, Conzatti 4288 (US); dry calcareous hills, Las Sedas, 1800 m. alt., 11 Aug. 1894, Pringle 4784 (ANSP, G, MBG, NY, US); Las Sedas, about 1900 m. alt., 27 June 1895, L. C. Smith 412 (G); Las Sedas, 1950 m. alt., 30 Aug. 1921, Conzatti 4194 (US). This is a low foliose species closely related to V. neomexicana and V. plicata, but distinguished from them by the semiamplexi- caul elongate leaves and rather characteristic compact habit. The specimens cited from Coahuila are excellent examples of phases intermediate between the species and the variety. The collection from Nevada seems somewhat out of range. Possibly it was carried in as a weed or escaped from cultivation. 29a. Var. Roemeriana (Scheele) Perry, new comb. V. Roemeriana Scheele, Linnaea 21: 755. 1848. Plants more densely hirsute, sparsely (if at all) glandular; floral bracts broadly ovate at base, abruptly acuminate, concave at anthesis, later slightly recurved, for the most part overtopping the calyx; dorsal surface of the nutlets reticulate-scrobiculate approximately two-thirds of their length, striate toward the base, nutlets usually more slender than in the species. Distribution: Texas and Mexico. )ecimens examine Texas: without ioonlity:, 1846, Lindheimer 500 (G, MBG, US); without locality, 1885, Reverchon 737 in part (MBG); Red River, Wichita Co., 10 April 1922, Tharp 1361 (US); sandy soils, Brown Co., Reverchon 1961 (G, NY); near Brownwood, Brown Co., 1 Nov. 1925, E. J. Palmer 29535 (MBG); Tom Greene Co., 1879, Tweedy (NY); Dove Creek, Tom Greene Co., May 1880, Tweedy 118, 246 (US); San Angelo, 13 May 1903, Reverchon 3903 (MBG); Legion Creek, Gillespie Co., Jermy 182 (MBG); gravel bars of Blanco River, Blanco, 5 April 1918, E. J. Palmer 13283 (MBG); 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 303 Austin, 16 May 1872, Hall 430 (F, G, MBG, NY, P, US); Austin, 1 April 1918, Young 112 (MBG); San Antonio to Austin, 18 April 1925, Small & Wherry 11969 (NY); Dryden, 22 April 1930, Jones 26226 (MBG, P); Langtry, Valverde Co., May 1913, Orcutt 6111 (MBG); rocky hills at Devils River, 10 Sept. 1900, Eggert (MBG); Lacey's Ranch, Kerr Co., 1 June 1916, E. J. Palmer 10002 (MBG); Kerrville, 7-14 May 1894, Heller 1782 (G, MBG, NY, US); San Marcos, 6 Nov. 1897, T'release (MBG); New Braunfels, April 1850, Lindheimer 294 (MBG), 1074 (F, G, MBG, NY, US); Ozona, 30 April 1930, Jones 26222 (P); Rock Springs, 17 April 1930, Jones 26223 (P); Del Rio, 20 April 1930, Jones 26225 (P); Fort Clark, Kinney Co., 27 Feb. 1893, Mearns 1237 (US); Fort Clark, 10 May 1893, Mearns 1456 (US); gravel and sand deposits along stream, Uvalde, 11 May 1918, E. J. Palmer 13563 (MBG); dry rocky ground along small stream, Uvalde, 24 April 1928, E. J. Palmer 33605 (MBG); near Uvalde, 30 April 1928, E. J. Palmer 33646 (NY); 16 km. west of Uvalde, 8 June 1931, Moore & Steyermark 3005 (MBG); near Verde Creek, near Hondo, San Antonio to Del Rio, 21 April 1925, Small & Wherry 11980 (NY); San Antonio, April 1911, Clemens & Clemens 9?4 (MBG, P), 971, 972, 973 (P); San Antonio, 16 March 1916, E. J. Palmer 9188 (MBG); on limestone hill, near Bracken, 27 June 1903, Groth 16 (F, G, NY); in fields near Rio Frio, July 1829, Berlandier 2054, 644 in part (G); El Jardin, 10 March 1924, Runyon 629 (US); woods near Colorado River, near Wharton, 12 April 1925, Small & Wherry 11826 (NY); stony hills near Goliad, 9 April 1900, Eggert (G, MBG); near Mathis, San Patricio Co., 5 April 1931, McKelvey 1710 (G); west San Diego, 31 July 1931, Clark 4041 (MBG). Mexico: NveEvo Leon: Walnut Grove, 27 May 1847, Gregg 791 (MBG, NY); Monterey, 26 May 1847, Wislizenus 825 (MBG); Monterey, 17-26 Feb. 1880, Palmer 1044 (ANSP, G, US); Pico Chico, near Monterey, 19 March 1900, Canby 194 (US); Monterey, 26 July 1926, Fisher 235 (US). TAMAULIPAS: vicinity of Victoria, 320 m. alt., 1 Feb.-9 April 1907, Palmer 82 (F, G, MBG, NY as no. 8, US); from Santander to Victoria, Nov. 1830, Berlandier 827 (G, MBG, NY); Jaumave, 1931, von Rozynski 13, 17, 185 (F). Differing from the species chiefly in its coarser less glandular pubescence and the broadly ovate abruptly acuminate bracts. A few of the above-cited specimens show transitional phases, but, as a whole, are closer to the variety. 30. V. subuligera Greene, Pittonia 1: 156. 1888. Stems probably several from a common base, procumbent, hirsute-pubescent; leaves ovate, gradually narrowed into a sub- petiolar base, 1.5 to 4 cm. long, pinnately cleft, larger segments incised or sharply dentate, appressed-hirsute on both surfaces but especially beneath on the somewhat prominently reticulated veins; spikes short-petiolate, elongate, conspicuously bracted, in fruit interrupted; bracts lanceolate-subulate, twice as long as the calyx, ascending-spreading, reflexed in age, hirsute; fruiting calyx about 3 mm. long, pubescent, hirsute along the nerves, [Vor. 20 304 ANNALS OF THE MISSOURI BOTANICAL GARDEN lobes very short, strongly connivent over the schizocarp; corolla- tube scarcely longer than the calyx; nutlets subcylindric-trigonous, 2 mm. long, only faintly striate; commissure muriculate or smoothish, not reaching the top of the nutlet. Distribution: Mexico. Specimens examined: Mexico: DunaNGo (?): "from the Sierra Madre, west of Durango," about 2400 m. alt., Sept. and Oct. 1881, Forrer (ANSP, F, G, NY, US), Tyre collection. Although the general habit of this species is much like that of V. bracteata, the latter differs in having obtusish leaf-segments and sharply reticulate-striate nutlets with commissural faces fully as long as the nutlet. 31. V. bracteata Lag. & Rodr. in Anal. Cienc. Nat. 4: 260. 1801. V. bracteosa Michx. Fl. Bor.-Am. 2: 13. 1803. V. squarrosa Roth, Catalect. Bot. 3: 3. 1800. V. canescens Chapman, Fl. Southeast. U. S. 307. 1860. V. bracteosa var. brevibracteata Gray, Syn. Fl. N. Am. 2!: 336. 1878. V. rudis Greene, Pittonia 4: 152. 1900. V. confinis Greene, Pittonia, l. c. V. bracteosa var. albiflora Cockerell in Daniels, Fl. Boulder, Colo. 204. 1911. V. imbricata Wooton & Standley, Contr. U. S. Nat. Herb. 16: 166. 1913. Zapania bracteosa Poir. in Lam. Encyc. 8: 843. 1808. Stems usually several from a common base, diffusely branched, decumbent or ascending, rarely erect, coarsely hirsute; leaves 1-4(-6) em. long, pinnately incised or usually 3-lobed (lateral lobes narrow and divaricate, middle lobe large, cuneate-obovate, incisely toothed or cleft), narrowed into a short margined petiole, hirsute on both surfaces, midrib and veins slightly prominent beneath; spikes terminal, sessile, comparatively thick, conspicu- ously bracted; bracts much longer than the calyx, spreading- ascending, recurved in age, coarsely hirsute, the lowermost often incised and leaf-like, the upper linear-lanceolate, acute-acuminate, entire; calyx 3-4 mm. long, hirsute particularly along the nerves, 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 305 lobes very short, connivent over the schizocarp; corolla-tube protruding slightly beyond the calyx, very finely pubescent outside the throat; corolla-imb 2.5-3 mm. broad; nutlets tri- gonous, approximately 2 mm. long, sharply raised-reticulate above, striate below; commissural faces reaching to the distal end of the nutlet, muricately scabrous. Distribution: waste places, Maine to Florida, west to California, and northern exico. Specimens examined: (Herb. Bot. Gard. Madrid Tyre, MBG phot.). Described from specimen culti- vated in the Botanical Garden of Madrid. Ontario: Point Edward, Lake Huron, 19 July 1901, Macoun (NY). Maine: Cumberland, 12 July 1902, Chamberlain 418 (G). MassAcHUSETTS: Dedham, 4 Aug. 1900, Rich (G); Worcester, 27 June 1904, Dewhurst (G). Connecticut: Winsted, Winchester, 5 Sept. 1909, Fernald (G); Bridgeport, 4 Sept. 1908, Hames 8137 (G New York: Montauk Paint. 23 July 1895, von Schrenk (MBG New JEnsEY: Weehawken, Aug. 1894, Van Sickle (US); Kaisa Point, Camden, 20 July 1866, Parker (G, MBG). PENNSYLVANIA: "Cementon," valley of Lehigh River, 20 Aug. 1923, Churchill (G); Lancaster, 29 Aug. 1900, Heller (US). Nortu Carona: mountains of North Carolina, June 1872, Leroy & Ruger (NY); Statesville, June 1872, Ruger (US GEORGIA: Cartersville, Ravenel (NY); Thompsons Mills and vicinity, Gwinnett Co., 14 May 1908, Allard 207 (MBG, US); Stone Mountain, 24 May 1897, Eggert MBG); Ocmulgee River, swamp palen Macon Island, 8-9 July 1895, Small (NY); Smithville, 18 Aug. 1885, J. D. Smith (MBG). LORIDA: without data, Buckley (MBG), Chapman (MBG); Apalachicola, 5 Aug. 1889, Biltmore Herbarium 1082b (NY, US). ALABAMA: roads and uncultivated fields, Sept. 1843, Rugel (MBG); Stevenson, 5 Sept. 1877, Ward (US); Valley Head, July 1898, Ruth 495 (NY), 518 (MBG); Wilcox Co., 18 May 1840, Buckley (NY); sandy ground, near Atalla, Etowah Co., 9 July 1898, Eggert (MBG, NY); Auburn, 13 May 1898, Earle & Baker (P, US); Mobile, 22 May 1884, Mohr (US). LouisiANA: vicinity of Alexandria, 5 June 1899, Ball 553 (G, MBG, NY, US). B Sandusky, 14 Aug. 1920, Moseley (G). Micuiaan: Vincent Lake, Cheboygan Co., 15 Aug. 1917, Ehlers 627 (G); Manistee, 8 Aug. 1882, Morong 1O D INDIANA: of Bass Lake, Steuben Co., 25 July 1906, Deam 1272 (MBG, NY, US); Michigan City, 6 July 1903, Mell & Ea (MBG). Kentucky: without data, Rafinesque (G); southern Louisville, 19 Aug. 1892, —€— vog ENNE : Kin ngston Spring, 19 Aug. 1897, Eggert (MBG); Sherwood, 9 June 1897, oed (MBC): Memphis, 22 June 1851, Fendler (G). Wisconsin: Fort Howard, 22 July 1887, Schuette (G, US); Madison, 27 Aug. 1893, Churchill (G); near Mirror Lake, Sauk Co., 13 July 1903, Eggert (MBG). [Vor. 20 306 ANNALS OF THE MISSOURI BOTANICAL GARDEN Inurors: Fountaindale, 1887, Bebb (G); Elgin, 27 Aug. 1912, Sherf 1797 (MBG); Havana, 18 Aug. 1904, Gleason (G); Monticello, 17 June 1886, Seymour & Waite (G); Athens, Hall (US); East St. Louis, 11 June 1879, Eggert (P). Minnesota: Princeton, July 1892, Sheldon (G, US); Collegeville, 29 July 1912, Chandonnet (MBG); hills between Minnehaha and Fort Snelling, 5 July 1888, Schuette (G); Fort Snelling, 26 July 1899, Mearns 524 (US); Lake City, 25 July 1883, Manning (G). Iowa: Lawler, 1890, Rolfs (G); Fayette Co., 30 July 1894, Fink (US); Black Hawk Co., 15 July 1929, Burk 542a (MBG); Ames, 5 Aug. 1896, Ball (G, MBG, NY, US); ‘lene, Aug. 1904, Fawcett 12 (G, MBG, NY, US); Council Bluffs, Vasey (G). Missourtr: Hannibal, Davis 2691, 3587, 9089 (MBG); Eolia, 26 May 1914, Davis 2414 (MBG); St. Louis, Aug.-Sept. 1838, Riehl 9 (MBG); St. Louis, Sept. 1841, Engelmann (G, US); Jefferson Ave., St. Louis, 2 July 1875, Eggert (MBG); Kimms- wick, 15 July 1885, F. Wislizenus 284 (MBG); Williamsville, Wayne Co., 27 June 1914, E. J. Palmer 6106 (MBG); Poplar Bluff, 12 July 1930, Kellogg 16276 (MBG); Sibley, 30 June 1906, Bush 4010 (G, MBG); Courtney, Bush 2109, 8537 (MBG); bluffs of Pomme de Terre River, Polk Co., 29 July 1891, Trelease 715 (MBG); vicinity of Gates, 26 Aug. 1912, Standley 9362 (US); Joplin, Jasper Co., 10 July 1897, Trelease 494 (MBG); Webb City, 1 July 1903, E. J. Palmer 454 (MBG); Swan, 28 Sept. 1899, Bush 601 (MBG). ANSAS: Monette, 14 June 1927, Demaree 3353 (MBG); Jonesboro, 15 June 1927, Demaree 8371 (G, MBG); Batesville, 1 Sept. 1897, T'release (MBG); Fayette- ville, Harvey 62 (MBG); Fort Smith, 1853-4, Bigelow (US); Fulton, 22 Sept. 1900, Bush 1034 (MBG). RTH DAKOTA: Leeds, 10 Sept. 1896, Lunell (G, US); Fairmount, 22 July 1912, Bergman 2370 (MBG); Mandan, 1915, Sarvis 122 (US); Dickinson, 14 Aug. 1908, Holgate (G, NY). SovrHu Dakora: Swan Creek, 20 July 1911, Visher 3859 (MBG); White River Valley, 8 July 1911, Visher 2028 (NY); Deadwood, 23 July 1913, Carr 179 (G, MBG, US); Lead City, 7 July 1892, Rydberg 934 (NY, US); Black Hills, near Fort Meade, 19 June 1887, Forwood sd (US). NEBRASKA: "very common on sandy banks of Yellowstone River," 1853-4, Hayden 10 (MBG); gt. James, 24 June 1893, Clements 2612 (G, US); Lincoln, 20 June 1888, Webber (US); near Dix, 28 Aug. 1926, Heller 14303 (MBG). Kansas: prairie, Riley Co., 18 June 1895, Norton 392 (G, MBG, NY, US); Kearney Co., 27 Aug. 1897, Hitchcock 1129 (MBG); bottom lands of Arkansas River, south of Kendall, 15 June 1929, Rydberg & Imler 1008 (MBG); Ulysses, 27 June 1893, Thompson 42 (MBG, NY, US); vicinity of Richfield, 20 Sept. 1912, Rose & Fitch 17106 (US). KLAHOMA: Sapulpa, 22 July 1894, Bush 436 (MBG); near Page, 20 June 1914, Blakeley 1462 (G); near Alva, 1913, Stevens 68014, 2889 (G); Woods Co., 21 June 1899, White 206 (MBG); vicinity of Fort Sill, 22 May 1916, Clemens 11750a (MBG); near Snyder, 23 June 1913, Stevens 1192 (G, NY); near Camp, 11 May 1913, Stevens 891 (MBG). Texas: Dallas, Reverchon 736, 2116 (MBG); Tarrant Co., Ruth 109 (G, NY); Fort Worth, Ruth 109 (US); Polytechnic, Ruth 109 (MBG); Abilene, 22 May 1902, Tracy 8001 (G, MBG, NY, US); west of Big Spring, Sept. 1881, Havard (US); Amarillo, 2 Sept. 1910, Ball 1676 (US); banks of Red River, Randall Co., 13 Aug. 1900, Eggert (MBG); Canyon, 2 June 1918, E. J. Palmer 18861 (MBG); Lubbock, 10 June 1917, E. J. Palmer 12496a (MBG); El Paso, 7 June 1917, Clemens (P); below 1933] PERRY— NORTH AMERICAN SPECIES OF VERBENA 307 Dofiana, Parry, Bigelow, Wright & Schott (US); valley of Rio Grande, 64-80 km. below El Paso, ? 1851, Wright 567 (G), pu (G, MBG, NY); valley of Rio Grande, St. Elizani, May-Oet. 1849, Wright 454 (G, US). ALBERTA: Crow's Nest Pass, Aug. 1897, Macoun 24268 (N Montana: about 20 km. above Glendive, 17 July 1883, bue (US); near Glendive, 19-21 Aug. 1884, J. Ball (G); Thompson's Falls, 27 Aug. 1892, Sandberg, MacDougal & Heller 972 (G, P, US). Wyomina: Yellowstone National Park, 1 Aug. 1902, Mearns 2698 (US); Mammoth Hot Springs, 22 Oct. 1902, Mearns 5009 (NY, US); Casper, 6 July 1901, Goodding 207 (G, MBG, NY, P, US); Blue Grass Hills, 8 July 1894, A. Nelson 320 (G, MBG, NY, US); Laramie Peak, 8 Aug. 1895, A. Nelson 1652 (P); Laramie, 24 July 1900, A. Nelson 7671 (G, MBG, NY, P, US). Coronapo: Fort Collins, 12 July 1884, Sheldon 38 (G); Denver, 10 Aug. 1910, Eastwood 25 (G, MBG, US); Manitou, 12 Aug. 1901, Clements 9 (G, MBG, NY, US); Cheyenne Canyon, Colorado Springs, 16 Aug. 1915, Drushel 4887 (MBG); vicinity of La Junta, 26 Sept. 1913, Rose & Fitch 17503 (MBG, US); Deer Run, 25 Aug. 1901, Baker 920 (G, MBG, NY, P, US); Cimarron, 29 July 1901, Baker 288 (G, MBG, NY, P); Arboles, 18 June 1899, Baker 564 (F, G, MBG, P), Tyre collection of V. rudis. EW Mexico: without locality, 1847, Fendler 587 (ANSP, G, MBG, US), 592 (MBG); Raton, 21-22 June 1911, Standley 6268 (US); Cimarron Canyon, 21-24 Aug. 1903, Griffiths 5560 (MBG); Farmington, 8 Aug. 1904, Wooton 2831 (US TYPE of V. imbricata); Las Vegas, 1927, Arsène 18543, 18593, 18634, 18885, 18961, 18964 (all P); vicinity of Las Vegas, 1926, Arsène 18399 (ANSP), 18407 (MBG); near Pecos, San Miguel Co., 20 Aug. 1908, Standley 5136 (MBG, NY, US); mouth of Indian Creek, Pecos River National Forest, 25 July 1908, Standley 4549 (G, MBG, NY, US); Upper Rio Tesuque, 24 July 1908, Standley 4458 (MBG, US); Santa Fe, 20 July 1926, Arsène & Benedict 15734 (MBG); vicinity of Bernalillo, 11 July 1926, Arsène & Benedict 16790 (ANSP); Sandia Mountains, 19 Aug. 1926, Arsène & Benedict 16600 (MBG); Sandia Mountains, July-Aug. 1914, Ellis 221 (MBG, US); Capelin Canyon, 14 Aug. 1914, Ellis 221 (NY); near Albuquerque, 20 June 1926, E. J. Palmer 31117 MBG); Roswell, 24 Aug. 1900, Earle 351 (MBG, P, US); Queen, 2 Aug. 1909, Wooton (MBG, US); near Fort Craig, 1 Aug. 1880, Rusby 336 (MBG, NY); Organ Mountains, Dona Ana Co., 30 Aug. 1897, Wooton 409 (G, MBG, NY, US), TYPE collection of V. confinis; Mesilla Valley, Dona Ana Co., 6 Aug. 1907, Wooton & Standley 3330 (F, MBG); Kingston, 18 June 1904, Metcalfe 1008 (G, MBG, NY, P, US); Berendo Creek, Sierra Co., 20 May 1904, Metcalfe 897 (MBG); Gila River bottom, near Cliff, Grant Co., 13 June 1903, Metcalfe 137 (G, MBG, NY, P, US). Arizona: Tuba Oasis, 15-31 July 1920, Clute 93 (G, MBG, NY, US); vicinity of Flagstaff, 13 July 1898, MacDougal 286 (ANSP, G, NY, US); Flagstaff, 13 Aug. 1922, Hanson A.147 (MBG); near Prescott, 26 July 1927, Peebles, Harrison & Kearney 4243 (US); 16 km. west of MeNary, 23 June 1930, Goodman & Hitchcock 1817 (MBG, NY). Ipano: valley of Clearwater River, Nez Perces Co., 30 May 1892, Sandberg, Mac- Dougal & Heller 264 (G, MBG, NY, P, US); Falk's Store, Canyon Co., 7 July 1910, Macbride 200 (MBG, US); alkaline flats, sink of Big Lost River, 16 Aug. 1895, Henderson 4070 (US); near St. Anthony, 4 July 1901, Merrill & Wilcox 776 (G, NY, US); Boise, July 1892, Mulford (G, MBG); Boise, 1911, J. A. Clark 55 (G, MBG, P, US), 279 (G, MBG, NY, P, US). Uran: Oquirrh Mountains, Salt Lake City, 9-16 July 1902, Pammel & Blackwood 8589 (G, MBG); Salt Lake City, June 1869, Watson 823 (G, NY, US); Gunnison, 25 [Vor. 20 308 ANNALS OF THE MISSOURI BOTANICAL GARDEN June 1875, Ward 256 (G, US); Greenriver, 22 June 1894, Jones 5478 (MBG, NY, P, US); along Bullion Creek, above Marysvale, 21 July 1905, Rydberg & Carlton 7048 (G, ule US); Milford, 4 June 1902, Goodding 1022 (MBG). NEvADA: Winnemucca, 1 Sept. 1897, Hillman (P); saline flats, Las Vegas, 6 May 1905, Goodding 2312 (G, MBG, NY); Colorado River bottoms, 24 km. east of Search- light, 6 June 1915, Parish 10288 (G, MBG). RITISH COLUMBIA: Kamloops, 13 June 1889, Macoun (NY, US). ASHINGTON: without locality, 1889, Vasey 468 (G, NY, US); Cascade Mountains to Fort Colville, about lat. 49°, 1860, Lyall (G); Oroville, 26 June 1911, Jones (P); Meyers Falls, 21 Aug. 1902, Kreager 4?4 (G, NY, US); near Egbert Spring, Douglas Co., 1 July 1893, Sandberg & Leiberg 340 (G, MBG, NY, US); Sentinel Bluffs, 15 July 1903, Cotton 1363 (G, US); Pullman, 21 July 1896, Elmer 324 (MBG, NY); Wawawai, June 1896, Elmer (P); near river, Prosser, 10 July 1902, Cotton 621 (G, MBG, US OREGON: without locality, 1871, Hall 398 (G, MBG, NY); margin of shallow lake, Eastern Oregon, 25 June 1898, Cusick 1967 (G, MBG); Snake River at mouth of Cache Creek, Wallowa Co., 28 May 1897, Sheldon 8202 (MBG, NY); about 5 km. above mouth of Clark’s Creek, 9 Sept. 1897, Sheldon 8856 (MBG, NY, US); Pilot Rock, 2 Sept. 1896, Brown 58 (MBG, US); Pilot Rock, July 1902, Grifiths & Hunter 18 (NY, US); Butter Creek, Umatilla Co., 14 Sept. 1894, Leiberg 902 (G, MBG, NY, US); about 16 km. west of Boardman, 17 July 1928, Thompson 4880 (G, MBG, US); between Mosier and Rowena, 28-30 July 1922, = 9487 (MBG, P); The Dalles, on the Columbia, 2 Aug. 1880, Englemann (MBG). CaLrFORNIA: near Monterey, Bolander 426 (G); Jolon, 1880, Vasey 512 (US); between Tulare and Tulare Lake, 25-30 Aug. 1892, Palmer 2699 (US); near Lone Pine, Inyo Co., 14 June 1891, Coville & Funston 953 (US); near San Bernardino, 3 June 1891, Parish & Parish 2171 (NY, US); Soldiers Home, 20 June 1902, Abrams 2574 (G, MBG, NY, P); near ponds, Laguna, Munz 2207, 6349, 6598 (P); Laguna Canyon, 26 July 1916, Crawford (MBG, P); Blue Lake, Imperial Valley, 1 April 1903, Abrams (MBG, NY). Mexico: Lower Cauirornia: Tia Juana, Orcutt (F, x 1228 (MBG); Ensenada, 31 Aug. 1886, Ming 1549 (MBG); Mancadero, 31 Aug. 1889, Orcutt 1546 (MBG). Sonora: Alamo, 20 May 1925, Kennedy 7088 2 (US). sunu tg 80 km. south of Juarez, 1911, Stearns 7 (F, US); Bolsom de Mapini, 13 April 1847, Gregg 483 (MBG Pio ILA: near San Juan, valley of Nazas, 11 May 1847, Gregg 686 (MBG); — of Nazas, near San Lorenzo, 11 May 1847, Gregg 623 (MBG); El Toro, near Movan July 1910, godes 4524 in part (US); San Lorenzo de Laguna and vicinity, inre west of Parras, 1-10 May 1880, Palmer 1048 (ANSP, G, US). A readily recognized and widely distributed species varying greatly in habit and pubescence as well as in length and position of floral bracts. The typical form is a sprawling hirsute plant with laxly ascending branches and spreading bracts. In New Mexico it appears as a nearly erect herb (V. imbricata Woot. & Standl.) with shorter and ascending or appressed bracts (var. brevibracteata Gray). Some phases of this variation occur 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 309 throughout the northwest range of the species, but since these tendencies to vary are apparently lacking in correlation, they do not seem to be worthy of nomenclatorial attention. 32. V. carnea Medic. Bot. Beobacht. 1783: 131. 1784. V. caroliniana Michx. Fl. Bor.-Am. 2: 14. V. carolinensis Small, Fl. Southeast. U. S. ed. 1, 1009. 1903, and ed. 2, 1913. Phryma caroliniensis Walt. Fl. Car. 166. 1788. Styleurodon caroliniamum Raf. Fl. Tellur. 2: 104. 1836. Cinereous perennial; stems simple or sparingly branched above, ascending, puberulent; leaves spathulate to oblong or the upper oblong-hastate, sessile, shallowly serrate-dentate, scabrous above, less harsh and spreading-pubescent along the prominently reticulated veins of the lower surface; spikes terminal, peduncu- late, chiefly solitary but occasionally in 3’s, slender with crowded tips, more open in fruit; bracts lanceolate, Mou one-half shorter than the calyx, glandular-pubetcetitl calyx about 5 mm. long, glandular-pubescent, lobes acute, unequal; corolla-tube slightly longer than the calyx, pubescent without; corolla-limb about 5 mm. broad, segments somewhat truncate; anthers glandless; schizocarp 3 mm. long, not readily separating into 4 nutlets, longitudinally sulcate and commonly scrobiculate on the upper part. Distribution: North Carolina to Florida, and west to Texas. Specimens examined NonTH Carouina: Pinehurst, 3 Sept. 1897, Katzenstein (G); open pine woods 3 km. south of James City, 11 July 1922, Randolph (G); Burgaw, Aug. 1878, Hyams (US), June 1879, Hyams (MBG) Soutu CAROLINA: Santee Canal, Ravenel (G); sandy open pine woods near Navy Yard, Charleston, 27 April 1912, Robinson 26 (G); Summerville, June 1891, Taylor (US); Bluffton, 1881, Mellichamp (US); sand bank near Orangeburg, 8 May 1907, House 8276 (NY); Aiken, July 1870, Ravenel (NY, US); sandy ground north of Graniteville, 22 May 1899, Eggert (MBG). Greonara: sandy hills north of Augusta, 22 May 1899, Eggert (MBG); Augusta, 14 May 1900, Cuthbert pe: A E Alexander, Ellis (P); Flint River at Albany, 24-28 May 1895, Small (F, FLORIDA: siht nS Buckley (G, MBG); without locality, Chapman (NY); Lake City, 27 Feb. 1893, Rolfs 191 (MBG); pine barren, Duval Co., 29 April 1902, Fredholm 5136 (G, MBG, US); Oakwood, Duval Co., 1 June 1893, Fredholm 110 (P); dry pine barrens, near Jacksonville, May, Curtiss 1959 (G, MBG, NY, US), 8 May 1884, Curtiss 4765 (G, NY, US), 29 April 1893, Curtiss 4386 (MBG, US); [Vor. 20 310 ANNALS OF THE MISSOURI BOTANICAL GARDEN Hibernia, March 1869, Canby (G, NY, US); St. Augustine, 1877, Reynolds (NY Gainesville, 11 May 1925, O'Neill 986 (US); dry sandy woods, Irvine, 29 April 1930, Moldenke 1091 (MBG, NY); Ocala, Marion Co., 2 April 1879, J. D. Smith (US); high pine land, vicinity of Eustis, 1-15 May 1894, Nash 601 (G, MBG, US); pine lands at Lake Helen, 28 April 1906, Deam 1799 (US); Fort Meade, March 1880, J. D. Smith (US); between Tallahassee and St. Marks, April 1843, Rugel (NY); sandy open dep near Tallahassee, 9 April 1929, E. J. Palmer 35235 (G, MBG); Aspalaga, Chapman (MBG); y T barrens, near Apalachicola, 11 Aug. 1889, Biltmore Miler: peri (G, NY, U LABAMA: without locality, 1859, aag (G); Auburn, Lee Co., 10 May 1896, Earle & Underwood (NY); Auburn, 10 Feb. 1897, Earle & Baker (NY); Pinewoods, Buckley (NY); woods, Spring Hill, 6 Aug. 1897, Bush 312 (MBG, NY, US); Spring Hill, 25 June 1915, Drushel (MBG); open barrens, Spring Hill, June-July 1919, Graves 621 (MBG, US); Mobile, May 1875, Curtiss (MBG); Mobile, June 1879, Mohr (US). Mississippi; sandy soil, southern Mississippi, 1859, Hilgard (MBG); Ocean Springs, May 1892, Seymour & Earle 118 (G, MBG); Ocean Springs, 8 May 1895, Skehan (MBG); Biloxi, 21 April 1898, Tracy 4981 (G, MBG, NY, US); Long Beach, 6 Aug. 1891, Joor (MBG); Harrison Co., 19 April 1927, Woodson & Anderson 1577 (MBG). UISIANA: vicinity of Covington, 30 April 1920, Arséne 12117 (US), 8 May 1920, Arsène 11820 (US). Texas: sandy open ground, Fletcher, Hardin Co., 25 April 1916, E. J. Palmer 9559 (MBG). One of the most distinct species of the genus in North America. It is readily recognized by the long slender graceful spike, the sessile elongate leaves, and the tardy splitting of the schizocarp. The last character is probably Walter's reason for placing this species in the genus Phryma. SECTION 2. GLANDULARIA Schauer 2. GLANDULARIA Schauer in DC. Prodr. 11: 550. 1847. Sterile style-lobe adjacent to stigmatic surface and protruding well beyond it; ovary in later stages definitely but shallowly lobed at the distal end, with the style apparently inserted in a depression between the lobes; schizocarp constricted along the lines of cleavage; connective of the upper anthers chiefly appen- daged. Herbaceous perennials with prostrate, decumbent, as- cending, and sometimes erect stems. Flowers usually doo: at first fascicled or somewhat corymbose, later spicate. Calyx usually more than twice as long as the schizocarp and constricted or contorted beyond it. Species 33-51. 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA dll KEY TO THE SPECIES A. Leaves tapering into a margined petiole or subsessile or sessile, at least not subauriculate and semiamplexicaul at the base. B. Nutlets with a very definite beak parallel to the axis of the schizocarp.. bee CEM noa Nen coe ce ie bios ve nS wo See 33. V. quadrangulata B. Nutlets with an indefinite beak or none. C. Nutlets suggesting a tendency toward a beak (a slight protrusion horizontal to the axis of the schizocarp); commissural face practi- cally reaching the tip of the nutlet................... 34. V. delticola C. Nutlets without a beak; commissural face not reaching the tip of the nutle D. Nutlets subovoid, lateral surfaces ventricose and smooth, not at all similar to the dorsal surface. ..................... 39. V. tumidula D. Nutlets Oe a E surfaces not ventricose and usually scrobiculate, similar e dorsal surfac E. Leaves shallowly Gee mi or oe F. Corolla-tube slightly longer than the calyx. G. Calyx villous-hirsute, somewhat glandular................ DUM NN Cr CAPE 47a. V. Gooddingii var. bise G. Calyx short-strigillose, not glandular........... 38. V. tampensis F. Corolla-tube at least one-half longer than the calyx. H. Plants prostrate-decumbent; leaves varying from cuneate to orbieular-ovate; calyx-teeth short (1-1.5 mm. long)...... 37. H: ies ascending, decumbent only at base; leaves ovate to gate-ovate; calyx-teeth long (2-3 mm.)....35. V. canadensis E. Leaves 3-dleft, inclsod-plumdd or bipinnatifid. I. Corolla-tube protruding well beyond the calyx. J. Corolla-tube approximately twice as long as the calyx. K. Plants somewhat hirsute; both surfaces of the leaves simi- lar in color; spikes elongating in age; calyx-teeth 2-3 mm. long. iusso EBEN MER d 0 S4 ex ns 85. V. canadensis K. Plants premio < lower surface of leaves lighter than the upper; spikes compact at maturity; calyx- teeth ie surpassing 2 mm. in len L. Plants prostrate-decumbent; spikes few-flowered..36. V. elegans L. edo Fine g decumbent only at base; spikes y-Bowered. ii. cenas en 36a. V. elegans var. asperata J. Cardia bibi kope to one-half longer than the cal M. Floral bracts equalling or exceeding the calyx; esiyx not glandular, hispid-hirsute N. Leaves bipinnatifid, ultimate segments linear-oblong.. vvv ENERO eor 40. V. bipinnatifida N. Leaves 3-cleft, with segments remotely incised or AE ultimate segments much broader than in the DTE ot 40a. V. Watana var. boil M. Floral braets shorter than the calyx; calyx for the most part somewhat glandular, hispid-hirsute to villous- pubescent. [Vor. 20 312 ANNALS OF THE MISSOURI BOTANICAL GARDEN O. Leaves sessile, strigose-hispid............. 46. V. Andrieuxit O. Leaves subsessile to short-petiolate, hispidulous-hirsute to hirtellous P. Plants coarse; leaves usually 2.5-4 em. long, hispid- ulous-hirsute; corolla-liimb 6-12 mm. broad; nutlets 2.5-3 mm. long Q. Calyx-teeth long (2-3 mm.). R. Plants usually tall (2-4 dm.); leaves bipinnatifid ith ultimate segments lanceolate; calyx-teeth not especially conspicuous in mature fruit. S. Calyx glandular................ 41. V. ambrosifolia Calyx not glandular.41a. V. ambrosifolia f. eglandulosa R. — usually low (1-2 dm.); leaves trifid with egments more or less incised; calyx-teeth mesas long in mature fruit........... UXWEREEREISATUKMERS 49a. V. ciliata var. longidentata Q. Calyx-teeth short (usually less than 2 mm.). T. Plants decumbent-ascending or prostrate densely hirsute; calyx somewhat glandular, hispidulous-hirsute, scarcely viscid U. Plants with decumbent-ascending loose habit; leaf-margin merid revolute........ 48. V. ciliata U. Plants with prostrate compact habit; le margin strongly sinus te. .43b. V. ciliata var. pubera T. Plants ascending-erect, more or less hirsut T aene plandular-hirsute and oi cid-pubescent.................0. 42. V. Wrightà P. Plants Neid esed smaller, 1-2.5 cm. long, "page sor tu 5-6 mm. broad; nutlets Aerated ads TAA EIE OER 44. V. racemosa I. Corolla-tube slightly ue than the c V. Calyx-teeth acute-subulate, short dg than half as long as the calyx-tube). W. Plants repent, slender; spikes few-flowered, scarcely pro- truding beyond the subtending leaves. X. Corolla-limb 5-7 mm. broad............ 46. V. teucriifolia X. Corolla-limb 3 mm. broad: 46a V. teucriifolia var. corollulata W. Plants decumbent-ascending, stouter; spikes many-flow- ered, subsessile or short-pedunculate. Y. Plants more or less hirsute, decumbent; corolla incon spicuous, limb 3-5 mm. broad............ 48. V. pue Y. T ke ap ig ascending; corolla con- , limb 8-9 mm. broad........ 47. V. Goodding& V. Clysleih m a elongate (approximately half s long as the calyx-tube Z. Plante decumbent-ascending, Kpaka leaves ovate, trifid with the segments coarsely dentate. 49. V. setacea Z. Plants erect, sparsely hirsute; leaves narrow ae oe _Mipianatisd, with the segments linear.......... 50. V. lilacina A. I I i at the ay er ees 51. V. amoena 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 313 3. V. quadrangulata Heller, Contr. Herb. Franklin & Heat College (Bot. Expl. S. Texas) 1: 84, pl. 6. 1895. V. pumila f. albiflora Standl. Field Mus. Publ. Bot. 4: 256. 1929 H ain PF. deg eae Small, Fl. Southeast. U. S. ed. 1, 1011. 1908, and ed. 2, 1 Low plant more or PH m branched from the base; stems prostrate-ascending, rooting at the lower nodes, hirsute; leaves 1-3 cm. long, broadly ovate, with cuneate base contracted into a very short margined petiole, incised-pinnatifid or 3-cleft, with lobes incised, strigose-hirsute on both surfaces; spikes compact, sessile or nearly so, terminal; bracts somewhat shorter than the calyx, narrowly lanceolate, hirsute, ciliate; fruiting calyx 5(-6) mm. long, hirsute, particularly along the ribs, lobes short, subulate-tipped; corolla-tube slightly longer than the calyx, practically glabrous without; corolla-limb 2.5-3 mm. broad, lobes emarginate; anthers unappendaged; ovary sur- mounted by subhemispheric-angulate stylopodium; schizocarp constricted along the lines of cleavage; nutlets 4 mm. long, crowned with a smooth obtusish beak, shallowly reticulate- scrobiculate above, longitudinally striate and somewhat abruptly enlarged at base. Distribution: Texas and northern Mexico. Specimens examined: Texas: sandy beach and open flats, Rio Bravo del Norte, Schott (NY); Devils River, Val Verde Co., 15 May 1913, Orcutt (MBG); creek between Del Rio and Comstock, 22 April 1925, Small & Wherry 12010 (NY); “Brackett” (? Brackettville), 22 Aug. 1900, Trelease 101 (MBG); Spofford Junction, 22 March 1900, Canby 193 (US); Uvalde, 28 April 1928, E. J. Palmer 33592 (G, MBG, NY); San Antonio, April 1922, Schulz 767 (US); 24 km. south of San Antonio, 28 April 1921, Schulz 475 (US); woods near Colorado River, near Wharton, 12 April 1925, Small & Wherry 11824 (NY); railroad north of Moore Station, Frio Co., 5 April 1901, Eggert (MBG); sandy open ground, Pleasanton, 16 May 1916, E. J. Palmer 1747 (MBG); El Jardin, 10 March 1924, Runyon 628 (US); Eagle Pass, April 1883, Havard (US); Kennedy- Beeville, 15 March 1929, Tharp 5538 (US); Kennedy-Portland, 17 March 1929, Tharp 5610 (US); Corpus Christi Bay, Dec. 1879, Palmer 1046 (G); Corpus Christi, 10 March 1894, Heller 1888 (G, MBG, NY), Type collection; San Diego, 1885-6, Croft 78 (NY); Kingsville, 27 March 1920, High 75 (MBG); between Laredo and Bejar, Feb. 1828, Berlandier 1485 =225 (G); Laredo, Feb. 1891, Dodge 154 (US); sandy ground, Laredo, 6 April 1901, Eggert (G, MBG); Laredo, 21 March 1903, Reverchon 3902 (G, MBG, NY, P, US); Laredo, 1913, Orcutt 5542, 5780 (MBG); San Antonio, 1926, Pagel 2208 (F); Guadalupe, 168 km. southwest of San Antonio, [Vor. 20 314 ANNALS OF THE MISSOURI BOTANICAL GARDEN Sept. 1879, Palmer 2039 (G); chaparral near cun ge 6 April 1925, Small & Wherry 11903 (NY); Brazos Santiago, 1889, Nealley (F, U Mexico: TaMAULIPAS: east of Matamoros, May on Berhondior 3018 —1518 (MBG). Verbena quadrangulata is an anomalous species often confused with V. pumila on account of the strong similarity in habit, but is perhaps more closely related to V. delticola, the only other known North American species with a tendency toward develop- ing a nutlet with a beak. The style is enlarged at the base into a persistent subhemispherical body; hence, when the nutlets split apart they appear as if elongated at the apex into a beak-like appendage. Although the anthers are not glandular and the flowers not showy, the species seems to belong to the section Glandularia. 34. V. delticola Small, n. sp. Stems decumbent to ascending, branched, more or less hirsute; leaves ovate-deltoid, 3-7 cm. long, with truncate-cuneate base narrowed into a margined petiole, obtusish or acutish, coarsely serrate-dentate, often trilobed, usually thin, sparsely appressed- hirsute on both surfaces, trichomes above often with minute bulbous bases; spikes peduncled, fascicle-like in anthesis, be- coming elongated in fruit; bracts linear-attenuate, shorter than or equalling the calyx; calyx 7-8 mm. long, sparsely glandular, hirsute, lobes slender, subulate, unequal; corolla-tube protruding well beyond the calyx, pubescent or glabrate without; corolla- limb probably 10 mm. broad, segments emarginate; nutlets 3 mm. long, subcylindrical, reticulate from the apex to the defi- nitely broadened base; commissural face practically reaching the tip of the nutlet, muricately scabrous. 2 V, delticola Small, spec. nov., herbacea verisimiliter perennis; caulibus decum- bentibus vel ascendentibus ramosis plus minusve hirsutis; foliis ovato-deltoideis basi cuneata in petiolum alatum attenuatis 3-7 cm. longis obtusiusculis vel acutis grosse serrato-dentatis saepe trilobis utrinque sparse adpresso-hirsutis; spicis pedunculatis; bracteis lineari-attenuatis calyce brevioribus vel subaequant tibus; calyce 7-8 mm. longo sparse glanduloso hirsuto; calycis dentibus subulatis inaequalibus; corollae tubo exserto extus pubescente vel glabrato; corollae limbo circiter 10 mm. lato segmentis emarginatis; coccis subcylindricis 3 mm. longis superne reticulatis.— Collected at Las Palmas Ranch, vicinity of Brownsville, Texas, 1-5 Aug. 1921 Ferris & Duncan 3161 (NY), TYPE. 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 315 Distribution: Texas and Mexico. Specimens examined: Texas: Edinburg, Hooker 5999 (US); Samfordyce, 1927, me 7224 (US); chaparral near Harlingen, 16 April 1925, Small & Wherry 11901 (NY); Brazos Santiago, 1889, Nealley 117, 118 (US); Reynoldsville, Cameron Co., 11 April 1905, Lewton 141 (US); Las Palmas Ranch, vicinity of Brownsville, 1-5 aay 1921, Ferris & conic 3161 (MBG, NY Type). MeExic NO Leon: Rinconada, 24 April 1847, Gregg 752 (MBG); Guajuco, March 1880, Palmer 1051 (G, US); Walnut Grove, near Monterey, 7 Feb. 1847, Gregg 202 (MBG); river gravel, near Monterey, 11 July 1888, Pringle 2228 (G); Sierra Madre, near Monterey, 25 Aug. 1903, Pringle 11843 (G, US); Loma del Obispodo, Monterey, 1 Feb. 1907, Safford 1221 (US); Monterey, Guadalupe, May 1911, Arsène 6129 (G, MBG, NY, US); Monterey, 10-11 March 1923, Tharp 1826 (US). TAMAULIPAS: Victoria, 23 May 1898, Nelson 4424 (G, US); vicinity of Victoria, 1 Feb.-9 April 1907, Palmer 39 (F, G, MBG, NY, US); San Fernando to Jimeney, 26-27 Feb. 1902, Nelson 6628 (? 6028) (G, MBG, NY, US); vicinity of Tampico, 1-31 Jan. 1910, Palmer 90 (F, G, MBG, NY, US San Luis Potost: La Hoya, Liebmann 11314 (US). PoEBLA: near Metlaltoyuca, 31 Jan. 1918, Goldman 42 (US). V Cruz: Wartenburg, near Tantoyuca, April 1858, Ervendberg 236 (G); Colipa, March 1841, Liebmann 11313 (US). Superficially, this species bears a strong resemblance to V. canadensis, but usually the leaves are not so deeply incised. Its distinctive character is found in the nutlet. Commonly, in the section Glandularia, the schizocarp is shallowly lobed at the apex; hence, the style appears to be attached in a very definite de- pression and ordinarily the commissural face does not reach the tip of the nutlet. In V. delticola, however, the depression is indefinite, the commissural face practically reaches the tip, and the separate nutlets viewed from the lateral or ventral surface suggest a tendency toward developing a beak. In Safford 1221 a small beak is present. ih V. canadensis (L.) Britton, Mem. Torr. Bot. Club 5: 276. Aiia Jacq. Hort. Vind. 2: 82, pl. 176. 1772; Linn. f. Suppl. 86. 1781; Bot. Mag. pl. 308. 1795; Bot. Reg. pl. 294. V: O TETA Retz. Svenska Vet. Akad. Nya Stockh. Handl. 34: 143, pl. 5. 1778. V. Obletia Medic. Act. Acad. Theod.-Palat. 3: 194, pl. 7. 1775. V. longiflora Lam. Tab. Encye. 1: 57. 9 [Vor. 20 316 ANNALS OF THE MISSOURI BOTANICAL GARDEN V. rubra Salisb., Prodr. 71. 1796. V. Lamberti Sims in Bot. Mag. pl. 2200. 1821. V. Aubletia var. Drummondi Lindl. Bot. Reg. pl. 1925. 1837. V. Lamberti var. rosea Sweet, Brit. Fl. Gard. II. 4: 363. 1838. V. Drummondii Hort. ex G. Don in Loud. Hort. Brit. Suppl. 2:080. 1839. V. Drummond (Lindl. Baxt. ex Small, Fl. Southeast. U. S. ed. 1, 1011. 1903, and ed. 2, 1913. V. canadensis var. Lamberti 'Thell. Fl. Advent. de Montpellier, 428. 1912. Buchnera canadensis L. Mant. 88. 1767. Billardiera explanata Moench, Method. 369. 1794. Anonymos caroliniensis Walt. Fl. Carol. 164. 1788. Glandularia caroliniensis J. F. Gmel. Syst. Veg. 22: 920. 1791. G. Aubletia Nutt. Trans. Am. Phil. Soc. 5: 184. 1837. Stems decumbent to ascending, rooting at the lower nodes, more or less branched, spreading-hirsute or glabrate; leaves ovate to elongate-ovate, 3-9 em. long, 1.5-4 em. broad, with truncate or cuneate base narrowed into a margined petiole, in- cised or incised-pinnatifid or 3-cleft, appressed-hirsute or glabrate on both surfaces; spikes pedunculate, fascicle-like in anthesis, becoming elongated in fruit; bracts shorter than (or occasionally as long as) the calyx, linear-attenuate, hirsute, usually ciliate; calyx glandular-hirsute, in fruit 10-13 mm. long, lobes very slender, subulate-setaceous, unequal, the posterior lobe much shorter; corolla-tube about twice as long as the calyx, glabrous without or finely pubescent or glandular; corolla-limb 11-15 mm. broad, segments emarginate; anther-glands not minute; schizo- carp at maturity constricted along the lines of cleavage; nutlets 3(-3.5) mm. long, subcylindrical with slightly broadened base, reticulate-scrobiculate; commissural face muricate-scabrous. Distribution: North Carolina to Florida, and west to Colorado and Texas. ed: NonTH CanoriNA: roadsides, Kittrell, 11 April 1889, Sturtevant (MBG). SourTH Carona: Abbeville District, Hexamer & Maier (G). GEoRGIA: by roads and margins of fields through lower Georgia and middle Florida, March 1843, Rugel (MBG). Fiorma: Fort King, Alden (NY); St. Nicholas, Duval Co., July 1898, Lighthipe 599 (NY); Mous Creek, St. John’s Co., 15 April 1879, J. D. Smith (US); rich woods in Daytona, 30 March 1906, Deam 1831 (US); Ocala, 4 April 1879, J. D. Smith (US). 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 817 ALABAMA: above Tuscaloosa, 8 May 1875, E. A. Smith (US); above Tuscaloosa, 11 April 1892, Ward (US). LovuisiANA: without locality, Hale (G, US), Short (NY); Minden, 15 April 1901, Trelease (MBG); Winnfield, 13 April 1912, Petersen (NY); Natchitoches, 16 March , E. J. Palmer 7001 (MBG); Chopin, Natchitoches Parish, 23 March 1915, E. J. Palmer 7071 (MBG); vicinity of Covington, 20 March 1920, Arsène 11982 (US); Jacksonville, Drummond (G); Cameron, 4 July 1903, Tracy 8707 (G, MBG, NY, US). Kentucky: Shelbyville, Flint (G). TENNESSEE: Cedar Glades, north of Lavergne, 4 May 1898, Eggert (MBG). Iowa: Council Bluffs, Vasey (G). Missourt: St. Louis, May 1833, Engelmann 334 (MBG); St. Louis, 8 July 1910, Sherff 333 (G); Windsor Springs, 30 May 1890, Hitchcock (MBG); St. Louis Co., 16 May 1879, Eggert (MBG, US); Allenton, 27 April 1887, Eggert (MBG); west of Kimmswick, 13 April 1918, Drushel 3659 (MBG); Kimmswick, 30 April 1905, Johnson (MBG); Hillsboro, 24 May 1885, F. Wislizenus 285 (MBG); Crystal City, 19 Aug. 1886, Eggert (MBG); sandy ground north of Crystal City, 6 May 1891, Eggert (MBG); Siliea, 18 April 1896, Eggert (MBG); Shot Hill, near Selma, 30 May 1923, Greenman 4231 (MBG); Victoria, 10 May 1890, Hitchcock (MBG); Old Mines, Washington Co., 19 Aug. 1928, Kellogg 1958 (MBG); Blackwell, St. Francois Co., 18 April 1897, T'release 714 (MBG); Central, 28 Aug. 1898, T'release 1161 (MBG); banks of Pilot Knob Creek, 9 Sept. 1859, Engelmann (MBG); Pilot Knob, July 1867, Engelmann (MBG); Granite Mountain, Iron Co., 24 May 1918, Greenman 4076 (MBG); Ironton, 4 May 1923, Epling 6134 (MBG); Shepard Mountain, Iron Co., 26 May 1918, Greenman 3870 (MBG); Shut In, Stout Creek, Iron Co., 28 May 1916, Drushel 2762 (MBG); near Silvermine, Madison Co., 20 May 1927, Greenman (MBG); Mine La Motte, 19 May 1927, Greenman (MBG); Grandin, 5 May 1901, Bush 318 (MBG); dry rocky banks, Shannon Co., 13 April 1889, Bush 1168 (MBG); Jerome, 5 April 1914, Kellogg (MBG); Jackson Co., 28 March 1864, Broadhead (MBG); Jackson Co., 28 May 1893, Bush 283 (G, MBG); Independence, 1 June 1895, Tindall (MBG); Dodson, 27 April 1904, Bush 1933 (MBG); rocky woods, Vale, 13 April 1908, Bush 4922 (MBG, US); barrens, Greenwood, 28 Oct. 1915, Bush 2901 (G, MBG, NY); Greenwood, 25 April 1911, Bush 6439 (MBG); Jasper Co., 16 Aug. 1893, Bush (MBG); Oronogo, 10 July 1910, E. J. Palmer 2996 (MBG); open banks, Webb City, 19 April 1903, E. J. Palmer 559 (MBG); Webb City, 4 May 1902, E. J. Palmer 304 (MBG); prairie, vicinity of Pearl, 22 Aug. 1912, Standley 9171 (US); Willard, 20 July 1919, Blankinship (P); Springfield, 31 July 1892, Dewart 74 (MBG); vicinity of Springfield, 29 Aug. 1911, Standley 8367 (US); barrens, Swan, 25 Sept. 1899, Bush 569 (MBG); Swan, 7-9 Oct. 1915, Eggles- ton 12240 (NY, US); Galena, Stone Co., 27 May 1914, E. J. Palmer 5682 (MBG); limestone slopes, ‘‘ Bald Joe," Stone Co., 30 April 1924, E. J. Palmer 24616 (MBG); near Seligman, Barry Co., 4 April 1926, E. J. Palmer 29804 (MBG). ARKANSAS: Eureka Springs, 19 April 1899, T'release (MBG); top of dolomite hill along White River, near Beaver, Carroll Co., 23 Oct. 1925, E. J. Palmer 29333 (G, MBG, NY); bluffs, Van Buren, 5 April 1929, Demaree 6411 (US); 14 km. west of Fort Smith, 1853-4, Bigelow (US); Fort Smith, 1 April 1928, Benke 4566 (G); Hot Springs, 31 Oct. 1899, T'release (MBG); Malvern, Letterman (MBG, US); Prescott, 8 April 1900, Bush 545 (MBG); clay barrens, Fulton, Hompetibd Co., 28 April 1914, E. J. Palmer 5407 (MBG); Red River, Pitcher Kansas: Princeton, 19 May 1919, Street (P); between Olathe and Pleasanton, Miami Co., 18 June 1929, Rydberg & Imler 22 (MBG, NY); open woods, Cherokee [Vor. 20 318 ANNALS OF THE MISSOURI BOTANICAL GARDEN Co., 1896, Hitchcock 790 (G, MBG, NY, US); vicinity of Edna, 28 June 1929, Ryd- berg & Imler 379 (NY). OKLAHOMA: near Miami, Ottawa Co., 26 Aug. 1913, Stevens 2840 (G, MBG, NY, US); Sapulpa, 22 July 1894, Bush 437 (G, MBG, NY, US); Perkins, 14 June 1893, Waugh 168 (MBG); onland woods, Muskogee, 30 Sone 1918, E. J. Palmer 14283 MBG); Norman, 17 April 1915, Emig 415 (MBG, US); near Page, 20 May 1914, Blakeley 1471 (G); near Paul's Valley, Garvin Co., 19 April 1913, Stevens 109 (G, MBG); Colbert Station, 19 June 1891, Sheldon 26 (US); near Idabel, 18 May 1916, Houghton 3643 (G, MBG). Texas: Denison, 22 April 1904, Reverchon (MBG); Denison, 15 March 1904, Reverchon (MBG); wet places, T. & C. Junction, Bowie Co., 5 Sept. 1900, Eggert G, MBG); sandy ground north of Longview, Gregg Co., 19 April 1899, Eggert (MBG); open ground, Longview, 21 April 1915, E. J. Palmer 7124 (MBG); sands, Dallas, Reverchon (G), 740 (MBG); Newland, near Dallas, 26 March 1901, Rever- chon 2533 (MBG, NY); Grapeland, Houston Co., 28 May 1917, E. J. Palmer 12056 (MBG); Grapeland, 8 June 1920, Tharp 838 (G, NY, US); Livingston, Polk Co., 10 April 1914, E. J. Palmer 5185 (MBG); Sabine River, 25 km. north of Orange, 18 April 1899, Bray 65 (US); Dayton, Liberty Co., 21 May 1917, E. J. Palmer 11979 (MBG); sandy open ground, near Conroe, Montgomery Co., 15 April 1929, E. J. Palmer 83319 (G, MBG, NY); Houston, Feb. 1842, Lindheimer (G); woods, Houston, 6 April 1872, Hall 436 (MBG, abi P, US); Houston, Fisher (MBG), 33, 40, 47, 707 (US); Galveston, Van Huff (M CoronaApo: Rocky neri pe 40-41’, 1868, Vasey (G, MBG). Verbena canadensis is one of the most easily recognized of the North American species belonging to the section Glandularia. It has a large corolla with tube about twice as long as the calyx and, at maturity, an elongated spike with somewhat remote truita and conspicuous calyx-teeth. The leaves are variously lobed or cleft. Although these variations have been used at sundry times to segregate species, they do not seem to be constant nor are they supported by definite geographic ranges, hence scarcely merit recognition. 36. V. elegans HBK. Nov. Gen. et Sp. 2: 273. 1818. V. moranensis Willd. ex Spreng. Syst. 2:750. 1825. V. canadensis subsp. elegans Thell. Fl. Advent. de Montpellier, 428. 1912 V. canadensis var. Ehrenbergii Thell. l. c. Stems prostrate-decumbent, tending to root at the older nodes, more or less branched, sparsely hispidulous-hirsute; leaves lanceolate-ovate with cuneate base narrowed into a margined petiole, 2-5 cm. long, incised-pinnatifid or trifid with the divisions less deeply cleft, somewhat appressed-hirsute on both surfaces, 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 319 dark green above, lighter beneath, often with branches or fascicles of smaller leaves in the axils; spikes few-flowered, pedunculate, terminal, fascicle-like; bracts mostly one-half to two-thirds as long as the calyx, lanceolate, attenuate, glandular-hirsute, ciliate; calyx 8-10 mm. long, glandular-hirsute, teeth short, subulate and unequal; corolla-tube 15-18 mm. long, finely but sparsely pubescent without, throat gibbous; corolla-limb 10-12 mm. wide; nutlets subcylindrical, 3.5-4 mm. long, reticulate- scrobieulate chiefly on the upper half, lower part striate; com- missural face muriculate. Distribution: vocum Mexico. Specimens examined Mexico: Hipauco: Sierra de Pachuca, 2700 m. alt., 22 July 1898, Pringle 6908 (ANSP, F, G, MBG, NY, P, US); Sierra de Pachuca, 21 and 22 July 1901, Rose & Hay 6556 (US); mountains, Pachuca, July 1905, Purpus 1433 (MBG, P); between Pachuca and Real del Monte, 19 July 1905, Rose, Painter & Rose 8668 (NY, US); Real del Monte, 9 May 1910, Clokey (MBG); Moran, Humboldt & Bonpland 4063 (Bot. Mus. Berl.-Dahl. type of V. moranensis, MBG phot.); under oaks and firs, Sierra de Ajusco, ca. 2800 m. alt., 26 Aug. 1902, Pringle 11092 (F, G, MBG, NY, US) Mexico: lomas de Santa Fe, Aug. 1927, Lyonnet 178 (US). Vera Cruz: near Chila, Distr. Ozuluama, April 1888, Seler 723 (G). Oaxaca: Tehuantepec, 1906, Gandoger (MBG). The above species is related to V. canadensis but is easily separable by its slender more or less prostrate habit, few-flowered compact spikes, and shorter calyx-teeth. The nutlets of the two are about the same size, but the reticulations in V. elegans are a little coarser than those in the more northern species. 36a. Var. asperata Perry,” n. var. Stems decumbent-ascending, hirsute-hispid; leaves variously cleft or lobed; spikes dense, many-flowered; calyx glandular, hispidulous-hirsute, teeth a little longer than in the species. Distribution: Sonora ane Chihuahua to San Luis Potosi. DECORUS examine XICO Sonora: ice Madre, 15 Dec. 1890, Lumholtz 445 (G); Badehuachi, 2 Dec. 1890, Ya 446 (G); Hermosillo, 1888, M. A. Crawford (G). 23 Var. asperata Perry, var. nov., caules ascendentes hirsuto-hispidi; foliis bi- pinnatifidis vel multifidis; spicis densis multifloris; calyce glanduloso hispidulo- hirsuto.—Colleeted at San Antonio, Coahuila, Mexico, 31 Aug. 1848, Gregg 356 (MBG), TYPE. [Vor. 20 320 ANNALS OF THE MISSOURI BOTANICAL GARDEN Curnuanva: southwestern Chihuahua, Aug.-Nov. 1885, Palmer 295 (G, US). DunaANco: San Ramon, 21 Apr.-18 May 1906, Palmer 191 (G, MBG, NY, US). Coanuita: Sierra de Parras, Oct. 1910, Purpus 4974 (F, G, MBG, US); 72 km. east of Saltillo, July 1880, Palmer 1052 (G, US); San Antonio, 31 Aug. 1848, Gregg 3856 (MBG type); Sierra Encarnaciore, 28 July 1896, Nelson 3896 (US). San Luis Porosi: near Morales, 1876, Schaffner 716 (G); near San Luis Potosi, 1878, Parry & Palmer 720 (ANSP, F, MBG, US); Alvarez, 5-12 Sept. 1902, Palmer 51 (F, G, MBG, NY, US). This variety, as compared with the species, is stouter and more erect with coarser pubescence and larger spike. Palmer differs from the other collections in having very shallowly lobed ovate leaves. Parry & Palmer ?20 shows much variation in pubescence, some plants being scabrous, others scarcely so at all. Purpus 4974 is lacking the coarse pubescence but appears more closely allied here than elsewhere. 37. V. maritima Small, Bull. N. Y. Bot. Gard. 3: 436. 1905. Stems branched at base, decumbent or prostrate, sparingly pubescent or glabrate; leaves cuneate to orbicular-ovate, 1-4(-6) em. long, tapering into a margined petiole, incised-dentate or somewhat lobed, sparsely pubescent or glabrate on both surfaces; spikes terminal, pedunculate, fascicle-like in anthesis, becoming elongate in fruit; bracts linear-lanceolate, about one-half as long as the calyx, acuminate, pubescent, ciliate; fruiting calyx 10-13 mm. long, appressed-pubescent, often glandular, teeth short, slender, subulate; corolla-tube at least one-half longer than the calyx, pubescent without; corolla-limb 10-15 mm. broad; anthers with or without glands; nutlets subcylindric with broadened base, 4 mm. long, scrobiculate; commissure narrow, muricately scabrous. Distribution: Florida. Specimens examined: FLoripa: sandy ridges bordering the ocean, eastern Florida, Curtiss 1963 in part (G, MBG, NY); sandy soil, Merritt’s Island, 9 Dec. 1929, Moldenke 219a (NY); near Cape Canaveral, 15 July 1896, Curtiss 5706 (G, MBG, NY, P, US); Cape Canaveral, 2-5 April 1904, Burgess 638 (NY); Fort Pierce, 8-9 April 1904, Burgess 713 (NY); sand dunes, Hobe Sound, 19 March 1921, Randolph 52 (G); West Jupiter, 10 April 1904, Burgess 783 (NY); near beach, Palm Beach, 19 Nov. 1914, Small 2124 (NY); dry hammock south of Palm Beach, 1 May 1918, Small 8509 (NY); beach opposite Miami, Nov. 1904, Small 2100 (NY); opposite Miami, Feb. 1911, Small, Carter & Small 3311 (NY); Cape Florida, 29 March 1904, Britton 296 (NY) ; pine lands, Small Island, northwest of Perrine, 16 Jan. 1909, Small & Carter 2994 (NY); 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 321 sandy soil, Redlands, 2 Feb. 1930, Moldenke 549 (MBG, NY); sandy soil, Golden Beach, 10 Feb. 1930, Moldenke 586 (MBG, NY); Camp Jackson, 25 March 1904, Britton 220 (F, NY); Everglades, west of Camp Jackson, Dade Co., 6-9 May 1904, Small & Wilson 1961 (NY); Royal Palm Hammock, 20 Feb. 1915, Small & Small 5422 (NY); Paradise Key and vicinity, 21-29 Sept. 1917, Safford & Mosier 210 (US); Hammer Key, Everglades, 12 May 1918, Small 8594, 8599 (NY); in sand near Cocoa Beach, 16 March 1930, O'Neill 6309 (MBG, US); between Cutler and Long- view Camp, 9-12 Nov. 1903, Small & Carter 1077 (NY). This native of the sand dunes and the hammocks of Florida resembles V. canadensis in inflorescence, but is readily distin- guished by the creeping habit and the cuneate or orbicular-ovate leaves. 38. V. tampensis Nash, Bull. Torr. Bot. Club 23: 104. 1896. Stems erect from a decumbent or creeping base, pubescent or glabrate; leaves ovate to oblong-ovate, 2-7 cm. long, 1.5-4 cm. broad, truncate at base, tapering into a margined petiole (2 or more cm. long), coarsely serrate-dentate or shallowly incised, strigillose or glabrate on both surfaces, paler beneath; spikes pedunculate, fascicle-like in anthesis, becoming elongated in fruit; bracts linear-subulate, about one-half as long as the calyx, strigillose, ciliate; fruiting calyx 12-13 mm. long, strigillose, lobes very slender, subulate, unequal; corolla-tube somewhat longer (about 2-4 mm.) than the calyx, pubescent without; corolla-limb 10-12 mm. broad; anthers glandless or with a minute gland on the connective of each of the upper pair; schizocarp at maturity slightly constricted along the lines of cleavage; nutlets subcylindrical with broadened base, 4 mm. long, reticulate- scrobiculate above, striate toward base; commissure minutely scabrous. Distribution: Florida. Specimens examined: FLoriwa: without locality, 1842-49, Rugel 305 (F, MBG, NY, US); Daytona, 24 Jan. 1907, Mell (MBG); hammock south of Daytona, 8 May 1918, Small 8686 (NY); pine lands east of Eustis, 7 May 1918, Small 8469 (NY); Titusville, 31 March 1914, Mattern (US); dry sandy soil along roadside, 3 km. east of Fort Christmas, Orange Co., 9 Dec. 1929, Moldenke 212 (MBG, NY); hammock along Indian River, Cocoa, 9 May 1918, Small 8732 (NY); low open woods, Indian River, June, Curtiss 1963 in part (G, MBG, NY, US); Okeechobee region, Brevard Co., 18 April 1903, Fredholm 6804 (G); above Fort Lauderdale, 12 March 1920, J. P. Young 737 (US); Indian River at Biscayne Bay, 1874, Palmer 6433 (MBG); Tampa, May 1876, Garber (US); Tampa, 24 Aug. 1895, Nash 2470 (G, NY), TYPE collection; Tampa, 7 March 1898, [Vor. 20 322 ANNALS OF THE MISSOURI BOTANICAL GARDEN Pollard (US); St. Petersburg, April 1921, Beckwith 772 (US); Palmetto, 8 May 1900, Tracy 6650 (G, MBG, NY, US); Braden River, 13 km. southeast of Manatee, 13 June 1918, Barrett 13 (US); dry sandy soil north of Venice, 23 April 1930, Moldenke 1039 (MBG, NY); Owanita, Lee Co., 18 March 1907, W. Kellogg (G); thicket along river, Pondilla, Lee Co., 8 March 1927, Standley (US). Verbena tampensis, apparently endemic in peninsular Florida, is very similar in habit to V. canadensis. It differs in having more shallowly incised or dentate leaves, a strigillose calyx, slightly longer nutlets, and essentially glandless anthers. The nutlets are enlarged at the base somewhat more than those of closely related species. 39. V. tumidula Perry,” n. sp. Stems 15-20 cm. long, decumbent-ascending, branching, pilose-hirsute; leaves broadly ovate with cuneate base narrowed into a margined petiole, 2-4 cm. long, trifid with segments incised or coarsely crenate-dentate, appressed-pubescent or strigillose above, hirtellous beneath; spikes short-pedunculate, protruding slightly beyond the uppermost leaves, compact; bracts ovate- lanceolate, acuminate, not exceeding the calyx in length, ciliate; mature calyx 8-9 mm. long, inflated at base, hirsute, finely glandular, teeth subulate, 1.5 mm. long; corolla-tube about 11 mm. long, puberulent without; corolla-limb 8-10 mm. broad; anthers not glandular; nutlets 3 mm. long, dorsal surface retic- ulate-scrobiculate, lateral surfaces ventricose and smooth; commissural face almost smooth. Distribution: Texas and New Mexico. — ger examine “from W. Texas to El Paso, New Mexico," May-Oct. 1849, Wright 456 m US); Barkedale, 7 May 1918, E. J. Palmer 18512 (MBG TYPE). New Mexico: collection of 1851-2, Wright 1503 in part (G, NY). V. tumidula Perry, spec. nov., herbacea; caulibus A vel ascen- dentibus ramosis piloso-hirsutis; foliis late ovatis basi cuneata in petiolum alatum contractis trifidis segmentis inciso- vel grosse crenato-dentatis Sei hirtellis supra adpresso-pubescentibus vel strigosis; spicis breviter pedunculatis paulo exsertis; bracteis ovato-lanceolatis acuminatis ciliatis calyce fere paulo brevioribus; calyce m. longo basim inflato nd et iini glanduloso; calycis dentibus 1.5 mm. longis; corollae tubo circiter 11 mm. longo extus puberulo; corollae limbo 8-10 m. lato; eonneetivo antherarum SUpAHanm inappendiculato; coccis subovoideis 3 mm. longis dorso reticulato-scrobiculatis lateribus ventricosis et glabris; diae subtiliter muriculata. Collected at Barksdale, Texas, 7 May 1918, E. J. Palmer 18512 (MBG), TYPE. 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 323 This species undoubtedly belongs to the section Glandularia, although apparently the anthers are glandless. Superficially, it resembles V. Gooddingii var. nepetifolia, but the spikes are not so large nor so showy. Its distinctive characters are the ventri- cose nutlets and the inflated fruiting calyx. The bracts, too, are broader and shorter than those of nearly related species of this group. 40. V. bipinnatifida Nutt. Jour. Acad. Nat. Sci. Phila. 2: 123. 1821; Schauer in DC. Prodr. 11: 553. 1847. Glandularia bipinnatifida Nutt. Trans. Am. Phil. Soc. N.S. 5:184. 1837. More or less diffusely branched from the base; stems loosely ascending, occasionally rooting at the lower nodes or from sub- terranean branches, hispid-hirsute; leaves petiolate, blades 2-6 cm. long, bipinnately parted or tripartite with divisions more or less bipinnatifid, lobes linear or oblong, appressed-hirsute on both surfaces, margin at times revolute; spikes pedunculate, fascicle-like in anthesis, becoming elongate in fruit; bracts mostly longer than the calyx, linear-subulate, hispid-hirsute, ciliate; fruiting calyx 8.5-10 mm. long, pubescent, hispid-hirsute along the nerves, lobes very slender, subulate-setaceous from a broader base, unequal; corolla-tube about one-half longer than the calyx, pubescent without; corolla-limb 8-10(-12) mm. broad, lobes emarginate; nutlets cylindrie with slightly broadened base, mostly 3 mm. long, reticulate-scrobiculate above, striate toward the base; commissure muricately scabrous. Distribution: South Dakota to Alabama, westward to Arizona and northern exico Specimens examined: a : between Cahaba and Beloit, Dallas Co., 26 April 1927, Harper 15 (G, N US): on limestone outcrops along small creeks, near Demopolis, Marengo Co., | May 1925, E. J. Palmer 27209 (MBG); Spring Hill, 1918, Graves 536, 1946 (MBG). Mississippi: Ocean Springs, 1892, Skehan 47 (G). UISIANA: vicinity of Alexandria, 19 May 1899, Ball 401 MBG, NY, US); Cotes S xS St. Mary Co., 17 July 1893, Langlois (MBG, US). Miss : prairies, Upper Misscurt 21 June 1839, Geyer (US); Courtney, 6 June 1894, Bush p^ (G, MBG); Courtney, 13 June 1906, Bush 4029 (G, MBG, NY, US). ARKANSAS: “Red River,” Nuttall (G, NY), TYPE collection; dry gravelly hills, Fulton, Hempstead Co., 17 June 1915, E. J. Palmer 8045 (MBG); near Homan, 10 [Vor. 20 324 ANNALS OF THE MISSOURI BOTANICAL GARDEN June 1898, Eggert (MBG); along railroads, Miller Co., 19 June 1908, Eggert (NY); Texarkana, Aug. 1880, Letterman (MBG, NY). Soutn Daxora: Fort Pierre, Nicollet’s Northwest Expedition (US); Pierre, 2 Sept. 1891, T. A. Williams (MBG, NY); near McClure, June 1910, O. E. White (MBG); Kennebec, Lyman Co., 1 July 1914, Over 3177 (US); Pine Ridge Reservation, Washabaugh Co., 22 Aug. 1911, Visher 21382 (F, NY); White River, 16 July 1896, T. A. Williams (NY); White River flood plain, 4 Aug. 1914, Over 2103 (US); White River Valley, Shannon Co., 8 July 1911, Visher 2177 (NY); Fall River Falls, 18 June 1892, Rydberg 985 (G, NY, US). NEBRASKA: Spencer, Boyd Co., 25 July 1893, Clements 2776 (G, US); Callaway, 4 July 1901, Bates (G). Kansas: Cloud Co., 19 May 1888, Carleton (MBG); near Osborne City, 19 May 1894, Shear 38 (G, NY, US); Manhattan, 24 May 1901, A. Nelson 8245 (G, MBG, NY, US); plains, Ellis Co., 16 July 1895, Hitchcock 893 (G, MBG, NY, US); Smoky Hill, June 1867, Parry 15? (G, MBG); vicinity of Hays, 20 July 1929, Rydberg & Inler 1212 (NY), 1253 (MBG, NY); prairies, Medicine Lodge, 12 Sept. 1890, Smyth 306 (US); Morton Co., July 1891, Carleton 177 (US); vicinity of Richfield, 20 Sept. 1912, Rose & Fitch 17105 (US). OkLAnoMa: Vinita, 3 Aug. 1877, Gurney & Monell (MBG); Grant Co., 20 June 1899, White 19? (MBG); Woods Co., 29 June ?1899, White 165 (MBG); near Alva, Woods Co., 28 Sept. 1913, Stevens 2860 (G); near Whitehorse, 31 May 1913, Stevens 699 (G, MBG, NY, US); about 17 km. north of Boise City, Cimarron Co., 22 Aug. 1927, Stratton 449 (MBG); near Shattuck, Ellis Co., 20 May 1914, Clifton 3085 (G, P); Hollis, Harmon Co., 21 June 1913, Stevens 1108 (G); 10 km. southwest of Hollis, 6 Aug. 1927, Stratton 817 (MBG); vicinity of Fort Sill, 27 April 1916, Clemens 11751 (MBG); on false Washita between Fort Cobb and Fort Arbuckle, 1868, Palmer 4? (NY); Arbuckle Mountains, Davis, 29 Oct. 1914, Emig $95 (MBG); near Davis, 23 June 1917, Emig 788 (MBG); Mannsville, Johnston Co., April 1916, Griffith 3456 (G); Price’s Falls, Murray Co., 30 April 1926, Stratton 10 (MBG); near Crusher Spur, Murray Co., 11 April 1913, Stevens 14 (G, MBG, NY, US); near Camp, Texas Co., 12 May 1913, Stevens 427 (G); Caddo, 22 June 1891, Sheldon (MBG), 48 (US). Texas: without data, Lindheimer 282 (MBG), 289 (F, G, MBG, NY, US), 307 (G); Canadian, Hemphill Co., 10 Aug. 1900, Eggert (MBG); Canadian, July 1903, Howell 111 (US); vicinity of Terrell, Kaufman Co., 4 May 1904, T'yler (US); Dallas, Reverchon ?89 (MBG, US), 1962 (F, G, MBG, NY, US); Dallas, Reverchon (G, NY, US); Dallas, 22 June 1899, Eggert (MBG); dry ground, near Garland, 24 June 1899, Eggert (MBG); Fort Worth, Ruth 107 (F, G, MBG, NY, US); Weatherford, 26 May 1902, Tracy 7999 (F, G, MBG, NY, US); Corsicana, Navarro Co., 1 June 1915, E. J. Palmer ?821 (MBG); near Granbury, Hood Co., 4 May 1900, Eggert (MBG); Granbury, 4 Sept. 1914, E. J. Palmer 6510 (MBG); near Abilene, 7 June 1900, Eggert (MBG); Abilene, 19 May 1902, T'racy 8000 (F, MBG); Killeen, 1 Oct. 1891, Ward (US); Walker Co., 6-12 May 1910, Dixon 561 (F, P); near Austin, Travis Co., 18 April 1929, E. J. Palmer 33389 (G, MBG); Crab Apple Creek, Gillespie Co., Jermy 183 (MBG); about 25 km. west of San Felipe, March 1844, Lindheimer 146 (MBG); San Marcos, 11 June 1897, Trelease (MBG); Comanche Spring, New Braunfels, Lindheimer 1072, 1073 (F, G, MBG, NY, US); northwest of New Braun- fels, 14 Sept. 1913, Pennell 5444 (NY); limestone soil mixed with sand, near Bracken, Groth 80 (F, G, NY, US); Rock Springs, 17 April 1930, Jones 26228 in part (MBG); Del Rio, 18 April 1930, Jones 26229 in part (P); Fort Clark, Kinney Co., Mearns 1252, 1274, 1894 (US); Eagle Pass, Rio Bravo del Norte, Feb.-March 1852, Schott 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 325 (NY); San Antonio de Bexar, Feb. 1828, Berlandier 1449189 (G); between Trinity River and Bexar, June 1828, Berlandier 429 —?1749 (G); San Antonio, 4 April 1901, Eggert (G); San Antonio, 1897, Wilkinson 9 (MBG); San Antonio, 15 Aug. 1906, Ball 909 (US); vicinity of San Antonio, April 1919, von Schrenk (MBG); San Antonio to Austin, 18 April 1925, Small & Wherry 11968 (NY); 16 km. west of San Antonio, 8 June 1931, Moore & Steyermark 8001 (MBG); along Trinity River, near Liberty, 11 April 1925, Small & Wherry 11781 (NY); Columbia, 27 March 1900, Bush 449 (MBG, NY, US); Columbia, 3 May 1900, Bush 205 (MBG, US); Cuero, 22 March 1907, Howell 308 (US); prairies, near Victoria, 10 April 1900, Eggert (MBG); Victoria, 27 April 1905, Mazon $808 (US); Victoria, 7 March 1916, E. "i Palmer 9098 (MBG); from Goliad to Bexar, May 1834, Berlandier 2428 2998 (G, MBG). ?MoNTANA: sandy soil along "Vellawatons and upper Missouri, 1853-4, Hayden (MBG). New Mexico: without data, 1850-2, Wright 1502 (G, NY, US), 1604 in part (ANSP); below Dofiana, Parry, Bigelow, Wright, Schott (NY); Santa Fe, 20 July 1898, Earle 84 (MBG, NY); Roswell, Aug. 1900, Earle & Earle 526 (MBG, NY). Arizona: Bill Williams Mountain, 22 July 1898, MacDougal 317 (ANSP, F, G, NY); by streams of Santa Catalina Mountains, 13 April 1881, Pringle (G); Santa Rita Mountains, 31 July 1927, Peebles, Harrison & Kearney 4549 (US); Miller’s Canyon, Huachuca Mountains, July 1909, Goodding 251, 495 (G, NY); road to Rustler’s Park, Chiricahua Mountains, 18-19 June 1930, Goodman & Hitchcock 1211 (MBG); Barfoot Park, Chiricahua Mountains, 31 Aug. 1906, Blumer 1345 (F, G, MBG, NY). Mexico: Cuinvanua: near Colonia Garcia, 21 July 1899, Townsend & Barber 139 (F, G, MBG, P, US); near Colonia Garcia, 1899, Nelson 6117 (G, US); Round Valley, Sierra Madre, 17 Sept. 1903, Jones (P). In the specimens from Arizona and Mexico, it should be noted that the pubescence is hispidulous and the spikes at maturity are, for the most part, dense rather than elongated. The typical form of the species is very easily recognized by the hispid-hirsute pubescence, particularly of the inflorescence, the lack of glands, the long floral bracts, the elongating spike, the deeply cut leaves, and a rather coarse habit. Apparently it intergrades with V. ciliata in northern Mexico, with V. ambrosi- folia in New Mexico, and at times with V. canadensis in the south- western part of its range. Nevertheless, in spite of the many intermediate phases, it seems preferable to maintain each group as specific entities. 40a. Var. latilobata Perry,” n. var. Leaves trifid with the segments remotely incised or lobed, * Var. latilobata Perry, var. nov., folia trifida; segmentis grosse incisis lobisve supra strigosis subtusque praesertim in nervis et venis hirtellis hirsutisve; spicis maturitate densis vel elongatis.—Collected between San Pedro and Fronteras, Sonora, Mexico, 20-24 Sept. 1890, Hartman 906 (G), TYPE. [Vor. 20 326 ANNALS OF THE MISSOURI BOTANICAL GARDEN strigillose above, hirtellous-hirsute, especially along the veins beneath; spike compact or somewhat elongated in age; bracts variable in length, usually equalling or exceeding the calyx; floral and nutlet characters as in the species. Distribution: southern Arizona and northwestern Mexico. Specimens examined: ONA: Galiuro Mountains, 9 July 1894, — (NY); Bisbee, Oct. 1908, Qoodding 37 (NY); Flagstaff, 17 June 1887, Mearns (NY); Fort Huachuca, Aug. 1893, Wilcox (NY); Fort Huachuca, Sept. 1894, Wilcox (US); Huachuca Mountains, 29 June-5 July 1903, Griffiths 4833 (US); Huachuca Mountains, 3 Sept. 1903, Jones (P); Miller's Canyon, Huachuca Mountains, 26 Aug. 1910, Goodding 883 (G, NY); Santa Rita Mountains, 20 Sept.-4 Oct. 1902, Griffiths & Thornber 159 (US); Santa Rita Mountains, 24 Aug. 1903, Jones (P, US). Mexico: CHIHUAHUA: at base of Mt. Mohinora, about 13 km. from Guadalupe y Calvo, 23-31 Aug. 1898, Nelson 4856 (G, US); Colonia Juarez, Sierra Madre Mountains, 11 Sept. 1903, Jones (P); San Diego Canyon, Sierra Madre Mountains, 16 Sept. 1903, Jones (P); without data, Wright 1503 in part (G, NY). Sonora: between San Pedro and Fronteras, 20-24 Sept. 1890, Hartman 906 (G TYPE); San Pedro, 14 Sept. 1890, Hartman 880 (G); Cananea, 27 Sept. 1908, Donnelly (P); Cananea, 20 Aug.-1 Sept. 1914, Murdoch (F). DunaNao: without data, Garcia 331 (US); Sierra Madre, west of Durango, Sept.— Oct. 1881, Forrer (F). The variety differs from the species chiefly in foliar characters. 41. V. ambrosifolia Rydb. in Small, Fl. Southeast. U. S. ed. 1, 1011. 1337. 1908, and ed. 2, 1913. More or less diffusely branched from the base; stems loosely decumbent-ascending, somewhat hirsute; leaves short-petiolate, blades 2-6 em. long, bipinnatifid with the ultimate segments lanceolate, appressed-hirsute on both surfaces, margin slightly revolute; spikes pedunculate or subsessile, fascicle-like in anthesis, somewhat elongating in fruit; bracts a little shorter than the calyx, lance-subulate, hirsute, ciliate; fruiting calyx 8-9 mm. long, glandular, pubescent, hirsute particularly along the nerves; teeth subulate; corolla-tube protruding one-third to one-half beyond the calyx, pubescent without; corolla-limb 6—8(-10) mm. wide; nutlets subeylindrie often with slightly broadened base, 2.5-3 mm. long, reticulate-scrobiculate above, striate below; commissure muricately scabrous. Distribution: Oklahoma and Texas to Arizona and northern Mexico. ecimens examine OKLAHOMA: near onion, Cimarron Co., 15 May 1913, Stevens 484 (G, MBG). 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 327 Texas: 16 km. east of Comstock, 9 June 1931, Moore & Steyermark 3009 (MBG); Alpine, 30 May 1928, E. J. Palmer 34240 (NY); prairies near Marfa, 14 May 1901, Eggert (MBG); foothills below McKittrick Canyon, 23 July 1931, Moore & Steyer- mark 3622 (MBG); Sierra Blanea, May 1913, Orcutt 6096, 619? (MBG); northeast base of Quitman Mountains, near Sierra Blanca, 4 July 1921, Ferris & Duncan 2474 (MBG). oLonADO: New Windsor, 18 June 1905, Osterhout 3165 (G); Boulder, July 1891, Penard 348 (NY); Denver, July 1892, Eastwood (NY); South Denver, 8 June 1891, E. C. Smith (MBG); foothills, below Colorado Springs, 29 May 1878, Jones 122 (P); Cafion City, May 1871, Brandegee (MBG); Pueblo, 10 July 1873, Greene 45 (G); road to Walsenburg, Pueblo, 13 June 1917, E. L. Johnston 986 (MBG); mesas near Pueblo, 14 May 1900, Rydberg & Vreeland 5676 (NY); Rocky Ford, Otero, 8 June 1900, Osterhout (G fragm., NY Type); vicinity of La Junta, 26 Sept. 1913, Rose & Fitch 17504 (MBG, US). New Mexico: collection of 1847, Fendler = in part (ANSP, F, G, MBG, US); south of Raton on the road to Taos, 30 June 1929, Mathias 545 (MBG); Tierra Amarilla, Rio Arriba Co., 18 April-25 May 1911. Eggleston 6644 (US); Ojo Caliente, ear Fairview, 18 July 1904, Wooton (US); in canyon near Santa Fe, 29 June 1846, Wislizenus 525 (MBG); about Santa Fe, 8-9 Sept. 1881, Engelmann (MBG); near Pecos, San Miguel Co., 15 Aug. 1908, Standley 4951 (F, G, MBG, NY, US); Pecos, 18 June 1927, Arsène 18615 (P); vicinity of Las Vegas, Anect 100, 143 (G), 179 (NY), Arsène 18793 (P); Nara Visa, 21 April 1911, Fisher (MBG); vicinity of Bernalillo, 12 July 1926, Arsène & Benedict 16306 (MBG); vicinity of Albuquerque, 6 Oct 1913, Rose & Fitch 17801 (MBG, US); near Albuquerque, 21 June 1926, E. J. Palmer 31192 (MBG); between Carrizozo and Socorro, 21 July 1928, Wolf 2890 (P); Socorro, May 1881, Vasey (F, MBG, US); White Mountains, Lincoln Co., Aug. 1897, Wooton 368 (G, MBG, NY, P); Capitan, 8-19 May 1902, Earle 688 (NY); Lincoln, 31 July 1900, Earle & Earle 286 (MBG, P); South Spring, 2-4 May 1903, Griffiths 4244 (US); Mimbres River, Grant Co., 20 Aug. 1904, Metcalfe 1231 (MBG); plains south of White Sands, Dofia Ana Co., 28 Aug. 1897, Wooton 642 (MBG, NY, US); Queen, 2 Aug. 1909, Wooton (MBG, US); Organ Mountains, Dofia Ana Co., 16 Aug. 1893, Wooton (MBG rentem Warsaw Mill, Pima Eo. 3 Dec. 1893, Mearns 2677 (MBG, NY, US); about 3 km. south of Holbrook, 17 June 1901, Ward (NY, US); Little Colorado, 1869, nisi (US). Mex See Candelaria Mountains, 1911, Stearns 104 (US). Nurvo Leon: Monterey, 17-26 Feb. 1880, Palmer 1050 (MBG, US). CoanvuILA: about 9 km. east of Saltillo, April 1880, Palmer 1050 (G); Sierra Mojada, 20 April 1892, Jones 634 (P); Sabinas, 21 May 1902, Nelson 6789 (NY, US); Carneros Station, 31 May 1890, Pringle 3467 (F); fields near Carneros Station, 10 Nov. 1904, Pringle 13157 (F, G, US An examination of the type specimens of V. ambrosifolia and V. Wrightii reveals differences probably sufficient to suggest distinct but closely related species. Typically the former is characterized by long calyx-teeth and coarser pubescence; whereas the latter has short calyx-teeth, finer and somewhat viscid pubescence. On the other hand, a study of the aggregate [Vor. 20 328 ANNALS OF THE MISSOURI BOTANICAL GARDEN specimens of the two shows a high variability in the combination of these characters; moreover, their geographic ranges practically coincide. For (hose reasons, V. ambrosifolia is maintained as a species with very doubtful status. 41a. Forma eglandulosa Perry,? new form. Differing from the species only in the more hirsute-hispid pubescence of the flower and the lack of glands. Distribution: Texas to Arizona, and Sonora, Mexico. Specimens examined: Texas: “El Paso to Monument no 53," Sept. 1892, Wagner 974 (US). New Mexico: Albuquerque, 21 June 1926, E. J. Palmer 31158 (MBG); 8 km. west of Magdalena, 20 July 1904, Wooton 2835 (US); Organ Mountains, 14 July 1897, Wooton (US); mesa just west of Organ Mountains, 13 June 1906, Standley (US); Santa Rita, 1 Aug. 1911, Holzinger (MBG rypx, US); Deming, 13 July 1917, Munz 1231 (P); Los Playas, 1854-5, Antisell 186 (NY); near White Water, 11 Sept. 1893, Mearns 2278 (US); Animas Valley, 2 Oct. 1893, Mearns 2486 (US). Arizona: Douglas, 13 May 1915, Carlson (G, US); Santa Catalina Mountains, 15 Sept. 1896, T'oumey (NY, US); Dragoon, 18 July 1920, W. W. Jones 185 (G); Cochise, 12 Oct. 1900, Griffiths 1919 (NY). Mexico: Sonora: San Pedro River, 15 Oct. 1892, Mearns 1138 (US). This is a phase of the species lacking glands and only separable from V. bipinnatifida by the short bracts. 42. V. Wrightii Gray, Syn. Fl. N. Am. 2!:: 337. 1878. Stems usually several from a common base, decumbent- ascending to erect, branched, sparsely hispid-hirsute; leaves sub- sessile or contracted into a short margined petiole, blades 2-3 (-4) em. long, bipinnatifid or trifid, with the divisions more or less deeply incised, ultimate segments narrowly lanceolate, hirtellous to hirsute on both surfaces; spikes short-pedunculate, in fruit more or less compact; bracts shorter than the calyx, lanceolate, acute or acuminate, hirsute-ciliate; fruiting calyx -9 mm. long, densely glandular, somewhat viscid-pubescent, nerves hirsute, lobes short, subulate, unequal; corolla-tube 11-12 mm. long, pubescent around the throat; corolla-limb 6-8 mm. broad, lobes retuse; nutlets 2.5-3 mm. long, reticulate- Borabieulate above, siate toward the base; commissure muri- cately scabrous. ** Forma eglandulosa Perry, forma nova; calyce hirsutiori-hispido oe — Colleeted at Santa Rita, New Mexico, 1 Aug. 1911, Holzinger (MBG), T 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 329 Distribution: bee Colorado, New Mexico, and Arizona. Specimens exa ; Texas: San picti Havard (US); Barstow, 16 April 1902, Tracy & Earle 61 (F, G, MBG); northeast of Grand Falls, 17 May 1917, Clawson 13923 (US); near Feodora, Terrell Co., 26 May 1928, E. J. Palmer 33583 (MBG); Upper Blue Creek Canyon, Chisos i cope 1 July 1931, Moore & iibi $324 (MBG); Fort Davis, 9-12 July 1921, is & Duncan 2647 (MBG, P); Davis Mountains, 23 April 1902, T'racy & EL 168i in part (F, G, MBG, La foothills, Davis Mountains, 15 April-3 May 1902, Tracy & Earle 162a (NY); canyon of Upper Limpia Creek, 10 June 1926, E. J. Palmer 30672 (MBG); Guadalupe Mountains, 9 Sept. 1916, Young (MBG); flats near Van Horn, 13 May 1901, Eggert (G, MBG); Vieja Moun- tain, Oct. 1883, Havard 97 (G, US); Canutillo, Franklin Mountains, 12 July 1911, Barlow (US); mountains near Fronteras, 22 March 1852, Wright (MBG, NY), 1504 in part (G Type); Hueco Mountains, March 1851, Thurber 143 (G); about 8 km. north of Shafter, 10 April 1919, Hanson 551 (MBG ). CoronRApo: Durango, 20 July 1898, Baker, Earle & Tracy 531 (G, MBG, P). New Mexico: without locality, 1851-2, Wright 1603 in part (MBG); Copper Mines, 28 July 1851, Bigelow (NY); Rabbit Ear Creek, 28 June 1846, Wislizenus 483 (MBG); Las Vegas, 26 June 1895, Mulford 87 (MBG, NY); hills, Santa Fe, 15 May 1897, Heller 3536 (G, MBG, P); Sandia Mountains, 6 Sept. 1884, Jones 457 (P); Sandia Mountains, 28 April 1914, Ellis 17 (MBG); about 3 km. east of Albuquerque, 1915, Kammerer 83 (MBG); Albuquerque, Sept. 1884, Jones (NY, P); Albuquerque, 5 Sept. 1909, Rusby (NY); mountains west of Grant’s Station, 2 Aug. 1892, Wooton (US); South Camp, Magdalena Mountains, July 1897, Herrick 704 (US); White Mountains, Lincoln Co., 20 Aug. 1897, Wooton 864 (G, MBG, P); near Lincoln National Forest, 1903, Plummer (US); Vaughn, 10 Sept. 1921, Harwood (P); plains, White Sands, Otero Co., 25 Aug. 1899, Wooton (US); High Rolls and vicinity, 21-28 May 1902, Viereck (ANSP); Alamogordo, 7 April-24 May 1902, Rehn & Viereck (ANSP, MBG, P); Kingston, Sierra Co., 9 July 1904, Metcalfe 1090 (G, MBG, NY, P, US); Gallina’s Canyon, Black Range, Grant Co., 17 Aug. 1915, Pilsbry (ANSP); Mangas Springs, about 29 km. northwest of Silver City, 11 June 1903, Metcalfe 126 (G, MBG, P); near Silver City, June 1880, Greene (MBG, P); Silver City, 8 May 1919, Eastwood 8468 (G); Rincon, 16 May 1890, Jones (P); Big Burros Ranger Station, Gila Forest, Grant Co., 11 Sept. 1920, Eggleston 17231 (MBG, NY); Organ Mountains, 8 May 1895, Wooton (P); Organ Mountains, 11 June 1906, Standley (MBG, US); Filmore Canyon, Organ Mountains, Dofia Ana Co., 25 May 1905, Wooton (US); Van Pattens, Dofia Ana Co., 29 Aug. 1894, Wooton (US); Slaughter Canyon, 12-20 Aug. 1924, Standley 40625 (US). RIZONA: Long H. Ranch to St. John’s, 6-15 Aug. 1913, Griffiths 5190 (US); between Springerville and St. John’s, Apache Co., 15 Sept. 1917, Marsh 14224 (MBG). On account of its strong variability, this species has been exceedingly difficult to define. It was first set apart as an erect annual, with anther-glands as high and almost as large as the anther-cells. In the series of specimens at hand, the anther- glands are neither constant in size nor in height relative to the anther-sacs. The main characters appear to be the glandular somewhat viscid pubescence of the calyx, the very short acute- subulate calyx-teeth, and the somewhat compact spikes. [Vor. 20 330 ANNALS OF THE MISSOURI BOTANICAL GARDEN 43. V. ciliata Denth. Pl. Hartw. 21. 1839. Stems several from a common base, decumbent, branched, hirsute-hispidulous; leaves short-petiolate or subsessile, subbi- pinnatifid or commonly trifid, with the divisions more or less deeply incised, ultimate segments linear-oblong with slightly revolute margin, hirsute-strigillose on both surfaces; spikes subsessile or short-pedunculate, in fruit more or less compact; bracts a little shorter than the calyx, lanceolate, subulate or acuminate, hirsute-ciliate; calyx 7-8 mm. long, somewhat glandular, hirsute-hispidulous, lobes short, subulate, unequal; corolla-tube 10-12 mm. long, pubescent without; corolla-limb 6-8 mm. broad, lobes retuse; nutlets 2.5-3 mm. long, reticulate- scrobiculate above, striate toward base; commissural faces muricately scabrous. Distribution: New Mexico, Arizona to southern Mexico. Specimens examined: Ne Ew Mexico: “Upper Corner Monument, parallel 31° 47," 3 May 1892, Mearns 108 (G, NY). ARIZONA: without locality, 1892, Towmey 305% (US); Ash Fork, 18 June 1901, Barber 106 (US); Ash Fork, 20 June 1903, Griffiths 4757 (US); Aguila, 3 April 1930, Jones 26231 (MBG, P); Snowflake, 30 July 1897, Zuck (US); Pinedale, 1-15 Aug. 1897, Hough 109 (US); Fort Grant, Bonita, 18 July 1917, Munz 1249 (P); between Tombstone and Bisbee, 23 May 1928, Peebles 5353 (US). Mexico: without data, Coulter (ANSP); without locality, 26 July 1885, Schu- mann 1071 (G, US). Sonora: San Bernardino, Aug. 1852, T'hurber 769 (G). CnurinvAnua: Saint Eulalia plains, 26 Sept. 1885, Wilkinson (US); highlands near Chihuahua, 28 April 1847, Gregg (G, MBG, NY); valley near Chihuahua, 13 Sept. 1846, Pringle 1117 (F, US); vicinity of Chihuahua, 8-27 April 1908, Palmer ?9 (F, G, MBG, NY, US); vicinity of Madera, 27 May-3 June 1908, Palmer 281 (G, NY, US). DvunaNao: Valley of Nazas, Bolson de Mapimi, 11 May 1847, Gregg 632 (MBG); near El Salto, 12 July 1898, Nelson 4559 (MBG, US); Durango and vicinity, April- Nov. 1896, Palmer 345 (F, G, MBG, NY, US); Durango, 1 Aug. 1898, Nelson 4604 (MBG, US); Otinapa, 25 July-5 Aug. 1906, Palmer 401 (F, G, MBG, NY, US). OAHUILA: near Saltillo, 1847, Gregg 57, 257 (MBG); Saltillo, 15-30 April 1898, Palmer 74 (F, G, MBG, NY, US); Saltillo, 24 Aug. 1926, Fisher 223 (US); Oro, 18 Aug. 1903, Rose & Painter 6431 (US); La Ventura, 2-5 Aug. 1896, Nelson 3920 (US); Parras, March 1905, Purpus 1095 (NY); Valley of Nazas, ius 49, Gregg 26 (MBG); El Toro, near Movano, July 1910, Purpus 4524 in part (F, MBG). SAN Lala Porosr: near San Luis Potosi, 1876, Vis Md 717 (ANSP, G); near San Luis Potosi, 1878, Parry & Palmer 719 (ANSP, ; ZACATECAS: Zacatecas, July 1904, Kuntze 424 jeg Hacienda de Cedros, 1908, Lloyd 159 (US); near Plateado, 3 Sept. 1897, Rose 2769 (US); near San Juan Capis- trano, 19 Aug. 1897, Rose 2435 (US). 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA Sol AGUASCALIENTES: near Aguas Calientes, 9 Oct. 1903, Rose & Painter 7742 (US). Nayarit: Tepic, 5 Jan.-6 Feb. 1892, Palmer 2055 (US); Ixtlan, 19 Feb. 1927, Jones 23244 (MBG, NY, P). Jauisco: near Guadalajara, 11 Aug. 1902, Pringle 11091 (F, MBG, NY, US); near Guadalajara, 23 Feb. 1907, Safford 1398 (US); Guadalajara, Aug. 1909, Furness (F, US); Rio Blanco, June-Oct. 1886, Palmer 35 (G, NY, US). GUANAJUATO: Leon, 1839, Hartweg 176 (G, NY), TYPE collection; Guanajuato, 1894, Duges 496, 496a (G); Obregon, 28 July 1926, Fisher 207 (US). QUERETARO: near San Juan del Rio, 17 Aug. 1905, Rose, Painter & Rose 9503 (US); Queretaro, 1910-13, Arsène & Agniel 10252 (F, G, MBG, yt US). Hripauao: above Pachuca, 23 July 1898, Pringle 7691 (F, VERA Cruz: Orizaba, 1853, Mueller 325, 1209, 1319 pals Oraibi 4 Aug. 1891, Seaton 150 (F, G, US); Maltrata, Jan. 1883, Kerber 255 (US); Maltrata, 16 Aug. 1891, Seaton 391 (F, G, NY, US). PuEBLA: Cholula, 1 Jan. 1899, Deam (F), 86 (G); Tehuacan, Galeotti 736 (NY, US); near Tehuacan, 8 Sept. 1906, Rose & Rose hee (G, NY, US); vicinity of San Luis sare umen Se July 1908, Purpus 3407 (G, Mexico: meadows near Lecheria, 12 d nx Pringle 13158 (F, G, US); Valley of Mexico, 1865-6, Bourgeau 120 (G); Eslava, 15 June 1901, Pringle 9313 (G, MBG, NY, US); Amecameca, 24 July 1924, Fisher (F), 319 (MBG, US); near San Angel, 9 April 1849, Gregg 648 (MBG). MicnoacaN: low valley, Zinapecuaro, 2 May 1849, Gregg 756 (MBG); Querendaro, 28 Oet. 1895, Seler 1174 (G); Los Reyes, 8-12 Feb. 1903, Nelson 6858 (US). DERE voleano of Colima, 14 July 1892, Jones 686 (P). OAXACA: Cerro de las Soledad, 20 Nov. 1895, Seler 1379 (G); Las Sedas, 8 Sept. 1894, c. L. Smith 221 in part (US). This is a variable and wide-ranging species closely related to the bipinnatifida-Wrightii-ambrosifolia complex and scarcely capable of sharp delimitation. Its best characters are the hirsute-hispid pubescence, the usually sessile compact spike, the medium-short somewhat glandular calyx subtended by a slightly shorter bract. The specimens, Seaton 150, Pringle 9318, and Gregg 648, have typical spikes, but the leaves are small and the stems tend to be repent, suggesting a possible intergradation with V. teucritfolia. 43a. Var. longidentata Perry,?’ n. var. Hispidulous-hirsute plant of variable size and more or less open habit; leaves bipinnatifid; calyx-teeth 2-2.5 mm. long. Distribution: BÉ Texas, New Mexico, and Tamaulipas. Specimens examined: Texas: Post, 22 ie 1925, Wooton (US); Lubbock, 14 Máy 1930, Demaree 7685 27 Var. longidentata Perry, var. nov., planta hispidulo-hirsuta; foliis bipinnatifidis; calycis dentibus 2-2.5 mm. longis.—Collected at Matamoros, Mexico, April 1836, Berlandier 3020 (MBG), TYPE. [Vor. 20 332 ANNALS OF THE MISSOURI BOTANICAL GARDEN (US); Vista, May 1892, T'release (MBG); woods near Colorado River, near Wharton, 12 April 1925, Small & Wherry 11884 (NY); hills near San Antonio, 4 April 1901, Eggert (MBG); El Jardin, 10 March 1924, Runyon 630 (US); Laredo, Feb.-March 1913, Orcutt 5657 (MBG); Corpus Christi, 11 May 1900, Bailey 258 (US); Corpus Christi, May 1913, Orcutt 5792 (MBG); Aransas Pass, 1922, Schulz 845 (US); Kingsville, 25 March 1920, High 59 (MBG); Brownsville, 4 May 1900, Bailey 221 (US). New Mexico: Hagerman, 26 April 1929, Benke 5046 (F, G, MBG). Mexico: TAMAULIPAS: Matamoros, April 1836, Berlandier 1520 (G), 3020 (G, MBG rypr, NY). Scarcely differing from the species except in the relatively very long calyx-teeth and the often somewhat less deeply cleft leaves. 43b. Var. pubera (Greene) Perry, n. comb. V. pubera Greene, Pittonia 5: 136. 1903. Plants with more or less prostrate compact habit; ultimate segments of the leaves linear-oblong, margins strongly revolute. Distribution: rop to Arizona. Specimens exam t EXAS: Davis om 23 April 1902, Tracy & Earle 162 in part (MBG), TYPE collection of V. pubera; Sierra Blanca, El Paso Co., 22 May 1913, Orcutt 6172 (MBG). New Mexico: low mountains west of San Antonio, 14 April 1908, Wooton 3849 (US); near Silver City, May 1880, Greene (MBG, P); Deming, 29 April 1884, Jones 3801 (NY, P, US); Deming, 9 April 1930, Jones 26232 (MBG, P); Lordsburg, 3 July 1891, Evans (MBG); east of Lordsburg, 5 — 1930, Jones 26227 (MBG, P); Hudson Hot Springs, 2 —— Ag diee 804 (N ARIZONA: Cameron's, 16 June 1929, Jones (P); pM of Flagstaff, 8 Aug. 1922, Hanson A146 (F. MBG, fo Peach Springs, 15 June 1930, Jones 25480 (MBG, P); Dewey, 16 July 1922, W. W. Jones 180 (G); Taylor, 1 Aug. 1897, Zuck (MBG, NY); Chiricahua Mountains, 5 April 1897, T'oumey (US). This variety is a low plant with several stems from a common base, short internodes, and a tendency to branch freely. The leaves are more finely dissected than in the species and the margins are more revolute. 44. V. racemosa Eggert, Torreya 2: 123. 1902. V. pulchella Greene, Pittonia 5:136. 1903. Stems several from a common base, ascending-erect, branched, pubescent; leaves bipinnatifid or trifid with segments deeply cleft, ultimate segments linear, somewhat pubescent or hirtellous on both surfaces; spikes subsessile, elongate but dense at maturity ; braets scarcely so long as the calyx, lanceolate, acuminate, ciliate; calyx 5-6 mm. long, sparsely glandular-hirtellous, teeth 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 333 short, acute-subulate; corolla-tube 7-9 mm. long, puberulent or glabrous without; corolla-limb 5-6 mm. broad; nutlets 2 mm. long, reticulate-scrobiculate almost to base; commissural faces muricately scabrous. Distribution: perg New Mexico? DT exam - Exas: Upton Co., 2 May 1929, Cory 664 (G); Stockton, June 1881, Havard (US); Beier of Davis Mauniatus, 21 April 1902, Tracy & Earle 106a (F, G); low valleys near Sierra Blanca, El Paso Co., 15 May 1901, Eggert (G, MBG); Sierra Blanca, 22 May 1913, Orcutt 6184 (MBG); valley of the Rio Grande below Dofiana, Parry jaa Wright & Schott in part (US). w Mexico: without locality, 1851-2, Wright ?1498 (NY), 1501 (G). Fine pubescence, short calyx-teeth, small leaves, and elongated profusely floriferous spikes are the distinctive characters of this species. It is probably a close relative of V. ciliata. 45. V. Andrieuxii Schauer in DC. Prodr. 11: 553. 1847. Stems diffuse; branches procumbent, sub4-angled, hispid; leaves 4 cm. long, cuneate at base, sessile, deeply trifid, rugose with veins impressed above, strigose-hispid on both sides, middle lobe pinnatifid, lateral lobes spreading, lanceolate, acutish, margin incised or entire, subrevolute; spikes terminal and axillary, short-pedunculate, oblong, crowded, hispid, subglandular; bracts one-third shorter than the calyx, subulate-lanceolate; calyx 8-9 mm. long, teeth subulate; corolla-tube one-half longer than the calyx, puberulent, corolla-limb medium; schizocarp one-half the length of the calyx. isi pe southern Mexico. Specimens examined: duo ?Mexico: between Mexico and Oaxaca, Andrieux 138 (DeCandolle Herb. TYPE, F phot., K, MBG phot.). Oaxaca: mountain ridge, west side of valley of Cuicatlan, 10 Nov. 1894, Nelson 1895 (US); Tieneguilla, 5 April 1895, L. C. Smith 348 (G); valley of Oaxaca, 2 Oct. 1894, Nelson 1613 (G, US) The above description is a close translation of the original. Even with the photograph of the type, the species is difficult to interpret without floral dissections. For the present, it seems preferable to maintain its status, referring here the above some- what similar collections. [Vor. 20 334 ANNALS OF THE MISSOURI BOTANICAL GARDEN 46. V. teucriifolia Mart. & Gal. Bull. Acad. Brux. 11°: 322. 1844. V. exilis Schauer in DC. Prodr. 11: 553. 1847. Stems several from a common base, prostrate or repent, more or less diffusely branched, 4-angled, glabrate or hirtellous; leaves 1.5-2.5 em. long, cuneate at base, narrowed into a margined petiole, pinnatifid-incised, segments linear-oblong, obtusish, often glabrous or sparsely strigillose on both surfaces, nerves impressed above, subglaucescent beneath; margin slightly rev- olute; spikes terminal, sessile or short-peduncled, scarcely emerging from the highest leaves, few-flowered; bracts approxi- mately one-half as long as the calyx, lanceolate, acuminate-acute, ciliate; calyx 5-7 mm. long, sparsely hirtellous or occasionally hirsute, lobes short, acute; corolla-tube a little longer than the calyx; corolla-limb 5-7 mm. broad; anther-glands minute; nutlets 2.5-3 mm. long, chiefly reticulate-scrobiculate; commissu- ral face muriculate or almost smooth. Distribution: Nexioe and Guatemala. Specimens examine Mkxico: San Luis Porosi: Huaxalote, Ehrenberg 131 (Bot. Mus. Berl.-Dahl. TYPE of V. exilis; MBG phot.). doce ye Leon, June-July 1841, Liebmann 11329 (US). Hiparco: near Real del Monte, 2 June 1899, Rose & Hough 4484 (US). Vera Cruz: sand hills, Perote, Halsted (NY); Perote, near Jalapa, 1894, C. L. Smith 1474 (G); peak of Orizaba, June-Oct. 1840, Galeotti 777 (K), TYPE collection; Vaqueria del Jacal and Orizaba, Oct. 1841, Liebmann 11328 (US); “Talvare,” Feb. 1843, Liebmann 11330 (US); Boca del Monte, 13 March 1894, Nelson 196, 223 (US). : common in dry volcanic sand, Chalchicomula, 8 April 1890, Stone (ANSP); Chalchicomula, 20 Feb. 1892, J. G. Smith 464 (MBG); near Chalchicomula, 15 March 1894, Nelson 242 (US); Popocatepetl, 7-8 Aug. 1901, Rose & Hay 6049 (US). Mexico: La Cima, Jalapa, Aug. 1904, Kuntze 23738 (NY); Sierra de las Cruces, Salazar, 13 Aug. 1896, Harshberger 41 (ANSP, G, US); lava beds, near Eslava, 19 July 1910, Rusby 124 (NY); near Toluca, 23 June 1889, Pringle 2927 (F, G); valley of Toluca, 15 Aug. 1892, Pringle 4180 (ANSP, F, G, MBG, NY, US); San Angel, 12 Aug. 1910, Orcutt 3551 (F, MBG, US). More tos: Toro, 5 Aug. 1924, Fisher (F, MBG), 337 (US); Tres Marias, 11 Aug. 1906, Pringle 13792 (G, US). Mıcnoacan: Morelia, 9 July 1910, Arsène (F). OAXACA: : Chinantle, May 1841, Liebmann 11308, 11316 (US). ENTRAL AMERICA: GUATEMALA: near Quezaltenango, 23 June 1882, Lehmann 1603 (US); Quezaltenango, Oct. 1886, von T'uerckheim 1068 (ANSP, G, US). 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 335 The specimens cited vary greatly in degree of pubescence. The larger number are sparsely hirtellous or strigillose. The following are densely hirtellous-hirsute: Liebmann 11329, Nelson 196, 223, Smith 1474, Halsted, Smith 464, Nelson 242 The Guatemalan collections appear to be conspecific, although both the cilia of the bracts and the trichomes along the nerves of the calyx are somewhat coarser and stiffer than those of the Mexican material. V. teucriifolia is a slender plant easily separated from its nearest relatives by its prostrate or repent habit and few-flowered spikes. 46a. Var. corollulata Perry;?? n. var. Usually finely and densely pubescent; calyx 5 mm. long, spreading-pubescent, lobes very short; corolla inconspicuous, tube barely protruding beyond the calyx, limb 3 mm. broad; nutlets 2.5 mm. long. Distribution: Mexico. Specimens examined: XICO: San Lurs Porosi: region of San Luis Potosi, 1878, Parry & Palmer 718 (G, MBG TYPE, US), 7/9 in part (ANSP, F, MBG, US). ZACATECAS: plains, LaHonda, 18 Aug. 1890, Pringle 8651 (G); Zacatecas, 1903, Purpus 187 (US). Hipauao: Hacienda Palmar, near Pachuca, 21 July 1905, Rose, Painter & Rose 8823 (G, NY, US); Telles, Sept. 1910, Orcutt 4146 (F, MBG). The variety differs from the species chiefly in its smaller corollas; perhaps it is only a dimorphic form. 47. V. Gooddingii Briq. Ann. Conserv. & Jard. Bot. Genève 10: 103. 1907. V. verna var. fissa Nelson, Am. Jour. Bot. 18: 437. 1931. Stems usually several from a common base, erect or decumbent- ascending, cinereous-green, branched, densely pilose or some- what villous, often glandular; leaves 3-5 cm. long, tapering at the base into a short margined petiole, 3-cleft, divisions coarsely toothed or incised, cinereous-green, more or less villous-hirsute 28 Var. corollulata Perry, var. nov., planta dense pubescens; calyce 5 mm. longo patenti-pubescente; corolla inconspicua, tubo vix exserto, limbo 3 mm. lato; coccis 2.5 mm. longis.—Collected in the region of San Luis Potosi, Mexico, 1878, Parry & Palmer 718 (MBG), TYPE. [Vor. 20 336 ANNALS OF THE MISSOURI BOTANICAL GARDEN on both surfaces, midrib and veins slightly prominent beneath, margins scarcely (if at all) revolute; spikes pedunculate, fascicle- like in anthesis, somewhat elongated in fruit; bracts usually a little shorter than the calyx, lanceolate, acuminate, villous- hirsute, long-ciliate; calyx 8.5-11 mm. long, villous-hirsute, more or less glandular, lobes slender, subulate, unequal; corolla-tube very little longer than the calyx, pubescent without; corolla- limb 8-9(-12) mm. wide, segments retuse; nutlets 3(-3.5) mm. long, subcylindrie, reticulate-scrobiculate except at the striate base. Distribution: Utah, Arizona to California, and Lower California. Specimens examined: Uram: southern Utah, 1876, Johnson (US). Nevapa: sandy soil, Charleston Mountains, about 1800 m. alt., May-Oct. 1898, Purpus 6061 (P, US); Trout Creek, Charleston Mountains, about 1900 m. alt., 1926, Jaeger (P); Calientes, about 1400 m. alt., 29 April 1914, Jones (P); Kernan, Valley Meadow Wash, 28 April 1902, Goodding 645 (F, G, MBG, P), Tyre collection. Arizona: without locality, 1891, MacDougal 603 (NY); Chloride, about 1350 m alt., 15 April 1903, Jones (P); Hackberry, 24 May 1884, Jones (F, P, US); Peach Springs, June 1884, Lemmon (US); Peach Springs, April 1893, Wilson 99 (US); Kingman, 11 March 1912, Wooton (US); Skull Valley, about 1300 m. alt., 28 April 1903, Jones (P); Hillside, about 1100 m. alt., 1 May 1903, Jones (P); near Williams, 11 April 1905, Wilcor (US); Jerome Divide, 25 July 1921, W. W. Jones (MBG); east of Jerome Junction, 1 May 1908, Tidestrom 898 (US); Rio Verde, Fort Whipple, 30 Aug. 1865, Cowes & Palmer 498 (MBG); Fort Verde, 1888, Mearns (NY); Natural Bridge, 23 April 1904, F. M. Chamberlain 56 (US); Bradshaw Mountains, 22 June 1892, Toumey 305a (US); 16 km. south of Black River, White Mountains, about 1800 m. alt., 22 June 1930, Goodman & Hitchcock 1297 (MBG, NY); Galiuro Moun- tains, 29 July 1894, Toumey (US); Tucson, 16 May 1896, Zuck (NY, US); washes, near Baboquivari Mountains, 24 Feb. 1923, Hanson A1021 (F, MBG); Baboquivari Mountains, 10 March 1926, T'hackery & Leding 1100 (US); south of Warren, Sulphur Springs Valley, 3 June 1915, Carlson (US); Lowell, May 1884, W. F. Parish 196, 197 (G); wash near Agua Caliente, 20 Jan. 1920, Bartram 255 (ANSP). CALIFORNIA: east side, summit of Providence Mountains, 29 May 1861, Cooper (US); vicinity of Bonanza King Mine, east slope of Providence Mountains, about 1200 alt., 21-24 May 1920, Munz, Johnston & Harwood 4252 (P); Rock Spring, San Bernardino Co., May 1876, Palmer (G), 539 (F, MBG, NY, US); southern part of San Diego Co., 1875, Palmer (G). Mexico: Lower Cauirornias: Pifione Forest, 25 July 1883, Orcutt (F). In habit V. Gooddingii is much like V. ciliata, but differs in the villous-hirsute pubescence and the larger flowers. It is readily recognized in its typical form by the long-ciliate character of the bracts, the short dense villous spikes, and the shaggy pubes- cence. 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 337 47a. Var. nepetifolia Tidestrom, Proc. Biol. Soc. Wash. 38: 15. 1925. V. arizonica Briq. Ann. Conserv. & Jard. Bot. Genéve 10: 102. 1907, not V. arizonica Gray, Proc. Am. Acad. 19: 95. 1883. V. verna Nelson, Am. Jour. Bot. 18: 436. 1931. Leaves broadly ovate, coarsely and unevenly dentate, at times lobed, abruptly narrowed into a cuneate base; pubescence varying from strongly pilose to densely villous. Distribution: Colorado, Nevada, Arizona, and northwest Mexico. Specimens examined: eae southern Colorado, 1867, Parry (US). Nevapa: El Dorado Canyon, Lincoln Co., 1880, T. W. Davis (MBG); El Dorado at Nelson, 30 April 1907, Jones (P). ARIZONA: Oatman, 4 May 1928, Thackery 348 (US); Hackberry, 7 March 1912, Wooton (US); Camp Verde to Prescott, Aug. 1896, Fernow (US); by streams of Santa Catalina Mountains, 13 April 1881, Pringle in part (F, G, MBG); Tucson, 1871-5, Rothrock (US); Tucson, 1881, Vasey (US); Tucson Mountain, 7 March 1901, Griffiths 2421 (NY); Tucson Mountain, 13 March-23 April 1903, Griffiths 3489 (MBG, US); Laboratory Hill, vicinity of Tucson, April 1908, Rose 11968 (US); Tucson, 1909, Parish (P); vicinity of Tucson, 1910, Rose, Standley & Russell 15176, 15181 (NY, US); Tucson, 1911, Beard (MBG); Tucson Mountain, oe 11, 1913, Greenman & Greenman 56 (MBG); Tumanoc Hill, Tucson, 25 July 1916, J. A. Harris C16375 (NY); rocky slopes west of Tucson, 26 Dec. 1919, Bartram b (ANSP, US); near Tucson, 17 March 1927, Harrison & Kearney 3646 (US); Sells, Papago Reservation, 24 Feb. 1926, Loomis & Thackery 910 (US); Huachuca Mountains, Aug. 1882, Lemmon (US); south Huachuca Mountains, 31 May 1930, Peebles 6755 (NY); Yucca, 14 May 1884, Jones 3901 (F, P, US), TYPE collection of V. arizonica Briq., 36 (G) Mexico: Lower CALIFORNIA: Pifione Forest, 6 Oct. 1882, Orcutt (US); San Jacinto, 18 July 1885, Orcutt 1303 (F, MBG, NY); Rosario, 3 May 1886, Orcutt 5141 (F, MBG, NY); Agua Dulce, 48 km. southeast of San Fernando, 9 Sept. 1905, Nelson & Goldman 7123a (MBG, US) ONORA: ‘‘Niggerhead Mountains, near Monument no. 82," Aug. 1893, Mearns 1904, 1914 (US); Alamos, 26 March-18 April 1890, Palmer 307 (G, US); Alamos, 27 Jan. 1899, Goldman 292 (G, NY, US); Sierra de Alamos, 14 March 1910, Rose, Standley & Russell 12840 (NY, US); Alamo, West Magdalena, 17 May 1925, Kennedy 7033 (US); Pinocate Mountains, 21 Nov. 1907, MacDougal 73 (US); La Cienega, 18 July 1911, Goodding 952 (US). SiNALOA: without locality, 1922, Ortega 4541 (US); sandy soil along river, vicinity of Fuerte, 25 March 1910, Rose, Standley & Russell 13449 (NY, US). The variety differs from the species in its much less indented and broader leaves. The pubescence is much more variable and at times harsh, suggesting an intergradation with V. bipinnatifida. [Vor. 20 338 ANNALS OF THE MISSOURI BOTANICAL GARDEN The leaf-forms recall V. canadensis, but the spike is scarcely distinguishable from V. Gooddingit. 48. V. pumila Rydb. in Small, Fl. Southeast. U. S. ed. 1, 1010. 1903, and ed. 2, 1913. V. inconspicua Greene, Pittonia 5: 137. 1908. V. brevibracteata Eggert, Torreya 2:124. 1902, not V. bracteosa var. brevibracteata Gray, Syn. Fl. N. Am. 2!: 336. 1878 Stems usually several from a common base, branched, de- eumbent-ascending, hirsute, often finely glandular; leaves 1.5-3 em. long, obtusely triangular with cuneate or truncate base contracted into a short narrowly margined petiole, trifid, occa- sionally lobed, divisions variously incised, appressed-hirsute on both surfaces; spikes short-peduncled to sessile, fairly compact; bracts almost as long as the calyx, linear-lanceolate, hispid- hirsute; fruiting calyx about 6 mm. long, pubescent, hispidulous along the nerves, at times finely glandular, lobes subulate, short; corolla-tube a little longer than the calyx, slightly, if at all, pubescent without; corolla-limb 3-5 mm. broad; anther-glands minute or wanting; nutlets 2.5(-3) mm. long, reticulate-scrobic- ulate except at base; commissural faces muriculate. Distribution: erect Texas, New Mexico; and Sonora and Sinaloa, Mexico. Specimens examine OKLAHOMA: Caton, 8 May 1895, Bush 1275 (MBG, NY mvrE); Oklahoma City, 11 May 1891, Carleton 134 (US); vicinity of Fort Sill, 6 May 1916, Clemens 11752 (MBG); Arbuckle Mountains, Davis, 1 April 1916, Emig 401 (MBG); Price’s Falls, Murray Co., 30 April 1926, Stratton 11 (MBG); near Crusher Spur, Murray Co., 12 April 1913, Stevens 28 (G, MBG, NY, US); near Tishomingo, 15 April 1916, Houghton 2570 (G). Texas: indefinite data, Lindheimer 501 (F, G, MBG, US); Dallas, 17 March 1876, Reverchon 788 (MBG, US), 1963 (F, G, MBG, NY), 2117 (MBG); Poly- technic, 10 April 1913, Ruth 110 (MBG); Tarrant Co., 10 June 1913, Ruth 110 (G); Haskell, 1898, Morton (MBG); sandy ground, near Granbury, Hood Co., 6 May 1900, Eggert (MBG); sandy ground, Big Spring, Howard Co., 11 June 1900, Eggert (MBG); Round Top Mountain, Comanche Co., 9 May 1900, Eggert (MBG); near Comanche, 10 May 1900, Eggert (G, MBG); San Saba River, Brady, 16 April 1926, Studhalter 1106 (US); San Saba, 8 May 1917, E. J. Palmer 11843 (MBG); plains west of Pecos, 20 April 1902, Tracy & Earle 106 (F, G, MBG, US); foothills, Davis Mountains, 23 April 1902, Tracy & Earle 178 (F, G “MBG, NY, US); sandy ground, near Bastrop, 17 April 1929, E. J. Palmer 33367 (MBG, NY); its hills, Austin, May 1872, Hall 428 (G), 431 (MBG, NY, P); Austin, Bray 98 (NY), Young (G, MBG), Young 114 (P), E. J. Palmer 9389 (MBG), Tharp 1862, 1364 (US), Armer 6881 (US); gravel bars of Blanco River, Blanco, 5 April 1918, E. J. Palmer 13281 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 339 (MBG); Big Branch, Gillespie Co., Jermy 184 (MBG); New Braunfels, March 1850, Lindheimer 434 (MBG), 1075 (F, G, MBG, NY, US); Lacey’s Ranch, Kerr Co., 1 June 1916, E. J. Palmer 10008 (MBG); Kerrville, 28 March 1916, E. J. Palmer 9278 (MBG); Hungerford, 4 March 1914, E. J. Palmer 4840 (MBG); Wharton, 18 March 1914, E. J. Palmer 4977 (MBG); Bexar Co., Jermy (MBG), 84, 85, (US); San Antonio, March 1882, Havard (US); San Antonio, Jermy (NY), Jermy 209 (G), Wilkinson 8, 108 (MBG), Trelease 99 (MBG), Eggert (MBG), Bush 1185 (MBG), Clemens & Clem- ens 976 (MBG, P); Rock Springs, 17 April 1930, Jones 26220, 26228 (P); gravel and sand bars, small streams, Uvalde, 11 May 1918, E. J. Palmer 13562 (MBG); near Vic- toria, 11 April 1900, Eggert (G, MBG); Riverside, Walker Co., 24 March 1918, E. J Palmer 13175 (MBG); Brownsville, 23 Jan. 1919, Hanson (MBG); Brownsville, 30 Jan. 1919, Hanson 322 (G, NY, US); Brownsville, 14-15 March 1923, Tharp 1871 (US); southwest Texas, 1851-2, Wright 1500 (ANSP, F, G, MBG, NY). New Mexico: without data, 1851-2, Wright 1501 (ANSP); Carlsbad Cavern, 5 May se pum 109 (US) MExic xod: Alamos, 26 March-8 April 1890, Palmer 326 (G); along an arroyo, vicinity of Alamos, 13 March 1910, Rose, Standley & Russell 12745 in part (NY, US). SiNALOA: San Blas, 2 Feb. 1927, Jones 23243 (MBG). Verbena pumila has often been confused with V. quadrangulata and, superficially, closely resembles it, but is readily separated on the beakless nutlets. 49. V. setacea Perry,’ n. sp. Pl. 15. Stems decumbent-ascending, soft-pubescent; leaves 3-5 cm. long, with cuneate-truncate base narrowed into a margined petiole, trifid with segments coarsely dentate, lateral lobes small, veins somewhat prominent beneath, soft-villous-pubescent on both surfaces; spikes short-peduncled; bracts linear-lanceolate, about three-fourths as long as the calyx, subulate-setaceous, soft-pubescent, ciliate; calyx 6 mm. long, villous-pubescent, teeth (2 mm.) about half as long as the tube, subulate-setaceous; corolla-tube 7 mm. long, pubescent without; corolla-limb about 6—7 mm. broad; anthers not glandular; mature nutlet not seen. 29 V, setacea Perry, spec. nov., herbacea vel basi suffruticosa; caulibus decum- bentibus ramosis ramis spoon iode piloso-pubescentibus; iie is basi cuneata in petiolum alatum attenuatis 3-5 cm. longis tri tatis lateralibus parvis subtus venoso-reticulatis utrinque molliter villoso-pubescentibus; spicis breviter pedunculatis; bracteis lineari-lanceolatis subulato-setaceis piloso-pubescent- ibus ciliatis calyce paulo brevioribus; calyce 6 mm. longo villoso-pubescente, calycis dentibus subulato-setaceis 2 mm. longis; corollae tubo 7 mm. longo extus pubescente; limbo fere 6-7 mm. lato; connectivo antherarum superiorum inappendiculato; fructu immaturo.—Collected at Calmalli, Lower California, Jan.-March 1898, Purpus 195 (P), TYPE. [Vor. 20 340 ANNALS OF THE MISSOURI BOTANICAL GARDEN Distribution: Lower California. Specimens examined: Mexico: Lower CALIFORNIA: rocks, Calmalli, Jan.-March 1898, Purpus 195 (P TYPE). The gross habit of this plant is very like that of V. Gooddingii var. nepetifolia. Unfortunately the inflorescence is immature and only a few flowers are in anthesis, hence it is rather difficult to say what are the characters of the spike or of the mature nutlets. Moreover, the corolla may be larger than appears in this speci- men. ‘The flower itself is similar to that of V. lilacina, but the two plants are so different in habit it would seem as if this were perhaps only a superficial resemblance. The species is readily distinguished by its general habit and the long setaceous calyx- teeth 50. V. lilacina Greene, Bull. Calif. Acad. Sci. 1: 210. 1885. Stems erect, much branched, 0.5-1 m. high, very sparsely hirsute or glabrous except just below the spike; leaves 3-5 cm. long, contracted at base into a margined petiole, bipinnatifid (upper pinnatifid), divisions remote, ultimate lobes chiefly linear and acute, somewhat scabrous and strigillose on both surfaces, rugose above, midrib prominent beneath; spikes fascicle- like, cymosely arranged, long-pedunculate, bracts somewhat shorter than the calyx, lanceolate-setaceous, pubescent, ciliate; calyx about 7 mm. long, appressed-pubescent, short-hirsute along the nerves, lobes unusually long (2.5-3 mm.), attenuate into subulate-setaceous teeth; corolla-tube protruding very little beyond the calyx; corolla-limb 10 mm. broad, segments emargin- ate; anthers not glandular; nutlets almost smooth, slightly enlarged at the base; commissural face muricate-scabrous, not reaching the tip of the nutlet. Distribution: Known only from Cedros Island. Specimens examined: XICO: Lower CALIFORNIA: Cedros Island: 29 April 1885, Greene (G); 1889, Palmer (G), 677 (F, NY); March-June 1897, Anthony 288 (F, G, MBG, US); 12 March 1911, Rose 16155 (NY, US). This is an anomalous species of uncertain relationship. It has an erect open habit with more or less glabrous stem and long internodes somewhat suggesting V. neomexicana (§ Verbenaca); 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 341 nevertheless, the short stout dense spike does not point to an affinity with this section. Although the anthers are not glandu- lar, the sterile style-lobe protrudes well beyond the stigmatic surface, and, in mature fruit, the style appears to have been inserted in a deep depression at the apex of the schizocarp; hence the species is regarded provisionally as a member of the section Glandularia. 51. V. amoena Paxton, Mag. Bot. 7: 3, pl. 1840. V. grandiflora Sesse & Mocino, Pl. N. Hispan. ed. 1, 6. 1887-90 [La Naturaleza, II. 1. App.]. Stem about 5 dm. tall, decumbent or ascending, retrorsely hispidulous; leaves 5-8 cm. long, bipinnatifid, with the lower part entire, forming a broadly margined subauriculate and semi- amplexicaul base, divisions remote, linear-oblong, sparsely incised, appressed-pubescent above, hispidulous beneath with midrib and veins prominent; internodes 4-6 cm. long; spike terminal, fascicle-like in anthesis, pedunculate; bracts linear- lanceolate, subulate, pubescent; calyx about 10 mm. long, densely pubescent, particularly along the nerves, glandular, teeth slender, subulate; corolla-tube slightly longer than the calyx, pubescent without, particularly around the throat; corolla- limb about 9 mm. broad; anthers unappendaged; mature fruit not seen. Distribution: Mexico. Specimens examined: Mexico: MicHoacan: Puruandiro, Sesse & Mocino 99 (Herb. Bot. Gard. Madrid TYPE, MBG phot.). Mexico: hills, Lecheria, 5 July 1904, Pringle 13434 (US). This unique species is readily recognized by its coarse habit, together with its bipinnatifid, subauriculate and semiamplexicaul leaves. Although Paxton’s description is not accurate, the specimen agrees well with his plate and with the photograph of V. grandiflora. DOUBTFUL OR LITTLE-KNOWN SPECIES ‘í VERBENA ERINOIDES Lam." This species has established itself in several places. It belongs to a South American species- [Vor. 20 342 ANNALS OF THE MISSOURI BOTANICAL GARDEN complex needing critical study to determine accurately its real identity. V. GRANDIFLORA Ortega, Hort. Matr. Dec. 2. 1797. Although the description is inadequate for positive identification, the phrase ‘‘Semina saepius duo" would seem to indicate that the species does not belong to the genus Verbena. V. nEPENS Spreng. Erst. Nachtr. d. Beschrieb. d. Bot. Gart. Univ. Halle, 40, no. 51. 1801. The literature on this species is very vague. The plant described is said to be a native of Santo Domingo. According to Lamarck,? the vervain of Santo Domingo is a species of heliotrope. V. TRIFIDA HBK. Nov. Gen. et Sp. 2: 273, pl. 134. 1818. This is a true Verbena, but none of the specimens at hand agrees with the description; and the floral drawings are insufficient to reveal its affinities. V. BARBATA Grah. Edinb. New Phil. Jour. 176. 1827. This is undoubtedly a member of the section Verbenaca, but the de- scription is too meagre to identify it. V. INCARNATA Raf. Atl. Jour. 154. 1832. Here again the description is too inadequate for specific identification. V. DELICATULA Mart. & Zucc. in Otto & Dietr. Allg. Garten- zeit. 2: 245. 1834. Unfortunately the essential characters, which would separate this species from its allies, are not clearly defined. The description is so much like that of V. barbata that the writer suspects that these two species are identical. V. MamruEsrn Turez. Bull. Soc. Nat. Mosc. 36?: 196. 1863. In the group to which this species evidently belongs, the units are very closely related; hence without more specifie definition the species is obscure. V. PAUCIFOLIA Turez. Bull. Soc. Nat. Mosc. 36?: 196. 1863. This may be V. carolina, although the leaves of the latter are somewhat broader; but until such time as the type may be studied it seems preferable to withhold any decision regarding it. V. INTEGRIFOLIA Sesse & Mocino, Pl. Nov. Hispan. 6. 1887 [La Naturaleza, II. 1. App.]. No known Mexican species of Verbena has entire leaves. 30 Lamarck, Encyc. 1, Suppl. 5: 469. 1817. 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA V. SCABRELLA Sesse & Mocino, l. c. 343 Apparently this species does not belong to the genus, but anything further regarding its identity is unknown to the writer. V. 12: 12 synonymy under V. canadensis. AUBLETIA var. LAMBERTI M. E. Jones, Contr. West. Bot. 1908. This combination, nomenclatorially, belongs in Its entity is clearly with the section Glandularia, but with which species it should be associated is unknown. LIST OF EXSICCATAE The collectors’ numbers are printed in italics. Unnumbered collections are in- dicated by a dash. The number in parenthesis is the species number used in this revision. Abrams, L. R. — (22, 31); 2481 (22); 2574 (31); 3406 (7); 3787 (22); 9487 (31). Abrams, L. R. & McGregor, E. A. & (22 Adole, Bro. 22 (29). Allard, H. A. 20? (31). Anderson, I. W. — (16). Anderson, J. P. — (15, 16). Anderson E. S. & Woodson, R. E., Jr. 4? (17); 1577? (32). Andrews, L. — (16). Anect, Bro. 5? (20); 100, 143, 179 (41); 211 (20). Anthony, A. W. 288 (50); 380 (4). Antisell, T. 186 (41a). Applegate, E. I. 2228 (22). Armer, * A. 5381 (48); 5385 (8). Arséne — (1, 5, 11, 46); 672 (16); 2798 b ; 8000 (11); 6129 (34); 9998 (7); 10626 (29); 11820 (32); 11831 (8); 11859 (2); 11982 (35); 12117 (32); 12242 (8); 12534 (2); 18399, 18407, 18543, 18593 (31); 18615 (41); 18684 (31); 18793 (41); 18885, 18961, 18964 (31). Arsène & Agniel, 10242 (7); 10252 (43). Arséne & Benedict, 15734 (31); 16306 (41); 16600, 16790 (31) Arthur, J. C. 24 (17). Ashe, W. W. — (17). Bailey, V. 221, 258 (43a). Bain, S. M. 328 P 444 (16). Baker, C. F. , 16, 17); 288, 564 (31); 565 (20): (31); 2591 (6). Baker, C. F., Earle, F. S. & Tracy, S. M. 531 (42 ` Ball, C. ; $44 (8); 401 (40); 553 n. 556 Eo. 605 (24); 909 (40); 1171 (25); 1585 (19); 1675 (31). Ball, J. — (31). Ballard, C. A. — (16). Barber, H. S. 106 (43). Barkelew, F. E. 231 (4). Bartlett, A. H. 1122 (1). artram, E. B. — (7); 255 (47); 256 (25); 257 (478); 1 1021 (17). Basile, b 99 (7). Bates, — "o 19, 40). Beard, Beattie, F br — (us 16). Beaumont — Beckwith, F. 48 (19); 138 (20); 772 (38). Benke, H. C. 4566 (35); 5046 (43a); 5164 19). Berg, H. K. — (14). Bergmann, L. S. — (31); 2385 (16); 2370 (31). Berkley, E. E. 1304 (16). Berlandier, J. 322 ES 429 —?1749 (40); 344 578 (28); 644 (25, 29a); 827 (29a); 1222 (12); 1449=189 (40); 1485 (= 225) in part (25, 33); 1511 (8); 1520 (43a); 1592 (8); 2054 (29a); 2428 = 998 (40); 2506 in part (25); 3016 (8); $018 — 1518 (33); 3020 (43a). Bernoulli, G. 127 v as (5). Bessey, C. E. — (15 Bigelow, J. M. Mer 2 31, 35, 42). Billings, F. H. 49 (1). Biltmore Herbarium, /082b (31); 8653a (19); 4759, 4759b (17); 4761a (32); 4762 (6). Bissell, C. H. — (16). Blakeley, O. W. 1462 (30; 1471 (35). Blanchard, F. — (15, 17). Blanchard, W. H. 26, ^ 161 (17). Blankinship, J. W. — (15, 17, 19, 22, 35). Blanton, F. S. 6598 (3). Blewitt, A. E. 14 (19). Blumer, J. C. 1345 (40); 1612 (28); 1783 (11); 1804, 2170 (26a). Boettcher, F. L. J. 176, 226 (15). Bogusch, E. R. 1236 (24). Bolander, H. N. ee (31). Botteri, M. 180 (11). Bourgeau, E. — (16); 119 (11); 120 (43); 860 (7); 861 (28); 547 (7). Boyce, H. T. E. — (1 — ki 8. — (11, B. 466 Brannon, i 0.1 Braunton, E. 99 ain 328 (22); 1263 (23). 16, 20, 41); Bray, W. L. 65 (35); 98 ay ET (292). Brewer, W. H. 31 a 229 (22). Brinton, J. B. — (17 Britton, N. L. — LM 220, 296 (37); 1627 (14); 3176 (1). Britton, N. L. & Brown, M. S. 28 (6); 98 (2); 158 (1); 878, 1681 (14). Britton, N. L. & Cowell, 10326 (6). i . & Hollick, 2704 (14). . & Wilson, P. 5771 (14). Britton, N. L., Britton, E. G. & Brown, M. S. 6019, 7037 (14). Britton, N. L., Britton, E. G. & Shafer, J. A. 292 (14) [Vor. 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN Britton, N. L., Brown, M. S. & Wortley, 1645 (1). Britton, N. L., Wilson, P. & Leon, Bro. 15256 (14). Broadhead, G. C. — (15, 35). Brown, H. E. 58 (3 E Brown, 8. 492 (6); 693 (1) Brown & Brit qiio Britton & Brown. Bryant, M. L. 88 (22) Bucholz, A. F. — (15). Buckley, S. B. — (8, 31, 32). Burgess, A. B. 638, 713, 783 (37). Burk, M. 542a (31); 591 (15). Burnham, 8. H. — (15, 17). Bush, B. F. — (15, 17, 35); 28, 77, 84 (8); 205 (40); 206 (17); 283 (35); 312 (32); 318 (35); 851 (40); 432 (15); 433 (17); 435 (19); 436 (31); 437 (35); 449 (40); 450 (17); 452 (15); 475 (19); 545 (35); 569 (35); 570 (19); 601 (31); 884 (15); $99 (24); 914 (17); 1034 (31); 1168 (35); 1185 (48); 1275 (coll. of 1900) (24); 1275 (coll. of 1895) (48); 1482 (15); 1933 (35); 2109 (31); 2214 (16); 2901 (35); 4010 (31); 4029 (40); 4922, 6439 (35); 7586a, 7647 (17); 8537 (31); 9165 (16). Butler, G. D. 1621 (22). Calderon, S. 729 D HM (11); 925 (5). Campbell, J. E. — (16). Campbell, J. R. 67 TA Canby, W. M. — (32); 193 (33); 194 (29a). Carleton, M. A. —, 177 (40). Carleton, M. L. 134 (48). Carlson, J. I. — (7, 23, 26a, 41a, 47). Carr, W. P. 118 wt 179 (81). Carter, W. R. — (16). Olubealt. E. : 418 (31). oe E. B. & Knowlton, F. H. 16). Chamberlain, F. M. 56 A Chandler, H. P. — (8, 16). Chandonnet, Z. L. — (15, 16, 19, 31). Chapline, W. R. 609 (26 Chapman, A. W. — (6, 14, 15, 16, 31, Chass, A — (15). Cheney, C. I. — (16). 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA Chestnut, V. K. — (22 Choussy, F. 36 (11). Churchill, J. R. — (31). Clark, J. A. 55 (31); 253 (16); vila (31). Clark, O. M. 4041 (29a); 4250 (27). Clawson, A. B. 13923 (42 Clemens, Mrs. J. — (22, 31); 11748 (19); 11749 (15); 11750a (31); 11761 (40); 11752 (48). Clemens, Mr. & Mrs. J. 969, 970 (8); 971, 972, 978, 974 (29a); 976 (48). Clements, F. 2612 (31); 2776 (40). Clements, F. E. & E. S. 9 (31 Clifton, R. L. 3085 (40). Clute, W. N. 93 (31). Clokey, I. W. — (36); 1 11). Collins, F. S. 45 (1); 53, pel EE 267 (6); 268 (14); 269 (1). Congdon, J. W. — (23). Conzatti, C. 4194 (29); 4207 (7); 4288 9 ). O. F. & Griggs, 148 (5). Cooper, J. G. — (47). Cooper, J. J. 5890 (5). Cory, V. L. 664 (44). Cotton, J. s. 621, 1363 (31); 1896 (16). Coues, E. & Palmer, Ed. 279 (16); 498 (47); 671 (26a). Coulter, — (43). Coville, F. V. & BE F. 968 (31). wen, J. H. Crandall, C. S. 25 (16). Croft, M. B. 78 (33 Culbertson, n W. 4810 (22) Curtis, C. D. — (15 Curtis, M. A. — (6). Curtiss, A — (6, 17, 32); 677 (6); i (17); 1959 (32); 1963 in part (2, 38); 4386, 4765 (32); 6111 (14); Bn (37); 649 E A Cusick, W. C. 1 31). Cuthbert, A. i. i 268 (32); 858 (2). Davidson, A. — (1 Davis, J. 16 (19); 73, 950 (15); 1202 (19); 127 4a (15); 1613 (19); 2414, 2691 (31); 345 $791, 2960, 2962 (19); 3192 (16); $227 (19); 3453 (15); 3565, 8578 (16); $587 (31); 3588 (17); 9589 (19); 3590 (16); 3591 (15); 3817 (16); 3852, 4450 (15); 4562 (17); 4461, 4487 (19); 4565 (15); 6870 (19); 8458 (15); 9039 (31); Day, M. A. 79, 147 (15); 149 (16). Deam, C. C. — (19); 86 (43); 1272 (31); 6180 (5); 13619, 20857 (17); 20857a (19). Deam, Mrs. C. C. 1799 (32); 1831 (35). Dean & Thomas, 4874 (16). Deane, W. — (15). Dehesa, M. P. 1561 (11). Demaree, D. 3353, 3371 S vae (19); 6411 (35); 7589 (25); 7685 (48a). Dewart, F. W. 15, 18 Wen d (17); ?4 (35). Deweys, Mrs. F. E. — (8). Dewhurst, B. L. — (31). Dixon, R. A. 473 (24); 561 (40). Dodge, C. K. 154 (33). Donnelly, L. L. — (40a). Drummond, T. — (14, 24, 35); 253 bis (24). Drushel, J. A. — (32); 2762 (35); 2842 (24); 3659 (35; hi D; 6351 (16). 6, 496a (43) E (id. H. — (19); 8137 (31). Eames, E. H. & McDaniels, L. H. 4870 (15). Earle, F. S. 84 (40); 387 (27); 427, 606, (25); 619 (27); 638 (41). Earle, F. S. & Baker, C. F. — (31, 32). Earle, F. S. & E. S. 236 (41); 361 (31); 387 (27); 526 (40). Earle, F. S. & paar — (32). Eastw : 3, 41); 25 (31); 41 (16); ps (19); s (23); 8468 (42). Eaton, D. C. — (15, 16, 17) Edwards, — (20). Eggers, H. F. A. von, 1828 (6). Eggert, H. — (6, 8, 15, 16, 17, 19, 24, 346 25, 29a, 31, 32, 33, 35, 40, 41, 42, 43a, 44, 48). Eggleston, W. W. — (17); 1581 (15); 4430 (17); 4837 (19); 4841 (15); 5237 (17); 6644 (41); 12240 (35); 15224 (19); 15551 (16); 17106 (20); 1?231 (42). Ehlers, J. H. 627 (31); 642 (19). Ehrenberg, C. 131 (46); 713 (10). Ellis, C. C. 17 (42); 221 (31); 258 (20). Elmer, A. D. E. —, 324 (31); 537 (16); 3846, 4045 (22); 4950 (23). Emig, W. H. 107, 365 (19); 395 (40); 401 (48); 415 (35); 716 (15); 787 (19); 788 (40 Greeters, G. — (16, 17, 19, 23, 31, 35, 41); 334 (35); wer an; 337 (16). Engelmann, H. — Epling, C. C. 5445 - 6134 (35). Ervendberg, L. C. 153 (10); 236 (34). Evans, W. H. — (43b). Evermann, B. W. 970 (19). Eyerdam, W. J. 201, 432 (14). Ezell, A. 5699 (8). Fassett, N. C. & Hotchkiss, N. 3322 (16). Fawcett, H. S. /2 (31). Faxon, C. E. — (16 Fendler, A. — (31); 586 iu 587 (31); 594 (25); 597 in part (19, 20). Fernald, M. L. — (16, 31); 296 (16). Fernald, M. L., Hunnewell, F. W. Long, B. 10262 (15). Fernald, M. L. & Long, B. 10268 (16). Fernald, M. L. & Parlin, J. E. 928 (15). Fernald, M. L. & Pease, A. 8. 25247 (16). Fernow, B. E. — (47a). Ferris, R. S. & Duncan, C. D. 2474 (41); 2607 (20b); 2647 (42); 2786 (25); 3161 (34); 3268 (24); 3337 (8); 8518 (16) Fink, B. — (31); 251 (19). Fisher, G. L. — (11, 16, 35, 41, 46); 19 (19); 33, 40, 47 (35); 122 (24); 168 (5); 207 (43); 212 (24); 223 (43); 235 (29a); 282 (25); 319 (43); 820 (19); 33 7 (46); 460, 625 (1); 707 (35). — (35). Forrer, A. — (30, 40a). [Vor. 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN Forwood, W. H. 298 (31); 299 (19). Fowler, J. — (16). Fredholm, A. 110, 5136 (32); 6416 (14); 5804 (38). Freiburg, G. W. — (15, 19). Fritchey, J. A. 29 (22). Frost, W. D. — (19) Fuertes, M. 391, 1758 (14); 1771, 1856 (6). Furness, D. R. — Galeotti, H. 736 (11); 736 (43); 737 (11); 777 (46); 778 (T); 781 (6); 795A (11). Gallegos, J. M. 2342 (18 Gandoger, M. — (36) Garber, A. P. — (38). Garcia, P. I. 331 (40a). Gardner, J. R. 553 (19). Garesche, E. E. — (19). Gates, F. C. 9994 (19). Gates, F. C. & M. T. 10708 (16). Gattinger, A. — (19). Gershoy, A. 583 (6). Geyer, C. A. — (16, 40). Gibbes, L. R. — (2, 15). Gilman, M. F. 1120 (26a). Glatfelter, N. M. — (8, 15, 17). Gleason, H. A. — (19, 31); 377 (19); 718 (15); 1945 (16). Goldman, E. A. 42 (34); 292 (47a). Goldsmith, G. ^d 9 (15 Goodding, L. N. 37? (402); ; 207 (81); 251 (40); 334 (26a); 495 (40); 645 (47); 883 egy 952 (47a); oi 2312 (31). Goodm oJ. iteheock, C. L 1211 (40); 1297 w^ Pe (31). phe ee LT A. 90 (16). Graham Grant, M. L. pa 73 (16). Graves, E. W. 525 (1); 536 (40); 621 (82); 1691 (15); 1946 (40); 1947, 1994 (19). Green, T. — (17). Greene, E. L. — (23, 26, 42, 43b, 50); 45 (41); 77 (20); e e ` Greenman, 44, 245 (15); 398, 1273 (16); 1376 y 1877, 1379 (16); 1380 (15); 1382 (16); 1877 (15); 1953 (16); 1981, 3640 (17); 3766 (15); 3870, 4076 (35); 4125 (17); 4231 (35); 4412 (15); 4578 (17 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 347 Greenman, J. M. & Greenman, M. T. 28 (25); 56 (47a). Gregg, J. — (8, 29, 43); 11 (7); 26, 57 (43); 99 (29); 202 (34); 246 (29); 257 (43); 265, 276 (7); 355 (86a); 406 (7); 483 (31); 545 (29); 623 (31); 632 (43); 686 (31); 648 (43); 740 (29); 752 (34); 756 (43); 764 (5); 791 (29a); 823 (7); 1001 (11) Griffith, F. 3456 (40). Griffiths, D. 1595 (25); 1919 (41a); 2421 (47a); 8431 (26a); 3489 (47a); 4244 (41); 4757 (43) 4833 (40a); 5050 (19); 6190 (42); 5560 (31 Griffiths & Hunter, 18 (31). Griffiths, D. & Thornber, J. J. 159 (40a). Grimes, E. M. 3716, 8788 (17); 3794 (15). Groth, B. H. A. 16 (29a); 80 (40); 75 (8); 187 (24). Guidroz, H. 3 (15). Gurney, J. & Monell, J. T. — (40). Hale, T. J. — (8, 15, 17, 19, 24, 35); 245 (8) Hall, E. — (19, 31); 393 (31); 428 (48); 429 (25); 430 (29a); 431 (48); 432 (8); 438 (1); 484 (24); 435 (35). Halsted, G. 176 (15). Hammond, E. W. 322 (22). Hansen, G. 964, 1823 (22); 2026 (5). Hanson, H. C. — (48); A146 (43b); A147 (31); A148 (20); 322 (48); 551 (42); 619 v pud (26b). Hanson, H. C . E. A1021 (47); A1130 (26a). Harger, E. B. — (16, 17). Harper, R. M. 16 (40); 242 (14); 385 (17). Harris, J. A. C16375 (47a). Harris, W. 9132 (2); 9937, 11808 (14); Harrison, G. J. 4778 (26a); 4897 (14). Harrison, G. J. & Kearney, T. H. 3646 (47a); 5796 (26a); 6144 (11); 6689 a). Harshberger, J. W. — (1, 2); 41 (46). Hartman, C. V. 94 (11); 608 (26a); 880, 906 (40a). Hartweg, T. 174 = 175 (7); 176 (43); 177 (29); 1924 (22). Harvey, F. L. " n 62 (31); 1958 (19). Harwood, R. D. — ( Havard, V. — jp 27, 31, 33, 42, 44, 48); 97 (42); 197 (27). Hayden, F. V. — (15, 16, 40); 10 (31). Haynes, D. — (22). Hedgcock, G. G. — (15, 16). Heller, A. A. — (15, 31); 638 (19); 1388 (33); 1419 (8); 1732 (29a); 5785, 5919, 6778 (22); 14222 (16); 14290 (19); 14303 (31 Heller, A. A. & E. G. 3536 (42). Henderson, L. F. 4070 (31). Herrick, C. L. 704 (42); 715 (26b). Hertel, H. — (15). Hexamer, A. C. & Maier, F. W. — (16, 17, 35). Heyde, E. T. 120, 477, 580 (11); 610 (5). Heyde, E. T. & Lux, E. 3018 (11); 3019 in part (5, 11); 4370 (5). Hicks — (15). High, M. M. 52 vn 59 (48a); 76 (33). Hilgard, E. — (8, 32). Hillman, F. H. — (31 Hitchcock, A. S. — (2, 14, 15, 16, 17, 32, 35); 269 (14); 393 (40); 790 (35); 791 (17); 972 E 1129 (31). Hoffstetter, G. — (15). Holgate, W. R. — ie Holm, T. — (15). Holzinger, I. M. — (19, 26, 41a). Hooker, L. H. 5999 (34). Hough, W. 109 (43). Houghton, H. W. 2570 (48); 3622 (17); 3643 (35); 3646 (17); 4000 (19); 4015 (15 Hosen, H. D. 875, 1363 (17); 8276 (32). Howell, A. H. 64 (19); 111, 308 (40); 313, 362 (8). Howell, T. 174 4 (22); = (16); 1249 (22). Hoyeradt, Humboldt, A. p Bonpland, A. — (11); 4063 (36); 4066 (12 Hyams, M. E. — (32 Jaeger, E. C. — (47). Jermy, G. — (48); 84, 85 (48); 182 (29a); 183 (40); 184 (48); 203 (8); 209 (48). 348 Jesup, H. G. — (16). Johnson, A. G. — (15, 16, 19, 35). Johnston, E. L. 507 (16); 986 (41) Johnston, I. M. — (14); 1320, 1608 (22). Jones, M. E. — (7, 8, 11, 16, 19, 20, 22, 23, 25, 20a, 40, 40a, 42, 43b, 47, 47a); 36 (47a); 122 (41); 372 (25); 457 (42); 487, 523 (16); 554 (29); 634 (41); 686 (43); 1487 (16); 2598 (22); 3801 (43b); 3901 (47a); 5478 (31); 6026 (20); 8963 (16); 23243 (48); 23244 (43); 25006 (11); 25480 (43b); 26220 (48); 26224, in part (26b); 26225, 26226 (29a); 26227 (43b); 26228 in part (40, 48); 26229 in part (8, 40); 26231 (43); 26232 (43 Jones, b Pd esu 180 (43b); 185 6a). Katzenstein, O. — (32). Keeler, H. D. — (14). Kellerman, W. A. — (17); 5825 (11). Kellogg, J. H. — (15, 16, 35); 496 (17); 498 (19); 1130 (17); 1131 (10); 1958 (35); 15276 (31); 15277 (15). Kellogg, W. — (38 Kelsey, J. A. 173 (10). Kempton & Collins — (11). Kendall, M. L. — (22). a G. G. — (15 nedy, P. B. 7032 ( on; 7033 (47a). Kerber, E. 255 (43); 311 (5). Killiam, O. L. 6933 (15). Killip, E. P. 3510 (5). z = S = Ez E A w Knowlton, C. H. — - (16). Korthoff, I. — (3). Kreager, F. O. 469 (16); 474 (31); 475 19). ( Krig, J. T. 3187 (14). Kuntze, O. 424 (43); 2109 (5); 23444 (28); 23738 sey 23811 (24). Lane, W. C. — 6). Lang, H. A. x 4). Langlois, A. ^i — (1, 24, 40); 123 (24). [Vor. 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN Lansing, O. E. — (16); 2805 (16); 2806 (15); 2810 (19); 2976, 3135 (17); 3164, (19); 3241, 3340 (16) Laybourne, W. A. — (15). Lea, M. C. — (6). Lee, D. W. 109 (48). Lehmann, F. C. 1603 (46). ond J. B. 902 (31); 2630 (16). , J. G. — (28, 47, 47a); 2857 (26a); 3075 (11). Leon, Frére — (6); 685 (14); 3910 (1). Leon, Frére & Edmunds, Frére — (6); 8719 (14). Leonard, E. C. 3939 (6). Leonard, E. C. & Killip, E. P. 534 (17). Leroy & Ruger — (31). Letterman, G. W. — (8, 15, 17, 19, 24, , 40). Lewton, F. L. 141 (34). Liebmann, F. M. 1/1308 (46); 11318, 11314 (34); 11316 (46); 11318 (13); 11328, 11329, 11330 (46); 11335 (10). Lighthipe, L. H. 599 (35). Lindheimer, F. — (8, 14, 16, 24, 35); 145 (24); 146 (40); 154 (24); 155 (8); 195 (15); 232, 289 (40); 294 (29a); 307 (40); 434 (48); 500 (292); 501 (48); 618 (14); 1072, 1073 (40); 1074 (29a); 1075 (48); 1076 (8); 1077 (14). Lloyd, C. E. 448 (11). Lloyd, F. E. 159 (43). Lloyd, F. E. & Tracy, S. M. — (6); 20 (3); 22 (14). Loomis, H. F. & Thackery, F. A. 910 (47a). Louis-Marie, Pére, 141 (16). Lumholtz, C. 445, 446 (36a). Lunell, J. — (16, 31). Lyall, D. — (31) Lyon, 6 (1 Lyonnet, E. — (7); 173 (30). Macbride, J. F. 200 (31); 304 (16). Macbride, J. F. & Payson, E. B. 781 (7); 850 (23). MacDaniels, L. H. 4873 (16). MacDougal, D. T. 73 (47a); 153 (7); 249 (20); 286 (31); 317 (40); 566 (16); 608 MacElwee, A. 760 (17); 873 (6); 882 (15). 1933] PERRY— NORTH AMERICAN SPECIES OF VERBENA M T K. K. 96 (8); 261 (19); 4218 (17); 4746 (15). del C. — (19). Macoun, J. — (16, 31); 1305 (17); 24268 (31 McAtee, W. L. 1953 (8); 2053 (19). McClatchie, A. J. — (14). McFarland, F. T. 92 (15); 109 (17). McFarland, F. T. & Anderson, W. 253 (19). McKelvey, S. D. 1710 (29a); 1756 (25); 2046 (26b). Manning, W. H. — (19, 31). Marble, D. W. 737 (2). Marsh, C. D. 14224 (42). Mathias, M. E. 545 (41) Mattern, E. S. & W. I. — (38). Maxon, W. R. 3808 (40). Maxon, W. R. & Harvey, F. L. 7951 (5). Maxon, W. R., Harvey, & Valen- nh 7356, 7478 (5). .R.& me E. P. 569 (2). Sera, E. A. — (40a, 47); 108 (43); 524 (31); 1013 (26a); 1116 (28); 1138 (41a); 1237 (29a); 1252, 1274, 1894 (40); 1456 (29a); 1887 (26a); 1904, 1914 (47a); 1918 (26a); 2278, 2486 o>) 2627 (11); 2677 (41); 2698, 9 (31). Fa C. D. — (38). Mell, C. D. & Knopf, E. C. — (81). Mellichamp, K. — (32). Meredith, H. B. — (6). Merrill, E. D. phy ox E. N. 776 (31). Merton, E. C. 2042 (26: Metcalfe, O. B. 126 42); 187 (81); 612 (26); 897 (31); 955 pud 1008 (31); 1090 (42); 1231 (41); 1 (27). Michener, C. A. & iden i5 T. 123 (23). Millspaugh, C. F. 866 (17). Mohr, C. — B 17, 24, 31, 32). Molby, E. 7224 (34). Moldenke, H. N. 148 (2); 212 (38); 219a, 549, 586 (37); 591, 599a, 972 (14); 1039 (38); 1091 (32); 1184 (3); 1330 (15); 1339 (16). Moore, A. H. 2569 (16); 2874 (14); 349 2939a (6); 2946 (2); 2947 (14); 2984 ; 4816 (16). Moore, J. A. & a J. A. 3001 (40); 3005 (29a); 3009 (41); 3277 (26b); 3299 (25); 3324 (42); 3611 (27); 3622 (41). Morales, R. 786 (11). Moron E+ a 7). Mueller, F. — (11); 325 (43); 766, 887 (11); 1209 (43); 1215 (29); 1319 (43). Mulford, A. I. — (31); 37 (42); 39 (20). Munz, P. A. 616 (15); 617 (16); 1149 (26a); 1231 (41a); 1249 (43); 1470, 607 (24); 2207, 6349, 6598 (31); m 9295 (22). z, P. A. rodar. I. M. 11289 (14). "unm P. A., Johnston, I. M. & Harwood, R. D. 4252 (47). Murdoch, J. — (40a). cos G. V. 601 (32); 1248 (14); 2470 Nea I G. C. — (33); 117, 118 (84). Nelson, A. 320 (31); 505, 538 (19); 1652 (31); 2258 (16); 7671 (31); 8245 (40); 8354 (20); 8489 (16). Nelson, E. . 196, 223, 242 (46); 1021 (11); 1105 (12); 1518, 1895 (45); 1943 (7); 3896 (36a); 3920 (43); 4424 (34); 4559 (48); 4577 (7); 4593 (28); 4604 (43); 4856 (40a); 6117 (40); 6161, 6271 (26a); 6628 (34); 6789 (41); 6858 (43). Nelson, E. W. & Goldman, E. A. 7123a (47a); 7425 (21). Nelson, J. C. 1429 (22); 1804 (16); 2669 (2 zn ess, H. — (24). dee cw — (11). Nichols, G. E. 163 (2). Nicolas, — (29). Nicollet's Northwest Expedition, — (40 idrico à A. 2685 (17). Nor — (6, 15, 17); 389 (15); pe Ne: E (19); 592 (31). 350) Nuttall, T. — (17, 40). Oakes, W. — Oersted, A. 8. 11084, 11324 (5). Olney, S. T. — O'Neill, H. — "s 986 ha 6309 (37). Orcutt, C. R. — (18, 31, 33, 47a); 909 (18); 1228 (31); - de 1302 (14); 1303 (472); 1546, 1549 (31); 2736 (2); 3321 (13); 3488 (11); 3551 (46); 3950 (12); 4045 (11); 4146 (46a); 5141 (47a); 5423 (10); 5542 (33); 5556 (25); 5657 (43a); 5717 (25); 5780 (33); 5792 (43a); 5867 (8); 6096 (41); 6111 (292); en (43b); 6184 (44); 6197 (41); 6235 (25). Ortega, J. G. p M; 4541 (47a). Osterhout, G. E. (15, 16, 20, 41); $8166 (40). Over, W. H. 2103 (40); 8177 (40); 14386 (19). eo L. O. — (15, 17). Pace, L. 22 (8). Pagel, L. E. 2208 (33). Palmer, Edward — (17, 26a, 41, 47, 50); 25 (29); 85 (43); 39 (34); 47 (40); 51 (36a); 62 (26a); 74 (43); 78 (8); 79 (43); 82 (29a); 90 (34); 135 (28); 141 in part (5); 141 in part, 158, 191 (coll. of 1898) (7); 191 (coll. of 1906) (36a); 200 (28); 268 (7); 281 (43); 295 (36a); S03 (29); 307 (47a); 308 (7); 809 (22); 310 (23); 312 (26a); 326 (48); 339 (coll. of 1876) (47); 339 (coll. of 1896) (11); 33914 (26a); 1g (23); 342 (22); 345 (43); 356 (7); 864 (11); 397 (14); 401 (43); 456 (28); 6?7 (50); 911 (28); 1040 (14); 1042 (7); 1043 (8); 1044 (29a); 1046 (33); 1047 (29); 1048 (31); 1050 (41); 1051 (34); 1052 (36a); 1156 (11); 2014a, 2019 (5); 2037 (10); 2038 (25); 2039 (33); 2040 (25); 2055 (43); 2057 (11); 2529a (22); 2699 (31); 6433 (38). Palmer, E. J. 197 (16); 198 (15, 16); 199 (19); 200 (17); 304 (35); 454 (31); 559 (35); 1085 (16); 1747 (33); 2980 (19); 2981 (17); 2996 (35); 8319 (17); 4229 (15); 4232 (17); 4840, 4977 (48); 6044 (8); 5185, 5407, 5682 (35); [Vor. 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN 5769 (17); 6106 (31); 610? (19); 6510 (40); 7001, 7071, 7124 (35); 7369 (8); 7566 (24); 7821 (40); 8000 (15); 8045, 9098 (40); 9111 (8); 9183 (29a); 9278, 9339 (48); 9485 (8); 9559 (32); 10002 (29a); 10003 (48); 10037 (8); 10568 (16); 10692 (14); 11715 (8); 11843 (48); 11979 (35); 12001, 12038 (1); 12056 (35); 12496a (31); 13176, 18281 (48); 13282 (25); 18283 (29a); 18512 (39); 13562 (48); 13563 (29a); 18664 (8); 13730 (25); 13861 (31); 14988 (35); 15262 (17); 24616 (35); 27209 (40); 27462 (14); 28738 (16); 29303 (2); 29333 (35); 29535 (29a); 29804 (35); 30523 (25); 30672 (42); 30791 (26b); 31117 (31); 81158 (41a); 31192 (41); 33319 (35); 33367 (48); 33389 (40); 33537 (25); 33583 (42); 33592 (33); 33604 (8); 33605, 33646 (29a); 34065 (26b); 34240 (41); 34766 (17); 355235 (32); 37627 (19). Palmer, W. — (16 Pammel, L. H. — (17, 19); 78 (16); 88 (19); 187 (22); 272, 1699 (17); 1806 (16 Pam nai L. H. & Blackwood, R. E. 3589 (31); 3638 (16). Parish, S. B. — (14, 47a); 2819 (22); pend 7149 (14); 10288 (31); 11590 ors W. F. 196, 197 (47). Parish, S. B. & W. F. 969 (22); 1043 (14); 2171 (31); 11148 (14). Parker, C. F. — (6, 17, 31). Parlin, J. E. — (16). Parry, C. C. — (47a); 167 (40). Parry, C. C., Bigelow, J. ia Wright, C. & Schott, A. — (26b, 31, 40, 44). Parry, C. C, & AMA J^ e p^ (22) Parry, C. C. & Palmer, E. 717 (7); a 719 in part (46a); 719 in part (43); 720 (36a); 722 (28). Patterson, H. N. — (19). Patzky, — (28). Pease, A. S. & boy B. 22350 (16). Peck, M. E. 8 Peebles, R. H. po a ; 6756 (47a). 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA Peebles, R. H., Harrison, G. J. & Kearney, T. H. 3790 (26a); 4243 (31); 4549 (40). Peirson, F. W. pi (7). Penard, E. 348 (41). Pennell, F. W. 4204, 4332 (1); 5444 (40). Perkins, J. R. 1024 (1); 102 (2). Peterson, N. F. — ( Phelps, O. P. 799 (18); 800 (15). Pilsbry, H. A. — Piper, C. V. 6160 d Pitcher, — (35). Pittier, H. 414 (29); 1909 (11). Plank, E. N. — (19). Plaskett, R. A. 142 (22). Plummer, F. G. — (42 Pollard, C. L. — (38); 1191 (14); 1312 17). Pollock, W. M. — (16). ( ; Pringle, C. G. — (11, 14, 25, 26a, 28, 40, 47a); 54 (28); 270 (26a); 1117 (43); 1599 (7); 1948 (10); 2228 (34); 2813 (11); 2927 (46); 3467 (41); 3551 (46a); 4180 (46); 4769 (12); 4784 (29); 4877 (5); 4892 (11); 5715 (7); 6539 (28); 6908 (36); 7590 (28); 7591 (43); 8534 (7); 9135 (28); 9312 (11); 9318 (43); 9529 in part (7, 11); 11091 (43); 11092 (36); 11093 (5); 11843 (34); 13157 (41); 13158 (43); 13159 (28); 13434 (51); 18597 (12); 13792 (46). Purpus, C. A. 13? (46a); 195 (49); 467 (29); 1094 (26b); 1095 (43); 1433 (36); 3406 (13); 3407 (43); 4524 in part (31, 43); 4974 (36a); 5451 (10); 5515 (29); 6061 (47); ia m Quaintance, A. L. — Rafinesque, C. S. — ap. Randolph, L. F. & F. — (32); 52 (37); 474 (15); 587, ib (6); 688 (1); 1004 (3); 1012 Ravenel, H. W. — (15, 31, 32). Read, A. D. — (14); 19 (20). Reade, J. M. — (1). Redfield, J. H. — (15, 16); 522 (19); 6430 (17). 351 Redfield, R. 4 (11). Rehn, J. A. & Viereck, H. L. — (42). Renson, C. 175 (11); 291 (5). Reverchon, J. — (15, 19, 35, 40); 42 (8); 118 (25); 782 (8); 734 (15); 735 (19); 736 (31); 737 in part (25, 29a); 738 (48); 739 (40); 740 (35); 834 in part, 1953 (25); 1961 (29a); 1962 (40); 1963 (48); 2116 (31); 2117 (48); 2118, 2532 (24); 2533 (35); 8902 (33); 3903 (29a); 4814 (25). Riddell, J. L. 1268 (8). Ridgway, R. 2431 (19); 2831 (15). Riehl, N. — (16); 9 (31); 185 (15); 196 (19); 456 (17). Robinson, B. L. — (19); 26 (32); 94 (2); 118 (6); 127 (2); 134 (15); 153, 196 (16); 559 (15); 672 (16). 11). ; 191 (32). Rose, J. N. 1677? Hew 1763 (7); 2485, 2769 (43); 11968 (47a); 16155 (50). Rose, J. N. & Fitch, W. R. 17105 (40); 17106, 17503 (31); 17504, 17801 (41). Rose, J. N., Fitch, W. R. & Parkhurst, 17717 (20). ose, J. N. & Hay, R. 5310 (12); 5488 (28); 5556 (36); 6049 (46); 6212 (29). Rose & Hough, = : 6). se, J. N. & Pai , 4. H, 6481 (43); 7741 (29); 7742 io 7799 (1). Rose, J. N., Painter, J. H. & Rose, J. S. 8382 (7); 8495 (28); 8496 (11); 8668 (36); 8753 (7); 8823 (46a); 8971 (29); 9077, 9151 (5); 9162 (29); 9503 (43). Rose, J. N. & Rose, J. S. 11398 (43). Rose, J. N., Standley, P. C. & Russell, P. G. 12451, 12934 (7); 12745 in part (48); 12840 (47a); 13130, 13422, 13447 (7); 18449, 15176, 15181 (47a). Rothrock, J. F. — (47a). Rovirosa, J. N. — (5). von Rozynski, H. W. 13, 17, 135 (29a). Ruano, J. M. 332, 03 (5). Rugel, F. — (15, 31, 32, - 121 (6); 127 (17); 156 a 305 (38 Ruger, M. — (17, 31). 352 Runyon, R. 628 (33); 629 (29a); 630 und E (15 ak H. — (20, 42); 124 (46); "T Ay 336 (31). Ruth, A. — (6); 107 (40); 108 (8); 109 (31); 110 (48); 495, 518 (31); ?31, ?40, 765 (17); 788, 833 (15); 1289 (25). Rydberg, P. A. — ; 154 (17); 932 (19); 934 (31); 935 (40); 1422 (19); 1515 (10); 1716 (15). Rydberg, P. A. & Carlton, E. C. 7043 (31). Rydberg, P. A. & Garrett, A. O. 9201 28). Rydberg, P. A. & Imler, R. H. 22 (35); 120 (17); 379 (35); 433 (19); 482 (15); 1008 (31); 1212, 1253 (40). Rydberg, P. A. & Vreeland, F. 5676 (41). St. John, Mrs. O. — (20). Safford, W. E. 216 fs 1221 (34); 1398 R. — (43). Safford, W. E. & Mosier, 210 (37). Salas, G. 32 (5). Salazar, F. — (5, 7, 10, 12, 28). Sandberg, J. H. & Leiberg, J. B. 340 (31). Sandberg, J. H., MacDougal, E. XT. & Heller, A. A. 264, 972 (31). Sanford, S. N. — (19). Sarvis, J. T. 122 (31). Savage, E., Cameron, F. E. & Lenocker, J. E. — Schaffner, J. G. 716 (362); 717 (43); 718 (5); 719 d 720 (28). Schiede, C. J. W. 8 Schoenfeldt, L. Pie (T. Schott, A. — (33, 40). von ^n H. — (6, 15, 16, 31, 40, 48). Schuchert, C. — (17). Schuette, J. H. — (15, 19, 3 Schulz, E. D. 255 (20); je (33); 679, 701 (24); 766 (8); 767 (33); 845 (43a). Scott, W. — (15). Schumann, W. 1071 (43). Seaton, H. E. 7 (7); 27 (5); 150 (43); 3822 (29); 391 (43). Seler, C. & E. — (27); 22 (28); 722 (10); ANNALS OF THE MISSOURI BOTANICAL GARDEN [Vor. 20 783 (36); 724 (13); atb 1379 (43); 4194, 4347 (13); 4560 (15). Sesse, M. & Mocino, J. M. 99 (51). Seymour, A. B. —, 48, 49 (17); 50 (8). Seymour, F. C. 511 (16); 601 (15); 1324 (16). Seymour, A. B. & Earle, F. S. 118 (32). Seymour, A. B. & obe — (81 Shannon, W. C. 3638 (11). Shear, C. L. 38 nit 191 (19); 202 (16); 3577? (20). Sheldon, E. P. — (31, 40); 26 (35); 38 (31); 48 (40); 49 (17); 8202, 8856 (31); 11167 (16). Sherff, E. E. — (19); 333 (35); 1649 (16); Short, C. W. — (17, 19, 35). Shull, G. H. 156 (16). Singer, J. W. 178 (17); 308 (15). Sintenis, P. 1074, 2010 (14). Skehan, J. — (14, 32); 20 (27); 46 (8); 4r (40); 109 (8). Small, J. K. — (6, 9, 15, 16, 17, 19, 31, 32); 2100, 2124 (37); 4020 (14); 5707 (3); 8469 ve 8509, 8594, 8599 (37); 8686, 8732 (38). Small, J. K. & Qe. J. J. 1072 (14); 1077, 2994 (37). Small, J. K., Carter, J. J. & Small, G. K. 38311 (3 Small, J. K. 2 Heller, A. A. — (9); 432 "e 484 T E p Small, G. K. 4337, 4341, ^ 4520 (14); 6422 (37). Small, J. K. & Wherry, E. T. 11774 (8); 11781 (40); 11818 (1); 11824 (33); 11826 (29a); 11834 (48a); 11901 (34); 11903 (33); 11968 (40); 11969, 11980 (29a); 11998 (8); 12010 (33 ie A K. & Age P, 1961 (37). Smith, s n part (43); 222 (5); "ig S e A (48): 173? (11). Smith, C. P. 3186 e Smith, E. A. — (35. - Smith, E. C. — (41). Smith, H. H. 5645 (19); 5654 (16); 5670, 5923 (17); 5952, 6028 (19). Smith, J. D. — (17, 31, 32, 35). 1933] PERRY— NORTH AMERICAN SPECIES OF VERBENA Smith, J. G. — 464 (46). Smith, L. C. 27 (7); 348 (45); 412 (29); Spencer, M. F. 971 (7); 994, 1164 (22); 1414 (7). Stabler, L. M. — (15, 16). Standley, P. C. — (27, 38, 41a, 42); 4223 (20); 4458, 4549 (31); 4927 (20); 4951 (41); 5136 (31); 6065 (20); 6268 (31); 8239 (20); 8367, 9171 (35); 9336 (17); 9362 (31); 14224 (20); 20617 (5); 21831 (11); 21486 (5); 22414 (11); 22842 (5); 23083 (11 23326, 24404 (5); 24443 (11); 32652, 33293 (5); 40625 (42); 40686 (27); 55944 (11); 56082 (5). Standley & Valerio, 45430 (5). Stanfield, S. W. — 2 24.) Stanford, E. E. ?21 (16). Stearns, E. S. ? PN 104 (41); 542, 358 “we (20 Steele, E. S. 106 (16); 314 (17). Stephenson, B. C. — (15 Stevens, G. W. 1 4 (40); 28 (48); 109 (85); 391 (31); ys (40); 484 (41); 680%, (31); 699 (40); 1034 (19); 1108 (40); 1192 (31); 135414 (19); 1673, 1742 (15); 1791 (16); 1821 (15); 1865 (19); 1980, 2080, 2299 (17); 2340 (35); 2850 (40); 2889 (31); 2901 (16); 2930 Stevens, W. C. — (17). Stevens, W. H. 28 (48). Stevens & Hess, 4260 (14). Steyermark, J. A. 292 (17). Stone, W. — (46). Stratton, R. 10 (40); 11 (48); 62 (8); 158 (15); 249 (19); 317 (40); 400 (16); 403 (19); 449 (40). Street, M. H. — (35). Studhalter, R. A. 1106 (48). Sturtevant, E. L. — n 16, 35). Taylor, A. 4244 (2 (2). 353 Taylor, K. A. — (32). Tejada, R. 57, 72 (11). Thackery, F. A. 348 (47a). Thackery & Leding, 1100 (47). Tharp, B. C. 667, 668 (24); 838 (35); 1361 (29a); 1362, 1364 (48); 1826 (34); 1871 (48); 2558 (2); 2818 (8); 3166 (24); 3682 (27); 5538, 5610 (33). Thompson, C. H. 154 (19). Thompson, H. 42 (31); 4412 (22); 4880 (31 Thone, F. H. 50 (15); 88 (19). Thornber, J. J. — (16). Thurber, G. — (6, 15); 143 (42); 446 (26a); 555 (7); 769 (48). Thurow, — (1). Tidestrom, I. 872 (26a); 898 (47); 11418 (15); 11551 (16); 11572 (15). Tindall, C. W. — (17, 3 Tonduz, A. 488 (11); 627 (5); 628 (11). Torrey, J. 154 (15); 416, 417 (22). Toumey, J. W. — (40a, 41a, 43b, 47); 805a (47); 305% T 306 (25). Townsend, C. & Barber, C. N. ed TU 192 do. Tra M. — (8, 14, 15); 4981 (32); 0050 jm 6652 ad 7588, 7996 (8); 7999, 8000 (40); 8001 (31); 8037 (15); 8706 (3); 8707 (35); 8708 (24); 8709 (8). F. S. 30, 41 (25); art (42, 43b); 162a (42); 178 (48). dA & mm Lloyd & Tracy. Trask, B. — ; Trelease, W. — (8, 15, 16, 17, 19, 29a, 31, 35, 40, 43a); pe (48); 101 (33); 494 (31); 718 (17); 714 (35); 715 (31); 716 (16); 717, 718 (19); 719 (15); 1161 ) (35). von Tuerckheim, H. 77 651 (11); 904 (5); 913 (11); 1068 (46); 8442 (11). Tweedy, F. —, 113, 246 (29a). Tyler, F. J. — (40 i? Umbach, L. M. — (19). Van Sickle, W. M. — (19, 31). Vasey, G. — (17, 22, 31, 35, 41, 47a); 468, 512 (31). [Vor. 20 354 ANNALS OF THE MISSOURI BOTANICAL GARDEN Velasco, L. V. 8848 ( Eu hen (11). Wislizenus, A. 150 (26a); 325 (29a); Victorin, M. 3/24, 1 15). 483 (42); 525 (A1). Viereck, H. L. — (4 E Wislizenus, F. 280 (17); 282 (15); 284 Visher, S. S. — (15); 1507 (19); 2028 (31); 285 pee 1214 (6). (31); 2132, 2177 (40); 2263 (15); Wolf, C. B. 2 8359 (31); 4439 (16); 4450 (19). Woodson, R. E 198 (35). Vreeland, F. K. 636 (20); 804 (43b). Wooton, E. O. — (20, 25, 26, 26a, 27, ui n F. 974 (41a). 31, 41, 41a, 42, 43a, 47, 47a); 187 (27); Wallac — (14). 208 (20); 363 (41); 364 (42); 409 Ward, L F. — (6, 24, 31, 35, 40, 41); (31); 642 (41); 646 (26); 2831 (31); 11 (20); 256 (31). 2835 (41a); 2836 (31); 3849 (43b); Warner, S. R. 43 (24). 8852 (26b). Watson, S. 823 (31). Wooton, E. O. & Standley, P. C. 3330 augh, F. A. — (15); 168 (35). (31). Weatherby, C. A. 5076 (16). Worthen, G. C. — (5). Webb, R. J. 5441 (16). Wright, C. — (15, 42); 454 (31); 4 Webber, H. J. — (16, 19). (39); 567 (31); 1496 (25); 1497 m —_ M. 165 (40); 174 (16); 197 (40); 1498 (16); 21498 (44); 1499 (31); 6 (31). 1500, 1501 in part (48); 1501 in part Sn O. E. — (40). (44); 1502 (40); 1508 in part (39, 40a, Wiggins, I. W. & Gillespie, D. K. 3977 42); 1504 in part (40, 42); 3658 (6); (23). 3659 (14) Wilcox, T. E. — (11, 26a, 28, 40a, 47). Wurzlow, E. C. — (3, 5). Wilkinson, E. H. — (8, 26a, 43); 8 (489); Young, H. A. — (17). 9 (40); 108 (48). Young, J. P. 737 (38). Williams, E. F. — (15, 16). Young, M. 8. — (8, 25, 27, 42, 48); ?? Williams, T. A. — (40). (8); 112 E 29a); 114 (48); 662 Wilson, N. C. 99 (47). (1); 1708 (26). Wilson, P. 161 (17). Zuck, M. — y 43b, 47). INDEX TO SPECIES New species, varieties, and combinations are printed in bold face type; synonyms, in italics; and previously published names, in ordinary type. Page Page Anonymos caroliniensis........... 316 UDerOWÍa. coord xen 251 Billardiera................ ees 251 Verbena.......... uses 251 € "RERE EEE A 36 («oS AED 316 JC MTM 257 Buchnera canadensis,............ 316 Alopecurus.... 0.0.00. e cece 285 Glandularia. ...... eer 251 ambrosifolia.................. 326 BINE in bs nh revu» STARS 316 forma eglandulosa .......... 328 bipinnatifida... a...an ee 323 amoena. r E EE AA 341 caroliniensis...... lessen 316 Andrieuxii..................05 333 Helleranthus quadrangulatus....... 313 angustifolia PIXWRAG PAGG EAT ERE 282 ee ee TT PETIT 251 NONE ur EFVEKEXE: 300, 337 Phryma ici i EG d USYE T 309 Aubletia....... e elles Pep PRETT IEEE 251 var. Drumm a, ss. 316 1933] PERRY—NORTH AMERICAN SPECIES OF VERBENA 300 Page Page PAAR 0S eee ee ees 342 gracilis, 6.0. <5 2% ee ed 300 bipinnatifida.................. 323 randi loras cisco ee ores me 341 var AMODA 60 dtd as 325 imr LER ie oS see CDM ER 342 USER NOLES fee tia eee oe ate ee 208: 3HB8l601. oa aera ee ee 265 DOUATIGNAIS sere Ae coches otra 254 Ha mae papaya Soak eee RE 257 bonariensis hastata: «2. sau cee E TA 278 VACS OT OUE renes a 257 hastata aa A E E 253 var. oblongifolia.............. 267 DIACOR OU E E N 304 VAT PONCO a 640 Gas 278 eos UR ENG box 304 TINS 0800. o eua 278 var 12) Poner EE 304 var. pinnatifida. oss s 278 var. ERAS No ATA AN 304 Mreta S E A E A ee 268 DFASHIODSIBS e a aA eso outs eph a 255 lax oer ONOCN. MI OOTOLO 0e o 338 s MUROS ACTEUR 342 ONE se sin aii ss RR 315 ACU. ooo NET te ee 338 fe gatos DD et ee 342 var. Ehrenbergii. ............ 318 UOTE eve CUM ERAN 316 BUDEP: INO, oo oi ees 318 VAT. T0860... eene 316 VET LOAIOCT ech no 316 LASTOGLOCIYS = as Ec 290 CATIOBDOIH seers Soe eee 301 leucanthemifolia...........0055 265 ar. Roemeriana............ 302 BUA, E ee eee 340 CUREHNCOH S Eder vd trs MER RR. 304 ey renee TT T 257 VAT: n60m6éxtcana., .... s ec eae 296 litoralis COP GC 0 EAE oe DN a 257 VAL Draschensis | Sd REEL 255 Den ecu me oo Sere e 309 VAT CO Gcasqna.. ee ck ere 257 ORTUDDUSR NS io conet dta 268 lit is COME. eee i 268, 309 B leptostachya. .............. 257 orma or var. pu "ee 2 a Dycnostachyja ..«.. b. eee 257 forma or var. recía. .......... 271 DUM. eis a A 315 COE derer verre ya 309 D ADMIRE TRITT 272 alibi p MAR PT 330 HOG Loser ee eme esta 293 var. longidentata...... ...... 331 Mae DONA. acosa does ris 288 ee a e ee 332 acDougalii mut. rosella. ...... 288 DAR E E ME 304 BEDA 6.43. epo EX ran 289 cuneiolua. v 286 IBS nek, PEPEE TE ea E E 320 delicatula.. aena a aaa 342 AE ar ai i AREA A ESTO 342 delticola 22 n 9 E E 314 menthaefolia. .... -iie e arrenak 263 aunn Ea POE E N A NE 275 CD RENE r 268, 286 dOTmATgensis- oos Se ee eee 262 SROVONCRAS « voveo bas a 31 DONON s ons oe ee re 316 TOOMGENIGANG . ov Fe os nests 296 Ehrenbergiana................ 267 var. Wirtella. : ....: 2.9 985 298 [MISC EDD EOS LS 318 warszylopoda....4.. 2.0 297 var. asperata................ 319 UDIIPNg. or ee eee 315 CHO HORRORE EET EET 254 DNENGE Lees RUE e 315 PBELIGRIGN ae erue sd se 341 OIDODAIDU o ore que EP eT 262 Aldi M M E 334 officinalis var. hirsuta. ......... 296 GOOG DEEP Eee ees oso orn o 335 DOFCULHanA - cuve. ud o rs 284 var. eile Ng 337 DREN, @ os 5:- - se vas STU 278 356 Page var. pinnatifida.......... LL. 278 DEUM ev rk a xao Ys 268 BINNEN. coxrkrrsa kv d Cao 268 paucifolia..............Luesue. 342 WOON PRIMETIME 299 pinnatifida chee awed sad eee ees 278 ee LE T ee 294 polystachya . . 6... ce cee ee 268 prostrata... 0... lessen 290 C CP PEERS POPE EPOR 332 aiino ee ee AARON 332 WER ONNE ERES EEE T TE E 338 ai forma albiflora.......... 313 quadrangularis... .......ssc.... 254 soci ia ETELE EP TENER 313 ^j líí MÀ 332 "dO: AUOD PEPCPRRATTPPMSE RM 271 TOMOA, onc xr er ea 300 jn. PEE 342 Ju S TERRE TETTE ITILTI 286 HN LL cocco XE eens 252 TAR Loo keh x4 xxv RERI 267 | a ry rrr eda 292 Roemeriana.........i sees. 302 PIPILL o cep TREE YEAR 316 xo ^ PROFERRE PLE TET 304 i rr 282 d EERTE TEET ET EEE E 272 Se ee ee ee T 272 penis TERE end ceed bred eaten 343 [Vor. 20, 1933] ANNALS OF THE MISSOURI BOTANICAL GARDEN Page BMRONA 5. 6 44 4s Pi ER RE 339 NND. ioo rr ne SAE ate rT 263 s; or^ AREE USC 282 ono MD 1730942351 ATEEN TT 256 EEEREN AT ET 262 ee ee ET 304 sion 93-8 32 E89 EARN XR REX BAS 285 gio e UIE S 12g E Eu 286 Pt Ms EEEE 286 forma arene in | him | 2 e m| a S516] 8 51A | e p = > 2 e^ [» e = z z 7 "d uonnpog ?|l$|2$|5|"*|*;$|"]|s|H|** |] BIE UALIT Hud SNVUSD NI SNOILOTOS SQO0IHVA I FIAVL JHL AO NOLLISOdIXOO 1933] WYND—NUTRIENT SOLUTIONS FOR ORCHIDS 365 per million, in table r1; and the growth data, in table rm. All solutions received 1 cc. of a M/200 suspension of ferric phosphate per liter prepared as described by Livingston (719). TABLE II COMPOSITION OF THE SOLUTIONS IN PARTS PER MILLION OF THE NUTRIENT COMPONENTS Ca | Mg K PO, | SO, | NO; | NH, quud Cl La Garde 360 | 98 287 | 698 | 390] 388 | 283 | 308 658 Crone 213 49 387 153 474 623 140 Knudson 169 25 112 137 462 525 137 225 Sachs 310 49 387 307 474 623 140 302 Pfeffer 135 20 186 140 78 545 123 Shive A 144 | 360 703 | 1710 | 1440 448 101 Schimper 291 | 42 260 | 235 | 168 | 1071 241 260 Shive B 208 | 480 563 | 1368 | 1930 | 646 145 Knop 169 | 25 134 | 175 98 | 525 118 58 Tottingham 399 | 347 700 | 1235 | 1390 | 1547 348 Hansteen- Cranner 202 | 60 129 | 314 | 1098 322 | 250 458 Zinzadze 49 | 17 62 47 | 112 51 15 23 58 TABLE III GROWTH DATA, BASED UPON THE AVERAGE OF 25 SEEDLINGS Solution Height in Diameter in| % Total pH at ps E ~ microns microns salt planting experiment La Garde 4243 1590 . 400 4.9 3.8 Crone 3760 1213 225 5.5 4.5 Knudson 3360 1145 . 200 5.0 4.5 Sachs 3258 1300 .300 D. 4.5 Pfeffer 2953 1135 .150 5.0 4.6 Shive A 2908 1250 .700 4.8 4.5 Schimper 2835 1223 .945 5.1 4.5 Shive B 2793 1213 .812 4.8 4.4 Knop 2360 1145 .162 5.0 4.5 Tottingham 1868 900 .820 4.9 4.5 Hansteen- Cranner No growth | No growth .295 4.9 3.9 Zinzadze No growth | No growth .052 5.1 3.7 The fact that no growth occurred on the solution of Hansteen- Cranner and of Zinzadze was probably due to instability of the or, 20 366 ANNALS OF THE MISSOURI BOTANICAL GARDEN pH of these solutions as shown by their high acidity at the end of the experiment. In this connection it is interesting to note that 7 § § 8 8 8 g 8$ F The relation of growth to caleium content. Zinzadze, Pfeffer, Shive A, Knop, Knudson, Hansteen-Cranner, Shive B, Crone, Schimper, Sachs, La Garde, Tottingham. | | 8 8 5 8 8 $ 8 8 § 8 Fig. 2. The relation of growth to magnesium content. Zinzadze, Pfeffer, Knop, Knudson, Schimper, Sachs, Crone, Hansteen-Cranner, La Garde, Tottingham, Shive A, Shive B. 1933] WYND— NUTRIENT SOLUTIONS FOR ORCHIDS 367 3500 4 2800 4 700 4 8 Š 8 8 S 8 8 8 Fig. 3. The relation of growth to potassium. Zinzadze, Knudson, Hansteen- Cranner, Knop, Pfeffer, Schimper, La Garde, Sachs, Crone, Shive B, Tottingham, Shive A $500 2.5 9 3$ 2 $134 73 Fig. 4. The relation of growth to phosphate. Zinzadze, Knudson, Pfeffer, Crone, Knop, Schimper, Sachs, Hansteen-Cranner, La Garde, Tottingham, Shive B, Shive A. [Vor. 20 368 ANNALS OF THE MISSOURI BOTANICAL GARDEN = > Y T - de cones xS aaa ver 3 5 3 8 3 P E § Po] Fig. 5. The relation of growth to sulphate. Pfeffer, Knop, Zinzadze, Schimper, La Garde, Knudson, Crone, Sachs, Hansteen-Cranner, Tottingham, Shive A, ive B. Shive 4200 | 3600 2800 | 2100 | 5 SO: @ E § P RP b Fig. 6. The relation between growth and nitrate. Zinzadze, La Garde, Shive A, Knop, Knudson, Pffefer, Crone, Sachs, Shive B, Schimper, Tottingham. La Garde's solution also was strongly acid at the end of the experiment, probably because of the greater growth of the numerous seedlings upon it. The use of ammonium carbonate by 1933] WYND—NUTRIENT SOLUTIONS FOR ORCHIDS 369 La Garde was probably for its buffer action against this unde- sirable change of acidity with growth. The effectiveness of the concentration of carbonate used as a buffering agent was tested by preparing La Garde's solution with and without carbonate and then comparing the titration curves obtained with N/100 hydrochloric acid. In both cases, the titration curves were $500 2800 2100} 1400[ i 5 : 5 ;_—, 8 8 2 3 3 3 3 Fig. 7. Relation between growth and ammonia. Zinzadze, Knudson, La Garde, Hansteen-Cranner. identical. The solutions were also analyzed for carbonate after autoclaving, but no positive test could be obtained. The auto- claving at 20 pounds pressure for 20 minutes at the initial pH necessary (4.25) undoubtedly destroyed the small amount of carbonate present. Examination of table 1 shows that the best growth occurred on the solution of La Garde, followed by that on Crone's and [Vor. 20 370 ANNALS OF THE MISSOURI BOTANICAL GARDEN Knudson’s. While the seedlings on La Garde’s solution were conspicuously larger than those on Knudson’s solution, the Knud- son seedlings showed a definitely superior root development. The Crone seedlings were actually larger than those on Knudson’s solution, but they did not appear to be so green. Of all the solutions tested, we would regard those of La Garde and Knudson to be the most satisfactory. 4200 | 3600 | 2100 } n Lond [2d to to CA CA o o 3 9 S 8 a Fig. 8. Relation between growth and total nitrogen. Zinzadze, Shive A, Knop, Pfeffer, Crone, Sachs, Shive B, Knudson, Schimper, Hansteen-Cranner, La Garde, Tottingham. The graphs indicate the comparative growth in relation to the concentrations in parts per million of each ion. It is apparent that the quality of the solutions is not related to a specific amount of any one ion. The author regards the superiority of La Garde’s solution to be due to the particular complex of nutritional factors, and not to the specific effect of any particular ion. The nature of its superiority might well be a favorable condition of permea- bility of the cells produced by chemical means not yet understood. : 933] WYND—NUTRIENT SOLUTIONS FOR ORCHIDS 371 4200 3500 2800 2100 1400 700 a g B 3} à 85 & 1n TEM Fig. 9. Relation of dia to chlorine. Zinzadze, Knop, Schimper, Sachs, Hansteen-Cranner, La Gar | | à E P o o 2 £ ost’ f Fig. 10. Relation of growth to per cent total salt. Zinzadze, Pfeffer, Knop, Knudson, Crone, Hansteen-Cranner, Sachs, Schimper, La Garde, Shive A, Totting- ham, Shive B. In all figures, the vertical distance represents the height of the seedlings in microns, and the horizontal distance represents the [Vor. 20, 1933] 312 ANNALS OF THE MISSOURI BOTANICAL GARDEN concentrations in parts per million of the ions indicated. The vertieal lines representing the growth on the various media are listed from left to right. CONCLUSIONS 1. Seeds of Cattleya Trianae Linden & Rchb. f., were germinated and grown on a number of published nutrient solutions. The best growth was obtained on La Garde's solution, which was closely followed by that obtained on the media of Crone and of Knudson. 2. No growth was obtained on the solutions of Hansteen- Cranner and Zinzadze. This was probably due to the unstable pH of these solutions. 3. The solutions studied differed in so many factors that the effects of any given species of ion were modified and obscured. Hence the nutritional value of the various solutions may not be interpreted as the effects of the concentrations of specific ions. This is clearly illustrated by the graphs. The author wishes to thank Dr. E. S. Reynolds, Plant Physi- ologist, Missouri Botanical Garden, for his coóperation, and also Dr. G. T. Moore, Director of the Missouri Botanical Garden, for a generous supply of orchid seeds. BIBLIOGRAPHY Benecke, W., and Jost, L. (24). Pflanzenphysiologie 1: 135-137. Jena, 1924. Duggar, B. M. (24). Plant physiology, p. 145. New York, 1924. Knudson, L. (22). Non-symbiotie germination of orchid seeds. Bot. Gaz. 73: 1-25. 1922. La Garde, R. V. (29). Non-symbiotic germination of orchids. Ann. Mo. Bot. Gard. 16: 499-514. 1929. Livingston, B. E, (19). A plan for coóperative research on the salt requirements of epresentative agricultural plants. Baltimore, 1919. maine D. T. (01). Plant physiology, p. 224. New York, 1901. Shive, J. W. ('15a8). A study of physiological balance in Aus media. Physiol. Res. 1: 327-397. 1915 , (15b). A three salt nutrient solution for plants. Am. Jour. Bot. 2: 157-160. 1915. Tottingham, W. E. (14). A quantitative chemical and physiological study of nutrient solutions for plant cultures. Physiol. Res. 1: 133-245. 1914. Wynd, F. L. (33). The sensitivity of orchid seedlings to nutritional ions. Ann. Mo. Bot. Gard. 20: 223-237. . Zinzadze, S. *. (26). Eine neue Náührlósung. Ber. deut. bot. Ges. 44: 461-470. Annals of the Missouri Botanical Garden Vor. 20 SEPTEMBER, 1933 No. 3 THE FOLIOSE AND FRUTICOSE LICHENS OF COSTA RICA. I CARROLL WILLIAM DODGE M ycologist to the Missouri Botanical Garden Professor in the Henry Shaw School of Botany of Washington University While there exists much literature on various groups of lichens from various parts of tropical America, at present I know of no recent work covering the whole area or that of any constituent country whereby the traveler may identify the lichens he meets, or the student in an herbarium of the temperate zone may readily determine any considerable portion of the miscellaneous material sent in by collectors. It is hoped that this work may furnish such a manual, although no one realizes better than I the huge amount of monographie work necessary before a book approach- ing completeness and accuracy can be published. I first came under the spell of the rich flora of Costa Rica during a two-weeks visit in the summer of 1925 and at that time formed the resolution to spend my first sabbatical year in that country in the study of its fungus and lichen flora. During the subsequent years much material in addition to my own col- lections came to me for identification. Plans were matured and in early September, 1929, my family and I landed in Costa Rica and made our headquarters in San José, the capital of the country. In the latter part of September I was joined by Mr. W. Stephen Thomas, a student at Harvard, who accompanied me on most of the collecting trips until the middle of March. Mrs. Dodge also spent some time in the field and much is owed to her keen observation and helpful care of the collections. Even the small daughter, at the age of four, added a number of specimens Issued October 6, 1933. ANN. Mo. Bor. Garb., Vor. 20, 1933 (373) [Vor. 20 374 ANNALS OF THE MISSOURI BOTANICAL GARDEN not otherwise easily obtained, through her skill in climbing guayabos (Psidiwm guajave L.) which were too small and too brittle to be climbed by an adult. We left Costa Rica in June, 1930, and after a brief stay at the Farlow Herbarium to sort over material and give tentative determinations I took the more puzzling specimens to Europe where the months from August to December, 1930, were spent in the larger herbaria. Very many persons and institutions have contributed to the success of this undertaking both at home and abroad, and to all ] tender my most hearty thanks in grateful acknowledgment of their many kindnesses. For financial assistance, I am grateful to the John Simon Guggenheim Memorial Foundation which appointed me as a Fellow to Costa Rica for the year 1929-30 and to Europe from October to December, 1930; to the President and Fellows of Harvard College and to the Farlow Herbarium, for equipment, incidental expenses, and for an assistant in the field; to my colleague, Prof. Greenman, and former colleagues, Professors Ames, Barbour, and Thaxter, and to Mr. Paul C. Standley of the Field Museum, for excellent advice in planning my trip and for helpful letters of introduction. Members of the staff of the United Fruit Company were also very helpful both in planning the trip and placing the facilities of that great company at my disposal in Costa Rica, especially Dr. J. R. Johnston of the Boston office, and Messrs. Kress, Fuller, Craw- ford, Stübbe, and George Catt, the director of its Botanical Garden and Experiment Station at Siquirres, who accompanied me on several trips in that vicinity. Among Costa Ricans, I am deeply indebted to the following for advice and hospitality, without which I should have been unable to cover so much territory so thoroughly in so little time: Bernado Yglesias Rodriguez, the Director of the Escuela Na- cional de Agricultura, who placed the facilities of the botanical laboratory at my disposal and enabled me to spend three very profitable weeks at the ancestral Yglesias finca, Guayabillos, on the upper slopes of the western face of Irazü; to Anastasio Alfaro, formerly director of the Museo Nacional (and later to his suc- cessor, the late J. Fidel Tristan), who placed the facilities of the library and herbarium (Wercklé) of the institution at my dis- 1933] DODGE—LICHENS OF COSTA RICA. I 375 posal and suggested the plan of my trip to Guanacaste, as well as contributing several specimens; to Dr. Alberto M. Brenes, the botanist at the Museo Nacional, who collected extensively for me in the vicinity of San Ramón, an interesting region which I did not have an opportunity to visit personally; to Ricardo Chavarría Flores, formerly city engineer of Cartago, who en- abled me to collect extensively at his finca in the vicinity of Santiago de Cartago and the Rio Birrís; Juvenal Valerio Rodri- guez, of the Instituto de Alajuela, who planned trips in the vicinity of Alajuela and to the Cerros de Zurquí; Otón Jiménez Luthmer, of the Botica Oriental, a well-known amateur botanist, who gave generously of his time in suggesting trips and making plans, although his business duties prevented his accompanying me in the field except for a short time on our trip to Guanacaste; to Ferdinand Nevermann, the coleopterist, for numerous pleasant collecting trips, including a three-weeks stay at fincas of which he is manager in Limón Province, and for much helpful information in connection with my large insect collections; to Ruben Torres Rojas, of Cartago, for specimens, although I did not have the pleasure of a personal contact with him during my short visits to Cartago; and to Charles H. Lankester, for helpful suggestions. Besides these who are more or less professionally interested in natural history, the following were generous in their hos- pitality and enabled me to visit regions which would have been otherwise quite inaccessible to the traveler: Gonzalo Volio and his son Carlos, Antonio Sobrado and his brothers, Carlos Collado, Fernando Castro, José Castro Araya, Antonio Gutiérrez, José Luis Sancho, Carlos Piedra and son, Alexander Ross and sons, Robert Hanckel, and H. J. Marks. While in Europe I was indebted to the curators, acting curators, and staffs of the following institutions for permission to study the herbaria under their care: the Art Galleries of the Glasgow Corporation (Stirton Herbarium); British Museum of Natural History; Royal Botanic Gardens at Kew; Universitetets Bot- aniske Museum, Oslo; Botaniske Trädgård, Upsala; Riksmuseet, Stockholm; Botaniske Museum, København; K. Botanische Garten, Berlin-Dahlem; K. Botanische Museum, München; Conservatoire Botanique and Herbier Boissier of the Université [Vor. 20 376 ANNALS OF THE MISSOURI BOTANICAL GARDEN at Genéve; and the Laboratoire de Cryptogamie at Paris, as well as the private herbarium of M. le docteur M. Bouly de Lesdain of Dunkerque. I am also indebted to the following lichenolo- gists, whom I met during my trip, for helpful suggestions and criticism: Miss A. L. Smith, Bernt Lynge, Einar DuRietz, the late Karl Schulz-Korth, J. Motyka, and Ove Hgeg. Since my return to America I have had access to or seen specimens from the following herbaria, for which I gratefully acknowledge my indebtedness to the curators: Farlow Herbarium of Harvard University, New York Botanical Garden, National Herbarium of the Smithsonian Institution, the Field Museum of Natural History, Chicago, the herbaria of the University of Michigan and of the Missouri Botanical Garden. I am also grateful to the Director of the Missouri Botanical Garden and to the Chan- cellor of Washington University for a leave of absence after the appointment to the staffs of those institutions, in order that I might complete certain studies in the Farlow Herbarium of Harvard University before I left that institution. Finally I wish to thank any others through whose hospitality and assistance this work has been made possible. GEOGRAPHY Topographically Costa Rica may be divided into four main phytogeographic regions: the Atlantic Coastal Plain, the Meseta Central, the Subalpine Region, and the Pacific Coastal Region, the latter less well known and less homogeneous. The Atlantic Coastal Plain is characterized by its gentle slope from the sandy beaches up to about 100 m. or even higher in the poorly known Llanos de Santa Clara and in the Sarapiquí, San Carlos and Rio Frío valleys. The beaches are interrupted only by shallow harbors at Cahuita, Limón, and Moin, the first being protected by a coral reef, the two latter being partially protected by the small island Uvita, now used as a quarantine station for Puerto Limón. Several large rivers rising in the highlands cross this plain and during their flood states overflow, forming spectacular anastomosing channels which leave small pools in the dry season. Most of the river mouths are closed by shifting bars which hinder or prevent the use of the rivers for COCKAYNE, BOSTON PEAT PME iA 9 NENE, | DRITTE le "Ud Ve cuz boat C AC Map 1 4 7^ i T EU ess) p A v. D u E S 4 " ^ - X AN d Vi rovs (Co [y a — « A j od NE AN nZ re s S PX T À À P » " y Y SI x O14 D. 4 Y ^s 3 X F á \ i wi AA ~ JA A ^ ES f, - KE? "UO di fe / G " A =~ - ~ x Ff » 4 A COSTA RICA g BOCA P NA D ; ADBOC A PA ZS M TRA Y 4 f EN ae BIS ie AB i Se ON. : EE NEK x N A NEUMA ay à SA : y ^ j 4 " "o EA / Wy; INS Sore Pert i "4| Ww. = sm e ~ js DODGE—LICHENS OF COSTA RICA ANN. Mo. Bor. GARD., Vor. 20, 1933. 1933] DODGE—LICHENS OF COSTA RICA. I 377 shipping. Just back of the beach there is an almost continuous series of lagoons, which may be dredged and may be used for transport of the farm produce to the railroads and ports. These lagoons extend most of the way from San Juan del Norte to Limón. The soil is quite fertile where drainage is sufficient and much of the forest has been cleared for agricultural purposes, in the Llanos de Sta. Clara and Limón. The rainfall is heavy and quite evenly distributed, although somewhat less during January, February, and March, the so-called dry season (verano). Dur- ing most of the year the sun shines until about noon, then showers may be expected for either a short time or for the rest of the afternoon. The nights are usually clear and cool. During the colonial era there was very little permanent settle- ment of this region although many attempts were made. Some cacao plantations were developed at Matina, and their product taken overland via Cartago to Honduras and Guatemala or quietly sold to the English pirates and smugglers who frequently entered the mouth of the Matina River for this purpose. With the advent of the Northern Railroad in 1870-1890 the banana industry spread rapidly over most of the available land from the Parismina to the Estrella River valleys and up the Sixaola from Bocas del Toro. When the banana wilt or Panama disease (Fusarium cubense E. F. Smith) and increasing soil exhaustion began to threaten this industry, coconuts, pineapples, and cacao were extensively planted in suitable localities. Coconuts were à prompt and complete failure owing to bud rot and other factors, although they still furnish enough for local consumption. Pineapple cultivation has recently been abandoned for economic reasons, although the product was much superior to that of Hawaii. Cacao cultivation has been only partially successful owing to high labor costs. Recently much land has been practi- cally abandoned and is gradually returning to second-growth forests. While small plantations of forest trees, as Hevea, Ochroma, etc., are still in an experimental stage, they seem to be successful botanically but owing to the high cost of labor they are still doubtful economically. Maize is sometimes grown on the Llanos de Santa Clara. [Vor. 20 378 ANNALS OF THE MISSOURI BOTANICAL GARDEN The Atlantic Coastal Plain penetrates far inland in the Tortu- guero, Sarapiquí, San Carlos and Río Frío systems, but since I have not had an opportunity to examine them personally and have seen very few specimens from them it is impossible to discuss them further. Before the railroad was built the Río Sarapiquí was the main route from San José to the Atlantie, and a few villages still remain. ‘There are a few scattered settlements on the Río San Carlos and Río Frío, as may be seen from the relations of Thiel and Cespedes, but except for a few specimens by Pittier from the upper Sarapiquí Valley and from the vicinity of San Rafael on the R. San Carlos, I know of no lichen collections from this region. 'The Atlantie Transition Zone from 100 to 900 m. is evident mostly in narrow river valleys, the only botanically well-known portions being the southeastern portion of the Llanos de Santa Clara around Guápiles, formerly an important farm of the United Fruit Company used for their earlier experimental plantings of forest trees, and the valley of the Reventazón. In the valley of the Reventazón, tertiary sedimentary rocks are evident in the cafions and railway cuts. The slopes are very steep, due to the great erosion caused by the heavy rainfall which is about the same as that of the Coastal Plain. Bananas and coffee form the principal crops in the broader side valleys, the banana being comparatively short-lived but much used as a shade for the coffee. Since this development is comparatively recent, it still remains to be seen whether this is à permanent stage or a temporary phase in the agricultural development of the region. Formerly sugar cane was extensively cultivated in this region and considerable areas remain in Juan Vifias and Santiago de Cartago. Pineapples are grown for the domestic markets near Turrialba and to a less extent in some other places. Owing to the rapid erosion it seems doubtful whether much of the land which has been cleared on such steep slopes can remain permanently in cultivation. "There are available a few collections from Angostura near Turrialba by Polakowsky and my own from the beautiful valley of the Río Pejivalle near the United Fruit Company's farm of that name, as well as the even more extensive ones from Santiago de Cartago and occasional specimens from 1933] DODGE—LICHENS OF COSTA RICA. I 379 Juan Viñas (Naranjo) by Ørsted and scattered numbers by Pittier. On the practically inaccessible slopes and in the valleys where transportation is too costly to make farms economically profitable yet, some forests are still to be found. These still resemble those of the Atlantic Coastal Plain with their abundant epiphytes, bromeliads being conspicuous, especially Tillandsia usneoides. There are many small areas among the hills where fogs keep the humidity much higher than in neighboring areas. These are quite evident by the abundance of the long pendulous species of Usnea which replace Tillandsia usneoides as a prominent feature of the landscape. Species of Leptogium, Parmelia, Anaptychia, and mosses are also very abundant. Cladonias begin to appear and replace the similarly colored moss of the lower elevations. In the top of the forest trees and in the isolated trees of the potrero (pasture), crustose species are abundant. In the lower levels of the forest, lichens are rarer and usually sterile, perhaps due to the small amount of light. Trees whose bark scales off rapidly, such as the guayabo (Psidium guajave L.), are rarely found to have lichens, except for a few fruticose species of Ramalina and Usnea. Nephroma was found in one of these Usnea-Tillandsia zones on an exposed ‘‘knife-edge’”’ pro- truding into the cañon of the Reventazón near Santiago de Cartago. The Meseta Central, occupying the huge central valley between the Cordillera Central and the Cordillera de Talamanca, consists of long gentle slopes between 1000 and 1700 m. cut by deep river gorges. The pass of Ochomogo and the Cerro de Carpintera divide this meseta into two portions, the eastern valley of Guarco, comprising the upper portion of the valley of the Reventaz6n and its tributaries, and the western valley of San José, comprising the upper part of the valley of the Rio Grande de Tarcoles and its tributaries. The valley of Guareo is bounded on the north by the upper slopes of the volcanoes Irazü and Turrialba and on the south by the continental divide in the Cordillera de Talamanca. Most of the large tributaries enter from the south and are not explored fully from a geographical standpoint. I have seen lichen collections from the valley of the Río Navarro and the [Vor. 20 380 ANNALS OF THE MISSOURI BOTANICAL GARDEN lower part of the Orosi. The Río Grande is by far the largest and longest of the tributaries but its flora is wholly unknown to me. Coffee is the principal crop though considerable amounts of garden vegetables, especially chayote and yuca, are raised to supply not only the local markets but also the demands of the Atlantic Coastal Plain. Cartago, the colonial capital, is the only city but there are numerous small towns. The collections of C. Wercklé, ‘‘near Cartago 1200-1500 m.," in San José and Berlin, seem to be from the region about Aguacaliente and the lower slopes of the Cerro de Carpintera, judging from species represented. The valley of San José is bounded on the north by the continental divide of the western portion of the Cordillera Central, the volcano Pods, and Barba with its lesser peaks, the Cerros de las Caricias, Las Lajas, and Zurquí, on the east by Irazú and the Cerro de Carpintera, on the south by the Cerros de Escasi and Piedra Blanca, on the west by the Cerros de Aguacate. In the valley of San José the slopes descend gently to about 650 m. This valley is much drier than the Guarco with a longer dry season from December to April and less rain in the other months. The winds are strong, especially in the dry season. Practically the whole region has been denuded of forests, partly in the interests of agriculture and partly owing to the fiendish glee of the in- habitants who deliberately set fires to the dry grass on windy days in order to watch it burn. By the end of the dry season practically the whole valley outside the cafetales and cane fields is black from the charred remains of grass fires. The underlying rock of both valleys is largely volcanic, while the soil above is largely of aeolian origin from the volcanic dust which is still being regularly emitted from Irazü and occasionally from Pods. In some places fields of voleanic boulders exist in such profusion that they make agriculture as difficult as do the glacial boulders in parts of New England. The greater portion of both valleys is devoted to coffee growing, with sugar cane and upland rice in the lower levels. Formerly considerable excellent tobaeco was grown but this crop has practically disappeared. The cities of Alajuela, Heredia, and San José, and many important towns are located in this valley. 1933] DODGE—LICHENS OF COSTA RICA. I 381 Crustose lichens are abundant in the cafetales both on the coffee and its shade, but practically no attempt was made to collect them since the thalli are mostly young and sterile and we hesitated to injure the trees by removing portions of bark. The living fence posts, mostly species of Erythrina (poró), were utilized, however, since they are regularly and heavily pruned back. These yielded most of the crustose species from this area. Another fruitful source was the roadside bank. These banks are caused by erosion and by cutting down with spades to secure material to fill in holes in the road. The depth of a road in most places is proportional to the slope and to its age, a recently laid-out road having only slight banks while an old road often has vertical sides rising 15-30 feet above the wheel tracks. These banks have a wealth of lichens, bryophytes, and fungi during the rainy season (invierno) but usually dry out, except in occasional moist shady spots, during the dry season. Con- sequently most of the species are annuals and disappear quite completely during the dry season. These banks are even more conspicuous and important in the Subalpine Region. The Subalpine Region probably includes most of the land in Costa Rica above 1700 m., but is known botanically only on the Cordillera Central and the northernmost mountains of the Cor- dillera de Talamanca as far as the Cerro de las Vueltas. Pods, Irazú, and Turrialba are largely cleared on their southern slopes and used for dairy farms. Barba is still largely wooded and inaccessible, but Standley and Maxon have explored the slopes of the lower peaks of the Cerros de las Caricias, las Lajas, and Zurqui. I also had the opportunity to spend two days on the lower slopes of Zurquí. Standley and J. Valerio secured an extensive series of lichens from Cerro de las Vueltas and neigh- boring peaks within reach of Santa Maria de Dota, in 1925-6. Various collectors have brought back small series of specimens from Pods. Standley in 1924 and I in 1929 collected on the south slope of Turrialba for a short period, and I spent about three weeks at Guayabillos on the west slope of Irazü, as well as climbing to the crater twice from the south side. Ever since the time of Ørsted many collectors have brought in a few speci- mens from lrazá. The alpine and subalpine areas of Irazü [Vor. 20 382 ANNALS OF THE MISSOURI BOTANICAL GARDEN are much less rich than the other mountains probably because of the constant dust. On my last ascent I did not find even a sterile thallus above 3000 m., while Ørsted reports lichens from the summit. Botanically this Subalpine Region is the most interesting, as it has a high percentage of Colombian and other South American species, as well as most of the endemic species. The Pacific Coastal flora is too poorly known to classify into geographical divisions. I have been told by lumbermen that there are many interesting local distributions of the forest trees but these observations are too fragmentary for generalizations. For present purposes we may subdivide into the province of Guanacaste the peninsula of Nicoya, the great river valleys of the R. Naranjo, R. General, and R. Grande de Térraba or Diquís, and the Peninsula of Osa. Of these regions I have personally visited only the first and the last. The peninsula of Nicoya has not been explored at all as far as concerns lichens, while our knowledge of the great river valleys is due solely to two journeys by Pittier, Biolley, and Tonduz, in January and February, 1891, and by Tonduz from October 18, 1891, to April 17, 1892 (see p. 388). The province of Guanacaste consists of a low coast range penetrated by the Bahías de Culebra and de Salinas, a broad valley of the R. Tempisque about 40 km. wide and 60 km. long, rising gently from sea level to about 200 m. and sloping again gently to the Lago de Nicaragua, with Liberia, the provincial capital, at its center, and the west slope of the Cordillera de Tilarán (or de Guanacaste) on its east. Ørsted crossed the plain from the Tempisque to the Bahía de Salinas and explored the region of R. Sapoa and the coast, but I have seen no lichens from this region. Pittier collected along the shore of the Bahía de Salinas and may have penetrated the low coast range. Other- wise it is unknown as far as lichens are concerned. "The plain of the Tempisque is used for grazing with some sugar cane grown in the southwest part in the vicinity of El Paso. The underlying rock and soil in Liberia is white cascajo, a soft pumice laid down in water and hence somewhat stratified. Further south there are clay deposits. Rainfall is high during the rainy season but the dry season is hot and dusty, unofficial temperature in the 1933] DODGE—LICHENS OF COSTA RICA. I 383 shade rising as high as 108°F. in the hottest part of the day and correspondingly cool at night. Wind velocities are high where the trade winds sweep through the passes between the volcanoes in the first half of the dry season, so that most of the perennial vegetation is that of the arid regions and may account for the similarity of the flora with that of Nicaragua and the north. Here grass fires are even more frequent in proportion to the population and more destructive, hence the region is seldom visited by botanists; in fact I have failed to find any collections from this region or any mention of it in botanical works. In 1930 I was able to spend about two weeks in the vicinity of Liberia. The western slopes of the Cordillera de Tilarán are used for grazing, with small cafetales located in sheltered, moister areas. 'The vegetation does not appear much different from the plain of the Tempisque, especially on the exposed slopes, but in the shel- tered ravines more delicate types survive. This region has not been visited by botanists to any great extent, although Standley and J. Valerio collected in the vicinity of Tilarán in 1926, and Valerio probably made studies of the local flora of Tilarán while he was stationed there as a teacher some years ago. Be- sides a week in Tilarán I spent about a week at 700 m. on the slopes of Santamaría, the next peak south of La Vieja, and a week at La Granadilla on the plateau between the R. de las Cafias and the R. San José which culminates at 960 m. in the Cerro de San José. In Santamaría and Tilarán, even during the dry season, much of the time the weather was characterized by mists and high winds and much rain, while a short distance lower and to the west there was neither mist norrain. However, the brief intervals of hot sun are sufficient to keep the xerophytic vegetation on the wind-swept prairie, while the sheltered ravines and shaded hillsides have mesophytie types. At Tilarán we crossed the continental divide into the headwaters of the Río Frío system with a change in vegetation even more striking than in the pass at Ochomogo in the Meseta Central. The Cerro de San José rises sharply above the surrounding peaks to 960 m., somewhat higher than the near-by peaks of the [Vor. 20 384 ANNALS OF THE MISSOURI BOTANICAL GARDEN continental divide. It seems to be composed of stratified sedimentary rock sloping 40? S, 70° W. To the west rises the lower exploded crater of Cerro Pelado, while to the east lie the lower peaks which form the junction of the Cordillera de Tilarán and the Cordillera Central. Though the region is considerably drier than the same elevations on the Cordillera de Tilarán, yet it maintains a luxuriant forest vegetation where it has not been cleared for potreros. Being an isolated wind-swept and fire-swept peak, the vegetation of the summit has a stunted, subalpine appearance. The Hacienda Granadilla on the plateau between the Río de las Cañas and the Río de San José at about 900-600 m. furnished much interesting material, especially as we were fortunate in finding two huge pochotes (? Bombaz elliptica HBK.) which had blown over so recently that the crustose lichens were still fresh and the clusters of orchids and bromeliads were still in bloom. These furnished about the only collections to show the flora of the tree tops in Guanacaste. The region between the Cerro de San José and Puntarenas has not been visited by botanists to my knowledge, although it is quite accessible on account of the Abangares and other gold mines which have been exploited more or less continually since colonial times. Since the region is of a different geological formation it would probably well repay investigation, as would also the calcareous Santa Catalina range between the R. Tempis- que and R. Bebedero at the head of the Gulf of Nicoya. The peninsula of Osa between Golfo Dulce and the Pacific consists of low mountain ranges, probably of sedimentary rock; at least such were the outcrops observed along the shore and in the gorge of the Quebrada de la Laja, a tributary of the Río Nuevo. G. Cufodontis, of the Austrian expedition of 1930, visited the region in the immediate vicinity of Puerto Jiménez (Santo Domingo, on many old maps and charts, on the small bay just above Puntarenitas). Two weeks later I visited the same region, also the Hacienda Sándalo between the Tigre and Terrones Rivers (marked Necki on the charts). The coastal plain about a mile wide slopes gradually from the beach to about 30 m. and then more steeply up the spurs of the range which follows the west coast of the Golfo Dulce. The forest is still virgin in most 1933] DODGE—LICHENS OF COSTA RICA. I 385 places and consists of a rather open palm and hardwood forest characteristic of well-drained areas. The coast between the tide levels has a characteristic mangrove swamp. Mr. Dunlap, of the United Fruit Company, visited this region a few years ago but I have not learned of any specimens taken. Mr. H. J. Marks, the proprietor of the Hacienda Sándalo, has sent a few specimens of wood and herbarium specimens of forest trees to Prof. 8. J. Record of Yale University. Practically nothing is known of the flora of the islands off the west coast. The only one which has been visited by bota- nists, Isla Coco, has so far yielded very few foliose or fruticose species. Dr. H. K. Svenson, botanist of the Astor expedition, one of the latest to visit the island, describes it as follows: ‘‘ The lichen flora must be very inconspicuous; I did not go out of my way to look for lichens but if any large forms had been in my path, I believe I would have picked them up. Cocos Island is entirely covered by a rank vegetation which the sunlight rarely pene- trates.” So far the only species reported are one Leptogium and two Parmelias, all wide-ranging species of the tropical lowlands. BoTANICAL EXPLORATION Of early botanical exploration in Costa Rica the little that is known apparently indicates that few or no cryptogams were collected. Martin Sessé and José Mariano Mociño probably visited the Pacific Coast between 1795 and 1804. The Sulfur," with George Barclay as botanist, entered the Golfo de Nicoya about 1840 and about the same time Emmanuel Friedrichstahl visited a portion of Nicaragua and Costa Rica though his collec- tions were not carefully labeled as to locality, all the sheets being marked Guatemala. Between 1845 and 1851, the “Herald” probably visited Pedregal, the harbor for Davíd, Panamá, as Seemann the botanist brought back from Boquete and the neighboring voleano of Chiriquí, lichens which were studied by Churchill Babington and published by Seemann (1852-1857). These specimens are in Babington's herbarium at Cambridge University and in the Royal Botanic Gardens at Kew. The first to make extensive collections was Anders Sandge [Vor. 20 386 ANNALS OF THE MISSOURI BOTANICAL GARDEN Ørsted (1816-1872), to whom we owe the first elaborate accounts of the natural history of Costa Rica. For details of his travels and lists of plants collected at various places, see Ørsted (1848, 1851, 1863), Hemsley (1878-1888), and Durand & Pittier (1891). He entered the country at Puntarenas, passing over the old carretera which ascended the Cerros de Aguacate thence via Alajuela to San José, where he made his headquarters for some time. From this point he visited C. Escazú and Pacaca, pene- trating to the mountains of Jarís to the south. He also visited Candelaria, probably in the vicinity of Tablazo, judging from the list of plants he collected in that locality. In May, 1847, he explored Poás and Barba. Earlier in that year he had spent some time studying the volcano Irazü, making his headquarters in Cartago. From this point he visited Aguacaliente and Ujarrás (probably the valley below the modern Paraíso), then made the very difficult trip to Moin over the old trail now largely aban- doned, passing Cervantes, Río Birrís, and Quebrada Honda to Naranjo (the modern Juan Viñas), thence to Turrialba, keeping along the shelf of land above the lake near Tunnel Camp, Bonilla, ete. to the eustoms house on the R. Reventazón (probably below the present La Junta) where he crossed the Reventaz6n in canoes and proceeded along the coastal plain to Moin. Another long trip was his visit to Guanacaste and adjacent Nicaragua, the details of which were published in 1848 from a letter he had sent home. He does not state in what year this trip was undertaken, but since the mention that Puntarenas was then a free port (this having been decreed in 1847) and since he was visiting voleanoes in February and May, 1847, it is probable that he made the journey in 1848. He was persuaded by govern- ment officials to explore a possible route for an inter-oceanic canal between the Lago de Nicaragua and the Pacific. The party, headed by Francisco Gutierrez, left San José on February 17 for Puntarenas, where they took a huge canoe up the Gulf of Nicoya and thence up the Río Tempisque. On March 8 they set out with a pack train from Santa Rosa for the Hacienda Sapoa just above the junction with R. Bolaños. Thence they visited H. Las Animas above the junction with the R. Gua- chepelin, ascending that valley and descending the R. Tortuga, 1933] DODGE—LICHENS OF COSTA RICA. I 387 the next river east of the R. Sapoa. @rsted may have visited some of the islands in the Lago de Nicaragua before he ascended the Río Sapoa to the junction of the Sansapote and overland to the H. Sapoa. Another trip was taken to the Golfo de Bolafios, thence over the height of land and down the Sansapote to its junction with the Sapoa. Ørsted also visited San Juan del Norte (Greytown), crossing the pass between Barba and Irazü at La Palma and descending along the Río Sucio and the R. Sarapiquí. It is not known whether he sailed from San Juan del Norte or from Puntarenas, which was frequently visited by Danish vessels at that period, thirteen having called there in 1853. Warscewiez, coming from Guatemala via El Salvador and Nicaragua, is said to have met Ørsted in Nicaragua before reaching Costa Rica via the R. Sarapiquí. He evidently visited Irazá and Turrialba, but I have seen no lichens from his collec- tions. Orsted's huge collections are to be found in Kjgbenhavn where their study has only recently been completed. "They were distributed about 1931. The fungi were sent to Elias Magnus Fries at Upsala and were published promptly by him in 1851, portions of the collections being found in both Kjøbenhavn and Upsala. The lichens were sent to Th. M. Fries but only a single number, Stereocaulon obesum 'Th. M. Fr., was published. Some have since been named and inserted in the herbarium at Upsala. Through the kindness of Professor N. Svedelius, the remaining unnamed specimens were turned over to me for study and are reported in this work. At present none of Orsted's lichens are at Kjøbenhavn but it is probable that a complete set of duplicates will eventually be deposited in that herbarium. Ørsted was soon followed in Costa Rica by a number of natu- ralists who have left us long accounts of geology and natural history. Many resided in Costa Rica for several years, but few appear to have collected lichens. During his visit, Polakowsky collected a few specimens studied by Nylander (1876) and apparently distributed quite widely, as I have seen a complete set in Berlin and scattered numbers in other herbaria. [Vor. 20 388 ANNALS OF THE MISSOURI BOTANICAL GARDEN A new era for Costa Rican natural history began with the establishment in 1887 of the Instituto Fisico-geografico Nacional de Costa Rica, with Henri Pittier as its first director. Due to his enthusiastic explorations and those of his fellow botanist, Tonduz, a large herbarium was assembled in San José and materials were sent to many foreign collaborators for study, resulting in the publication of Durand and Pittier (1891-1901). The lichens which were studied by Müller Argau were largely from two extensive expeditions to the southwestern portion of the country and scattered numbers collected by others in various parts of the country. Pittier (1891) described these expeditions in detail. As these collections will be frequently mentioned I have thought it wise to include an abstract of their itinerary. On January 15, the first expedition reached San Marcos de Dota, the capital of the canton of Tarrazü, the next night Santa María de Dota, and the C. del Angelo, small valley of El Copey 1790 m., and El Roble 2670 m., on the following day. On January 18 they passed Alto de la Baraja 2933 m., Rancho de las Vueltas, C. de las Vueltas 3019 m., and Ojo de Agua 2760 m.; on January 19 the Paramo de las Vueltas and the C. de Buena Vista 3299 m., and La Muerte at 3130 m.; on the 20th, southward along the slopes of Buena Vista to Lagunilla 1857 m., the Alto del Palmital at 1211 m., reaching El Géneral on the evening of January 21 where they stayed until January 28, 1891, and whence Biolley returned to San José. Pittier and Tonduz continued to Buenos Aires, Térraba, and Boruca, whence Pittier returned by a somewhat different route to San Marcos and San José, arriving on February 23, while Tonduz stayed in the field until March 8, 1891. The following season Tonduz left San José on October 18, 1892, more or less retracing the journey of the previous season to Buenos Aires, thence down the R. Grande de Térraba or Diquís, stopping some time at Boruca, finally down the river to the coast and back by sea to Puntarenas, stopping at several points, arriving on April 17, 1893. The material from these two trips furnished the bulk of material upon which Müller Argau (1891, 1893) based his account of the lichens. 'These collections are to be found in the Herbier Müller Argau, 1933] DODGE—LICHENS OF COSTA RICA. I 389 which was purchased by the Herbier Boissier and is now at the Université de Genéve. I have been told that a duplicate set exists in Brussels, but apparently none was returned to the Museo Nacional in San José. C. Wercklé, an engineer who resided in Costa Rica for many years, collected much in the vicinity of Cartago. His specimens were determined by Lindau and may be seen in Berlin and in Wercklé’s herbarium at the Museo Nacional de Costa Rica. Wercklé was not at all discriminating in selecting samples for Lindau, so that most of the material in San José needed revision. The Calverts (1917) spent a year, 1909-1910, in Costa Rica and have given us a good account of its natural history. They brought back some plants which are in the herbarium of the University of Pennsylvania, but I have seen only one lichen from them (at the Farlow Herbarium). Maxon and assistants have visited Costa Rica for ferns and brought back some lichens which were sent to G. K. Merrill for study and are now to be found in the Farlow Herbarium (which acquired the Merrill herbarium) and in the National Herbarium of the Smithsonian Institution at Washington. Paul C. Standley, in 1924 and again in 1925-6, in company with Juvenal Valerio Rodriguez and Ruben Torres Rojas, made very extensive collections of lichens although his primary interest was angiosperms. These lichens were likewise sent to G. K. Merrill who was studying them at the time of his death. As curator of the Farlow Herbarium I continued this study in preparation for my trip to Costa Rica and am here reporting these collections. In 1925, during a two-weeks’ visit, I secured a large amount of material from Cerro Carpintera and again in 1929-30 I secured several thousand specimens of cryptogams which are still being studied. The most complete set of these plants is in my personal herbarium and the next set, lacking only a few numbers of the commoner species, is at the Farlow Herbarium. Karl Danielson, an assistant to H. E. Stork, in 1928 collected a few numbers of lichens which are now at the University of Michigan. These are also reported here for the first time, duplicates being found in the Farlow Herbarium and in my own herbarium. [Vor. 20 390 ANNALS OF THE MISSOURI BOTANICAL GARDEN In 1920 and 1930 the Costa Rican government botanist, Alberto M. Brenes, collected many lichens in the vicinity of San Ramón and a few from other parts of the country. These were also turned over to me for study and are to be found in the Museo Nacional, with duplicates in my own herbarium and in the Farlow Herbarium. The Oesterreichische Costa Rica Expedition, under the leader- ship of Otto Porsch (1932) with G. Cufodontis as assistant botanist, spent the late verano and early invierno in Costa Rica, visiting much the same territory as I did. If lichens were collected they were probably studied by Zahlbruckner but I have seen no publication on this group, although many of the other groups are already published. FLonisTICS It is too early to draw any very definite conclusions regarding the distribution of plants and the affinities of the flora, but certain typical distributions are evident. Families and genera which are abundant in the northern United States and are often widespread in the North Temperate Zone are usually found at the higher altitudes and reach their southern limits in the mountains of Colombia, except for oc- casional isolated species farther south. Such an example is the Peltigeraceae, which are widespread and abundant in the northern part of the United States. Only two genera, Nephroma and Peltigera, reach the American tropics. Nephroma is rep- resented by a single species seen in only one restricted locality in Costa Rica. Peltigera is represented by about eight species, mostly strictly tropical with only one undescribed species reaching Magellanes, Chile. Another typical example is a northern family, which is char- acteristic of the Atlantic Coastal Plain, the Collemaceae. Here genera are mostly common to the eastern and western hemispheres but subgenera and species are much more restricted in distri- bution. These are mostly found on the coastal plains of Costa Rica, being characteristic on the plains of Guanacaste and occasional on the Atlantic Coastal Plain but not yet reported from Panama. The species belonging in these genera from the 1933] DODGE—LICHENS OF COSTA RICA. I 391 higher elevations have relations with those of the mountains of Colombia and not with those north of the Isthmus of Nicaragua. A third type of distribution is found in those families which are predominantly southern, as in the Stictaceae. Here we find them highly developed and characteristic in the highlands, relatively rare in the lowlands. The genera and subgenera are world-wide in distribution and the species also have wide ranges, becoming less frequent in numbers and in species in the North Temperate Zone. In this group, however, species of the moun- tains of Costa Rica may extend into North America along the Atlantic Coastal Plain, even as far as the Avalon peninsula of Newfoundland, while the species of the Coastal Plain in Costa Rica are southern in their ranges. Finally some cosmopolitan families in which it is difficult to determine a center of distribution have genera or subgenera which are strictly limited to tropical America. Further consideration of problems of distribution may be deferred until the systematic enumeration is complete. In general the affinities of the flora seem to be with eastern Brasil but since we have very few collections available from the inter- mediate mountain ranges of eastern Colombia and southern Venezuela and British Guiana, it is likely that stations of many of these species will eventually be discovered in these regions. Relatively few species are common to Mexico and Costa Rica or to Perú and Costa Rica, and practically no Costa Rican species extend as far south as Chile. In the following systematic enumeration I have included in the keys to species all those which have been described from tropical America, whether they have been found in Costa Rica or not. I have not included species which were not originally described from this area, although they may have been reported from it, since often these reports have been based on misdeter- mined specimens. While it is evident that political boundaries do not always coincide with natural floristic divisions, for the purposes of this study tropical America may be defined as Mexico and the West Indies southward to Perú, Bolivia, Para- guay, Brasil, and Uruguay. These are quite natural in the pres- ent state of our knowledge, as the lichen flora of northern [Vor. 20 392 ANNALS OF THE MISSOURI BOTANICAL GARDEN Mexico which might be expected to show affinities with that of the southwestern United States is practically unknown. Simi- larly, that of northern Argentina, which might be expected to be similar to that of adjacent Paraguay, has been little explored for cryptogams. Northern Chile, largely desert, has also been little studied, while the flora south of the desert has relatively few species which have been reported from tropical America as here defined. The greatest difficulty occurs in the exclusion of southern Florida where occur many species which are tropical rather than northern in their affinities. Fortunately for our purposes comparatively few species have been described from this region. KEY TO TROPICAL AMERICAN FAMILIES O FOLIOSE AND FRUTICOSE LICHENS Apothecia more or less exposed, paraphyses growing above the asci to = a tissue, asci early evanescent, leaving an enclosed — spore mas Gidea um).. ° CONIOCARPINEAE Thallus foliose or -frutiooss, corticate, apothects enel. — Apothecia linear, elongate-elliptic or angular, not forming à a ‘masadic cu da dad Cannas celcy i040 kine od ede E EE A NE GRAPHIDINEAE Thallus fruticose, erect or dependent, with a basal layer attached to the substrate, corticate [not yet reported between Lower California and Perú]; See rr ee er en eee ree ree RoccELLACEAE Apothecia poten not forming a mazediu ..CYCLOCARPINEAE Thallus strictly crustose, but aang piede appearing fruticose owing p the proliferation of the apothecia from the margins, easily mistaken for a hepatic on cursory examination; Guiana........ Polystroma Thallus loosely byssine from filaments of Cladophora or Trentepohlia, apo- thecia with light-colored parathecium.................. COENOGONIACEAE Thallus filiform or dwarf fruticose, or squamose, very — iam foliose [deferred for further treatment with the crustose speci gr symbiont Scytonema or Stigonema, apothecia more or m sunk llus, small and easily overlooked.................. EPHEBACEAE Algal avoddon ot DIR RP ENERO ECRES PYRENOPSIDACEAE Algal symbiont Révularía......... sese LICHINACEAE Thallus gelatinous, swelling greatly when moistened, dwarf fruticose, squamose or large foliose, algal symbiont usually Nostoc........ OLLEMACEAE Thallus definitely foliose or fruticose, or if small, the form of the thallus not due to the algal symbiont. Hypothallus and rhisoids highly — thallus squamose to small foliose, upper gal symbiont usually Nostoc. eae Wypothallus evanescent. Thallus large foliose, algal symbionts Nostocaceae or Palmellaceae, spores fusiform to acicular, 2 or more cellec 1933] DODGE—LICHENS OF COSTA RICA. I 393 Both surfaces corticate, lower surface interrupted by more or less highly developed breathing pores, apothecia with well-de- veloped parathecia or amphithecia, stipitate or sessile... . .STICTACEAE Only upper cortex developed, the lower surface tomentose with more or less highly developed network of veins and tufted rhizoids, apothecia attached to the thallus over their whole under surface without true parathecium or amphithecium aad surrounded fe the torn remains of the tissue cover- ing the young apothecium...................00005 PELTIGERACEAE Thallus oni or small foliose, only upper surface corticate, algal symbiont Pleurococcus, ups ecia is well-developed light- colored parathecium, spores va No secondary thallus or podetia pu apothecia sessile. ...... t ope op OST HYLLOPSORACEAE Secondary thallus or podetia present, varying from simple sta apothecium without algae to highly developed branched or infundibuliform structures, primary thallus squamose or small foliose, hie evanescent or degenerate, perhaps crus- tose in qome species s on Taal eds sero oe LADONIACEAE Thallus large foliose or Ee. algal symbiont Pleurococcus, spores ne- or two-celled or muriform, never elongate-fusiform to acicular. Apothecia with black parathecium, spores one-celled [in Gyrophora aplocarpa from Pert, the only species of the family so far re- Orted|. c anaa e EDD RUNE a ae GYROPHORACEAE ieee with bright-colored MESS ds Spores not placodiomorphou Thallus PEON ai piga Rae PRAE PARMELIACEAE Thalis Irulioo8b. ..... zx e E TUE. USNEACEAE Spores dimora Spores hyaline, thallus usually bright yellow, at least the epi THE CIUM OTaANgE « eoim Se G ELOSCHISTACEAE Spores brown, balie usually more or less glaucous, epithe- Mum brown OF DrUInoso. o NEL e HYSCIACEAE SPHAEROPHORACEAE Thallus foliose or fruticose, corticate on both surfaces or the lower surface incompletely corticate, with Protococcus. Apo- thecia sessile on the margin, or on the lower side of the thallus, open at first or enclosed by an amphithecium. Only two genera have been reported from tropical America, both fruticose with solid axis and terminal apothecium. Acro- scyphus sphaerophoroides Lév. without an amphithecium has been reported from Mexico and Perú but has not been found in Costa Rica. Sphaerophorus is typically Subarctic-Antarctic in distribution, coming southward on the higher mountains with [Vor. 20 394 ANNALS OF THE MISSOURI BOTANICAL GARDEN a single species on the higher mountains of tropical America. The whole family is greatly in need of revision. SPHAEROPHORUS MELANOCARPUS (Sw.) DC. apud Lam. & DC., Fl. Franç. ed. 2, 6: 178. 1805, excl. syn. Lichen melanocarpus Swartz, Nova Gen. Sp. Pl. Prodr. 147. 1788. Type: Jamaica, Swartz. Thallus ashy glaucescent or pale above, pale below, partly terete, partly compressed, smooth, not rugose, highly branched, 1-6 cm. high, primary branches 1-4 mm. thick, KOH yellow above, KOH - below, CaOCL, — ; cortex 20 y thick, subpellucid, of thick-walled conglutinate irregularly woven hyphae; medulla I —, composed of hyphae 4-8 u thick; mazedium oblique in the ends of the thicker terete branches, becoming disciform, exciple lacerate, ascospores 7-11 y in diameter. Spermatogonia immersed in the tips or in the lower surface of the branches, opening by a blackened verruciform mouth; spermatia oblong, 1 x 3 y [description from Nylander and Vainio]. The Costa Rican material does not altogether agree with the above description, but there is so much variation in macroscopic appearance that only a careful monograph can settle the number of species and their relationships. Cartago: F. V. Turrialba, 2000-2400 m., Standley 35313. San José: L. la Chonta, n. e. Sta. María de Dota, 2000-2100 m., Standley 42290; C. de las Vueltas, 2700-3000 m., Standley & J. Valerio 43832. Heredia: C. Zurquí, 2000-2400 m., Standley & J. Valerio 50436. ROCCELLACEAE This strictly maritime family has not yet been reported between Lower California and Perú, probably because no suitable mari- time habitats have been visited by botanists, for the group is abundant in such strictly tropical habitats as the Galápagos Islands. I did not have an opportunity to visit such myself, but had I been able to visit the cliffs along the various headlands between Puntarenas and Golfo Dulce, it is possible that I might have secured some representatives of this family. COENOGONIACEAE Thallus spongy byssoid, either adnate or forming dimidiate shelving masses, homoeomerous with Trentepohlia or Cladophora, 1933] DODGE—LICHENS OF COSTA RICA. I 395 whose filaments are partially surrounded by hyphae. Apothecia with pseudoparenchymatous parathecium; asci 8-spored; spores hyaline, one- or two-celled; spermatia exobasidial. Thallus with Trenlepohlia ee ee JE MI Coenogonium Thallus with Cladophora, apothecia unknown ............. sess. Racodium COENOGONIUM Ehrenb. CorNocontum Ehrenberg apud Nees ab Esenbeck, Horae Physicae Berol. 120. 1820 Thallus loosely spongy byssoid, either adnate or forming dimidiate shelving masses (suggesting a thin species of Poly- stictus), homoeomerous with Trentepohlia which is partially surrounded by a network of hyphae. Apothecia scattered on the upper surface, scutiform, usually with a short stipe, with a parathecium of thin-walled pseudoparenchyma without a me- dulla; paraphyses unbranched, often with swollen tips; asci 8- spored; spores hyaline, fusiform, ellipsoidal to elongate, one- or two-celled; spermagonia spherical, spermatia exobasidial, fusi- form, straight; spermatiophores mixed with anaphyses. Clements segregated the species of Coenogonium with uni- cellular spores as Holocoenis. However, his proposal has not been followed by other authors, although Vainio and Zahlbruckner have proposed sections on this basis, Vainio using Coenobiatora and C torina, while Zahlbruckner incorrectly adopted Holocoenis and Coenobiatorina. This family is very aberrant among lichens in several respects and perhaps would be better dropped and the species distributed among the fungi and algae. It seems more logical to regard the group as a case of parasitism of fungi on algae, as neither fungus nor alga seems to influence the development of the other, and there has been no further evolution resulting from the acquisition of a photosynthetic unit by the fungus. Taking this view, one would recognize about four or perhaps five species of fungi from tropical America, one species with unicellular spores about 6-10 x 2.5—4 y. belonging in Patinella; and three or four species with 2-celled spores belonging in Orbilia or a segregate from that genus. Among the algae one would recognize a dozen or more species of Trentepohlia which are often subject to the attacks [Vor. 20 396 ANNALS OF THE MISSOURI BOTANICAL GARDEN of Patinella and Orbilia. In the following key to the tropical American species of Coenogonium, as it is commonly understood by lichenologists, it will be noted that out of nineteen names proposed, four were based on specimens in which no apothecia were found. Four other species originally described in Coeno- gonium have already been transferred to Trentepohlia. I have earefully investigated the ascocarps of the fertile Costa Rican specimens and can find few characters which would separate species from each other, except spore size and septation. For lack of time to investigate this question thoroughly, I have here followed the traditional arrangement, as to do otherwise would involve serious nomenclatorial difficulties. To adopt the view that this group should be excluded from the lichens would invalidate Coenogonium, as based on a combination of characters of parasite and host and involve redescription of its species in Patinella, Orbilia, and Trentepohlia. KEY TO THE TROPICAL AMERICAN SPECIES OF COENOGONIUM Apothecial dise livid, cell walls aliis spores unicellular; algae 11-16 u in diameter; French Guiana. ............. 0.0 ccc eee eee eee ee ees C. Leprieurit Apothecia carneous, yellowish or waxy white, never livid. Thallus dimidiate, not adnate to the substrate Thallus thick, rigid, with white subsilky villum, zonate, sterile; Brasil. . "EE eee eee Te eT rer ey Terre re . Echinus Thallus thinner. Filaments 7-9 & in diam., cells not very distinct, greenish, sub- iffuse, WAROROIE So oi dics bie oy ERE PRAE REaMESA PERF subvirescens Filaments 12 u in diam., rigid: BINE 5554044400944 REPE AU RR C. Linkit Filaments thicker, 16-20 » in diam., cen m rigid. Cells 4-5 times as long as broad; Mexico................. C. confervoides Cells about twice as long as broad; spores T. 8.5 x 2-3 u; Brasil. . ————————————— eee C. aorocephalum Filaments and paraphyses thick, imperfectly known............. C. andinum Thallus adnate to effuse Filaments moniliform, 18 4 in diam.; thallus glaucous green to tawny brown, margins paler; spores 2-celled, 3-4 times as long as broad: CUDA: 1422453 09 X493 EEEIEE SOREN ANNETTA C. rica ais Filaments partially moniliform, partly cylindrical, cells 20- m.; € M Luo oa ERUDEERCOIERR TE SASL ES C. Missis Filaments uniformly ia Filaments 23-36 u in diam.; Sion apothecia 0.6-0.7 mm. in diam.; spores u- 15 x 25-35 u OBOE okt ies Se yn C. disjunctum filis 20-28 uw in diam.; ijon niae A 6-10 x3 à; or- bonia or ionii —— —— PIT E . interpositum 1933] DODGE—LICHENS OF COSTA RICA. I 397 Filaments 17-20 y in diam.; cells 30-50 » long; spores 8 x 2.5 u, 2- edad: Costa Bi .....eoeoe eorr err ERAS P E C. interponendum Filaments b m in Pc thallus ashy green, pulvinate, n^ to 2 eu DONE TIG. eie ores rre S vL iiio Filaments 12-16 n in pon thallus yellowish; spores 8-10 x 3 u Meek Kaan 85:5 0k os os ee 1. i C. interplerum hace 12-14 yw in diam.; thallus yellowish to green; St. Vince DCN SY es errr ere ee ee eprieurii v. panne Filaments 8-11 » in diam., conglutinate in fancidion 70-90 u diam.; e 7-10 3.5-4.5 y; Bolivia. -cees vats ss ade a C. paie ans Filaments 4-8 Filaments age aps spermatia 7-9 x 15 u; paraphyses cla- o PAR, oon ee OS eee C. dialeptizum Filaments articulate; spermatia not reported. Cells 2.5-4 times as long as broad, paraphyses obovoid; C. aiT AEE E esa EE ee ae ee rn pannosum ells 1.5-3 times as long as broad, sterile; Brasil......... C. depressum Filament size not given, rugulose, articulate-branched, coalesced into fascicles; Venesuels..... ... <0 610 s00 eto RR VH C. Tuckermani Cornocontum LEPRIEURII Nyl., Ann. Sci. Nat. Bot. IV, 16: 89. 1862. C. Linkii var. Leprieurii Mont., Ann. Sci. Nat. Bot. III, 16: 47. 1851 Holocoenis Leprieurii Clements, Genera of fungi, 174. 1909. Type: Guayane Française, Leprieur, sine no. Algae loosely woven, yellowish, 11-16 u in diameter. Apo- thecia plane or convex, excluding the white margin, dise livid from the first; asci about 20 u in diameter; spores oblong to fusiform or ellipsoidal, unicellular, 6-10 x 2.5-4 u; paraphyses slender, apices clavate. I have not seen this species from Costa Rica but Dodge & Nevermann 7399, from jungle at Castilla farm, 20 m., Limón Province, seems to be a variety of this species. The fungus is not well developed but has a livid disc, the margin is still thick and white, and the alga is of much smaller diameter. The exciple is composed of isodiametric cells with colored walls. It appears to be the same fungus attacking a much smaller alga, and should be referred to Patinella. COENOGONIUM SUBVIRESCENS Nyl., Flora 57: 72. 1874. C. Leprieurii var. subvirescens Nyl., Ann. Sci. Nat. Bot. IV, 16:89. 1862. Type: Brasil, Amazonas, Rio Negro, Spruce 28, is type of the [Vor. 20 398 ANNALS OF THE MISSOURI BOTANICAL GARDEN species, while C. Leprieurii var. subvirescens was described from rench Guiana without collector. Algae somewhat interwoven but more or less parallel, 7-9 y, forming a loose dimidiate to subcircular thallus depending upon position and shape of substrate, not or obscurely zonate, pale to dark olive buff, cells about 21-24 y long, septa not very evident, other cell walls thick, chloroplasts appearing somewhat in the shape of a dumb-bell with the pigment more or less massed at the ends of the cells. Apothecia about 800 y. in diameter, tapering below to a short stipe about 250 y. in diameter, composed of thick-walled pseudo- parenchyma with the outer 40 u composed of a palisade of elongate cells, perpendieular to the surface; parathecium about 80 y thick, homogeneous with the rest of the apothecium, the hypo- am very thin, filamentous, of cells with thinner walls. The hymenium is about 80 y tall and is composed of filiform para- physes with clavate tips and slender clavate asci. The asco- spores are not mature in my specimens but seem to be slender, 2-celled, probably about the size of most of the other spores in the genus except C. disjunctum. Limén: Cadiz, Dodge 7452; Castilla farm, 20 m., Dodge & Nevermann 7451; Monte Verde, K. Danielson 73; Siquirres, I m., Dedin Catt & Thomas 5574; La bier 70-80 m., Standley 368 Car : R. Pejivalle, 650-800 m., “re & Thomas 4408; El Muñeco, on R. aan 1400- 1500 m., Standley & J. Valerio 61055; C. Carpintera, 1500-1850 m., Standley 35596. Puntarenas: Osa, Golfo Dulce, Dodge 7450. Guanacaste: Q. Serena near Tilarán, 700 m., Standley & J. Valerio 46242; Santa- maría, 720-850 m., Dodge 7015. CoENoGONIUM Linxir Ehrenberg ap. Nees ab Esenbeck, Horae Phys. Berol. 120. 1820. C. controversum Pers. ap. Gaudich. in Freycinet, Voy. Uranie, Bot. 214. 1826 (nom. nud.). Type: Brasil, Sta. Catharina, Chamisso. The type of C. controversum came from the walls of the aqueduct of Corcovado, Rio de Janeiro, Gaudichaud. Algae soriewliag interwoven but more or less parallel, about 12 v in diameter, forming a somewhat rigid dimidiate to sub- orbicular thallus, not or obscurely zonate, pale to dark olive- buff, cells 40-60 y. long, septa very difficult to observe, chloro- 1933] DODGE—LICHENS OF COSTA RICA. I 399 plasts small and scattered, outer cell walls thick, partially covered with anastomosing fungal hyphae. Apothecia warm buff, about 850 y in diameter, tapering sharply to a short stipe about 240 y. in diameter and up to 300 y. long, pseudoparenchymatous, without palisade layer at the sur- face; parathecium about 100 u thick, homogeneous with the rest of the apothecium, the hypothecium very thin and soon evanes- cent; the hymenium about 80 u tall, composed of filiform para- physes with swollen tips and cylindrical asci. The ascospores are not fully mature in my specimens but probably the common size for the genus. Cartago: C. Carpintera, 1800 m., K. Danielson 9. Alajuela: La Palma de S. Ram in. 1100 m., Brenes 175. Guanacaste: El Silencio, near Tilarán, 750 m., Standley & J. Valerio 44619. COoENOGONIUM CONFERVOIDES Nyl., Flora 41:380. 1858. ? C. andinum Karsten ap. Nyl., Bot. Zeit. 20: 178. 1862. Type: Mexico, Orizaba, Fr. M üller (Schimper Herb.). Type of C. andinum from Colombia, 2000 m., Lindig 2560, also Perú, Weddell, and southern Brasil, Guillemin. Since I have been unable to see the types, the interpretation of this species is difficult. In Ann. Sci. Nat. Bot. IV, 11: 242. 1859, Nylander amplified his very inadequate description, basing his emendations on a specimen from Tahiti, Lépine 14. When he monographed the genus in Ann. Sci. Nat. Bot. IV, 16: 91. 1862, he cited Lepine 14 from Tahiti first, adding Guadeloupe Island, Duchas- saing, Brasil, Gaudichaud, and Weddell, and Chile, Gay, not citing the original Mexican type. Since none of these descriptions was based on specimens having apothecia its application is further confused. Since the descriptions only state how it differs from C. Linkii, I have assumed that the thallus is more or less dimidiate with a larger alga than in that species. Nylander also notes that a specimen from Colombia, Bogota, 2700 m., Lindig 887, has algal filaments 12-18 » in diam. Unfortunately I have been unable to see any of the above-mentioned specimens. The following description is based on Costa Rican specimens. Algae somewhat interwoven but more or less parallel, 16—20 (-28) u in diameter, forming a soft dimidiate or suborbicular thallus, not zonate, pale to deep olive buff, cells about 60-100 y. [Vor. 20 400 ANNALS OF THE MISSOURI BOTANICAL GARDEN long, septa plainly visible, chloroplasts rather indistinct in these specimens, often alternate cells more or less collapsing, giving a somewhat moniliform appearance to old filaments, outer walls thick, not densely covered with hyphae. The collapsed cells sometimes have the coloring matter collected at the ends, giving the dumb-bell appearance to the chloroplast as in C. subvirescens. Apothecia warm buff, 560 x in diameter, patelliform, with stipe about 160 u in diameter and 300 ų long, pseudoparenchymat- ous, of very thick-walled cells which tend to arrange their long diameters perpendicular to the surface of the apothecium, but not forming a definite palisade as in C. subvirescens; the para- thecium 40-50 x thick, homogeneous with the rest of the apo- thecium, the hypothecium 20-30 y. thick of interwoven hyphae, the hymenium about 60 u tall, composed of filiform paraphyses with spherical tips and slender cylindrical asci about 5 y in di- ameter; ascospores distichous, 4-celled, 10-12 x 3 u. [The asco- spores are immature and so closely packed in the ascus that I am not certain of size and septation, as I have been unable to free spores from the ascus.] Limón: Waldeck, Dodge 7453. Cartago?: Morpho Valley, 1420 m., K. Danielson 51 [This locality is uncertain and Danielson's barometrie readings are apt to be a little high, but it probably in the vieinity of the upper Reventazón below Cartago, where Lepidoptera of the genus Morpho are not uncommon.] COENOGONIUM HETEROTRICHUM Müll. Arg., Bull. Soc. R. Bot. Belg. 32: 162. 1893. Type: Costa Rica, San Marcos de Dota, Tonduz 6115. Thallus pulvinate, cespitose-effused, deep olive buff to dark olive, algae somewhat interwoven, dimorphic, larger filaments 20-25 y. in diameter, cells about 40 u long, chloroplasts disciform, scattered; smaller filaments submoniliform, the cells of smaller diameter being about 7-8 y. and 12 y. long, those of larger diameter being about 10-12 y. and of about the same length. Fungus unknown. In Brenes 29a the fungal hyphae are brown instead of hyaline. San José: S. Marcos de Dota, 1200 m., Tonduz 6115, type. Alajuela: La Palma de S. Ramón, 1250 m., Brenes 29a. COENOGONIUM INTERPONENDUM Nyl. ap. Polakowsky, Jour. Bot. Brit. & For. 15: 225. 1877. 1933] DODGE—LICHENS OF COSTA RICA. I 401 Type: Costa Rica, Cartago, Angostura, H. Polakowsky 496. Algae loosely interwoven, forming an adnate irregular thallus, deep olive buff to citrine drab, cells about 17-20 y. in diameter, 30-50 u long, pigment quite uniformly dispersed throughout the cell, yellowish green. All my Costa Rican material is sterile. Nylander states that the spores are fusiform, 2-celled, 8 x 2.5 y. Limón: Carmen, Dodge 7449. Cartago: Angostura, H. Polakowsky 496, TYPE. Alajuela: La Palma de S. Ramón, 1250 m., Brenes, 53a, 116, 3 Guanacaste: H. Q. Azul on lower slopes of v. Tenorio, 400- she m. l Dédye & Thomas 042. CoENOGONIUM INTERPOSITUM Nyl, Ann. Sci. Nat. Bot. IV, 16:91. 1862. Type: Borbonia, Lepervanche, Meziéres; Louisiana, Hale. Thallus loosely tomentose intertangled, more or less pulvinate, algal cells 20-28 u in diameter and 60-88 y long, chloroplasts appearing the shape of a dumb-bell with the pigment massing at the ends of the cells. [Apothecia pale carneo-luteous; spores oblong, simple, 6-10 x 3 u; paraphyses slender or medium.] As all my Costa Rican material is sterile the reference of it to this species is doubtful. It quite possibly belongs with C. disjunctum Nyl., based on specimens from Martinique and Cuba, Wright 170, which reaches a slightly larger diameter of algal filament and much larger 2-celled spores. Both species are said to resemble C. confervoides and perhaps should be placed in the C. confervoides group. Cartago: C. Carpintera, 1560-1700 m., Dodge & Thomas 7920. Alajuela: La Palma de S. Ramón, 1100-1250 m., Brenes 53, 115. COENOGONIUM INTERPLEXUM Nyl., Ann. Sci. Nat. Bot. IV, 16: 92, pl. 12, f. 20, 21. 1862. Type: Colombia, 2200 m., Lindig 2561. Thallus loosely interwoven, intricate, algal cells 12-16 » in diameter and 30 y long, chloroplasts more or less disciform, scattered. [Apothecia fleshy-yellow, plane, 1 mm. in diameter, margin waxy, fleshy to white; spores short-fusiform, 2-celled, 8-10 x 3 u; paraphyses medium, apex clavate.] Alajuela: Piedades de S. Ramón, 900 m., Brenes 408. Puntarenas: Boruea, 560 m., T'onduz 6114. [Vor. 20 402 ANNALS OF THE MISSOURI BOTANICAL GARDEN COENOGONIUM PANNOSUM Müll. Arg., Flora 64: 234. 1881. Type: Brasil, São Paulo, Apiahy, Puiggari 1026. Thallus loosely interwoven, dark olive buff to citrine drab, more or less adnate to substratum, algal cells 6-8 » in diameter, 28-32 y long, pigment migrating toward the ends of the cells. [Apothecia 350-500 y. in diameter, pale or orange white, plane, with paler and thinner margin, finally immarginate and slightly convex, becoming more flesh color; hymenium hyaline, para- physes very slender with obovoid head about three times as thick as the stalk; asci slender, cylindrical, 8-spored; spores 2- locular, 6-7 x 2 y, fusiform with acute ends.] Since my Costa Rican material is sterile its reference here is uncertain and it may be only very young algae of C. subvirescens Nyl., with which it agrees in general appearance, but there is no trace of the formation of a dimidiate thallus. Guanacaste: H. Q. Azul, on slope of V. Tenorio, 500-600 m., Dodge & Thomas 6654. COENOGONIUM DEPRESSUM Müll. Arg., Flora 64:525. 1881. Type: Brasil, São Paulo, Apiahy, Puiggari 1034. Thallus adnate, pale olive buff, filaments short, flexuous, depressed, subintricate; cells 4-8(-13) y in diameter, 16-20 y long, somewhat inflated in the middle but scarcely enough to give a moniliform appearance, chloroplasts no longer distinct. As I have not compared the Costa Rican material with the type I cannot be certain of the identity. Cartago: alpine region of Irazüá, Ørsted. COENOGONIUM IMPLEXUM Nyl., Ann. Sci. Nat. Bot. IV, 16: 92. 1802. Type: Australia, Victoria, Jarvin, Ferd. M üller. [Similar to C. interplexum but algal cells a little smaller, 10-13 y in diameter, spores a little larger, 8-11 x 3.5-4.5 u, paraphyses thicker.] The following collection has been referred here by Müller Argau, but I have not studied it. Puntarenas: Boruca, 560 m., T'onduz 6118. EPHEBACEAE Thallus dwarf fruticose, branched, more or less filiform, without rhizinae, crustose or small squamose, with Scytonema 1933] DODGE—LICHENS OF COSTA RICA. I 403 or Stigonema. Apothecia small, often with very small puncti- form disc, scarcely visible; paraphyses well developed or absent; asci 8-spored; spores hyaline, 1—2-celled. The individuals of this family are so small that they are easily overlooked both in collecting and in sorting material, and so the few species here reported are probably not representative of this group in Costa Rica. KEY TO TROPICAL AMERICAN GENERA OF EPHEBACEAE Thallus crustose to small squamose, homoeomeric; Brasil........ Pterigyopsis atra Thallus dwarf fruticose, dark, thickly branched. Apothecia sunken in swellings of the thallus, single or gregarious; spores unicellular; paraphyses present; Brasil.....................lsuu. Ephebeia Apothecia sessile on the thallus, either lateral or terminal. Thallus without pseudoparenchymatous cortex or medulla; paraphyses filiform, simple; asci 8-spored; spores ovoid or spherical, unicellular Fu M cnc De TIME ce. UE Ade Thermutis Thallus with pseudoparenchymatous cortex and medulla. PENNE hv e or REP MISSAE EPA REND Leptogidium byssoides Loc 2 bono. C oe 585s 0 OS Rh a Oe ea 0 ee Polychidium THERMUTIS E. Fr. THERMUTIS E. Fr., Syst. Orb. Veg. 1: 392. 1825. Gonionema Nyl., Mem. Soc. Sci. Nat. Cherbourg 3: 163. 1855. Thallus dwarf fruticose, thickly branched and filiform, without rhizinae, Scytonema present with the hyphae running in the gelified sheath. Apothecia small, lateral, saucer-shaped to almost spherical; parathecium often highly developed and thick; hypothecium light-colored; paraphyses unbranched, filiform, tips not swollen; asci clavate, thin-walled, 8-spored; spores hyaline, ellipsoidal, elongate, unicellular, thin-walled. Sperma- gonia lateral or terminal, sessile, more or less spherical; spermatia small, ovoid, or elongate. THERMUTIS VELUTINA (Ach.) Fw., Linnaea 23: 170. 1850. Lichen velutinus Ach., Lichenog. Supe: Prodr. 218. 1798. Gonionema velutina Nyl., Act. Soc. Linn. Bordeaux 21: 262. 1856. This species has been reported by Müller Argau from the following localities in Costa Rica. The specimens were not studied while I was in Genève. San José: San Marcos de Dota, 1200 m., Tonduz 5378. Guanacaste: Boruca, 560 m., T'onduz 5371; Térraba, Tonduz 5372. [Vor. 20 404 ANNALS OF THE MISSOURI BOTANICAL GARDEN A small sterile specimen from Guanacaste: near Tilarán, 500- 690 m., Dodge & Thomas 8081, may belong in Polychidium but the specimen is not definitely determinable as to genus. COLLEMACEAE Thallus gelified, crustose to foliose or dwarf fruticose, with or without rhizoids, sometimes umbilicate, homoeomerous with Nostoc. Apothecia from sunken almost perithecia to sessile apothecia usually with an amphithecium, occasionally with parathecium; paraphyses simple; asci 8-spored; spores hyaline, spherical to acicular, straight or twisted, 1-celled to muriform, usually with a thin wall (except in Physma and section Lem- phospora of Lempholemma). This family is widely distributed in the temperate and tropical zones, the more highly developed members occurring in the latter. It is possible that some of the smaller and less conspicuous genera have been overlooked or have been included in the crustose material in the preliminary sorting. So far only four genera have been found in Costa Rica. KEY TO TROPICAL AMERICAN GENERA OF COLLEMACEAE Spores unicellular. Thallus crustose, not gelified, parathecium present; spores ellipsoidal, thin-walled. Brasi FORD REITI TERI ER x Wa REA FEX EXXTAMESVEUSURHS C cats L. Finkii Thallus squamulose, gelified; spores soberiol, thin-walled but surrounded by a thick vire sheath. Thallus not corticate.......... 0. ccc ccc cece cece cc eceaseues Lempholemma Thallus corticate, wholly pseudoparenchymatous; Juan Fernandez ECER TEE EE IE a on EE eee Lemmopsis PER TETT E Thallus foliose, gelified; spores ellipsoidal to fusiform, thick-walled, h surrounded by a gelified sheath.............. 0.000. essere Physma Spores phragmospores or dictyospores. Cortex not developed; apothecia with parathecia only................ Collema Cortex of pseudoparenchyma; apothecia with amphithecia (parathecia often also present)............ 0c ccc ccc c cece cece hr eptogium LEMPHOLEMMA Körb. LEMPHOLEMMA Körber, Syst. Lich. Germ. 400. 1855. Type species: Lempholemma compactum Körber. Thallus from verrucose, squamulose, dwarf fruticose to foliose in our species, gelified when moist, clothed with rhizinae below, 1933] DODGE—LICHENS OF COSTA RICA. I 405 homoeomerie, without cortex, with Nostoc. Apothecia superfi- cial or terminal, mostly sunken in the thallus (not in our species) ; amphithecium either with or without cortex; parathecium color- less, urceolate; hypothecium colorless; paraphyses filiform, simple; asci 8-spored, clavate, often twisted below; spores hyaline, fusiform, ellipsoidal-ovoid or spherical, smooth. The genus has been divided into eight sections of which only one section Lemphospora is tropical. This section is characterized as having apothecia without cortex, spores more or less spherical with a thick gelified sheath. Lempholemma Dussii (Vainio) Zahlbr., Cat. Lich. Univ. 3: 23. 1925 [Collema Dussii Vainio, Ann. Acad. Sci. Fenn. A6': 114. 1915] has been described from Guadeloupe in the Antilles. LEMPHOLEMMA (LEMPHOSPORA) oblique-peltatum (Eschw. ap. Martius) Dodge, comb. nov. Collema oblique-peltatum Eschw. ap. Martius, Icon. Pl. Cryptog. Fasc. 2: 27, pl. 11, f. 2 [between 1828 and 1833]. Type: Brasil, near Pará, Martius. Thallus foliose, up to 5 cm. in diameter, drying grayish olive to dark olive gray, margin thick, broadly subdichotomously lobed, lobes up to 3 mm. broad and 2-3 mm. long, smooth, surface minutely pitted although appearing practically smooth to the naked eye, without soredia or isidia; thallus 200—400 y. thick, homoeomerous, without cortex, but provided with a thin layer of tangled rhizinae below. Apothecia subsessile, often appearing obliquely attached to the thallus in section, up to 2.5 mm. in diameter, margin of the same color as the thallus, with periclinal folds and wrinkles, disc tawny to mars brown; amphithecium appearing lobulate in section (due to the periclinal wrinkles), homogeneous with the thallus, 200-300 y thick but extending beyond the parathecium as much as 800 u; para- thecium pseudoparenchymatous, highly developed, about 100 yu thick below, thinning out above to about 50 y; hypothecium about 40 y. thick, of highly gelified hyphae; thecium about 100 y. 1 I have been unable to locate the date of the second fascicle of this work. The cover of the copy in the Missouri Botanical Garden gives the pages of text and plates in each fascicle but only the dates 1828-1834. Martius in his ‘Flora Brasiliensis’ 1: 233-234. 1833, quotes from the ‘Icones’ citing page, hence it must have been issued before that date. In this work he placed it in the subgenus Enchylium. [Vor. 20 406 ANNALS OF THE MISSOURI BOTANICAL GARDEN tall; paraphyses about 1-2 y. in diameter, clavate above, septate, cells 4-6 u long, with walls highly gelified so that they appear as slender rows of cells imbedded in a gel; asci clavate-cylindrical, about 80 x 16 u, containing 8 monostichous spores; ascospores broadly ellipsoidal, 12 x 8 y. while still in the ascus, with a thick gelified wall. While I have not studied the type specimen of this species, a careful study of the description and the figures shows that it belongs in Lempholemma section Lemphospora rather than in Collema. Judging from determinations observed in various herbaria, this species seems to have been confused with Lepto- gium vesiculosum (L. bullatum), L. tremelloides, and L. foveolatum, although the abundant development of rhizoids below should have separated it. While it is minutely scrobiculate-foveolate (visible under hand-lens), it is not conspicuously so to the naked eye as in L. foveolatum. Besides my Costa Rican material I have also seen it from Eustis, Lake Co., Florida, Nash 2024, determined by Eckfeldt as L. faveolatum (sic) Nyl. Limón: Waldeck, Dodge 7404; near Siquirres, 70-170 m., Dodge, Catt & Thomas 5588, 8024; Hamburg, 55 m., Standley & J. Valerio 48767. Guanacaste: H. Santamaría, = m., Dodge & Thomas 6898; near Tilarán, 340- 670 m., Dodge & Thomas 6644, 80 Puntarenas: near Corozál, dd Dodge 8023; Puerto Jiménez, Brenes 839. LEMPHOLEMMA (LEMPHOSPORA) dichotomum Dodge, sp. nov. Type: Costa Rica, Guanacaste, H. Granadilla, Dodge & Thomas 6736. Thallus foliosus, ad 5 em. diametro, 400-650 y crassitudine metiens, griseo-olivaceus, margine crassiuscula, integra, lobatus, dichotomus, lobi 1-1.5 mm. lati, divergentes, superne longi- tudinaliter rugosus, sine sorediis isidiisque, inferne rhizinis nigris densissimis intertextis obsitus, homoeomerus, nostocaceus, sine strato corticali. Apothecium peltatum, basi constrictum, ad 2.5 mm. diametro metiens, margine concolori, rugosa, disco fulvo castaneove; amphithecium sectione lobatum, homogeneum, 150-300 y crassitudine inferne, sed ad 500 y superne et ultra marginem parathecii; parathecium pseudoparenchymatieum, ad 160 uw crassitudine inferne, attenuatum ad 80 y superne; hypothecium filamentosum, ad 80 y crassitudine, hyphis geli- factis; thecium ad 120 y altum; paraphyses filiformes, apicibus 1933] DODGE—LICHENS OF COSTA RICA. I 407 decompositis brunneis, in materia gelata fixi, cellulis cylindricis, l a diametro, 4-6 y longitudine metientibus; asci 60 x 10 y, cylindriei; ascosporae octonae, monostichae, late ellipsoideae, 8 x 12 y. (dum in ascis sunt), membrana incrassata, gelifacta. Thallus foliose, up to 5 cm. in diameter, 400-650 u thick, grayish olive, margin thick, smooth, dichotomously lobed, lobes 1-1.5 mm. broad, divergent, upper surface longitudinally wrinkled, without isidia or soredia, lower surface covered with a dense black nap of rhizinae, homoeomerous, with Nostoc, without cortical layer either above or below. Apothecium constricted at the base, up to 2.5 mm. in diameter, margin concolorous, irregularly wrinkled, dise tawny to chestnut; amphithecium in sections appearing lobed, homogeneous with the thallus, 150- 300 u thick below but extending up to 500 u above and beyond the margin of the parathecium which is pseudoparenchymatous, about 160 » thick below, thinning out to 80 y at the margin; hypothecium filamentous, 80 » thick, of gelified hyphae; thecium about 120 u tall, paraphyses filiform, the apices decomposing brown, imbedded in a gel, cells cylindric, about 1 y in diameter, 4—6 y. long; ascospores 8 per ascus, monostichous, broadly ellip- soidal, about 8 x 12 y. (while still in the asci), with a thick gelified wall. This species has also been seen from Barro Colorado Island, Gatun Lake, Panamá. Guanacaste: H. Granadilla, 540 m., Dodge & Thomas 6737. PHYSMA Mass. PuysMA Massalongo, Neag. Lich. 6. 1854. Dichodium Nyl., Bull. Soc. Linn. Normandie II, 2: 43. 1868. Type: Physma Boryanum Massalongo. Thallus foliose with rhizinae beneath, corticate with several layers of pseudoparenchymatous cells; algae Nostoc. Apothecia superficial, lecanorine, with broad dise, thick margin; hypo- thecium light-colored; paraphyses filiform; asci 8-spored; spores colorless, ellipsoidal or fusiform, unicellular with thick almost warty wall or with a gelified sheath. Spermagonia sunk in the thallus, showing above by the dark swelling, surrounded by pseudoparenchyma; spermatiophores simple or forked, septate, cells short; spermatia short, straight. [Vor. 20 408 ANNALS OF THE MISSOURI BOTANICAL GARDEN Only two species are known from tropical America, P. chilensis Hue from Chile and P. pruinosum Vainio, from the Antilles. Material reported from America as belonging to P. byrsinum (Ach.) Müll. Arg. (P. byrsea (Ach.) Tuck.) is probably mis- determined. Apparently the genus is most highly developed and wide-spread in Oceania and adjacent Asia. PHYSMA PRUINOSUM Vainio, Ann. Acad. Sci. Fenn. A67:112. 1915. A single sterile specimen collected near Cartago, 1500 m., C. Wercklé, May 1900 (in Mus. Nac. 17233, and in Herb. Bot. Gard. Berlin), was referred to Physma byrsinum by Lindau. It is possible that this species belongs to Physma pruinosum Vainio or it may possibly be a very young thallus of Leptogiwm marginel- lum (Sw.) S. F. Gray. Its spermagonia are marginal and its spermatia are ellipsoidal and small. COLLEMA Wigg. CornLEMA Wiggers, Primit. Fl. Holsat. 89. 1780. ? Gabura Adanson, Fam. Pl. 2:6. 1763? Scylenium S. F. Gray, Nat. Arrang. Brit. Pl. 1: 398. 1821. Type species: Collema Lactuca (Web.) Wiggers [Lichen crispus L.]. Thallus foliose or squamulose to almost crustose, gelified when moist, lying on the substrate without rhizinae, homoe- omerous, not corticate, hyphal system loose; algae Nostoc. Apothecia at first sunken, erumpent, sessile or scutellate and constricted below, with amphithecium; parathecium either present or absent, both parathecium and hypothecium either of interwoven hyphae or pseudoparenchymatous; paraphyses simple, adherent, mostly septate; asci 8-spored; ascospores colorless, cylindric, acicular, fusiform, long-ellipsoidal or ovoidal to almost cubical, ends obtuse or acute, sometimes becoming muriform, thin-walled, without gelified sheath. Spermagonia sunken in the thallus or in thalline warts, with light-colored wall; spermatiophores simple or branched, without sterigmata, septate with short cells; spermatia short, oblong to ellipsoidal, straight. KEY TO TROPICAL AMERICAN SPECIES OF COLLEMA Amphithecium corticate; spores acicular........... isses COLLEMODIOPSIS [C. lherminieri Hue, from Guadeloupe, is the only species of this typically northern group so far reported. C. Granadillae is here reported from Costa Rica.] 1933] DODGE—LICHENS OF COSTA RICA. I 409 Amphithecium not cort Spores oblong to amet more or less muriform; thallus laciniate, Tore Or NM EEUE. iL soos Sivan eR ern nen sete BLENNOTHALLIA Lobes n than 0. 5 mm. broad; apothecia unknown; Chile....... C. millegranum Lobes more than 0.5 mm. broad; apothecia well developed; Mexico. M rr NEMESIS C. masioandes TE clavate, ins esae to fusiform, siguen rarely up to 3-septate, thallus narrowly laciniate or incised...... DICOLLEMA nd clavate, 14-21 x 4.5-5.5 u; Brasil, soon , and Argentina.... mH m TIN. C. corynesporum Spores ellipsoidal or fusiform. Margins of laciniae nodose-granular................00e00 00s C. pycnocarpum po. PLI p; Brasil. orii oh oe bk na ees o rr v. Minarum Sporos 10-44 x 6-4 u; Brasil oi ces se. Ace ei ee pas One v. Malmei Spores 10-15.5 x 4-5.5 u; Brasil and Colombia.......... v. crassiusculum Spores 14-17 x 4-5 u; New York and New Jersey............ v. typicum Spores 16-25 x 3-7 u, 3-septate; Virginia southward.......... C. cyrtaspis Spores 16-30 x 4-5 yu, 1-3-septate; S Pr be EM C. pycnocarpoides Margins of laciniae not nodose-granula Laciniae 1-3 mm. broad, spores 12- 14 x 5-6 u; Paraguay...... C. crenatum Laciniae under 1 mm. broad, more or less canaliculate below; Cuba. pe gt a ee en erm n ar ROSE . stellatum Laciniae thin, flat; spores 11-17 x 4-5 w................ C. solenarium Spores fusiform to acicular, more than 3-septate, over 30 » long; thallus broadly laciniate, often fenestrate...............00. SYNECHOBLASTUS Thallus isidiose; spores over 100 x in length; Brasil Parathecium present, spores 125-175 x 8-4 u........... sun. d rade Parathecium absent, spores somewhat over 100 nu inlength........ C. Thallus not isidiose; spores not over 95 y in length. Apothecia white-pruinose. Apothecia small, crowded on short bullate prominences; spores 0-80 x 3-5 wu; thallus pustulate, not fenestrate; southern United States and southward along coastal plains. ...... E OS ae sessile or nearly so, larger; thallus more or less fen 6; tropical highlands. ELDER ese C. Howophtilmsum Piat 77-92 x 624 iy MOXIOO s a a a a eese typicum Spores 55-74 x 5-7 m, 7-1l-septate; Colombia............ v. pease Spores 42-58 x 4-5 u, 6-9-septate; Brasil................ v. brasiliense Spores 36-50 x 7 p, only 5-septate. ............... sss. C. leucocarpum Apothecia not white-pruinose. Spores 48-66 x 6-8 u; Colombia............... eee eee C. implicatum Spores 38-50 x 6-8 p; Chile... REEL ahs ss hee tases C. chilenum E 40-45 x 5-6.5 u; apothecia sessile, margins turgid; Mex- E vier s UE O Gea lk tol Rae Sos o eB C. turgidulum bas, 30-37 x 2.5-3 pu; apothecial margins thin, blackening; DIOR o eonun ere aa ee . baculiferum CoLLEMODIOPSIS Vainio, Étude Lich. Brésil 1: 234. 1890. Type species: Collema nigrescens (Huds. [Vor. 20 410 ANNALS OF THE MISSOURI BOTANICAL GARDEN Spores slender, several-celled phragmospores, never muriform, apothecia with a pseudoparenchymatous cortex. This subgenus is typically northern in its distribution. Only C. lherminieri Hue from Guadeloupe has been described previ- ously from the American tropics. CoLLEMA (CoLLEMopIopsis) Granadillae Dodge, sp. nov. Type: Costa Rica, Guanacaste, H. Granadilla, Dodge & Thomas 6576. Thallus irregulariter subpinnatifide lobatus, lobis tenuibus, ad 1 mm. latis, viridi-nigricans, superne elevato-rugosus verruco- susque ad subisidiosus, inferne reticulatim elevato-rugosus, cinerascens, ad 400 y crassitudine, decorticatus, rhizinis desti- tutus. Apothecium planum, peltatum, basi constrictum, 0.5- 1.0 mm. diametro metiens, omnino nigrum, margine verrucosa; amphithecium 150-250 yu crassitudine, pseudoparenchymatice corticatum, intus algis nostocaceis subrectis; parathecium grosse pseudoparenchymaticum, bene evolutum, 100—120 y crassitudine inferne, attenuatum ad 25-35 y superne; hypothecium hyphis tenuibus dense contextum, 25-30 y crassitudine; thecium 150- 170 y altitudine, paraphyses arcte cohaerentes septati, cylindrici, filiformes, ad 1 u diametro metientes, clavato-capitati; ascosporae octonae, polystichae, rectae, fusiformes vel aciculares, 5-septatae, 97-42 x 4-5 y. Thallus irregularly subpinnately lobed, lobes slender, about 1 mm. broad, dark greenish black, with elevated folds and wrinkles above with masses of small subisidiose warts, below reticulately deeply scrobiculate wrinkled, more or less ashy, about 400 y thick, without cortex or rhizoids. Apothecium peltate with constricted base, plane, about 0.5-1 mm. in diameter, wholly black, margin verrucose; amphithecium 150-250 y thick, corticate with several layers of pseudoparenchyma below, within the filaments of Nostoc more or less straight and loosely inter- woven; parathecium of large-celled pseudoparenchyma, about 100-120 y thick below, thinning above at the margin to 25-35 y; hypothecium of densely woven slender hyphae, 25-30 y thick; thecium 150-170 u tall; paraphyses closely adherent, septate, cylindrical, filiform, about 1 u in diameter, clavate, capitate at the apex; ascospores 8 per ascus, polystichous, straight, fusiform or acicular, 6-celled, 37—42 x 4-5 u. 1933] DODGE—LICHENS OF COSTA RICA. I 411 Guanacaste: H. Granadilla, Dodge & Thomas 6576; near Tilarán, Standley & J. Valerio 44529, 640—660 m., Dodge & Thomas 6560. CoLLEMA LEUCOPEPLA (Tuck.) Schneider, Guide Study Lich., 181. 1898. Collema nigrescens var. leucopepla Tuck., Syn. N. Am. Lich. 1:148. 1882. Type: not stated in original description, based on material from South Carolina, Georgia, Florida, Alabama, and Louisiana. It is probable that material collected in Puntarenas, Boruca, 560 m., Tonduz 5373, and determined by Müller Argau as C. nigrescens var. caesium Ach., belongs here but I did not have time to study it critically while I was in Genéve. This species of the lower Atlantic and Gulf Coastal Plain in the United States has also been reported by Vainio from Mexico. DICOLLEMA (Clements) Dodge, n. m ^ Dicollema Clements, Gen. Fung. 74. Type species: Collema pycnocarpum Ny i Thallus narrowly laciniate or incised, spores clavate to fusiform, typically 2-celled, rarely 4-celled. The material referred to the type species of this genus by different authors is quite variable and the whole group needs a thorough revision. Pending such a study, I am proposing several varieties of C. pycnocarpum Nyl. to cover the more conspicuous variations which seem to have separate geographical ranges, those with the smaller spore sizes being rather more southern in their distribution. COLLEMA PYCNOCARPUM Nyl., Syn. Meth. Lich. 1: 115. 1858. Type: United States [on journey from New York to Philadel- phia], Moré. Thallus dark or pale green, medium in size, granulate-nodose. Apothecia rufous, crowded, almost contiguous, at first plane, then somewhat convex, quite small, 0.5 mm. or a little larger; spores 8 per ascus, oblong or oblong-ellipsoid, simple or uniseptate, 14-17 x 4-5 y. Northern Atlantic Coastal Plain. Var. Minarum Dodge, var. nov. Collema pycnocarpum Vainio, Etude Lich. Brésil 1: 238. 1890. Type: Brasil, Minas Geraes, Sitio, 1000 m., Vainio 734. [Vor. 20 412 ANNALS OF THE MISSOURI BOTANICAL GARDEN Sporae oblongae, ellipsoideae vel fusiformes, 1-septatae, 6-0 x 3 y. Var. Malmei Dodge, var. nov. C. pycnocarpum Malme, Ark. f. Bot. 198: 7. 1924. Type: Brasil, Rio Grande do Sul, Canóas near Porto Alegre, Malme 536. Sporae oblongo-ellipsoideae vel ellipsoideae, uniseptatae, 10— 14 x 3-4 y. Var. crassiusculum (Malme) Dodge, comb. nov. Forma crassiusculum Malme, Ark. f. Bot. 198: 7. 1924. Collema pycnocarpum Nyl., Acta Soc. Sci. Fenn. 7:428. 1863. Type: Brasil, Matto Grosso, Corumba, Malme. Sporae ellipsoideae, utroque apice obtusae, raro acutae vel altero obtusae, altero acutae, 10-15.5 x 4-5.5 y. The specimens cited by Nylander from Colombia, Bogotá, 2600 m., Lindig 2872, have spores 10-14 x 4.5-5.5 u. COLLEMA CYRTASPIS Tuck., Proc. Amer. Acad. Arts & Sci. 5:387. 1802. Type: no type mentioned in the original description. Thallus deeply and irregularly lobed, verrucose, deep blue green, darkening on drying, verrucae much larger than in C. pycnocarpum, about 400 y thick; algal layer about 100 y thick, composed of coiled and tangled filaments of Nostoc imbedded in a gel, cells spherical, about 4 u in diameter; medulla gelified, traversed by very loosely tangled hyphae about 2-3 y in diameter and occasional straight filaments of Nostoc. Apothecia immersed in the verrucae when young, becoming plane or even convex, up to 2 mm. in diameter, margin slightly verrucose when young, practically disappearing as the apothecium becomes increasingly complex; amphithecium about 160 » broad, of the same texture as the thallus; parathecium lacking; hypothecium filamentous, quite highly developed, about 40 y thick, not conspicuously thinner toward the margins, hyphae 2-3 » in diameter, very loosely woven; thecium about 80 y tall; paraphyses filamentous, clavate, expanded above, imbedded in a gel; asci clavate, about 40 x 8 u, 8-spored; ascospores subfusiform, 2-4-locular, 16-25 x 3-7 u [16-20 x 4-4.5 u in Costa Rican specimens]. This species is widely distributed in the southern Atlantic 1933] DODGE—LICHENS OF COSTA RICA. I 413 and Gulf Coastal Plain in the United States, extending up the Mississippi Valley to Iowa. Guanacaste: Liberia, 100 m., Dodge & Thomas 8016, 8027. SYNECHOBLASTUS (Trevis.) Vainio, Étude Lich. Brésil 1: 234. 1890. Lathargium S. F. Gray, Nat. Arrang. Brit. Pl. 1: 399. 1821, p. p. Synechoblastus 'Trevis., Caratt. Tre Nuov. Gen. Coll. 3. 1853, p. p. Type species: as subgenus based on C. glaucophthalmum N yl. Apothecia without pseudoparenchymatous cortex, spores fusi- form to more or less acicular, many-celled, mostly over 30 y long. This subgenus is typically northern. Only two species are known from Costa Rica, where they are temperate species at elevations between 1000 and 1800 m., and are also found in the highlands of Colombia at somewhat greater altitudes. COLLEMA GLAUCOPHTHALMUM Nyl., Syn. Meth. Lich. 1: 114, 115. 1858. Type: Mexico, Orizaba, Fr. M üller. Thallus olivaceous-fuscous, medium size, expanded, more or less fenestrate and dissected, scrobiculate, and often granuliferous. Apothecia glauco-lilac colored, plane and somewhat concave, margin prominent, thin; ascospores 77-92 x 6-7 y. Var. GRANATENSE Hue, Jour. de Bot. [Morot] 20:12. 1906. Collema glaucophthalmum Nyl., Acta Soc. Sci. Fenn. 7: 428. 1863, non loco alio. Type: Colombia, Choachí, 2600 m., Lindig 813. Thallus dusky yellowish green to almost black, irregularly lobed and fenestrate, smooth or somewhat scrobiculate, coarsely wrinkled and subverrucose, especially toward the tips, margins thick and rounded, 800-1000 u thick, homoeomerous, of loosely tangled filaments of Nostoc, cells 5-6 x 2.5-3 y, ellipsoidal. Apothecia peltate, constricted at the base, up to 3 mm. in di- ameter, margin at first prominent, smooth, becoming thinner, less prominent and verrucose at maturity; disc chalky white to vinaceous-russet; amphithecium about 160 y. thick, homogeneous with the thallus; parathecium pseudoparenchymatous, 40 u thick, thinning to 20 u or even disappearing above at the margin; [Vor. 20 414 ANNALS OF THE MISSOURI BOTANICAL GARDEN hypothecium filamentous, of large very densely interwoven hyphae about 40 y thick; thecium about 120 y tall; paraphyses filiform, about 1 u in diameter, not swollen above; asci clavate, about 20 y in diameter, 8-spored; ascospores polystichous, fusiform to acicular, 55-74 x 5-7 u. The whiteness of the apothecial dise is very variable on the same thallus, in general being more pronounced on young and rapidly growing apothecia and gradually disappearing on older and more exposed ones. It seems to be a semi-crystalline deposit which slowly dissolves away, disclosing the vinaceous- russet disc formed by the discoloration of the upper portion of the gel surrounding the paraphyses. This variety has previously been reported only from Colombia. Cartago: near R. Birrfs above Santiago, 920-1340 m., Dodge 8011, Dodge & Thomas 8015; Carpintera, 1700 m., K. Danielson 102. San José: Sta. María de Dota, 1500-1800 m., Standley & J. Valerio 44150. CoLLEMA IMPLICATUM Nyl., Acta Soc. Sci. Fenn. 7: 428. 1863. Type: Colombia, Villeta, 1100 m., Lindig 749; Cundinamarca, Bogotá, 2400-2600 m., Lindig. Thallus dark greenish black, broadly lobed and fenestrate, smooth near margins, becoming deeply reticulate-scrobiculate in older portions of the thallus, verrucae scattered, more elevated, often subisidioid, margins thin and semipellucid, homoeomerous, of loosely tangled filaments of Nostoc, cells 4 x 5 u, heterocysts 5x8 y, much more closely tangled in an outer zone about 150 y thick. Apothecia peltate, constricted at the base, up to 4 mm. in diameter, margin at first prominent, verrucose, becoming thinner, less prominent, scarcely visible as the apothecium becomes expanded and convex, dise cameo-brown, without pruina in young apothecia; amphithecium about 100 p thick, homogeneous with the thallus; parathecium pseudoparenchy- matous, about 80 à thick below, thinning out to a single layer of thick-walled cells 6 & wide above at the margin; hypothecium of densely woven hyphae about 20 u thick; thecium about 100 y tall; paraphyses about 2 y in diameter, with clavate brown tips; asci clavate, 8-spored, 12-16 y in diameter; ascospores poly- stichous, fusiform to acicular, 48—66 x 6-8 u, 7—9-septate. This species is very closely related to C. glaucophthalmum 1933] DODGE—LICHENS OF COSTA RICA. I 415 and has been reduced to synonymy by some authors. It differs from the previous species in several minor details of proportion of measurements and in the lack of a well-developed pruina. It is to be hoped that some one will be able to study the two forms in the field to determine the constancy of this character. In a series of specimens one can find considerable variation in the amount of pruina in C. glaucophthalmum. It should be noted that the two species occupy the same areas in Costa Rica. Cartago: above R. Birrís at Santiago, 1140-1180 m., Dodge 8017; Carpintera, 1700 m., K. Danielson 101c. The following species has not been found in Costa Rica but it is closely related to this group. COLLEMA (SYNECHOBLASTUS) Ramboi Dodge, sp. nov. Type: Brasil, Rio Grande do Sul, Porto Alegre, B. Rambo 74. Thallus parvus, adscendens, obscure viridi-nigricans, angusti- lobatus, fenestratus clathratusque, verrucis elevatis isidioideis, marginibus plus minusve integris, tenuibus, pellucidis, 130-140 y crassitudine, homoeomerus, filamentis nostocaceis laxe implexis, in zona exteriori 40 y crassitudine dense contextis, cellulis ad 3x6 yu. Apothecium peltatum, basi constrictum, planum, 0.5-0.6 (— 1.0) mm. diametro, margine tenui, verrucosa sub- crenulatave, disco obscure castaneo nigroque; amphithecium 20 p crassitudine, cum thallo homogeneum; parathecium deest; hypo- thecium 20 u crassitudine, hyphis tenuibus dense contextum; thecium 160 y altitudine; paraphyses filiformes, ad 1 y diametro, apicibus inflatis; asci clavati, 120 x 8-10 »; ascosporae octonae, polystichae, aciculares, immaturae, plus quam 100 y. longitudine, multiloculares. Thallus small, elevated, dark greenish black, narrowly lobed and fenestrate, verrucae elevated and isidioid, margins more or less smooth, thin, pellucid, 130-140 u thick, homoeomerous, of loosely tangled filaments of Nostoc, cells about 3x 6 y, much more closely tangled in an outer zone about 40 y thick. Apothecia peltate, constricted at the base, 0.5-0.6 (- 1.0) mm. in diameter, plane, margin very thin, verrucose, subcrenulate, finally almost disappearing while the expanded apothecium remains plane, dise dark chestnut to black; amphithecium 20 y broad, homo- geneous with the thallus; parathecium absent; hypothecium 20 y. [Vor. 20 416 ANNALS OF THE MISSOURI BOTANICAL GARDEN thick, of slender densely woven hyphae; thecium about 160 u tall; paraphyses filiform, about 1 y in diameter with greatly nidi tips, imbedded in à hymenial gel; asci clavate, 8-spored, 120 u long, 8-10 u wide; ascospores polystichous, 8 per ascus, acicular, immature and very difficult to measure but somewhat more than 100 u long, many-celled. LEPTOGIUM 5$. F. Gray Leproacium S. F. Gray, Nat. Arrang. Brit. Pl. 1: 400. 1821. Type species: Leptogiwm tremelloides S. F. Gray, excl. syn. Thallus foliose in the tropical species, gelified, below naked (covered with rhizinae in section Mallotium), with pseudo- parenchymatous cortex above and below, algae Nostoc. Apo- thecia sunken in the thallus at first, emerging and often short- stipitate; amphithecium often corticate with several layers of pseudoparenchyma below, usually of a single layer above; parathecium usually present and pseudoparenchymatous, often thinning out above at the margin, sometimes of large, more or less parallel hyphae; paraphyses simple and filiform, often with apex variously thickened; asci clavate to cylindrical, 8-spored; ascospores usually imbricately monostichous, occasionally distich- ous, hyaline, usually broadly fusiform with acute to acuminate ends (acicular in section Leptogiopsis), usually muriform (except in Leptogiopsis), with thin walls. Four of the seven sections into which Leptogium is divided are found in tropical America, all four of them in Costa Rica. Lep- togiopsis differs from all the other sections in its acicular, never muriform spores, and perhaps should be treated as a separate genus, more or less related to section Synechoblastus of Collema. It is endemie in tropical America, with a single unidentifiable fragment yet found in Costa Rica. Mallotiwm, to which may also be referred the section Leptolobaria of Vainio based on L. callithamnium from Chile and the Antilles, has highly developed rhizinae below, and often more or less tomentum above, especially in the vicinity of the apothecia. It is widespread in its distribu- tion, being found in the temperate zones and evidently preferring the colder regions of the tropies, not occurring below about 1300 m., in Costa Rica not seen below 3000 m. Diplothallus, with 1933] DODGE—LICHENS OF COSTA RICA. I 417 its two separate layers of thallus, each with its own cortex, connected by pillars of pseudoparenchyma which on drying cause punctate depressions in the thallus, is endemic in tropical America and confined to the temperate regions, in Costa Rica occurring between 1200 and 1700 m., coming down to lower levels in the mountain near H. Santamaria of the Cordillera de Tilarán in Guanacaste. Euleptogium, which contains the most species, is found from sea level to about 1800 m. in Costa Rica, although most of the species of this section have a narrower altitudinal range. The lowland species ascend the river valleys but mostly drop out at about 1000 m. These are usually wide- spread at low elevations, extending from the Atlantic and Gulf Coastal plains of the United States to southern Brasil and Paraguay. The species of the temperate region of Costa Rica are mostly confined to the mountains from southern Mexico to the northern Andes and the mountains of eastern Brasil. The morphology of the apothecium is not altogether clear and considerable confusion exists in the nomenclature of the various parts. In the following discussion, I have treated the tissue, usually well differentiated, underlying the hypothecium, as the parathecium, whether it extends to the surface of the thecium or not and whether or not it is pseudoparenchymatous. In a very few species the parathecium as thus defined thins out and disappears near the edge of the thecium, and the hyphae of the hypothecium extend up to the surface of the thecium. It is possible that some would prefer to regard the parathecium as formed of two layers, the lower of which is pseudoparenchymatous and the upper filamentous, with only the upper extending to the surface of the thecium. The amphithecium (thalline margin of earlier lichenologists) is essentially homogeneous and continuous with the thallus, although the algae are sometimes more densely tangled and in many species the cortical cells on the under side of the amphithecium (morphologically continuous with the upper surface of the thallus) or even the under side of the thallus below the apothecium, may proliferate, forming a pseudoparen- chyma of several layers of cells sometimes even thicker than the algal zone. [Vor. 20 418 ANNALS OF THE MISSOURI BOTANICAL GARDEN KEY TO THE TROPICAL AMERICAN SPECIES OF LEPTOGIUM Spores acicular, never mu Thallus efiulateseroieulate; French Guiana. [Spores 6-celled, 35 x 10 y, teste Leighton, in specimen from Amazonas, Brasil].......... L. reticulatum Thallus far reticulate-scrobiculate. Spores 48 u long, 8-10-celled; Florida.......... L. fusisporum (Tuck.) Dodge Spores 50-60 x 7-8 u, 10-12-celled; Mexico................... pressum Spores IT x 3.5-5 yu, 10-12-celled; Brasil.............. L. iégapols vou" Spores murifor Thallus of im lene each with its own cortex, attached to each other by columns of tissue... nnne DIPLOTHALLUS AB ME TI oi 9555055 9-5 0 0 ee teens L. diaphanum Thallus of a single lamella, not as above. Thallus with a thick tomentum of rhizinae below............ MALLOTIUM Apothecia on lower — lobes of thallus broad, imbricate. . . . L. resupinans Apothecia on upper su Upper surface of thallus tomentose, at least near apotheci Margins of sinuses “iy lobes medium, sinuate; spec margins microphylline................. 0.00000 e eee L. inflecum Margins of thallus dios dissected. Upper v of th and more g MEN A ERATE E E E T L. papillosum -— n of the thallus arachnoid-tomentose, lobes nutely dissected, not otherwise isidiose; Chile. L. callithamnium Upper parma of thallus glabrous except in the vicinity of the apothecia; apothecial margin isidiose, scattered groups of isidia sometimes on the upper ica of the thallus. POC TET CETELE TT ee TETE ET inflecum v. isidiosulum Thallus corticate with a single layer of cells on both yi without ae Po. me ee Tee Seen ee bree Te EULEPTOGIUM Thallus scrobiculate. Margin of thallus papillate-denticulate, olivaceous, thallus deeply reticulate-scrobiculate; parathecium filamentous; spores 22- MINE. rire a4 5895 05 659004959440858 050 EXER L. olivaceum Margin of thallus smooth, thallus plumbeo Thallus deeply reticulate-scrobiculate; nab 30-40 x 12-16 y. — rre L. foveolatum Thallus shallowly scrobiculate; spores 20-30 x 8-10 y. . .L. microstictum Thallus variously wrinkled on drying, but not scrobiculate, usually h when moist Apothecia on tips of long swollen processes of the thallus. Thallus lobes narrow, dichotomously branched, wrinkles ir- TOA ry ia sn E he ae doa aoe ak eaten R RITE ES . stipitatum Thallus lobes broader, not conspicuously dichotomously branched, wrinkles predominantly go ied or radial, secondary wrinkles mostly periclinal to m — radially wrinkled to node spores 23-32 x MIB TCI" . vesiculosum "mE. with periclinal wrinkles, foliolate or isidiose. 'Thallus and apothecia not isidiose. 1933] DODGE—LICHENS OF COSTA RICA. I 419 Apothecia small, spores 22-30 x 10-12 u, margin wrinkled. Thallus lobes broad c.s oou rre) L. depuis Thallus lobes narrower, margins crisped........... ampestre Apothecia rather large with more or a foliolate margins. PATE i Nie hn Gre anki ob ps anne RS See v. macrocarpum Thallus and apothecia isidiose, often densely so; spores BO-40 (65) x 12-17 poss coo L. coralloideum Apothecia marginal, sessile or on short solid stalks, very minute. Thallus dark green to black, exciple isidiose [spores 24-30 ads e u, transverse septa 5, fide Malme]; CTC ALC: ERO IINE D I IE L. chloromelan Thallus weil to plumbeous. Exciple isidiose, parathecium 150 y thick; thecium 150- y. CD LE. . marginellum Exciple verruculose, iiio 100 & thick; thecium 120-140 y tall; Paraguay......... eee L. pilcomayense Exciple crenulate, esie ink 30 u thick; thecium 110- 0 a tall; Paragüsy. o n e eR . microcarpum ay Apothecia scattered on upper surface of the thallus. Thallus thick, 250 4 or more, roughened bi by more or less conical outgrowths but rhizinae absen d Ach or slightly wrinkled, lobes E A obo- eet 2 oscó i Eia eg noe UR EE xi L. hypotrachynum EI wrinkled above, lobes appressed, short, wide margins sometimes crisped; Brasil.......... . mattogrossense Smooth or only slightly wrinkled below................. L. sessile Thallus thinner, seldom reaching Thallus more or less isidiose. Isidia blackening, cylindrie; parathecium well developed. Spores 18-23 x 8-10 u, transverse septa 3, rare Brasil and Paraguay........ pichneum Malme excl. syn. Spores 22-30 x 11-13 y, transverse Aaa Donde et: MERE EE simplicius v. pichneoides Isidia concolorous, parathecium not well developed. . Nr ege E E ERI RIT L. doninta Nyl. Thallus 100 „ thick; apothecia not isidiose; spores u, transverse septa usually 5; isidia coralloid branched or somewhat flattened; Brasil. k^ woe Fuse For AAA L. austroamericanum (Malme) Dodge Thallus about 30 u thick; apothecial margins verrucose, spores 16-23 x 8-9 m, transverse septa 3; isidia often flattened and microphylline.............. eres ee L. denticulatum Malme non Nyl. Isidia of hemispheric concolorous granules, thallus rigid, eranl undulate, wrinkled toward the margin; Paraguay: sse o. as NARRA e i HEAR A Y ap L. granulare Thallus not isidioe Parathecium Pee er or, if pseudoparenchymatous, very thin and inconspicuous. [Vor. 20 420 ANNALS OF THE MISSOURI BOTANICAL GARDEN Cortex of amphithecium a single layer i — thallus 50-75 u thick; apothecia rarely 1 in diam.; spores 15-19 x 6-8 p with 3 federa septa; DM ariista $e aes POR S ed L. Puiggarit Cortex of amphithecium a single layer of cells above, becoming thick, pseudoparenchymatous below; apothecia 1-3 mm. in diam. Thallus greenish black, 80-100 u thick; spores 15-22 X 6-8 u, with 3-5 transverse septa; Brasil. . . L. brasiliense Thallus ashy, Mic or bluis Thallus papulose and wrinkled above, opaque, 130-160 » thick; spores 20-30 x 7-12 u.. L. pulchellum Thallus minutely wrinkled above, opaque, 150- 170 u thick; spores 24-28 x 12 w........ L. Standleyi Thallus minutely wrinkled above, margins in- rolled; spores 38-46 x 16-18 w.......... L. dimorphum Thallus smooth or minutely wrinkled on drying, pellucid, less than 100 x thic E 22-28 x 8-10 4; thecium 120-150 u tall; thallus 80-100 p thick... L. azureum Spores 18-22 x 10-12 „, with 3-5 transverse — LM 90-120 ,& tall; thallus 35- OO atbiok. oor boa 6s 40a RS L. Tuckermani Spores 18-24 x 8-9 u with 3-5 transverse septa; thecium 140-160 u tall; thallus 35-50 pu ge | es ee LER "A Schiffneri Parathecium peeudoperenchy matous. Spores 35-40 x 16-20 4; thallus plumbeous to — amphi thecium smooth; Bolivia.................. L. laevius (Nyl ) Dodge Spores 30-40 x 12-17 u; thallus rather dark plumbeous to bla Laciniae 1-1.5 mm. broad, apothecia not over 1.5 mm. in Pep thecium 170 4 tall; paraphyses 1.5-2 u thick; Brasil.... ETT Tee ee AMOUR EATER KORR EA E SAAREEN TS L. Lafayetteanum Laciniae up to 8 mm. broad, apothecia 1-2.5 mm. in diam.; thecium 190-220 , tall; paraphyses 2.5-3 u thick; Brasil. DES ARRXVECECNRELERE SUE X RO EY ESO En eke L. pachycheilum Spores 25-35 x 12-15 4; apothecia 1-2.5 mm. in diam.; thecium 150-180 4 tall; paraphyses 2-2.5 4 thick; Brasil.......... Sins bak Ev OURS s 400440 EO RUE E VR L. ulvaceum perius non Pers. Spores 22-30 x 11-13 u; apothecia 1.5-2.2 mm. in diam.;thec 150-170. 5 Illo sisse ERE Wr Eo tn Eee bs Ls simplicius Spores 18-34 x 9-15 u; thallus dark green to plumbeous; amphi- thecium granulate to wrinkled. Thalline lobes ascendant, gyrose-plicate. . L. conchatum (Tuck.) Dodge oromelan auct. non Sw.) Thalline lobes narrower, edges erect, crisped. L. stellans (Tuck.) Dodge DrirLorHALLUS Vainio, Étude Lich. Brésil 1: 222. 1890. Thallus of two lamellae connected by pillars. Each lamella 1933] DODGE—LICHENS OF COSTA RICA. I 421 has a cortex of a single layer of isodiametric cells above and below, drying impressed-punctate. Rhizinae absent, spores muriform. LEPTOGIUM DIAPHANUM (Sw.) Mont., Ann. Sci. Nat. Bot. III, 10: 134. 1848. Lichen diaphanum Sw., Nov. Gen. Sp. Pl. Prodr. 147. 1788. Parmelia diaphana Ach., Meth. Lich. 223. 1803. Collema diaphanum Ach., Lichenog. Univ. 654. 1810. Leptogium punctulatum Nyl. ap. Fournier, Mexic. Pl. 1: 1. 1872. Leptogium tremelloides var. impressopunctatum Tuck. ap. Williams, Amer. Nat. 29: 482. 1895. Type: Jamaica, Swartz, in Riksmuseet, Stockholm, carefully described by Malme, Ark. f. Bot. 198: 27. 1924. Thallus mineral gray, lobes rounded, ascending, crowded, impressed-punctate, otherwise quite smooth, consisting of two layers, each about 40 u thick, composed of a row of pseudo- parenchymatous cells about 8 u in diameter on each surface and an algal layer of Nostoc between, the layers connected by pseudoparenchymatous pillars whose contraction in drying form the depressions. Apothecia borne on the upper layer, peltate, constricted at the base, 1-1.5 mm. in diameter, margin light buff, smooth, dise chestnut; amphithecium corticate with thick pseudoparenchyma below, about 80 y thick, thinning to two layers of cells above next the disc, algal layer about 30 y thick below the thecium; parathecium absent; hypothecium of densely woven, large, thick-walled hyphae about 50 y thick; thecium 120-130 x tall; paraphyses filiform, 1-2 y thick, with clavate tips, forming a brown epithecium imbedded in a gel; asci clavate to cylindric, about 12-16 y in diameter, thin-walled tips thickened and staining deep blue with iodine; spores 8 per ascus, imbricately monostichous, fusiform, muriform, 16-25 x 7-8 y, with 3-5 transverse septa. This species is found in the mountains from Mexico and the West Indies, Dominican Republic, and Jamaica (900-1200 m.) to Minas Geraes (1400-1500) m., Bolivia (1900 m.), and Perü (1700 m). In Costa Rica it is found from 1200 to 1700 m. and descends to 800 m. on the mountain back of the farmhouse at [Vor. 20 422 ANNALS OF THE MISSOURI BOTANICAL GARDEN H. Santamaría where the peculiar weather conditions enable many species to flourish below their normal altitudes. Ap- parently it needs high humidities as it is found mostly in localities of frequent and long-continued fogs (see p. 379). Cartago: R. Birrís above Santiago, 1220-1340 m., Dodge & Thomas 7938; Cartago, R. Torres R. 143; Carpintera, 1700 m., K. Danielson 103. San José: Sta. María de Dota, 1500-1800 m., Standley & J. Valerio 43208. Heredia: C. Central de Zurquí, 1600-1700 m., Dodge, J. Valerio & Thomas 4624. Guanacaste: H. Santamaría, 720-850 m., Dodge, Jiménez & Thomas 7017. MarroTiUM Ach., Lich. Univ. 644. 1810. Mallotium Gray, Nat. Arrang. Brit. Pl. 1:399. 1821. Type: Collema saturninum (Dicks.) Ach. Thallus foliose, cortex of pseudoparenchyma above, below tomentose with rhizinae; spores muriform. This subgenus, considered as a separate genus by many authors, is cosmopolitan in distribution, mostly along cold foggy coasts or in mountains, being abundant in species in northern Europe and in Patagonia, Tierra del Fuego, and Ant- arctic Islands. Our tropical species seem confined to much higher elevations (above 1800 m. in Costa Rica) than the other members of Leptogium. It is quite distinct in appearance, often tomentose above, so that sterile specimens might be taken for Erioderma or Umbilicaria on macroscopic examination. Our two species, L. inflerum and L. papillosum, are both Mexican, the former extending to Pert, the latter not known south of Costa Rica. Leprocium papillosum (Bouly de Lesdain) Dodge, comb. nov. Leptogium Hildenbrandii var. papillosum Bouly de Lesdain, Lich. Mexique, 30. 1914. Type: Mexico, Michoacan, Puebla, H. Batan, G. Arséne Brouard 4212; Morelia, C. Azul, G. Arséne Brouard 3999. Thallus mineral gray above, dark olive buff below, lobes subpinnatifid, with small, lacerate, almost isidioid margins, papillate-granulose, sometimes isidiose above, densely tomentose below, algal layer about 60 y thick, corticate on each surface, with cells 5-6 w in diameter, the cortex of the lower surface giving rise to a dense covering of rhizinae. Apothecia sub- marginal but too immature in our specimens to show details of structure well. 1933] DODGE—LICHENS OF COSTA RICA. I 423 San José: L. de la Chonta, n. e. Sta. María de Dota, 2000-2100 m., Standley 42285; near Sta. María de Dota, 1500-1800 m., Standley 41643. LEPTOGIUM INFLEXUM Nyl., Flora 41: 377. 1858; Syn. Meth. Lich. 1: 132. 1858. Type: Mexico, Orizaba, Fr. M üller. Thallus deep glaucous gray, dark olive buff below, lobes ir- regular, rounded, margins crisped and subascending, smooth above except in the vicinity of the apothecia, densely tomentose below, algal layer about 60 y. thick, corticate on both sides with two layers of cells about 8-10 y in diameter, the lower surface giving rise to rhizinae, as also the outer cortical cells in the vicinity of the apothecia. Apothecia large, up to 4 mm. in diameter, margin more or less tomentose below, phyllophorous, concolorous with the thallus, disc orange rufous to auburn, edge of para- thecium lighter; amphithecium with pseudoparenchymatous cortex about 100 u thick, thinning to about one or two cells thick above, algal layer of Nostoc about 80 y thick; parathecium pseudoparenchymatous, about 20 y. thick below the hypothecium, expanding upward to thickness of 100 u above; hypothecium 30 u thick, of densely woven hyphae; thecium 240 y tall; para- physes filiform, 1-2 u in diameter, with swollen clavate tips forming a brown epithecium; asci 20-25 y in diameter, cylindrical; ascospores monostichous, 8 per ascus, broad, fusiform, muriform, with 7 transverse septa, about 35-40 x 13-17 y. Superficially the thallus suggests the texture of L. tremelloides, while the apothecia might easily be mistaken for L. phyllocarpum, although the rhizinae below should be easily observed. Cartago: F. Voleán de Turrialba, 2000—2400 m., Standley 34966, 35174. San José: Quebradillas, 7 km. n. Sta. María de Dota, Standley 42899; C. de las Vueltas, 2700-3000 m., Standley & J. Valerio 43823a. Var. iSiDIOSULUM Nyl., Acta Soc. Sci. Fenn. 7: 429. 1863; Ann. Sci. Nat. Bot. IV, 19: 289. 1863. Type: Colombia, Paramo Choachí, 3600 m., Lindig. This variety differs in the thallus being partly isidio-furfu- raceous, especially the margins of the lobes; apothecial margin isidiose instead of phyllophorous, subnude below; spores 36—44 x 18-25 y. This variety would seem to differ from L. papillosum in that the isidia are in scattered groups or marginal, much more highly [Vor 20 424 ANNALS OF THE MISSOURI BOTANICAL GARDEN developed and more denuded below. A single sterile specimen has been referred here. San José: C. Zurquí, 2000-2500 m., Standley & J. Valerio 48292. EvLrEPTOGrUM Tuck., Gen. Lich. 95. 1872. Stephanophorus Fw., Linnaea 17: 290. 1843. Type species: no species cited. Thallus foliose, monophyllous, without rhizinae, upper and lower cortex present, each of a single layer of isodiametric cells; spores muriform, many-celled. This subgenus is predominantly tropical with many species in tropical America. Leproarum OLIVACEUM (Hook.) Zahlbr., Cat. Lich. Univ. 3: 146. 1924, excl. syn. Collema olivaceum Hook. ap. Kunth, Syn. PI. Aequinoct. Orb. Nov. 1:38. 1822. Type: Colombia, Cauca, between Popayan and Almaguer, Humboldt 252. Thallus isabella to light brownish olive, lobes broad and round- ed, surface deeply reticulate-scrobiculate with sharp ridges and margins granulate-isidiose, or ridges rarely microphylline; about 200 u thick, corticate above and below with an algal layer of loosely tangled filaments of Nostoc corresponding to a medulla, separated from the cortex by a palisade layer about 30 y thick. Apothecia sessile, margin verrucose, concolorous, disc darker; amphithecium thick, homogeneous with the thallus; parathecium about 80 u thick, of very densely woven hyphae; hypothecium about 25 y thick, very deeply staining; thecium about 100 y tall; paraphyses filiform, about 1.5 u in diameter, ending in the epithecial gel. The apothecia are very rare and too immature in the one specimen from Costa Rica to give all the characters, but it seems quite distinct from L. reticulatum and L. foveolatum. Cartago: R. Reventazón below Santiago, 740—750 m., Dodge 8010. Var. granulosum Dodge, var. no Type: Costa Rica, San José, R. Virilla below El Brazil, Dodge 8030. Thallus similis speciei sed nigro-granulatum ambobus super- ficiebus. Apothecia obscure brunnea, nigricantia, sessilia, basi 1933] DODGE—LICHENS OF COSTA RICA. I 425 constricta, disco concolore; amphithecium 190 y crassitudine, corticatum cellulis isodiametricis; parathecium 130-140 y. crassi- tudine, ad marginem 40 y. attenuatus, hyphis dense contextum; hypothecium 30 uw crassitudine; thecium 120-130 y altum; paraphyses similes speciei; asci 12-16 y diametro metientes; ascosporae imbricatim monostichae, octonae, fusiformes, 22-24 x 12-14 y. Thallus similar to that of the species but black-granulate on both surfaces. Apothecia dark brown, blackening, sessile, constricted at the base, disc concolorous; amphithecium 190 y thick, homogeneous with the thallus, corticate with a single layer of isodiametric cells; parathecium 130-140 y thick, thinning toward the margin to 40 u, of densely woven hyphae; hypothecium 30 u thick, so deeply staining that its structure is not clear; thecium 120-130 y tall; paraphyses as in the species; asci cylindric, 12-16 y in diameter; ascospores imbricately monostichous, 8 per ascus, fusiform, 22-24 x 12-14 y. San José: R. Virilla below El Brazil, Dodge 8030. LEPTOGIUM FOVEOLATUM Nyl., Syn. Meth. Lich. 1:124. 1858. Type: not stated, specimens from Bolivia, Weddell, and Mexico, Fr. M üller, are mentioned. Thallus mineral gray or darker, often discolored light yellowish olive or darker, lobes rounded, smooth, surface deeply foveolate with sharp wrinkles both above and below; very variable in thickness up to 250 u, Nostoc very sparsely scattered throughout, rather denser and tending to form a palisade layer just below the surface, but this layer much thinner and less definite than in L. olivaceum. Apothecia about 1.5 mm. in diameter, constricted at the base, margin smooth, light buff, dise cinnamon rufous; amphithecium about 130 u thick, similar to the thallus in structure but algae much more abundant and closely tangled and cortex pseudoparenchymatous, about 50 y thick; parathecium about 40 y thick below, thinning out and disappearing above, pseudo- parenchymatous with small cells; hypothecium about 80 y. thick, very deeply staining; thecium about 170 y» tall; paraphyses filiform, about 2 y in diameter, tips clavate; asci 8-spored, cylindrical, about 12 y in diameter; ascospores imbricately [Vor. 20 426 ANNALS OF THE MISSOURI BOTANICAL GARDEN monostichous, 30-40 x 12-16 u [only 20 x 10 u in Costa Rican specimens, but obviously immature]. San José: R. Virilla below El Brazil, Dodge 7783. LEPTOGIUM MicnosTICTUM Vainio, Dansk Bot. Ark. 4": 18. 1926 Type: not stated, specimens cited, Mexico, Palenque, Liebmann 7415, 7416, 7428; Papantla, Liebmann. Thallus light glaucous blue or a little darker, lobes rounded, surface smooth but irregularly impressed, margin thin and smooth; about 225 y thick, corticate on both surfaces with cells about 4 u in diameter, the filaments of Nostoc very loosely tangled, more or less parallel to the surface and very scanty in the middle. Apothecia 0.5-1.0(— 1.8) mm. in diameter, sessile, constricted at the base, margin thin, whole, concolorous with the thallus, dise pale to tawny; amphithecium about 60 u thick, cortex pseudoparenchymatous, about 25 y thick below, thinning out above to a double layer of cells, algal layer of tangled filaments of Nostoc; parathecium pseudoparenchymatous, about 40 y thick, thinning above to about three layers of cells; hypothecium 30 y. thick, of densely tangled rather large hyphae suggesting pseudo- parenchyma but cells much smaller than in the parathecium; thecium about 100 y tall; paraphyses filiform, 1 u in diameter, ending in the epithecial gel; asci clavate to cylindrical, 12-16 u in diameter; ascospores distichous, 8 per ascus, fusiform, with about 5 transverse septa, 20-30 x 8-10 u. A single small specimen from Costa Rica is doubtfully referred here. The thallus is somewhat more finely impressed and oc- casionally near the margin it appears very minutely scrobiculate. This specimen is from a greater elevation than most other mem- bers of this subgenus. San José: Guayabillos and Cabeza de Vaca, 2150-2350 m., Dodge & Thomas 7418. LEPTOGIUM VESICULOSUM (Sw.) Malme, Ark. f. Bot. 198: 14-15. 1924. Lichen vesiculosus Sw., Nov. Gen. Sp. Pl. Prodr. 147. 1788. Lichen bullatum Ach., Lichenog. Suec. Prodr. 137. 1798. Leptogium bullatum Mont., Ann. Sci. Nat. Bot. II, 16: 113. 1841. Type: Jamaica, tops of mountains, O. Swartz. L. bullatum Ach. based on the same material. 1933] DODGE—LICHENS OF COSTA RICA. I 427 Thallus plumbeous, lobes rounded, more or less appressed, longitudinally wrinkled with shallower cross ridges; about 100 y. thick between the ridges, corticate on each surface with a single layer of isodiametric cells, homoeomerous, of very loosely tangled filaments of Nostoc running more or less parallel to the surface. Apothecia sunk in the tips of podetiiform elevations of the thallus, margin wrinkled but not isidioid or phyllophorous, edge of parathecium conspicuous light buff, disc concave, rufous; amphithecium continuous with the thallus without differentiation, cortex a single layer of cells both above and below; parathecium 100 u thick below, thinning out to 40-60 u above at the margin, pseudoparenchymatous; hypothecium thick, 80 y, of densely woven slender hyphae; thecium 120-140 y tall; paraphyses slender, filiform, cells almost isodiametric above, disintegrating into the brown epithecial gel; asci cylindrical, 20 » in diameter, 8-spored; ascospores imbricately monostichous, broadly fusiform, 23-32 x 10-13 y, with 5 transverse septa. This species seems to be widespread in the American tropics, from 700 to 2100 m., rarely also below these levels. L. stipitatum, with which this species is easily confused, seems confined to the lower levels, reaching up to about 800 m., and has so far only been reported from Guadeloupe and Costa Rica. Limón: Waldeck, Dodge & Nevermann 7938; Marta, 20 m., Dodge 7427. Cartago: Angostura, Polakowsky 456; Juan Vifias, 1000 m., Calvert 68; Santiago, 1140-1180 m., Dodge 4555; R. Birrís, 1220-1340 m., Dodge & Thomas 4628; Agua- caliente, 1240-1460 m., Dodge & Thomas 7082; Carpintera, 1700 m., K. Danielson 101a. San José: S. Pedro de Montes de Oca, 1200 m., Thomas 4717; S. Juan Tibas, 1000-1100 m., Dodge 4303. Guanacaste: H. Santamaría, 640-680 m., Dodge & Thomas 6797. Puntarenas: Boruca, 560 m., T'onduz 5380. Var. DIGITATUM Eschw. ap. Martius, Fl. Brasil. 1: 238. 1833. Collema bullatum Raddi, Atti Soc. Ital. 18: 36, pl. 4, f. 2. Collema bullatum var. dactylinoideum Nyl., Flora 41: 338. 1858; Syn. Meth. Lich. 1: 129. 1858. Type: Brasil, between Mandiocca and Morro do Frade, Raddi, type of var. dactylinoideum not cited, but specimens from Mexico, Fr. Müller, Colombia, Tolima, Goudot, and Bolivia, Weddell, are mentioned. This variety is stated by Eschweiler to have a thicker thallus, [Vor. 20 428 ANNALS OF THE MISSOURI BOTANICAL GARDEN deep green instead of lead color, densely wrinkled instead of granular; podetia cylindrical instead of ventricose, 6-12 mm. instead of 2-6 mm. tall, about 6 mm. in diameter. From the description this variety does not seem distinct from the species and I have seen no material referable here from Costa Rica. It is possible that Eschweiler was attempting to separate this species from L. phyllocarpum which he called Collema (Leptogium) bullatum var. sertatum. LEPTOGIUM STIPITATUM Vainio, Hedwigia 38: (255). 1899. Type: Guadeloupe Island, near Gourbeyre, P. Duss 434. Thallus plumbeous, soon discolored buffy citrine to Saccardo’s olive, irregularly dichotomously branched, ultimate lobes 2-7 mm. broad, then crowded and confluent, rounded, acutely and irregularly wrinkled above and below; about 200 » thick between the ridges, with Nostoc filaments densely coiled and tangled near the surface, mostly parallel to the surface within, much more abundant than in L. vesiculosum, with cortex of isodiametric cells on each surface. Apothecia on short thalline pustules, up to 2.5 mm. in diameter, margin thick with periclinal folds (never radial as in L. vesiculosum), disc chestnut; amphithecium about 200 y thick, of the same texture as the thallus; parathecium 80 u thick, disappearing above, of large-celled pseudoparenchyma; hypothecium 80 y thick, of more or less parallel hyphae continued upward as a false parathecium; thecium about 140 y tall; para- physes filiform, about 1.5 » in diameter, septate, with capitate tips in the brown epithecial gel; asci cylindrical, 16-20 y in diameter; ascospores distichous, broad-fusiform with acute ends, 24—32 x 10-12 y. Ban José: cafion of R. Virilla, below El Brazil and Sta. Ana, oe 8019. Puntarenas: Peninsula of Puntarenas, 3-5 m., Dodge LEPTOGIUM PHYLLOCARPUM (Pers. ap. pie e Mont., Ann. Sci. Nat. Bot. III, 10: 134. 1848. Collema phyllocarpum Pers. ap. Gaudich. in Freycinet, Voy. Uranie, Bot. 204. 1820. Collema bullatum var. sertatum Eschw. ap. Martius, FI. Brasil. 1: 239. 1833. Type: Brasil, Rio de Janeiro, Gaudichaud. Thallus plumbeous but frequently becoming black, especially 1933] DODGE—LICHENS OF COSTA RICA. I 429 in specimens which have been dried slowly or repeatedly wetted and dried after collection, lobes broad, surface deeply and acutely wrinkled, the wrinkles predominantly longitudinal or radial, the secondary wrinkles very irregular; more than 200 y thick between the wrinkles, corticate on both surfaces by a single layer of isodiametric cells, the filaments of Nostoc loosely tangled, more or less parallel to the surface and much more coiled and densely tangled just under the upper cortex. Apothecia about 2 mm. in diameter, margin with crowded periclinal wrinkles, dise rufous, immersed in hollow protuberances as in L. vesi- culosum; amphithecium appearing lobed in section up to 500 y thick, homogeneous with the thallus but algae a little denser; parathecium about 100 y. thick below, thinning out to about 20 y. above at the margin, of large-celled pseudoparenchyma; hypo- thecium about 30 y thick, of slender densely woven hyphae continuing up to the margin at about the same width; thecium about 140 y tall; paraphyses filiform, not dilated above, epithecial gel brown; asci cylindric, about 20 y in diameter, 8-spored; ascospores broad-fusiform with acute ends, monostichous, with 5 transverse septa, 22-30(- 35) x 10-12(- 14) y. Apparently this species is very widespread and common in the American tropics from low elevations in the southern United States to 2500 m. in Perú. In Costa Rica it ranges from 500 m. to about 1900 m., being characteristic in the moister fog-bathed areas of the temperate region. Cartago: Santiago, 1140-1180 m., Dodge 8000; R. Birrís, 1220-1340 m., Dodge & Thomas 7994. San José: Cañon of R. Virilla below El Brazil, Dodge 7782; Zapote, 1200 m., Standley 40273; La Palma, 1500-1700 m., Maxon & Harvey 7891; Quebradillas, near Sta. María de Dota, 1800 m., Standley 42977; Sta. María de Dota, 1500-1800 m., Standley 42424; S. Marcos de Dota, 1200 m., Tonduz 5374; R. Naranjo, Tonduz 5375; S. Gabriel, Tristan 5229; S. José, Polakowsky 113. Alajuela: R. Ciruela, 920-980 m., Dodge & J. Valerio 4858; Santiago de S. Ramón, 1000 m., Brenes; Viento Fresco, 1600-1900 m., Standley & Torres 47771. ? Puntarenas: Corozal, 5-50 m., Dodge 8025. Var. MACROCARPUM Nyl., Syn. Meth. Lich. 1: 130. 1858. Type: not cited but specimens from Aequatorial America, Humboldt, Chile, Perú, and Venezuela, Lind. 1092, are mentioned. Similar to the species but apothecia large, 5-9 mm., margins lobulate, foliose. American material referred to var. daedaleum [Vor. 20 430 ANNALS OF THE MISSOURI BOTANICAL GARDEN Nyl. probably belongs here. Distribution in Costa Rica es- sentially similar to that of the species but not yet collected in the higher elevations. Cartago: Santiago, 1140-1180 m., Dodge 8001; R. Birrís, 1220-1340 m., Dodge & Thomas 4632; Carpintera, 1700 m., K. Danielson 101b. San José: R. Naranjo, Tonduz 5377. Alajuela: La Palma de San Ramón, 1250 m., Brenes 91. Guanacaste; H. Santamaría, 640—780 m., Dodge & Thomas 6776, 8002; Tilardn, 500—690 m., Standley & J. Valerio 44803, Dodge & Thomas 6562. Puntarenas: Corozal, 5-50 m., Dodge 7529. Var. CAMPESTRE Malme, Ark. f. Bot. 198: 13. 1924. Type: Brasil, Bahia, R. Vermelho, Malme 29, 30; Matto Grosso, Cuyaba, Malme 2715c, etc.; Serra da Chapada, pr. Bocca da Serra, Malme 2245B. Differs from the typical form by narrower crowded lobes, margins crisped; apothecia 1-2 mm. in diameter, margin rugulose, thin; amphithecium corticate with a single layer of cells; para- thecium 60-70 » thick, reaching the margin above; thecium 100- 125 y tall, asci up to 20 u thick, 8-spored; ascospores irregularly distichous, broad-fusiform, acute ends, 22-27 x 8-10 y, usually with 5 transverse septa. San José: Turrdcares, 540-600 m., Dodge & Thomas 7481. Alajuela: R. Ciruela, 920-980 m., Dodge & J. Valerio 7921; Santiago de S. Ramón, 1000 m., Brenes. LEPTOGIUM CORALLOIDEUM Vainio, Ann. Acad. Sci. Fenn. A 67:110. 1915. Leptogium diaphanum f. coralloideum Mey. & Fw., Nova Acta Acad. Leopold. Carolin. 19: Suppl. 226. 1843. Leptogium phyllocarpum var. isidiosum Nyl., Syn. Meth. Lich. 1: 130. 1858. Leptogium phyllocarpum var. coralloideum Hue, Nouv. Arch. Museum Paris III, 10: 228. 1898. The type of var. isidiosum is from Mexico, Orizaba, Fr. M üller. Thallus as in L. phyllocarpum but densely isidiose, especially along the wrinkles, somewhat thicker, semi-pellucid. Apothecia large when present, very rare, margin densely coralloid-isidiose, dise rufous; amphithecium highly developed, homogeneous with the thallus; parathecium 140 y thick, disappearing above, of large thin-walled pseudoparenchyma; hypothecium 60 y thick, reaching the margin above, of densely woven slender hyphae; 1933] DODGE—LICHENS OF COSTA RICA. I 431 thecium 200 y. tall; paraphyses filiform with clavate apices; asci cylindrical; ascospores 8 per ascus, imbricately monostichous, ellipsoid to fusiform, apices acute, 30-40(-45) x 12-17 u, with 7 transverse septa. This species is perhaps only a large isidiose variety of L. phyllocarpum, essentially similar in morphology but larger in most dimensions. It occupies the same general region as L. phyllocarpum. It is usually sterile and hence not easily distin- guished from sterile L. marginellum var. isidiosellum. Material has been seen from Mexico, Pert, and Brasil. Limón: Hamburg, 20-30 m., Dodge 7420. Cartago: Pejivalle, 650-900 m., Standley & J. Valerio, Dodge & Thomas 4337; Cañon of R. Reventazón below Santiago, 920-1000 m., Dodge & Thomas 8018, Dodge 8004; R. Birrís, 1220-1340 m., Dodge & Thomas 7954; C. Carpintera, 1320— 1700 m., Dodge 3758, Dodge & Thomas 4762, K. Danielson 98; Aguacaliente, 1240— 1460 m., Dodge & Thomas 8003; R. Reventado, 1460-1650 m., Standley & J. Valerio 614. one José: R. Virilla below El Brazil, Dodge 6496; Zurquí, H^. d m., Standley & J. Valerio 48255, 48266, 48146, Dodge, J. Valerio & Thomas 6042. Alajuela: near Fraijanes, 1500-1700 m., Standley & Torres 47425, 47462 Guanacaste: H. Santamaría, 640—680 m., Dodge & Thomas 8005; Tilarán, 500-650 m., Standley & J. Valerio 44367, 44403. LEPTOGIUM MARGINELLUM (Sw.) S. F. Gray, Nat. Arrang. Brit. Pl. 1: 401. 1821. Lichen marginellus Sw., Nov. Gen. Sp. Pl. Prodr. 147. 1788. Collema vesicatum Taylor, London Jour. Bot. 6: 196. 1847. Leptogium corrugatulum Nyl., Syn. Meth. Lich. 1: 182. 1858. Type: Jamaica, Blue Ridge Mt., Swartz. The type of C. vesicatum Taylor is from St. Vincents, West Indies. The type of Leptogium corrugatulum is from Mexico, Jalapa, Galeotti 9630. Thallus plumbeous, round-lobed, margin smooth or occasionally microphylline, pellucid, surface of undulating longitudinal wrinkles less acute than in L. phyllocarpum, of variable thickness, about 160 x thick, corticate on both surfaces with small isodia- metric cells. Apothecia minute, up to 0.3 mm. in diameter, margin densely isidiose, disc rufous; amphithecium thick, with pseudo- parenchymatous cortex below, a single layer of cells above, as are also the isidia; parathecium 150 y. thick, pseudoparenchymatous; thecium 150-180 y tall, epithecium tawny; asci 8-spored; spores imbricately monostichous, ellipsoid-fusiform, ends acute, 25-35 (-40) x 10-13 y, with 5 transverse septa. [Vor. 20 432 ANNALS OF THE MISSOURI BOTANICAL GARDEN This species rarely matures spores although rudimentary apothecia are nearly always abundant. ‘The microscopic details of the apothecium are largely taken from Malme as I have seen no mature asci in my Costa Rican material. There is consider- able variation in the amount of isidia, as they easily break off leaving minute foveoles on the margin. The species is quite distinet and is widespread in the American tropies. I have seen material from Florida, Alabama, Bermuda, various islands of the West Indies, Mexico, Brasil, and the Galápagos Islands. It seems to range from sea level to 1500 m. From the nature of the isidia it seems likely that var. isidio- sellum Riddle, Brooklyn Bot. Gard. Mem. 1: 115. 1918, should be referred to L. coralloideum, although sterile isidiose states are very difficult to place. Limón: Marta, 20 m., Dodge & Nevermann 7986; Carmen, Dodge 7422; Waldeck, Dodge & Nevermann 7421. Cartago: Pejivalle, 600-900 m., Dodge & Thomas 8014, Standley & J. Valerio 46769; R. Birrís, 920-1100 m., Dodge 7988; Santiago, 1140-1180 m., Dodge 7987; Aguacaliente, 1240-1460 m., Dodge & Thomas 7937; Carpintera, 1500 m., Dodge 3948. Alajuela: Santiago de S. Ramón, 1000 m., Brenes 242; S. Pedro de S. Ramón, 700 m., Brenes. Guanacaste: H. Santamaría, 640—720 m., Dodge & Thomas 6867, 7989; H. Grana- dilla, above R. 8. José, 480 m., Dodge & Thomas 7936; H. Q. Azul, on lower slope of V. Tenorio, 400—600 m., Dodge & Thomas 6653; Tilardn, 500-690 m., Dodge & Thomas 61. Puntarenas: R. Terrones, 30 m., Dodge & Marks 7980; Corozal, 5-50 m., Dodge 7583. LEPTOGIUM DENTICULATUM Nyl., Ann. Sci. Nat. Bot. V, 7: 302. 1807. L. tremelloides v. leptophyllinum Mey. & Fw., Nova Acta Acad. Leopold. Carolin. 19: Suppl. 228. 1843. L. tremelloides v. azureum f. isidiosum Müll. Arg., Bull. Soc. R. Bot. Belg. 29: 49. 1891, non Flora 65: 292. 1882. L. leptophyllinum Zahlbr., Cat. Lich. Univ. 3: 136. 1924. Type: Colombia, San Jil, 1300 m., Lindig. Thallus mineral gray or darker, large, broadly lobed, lobes round, edges smooth, somewhat crisped, surface slightly wrinkled, 100-120 u thick, microphylline, isidiose, corticate on both sides with a single layer of cells, the short filaments of Nostoc with cells 4-5 y in diameter mostly assembled just under the cortex, 1933] DODGE-—LICHENS OF COSTA RICA. I 433 leaving the central portion relatively free of algae. Apothecia relatively rare but when present usually numerous, scattered over the upper surface, up to 2 mm. in diameter, margin granular, thick, becoming isidiose in age, buff, disc rufous; amphithecium 220 y thick, with a layer of pseudoparenchymatous cortex below about 160 u thick, not thinning out much above and merging with the parathecium, algal layer 60 y thick below, thinning out to 30 u above; parathecium 100 y. thick below, thinning out to about 40 u above, of large thin-walled hyphae at times sug- gesting pseudoparenchyma; hypothecium 20 » thick, of more slender deeply staining hyphae; thecium 160 y tall; Bus slender, ending abruptly in the epithecial gel; asci 10-12 y in diameter, cylindrical, 8-spored; ascospores imbricately mono- stichous, fusiform with acute ends and about 5 transverse septa, about 24 x 10 y. This species seems widespread in the American tropics, al- though it is so rarely fertile that it is difficult to be sure of the determination of much of the sterile material, especially to separate it from L. cyanescens var. austro-americanum Malme. As here described, this species is close to, although not identical with, Malme's variety, while the L. denticulatum Malme seems to be an isidiose state of or closely related to L. Tuckermant. Only a careful comparison of all the fertile specimens from tropical America in the various herbaria of Europe can settle these problems. Limón: Marta, 20 m., Dodge & Nevermann 7428; Hamburg, Standley & J. Valerio 48750; Castilla, 20 m., Dodge & Nevermann 7996; Waldeck, Dodge & Nevermann 7997, Dodge 8012. rtago: R. Reventazón below Santiago, 740-750 m., Dodge 4626, 920-1000 m., Dodge & Thomas 7993; Santiago, 1100-1140 m., Dodge 7939; R. Birrís, 1220-1340 m., Dodge & Thomas 7992; Estrella, 1600 m., K. Danielson 162; near Cartago, 1500 m., K. Danielson 22; C. Carpintera, 1500 m., Dodge 3978 San José: San José, Tonduz 5247. Heredia: C. Zurquí, 1600-1700 m., Dodge, J. Valerio & Thomas 8048. Alajuela: C. Mondongo de il PE 750-800 m., Brenes 275; Alto de la Palma de S. Ramón, 1250 m., Brenes 377. Guanacaste: H. tannin, 680-780 m., —— E poe 6870, 7014; H. Grana- dilla above R. Las Cañas, 500-600 m., Thom Puntarenas: Osa, Corozal, 5-50 m., Dodii m. LEPTOGIUM AZUREUM (Sw.) Mont. ap. Webb, Hist. Nat. Îles Canaries 3!: 129. 1840. [Vor. 20 434 ANNALS OF THE MISSOURI BOTANICAL GARDEN Lichen azureum Sw. ap. Ach., Lichenog. Suec. Prodr. 137. 1798. Type: mountains of Jamaica, O. Swartz. Thallus pellucid, greenish-glaucous-blue to puritan gray, broadly sinuate-rounded lobes, smooth; margins smooth, 80- 100 uw thick, of loosely coiled and tangled filaments of Nostoc with ellipsoidal cells 2.5 x 4 u. Apothecia scattered, 1-2 mm. in diameter, margin smooth, light buff, disc rufous, sessile or on a short stalk; amphithecium 320 y thick, below with a pseudo- parenchymatous cortex about 100 » thick, above with a single layer of isodiametric cells in the cortex; parathecium filamentous, 40 y. thick, poorly developed, of slender hyphae not reaching the margin; hypothecium 20 y thick, deeply staining; thecium 120- 150 y tall; paraphyses filiform, 1.5-2 u thick, thickened at the tips to about 4 u; asci clavate-cylindrical, about 18 y in diameter, 8-spored; ascospores imbricately monostichous or somewhat distichous, fusiform ends acute, with 3 or 5 transverse septa, 22-28 x 8-10 y. Only a careful study of all the types involved can settle no- menclature of the group of species centering around L. tremel- loides. ‘This species was originally collected in South Africa by Thunberg, and very briefly described by Linné fil. 'The early lichenologists of the west coast of Europe identified their material as this species and later reduced Swartz’ Jamaican L. azureum to synonymy. There is apparently a whole group of species having approximately the same macroscopic appearance but wholly different microscopic structure which are at present referred to this species by various workers. As we have here defined it (agreeing closely with L. tremelloides Malme), it is a species of the lowlands coming up the river valleys under favor- able conditions to about 1000 m. in Costa Rica. It seems widely distributed at lower elevations in tropical America. Limón: Marta, 20 m., Dodge & Nevermann 7414, 7426; Hamburg, 20-30 m., Dodge & Nevermann 7417; Castilla, 20 m., Dodge 7209, 7419, 7424; Carmen, Dodge 7415; Indiana I. near Siquirres, 60—70 m., Dodge & Thomas 6572; along R. Siquirres, 70-350 m., Dodge, Catt & Thomas 8006, 8007, 8008; Livingston, 80-100 m., Dodge, Catt & Thomas 6573; Waldeck, Dodge & Nevermann 7412, 7413, 7416; Guápiles, 300- 500 m., Standley 37137. Cartago: Turrialba, Ørsted; R. Pejivalle, 650-800 m., Dodge & Thomas 4414; R Reventazón, 920-1000 m., Dodge & Thomas 4604. 1933] DODGE—LICHENS OF COSTA RICA. I 435 Guanacaste: H. Santamaría, 760-900 m., Dodge & Thomas 6876; Tilarán, 500-690 m., Dodge & Thomas 8009. Puntarenas: Osa, C. Guaca, 85 m., Dodge 7505; Corozal, 5-50 m., Dodge 7530. Isla Coco: Snodgrass & Heller. Leptocium Standleyi Dodge, sp. nov. Leptogium tremelloides Auct., non Linné fil. Type: Costa Riea, La Hondura, Standley 37870. Thallus plumbeus, crassus, laevis vel minute rugulosus, lobis rotundatis, latis, marginibus integris, 150-170 y crassitudine, filamentis nostocaceis sub cortice dense implicatis, centro laxe aut sparse implicatis, corticatum e serie simplice cellularum. Apothecia ad 3 mm. latitudine, plana, margine integro, tenui, laevi, fulva, disco rufo; amphithecium inferne 160 u, ad 100 y. crassitudine superne in margine attenuatum, cortice pseudo- parenchymatico, 90 v. crassitudine inferne ad una serie cellularum superne attenuatum, filamentis nostocaceis dense contextum; parathecium 60 y. crassitudine, hyphis tenuibus dense contextum; hypothecium flavum, 20-30 y crassitudine, hyphis tenuibus dense contextum; thecium ad 180 y altitudine; paraphyses filiformes, tenues, apice non incrassata; asci cylindrici, 20 u diametro metientes; ascosporae octonae, imbricatim monostichae vel distichae, fusiformes, 24—28 x 12 y. Thallus mineral gray or darker, thick, smooth or minutely and shallowly wrinkled, lobes broad, rounded, margins smooth, 150-170 u thick, filaments of Nostoc densely tangled next the cortex, loosely tangled and scattered in the center of the thallus, cortex of a single series of cells on both surfaces. Apothecia 3 mm. broad, plane, margin smooth, thin, tawny, dise rufous; amphithecium 160 » thick below, thinning to 100 y. above at the margin, corticate below, 90 y thick, pseudoparenchymatous, above becoming a single layer of cells, filaments of Nostoc densely tangled; parathecium 60 y. thick, of slender more or less parallel hyphae; hypothecium yellowish, 20-30 y thick, of slender densely woven hyphae; thecium 180 y. tall; paraphyses filiform, slender, apices not thickened; asci cylindrical, 8-spored, 20 y. in diameter; ascospores imbricately monostichous or distichous, fusiform, 24-28 x 12 y. This species is essentially close to L. azureum in its microscopic characters but differs in its much larger dimensions, its thicker, [Vor. 20 436 ANNALS OF THE MISSOURI BOTANICAL GARDEN opaque thallus, and larger apothecia. Superficially it some- what resembles L. pulchellum but is much less coarsely wrinkled and differs in microscopic details. In Costa Rica it seems to be confined to the higher elevations reaching 1700 m. It is probable that material determined as L. pulchellum from Costa Rica by Müller Argau belongs here, although I did not have time to section it. It certainly is not typical L. pulchellum. Cartago: Turrialba, Ørsted; Orosf, 1000-1100 m., Standley 39791. San José: San José, 1130 m., Standley 41233; La Hondura, 1300-1700 m., Standley $7870, TYPE. Alajuela: Pods, Tonduz 5301. Leprocium Tuckermani Dodge, sp. nov. Leptogium tremelloides var. minor Tuck. in herb. ? Leptogium moluccanum Vainio, Étude Lich. Brésil 1: 223-224. 1890, non Collema moluccanum Pers. ap. Gaudich. in Freycinet, Voy. Uranie, Bot. 203. 1826. Type: Cuba, Monte Verde, C. Wright 56. Thallus plumbeus vel obscurior, lobis confertis, rotundatis, superne inferneque sat laevigatus, tomento isidiisque destitutus, tenuissimus, 35-60 y crassitudine, strato nostocaceo 25 y, corti- cibus pseudoparenchymaticis cellulis magnis. Apothecium minu- tum, 0.5-0.9 mm. diametro metiens, peltatum, basi constrictum, margine integro, pallescente, disco castaneo; amphithecium corticatum cellulis pseudoparenchymaticis, 60 u crassitudine inferne, attenuatum ad 12 y superne, strato nostocaceo 40 u crassitudine; parathecium pseudoparenchymaticum tenue, 10- 12 » crassitudine, cellulis parvis; hypothecium 20 y crassitudine, hyphis tenuibus dense contextum; thecium 90-120 y. altitudine; paraphyses filiformes, 2 y diametro, apicibus incrassatis; epi- thecium brunneum; asci cylindrici, 12 & diametro metientes; ascosporae octonae, imbricatim monostichae, apicibus acutis, murales, septis transversalibus 3-5, cellulis haud numerosis, 18-22 x 10-12 y. Thallus mineral gray or darker, lobes crowded, more or less elevated and crisped, rounded, smooth on both sides, 35-60 y thick, algal layer of Nostoc about 25 u, with pseudoparenchyma- tous cortex above and below of large cells. Apothecia minute, 0.5-0.9 mm. in diameter, constricted at the base, margin whole, smooth, pale, dise chestnut; amphithecium with a thick pseudo- 1933] DODGE—LICHENS OF COSTA RICA. I 437 parenchymatous cortex below, about 60 y, thinning to two rows of cells about 10-12 u thick above; parathecium thin, pseudo- parenchymatous, 10-12 u thick, of small cells, inconspicuous and often reported absent; hypothecium 20 » thick, of slender densely woven hyphae; thecium 90-120 u tall; paraphyses filiform, 2 v in diameter, with clavate apices; epithecium brown; asci cylindric, 12 y in diameter; ascospores 8 per ascus, imbricately monostichous, ends acute, muriform, with 3-5 transverse septa, cells not numerous, 18-22 x 10-12 y. This species has long been confused with L. moluccanum, L. Mariannum, and L. diaphanum. It differs from the two former in habit and color and from the latter by its structure. It is apparently widespread in the American tropics, being reported from Paraguay, and from Brasil in the states of Rio Grande do Sul and Matto Grosso by Malme, and in Minas Geraes by Vainio. Limón: Hamburg, Dodge 7425. Cartago: Pejivalle, 680 m., Dodge & Thomas 4556; Santiago, 1140-1180 m., Dodge 4555, 8044; Las Concavas, 1350-1450 m., Dodge 6842. San José: hills above Sta. Ana, Dodge 7781. Guanacaste: H. Santamaría, 680—780 m., Dodge & Thomas 8043; H. Granadilla, 480 m., Dodge & Thomas 8045. LEPTOGIUM SIMPLICIUS Vainio, Ann. Acad. Sci. Fenn. A6’: 109. 1915. Type: S. Domingo, La Cumbra, C. Raunkiaer. Thallus between light drab and light grayish olive, with irregularly incised lobes 5-7 mm. broad, not isidiose, with slightly undulate elevated acute wrinkles on both surfaces, corticate with a single layer of isodiametric cells. Apothecia broadly adnate, constricted at the base, sessile, 1.5-2.2 mm. in diameter, disc plane, rufous, margin thin, smooth, lighter than the thallus; amphithecium 140-150 u thick, corticate with a single series of cells above, pseudoparenchymatous below, 90-100 wœ thick; parathecium 120 u thick below, thinning out to 20 u thick at the margin above, pseudoparenchymatous; hypothecium fila- mentous, 30 y thick; thecium 150-170 u thick; paraphyses filiform, tips somewhat thickened in the brownish epithecial gel; asci cylindric, 14-16 y in diameter, 8-spored; ascospores imbri- cately monostichous, fusiform, thin-walled, with 5 transverse septa, 22-30 x 11-13 y. [Vor. 20 438 ANNALS OF THE MISSOURI BOTANICAL GARDEN This species previously reported only from the type locality has yet been found only at low elevations on the Pacific slope. San José: Turrücares, 540—600 m., Dodge & Thomas 8047. Guanacaste: Liberia, 100 m., Dodge, Alfaro & Thomas 6586, Dodge & Thomas 8028, Tilarán, 500-690 m., Dodge & Thomas 8046. Var. pichneoides Dodge, var. nov. Type: Costa Rica, Guanacaste, H. Santamaria, Dodge & Thomas 8029. Thallus isidiosus, marginibus microphyllinis aut isidiosis. Apothecium margine tenuiore, laevi, thecium 130-150 y altitudine. Thallus isidiose, margins varying from dentate to short isidiose or slightly microphylline. Apothecia with thinner, smooth margins, thecium 130-150 y. tall. Only the type has been found fertile and here there are very few apothecia. The other collections cited probably belong here, but it is always difficult to determine sterile isidiose speci- mens in this genus. Alajuela: R. Ciruela, 920-980 m., Dodge & J. Valerio 4894. Guanacaste: H. Santamaría, 640-680 m., Dodge & Thomas 8029. Besides the species cited above, a single small fragment was found which seems to belong in Leptogiopsis, but more material is necessary for a satisfactory disposition. PANNARIACEAE Thallus squamose or foliose, not gelified; hypothallus and rhizinae usually highly developed, heteromerous, cortex of erect, irregular or periclinal hyphae, usually more or less pseudo- parenchymatic; medulla well developed, with Nostoc, Scytonema, or Dactylococcus; lower cortex of thick-walled periclinal hyphae often thin or wholly absent. Apothecia marginal or scattered over the upper surface, biatorine or lecanorine; paraphyses unbranched; asci 8-spored; spores hyaline, unicellular, rarely 2-4-celled ; spermatia short, straight. 'This family seems to be of southern origin with some genera confined to the southern hemisphere, while a few species come northward until they are cireumpolar. In general, the northern species are much smaller and more depauperate than the southern species. In many ways the family seems very homogeneous, so that 1933] DODGE—LICHENS OF COSTA RICA. I 439 genus distinctions are often rather arbitrary. Hue, in 1912, reduced the whole family to a single genus, but his sections and subsections correspond more or less closely to the genera of earlier workers. Zahlbruckner has recognized twelve genera, although the generic delimitation is not altogether clear. Fol- lowing this treatment I have tentatively recognized seven genera from Tropical America. Psoroma, with Dactylococcus algae, is largely confined to the southern hemisphere, and by many authors has been placed in the Lecanoraceae. Lepidocollema, with a thin upper cortex and almost homoeomerous, is very close to the Collemaceae and is known from a single collection in Brasil. The other five genera are much better known and widespread in the American tropics. The morphology of the thallus is not altogether clear and the nomenclature of the parts is rather confusing. In the ideal case, the uppermost layer of the thallus, here called ‘‘tomentum,”’ is a mass of loosely woven hyphae formed by proliferation of deeper layers. In Erioderma it is soft and spongy, like a loose felt, while in Malmella and in some of the isidiose members of Pannaria it has the appearance of a fine blotting paper, or in Coccocarpia the hyphae are more conspicuously parallel and conglutinate, suggesting a frayed cotton yarn. In some species of each genus the tomentum is partly evanescent, or occasionally it is conglutinated into spines or warts. Below the tomentum is an upper cortex of pseudoparenchyma, usually formed by the partial disintegration of a palisade of large thick-walled hyphae which occasionally penetrate between the algal filaments. In Coccocarpia the pseudoparenchyma is thin and scarcely distin- guishable from the tomentum or even absent. The algal zone varies from a dense palisade of filaments of Scytonema and hyphae in some species, to ellipsoidal colonies of Nostoc sur- rounded by hyphae. The medulla in most genera consists of loosely tangled hyphae, occasionally not even well differentiated from the algal layer, while in Coccocarpia it is formed of septate periclinal hyphae, scarcely to be distinguished from the lower cortex except in color and the size of the intercellular spaces. The lower cortex may be completely absent or consist of one or more layers of thick-walled periclinal hyphae which give rise to [Vor. 20 440 ANNALS OF THE MISSOURI BOTANICAL GARDEN tufts of rhizinae. The so-called hypothallus in most species is a thick dense mat of rhizinae, which may be absent in Erioderma, or occasionally only slightly developed in certain species of the other genera. The morphology of the apothecium is equally confusing. In Pannaria and Malmella an amphithecium is present, usually with a pseudoparenchymatous cortex, and a thin algal layer which may be separated from the parathecium by a thin medulla. The parathecium is pseudoparenchymatous, formed from thick- walled periclinal hyphae in Lepidocollema, Parmeliella, Cocco- carpia, Erioderma, and the smooth species of Pannaria, while in the isidiose species of Pannaria and in Malmella the parathecium is filamentous, in the former of large thin-walled cells, in the latter of slender hyphae. The hypothecium is differentiated somewhat in the species of Pannaria but practically undiffer- entiated in the other genera. Instead of the above interpretation those with a filamentous parathecium might be conceived as lacking a parathecium and having a somewhat highly developed hypothecium, or perhaps the so-called parathecium is really only a compact medulla of the amphithecium. Further study of the morphology of the developmental stages, also more representa- tives of some of the genera, are necessary before the correctness of any interpretation can be assured. KEY TO THE TROPICAL AMERICAN GENERA OF PANNARIACEAE Thallus with prm apothecia lecanorine, rhizinae little developed es a es oe KG eae a eee hee EE TELE LU Psoroma cinchonarum Thallus with het rhizinae usually well developed. Upper cortex thin, pseudoparenchymatous, algal layer occupying most of the thallus; apothecia biatorine; Brasil.......... Lepidocollema carassense Upper cortex well developed, peeudoparenchymat Upper cortex of two or more rows of cells; bros etd to the surface, medulla loosely woven. Apothecia lecanorine. m absent or very early evanescent, algae definitely Nostoc, colonies — separate; parathecium usually tou Tomentum thin but kai algae Nostoc ? or perhaps Scytonema; parathecium of slender hyphae................. 0000000 Malmella Apothecia biatorine, pseudoparenchymatous from periclinal hyphae. Tomentum absent, algae Nostoc; rhizinae well developed... .Parmeliella Tomentum present and highly developed, algae Shiai or Nostoc; lower cortex little developed; apothecia large, mar- | e ee E err E ree rioderma 1933] DODGE—LICHENS OF COSTA RICA. I 441 Upper cortex of septate periclinal hyphae which may simulate pseudo- parenchyma, tomentum usually absent; apothecia biatorine, medulla of periclinal conglutinate hyphae appearing pseudoparenchymatous, des e e 5 5 ck enh s 8s cc res ceves ces oe aes Oe aR Coccocarpia PANNARIA Del. ap. Bory PANNARIA Del. ap. Bory, Dict. Class. Hist. Nat. 13: 20. 1828. Type species: Pannaria rubiginosa (Thunberg ap. Ach.) Del. ap. Bory. Thallus granular, squamose to foliose with a well-developed bluish black or black hypothallus, rarely with dark, more or less tangled rhizinae below, heteromerous, upper surface corticate with large-celled pseudoparenchyma formed from a palisade of hyphae; algae Nostoc, medulla single or double, in the former case arachnoid, in the latter the upper portion of more or less parallel, thin-walled hyphae, loosely woven, and below of densely tangled hyphae, without lower cortex. Apothecia at first sunk in the thallus, finally sessile or peltate, superficial; amphithecium pseudoparenchymatous with a few algae in the center; hypo- thecium hyaline; asci clavate, 8-spored; ascospores hyaline, elongate, ellipsoidal to almost fusiform with a somewhat thick- ened and finely verrucose wall. Spermagonia in hemispherical warts, spermatiophores septate with short, broad cells; spermatia straight or very slightly curved, elongate, cylindrical. This genus seems one of the most variable and widespread of the family, occurring from the Arctic to the Antarctic, in Costa Rica occurring from sea level to 1700 m. In tropical America the genus separates easily into two sections, the isidiose species with a filamentous parathecium, and the smooth species with a pseudoparenchymatous parathecium. KEY TO THE TROPICAL AMERICAN SPECIES OF PANNARIA Hypothallus not well developed, pale with pale rhizinae below. Thecium 80-100 x or less tall; parathecium filamentous; hypothecium not differentiated; spores 11-12 x 4-5 u, more or less constricted at the middle and appearing 2-celled; S. Afriea............... suus P. rubiginosa Thecium 100-120 u or more tall; parathecium of septate hyphae; hypo- thecium pseudoparenchymatous; spores 13-16 x 7-9 u; Rio Grande do Sul, Brasil [P. rubiginosa Malme non aliorum]. ............... P. Malmei Hypothallus developed, sometimes sparingly. Hypothallus pale, thallus squamulose, pale olivaceous to rusty yellow, margin cut-crenate, irregularly lobulate; spores 8-9 x 2.5-3 u; Mexico. Ee ate ce s uet a iho whe 24 M Cee TOOT er e P. applanata [Vor. 20 442 ANNALS OF THE MISSOURI BOTANICAL GARDEN Hypothallus black. Thallus isidioid. Margins and surface with cylindrical, coralloid branched c spores 14-19 x 7-8.5 y; Antilles and Brasil............... . stylophora Margins isidio-lacinulate; spores 8-10 x 5-6 u, with thick epispore; Minos Gurus, BEAD... oee khao hmm ont ct COR ead P. isidioidea Margins with verruciform isidia, lobes under 1 mm. broad; apothecia 1-1.5 mm. in diam.; spores 12-14 x 8-10 4, wall thin and smooth. Thallus without isidia or soredia Thallus livid fuscescent, adini broad-lobed; spores 14 x 6 4; hypo- thallus less developed, center glebulose-lobed; São Paulo, Brasil. Duns Notas EMO EE T E TESTEN EE . imbricatula — pallid or ashy or pale glaucous, lobes subeuneate, up to m. broad; spores 12-15(-18) — ig 10) u; parathecium and iS »othecium pseudoparenchymatous.............6 2000.00: P. Vainii Thallus coeruleo-ashy, lobes tna peat epithecium black aeruginous; spores 12-20 x 7-8 yu, surface subplicate; São Paulo en, Temi P. caeruleo-nigricans Thallus livid glaucescent or lurid pallid, laciniae very Peeler radiate-stellate, even subimbricate............. 00.00.00 eee . radiata Thallus pale fuscescent, lobes 2-2.5 mm. broad and about 6 mm. long; epithecium rufous; spores 12-19 x 8-10 yw; parathecium Bici arose vndis hypothecium of conglutinate vertical ivl: DM A ooeviséveré VC he E443 443 Axio ERRORES P. brasiliensis PANNARIA ISIDIOIDEA Vainio, Ann. Acad. Sci. Fenn. A6’: 102. 1915. Pannaria Mariana var. isidioidea Vainio, Étude Lich. Brésil 1: 206. 1890, excl. syn. Type: Brasil, Minas Geraes, Sitio, 1000 m., Vainio 669, 983 (sterile). The following description is based on fertile specimens from Costa Rica. Hypothallus of dark green rhizinae extending only slightly beyond the thallus; thallus dark olive buff, margins lighter, surface very minutely tomentose, pinnatifid dissected, lobes somewhat cuneate, about 1 mm. wide, ultimate lobules rounded, 0.25—0.5 mm. broad, toward center becoming microphylline and subisidiose but idis somewhat flattened, rarely cylindric, 150- 160 u thick, upper 40 u of pseudoparenchymatous cortex, algal layer of Nostoc about 40 y thick, and medulla 40 y thick, of loosely woven hyphae with occasional small chains of Nostoc; rhizinae thick-walled, black, 6-8 yw in diameter. Apothecia crowded, peltate, constricted at the base, about 1 mm. in di- ameter, margin prominent, crenulate, almost lobulate, disc 1933] DODGE—LICHENS OF COSTA RICA. I 443 chestnut; amphithecium 200 y thick, of the same texture as the cortex and algal layer of the thallus; parathecium 60 y thick, filamentous, of large thin-walled periclinal hyphae, ending above in a false pseudoparenchyma; hypothecium about 20 y thick, of similar structure but more deeply staining; thecium 80-100 y tall; paraphyses slender, filiform, somewhat clavate above in the brownish epithecial gel; asci clavate, 12 » in diameter; ascospores ellipsoidal with a thick epispore, 8-10 x 5-6 u (perhaps still immature). This species is somewhat variable in appearance and perhaps should be separated into varieties and forms but I have not seen sufficient material to do so properly. Sterile material has been referred here solely on the structure of the thallus. Dodge & Thomas 6558 from Tilarán has a white reticulate cortex. Cartago: R. Birrís, 1220-1340 m., Dodge & Thomas 4559. Heredia: C. Central de Zurquí, 1600-1700 m., Dodge, J. Valerio & Thomas 6048. Alajuela: Piedades de S. Ramón, 900 m., Brenes 375. Guanacaste: H. Santamaría, 640-680 m., Dodge & Thomas 6823, 6911, 6988; Tilarán, 500-690 m., Dodge & Thomas 8059, 6558. Var. pulvinata Dodge, var. nov. Type: Costa Riea, Alajuela, C. de Pata de Gallo a S. Rafael de S. Ramón, 1200-1250 m., Brenes. Thallus ut in P. isidioidea sed isidiosissimus, marginibus adscendentibus, cinerascentibus; isidia coralloidea, elongata, pulvinum formantia. Thallus as in P. tsidioidea but densely isidiose with ascending, cinerascent margins; the isidia coralloid, repeatedly branched, erowded, forming a cushion up to 2 cm. in diameter and 3-4 mm. thick. This variety might easily be mistaken for a sterile pulvinate Stereocaulon or a Siphula, were it not for the occasional lobe at the margin which shows its relationship. Alajuela: C. de Pata de Gallo a S. Rafael de S. Ramón, 1200-1250 m., Brenes. PANNARIA Moseni Dodge, sp. nov. Pannaria rubiginosa Malme, Ark. f. Bot. 20%: 7-8. 1924, pro parte. Type: Brasil, São Paulo, Sororocaba near Santos, Mosen 3284. Thallus albidus, subpinnatifidus, lobis 3-5 mm. longitudine, 0.8 mm. latitudine, apicibus rotundatis, marginibus loborum [Vor. 20 444 ANNALS OF THE MISSOURI BOTANICAL GARDEN verrucosis, vel etiam in centro thalli isidiosus, ad 140 y crassitu- dine, cortex superior 30-35 y crassitudine, pseudoparenchymaticus ex hyphis perpendicularibus formatus, zona nostocacea coloniis subsphericis 35 y diametro metientibus, medulla ad 60 y crassi- tudine hyphis laxe implexis, rhizinae brunneo-nigricantes, hyphis nigris pachydermaticis 3-4 u diametro, conglutinatis. Apothecia 1-1.5 mm. diametro, marginibus tenuibus, crenato-lobatis et subisidiosis; amphithecium inferne 120 y. crassitudine superne ad 60 y attenuatum, cortice 40 y crassitudine, pseudoparenchy- matico; parathecium bene evolutum, inferne 40 u superne ad 60 u crassitudine, hyphis ad 4 y diametro leptodermaticis con- textum; hypothecium 25 y crassitudine, hyphis periclinalibus dense compactum; thecium ad 100 y altitudine; paraphyses filiformes 1-2 u diametro apicibus clavatis; asci clavati, 25 u diametro; ascosporae octonae, distichae, 12-14 x 8-10 u, epi- sporio tenuiusculo. Thallus pearl gray, subpinnatifid, lobes 3-5 x 0.8 mm., with rounded tips, margins of lobes verrucose or even occasionally cylindric, isidiose toward the center, about 140 y thick, upper cortex 30-35 u, pseudoparenchymatous from the breaking up of a palisade layer, algal zone of subspherical colonies of Nostoc about 35 u in diameter, medulla about 60 u thick, of loosely woven hyphae; rhizinae of brownish black thick-walled conglu- tinate hyphae 3-4 y in diameter. Apothecia 1-1.5 mm. in diameter, margins thin, crenate-lobulate or even somewhat isidiose; amphithecium about 120 y thick below, thinning to about 60 u above with a cortex about 40 u thick of pseudoparen- chyma; parathecium well developed, about 40 u below, spreading to 60 u above, of interwoven thin-walled hyphae about 4 u in diameter; hypothecium about 25 y thick of dense periclinal hyphae, staining very deeply; thecium about 100 y tall; para- physes filiform, 1-2 y in diameter, with clavate tips; asci clavate, 25 y in diameter, 8-spored; ascospores distichous, 12-14 x 8-10 y, with a somewhat thinner epispore. It is with some hesitation that I have described this species as new. ‘The thallus is very suggestive of the sterile P. stylophora, although it has a somewhat different color and lobing and rather larger dimensions of the parts. It does not seem to be closely 1933] DODGE—LICHENS OF COSTA RICA. I 445 related to P. rubiginosa, originally described from Thunberg’s South African collections, and is certainly not closely related to the material referred to that species in the North Temperate zone. So far I have only seen this species from the Atlantic Coastal Plain of Costa Rica. Limón: Hamburg, Standley & J. Valerio 48702, 48717; R. Siquirres, 70-200 m., Dodge, Catt & Thomas 8037. PANNARIA STYLOPHORA Vainio, Ann. Acad. Sci. Fenn. A6’: 102. 1915. Type: Antilles, Guadeloupe, Sofaga, P. Duss 1387 (sterile). Hypothallus black, extending slightly beyond the thallus; thallus with pale olive gray margins and olive gray center, lobes long, slender, oblong, 0.5-0.8 mm. wide, and primary lobes about 15 mm. long, irregularly more or less pinnatifid, smooth above around the margin, the center with scattered to crowded cylindrical isidia which are rarely dichotomously branched, and concolorous, about 125 y thick, upper cortex 20-25 yu, pseudo- parenchymatous from the development of a palisade layer; algal layer about 40 y thick, consisting of ellipsoidal colonies of Nostoc about 30 x 40 u; medulla about 60 y. thick, loosely woven; lower cortex absent; rhizinae of black thick-walled hyphae about 3- 4 y in diameter. Limón: R. Siquirres, 70-200 m., Dodge, Catt & T'homas 8036. PANNARIA Vainii Dodge, sp. nov. Pannaria rubiginosa Vainio, Étude Lich. Brésil 1: 204—205. 1890; Malme, Ark. f. Bot. 20%: 7-8. 1924, quoad spec. Dusenia- num. Type: Brasil, Minas Geraes, Sitio, 1000 m., Vainio 606. Hypothallus black, extending 1-2 mm. beyond thallus; thallus between olive buff and smoke gray, irregularly and repeatedly laciniate, lobes 1-2 mm. broad, ultimate lobes short, broader ones up to 4 mm. long, smooth above, not isidiose, about 160- 170 u thick, upper cortex 20-30 y thick of 2-3 layers of large isodiametric cells; algal layer of spherical to ellipsoidal colonies of Nostoc about 60-65 y. in diameter, algae rather large and some- what angular by mutual pressure, medulla about 80 y thick of large, thin-walled hyphae, loosely interwoven; rhizinae of black, thick-walled, conglutinate hyphae. Apothecia peltate, 0.8-2 [Vor. 20 446 ANNALS OF THE MISSOURI BOTANICAL GARDEN mm. in diameter, margin thick, crenulate, dise rufous; amphi- thecium 130-440 y. thick, with cortex 20-30 » thick; parathecium well developed, about 60 y thick of periclinal large, thin-walled, hyphae simulating pseudoparenchyma; hypothecium about 20 u thick, of similar structure but of more slender less frequently septate hyphae; thecium 90-100 y tall; paraphyses 1.5-2.5 w in diameter, filiform, slightly clavate at the tips in a yellowish brown epithecial gel; asci 14-16 » in diameter, 8-spored; asco- spores distichous, broadly ellipsoidal, 12-18 x 8-10 y, with thick epispore. This species is another of the group which in various herbaria is often more or less indiscriminately determined as P. rubiginosa, P. Mariana, or even P. nigrocincta, usually without much critical attention to microscopic details. Guanacaste: near Tilarán, both Atlantic ye Pacific slopes, 500-690 m., Standley & J. Valerio 44810, Dodge & Thomas 6564, 6565 PANNARIA radiata (Vainio) er pn inei nov. Pannaria Mariana f. radiata Vainio, Ann. Acad. Sci. Fenn. A6': 101-102. 1915, excl. syn. Type: Antilles, Guadeloupe, P. Duss. Hypothallus brownish black, extending about 1.5 mm. beyond the margin of the thallus, thick; thallus mineral gray with ends of lobes pale smoke gray, dichotomously (rarely trichotomously) divided, laciniae 1 cm. or more long, about 1 mm. broad, tips rounded, reflexed, wholly discrete in outer centimeter, thallus about 90 y thick, upper cortex of large thin-walled pseudo- parenchyma 20-30 y thick, algal zone of Nostoc colonies ellip- soidal, about 40 y in long diameter, medulla about 20 y. thick of loosely woven hyphae; rhizinae of conglutinate black thick- walled hyphae very closely woven at the lower side of the medulla and performing the function of a lower cortex. Apothecia mostly about 1 mm. in diameter, occasionally up to 2 mm., peltate, margin crenate, of medium prominence, dise rufous; amphi- thecium 160 y thick, cortex about 60 y thick, pseudoparenchy- matous, Nostoc colonies smaller than in the thallus, separated by strands of pseudoparenchyma; parathecium pseudoparenchyma- tous, 60-80 y. thick, filamentous, of large very thick-walled hyphae appearing pseudoparenchymatous; hypothecium 15-20 y thick, 1933] DODGE—LICHENS OF COSTA RICA. I 447 of periclinal, thin-walled, slender hyphae; thecium 120-130 u tall; paraphyses slender, filiform, septate, without clavate tips; asci 10-12 y thick, 8-spored; ascospores irregularly distichous, ellipsoidal, 15-16 x 7-8 u, with a very thick epispore. This species shows little relationship to P. Mariana Fr. of the Old World, a species to which it has been referred. It seems closest to P. Vainii but differs in microscopic details and habit. Limón: T Standley & J. Valerio 48768; R. Siquirres, 70-200 m., Dodge, Catt & Thomas 5 MALMELLA Dodge Malmella Dodge, gen. nov. Type species: Erioderma physcioides Vainio. Thallus squamosus aut microfoliosus, rhizinis nigris bene evolutus, tomentum superficiei superioris parce evolutum tenue, hyphis implexis sed non conglutinatis ut in Coccocarpia; cortex superior pseudoparenchymaticus, algae nostocaceae; medulla hyphis laxe implexis, cortex inferior tenuis hyphis nigris con- glutinatis aut nullus. Apothecia lecanorina, sessilia, sparsa; amphithecium adest; parathecium non pseudoparenchymaticum; paraphyses filiformes; asci clavati; ascosporae octonae, ellip- soideae, episporio crasso. Thallus squamose or small foliose, with a well-developed hypothallus of black rhizinae; tomentum of the upper surface less well developed than in Erioderma, of slender hyphae loosely woven, not conglutinate as in Coccocarpia; upper cortex pseudo- parenchymatous, algae Nostoc ?, medulla of loosely tangled hyphae, lower cortex of black hyphae as in Coccocarpia or wholly absent. Apothecia lecanorine, sessile, scattered over the upper surface, not marginal as in Érioderma; amphithecium present; parathe- cium of slender hyphae, not pseudoparenchymatous; paraphyses filiform; asci clavate, 8-spored; ascospores ellipsoidal, with a thick epispore. This genus seems intermediate between Pannaria, Erioderma, and Coccocarpia, and collections have been previously referred to each of these genera. In structure of the thallus it seems intermediate between Pannaria and Erioderma, while in super- ficial appearance it suggests the more laciniate species of Coc- cocarpia. [Vor. 20 448 ANNALS OF THE MISSOURI BOTANICAL GARDEN KEY TO THE TROPICAL AMERICAN SPECIES OF MALMELLA Thallus becoming densely isidiose. Thallus olivaceous with isidia becoming pruinose and subsorediate, then Done 3 o1 EET ORNETTE PERPE RSS DDR . Randii Thallus white, more laciniate............. 00.000 nnn BIA M. albida Thallus not isidiose, medulla not well differentiated from algal zone. Thallus white, 120 u thick, thecium 140-160 y tall.............. M. Santamariae 'Thallus olive buff or darker Thallus olive buff, 120 4 thick, thecium 100-120 à tall........... M. physcioides Thallus dark olive with white pruina about margins, 200 y thick, thocium 80-80 p Wl. s ioosasaso uae e aH rera 1. caestocinerea MALMELLA Randii Dodge, n. sp. Type: Maine, Mt. Desert Island, E. L. Rand. Hypothallus rhizinis viridi-nigricantibus; thallus olivaceus, lobis subdichotomis aut irregularibus, 1.0-1.3 mm. latis, margini- bus crispatis elevatisque, verrucosis et etiam isidioideis, glaber in lobis marginalibus, centro isidiosissimus, apicibus isidiorum pruinosis et subsorediosis; thallus 140-160 u crassitudine, to- mentum 12 y crassitudine hyphis laxe implexis, tenuibus, sub- dilabentibus; zona nostocacea [vel scytonematica] 100 y crassi- tudine, filamentis dense convolutis, in coloniis ellipsoideis, cellulis sphericis vel angularibus; medulla hyphis laxe implexis 2-3 u diametro, aut plus minusve hyphis subparallelis ut in Coccocarpia pellita, frequenter ad margines; cortex inferior hyphis conglutinatis nigris tenuis. Hypothallus of well-developed greenish black rhizinae extend- ing about 0.5 mm. beyond the thallus; thallus ecru olive to buffy olive, main lobes subdichotomous or wholly irregular, 1.0-1.3 mm. broad, margin crisped and elevated, becoming verrucose and finally isidioid, surface appearing like surface of blotting paper under microscope, smooth on marginal lobes, becoming densely isidiose toward the center, the tips of the isidia soon becoming pruinose and subsorediate, giving a puritan gray ap- pearance to the plant; thallus about 140—160 u thick, the tomen- tum 12 y thick of loosely appressed, somewhat tangled, slender hyphae which more or less disintegrate; the upper cortex 20 u thick, of about 2 layers of cells from the palisade; algal zone about 100 » thick, of closely coiled filaments, forming ellipsoidal colonies, cells spherical to angular from mutual pressure, Nostoc or perhaps Scytonema; medulla of loosely woven hyphae 2-3 y 1933] DODGE—LICHENS OF COSTA RICA. I 449 in diameter, in places more or less parallel as in Coccocarpia pellita, especially at the margin of the thallus with about 1 row of agglutinated black thick-walled hyphae functioning as a lower cortex, giving off rhizinae either as single hyphae or as small fascicles of hyphae; sterile. This species has not been seen fertile and its systematic position is uncertain. It has somewhat the appearance of an isidiose Malmella physcioides but differs from the latter in a less well- developed tomentum, a thinner upper cortex, and a thicker algal zone. It differs much more from Coccocarpia cronia in the loosely tangled tomentum layer which is present in Coccocarpia but closely appressed and formed of longitudinal hyphae, in the loosely tangled medulla (in Coccocarpia almost pseudoparenchy- matous of periclinal hyphae), and the much less developed lower cortex. MALMELLA albida Dodge, sp. nov. Type: Costa Rica, Limón, R. Siquirres, 70-200 m., Dodge, Catt & Thomas 8060. Hypothallus crassus, niger, bene evolutus; thallus albidus, cinerascens, laciniatus, di- vel trichotomus aut irregularis, laciniis linearibus, 0.2-0.5 mm. latis, et superficiei marginibusque dense isidiosis, coralloideis, inferne niger, rhizinis nigricantibus; thallus 80-100 u crassitudine, cortex superior hyphis periclinalibus 5 u et cellulis pseudoparenchymatibus 20 y. metiens, zona nosto- cacea (scytonematica?) 60 y crassitudine, filamentis contortis, cellulis sphericis, 4-5 u diametro; medulla 40-45 y. crassitudine, hyphis laxis implexis; cortex inferior 8-12 y. crassitudine, hyphis nigris pachydermatibus conglutinatis compactus. Sterilis. Hypothallus thick, black, well developed; thallus chalk- white, becoming ashy from isidia, laciniate, di- or trichotomous or wholly irregular, laciniae linear, 0.2-0.5 mm. broad, upper surface and margins densely isidiose, coralloid, below black with black rhizinae; thallus 80-100 » thick, upper cortex about 25 y thick, the outer 5 v. of periclinal hyphae, below which are about two layers of pseudoparenchymatous cells; the algal layer of Nostoc (Scytonema ?) about 60 » thick, of contorted filaments of spherical cells 4-5 y. in diameter; medulla 40-45 y. thick, of loosely tangled hyphae; lower cortex 8-12 y thick, of thick-walled, [Vor. 20 450 ANNALS OF THE MISSOURI BOTANICAL GARDEN black, conglutinate hyphae; sterile. Known only from a single small collection in Costa Rica. Limón: R. Siquirres, 70-200 m., Dodge, Catt & Thomas 8060. MALMELLA Santamariae Dodge, n. sp. Type: Costa Rica, Guanacaste, H. Santamaría, 640-680 m., Dodge & Thomas 8049. Hypothallus nullus sed rhizinae coeruleo-nigricantes, dense implexi, ei simulantes; thallus albidus, minute tomentosus, palmatim lobatus, lobis oblongis, 0.5-1 mm. latitudine, 120 y crassitudine, cortice filamentosa, 15-20 y crassitudine hyphis pachydermaticis periclinalibus; strato nostocaceo 40 u crassitu- dine in medulla gradatim transeunte, rhizinis nigris pachy- dermaticis 4-5 y. diametro metientibus. Apothecia 0.5-0.75 mm. diametro marginibus crassisimis crenulatis inflexis, disco nigro; amphithecium inferne 140 y. crassitudine superne ad 100 y attenuatum, cortice 40 y crassitudine pseudoparenchymatico; parathecium 20-30 y. crassitudine, hyphis tenuibus periclinalibus dense implexis; thecium 140-160 y altitudine; paraphyses tenues, filiformes, apicibus in gelatina dilabentibus; asci clavati, 12 y. dia- metro; ascosporae distichae, octonae, 10-11 x 7-8 y, ellipsoideae, episporiis crassis. Hypothallus of densely tangled bluish black rhizinae; thallus white, minutely tomentose, palmately lobed, lobes oblong, 0.5-1.0 mm. broad, 120 y thick; cortex filamentous, 15-20 y thick, of thick-walled hyphae, ind layer of Nostoc qradunlly merging into the medulla; rhizinae thick-walled, black, 4-5 y in diameter. Apothecia 0.5-0.75 mm. in diameter, with very thick inrolled margins and blackening disc; amphithecium 140 u thick below, thinning to 100 & at the margin, cortex 40 u thick, pseudo- parenchymatous; parathecium 20-30 y thick, of densely ap- pressed periclinal hyphae not clearly differentiated from the hypothecium; thecium 140-160 y. tall; paraphyses very slender, filiform, degenerating into the dark brown epithecial gel; asci clavate, 12 u in diameter, 8-spored; ascospores distichous, ellipsoidal, with thick epispore, 10-11 x 7-8 y. While eventually this species may be found to have a much wider distribution, it seems fitting that the excellent H. Santa- maría, with one of the richest lichen floras known to me, should be 1933] DODGE—LICHENS OF COSTA RICA. I 451 commemorated in botanical literature. The hacienda in turn preserves the name of the national hero of Costa Rica not far from the scene of his famous exploits. Guanacaste: H. Santamaría, 640-680 m., Dodge & Thomas 8049. MALMELLA physcioides (Vainio) Dodge, n. comb. Erioderma physcioides Vainio, Jour. Bot. Brit. & For. 34: 70. 1896. Type: West Indies, St. Vincent, Boxwood, W. R. Elliott 252. Hypothallus sparingly developed, strigose-pannose, greenish black; thallus deep olive buff, radiately subpinnatifid, ultimate lobes 1-1.5 mm. broad, short and convex, surface toward the margin slightly spongy, very minutely verrucose, about 120 y thick, upper cortex about 40 y thick, of large thin-walled pseudo- parenchyma, the rest of loosely woven hyphae with very small, separate colonies of Nostoc, the lower 10 u with more densely woven, more or less parallel hyphae. Apothecia circular to quite irregular by mutual pressure, crowded, about 1 mm. in diameter, margin concolorous with the thallus, crenulate, disc chestnut; amphithecium about 200 u thick, the outer 60 y of tomentum consisting of slender, thin-walled hyphae more or less perpendicular to the surface and somewhat tangled. Within this is the true cortex about 30 y thick of large-celled pseudo- parenchyma, thinning out to a single layer above, and the algal layer of slender periclinal thin-walled hyphae and colonies of Nostoc about 100 y thick below and about 80 v above; parathecium about 40 yu thick, of densely woven periclinal hyphae, staining very lightly and reaching the margin above; hypothecium deeply staining, 15-20 y. thick, of densely woven hyphae; thecium 100- 120 y. tall; paraphyses filiform, 2 » in diameter, frequently septate, disintegrating into the brownish epithecial gel; asci clavate, 6-8 u in diameter, 8-spored; ascospores broadly ellipsoidal, 10- 12 x 6-8 y, with a gelified wall, minutely roughened within. Previously reported only from the type locality. I have previously distributed duplicate material of this species under the name P. imbricatula Müll. Arg., which was originally briefly described from southern Brasil. Only a careful study of both types can settle its identity and relationships. Alajuela: Fraijanes, 1500-1700 m., Standley & J. Valerio 47651; S. Isidro de Alajuela, 980-1300 m., Dodge, J. Valerio & Thomas 4883. Guanacaste: H. Santamaría, 640—780 m., Dodge & Thomas 8034, 8035. [Vor. 20 452 ANNALS OF THE MISSOURI BOTANICAL GARDEN MALMELLA caesiocinerea (Vainio) Dodge, n. comb. Pannaria rubiginosa f. cinerascens Vainio, Jour. Bot. Brit. & For. 34: 70 [10]. 1896, non Nyl. Pannaria rubiginosa f. caesiocinerea Vainio, Acta Soc. Sci. Fenn. A6': 102-103. 1915. Type: not stated although several specimens mentioned. P. rubiginosa f. cinerascens Vain. based on Antilles, St. Vincent, Mt. St. Andrews, 320 m., Elliott 15. The following description is based on Costa Rican material. Rhizinae bluish black, highly developed and tangled; thallus buffy brown in the center, lobes dark olive with a white pruina about the margins, tomentose, lobes rounded, 1-4 mm. broad, sinuate, incised, the lobules oblong, about 0.5 mm. broad, convex, margins thick, inflexed below, 200 u thick, the upper cortex a palisade of thick-walled hyphae about 20 u thick, homoeomerous, with coiled filaments of Nostoc distributed throughout a very loosely tangled layer of hyphae, fraying out below without lower cortex. Apothecia up to about 0.5 mm. in diameter, margin very thiek and inrolled, white, crenulate, disc black; amphithecium thick, continuous with the thallus, about 140 y thick, with cortex about 60 » thick below and with algae much more abundant and closely tangled than in the thallus; para- thecium thin, filamentous, 20 y. thick, not differentiated from the hypothecium; thecium 80-90 y. tall; paraphyses filiform, septate, ending in the hypothecial gel which is deep brown and deeply staining; asci clavate, about 20 u in diameter, 8-spored; asco- spores distichous, ellipsoidal, 11-12 x 6-7 y, with a gelified epispore about 2 u thick, sometimes appearing 2-celled or even 4-celled with 2 transverse septa and 1 longitudinal septum in the middle, probably due to oil globules or other highly refractive substances, as other spores in the same ascus seem to be definitely unicellular. This species suggests a close relationship with Lepidocollema but the cortex is much better developed and the apothecium is lecanorine. Guanacaste: H. Santamaría, 640—680 m., Dodge & Thomas 8038. PARMELIELLA Müll. Arg. PARMELIELLA Müll. Arg., Mem. Soc. Phys. Hist. Nat. Genéve 16:376. 1862. 1933] DODGE—LICHENS OF COSTA RICA. I 453 Type species: P. triptophylla (Ach.) Müll. Arg. [P. corallinoides (Hoffm.) Zahlbr.] Thallus squamose, becoming elongate and subfoliose at the margins with a well-developed dark hypothallus, algae Nostoc; upper cortex of thick-walled pseudoparenchyma; medulla loosely woven; lower cortex an appressed layer of rhizinae. Apothecia superficial; amphithecium absent; parathecium pseudoparenchy- matous; asci 8-spored; spores hyaline, unicellular, elongate- ellipsoidal with thin walls; spermatia short, straight. This genus seems world-wide in its distribution, perhaps predominantly of the southern hemisphere but widespread in the North Temperate zone and in the tropics, preferring the higher elevations. Only P. pannosa has yet been found in Costa Rica. The other species have been incompletely described but have characters which should make them recognizable if found. KEY TO THE TROPICAL AMERICAN SPECIES OF PARMELIELLA Hypothallus indistinct, thallus glauco-cinereous or pale, squamules c fluent, crenate, granulate; spores small, 8-9 x 4-6 u; Chile... . P. scored likes Hypothallus forming a black margin about the thallus Hypothallus very thin, blue green; spores 15-19(-21) x 6.5-8 u; surface smooth; thecium 100 u thick; Chile................ 0 ee ee eeeee . nigrocincta Hypothallus blue black, densely woven, cortex a single layer of cells; spores 11-17 x 5-8 u; thecium 110 y thick; Mexico.......... P. miradorensis Hypothallus brownish black, densely woven and thick, cortex several layers of pseudoparenchyma; spores 12-15(-18) x 6-8 u; thecium ij, | rr NEU IU vos ir on P. pannosa Hypothallus very thick; apothecia 1-1.5 mm. in diam.; spores 16-25 x 8-10 »; BOWER... 05 one cc ee ee n ie eee P. nigrocincta var. Weddellit PARMELIELLA PANNOSA (Sw.) Miill. Arg., Flora 64: 86. 1881. Lichen pannosus Sw., Nova Gen. Sp. Pl. Prodr. 146. 1788; Fl. Ind. Occ. 3: 1888. 1806. Type: Jamaica, O. Swartz. Hypothallus brownish black, extending 1-2 mm. beyond the thallus; thallus tawny olive to isabelline, irregularly somewhat pinnatifidly divided in the margin, the central portion of the thallus becoming appressed microphylline and even somewhat isidioid, margins of lobes somewhat lighter in color, otherwise smooth above; thallus about 100 y. thick, the upper cortex 40 y thick, of large thick-walled spherical cells forming a pseudo- parenchyma, algal zone thin, about 20 y thick, of small somewhat [Vor. 20 454 ANNALS OF THE MISSOURI BOTANICAL GARDEN confluent colonies, medulla about 40 y thick, of large thin-walled, loosely woven hyphae, the lower cortex 10-12 u thick, of con- glutinate, black, thick-walled hyphae. Apothecia abundant, about 1 mm. in diameter, round to irregular in shape, the margin lighter colored when young, appearing to be thalline but not so, darkening in age but always slightly lighter than the disc and not exceeding the epithecium in height, disc rufous; amphi- thecium absent; parathecium 100-120 » thick, pseudoparenchy- matous, of large thick-walled cells extending to the top of the thecium; hypothecium about 40 y thick, of densely woven hyphae, deeply staining; thecium 120-140 y thick; paraphyses branched above, 1.5-2 y. in diameter, with thin walls; asci clavate, about 15 y in diameter, 8-spored ; ascospores irregularly distichous, ellipsoidal, 12-15 x 6-8 y, epispore thick. This species seems to be the commonest one in tropical America, in our area confined largely to higher elevations than Pannaria, being most abundant above 650 m., but rarely found at lower elevations. Without locality: Ørsted 8. Limón: Waldeck, Dodge 7403. Cartago: R. Pejivalle, 600—750 m., Dodge & Thomas 4341, 8038; Cartago, 1200— 1500 m., C. Wercklé (Mus. Nac. 17218). San dedi Sta. María de Dota, 1500-1800 m., Standley & J. Valerio 43394, 44057, 48444, 434406, 43472; La Hondura, 1300-1700 m., Standley 36458b. Heredia: Barba, El Gallito, R. Torres 150; C. Central de Zurquí, 1600-1700 m., Dodge, J. Valerio & Thomas 8039. Alajuela: 8. Ramón, La Palma, 1100 m., Brenes 64; Piedades, 900 m., Brenes 380; Alto Calera, 750-800 m., Brenes 265. Guanacaste: H. lantaa 680--780 m., Dodge & Thomas 6892; H. Granadilla, 500-600 m., Dodge & Thomas 6746. Var. coralloidea Dodge, var. nov. Type: Costa Rica, Heredia, C. Zurquí, 1600-1700 m., Dodge, J. Valerio & Thomas 8061. Thallus iteratim laciniatus sed non microphyllinus, isidiis coralloideis marginalibus rare sparsis, obscurior, 140 y. crassitu- dine, cortex superior 25-30 y crassitudine, duobus seriebus cellularum, serie superiore fusca, inferiore hyalina; stratum nostocaceum ad 60 u crassitudine, filamentis dense contextum; medulla hyphis magnis laxe implexa; cortex inferior ad 12 y crassitudine, hyphis nigris, pachydermaticis, conglutinatis, com- pactus; sterilis. 1933] DODGE—LICHENS OF COSTA RICA. I 455 Thallus repeatedly laciniate but not microphylline, isidia coralloid, mostly marginal, rarely scattered over the surface, darker, 140 y thick, the upper cortex 25-30 y thick, of two layers of thick-walled cells, the outer layer fuscous, the inner hyaline; the algal layer about 60 x thick, of densely woven filaments of Nostoc; medulla of large loosely tangled hyphae; lower cortex about 12 y thick of large black, thick-walled hyphae; sterile. Heredia: C. Central de Zurquí, 1600-1700 m., Dodge, J. Valerio & Thomas 8061. ERIODERMA Fée ERIODERMA Fée, Essai Cryptog. Ecorc. Offic. 146. 1824. Type species: Érioderma polycarpum Fée. Thallus foliose, attached to the substrate by either marginal rhizinae or those from the under surface; tomentum highly developed, cortex pseudoparenchymatous, formed by a palisade layer, the algal zone of short chains of Scytonema with thin sheaths or Nostoc; medulla loosely woven of thin-walled hyphae; no lower cortex, the lower surface somewhat veined and in some species with a dense covering of rhizinae. Apothecia marginal or superficial, peltate, constricted at the base; amphithecium absent; parathecium well developed, of large thick-walled cells, pseudoparenchymatous, with a loosely woven medulla; asci 8-spored; spores hyaline, unicellular, ellipsoidal, occasionally fusiform or subspherical. Spermagonia marginal, small, black- ish warts; spermatiophores septate; spermatia short, cylindric, straight. KEY TO THE TROPICAL AMERICAN SPECIES OF ERIODERMA Dense blackish fuscous tomentum below, tier hypothallus. Lobes narrow, deeply sinuate divided margins.................... E. Wrightii Lobes broader, margin with irai caesious soredia below, and white-villos6.......... rest re a a aa EOS . limbatum Nearly nude and veined, white below, although rhizinae bluish black or white, verruculose and villose above; Brasil.................. verruculosum SUME m below with white veins and eee coarsely éhorlipiiien e; apothecia up to 0.8 mm. in diam.; Jamaica.............. E. microcarpa Bluish white with fleshy white veins and incidi, m tufted d giving an net appearance when dry; apothecia 3-4 mm iam.; E E a E E R iD . pulchrum White Au thallus isidiose; apothecial margin isidiose, ashy virescent UL [go oo oo 8 8 ba nce 4 oe 0g 24 eee ss > cw E. Leylandi EnropERMA Wricuti Tuck., Amer. Jour. Arts & Sci. II, 25: 423. 1858 [Vor 20 456 ANNALS OF THE MISSOURI BOTANICAL GARDEN Type: Cuba, top of Loma del Gato, C. Wright. Hypothallus none, but dense brownish-black rhizinae clothe the under surface to within 3 mm. of the margin which is arach- noid-tomentose, white; thallus deep olive buff to citrine drab, sinuate-lobed, upper surface tomentose, margins with tufts of dark brown rhizinae, thallus about 120 y thick exclusive of tomentum on upper surface and rhizinae below; cortex pseudo- parenchymatous, 30—40 y thick, of large cells, algal layer very thin, 10-15 u thick, of densely coiled filaments of Scytonema; medulla very loosely woven, similar to the tomentum above; no lower cortex. Apothecia marginal with white densely pilose exciple and black disc; amphithecium absent; parathecium 200 y. thick, corticate with pseudoparenchyma about 25 y thick below, a little less above, formed from a palisade of hyphae perpendicular to the surface, the cortex covered with tomentum as in the thallus, the rest of the parathecium of large, loosely woven hyphae similar to the medulla but somewhat more closely woven; hypothecium not well differentiated from the parathecium, rather more densely woven; thecium about 120 y tall; paraphyses filiform, slender, branched above, ending in dense brown epithecial gel; asci clavate, 8-spored, 14-16 y in diameter; ascospores distichous, 12-14 x 6-7 y, ellipsoidal, with a thick epispore. This very distinct species resembles the Peltigeraceae in habit much more than the Pannariaceae, but it has the typical uni- cellular spore with the thick epispore of the latter family. It seems rather rare in tropical America but it is quite abundant in the few localities where it has been found. In Costa Rica it occurs from 680 to 1100 m. Alajuela: La Palma de S. Ramón, 1100 m., Brenes 50. Guanacaste: H. Santamaría, 680-780 m., Dodge & Thomas 6920, 6982. ERIODERMA limbatum (Nyl.) Dodge, n. comb. Erioderma Wrightw var. limbatum Nyl., Flora 52: 119. 1869. Type: Brasil, Rio de Janeiro, Serra dos Orgãos, Glaziou 2004. Thallus cinnamon buff to chamois, lobes rounded-sinuate, slightly excised, 5 mm. broad, margins crenate, thick, upper surface only slightly tomentose, lower surface with margins caesious-sorediate, white; rhizinae greenish black, very dense and thick; thallus about 300 » thick, upper tomentum not highly 1933] DODGE—LICHENS OF COSTA RICA. I 457 developed; cortex 25-30 y thick, pseudoparenchymatous from the disintegration of a palisade; algal layer about 60 u thick, of irregularly tangled filaments of irregular somewhat angular cells about 8 u in diameter (Scytonema ?); medulla about 60 y. thick next the cortex, of rather densely tangled hyphae, gradually becoming looser to form another indefinite layer 120 y. thick, of very large and very loosely tangled hyphae 10-12 y in diameter, with black hyphae of about the same size forming a more dense mat which serves as a lower cortex. Apothecia 3 mm. in diam- eter, stipitate, stipe 1-2 mm. tall, margin thin, erect or incurved, densely tomentose without, the hairs tangled, giving a granular appearance, dise chestnut, soon blackening; no amphithecium; parathecium highly developed, cortex a palisade of thick-walled infrequently septate hyphae proliferating to form the tomentum and medulla, 200 u or more thick below, thinning out at the edge of the thecium so that only the cortex about 100 u thick surrounds the upper portion of the thecium at the margin; hypothecium not differentiated; thecium about 90 y tall; paraphyses slender, filiform, apex not enlarged; asci cylindrical, about 8 y in diameter, 8-spored; ascospores monostichous, at least when found, thick- walled, hyaline, ellipsoidal, 10-12 x 5-6 p. This species, from its habit and color, might easily be mistaken for Sticta Weigelii or S. rufa except for a somewhat duller surface, but is distinguishable on the color and density of the rhizinae and the lack of cyphellae. Known from Costa Rica from a single collection with scant data by Brenes . . . a, probably from the region of S. Ramón in Alajuela Province. COCCOCARPIA Pers. ap. Gaudich. CoccocarPIA Pers. ap. Gaudich. in Freycinet, Voy. Uranie, Bot. 206. 1826. 'Type species: Coccocarpia molybdaea Pers. ap. Gaudich. Thallus squamose to foliose, with dark or light-colored rhizinae, corticate on both surfaces, upper surface pseudoparenchymatous from thin-walled large hyphae parallel to the surface, without tomentum; algal layer of coiled chains of Scytonema in thin sheaths; medulla of thin-walled somewhat conglutinate hyphae not sharply differentiated from the lower cortex, of more or less [Vor. 20 458 ANNALS OF THE MISSOURI BOTANICAL GARDEN septate hyphae running parallel to the surface. Apothecia superficial, sessile or somewhat constricted below; amphithecium absent; parathecium corticate with large-celled pseudoparen- chyma, within, of large septate periclinal hyphae, medulla lacking; hypothecium either light or dark; asci 8-spored; asco- spores hyaline, unicellular, spherical to ellipsoidal-fusiform, thin- walled. Spermagonia in warts on the thallus, spermatiophores frequently septate; spermatia straight, elongate-cylindric. This widespread tropical genus has two very widespread species or species groups and a number of rare and seemingly localized species scattered over the tropical regions. KEY TO THE TROPICAL AMERICAN SPECIES OF COCCOCARPIA Under side reticulate, black-fibrillose; lobes obtuse, ciliate, margin of apo- thecia crenulate, disc yellow..................ecececeeeees C. portoricensis Under side densely black-tomentose, forming a hypothallus. Lobes bipinnately divided, linear, ciliate, aeruginous-green........... C. subtilis Lobes cuneate to flabelliform, bluish to lead color. Thallus isidiose or microphylline in the center...................... C. cronia Thallus not reddish ferruginous within. Isidia concolorous, laciniae cuneate, 1.5-10 mm. broad. Laciniae 5-10 mm. broad, isidia applanate to microphylline. . . . TRCN SSM P SE ERKRRAJAKEC-E CREARE RE RR eons eh aan eh Ee Eas v. prolificans Laciniae 1.5-5 mm. broad, isidia terete; apothecia eni or rufo-fuscous, exciple white-ciliate................... v. isidiophylla Isidia usually darker at the tips and blackening, laciniae narrower and sublinear. Laciniae 1-1.5 mm. broad; hypothallus little developed; apo- thecia fuscous, blackening, exciple white-ciliate; Brasil. .v. camporum — 0.5-3.0 mm. broad, narrow granulose to lobulate in the er; apothecia black; hypothallus well developed. . . . v. granulosa Thallus hee ferruginous within; Cuba................ v. erythrocardia Thallus not isidiose in center... 2.2.0.0. .6. 00 ccc cece cee eee eee C. pellita Laciniae cuneate or broad-spathulate Hy ypot eg bie highly develaped; apothecia yellow to tawny POE PTT COPE ROLXXXAXERS AERIS v. pannosa es hs but much less well developed. Apothecia black from the first, — ere v. parmelioides Apothecia yellow, rufous, or fuscous. Thallus = thin, lobes anal, approaching microphylline C. cronia w. prolificans; apothecia finally TOT, «desees PENES S an ac ef en v. lividorufa Tui tiie lobes broad, 4-8 mm.; apothecia not dark- NE MOL. ee ree: v. pyrrhichocarpa Laciniae ststinene, up to 4 mm. broad, short, rounded. Apothecia black; Jamaica.................. ccc cece nnn v. genuina Apothecia fuscous, ciliate...........0000 0. cece eee ee eee v. strigdso 1933] DODGE—LICHENS OF COSTA RICA. I 459 Under side of thallus white with black-fibrillose margins, laciniae dick e. ri dM TT""—"-— C. fibrillosa Under ade of thallus white, rhizinae white. Thallus isidiose. Laciniae very narrow, 0.1—0.2 mm. broad, adnate............. C. dominguensis Laciniae broad and flabelliform, up to 10 mm. broad................ C. albida Thallus not isidiose. Laciniae 150—200 u broad, loosely adnate, ja rhizinae penicilloid, projecting beyond margin, dichotomous.................+ C. tenuissima Laciniae 250 u broad, tips ascending, "unt long white rhizinae, palmate with lacinulae pinnatifid; apothecia carneous, rigidly hi os54 00. P E E T E a ev . elegans Laciniae broader. Thallus greenish blue, rhizinae rigid, white; flattened apothecia , MMC 55 hia ade 4a eR NAE Pen epiphylla Thallus lead-color or slightly bluish, onea 28 bricate, pterygoid; apothecia testaceous, white, ciliate...... "C. asterella CoccocaRPIA PELLITA (Ach.) Müll. Arg., Flora 65: 320. 1882. Parmelia pellita Ach., Lichenog. Univ. 468. 1810, Sw., Lich. Amer. 7, pl. 6. 1811. Type: West Indies, O. Swartz. Laciniae of thallus rather narrow, about 4 mm. broad, pin- natifidly incised, either discrete or approximate, smooth above; hypothallus well developed and extending beyond the thallus. Apothecia deep fuscous. This species seems widely distributed and very variable in the tropics. Only a monographic study by one who has access to the types of all the proposed varieties as well as considerable experi- ence in the field in the principal floristic areas of the tropics can finally decide the validity of the proposed varieties and forms and provide adequate descriptions so that others may recognize them. Nylander, Miller Argau, Hue, Vainio, and Malme have each attempted more or less elaborate revisions but none seems wholly satisfactory. While probably artificial the separation of C. cronia for all the isidiose varieties of this species is relatively easy to apply, and has been followed here. Var. PARMELIOIDES (Hook. ap. Kunth) Müll. Arg., Flora 65: 320. 1882. Lecidea parmelioides Hook. ap. Kunth, Syn. Pl. Aequinoct. Orb. Nov. 1: 15. Type: Colombia, Cumana, Bordones, and Nueva Barcelona, Humboldt & Bonpland. [Vor. 20 460 ANNALS OF THE MISSOURI BOTANICAL GARDEN Thallus light olive gray, surface smooth, lobes cuneate, about mm. wide, sometimes once-cleft, closely appressed to the sub- strate, held by a few dark brown to black rhizinae composed of fascicles of hyphae, about 200 u thick; upper cortex about 25 u thick, the outer 5 y formed from the disintegration of a very thin layer of periclinal hyphae, the rest pseudoparenchymatous from the shifting of cells in a palisade layer, cells 10—12 y in diameter, rather thick-walled; algal layer a palisade of hyphae and filaments of Scytonema about 100-120 u thick; medulla about 40 y thick, of large septate periclinal hyphae 5-8 y in diameter; lower cortex of dark, periclinal conglutinate thick-walled hyphae 3-4 y in diameter. Apothecia immarginate, appressed and more or less adnate, 2 mm. or more in diameter, disc black; no amphithecium; parathecium of periclinal thick-walled hyphae, giving a pseudo- parenchymatous appearance, 100-120 y thick, thinning out to about 20 u at the edge; hypothecium not differentiated from the parathecium; thecium about 160 u tall; paraphyses 2-3 y in diam- eter, frequently septate, clavate tips black, about 4 u in diameter; asci cylindric to somewhat clavate, 8-spored, about 60 x 8-12 u; ascospores monostichous at first, gradually becoming distichous, ellipsoidal with acute ends, very thick-walled, unicellular, often with some deeply staining material in the middle with a slight thickening of the cell wall at that point, giving the appearance of a two-celled polarilocular spore with a wide isthmus, occasionally two other constrictions near the tip slightly suggesting a 4-celled condition, 11-12(-16) x 6-7 y. 'This variety seems widespread in the lower elevations although in absence of apothecia it is practically indistinguishable from var. pyrrhichocarpa. In Costa Rica var. parmelioides occurs in suitable habitats up to about 800 m., while var. pyrrhichocarpa occurs from 700 to 1340 m., but seen from Brasil up to 2000 m. in the state of São Paulo. Limón: Waldeck, 40 m., Dodge 7402. Guanacaste: H. Santamaría, 680-780 m., Dodge & Thomas 4709; C. San José de Libano, 500-960 m., Dodge, Hanckel & Thomas 6684; R. San José, 460—480 m., Dodge & Thomas 6581. Puntarenas: Osa, Puerto Jiménez, Brenes 825a; R. Sándalo, Dodge 7742. Var. PYRRHICHOCARPA Hue, Bull. Soc. Bot. France 48: lx. 1901 [1902]. 1933] DODGE—LICHENS OF COSTA RICA. I 461 Var. smaragdina Müll. Arg., Flora 65: 320. 1882, excl. syn. quoad specimina americana; Vainio, Etude Lich. Brésil 1: 210. 1890; Malme, Ark. f. Bot. 20°: 19. 1924. Type: Brasil, São Paulo, near São Paulo, Azevedo Sampaio. Thallus between yellowish glaucous and light mineral gray, lobing of thallus similar to var. parmelioides but much more irregular and lobes rather smaller, smooth above, 150-160 y thick, upper portion of cortex 4—6 y thick, of longitudinal hyphae soon disintegrating and disappearing, leaving a palisade 25-30 y. thick, which in turn becomes a somewhat irregular large-celled pseudoparenchyma; the algal layer 40-50 u thick, of loosely coiled and disintegrating filaments, perhaps Scytonema, but the cells rounding up and suggesting Nostoc, about 8 y in diameter, the medulla 50 y thick, of large septate periclinal hyphae, some- what less compact than in var. parmelioides, the lower cortex about 25 y thick, of black somewhat smaller conglutinate hyphae; rhizinae much more abundant, smaller conglutinate fascicles of hyphae similar to those of the lower cortex. Apothecia irregular, 3-3.5 mm. in diameter, carnelian red to vinaceous rufous, only slightly darkening, never dark brown or black, immarginate; amphithecium absent; parathecium adnate to the upper cortex, about 100 » thick, of large periclinal hyphae appearing almost pseudoparenchymatous at times; thecium 60 y. tall; paraphyses 2-3 y in diameter, filiform, apex not enlarged; asci clavate, about 12 y in diameter; ascospores 10-14 x 4-5 y. This variety is occasional at the medium elevations of the temperate zone in Costa Rica, ranging from 700 to 1400 m. It should not be confused with C. smaragdina Pers. or C. molybdaea Pers., both from the Old World tropics. Cartago: R. Birrís near Santiago, 920-1340 m., Dodge 4708, 8050; Cartago, C. Wercklé (Mus. Nac. 172186 p. min. p.). Alajuela: Santiago de S. Ramón, 1000 m., Brenes 238. Guanacaste: H. Santamaría, 680—780 m., Dodge & Thomas 6813. Var. srRIGOSA Müll. Arg., Flora 65: 326. 1882. Coccocarpia molybdaea v. cronia Nyl., Acta Soc. Sci. Fenn. 7:441. 1863, non Tuck. Type: Colombia, Cune, 1200 m., Lindig 2663 p. p. Thallus between pale olive buff and yellowish glaucous, lobes cuneate but narrower than in v. parmelioides, slightly incised, [Vor. 20 462 ANNALS OF THE MISSOURI BOTANICAL GARDEN smooth and shining above, thallus about 100 y. thick, outermost layer of hyphae about 4 y thick, soon disintegrating, upper cortex pseudoparenchymatous, 10-12 u thick, of about two layers of pseudoparenchymatous cells not part of a palisade, algal layer 32-40 y thick, of curved hyphae in a disintegrating palisade (intermediate between v. parmelioides and v. pyrrhichocarpa), medulla 25-30 y thick, of large periclinal hyphae, lower cortex 12-15 y thick, of large black periclinal hyphae, closely agglu- tinated rhizinae abundant but hyphae less fasciculate. Apo- thecia about 2 mm. broad, ochraceous tawny at first, becoming Prout’s brown or darker, immarginate; no amphithecium; parathecium about 100 y thick, of large periclinal hyphae, at the margins many hyphae projecting beyond the general level as stiff hairs; hypothecium not differentiated and not deeply staining; thecium about 40 » tall; paraphyses filiform with acu- minate tips which extend 1-2 y above the epithecial gel; asci clavate, 30x 8 y, with 8 spores; ascospores fusiform-ellipsoidal, 8 x 4 y, thick-walled, probably still immature. In the only collection from Costa Rica the thallus may not be quite mature, as no mature ascospores were found, hence the measurements may be a little small. Reported elsewhere only from Colombia. Cartago: R. Birrís, 920-1100 m., Dodge 8051. COCCOCARPIA CRONIA (Tuck.) Vainio, Ann. Acad. Sci. Fenn. A6': 103. 1915. Parmelia cronia 'Tuck., Proc. Amer. Acad. Arts & Sci. 1: 228. 1848; Syn. Lich. New England, 36. 1848. Coccocarpia molybdaea v. cronia Nyl., Syn. Meth. Lich. 2: 41. 1863. Coccocarpia pellita v. cronia Müll. Arg., Flora 65: 321. 1882. Coccocarpia parmelioides v. cronia Hue, Bull. Soc. Bot. France 48: lx. 1901 [1902]. Type: United States, Massachusetts, Lynn Hills, and West Cambridge, on mossy rocks, T'uckerman. Thallus mineral gray or darker, lobes rounded or short, broad- linear, more or less appressed, growing over mosses, about 130- 140 y thick; upper cortex 15-20 » thick of large thin-walled hyphae often appearing pseudoparenchymatous; algal zone 1933] i DODGE— LICHENS OF COSTA RICA. I 463 30-40 y. thick, with loosely woven hyphae in which are imbedded solid little groups 15-20 y in diameter of algal cells of closely coiled Scytonema; medulla 30-40 y thick, partly of close peri- clinal hyphae, partly more loosely woven, the lower cortex of two layers of black thick-walled hyphae giving rise to rhizinae, sterile; isidia terete, concolorous, or somewhat darkened at the tips. ‘Tuckerman noted on a specimen from Alabama, ‘‘spores 9-12 x 3-5 y, 2-celled?" Var. LIVIDORUFA (Mey. & Fw.) Zahlbr., Cat. Lich. Univ. 3: 287. 19265. Parmelia lividorufa Mey. & Fw., Nova Acta Acad. Leopold. Carolin. 19: Suppl. 222, pl. 4, f. 2. 1843. Coccocarpia molybdaea var. tenuior Nyl., ap. Krmplhbr., Flora 59:76. 1876 (nom. nud.). Coccocarpia pellita var. tenuior Müll. Arg., Flora 65: 321. 1882. 'Type: Brasil, Rio de Janeiro, Meyen (Bot. Mus. Berlin), type of var. tenuior; same locality, Glaziou 2026 (Univ. Genéve). 'Thallus mineral gray, lobes small, mostly about 1 mm. broad, quite irregular, smooth, about 100 y thick, structure as in the previous variety. Apothecia immarginate, irregular, convex, dise snuff brown to bister; amphithecium absent; parathecium 130—140 y thick; hypothesi not differentiated; theifiutu 80-85 u tall; paraphyses filiform, ending in the buitisecin] gel; asci 8- spored, clavate to cylindrical, 12 » in diameter, ascospores 11-12 x 5-6 y, ellipsoidal, thick-walled, distichous. In this variety the whole thallus approaches a microphylline squamose state. Evidently a lowland variety seen also from British Honduras and Nicaragua. Limón: R. see Sebi 70-200 m., Dodge, Catt & Thomas 8058. Alajuela: La Palma de S. Ramén, 1100 m., Brenes 43. Guanacaste: Tilarán, 650—690 m., Dodge & Thota 6557. Var. isidiophylla (Müll. Arg.) Dodge, comb. nov. Coccocarpia pellita v. wsidiophylla Müll. Arg., Flora 65: 321. 1882. Coccocarpia pellita v. cronia Vainio, Étude Lich. Brésil 1: 209. 1890, et auct. recentior., non Tuck. Coccocarpia cronia v. primaria Vainio, Ann. Acad. Sci. Fenn. A6':103. 1915. Type: Brasil, Rio de Janeiro, Glaziou 2025. [Vor. 20 464 ANNALS OF THE MISSOURI BOTANICAL GARDEN Thallus mineral gray or darker, laciniate, flabelliform or cuneate with rounded margin about 4 mm. broad, sometimes somewhat incised along the margin, upper surface isidiose with either simple or slightly branched isidia which are confined to the central portion of the thallus and to a lesser extent of the lobes; upper cortex of large thin-walled hyphae perpendicular to the margins of the lobes, up to about 20 u thick, closely septate; algal zone 60 y. thick of filaments of Scytonema about 12 y. thick with cells about 8 u in diameter, in the lower portion the algae less crowded, leaving some air spaces, medulla not differentiated from the algal zone; lower cortex 30 y thick, of conglutinate, black, thick-walled hyphae, giving rise to scattered rhizinae as well as penicillate fascicles of hyphae. Apothecia yellowish- rufous, becoming fuscous, somewhat lobulate, immarginate, adnate; parathecium 55-60 y in the center, thinning out at the margin, pseudoparenchymatous, of large somewhat periclinal hyphae; hypothecium not differentiated, about 60 y thick; paraphyses 2-3 y in diameter, filiform, with swollen tips in the very dark epithecial gel; asci clavate, 8-spored, 8-10 y in diam- eter; ascospores distichous, fusiform, immature. This species is close to C. pellita and perhaps should be con- sidered a variety of that species. It has approximately the same distribution, reaching from sea level to about 1700 m. in Costa Rica, 2700 m. in Pert. Cartago: Aguacaliente, 1240-1460 m., Dodge & Thomas 7087. Heredia: C. Central de Zurquí, 1600-1700 m., Dodge, J. Valerio & Thomas 6087; C. de las Caricias, 2000-2400 m., Standley & J. Valerio 55282. Alajuela: La Palma de S. Ramón, 1100 m., Brenes (fertile). Puntarenas: Osa, between R. Sándalo and R. Tigre, 1-2 m., Dodge 8057. Var. prolificans (Malme) Dodge, n. comb. Coccocarpia pellita var. prolificans Malme, Ark. f. Bot. 20°: 19. 1924. Type: Brasil, Rio de Janeiro, Regnell 65. Thallus mineral gray or darker, lobes flabelliform, somewhat zonate, 5-10 mm. wide, margin rounded, more or less lobulate along the incised portions between lobes, upper surface lobulate and more or less microphylline, lobules concolorous, about 70 u thick, upper cortex about 16 » thick, of large thin-walled longi- tudinal hyphae; algal zone about 30 y thick, of a loose palisade 1933] DODGE—LICHENS OF COSTA RICA. I 465 of Scytonema filaments, medulla about 20 u thick, of large thin- walled periclinal hyphae with some air spaces; lower cortex about 15 y thick, of thick-walled black hyphae, giving off fascicles of rhizinal hyphae. Sterile. This variety is rather common at elevations from 700 to 1400 m. in Costa Rica. Cartago: R. Birrís, 1220-1340 m., Dodge & Thomas 4710. Alajuela: La Palma de S. Ramón, 1250 m., Brenes 38a; Piedades de S. Ramón, 900 m., Brenes 385. Guanacaste: H. Santamaría, 680-780 m., Dodge & Thomas 6987, 6999. Var. granulosa (Müll. Arg.) Dodge, comb. nov. Coccocarpia pellita var. granulosa Müll. Arg., Flora 65: 322. 1882. Type: Brasil, São Paulo, Apiahy, Puiggari 244, and Colombia, Bogotá, 2700 m., Lindig 2538. Thallus thicker, greenish-glaucous-blue or darker, lobes up to 3 mm. broad, the larger lobes more or less incised, obovoid to almost linear, soon blackish granular to almost isidiose in the center, about 140 y thick, upper cortex a single layer of longi- tudinal hyphae 4-5 y in diameter, algal zone of Scytonema filaments 40-50 u thick, forming a palisade interspersed with occasional rows of fungal cells, medulla of large, septate, peri- clinal hyphae about 60 y thick, the lower cortex 6-8 y thick, of thick-walled dark hyphae, not well developed. Apothecia less adnate and somewhat isidiose; no amphithecium; para- thecium 150 y thick, pseudoparenchymatous, some of the outer layer of cells growing downward to form a ciliate to tomentose surface below; hypothecium not differentiated, thecium about 60 y. tall; paraphyses filiform, tips not enlarged, ending in a dark brown epithecial gel; asci clavate, rather immature. This variety as here defined includes some of Malme's var. isidiosa and some material may have been distributed under that name. In Costa Rica this variety is widespread in the temperate regions from 100 m. to 1500 m. Limón: R. Siquirres, 70-200 m., Dodge, Catt & Thomas 559. Cartago: Santiago, 1140-1180 m., Dodge 4638, 8054; C. istud 1320-1500 m., Dodge & Thomas 4763; Cartago, C. Wercklé (Mus. Nac. 17202). San José: Turrdcares, 540-600 m., Dodge & Thomas 8056; R. Virilla below El Brazil, Dodge 7777. Alajuela: C. Pata de Gallo a S. Rafael de S. Ramón, 1200-1250 m., Brenes 215; S. Pedro de S. Ramón, 700 m., Brenes 459. [Vor. 20 466 ANNALS OF THE MISSOURI BOTANICAL GARDEN Guanacaste: Liberia, 100 m., Dodge & Thomas 6582; H. Santamaría, 640—780 m., Dodge & Thomas 6900, 7007, 8055; Tilarán, 650-690 m., Dodge & Thomas 6556, 6559, Standley & J. Valerio 44429, 44520; H. Granadilla, between R. Las Cañas and R. S. José, Dodge & Thomas 6719. CoccocaAnPrA albida Dodge, sp. nov. Type: Costa Rica, Limón, Hamburg, Dodge & Nevermann 7401. Thallus albidus, isidiosus, lobis rotundatis, flabelliformibus, 10 mm. latis, zonatus, inferne albus rhizinis albidis, apicibus rhizinarum nigricantibus, thallus 120-130 y crassitudine, cortex 15 u, hyphis longitudinalibus septatis, 5-6 uœ diametro; zona scytonematica ad 40 y crassitudine, filamentis laxis implexis, medulla 40-60 y crassitudine hyphis periclinalibus compactis; cortex inferior hyphis hyalinis dense compactis; rhizinis albidis fasciculatis penicillatis. Thallus mineral gray or lighter, densely isidiose in the center, lobes rounded, flabelliform as in C. pellita v. parmelioides, 10 mm. broad, more or less zonate, below white with white rhizinae, the latter sometimes tipped brownish black; thallus 120-130 y thick, cortex 15 u thick, of longitudinal septate hyphae 5-6 y in di- ameter; algal zone 40 y thick, of loosely tangled filaments of Scyionema, medulla 40-60 u thick, of compact periclinal hyphae; the lower cortex of hyaline hyphae more or less agglutinated; rhizinae white, fasciculate, and penicillate. The lower surface resembles that of Sticta sp. quite closely, but cyphellae are absent and the structure is typical of Cocco- carpia. Perhaps it is an albino variety of C. cronia var. isidio- phylla. It has much broader lobes than C. dominguensis. Limén: Hamburg, 20-30 m., Dodge & Nevermann 7401. CoccocaRPIA ELEGANS Müll. Arg., Flora 64: 507. 1881. Type: Brasil, Sáo Paulo, Apiahy, Puiggari. No hypothallus, thallus terre-verte, small, linear, dichotomous (rarely trichotomous at the ultimate branches), larger laciniae 0.1-0.2 mm. broad, ultimate branches much narrower, dull and smooth above, white with white rhizinae below, thallus 80-85 y thick, upper cortex 8-12 y thick, decomposed, apparently of slender longitudinal hyphae; algal layer about 40 y thick, ir- regular, of loosely woven fungus hyphae supporting short irregular chains of angular to subspherical cells 8-12 y in diameter, medulla not well differentiated from the algal layer but of very loosely 1933] DODGE—LICHENS OF COSTA RICA. I 467 woven slender hyphae; lower cortex 10-15 u thick of closely woven hyaline, longitudinal hyphae, the rhizinae 40-50 y in diameter, large fascicles of hyaline hyphae. Apothecia immar- ginate, 0.1-0.5 mm. in diameter, chestnut, with long fascicles of hyphae forming stiff bristles about the margin, with stipe about 200 u in diameter and 40-50 y tall; no amphithecium; para- thecium 60 y thick, of thick-walled hyphae forming pseudo- parenchyma; thecium 60 y tall; paraphyses slender, filiform, not enlarged at the tips; asci clavate, 4—6 u in diameter, immature. The above description is based upon material from Brasil, Rio de Janeiro, Glaziou 18071, determined by Müller Argau, who also reports it from Puntarenas, Boruca, 560 m., Tonduz 5462. As this is a small epiphyllous species, a large thallus being only 1 cm. in diameter, it is quite possible that it has been overlooked as my epiphyllous material has not yet been examined carefully. A NEUTRAL (?) STRAIN OF MUCOR SPHAEROSPORUS FROM MISSOURI MORRIS MOORE Formerly Rufus J. Lackland Research Fellow in the Henry Shaw School of Botany of Washington University In the caves of the Ranken estate in Missouri, where the atmosphere is moist and the temperature approximately 16° C., was found a segment of a twig matted with a growth of mycelial threads, white in a mass, and hyaline when single. The appearance of the latter was that of many fine root hairs. When transferred to agar substrates, numerous sporangia, sporangioles, oidia, and chlamydospores developed, with no evidence of zygo- spores. The characteristics and description were similar to those given by Lendner (’08) for Mucor sphaerosporus Hagem 1908. This fungus has been reported chiefly from Europe, and in this country by Waksman (’16) and Gilman and Abbott (’27), as isolations from soil. As far as the author could determine, this is the first report of the above organism from Missouri. A culture sent to Dr. A. F. Blakeslee at the Carnegie Institu- tion of Washington, Department of Genetics, was tested with strains of presumably the same species, one being a minus isolated by Waksman in 1915, and another a plus strain probably from the Centralstelle, of unknown origin. No zygospores were produced. An attempt was made to obtain imperfect hybridiza- tion by treating this strain with some of the strong testers in the above laboratory, but no sexual reaction resulted. Growth on various media produced changes in the amount of mycelium, size and number of sporangia, sporangioles, hyphae, oidia and chlamydospores, and in chromogenesis. Thanks are due Dr. Blakeslee and Miss Satina of the Carnegie Institution of Washington for their interest and assistance. BIBLIOGRAPHY Gilman, J. C., and E. V. Abbott eile A summary of the soil fungi. Iowa State Coll. Jour. Sci. 1: 225-343. — A. ('08). Les Mucorinées fi la Suisse. Matériaux pour la flore crypto- mique Suisse 3!: 1-177. pl. 1-3. 8. Walsman. S. A. (16). Soil fungi and their activities. Soil Sci. 2: 103-156. 1916. Ann. Mo. Bor. Garp., Vor. 20, 1933 (469) A STUDY OF ENDOMYCES CAPSULATUS REWBRIDGE, DODGE AND AYERS: A CAUSATIVE AGENT OF FATAL CEREBROSPINAL MENINGITIS! MORRIS MOORE Formerly Rufus J. Lackland Research Fellow in the Henry Shaw School of Botany of Washington University INTRODUCTION The probable taxonomic relationships of the fungi known as yeast-like organisms have for many years attracted much atten- tion and discussion among mycologists, and to a certain extent among medical men. Considerations of their general physio- logical properties, modes of reproduction, their cytological differentiations, and varied pathogenic abilities were involved. The group comprises several families and many genera and species, with subdivisions of these, which from time to time call forth long and expounding dissertations on the mechanism of reproduction (the perfect stage). The problem of classification is unfortunately a very complex one, since it has been found, at least in the present work and with other closely related fungi which will be considered later, that several forms of development may be present in the same culture. Such phenomena are rather rare but render incorrect any means of classification based on one phase only. It is well known that the fungi are divided, on the basis of their morphology, cytology, and sexual development, into three major divisions: Phycomycetes, Ascomycetes, and Basidio- mycetes, and to these is appended a fourth group, the Fungi Imperfecti, the life history of which is not at all or incompletely known. The yeast-like organisms are considered to constitute a branch of both the Ascomycetes and the Fungi Imperfecti, the former where the production of asci is found, and the latter where no sexual development has been determined. More specifically, the perfect forms are considered to be members of the Endomycetaceae (Gaiimann and Dodge, ’28; Rewbridge, 1 An investigation carried out at the Missouri Botanical Garden in the Graduate Laboratory of the Henry Shaw School of Botany of Washington University, and submitted as a thesis in partial fulfillment of the requirements for the degree of doc- tor of philosophy in the Henry Shaw School of Botany of Washington University. Ann. Mo. Bor. GARD., Vor. 20, 1933 (471) [Vor. 20 472 ANNALS OF THE MISSOURI BOTANICAL GARDEN Dodge, and Ayers, ’29; Moore, '33), and Coccidioideaceae (Moore, '32), including such genera as Endomyces, Coccidioides, Rhinosporidium, and Pseudococcidioides, while the imperfect forms, as Monilia, Cryptococcus, Oidium, M ycoderma, and several others have features which are somewhat similar to the above groups. ‘This of course does not include the true yeasts which are included under the Saccharomycetaceae. It was because of the imperfect knowledge existing in the field that the author undertook the study of the cytology of Endomyces capsulatus, which had been isolated from a case of meningo- encephalitis. During the course of the work it was found desirable to learn more about the phenomena which might be associated with the pathogen, and hence attention was paid to its physiological and other properties. In addition, the occurrence of other infections by closely allied organisms, E. capsulatus var. isabellinus (Moore, '33), and FE. dermatitidis (Moore, 733a), have yielded the fungi which have been used in à comparative study, so that what may pertain to E. capsulatus has been found to hold true for the latter, with some minor exceptions as chro- mogenesis and cultural characteristics. HisronicAL REVIEW Occasionally debatable points in the field of medical mycology are cleared up by the isolation of some fungus. Such an organism, E. capsulatus, was cultured in 1928 from a case of meningitis. The patient was a white, married male, 48 years of age, a furrier by trade. He entered the Boston City Hospital in August, 1928, complaining of having had continuous headaches in the frontal and temporal lobes for a period of three and a half months. His illness began with a hard, tender, slightly movable mass on the medial surface of the lower third of the left thigh, about five months before entry. This mass subsided within four weeks, but a similar mass then appeared beneath the anterior surface of the neck. He developed defects in memory, losing himself on familiar streets, and could not keep track of time. He lost interest and was unable to understand questions or conversation. The cyst which had been present at the level of the thyroid isthmus between the trachea and anterior border of the sterno- 1933] MOORE—ENDOMYCES CAPSULATUS 473 cleidomastoid muscle was aspirated. Thirty cc. creamy, tena- cious, yellow pus was obtained, from which E. capsulatus was cultured. Similar organisms were obtained from the pus aspi- rated from the nodule of the left thigh. About a month later, the patient’s condition became worse. His temperature increased rapidly, as did his pulse and respi- ration, and he finally died. The clinical diagnosis was mycotic meningo-encephalitis (?), and an X-ray diagnosis showed bi- lateral pulmonary tuberculosis. It has been the author’s experience that an infection of this sort which involves the lungs is often mistaken for tuberculosis, and the diagnosis made here may have been incorrect. The organism from this case has been used in these studies for the greater part, having been kept as a stock culture by alternate transfers on nutrient and Sabouraud’s agar. The second case due to the variety isabellinus (pl. 21) occurred in St. Louis and was reported by MacBryde and Thompson. The patient, a white male, 28 years of age, had been a plumber at the time of the appearance of the skin lesions. He was first admitted to the Barnard Free Skin and Cancer Hospital in February, 1930, with a palm-sized annular ulceration on the left forearm just above the wrist, which involved the dorsal surface chiefly and showed clinical signs of blastomycetic dermatitis. Endomycetes were isolated from the pus of this lesion. About 3 cm. to the right of the anus was another lesion about 3 em. in diameter and involving the anal margin. On the upper part of the right arm was a scar of a lesion which had healed spontane- ously. All active lesions were pruriginous and had a foul smell. The patient later, in January, complained of coryza, with a mild cough, pain in the right temporal region, accompanied by a rise in temperature. His headache became universal and intense, posterior to the eyes and extending to the occiput and base of the neck. There was a marked bilateral Kernig reflex action. He had nauseation and vomiting spells. A lumbar puncture gave a cloudy fluid under slight pressure, in which were seen the organ- isms found in greater abundance with later punctures. At this time he received a large dose of iodine solution which cleared the spinal fluid and decreased the number of budding cells, improving [Vor. 20 474 ANNALS OF THE MISSOURI BOTANICAL GARDEN his health greatly. However, two weeks later he had a relapse, became comatose, and three days before death developed three small subcutaneous abscesses. One was on the right wrist, one on the sacrum, and one on the left side of the chest, and all contained pus and E. capsulatus on culture. The patient died forty-six days after the onset of the meningitis symptoms and about three years after the first appearance of the skin lesions. The morphology and cytology of the organism of this case was found to be the same as the first fungus, having a very light cinna- mon color as contrasted with the hyaline or white, in a mass, of E. capsulatus, and cultural characteristics similar in detail. The third organism studied is the one which has gone under many generic names, as Blastomyces, Oidium, Saccharomyces, Cryptococcus, and Mycoderma, because its life history was in- completely known and the sexual act was not definitely estab- lished, or the presence of the ascus and the actual number of spores determined. Historically this fungus is important be- cause it causes a disease known as blastomycosis which has become fairly widespread, both in this country and Europe. The organism studied by the author was isolated from a case which occurred at the Barnard Free Skin and Cancer Hospital of St. Louis and has been dealt with in detail in à previous paper (Moore, 33). Suffice it to say here that E. dermatitidis, the organism causing a clinical condition known as blastomycosis, is similar to the above organisms, differing in its chromogenesis, being a dark cinnamon to brown in culture and having a some- what different cultural reaction, to be mentioned later. In addition to the above three forms, an organism obtained from a case of the so-called European blastomycosis or general- ized torulosis reported by Urbach and Zach (730) (pl. 19, figs. 13-25) has been studied to a certain extent. The case in brief was that of a 27-year-old shoe-worker who had never left Europe. He entered the University Clinie for Syphilology and Dermatol- ogy at Vienna with a swelling of the gums of the lower jaw which was excised, and a half year later he returned with a similar swelling of the gums of the upper jaw. About one year later the patient noticed an abscess-like swelling on the left side of the abdominal region, two months later a similar disorder on 1933] MOORE—ENDOMYCES CAPSULATUS 475 his left thigh, and eight weeks later the same thing was found on his neck. The lesion on his left thigh was aspirated and from it 200 cc. purulent substance was obtained. A short time after that an inflammation of the lungs developed which healed spontaneously, but left him with a cough and vomiting spells. Yeast-like cells were cultured from the purulent secretion of the left thigh. The disease was clinically identical with syphilis, tuberculosis, leukemia, and several others. Inoculation of an extract of the yeast-like cells (blastomycin) brought forth strong local, focal, and generalized reactions, giving evidence of the presence of blastomycosis. The lesions were somewhat superficially healed by treatment. The patient later developed high temperatures in the evening, oral pains, deafness, violent coughing at night, and ruby-red expectorations. A tumor of the nasal septum, very suggestive of rhinosporidiosis, had developed and a catarrh of the Eusta- chian tube as a result of the tumor. There was an ulceration of the right tonsil and other lesions close by, and finally a heavy exudate in his lungs. Treatment gave him relief temporarily, but he returned to the clinic a short time later with a continuous headache, high temperature, and rapid spread of the mouth lesions. Paralysis of the optic musculature set in, with un- consciousness, and death ensued. The organism from this case has been cultured and subcultured and as yet has shown no final stage as is present in the three above forms. However, some organisms require a long time before the perfect stage is obtained. Furthermore, the histo- logical and pathological condition of the patient as investigated by Chiari (30), shows the identical tissue reaction present in cases of E. dermatitidis. With this brief review of the history of the organisms in- vestigated, let us now turn to a study of E. capsulatus. TECHNIQUE E. capsulatus was kept growing by subculturing on nutrient and Sabouraud’s agar as a stock culture. In the work done here, the organism was grown on nutrient agar, a product of the [Vor. 20 476 ANNALS OF THE MISSOURI BOTANICAL GARDEN Digestive Ferments Co., at pH 6.8. In cases where an abundant growth was necessary, with a thick mycelium, Sabouraud's agar, pH 5.6, was employed. For studying the cytology with regard to the nuclear changes in the sexual act, glycerine agar (beef extract agar plus 6 per cent glycerine, pH 7.1) was used. This medium gave the greatest number of asci in culture, in addition to a thick growth and many important diagnostic features, as chlamydospores, conidia, racquet mycelium, etc. Cultures on nutrient agar and Sabouraud’s agar were also used, but the cells were not so good for details as on the above medium. The cultures were fixed with a number of agents. Flemming’s stronger solution shrunk the material rather noticeably, but the weaker did not give very good results either. Bouin’s picro-formalin solution, which has been advocated by Kater (27) and other cytologists working with yeasts, was of no value in this work. Benda’s fluid, which is a modification of Flem- ming's stronger solution, caused the fungus to have too great an affinity for the stains, and inasmuch as it is difficult to destain the organism without causing some sort of damage, it had to be used with caution. Hermann’s fluid, which is probably the most expensive of the fixing agents, was also used. This is a variation of Flemming’s fluid, the chromic acid being replaced by platinie chloride, and although not particularly recommended by Chamberlain (732), the author has found it to give the best results. Various other fixing agents were employed, but the results were not worthy of note. The embedding proved to be a problem because the material had to be fixed and embedded while on the agar. Paraffin was used at first, but in the glycerine agar cultures it would not ad- here to the agar substrate and sectioning could not be done without damage to the material. Gradual and repeated changes of paraffin did not remedy the condition. A medium was then tried which contained very little protein and carbohydrate, but no suitable growth could be obtained. Finally the cultures were embedded in Dupont parlodion, using the Jeffrey technique. This method has been recently outlined by Wetmore (32) and with some modification was applied here. It consists in fixing 1933] MOORE—ENDOMYCES CAPSULATUS 477 the material, in this case either with Hermann’s or Benda’s fluid, pumping and dehydrating, then passing it through an intermediate stage of ether-alcohol, and then through a series of concentrations of the celloidin or parlodion. The first concentra- tion was a 2 per cent solution and each series increased 2 per cent until the final concentration was 12 per cent. The material on the agar was taken from the test-tube after fixing, cut into convenient pieces, and put through the above procedure. The pieces were kept in a tightly plugged bottle in an incubator at 45° C. and changes were made daily. After embedding, sections were cut to a thickness of 10 u. These were then stained and mounted. The best stain for this procedure was found to be iron-alum haematoxylin, using Heidenhain’s haematoxylin, although Ehr- lich’s gave just as good results. A combination of Benda’s fixing agent and iron-alum haematoxylin showed the reticulated network and the metachromatic material very well, whereas Hermann’s fluid plus Heidenhain’s iron-alum haematoxylin was best for nuclear structure. Methylene blue and Hermann’s fluid brought out the vacuoles and volutin and the metachro- matic material very clearly, as seen in pl. 23, figs. 15-18. For morphological work, hanging-drop cultures were made of 2 per cent proteose peptone and 2 per cent bacto-peptone, as well as lactose broth cultures. Material was also placed in a drop of a 1 per cent solution of crystal violet (aqueous) in glyc- erine, the dye being added to the desired intensity. The prep- aration was allowed to stand from 15 to 30 minutes for a sufficient clearing. Aman’s lactophenol was also applied, as was carbol fuchsin and very dilute solutions of methylene blue, eosine, and crystal violet. In addition, iodine potassium iodide (saturated solution) was used for studying the glycogen contents of the cells and chondriosomes, as advocated by Guilliermond. Osmic acid, platinic chloride, and iodine green were also used, as well as neutral red, but these materials and other methods of pro- cedure will be explained later in the text. CYTOLOGY The cytology of the lower Ascomycetes, particularly the yeasts [Vor. 20 478 ANNALS OF THE MISSOURI BOTANICAL GARDEN and the yeast-like organisms, had been for a number of years a topic of discussion among the older cytologists. The question as to whether or not a nucleus was present was indeed a serious one, judging by the numerous and lengthy dissertations on the subject. As is customary, there were two sides to the argument. One group maintained that the cells, of the yeasts in this case, were made up of a mass of protoplasm and of nuclein, without a true nucleus, and that the nuclein is differentiated at times in the cytoplasm in the form of granules which assume a definite color on staining. The other group favored the presence of a definite nucleus. The first reference to a nucleus in yeasts was perhaps made in 1844 by Nägeli, who decided that “ʻa little nucleus of whitish mucus, lying on the membrane, regularly in each cell” (Wager, ’98), was often found in the yeast cell. Whether Nägeli actually saw a nucleus in the fresh condition is rather doubtful, for an oil globule has often been mistaken for the nucleus even to-day, with our more advanced knowledge of cytology. About five years later, in 1849, Schleiden, by treating cells with ether-alcohol or potash, was able to find a rounded structure with a clear cell wall which contained delicate granules either singly or in groups, and in addition, a large flat body, which he called a cytoblast. Following this work, Brücke, in 1861, asserted that in living material, as well as dead cells treated with iodine and acetic acid, no definite nucleus could be seen and he reprimanded the former workers by saying that no one was justified in taking bodies of various sizes and numbers, such as often occur, for nuclei. In 1879 Schmitz, using haematoxylin, was able to demonstrate a nucleus in the cytoplasm of the cell near the vacuole, while Strasburger, repeating this work in 1884 and 1889, using picric acid and haematoxylin, was able to confirm these observations, finding that the nucleus was not demonstrable in unstained material. Krasser, in 1885, contradicted these findings by saying that granules existed in the yeast-cell, but no nucleus. His main argument was that there was no specific staining reaction for nuclei, and furthermore that the absence of a definite nucleus 1933] MOORE—ENDOMYCES CAPSULATUS 479 in the yeast-cell was supported by the rapid growth of the organism. For this reason he believed that there was nuclein in the cell which was distributed through the protoplasm very much as is generally held to be the case with bacteria to-day. Other workers who held the view that a nucleus, or at least a “corpuscle” as some called it, was present, the biochemical properties and physiological functions of which were analogous to those found in the cells of plants and of animals, were Hansen, Strasburger, Zacharias, Moeller, Buscalioni, Henneguy, Dan- geard (93), Janssens (02), and Janssens and Leblanc ('98). In fact, the latter writers made the following statement, ''La cellule de la levure peut étre considérée comme formé sur la type général de la cellule. On y trouve en effet, un noyau, un proto- plasme et une membrane." To these workers the nucleus had a vacuolar appearance with a very differentiated structure, to others it was a homogeneous body. Wager (’98), in a detailed bit of work, described it as being a vacuole filled with chromatin granulations (perhaps somewhat as is shown in pl. 23, figs. 17-18) which had been taken for a nucleus by Janssens and Leblanc, and furthermore that a spherical and homogeneous body, con- sidered by some as the nucleus, was always adjacent to the vacuole and could be compared with a nucleolus. He considered the vacuole plus the eccentric nucleolus as a primitive stage in the phylogeny of the development of the nucleus. This belief was also based on the observation that the small body and the vacuole divide simultaneously by budding. Guilliermond, in 1902, demonstrated that a definite nucleus was present, and that the vacuole was independent of the nucleus, being filled with granulation products which at the present time he holds to be nuclear decomposition substances (nucleic acid derivatives with some unknown base). For the granules he retained the name ‘‘corpuscules métachromatiques," a term which had been applied previously and which is used even now. Those who denied the presence of a nucleus but admitted the presence of a nuclein substance as described above were Raum, Roncali, Hieronymus, Macallum, and others. In accordance with the views of the older writers, the author has found that in Endomyces as treated here, when first isolated [Vor. 20 480 ANNALS OF THE MISSOURI BOTANICAL GARDEN from the tissue of the host, no nucleus exists, but simply a distribution of the chromatin material or nuclein or metachro- matic corpuscles throughout the cell, as may be seen in pl. 19, figs. 13-14, pl. 20B, pl. 21, figs. 1-3, and several others. This feature is a quite constant character of freshly isolated cells. However, when kept on agar or an artificial substrate, the cells not only change their form, passing through what may be termed the secondary stage, that is changing from a yeast-cell to a hyphal form, but also undergo a change in nucleoplasmic make-up. It has been found that even though the yeast-like cells retain the same morphological characteristics on agar media, their chro- matin material or nuclein is converted into a definite nucleus. This was clearly shown to be the case in E. dermatitidis (Moore, '83). What the mechanism involved here might be, it is difficult to say, but we may conjecture that it is linked up with the adap- tation to a changed environment and a different mode of devel- opment to which it must become accustomed, as shall be pointed out later. With such views being held in the past, it was of course to be expected that the presence of any phenomena which might in- volve the nucleus, as mitosis, would also be in dispute. In yeasts the question of nuclear division has received considerable attention, particularly by Guilliermond, Dangeard, and several others, the former writing many and long papers on the subject, and it has been said by some present-day workers, and the fol- lowing is a direct quotation from the papers of one of them, that ''the ideas of Guilliermond . . . which gain weight by the mere bulk of his work on yeast, seem to meet with more favor." The problem of the division of the nucleus in yeasts is perhaps more complex than would seem offhand. "The same writers who held that a nueleus was present in the cell first described nuclear phenomena. Wager (’98) interpreted an amitosis in the yeast with perhaps evidence of chromosomes. Janssens and Leblanc (98), working with Saccharomyces cerevisiae and S. Ludwigii, as well as Schizosaccharomyces octosporus and S. Pombe, described a form of mitosis which, in the light of other works, seemed very suspicious and they made the following statement, ‘‘On peut dire que tout ce qui diminue la vitalité d'une cellule tend à réduire la complication des phenoménes de division." 1933] MOORE—ENDOMYCES CAPSULATUS 481 Dangeard (093) found that the nucleus elongated and the nucleolus divided in two by an elongation to a thread-like process within the nuclear wall. Swellengrebel (’05) and later Fuhrmann (06) ascribed a definite mitosis to the nucleus, the latter studying Saccharomyces ellipsoideus I. Hansen and presenting a typical karyokinetie sequence with the formation of four chromosomes. There were several authors who held to this view, and at present there are those who believe that definite mitotic phenomena are present. Kater (27), using a smear technique with S. cerviciae [sic], demonstrated a mitosis, with the formation of about eight chromosomes and a definite spindle, instead of divi- sion by constriction. It should be pointed out here that smear methods as applied to yeasts and particularly as carried out by Kater have never yielded any results which might be considered reliable. The act of smearing, no matter how good or how careful the application of the fixative might be, usually allows for some action on the cell wall that obliterates the correct phenomena and substitutes artifacts, so that vacuolar constit- uents or secretion products of the cytoplasm have been mis- interpreted. On the other hand, Guilliermond (17), in a summary of his work, found amitosis to occur where budding was present. This process was characterized by an elongation of the nucleus which quickly divided by the resorption of the thread-like portion that separated the two segments. Guilliermond also observed that it was impossible to see the nuclear phenomenon clearly due to the abundant products of secretion which covered the nucleus. The observations of Kohl, of Wager and Peniston (710), and of Pénau confirmed Guilliermond's results, and even went so far as to actually claim definite amitosis. Whether or not mitosis or amitosis actually occurs is still a matter of dispute, despite the evidence that either side may advance. "There are several factors which must be considered in studying the process, or processes, in nuclear division. First, does the nucleolus have a single morphological characterization, such as exists in the higher forms, or does it constitute the total nucleoplasmie material, as some call it, which has an affinity for iron haematoxylin, for example, as in pl. 22, fig. 18? Second, [Vor. 20 482 ANNALS OF THE MISSOURI BOTANICAL GARDEN can the nucleus be seen so clearly that there is no mistaking the change that may take place during the actual process? The first question is difficult to answer because it has been found by several workers studying the chemistry of the cell, yeast or yeast-like or even that of the higher Ascomycetes, that certain materials take the same dye. Then, the linin network with the chromatin material or the chromomeres, as some call them, usually cause confusion. The second question is really serious, because we know that our most precise methods of technique are still too crude for the minute nuclear make-up of these lower Ascomycetes. Cell substances, or secretion products as Guilliermond calls them, usually take stains which are peculiar to the nucleus and mask any clear-cut pictures of nuclear phe- nomena, and nucleic acid substances, derivatives of excretion products of the nucleus, probably nucleophosphates and the like, are also substances that must be considered in detail. Besides, the size of the reactors in the process is too small for any precise determinations with the means at hand. On the other hand, according to the theories of modern genetics there must be some mechanism whereby chromatic substance or material is distributed through the agency of mitosis, whereas the act of amitosis is simply a means of increasing the nuclear sur- face or spreading the nuclear material through a cell and is comparable to the fragmentation or lobulation of nuclei. The latter is not at all reproductive in this respect. However, there are many acts taking place in the daily life of a fungus, particularly in these lower groups, which are impossible to explain with our present knowledge of cytology or genetics, and theories have been built up only to be torn down and built up again. This does not mean that genetics is not standing on firm ground, but that much more must be learned about mycological phenom- ena before any generalizations can be made. It is difficult to understand that the nucleus in the hyphae of yeast-like organisms, E. capsulatus or E. dermatitidis, divides by direct division, at least as far as can be made out from the nuclear appearance. The process is evidently very rapid, for an examination of a great many slides failed to show any condition other than that shown in pl. 22, figs. 3-5. In the 1933] MOORE—ENDOMYCES CAPSULATUS 483 ascogenous hyphae, or perhaps the antheridium and ascogonium (pl. 22, figs. 6-9), the division is somewhat slower, and here an elongation of the nucleus and nucleolus may be seen with a deeper staining central portion indefinite as to character but perhaps analogous to chromosomes. In a discussion of nuclear phenomena, the next thing to be considered is the mode of development or reproduction. Guillier- mond (705, '05a, '05b, "08, "09, '09b, 710, "10a, "10b, TL '11a, 712, '13,717,719,'20), Dangeard (793, '94, '94a, '94b, '97), Hansen (704), and several others, in a long series of investigations, have de- scribed the sexuality of the lower yeasts and some of the yeast- like fungi. It is known, of course, that in the Ascomycetes, particularly the lower forms, there may be three forms of devel- opment, heterogamy, isogamy, and a reduction to partheno- genesis. In the first case, two gametes, usually of unequal size, a small one generally representing the antheridial cell and a larger one the ascogonium (this is not a hard and fast rule), send out one, or sometimes several, small tubes which copulate and fuse. The cell contents of the antheridium then pass into that of the ascogonium. The two nuclei, one from each gamete, fuse, and finally an ascus develops through the subsequent division of the fusion nucleus, the presumptive mother spore. In the case of isogamy in the yeasts, two morphologically alike gametes fuse in like manner. Finally in the series, there is a reduction to a condition in which a cell may suddenly produce spores without copulation. This condition has been found by Mangenot (’19) for E. Lindneri and by Guilliermond for several other species, as E. fibuliger and Zygosaccharomyces Pastori. It occurs also when an ascogonial cell sending out a tube or tubes fails to copulate and hence produces spores parthenogenetically. In some species the above phenomena may be found separately, or two processes or even all three may be present in the same culture. This latter condition has been found to be true for E. capsulatus. In addition to the parthenogenetic formation of asci and asco- spores, there is à non-sexual reproductive structure which has not been given too much attention in the past. This is the conidium (pl. 23, fig. 10), which is à non-nucleated structure [Vor. 20 484 ANNALS OF THE MISSOURI BOTANICAL GARDEN filled with chromatin granules or the metachromatic corpuscles emphasized by Guilliermond. It occurs usually near a septum of the hypha, and measures approximately 5 y in diameter. Struetures of this sort may and usually do spring up from the hyphae when there is a lack of nutrient material, or when the hydrogen-ion concentration is fairly high (pl. 19, figs. 4, 7-8, 11-12). These may be pyriform (pl. 18, fig. 10) or round (pl. 18, fig. 7), sessile or on a short pedicel, with a thick wall (pl. 23, fig. 10) ora thin wall. There is a heavy reticulated network with the granules mentioned above usually occurring at the nodes of the threads. These particular structures break off easily and may serve as resting cells or chlamydospores which, when placed in favorable media, germinate, form nuclei by an accumulation perhaps of the chromatin material, and develop a normal growth. They represent probably a degeneration from an ascus, or, on the other hand, an advanced character of reproduction. To get a clearer understanding of the life history of E. cap- sulatus, it would be advisable to begin with the organism as it is found in the parasitized host and explain the mechanism by which it and related fungi propagate themselves. In the tissue the organism grows as a yeast 6-8 y. in diameter, and reproduces itself by budding. There is no nucleus present, at least it could not be demonstrated in tissue sections or in freshly isolated cells, but instead there is a distribution of nuclein material throughout the organism, much the same condition as exists in bacteria, Myxophyceae, or in the conidium shown here. When isolated and grown on an artificial substrate as on agar, the chromatin or nuclein material seems to become larger, and there is usually formed a nucleus and many large granules which may be presumptive nuclei. This latter condition is based on circumstantial evidence (pl. 21, figs. 4-9; pl. 20B). The cells then pass through this stage, the yeast-cells, to stage two which consists of large irregular cells, attaining a condition as seen in pl. 19, figs. 14-15, 20-25, and have been clearly shown for E. dermatitidis. At this stage the chromatin material is spread throughout the cell, but the reticulated network seen in the later stages is not very clearly established. However, it seems to be rapidly developing and nuclei are clearly distinct. The stage 1933] MOORE—ENDOMYCES CAPSULATUS 485 following is probably the most complex of all, inasmuch as there is à diversity in morphology and cytology, changes due probably to hydrogen-ion concentration and temperature and several other factors among which the changed habitat is outstanding. In this third and final stage, we find that sexuality has developed and asci with ascospores are produced. To follow the life cycle on an artificial substrate, it would then seem best to follow the development from the germinating ascospore. In doing this, it is necessary to consider first normal growth on some favorable medium, and second to correlate the nuclear changes in material fixed in hanging-drop preparations (Van Tieghem cells) and stained, with the nuclear phenomenon found in the material fixed and stained as outlined in the technique. It was noticed that from three to fifteen days were necessary for the life cycle of the fungus to be completed, depending on the broth used in the hanging-drop preparation. It was found that proteose peptone broth gave good growth as did lactose broth, but that bacto- peptone plus bacto-beef gave quickest development of asci, from three to five days usually. The material was fixed by the addition of two or three drops of a fixative described previously, and after a period of from six to eight hours, or over night, the fluid was drawn off with filter-paper and the material was care- fully washed with distilled water and then mounted in lacto- phenol plus crystal violet, or stained with iron haematoxylin. By keeping several of these hanging-drop cultures growing, it was possible to obtain different stages of growth, even in one preparation. The method was of course not entirely accurate cytologically, but it served the purpose as demonstrated here. The single spore (2-215 u) is found to be made up of densely staining granular material, but the granules are so small that it is difficult to discern them. On germinating, a tube is sent out or the spore simply elongates, taking with it the chromatin material (pl. 22, figs. 1-2). When somewhat older, the devel- oping hyphae produce nuclei which divide rapidly as pointed 2 Several solutions were used: lactose broth (product of Digestive Ferments Co.), pH 6.8; 2 per cent bacto-peptone broth, adjusted to pH 7.0; 2 per cent proteose peptone broth, pH 7.0; bacto-peptone plus 6 per cent glycerine; meat extract broth, pH 7.1; 2 per cent bacto-peptone plus 5 per cent bacto-beef (dehydrated), pH 7.2. [Vor. 20 486 ANNALS OF THE MISSOURI BOTANICAL GARDEN out previously, are distributed in pairs very close together, and then separate (figs. 3-4). At this time the reticulum is well developed and the chromatin granules, or basophilie grains, as Guilliermond calls them and which he believes probably rep- resent albuminoid bodies playing the róle of products of nutri- tion in the form of perhaps zymogen or reserve material, are very evident and take a deep stain with haematoxylin. At the same time, the wall increases in thickness and becomes fairly evident as in fig. 4. The number of nuclei in the hyphae varies, as many as thirty-seven having been counted on one hypha, but this was apparently rare for the amount usually varied between seven and fifteen with nine the most common, as in fig. 4. With the elongation of the new hypha, septa are laid down (fig. 5) and the formation of conidia as described previously (pl. 22, fig. 5; pl. 19, fig. 4) may be and usually is associated with this act. Following this, the hyphae, varying from 11% to 4 u on different media, may grow out to form branches or reproductive structures, usually lateral or terminal. However, on acid media or on media where there is a definite lack of some suitable protein or protein product in the form of peptone, peptides, or amino acids, there is an excessive formation of chlamydospores, either terminal (hypnospores), 415-6 x 9-12 u, or lateral, 5x 7 u, or interealary, approximately 5 y in diameter, and conidia (pl. 18, fig. 7; pl. 19, figs. 7, 11-12). With the apparent development of a hypha, there is an accompanying formation of sexual cells. This process may be slow or rapid, depending on conditions. On glycerine agar it seemed to be the most rapid, these particular organs being produced in three to five days. The antheridial or male cell contains two or three nuclei as a rule (pl. 22, figs. 6-8), but cases of four have also been seen. The ascogonium or female cell usually contains three to four nuclei (figs. 9-11), but five (fig. 12) have also been noticed. When two sexual cells are ready for copulation, the antheridium bends over to meet the ascogonium as seen in pl. 22, figs. 10-11. The two gametes may be on the same hypha or on separate hyphae. Both may be terminal (figs. 13-17), or both lateral (fig. 16), or one may be lateral and the other terminal (fig. 12), or there may even be variations of these. Nevertheless it is 1933] MOORE—ENDOMYCES CAPSULATUS 487 quite apparent that the male cell goes to meet the female cell. With the approach of the copulation branches, there is an action which may be termed tropistic and which may be explained, if analogies are acceptable, by a hormone action or physical stimulus as is found in zoological specimens, which calls forth the produc- tion of a beak, the copulating tube, as has been found in the lower yeasts. These two beaks meet and fuse. In the meantime, one of the nuclei proceeds to the tip of the beak. There is probably no specific nucleus involved in this act, because as far as can be made out all the nuclei are the same morphologically, and any nucleus that happens to be nearest the beak becomes sensi- tized to the fertilization reaction. With the progression of the fertilization nucleus, there is an accompanying retrogression of the remaining nuclei. Whether these disintegrate or not has not been definitely determined. In several cases where fertilization had taken place, it was noticed that the antheridial cell and basal portion of the ascogonium merely existed very much like functionless stamens in a flowering plant, which in the course of time disintegrated or perhaps degenerated. The next step in the cycle involves the fusion of the nuclei, or syngamy perhaps. The walls between the two tubes are dissolved or diffuse and the two nuclei copulate or fuse in the tube (pl. 22, figs. 13-16). This is rather contradictory to Dangeard's ('04, '04a, '94b, '97) observations that the two nuclei fused in the ascus, and Harper's (96, ’97, '99, '00), that in the higher Ascomycetes the fusion occurred in the archicarp after the entrance of the contents of the antheridium. The fusion nucleus is then transferred or migrates (whatever the process may be it is analogous to that by which the nuclei copulate) into the asco- gonium which is now known as the archicarp (pl. 22, fig. 17). The fusion tube dissolves in some cases or breaks away in others, with the broken portion being either resorbed by the cells or degenerating. At any rate, the archicarp repairs its walls by the resorption of the cellular contents, and the large fusion nucleus is ready for division (fig. 18). The nucleus divides by a process of amitosis as previously mentioned. Although difficult to interpret theoretically, it presents no clear-cut chromosomal formation. With the first [Vor. 20 488 ANNALS OF THE MISSOURI BOTANICAL GARDEN division, there is apparently the formation of a wall (pl. 22, fig. 19). The daughter nuclei go through a synchronous division (fig. 20) to form four nuclei. In several of the yeasts and in Æ. Magnusii, the process halts here and each nucleus develops into a spore, thus four spores. Here, however, the four nuclei undergo a synchronous division, the second for the ascus, and eight nuclei are produced, which number is à mean for the species of this genus. In other species there may be twelve or sixteen nuclei. In the higher Ascomycetes the problem of ascus formation is rather confusing, and the reader is referred to the work of Atkin- son ('15), Bagchee (725), Blackman and Fraser (05), Brooks (10), Brown (710, '11), Carruthers (11), Guilliermond, Mangenot and Plantefol (33), Faull (05), Fraser and Brooks (709), Fraser and Welsford ('08), Gaümann and Dodge (28), Harper (796, '97, '99, '00), and Maire (’04, '05), to obtain full information and various opinions on the subject. After the formation of the nuclei and the abjunction of the future ascus (8-14 y. in diameter) from the basal portion of the ascogonium, now a basal cell as seen in pl. 22, figs. 21-22, the devel- opment of the spores begins. Associated with this phenomenon are the secretion products which, as has been pointed out pre- viously by Guilliermond, nourish the spores and produce the spore wall. It should be pointed out here that, in addition to this hetero- gamous form of copulation, isogamy sind finally parthenogenesis may also be present. The latter is quite apparent and has been found time and time again. The process may begin by the jutting out of what is apparently a conidium. "This in turn develops and a nucleus becomes visible. The nucleus divides as noted in the above reaction and an ascus with eight ascospores forms (pl. 23, figs. 1-9). Also a terminal cell often produces a pyriform ascus parthenogenetically as seen in fig. 11. In all cases, the spores serve the same function and are equally as capable of producing new hyphae. The sporogenous plasma of the ascus, part of the cytoplasm, contributes greatly to the formation of the mature spore. The life cycle of E. capsulatus on an artificial substrate might be graphically illustrated as in diagram 1. 1933] MOORE—ENDOMYCES CAPSULATUS 489 Cellular contents.— The remainder of the ascus constitutes the epiplasm which is made up of reserve products in the form of lipoids, glycogen, the metachromatie granules of Guilliermond or nuclear decomposition products as volutin, oil globules, and nucleie acid substances and probably other protein derivatives and carbohydrates. All of these materials have been shown to be utilized by the spores in their growth. Several authors also present evidence that during the maturation process of the spores, £ antheridium | bot — ,hyphac > Spscus | pon E ascogenous hypha chlamydospores hvpnospores —+ hyphae ‘conidia -—— Diagram 1. the substances constituting the epiplasm are broken down, a part being absorbed and clearly demonstrable in the granular contents at maturity, and a part reserved for their germination. Volutin.—The cellular contents will be considered in the order studied. The first is what has been variously called ''Neisser's granules" found by Neisser in bacteria; ‘‘sporogenous grains" found by Ernst; ‘‘metachromatic corpuscles’ of Babés, who found them in the diphtheria bacillus and so named them because of their metachromatic action; ‘‘red grains" of Bütschli, so named because they took on a red coloration with many of the stains; and finally, one of the most common of all, volutin, as named by Meyer. In the fresh state these granules stain blue with methylene blue (pl. 23, figs. 16-18), and when fixed their color is red to violet. They are small at first, and then enlarge to assume various forms and sizes. They may have a heavy staining outer portion and [Vor. 20 490 ANNALS OF THE MISSOURI BOTANICAL GARDEN a paler center, probably due to their refractive powers. They are seen quite often in vacuoles as represented here and assume fantastic features. ‘They appear in the cell along the reticulated network, at the nodes, and stain heavily there. Volutin is very common in the younger cells, but not so abundant in the older hyphae. ‘The substance may also be stained in vivo with neutral red. Wager and Peniston (710) found Gram's aniline violet to be a useful stain for volutin. Zikes (22), studying the nature of volutin and of the physio- logical factors concerned in its production, found that it is present in moderately large amounts in nearly all fungi. Peptone when added to the culture media stimulated its production while ammonium sulphate and asparagin were not so favorable. Phos- phorus was found to be essential for its formation, and glucose and fructose were more favorable than the carbohydrates of higher molecular weight. He concluded that volutin is an albuminous substance similar to the nucleo protein, as it contains both phosphoric acid and nuclein. The general opinion seems to be that volutin is a nucleic acid substance with an unknown base, probably organic in nature, that it exists as a colloid, and that, due to fixatives, it is pre- cipitated in the vacuoles. It is supposedly a secretion from the nucleus and is used in nourishing the spores. Glycogen.—Another substance which supposedly acts as a reserve material and is used up for the maturation of the spores in the ascus is glycogen. This theory has been substantiated by the fact that glycogen is used up in the cell, particularly in maturing asci, as evidenced by its slow disappearance. Kohl (07) considered glycogen not as a reserve material, but as a regulator for the intake of sugar in the cell, in that it is the sub- stance formed, not being able to diffuse out from the membrane. He found it to be lacking in spores, and since reserve materials should be present, he presented this as evidence that glycogen is not a reserve material. Guilliermond, however, claims to have demonstrated that glycogen is absorbed by the ascospores at the time it is found to disappear from the epiplasm. Glycogen is found in the organism practically from the begin- ning of growth, first as small droplets perhaps, but later in 1933] MOORE—ENDOMYCES CAPSULATUS 491 fairly large masses (pl. 23, figs. 19-30). It has been estimated as constituting 32 per cent of the dry weight of yeast cells. It is associated with the nutritive condition by some workers and found to be almost entirely used up or greatly accumulated according to the condition of nutrition and growth. According to some workers it appears somewhat in the form of a vacuole and it has thus been called a glycogen vacuole. It is not a true vacuole, but perhaps a colloidal substance, as volutin is considered to be, forit takes a stain much as thelatter does. It may be stained in vivo with neutral red (figs. 26, 29-30), in which case it is not so clear, at least in E. capsulatus, as with a saturated solution of iodine potassium iodide, a modification of Gram's stain (figs. 19-25, 27-28). This substance is a glucoside, a polysaccharide which, like starch, is made up of n molecules of glucose to have the formula (CeHiOs)n. It is often compared with starch since it requires phosphorus for its formation. In fact, it has often been referred to as an amylopectin, which makes up the superficial part of the starch grain. By hydrolysis, glycogen is broken down to dextrin, then maltose, and finally glucose. On heating a saturated solution of glycogen stained with iodine potassium iodide it becomes pale but regains its intensity on cooling. It is insoluble in alcohol. Vacuoles.—The question of vacuoles in fungi has been given considerable attention in the past, and the author does not intend to enter into a discussion of them. However, in view of the many theories advanced as to their character and presence, mention of them in E. capsulatus would not be out of place. They are easily demonstrable with many dyes and present varied characteristics. With methylene blue, they are easily made visible (pl. 23, figs. 15, 17-18). They appear to be formed by the reticulum and to take the stain. They may be demon- strated fairly clearly with iodine green (fig. 14), and the standard dye seems to be neutral red, in which case small crystalloid bodies, often described as ‘‘dancing bodies," probably Brownian movement, may be seen (figs. 26, 28). Haematoxylin also brings them out distinctly by staining the surrounding network (figs. 1-2). What the function of vacuoles may be is difficult [Vor. 20 402 ANNALS OF THE MISSOURI BOTANICAL GARDEN to understand. It is known that they are colloidal solutions of cytoplasmic materials, formed by the absorption of water, that they may enlarge or disappear in a cell, and Guilliermond, Mangenot and Plantefol, say this of them, * Les vacuoles rep- résentent la phase aqueuse du cytoplasme et jouent certaine- ment le róle important en réglant la teneur en eau de la cellule." Chondriosomes.— he bodies designated as mitochondria or chondriosomes have been demonstrated in animal and plant tissue. What their particular function may be is not definitely known. In plants they have been associated with the elaboration of chlorophyll and starch through the agency of the chloroplasts. In animals they are supposedly associated with the elaboration of secretions, but no definite evidence has been produced along this line. They have been demonstrated in a number of lower plants, lower fungi and the higher Ascomycetes, and in fact, in practically all of the groups, as well as in animal cells. For further information the reader is referred to Guilliermond ('11b), who has traced the development of these particular bodies in many organisms. Chondriosomes (Cowdry, 17) have been shown in both animal and plant cells to be almost completely soluble in alcohol, ether, chloroform, and dilute acetic acid (organic solvents) and to be insoluble by chromization and treatment with formalin, at least in most cases. In the animal cell they do not stain with Sudan III or IV and are sometimes blackened with osmic acid. In plants, the whole cell blackens with osmie acid so readily that it is impossible to know just to what extent the chondriosomes themselves are affected. In studying E. capsulatus, the author made use of the knowledge that chondriosomes are very clearly demonstrated by the addition of iodine potassium iodide to a preparation of an organism. They are fairly evident as lightly staining rod-like and short bodies (light yellow with iodine), but one must pay particular attention to distinguish them (pl. 23, figs. 19-25). "There are short and rod-like bodies in the terminal hypnospores and some long forms too, as well as small point-like reflective granules. The rods are longer in the hyphae (figs. 23-24). Chondriosomes are scattered throughout the asci as fine, small, rounded bodies. 1933] MOORE—ENDOMYCES CAPSULATUS 493 It can thus be stated that by the iodine potassium iodide method, bodies comparable to those described as chondriosomes for other fungi are demonstrable. Fat, lipoidal substances.—In addition to the many substances mentioned, there are also fats, lipoidal substances, and other materials called reserve materials, secretion products, and also excretion products. These materials are found in varying amounts in many of the specialized portions of the organism, being abundant in the asci and chlamydospores and in very small amounts in young hyphae and younger elements. They can be demonstrated comparatively easily in spores and in yeasts particularly. The nature of many of these substances has not been determined as yet, but it is generally known that fatty acids, glycerides and sterides, glycerol, phospholipides and phosphoaminolipides (complex lipides) may be present in certain amounts. Several agents were used for the study of these materials, each giving some degree of difference in distinguishing them. A 2 per cent osmic acid solution which reduces fats and gives them a black coloration was the first tried. When applied to the living mycelium, the fats and lipoidal substances appear as small refraetile bodies or droplets but in some cases they are very much larger (pl. 23, figs. 31-36). "They are seen rather abund- antly in old hyphae (fig. 36) and in very small amounts in young hyphal tips (fig. 35). With platinic chloride (5 per cent solution) they appear as blackened granules much the same as with osmic acid (figs. 37-38). Iodine potassium iodide as applied for glycogen and chondriosomes shows lipoidal substances, as oil droplets, equally as well. These are very small, highly refractile, and hyaline, and to be seen require careful focusing and adjust- ment of the microscope (figs. 19-20, 23-25). Neutral red has also been used, but not much attention was paid to the reaction outside of what has already been mentioned with regard to glycogen and the vacuoles. Iodine green also has an affinity for lipoidal substances, but not enough attention was given to it to prove its value here. [Vor. 20 404 ANNALS OF THE MISSOURI BOTANICAL GARDEN CULTURAL CHARACTERISTICS In studying the cultural characteristics of E. capsulatus, all the media available at the time were utilized. This was considered essential, inasmuch as it has become an unfortunate custom with some mycologists to name as new species organisms which show a physiological variation on a different medium. Since the fungus had shown on previous occasions of culturing that it favored protein substrates, a variety of protein materials was chosen. In addition, regular routine media were used, some with much carbohydrate and some with very little. For some years the question of hydrogen-ion concentration has become an im- portant factor in work of this sort, and media with a wide range of pH were chosen within which the possibilities of growth had at one time or another been emphasized. All cultures were grown at a temperature of approximately 25? C. The following media arranged in the order of their decreasing hydrogen-ion concentrations were used: Raulin’s Solution Agar (Raulin’s solution plus 1.5 per cent agar, pH 4.1).—(pl. 18, fig. 7; pl. 19, figs. 8, 10-12). Growth poor, being present only around inoculum after 18 days and having a diameter of only 0.7 em. and 1.6 em. at end of 30 days. Color white. Hyphae grown heaped up from center in a loose fashion. Microscopically, abundance of conidia, pedicelled or sessile, 5 y. in diameter; endo-chlamydospores 5-7 y. in diameter; terminal hypnospores 4-6 x 8-11 u; hyphae 2 y in diameter; racquet mycelium in small amount, not very noticeable; asci 9-10 v in diameter. Richards’ Solution Agar (Richards’ solution plus 1.5 per cent agar, pH 4.3).—(pl. 19, figs. 5-7, 9). Growth slow, very sparse and cottony, with long and narrow hyphae 2 y in diameter, projecting loosely and irregularly from the edge of the colony which attained a diameter of approximately 1.8 em. at end of 18 days and 3.5 em. at end of 30 days. Round conidia very nu- merous, approximately 4 y. in diameter; pyriform conidia sessile or on short stalks, several, 314 x 7 w in diameter; round, thick-walled chlamydospores on short pedicels, 5 u in diameter; asci few, 7-9 y. in diameter; endo-chlamydospores approximately 6 y in di- ameter. 1933] MOORE—ENDOMYCES CAPSULATUS 495 Czapek’s Agar (pH 4.4).—Growth of very loose and sparse mycelium which spreads over the surface of the agar with thin hyphae. Macroscopically the culture is barely visible except by reflected light. Colony approximately 3 cm. in diameter at end of 18 days and 6.7 cm. at end of 30 days. Color white. Hyphae 214-31% y in diameter; racquet mycelium present, 6-7 y. in diameter at swollen portion and 214-3 y at thin portion; terminal hypnospores 7 x 12 v approximately; conidia many, 5 y in diameter; round chlamydospores 6 y in diameter; lateral chlamydospores 6 x 11 yu; asci several, 10 » in diameter. Wort Agar (Product of Digestive Ferments Co., pH 4.6).—(pl. 18, figs. 6, 8-10, 15). Growth at first slow, none on several of the cultures, in general thin and loose. One of the cultures showed good growth, attaining a diameter of 3.2 cm. at end of 18 days and 5 em. at end of 30 days. This was unusual and may be ascribed to too great an inoculation and to a dissemination of spores as a result of shaking. Color white. Terminal hypno- spores many, 7 x 12 y; hyphae 314-4 y. in diameter; chlamydo- spores 614-7 u in diameter; racquet mycelium abundant; conidia numerous, 5-6 y in diameter; asci 10-11 y in diameter. Malt Extract Agar (pH 5.1).—(pl. 18, figs. 1-3). Growth not so abundant, colony 1.6 cm. at end of 30 days. Color slightly brown, due to the malt extract. Mycelium of numerous swelled cells; hyphae 3-4 x in diameter; terminal hypnospores 6 x 16 y; hyphal swellings or chlamydospores 9 x 12 u; numerous conidia, pyriform and round, the round being approximately 5 y in di- ameter; asci very few, 10 u in diameter. Malt Extract Broth (The above minus the agar).—Very little growth at end of 30 days. Characteristics same as above. Maltose Agar (pH 5.4).—Growth slow. Colony coremium- like, with a diameter of 1.5 em. at end of 30 days. Hyphae short, 3-4 u in diameter. Many round chlamydospores 5 y in diameter; terminal hypnosp numerous, 4-6 x 9-11 u; asci few, 8-10 y in diameter. Sabouraud's Broth (Sabouraud's dextrose agar minus the agar, pH 5.5).—(pl. 21, fig. 21). Culture consists of submerged colonies of mycelium varying from V4 to 3 cm. in diameter at end of 18 days, with a great mass at the end of 30 days as a result of the [Vor. 20 496 ANNALS OF THE MISSOURI BOTANICAL GARDEN coalescence of all the colonies. The large flakes are grey in color when moist, but with white mycelium on the surface, fairly dry. Submerged hyphae 214 y in diameter with none or very few morphological characteristics. Aerial or dry mycelium, however, similar to that on the agar culture of the same medium. Sabouraud’s Agar (pH 5.6).—(pl. 16, figs. 12-14; pl. 19, figs. 13-25; pl. 21, figs. 14, 18, 28). Thick cream-colored growth with a diameter of 4.5 em. in 18 days and 7 em. in 30 days. Hyphae 2-3'5 y in diameter, in needle-like projections from the surface of the mycelium which appears very cottony. Color white. Older cultures show a felt-like matting with a tendency towards ridge formation, a condition found to a certain extent in Microsporon Audouini Ota and Langeron. Racquet mycelium present with swollen portions 4-6 » and narrow section 3 y; many lateral chlamydospores, 5-7 x 10-12 u; terminal hypno- spores 5 x 11 u; and many conidia, pyriform or round, approxi- mately 5 u in diameter; numerous asci, 10-12 y in diameter, with 8 ascospores 2-216 y in diameter. Oat- M eal Agar (Decoction of oat-meal plus dextrose and agar, pH 5.9).—Growth diffuse and slow, 5 em. in diameter after 30 days. Culture loose and cottony. Hyphae 2-214 y in diameter; numerous thick-walled, round resting cells or chlamydospores, 4—6 p in diameter; terminal hypnospores 4 x 7 u; conidia numerous, 4 y in diameter; asci several, approximately 8 u in diameter. Corn-Meal Agar (Product of Digestive Ferments Co., pH 6.0).— (pl. 16, figs. 8-9; pl. 21, fig. 24). Growth very loose and thin, with the hyphae projecting from the colony, appearing as threads of silk. Colony barely visible except from a lateral view. Hyphae long and thin, 2-214 y in diameter, with colony attaining a diameter of 5 cm. at end of 21 days and 6 cm. at end of 30 days; chlamydospores 6 y in diameter; conidia many, 4144-5 y in di- ameter; terminal hypnospores 5 x 7 u; racquet mycelium present but reduced in size, 3-4 y. at swollen portion; asci several, 7-8 y. in diameter. Potato-Dextrose Agar (Decoction of potatoes plus dextrose and agar, pH 6.2).—(pl. 21, fig. 22). Colony 3.6 em. in diameter at end of 30 days, edge smooth and round, no striations or other cultural changes as ridges. Growth thick and cottony at in- 1933] MOORE—ENDOMYCES CAPSULATUS 407 oculum, with a thin periphery about 2-3 mm. in width. Color white. Hyphae 2-214 y in diameter and long; conidia numerous, 5 y in diameter; terminal hypnospores few, 4x 7 v; chlamydo- spores rare; asci des: 8-10 y in diameter. Two Per Cent Aqueous Bacto-Peptone (Hanging-drop culture, pH 6.2).—(pl. 21, figs. 1-8, 25, 32). Mass of thick-walled hyphae ramifying and branching, 2-215 y. in diameter; racquet mycelium not evident; numerous chlamydospores, 4 y. in diameter; asci 4—6 y. in diameter; few terminal hypnospores, 3.5-5 x 6-7 y. Two Per Cent Aqueous Bacto-Peptone Plus Five Per Cent Meat Extract (Hanging-drop culture, pH 6.2).—(pl. 19, figs. 1-4; pl. 21, figs. 9-13). Growth profuse, covering the drop in 4 days. Hyphae 2-214 y in diameter, with an abundance of racquet mycelium intertwining and branching; thick-walled cells, chlamy- dospores in abundance, as well as asci 10 » in diameter. Two Per Cent Proteose Peptone Plus Six Per Cent Glycerine (Hanging-drop culture, pH 6.2).—(pl. 16, fig. 7; pl. 17; pl. 21, figs. 15, 30). Growth similar to that on bacto-peptone broth, but with an abundance of chlamydospores and racquet mycelium. Lactose Broth (Product of Digestive Ferments Co., pH 6.8).— (pl. 21, fig. 17). Growth of submerged large flakes of colorless mycelium measuring approximately 2 cm. in diameter at end of 18 days. These later coalesced or intertwined into a mat which grew up the sides of the flask to form a white mycelium. Sub- merged hyphae approximately 3 u in diameter, branching, inter- twining, with cross-walls. Swellings, chlamydospores, terminal hypnospores, and asci few in number and reduced in size as compared with those on agar. The aerial mycelium above the surface of the broth showed an increased number of characteristics which simulated those found on the agar. Lactose Agar (The above medium plus 1.5 per cent agar, pH, 6.8).—(pl. 18, fig. 11; pl. 19, fig. 20). Growth rapid, thick and cottony, colony attaining a diameter of 3.6 cm. at end of 18 days and 6.2 em. at end of 30 days. Culture showed 4 ridges radiating from a thick inoculum to a thick cottony circular periphery surrounded by a thin growing rim of hyphae. Culture similar to that on nutrient agar, macroscopically. Hyphae 214-3) y in diameter; conidia many, 414-5 y. in diameter; terminal hypno- [Vor. 20 498 ANNALS OF THE MISSOURI BOTANICAL GARDEN spores 4 x 715 u; racquet mycelium present in abundance; many chlamydospores, varying from 6 y in diameter (round) to 5x 8 u (pyriform); asci many, 8-9 y in diameter. Nutrient Agar (Product of Digestive Ferments Co., pH 6.8).— (pl. 21, fig. 19). Growth good, with the colony having a diam- eter of 3.5 em. at the end of 20 days and 7.0 cm. after 30 days. Growth loose and cottony, showing concentric circles, evidently due to a periodical formation of asci and the liberation of the spores which germinated. Single spore colonies 2 cm. in diameter after 15 days. Colony flat with age, and white. Hyphae 2-3 u in diameter; conidia many, sessile or pedicellate, 5 y in diameter; chlamydospores several, 6 ». in diameter when intercalary, 5 x 7 y when lateral, 6 x 11 u when terminal (hypnospores); racquet mycelium present, 5 x 3 y. Nutrient Broth (Meat extract, pH 6.8).—(pl. 21, fig. 31). Cul- ture similar to that on lactose broth. Uschinsky’s Protein-free Medium (pH 6.8).—No growth. Beef Extract Agar (Liebig's extract of beef, pH 7.0).—(pl. 16, fig. 15). Growth similar to that on lactose agar. Radiating ridges present. Eosine-Methylene-Blue Agar (Product of Digestive Ferments Co., pH 7.0).—(pl. 21, fig. 27). Growth good, attaining a diameter of 5 em. after 26 days. Color pink in younger portion of culture (periphery), darker pink to blue towards the inoculum, this being due to an absorption of the dyes by the mycelium. Growth ceased and colony became flat. Hyphae in growing culture 2-316 y in diameter, with morphological characteristics similar to those on Sabouraud's agar. Glycerine Agar (Nutrient agar plus 6 per cent glycerine, pH 7.1).— (pl. 16, figs. 1-3, 11; pl. 21, figs. 16, 23, 26, 29). The colony on this medium grew as a cerebriform, very thick, creamy culture, having a diameter of 6 cm. at the end of 30 days. Medium best for study of organism because of its nutrient constituents, pres- ence of abundant protein and carbohydrate, and an abundance of endo-chlamydospores, 515-6 x 11 v; intercalary chlamydo- spores 4-5 x 5-7 v; terminal hypnospores 4-6 x 9-12 u; hyphae 2V5-4 u in diameter; round chlamydospores, either terminal or lateral, approximately 6 y in diameter; racquet mycelium abund- ant; asci numerous, 10-14 y in diameter. 1933] MOORE—ENDOMYCES CAPSULATUS 499 Blood Agar (pH 7.2).—(pl. 16, figs. 4-6). Growth similar to that on nutrient agar. Chocolate Agar (Blood agar heated to about 75? C. until the blood became chocolate colored, pH 7.2).—Growth similar to the above. Serum Agar (Beef extract agar plus 10 per cent dehydrated blood serum, pH 7.2).—(pl. 18, figs. 4-5, 12-14, 16). Growth moist, showing a diameter of 1.4 cm. at end of 10 days and 3.5 em. at end of 30 days. Colony flat and even. Many yeast-like cells seen; hyphae few and several large cells. Center of colony cerebriform. With age the culture shows prickly forms of mycelium (dry) which covers the greater part of the culture. This evidently is a reversion to the yeast form, to a certain extent. Calcium Carbonate Agar (pH 7.4).—(pl. 16, fig. 10). Growth slow at first, then rapid, with a diameter of 7.4 cm. in 30 days. Colony compact, with a cerebriform central portion, and a loose, cottony outer zone. Color white. Hyphae 2-214 y. in diameter; conidia numerous, 414-5 y. in diameter, being spread throughout the entire culture, mostly round, several pyriform on short peduncles, others sessile; round chlamydospores many, 6 y in diameter, and terminal hypnospores 4-5 x 7-9 v; racquet my- celium not very evident; asci many, approximately 8 & in di- ameter. Endo's Agar (Product of Digestive Ferments Co., pH 7.5).— Growth slow, not quite so rapid as that on eosine-methylene-blue agar. Colony white at first, then, due to an absorption of the dye from the agar, pink to red, growing in coremium-like masses of straight hyphae with very few or no conidia. Hyphae short and thick-walled, 315-4 » in diameter. Few conidia in culture, 6 u in diameter; hypnospores very few, as well as round chlamydo- spores; asci several, 10 u in diameter. Litmus Milk.—Milk was heated in a flask for fifteen minutes in steam, then set away over night in the ice-chest to allow the cream to rise. The cream was then siphoned off and the milk diluted in the ratio of 1 part milk to 4 parts water, with enough litmus as an indicator. Tubes of this solution were sterilized by steam, then inoculated, and kept at 25° C. No growth resulted after 20 days. [Vor. 20 500 ANNALS OF THE MISSOURI BOTANICAL GARDEN Reaction to Temperature.—In studying an organism, it is necessary, especially if quantitative results are desired, to deter- mine conditions under which optimum growth can be obtained. One factor which enters into such considerations with fungi is temperature. It is known that various organisms grow best at certain temperatures and when placed in other conditions, growth will either be inhibited or retarded. With this point in mind, it was decided to grow a number of cultures on the same medium and same pH at various temperatures. As a medium, Liebig’s beef extract agar at pH 7.2, which on previous occasions had shown a qualitatively and quantitatively good growth, was employed. E. capsulatus, when first isolated from the lesions in its yeast- like form, showed very good growth at body temperature, 37.5? C. After having been kept in culture in its filamentous form for about two years, the optimum temperature was shown to be approximately 30? C. Continued growth on an artificial medium again reduced its optimum temperature, to approximately 25? C., as may be seen in table r, or figs. 1-2. Several cultures in petri dishes were grown at the various temperatures indicated in table r. The experiment was per- formed at three different times, and due to the fact that daily measurements required handling of the cultures, several of the plates showed contamination. A procedure such as this usually brings in some foreign spores, especially in a large laboratory where many open cultures may be found. 'The plates of the first two temperatures, —0.7? C. and 8.0? C., were kept in a Kelvinator, the former in the freezing unit where a fairly constant temperature was maintained, and the latter in a compartment some distance from the cooling device. The third set of plates, 16.0? C., was kept in a glass jar with running water around it. The temperature fluctuated somewhat, but not enough to affect the experimental data seriously. The plates of the other temperatures were kept in regulated incuba- tors. All the plates were inoculated with a loopful of a suspension mashed up with a sterile needle and contained 5 cc. of a beef extract broth solution of pH. 7.2. In some cases the inoculum proved greater than others. 1933] MOORE—ENDOMYCES CAPSULATUS 501 TABLE I MEAN DIAMETERS IN CENTIMETERS OF n GROWN AT VARIOUS TEMPERATURE Beef extract agar (Liebig's extract of beef) pH 7.2 Temperature in degrees Centigrade [xn o =) | © ~J oo © pi [en] © bo mÓ © bo or © Oo — © OO ~J © © o * > w [vv] m eememuono0tuodoÀuB6d]ouogdouooecocdge-—-: À C1 Ct di» 02 tO -1 Oo o m= for) oOoOoOoocoOooooouoonooooooooooooooooococooc.- Q3 03 03 02 00 Q2 Q2 Q9 Q2 Q2 Q2 Q2 r2 EO F2 bt2 t2 t2 — RRR OOCOCOCCOooOoCc oO o d d d d d 00-1c» c0 d -1C01C0020——0c 000] Hr 00cm 020200202 ANON OQ» C1 t2 OC 00 C1 t2 OC -1 ANO 00 CO» m B2 ON H8 HL ON OR PR DO Qo -1-1-*1-10: C: 0; C1 O1 Ct Ct Ct PRR WWWWNNN NRK KH OC CO NOON RPK 000€ t2ddo^ -1Cvt20cn -1H:»— 00v O2 -I1OCt tc O0: 02 OO 0o Ot d Cc»: 0:0 0v OU Ov Ct Ct Ct i Hu HR OC)0)02 02 b2 t2 t2t2t2 RK kK COO CO b3Q b2 OC d O0 -1 Ct Q2 ON ANHODAN ON C' t2. O 00 O» Q2 r— ON Hg» d m MIÍIRMOOOOOOOOCOCG Nr CO (d 000» Cv HS WW (e>) -I @ o *( indicates no growth Readings were made daily for 31 days, at approximately the same hour. The mean results of a representative series are given in table r. "Three plates were kept in the final table as representive of the experiment although several additional plates were used. [Vor. 20 502 ANNALS OF THE MISSOURI BOTANICAL GARDEN An analysis of the results of this experiment shows that no growth takes place at —0.7° C. and at 40.0? C. At 8.0? C. growth does not begin until the twenty-third day and then proceeds slowly, with à maximum diameter of 1.2 cm. which would rep- resent a total growth of 0.9 em. At 16.0? C. the culture does not show growth until the fourth day, and grows slowly but faster than that at 8.0? C., to show a final diameter of 3.9 cm. or 3.6 em. actual growth. At 21.0? C. there is a final diameter of 6.7 em., with a total growth of 6.4 em. At 25.0? C. the di- ameter of the colony reached an optimum of 8.3 em. and an actual growth of 7.9 em., and at 31.0? C. the diameter of the colony was 6.2 em. with a total growth of 5.8 em. The culture at 37.0? C. grew fairly well for the first ten days but after that no further growth. Transfers from these last plates to fresh medium produced no growth, hence it could be concluded that the heat finally killed the organism. In fig. 1, the diameter of the colony in em. was plotted against the time in days to obtain the rate of growth for each temperature. It may be seen here how the growth was affected. At optimum temperature, after the initial lag period of 1-2 days, growth proceeded in practically a straight line. With a deviation from the optimum temperature there is accordingly a decrease in the growth rate which corresponds to the temperature affecting it. At 16.0° C. we find a regular sigmoid curve which represents an initial lag period followed by a period of increased activity and finally a return to the lag period. When the maxima points of total growth are plotted, fig. 2 (diameter of colony against temperature), we have the relation- ship of the total growth to the various temperatures outlined. These results seem to indicate that the optimum growth for this organism has changed from approximately 37.5? C. in its parasitic condition to 25? C. in a saprophytic condition. This latter fact will be demonstrated in the animal inoculation experi- ments. It also shows that the longer an organism is kept in culture the greater will be its change in physiological phenomena to an optimum point for its changed environment. Hydrogen-Ion | Concentration.—The problem of hydrogen-ion concentration has been of interest for several years and its 1933] MOORE—ENDOMYCES CAPSULATUS 503 Ko] © o » 5 d d (2 o U 9 ped ^ ae © : . a N C M S O n B Q Q Q O b Q Q Q 12 > Q QQ Y R 3 v F ; 8 9 ei ; N H a 5 + i! le N f : E] wt 8 N e n » o0 a = ov 38 “ c £ Ll ^ Qo a e 8 E E S " E P" £z s A 2 da a vs E - e a0 2 "S o + [4] 9 ed v e tb 4 cay + N o o p to N - n o "uo UT SojuO[o9o Jo S202 oWeTd consequences especially emphasized by von Mallinckrodt-Haupt (32). It has been demonstrated that the change of pH in a medium plays an important part in the cultural characteristics [Vor. 20 504 ANNALS OF THE MISSOURI BOTANICAL GARDEN Temperature in degrees Centigrade Fig. 2 Maximum growth at various temperatures on beef extract agar with same pH. rr ra) + M "moO UT 89[uoTo9 JO si939uverq of the fungi grown. These changes may be brought about in their nutrient medium through the formation of acids which are 1933] MOORE—ENDOMYCES CAPSULATUS 505 present in large amounts, by the breakdown of the sugar molecule in its utilization by the organism, by the decomposition of the proteins or the products, and by the cracking of fats or lipoids as a result of growth. As is perhaps fairly evident in the case of the carbohydrate reactions to be explained later, we have the forma- tion of alkaline bases which are observed in the course of the utilization of the proteins. From a study of the cultural characteristics mentioned pre- viously and from further observations, a decided difference in growth on acid and alkaline media may be pointed out. In strongly acid media there is abundant budding and a formation of numerous conidia, and on strongly alkaline media a tendency towards shorter, thicker cells and a yeast-like cell formation. On the other hand, on weakly acid or alkaline media there is a favor- able condition established for mycelial and hyphal formation. The effect of alkalinity and acidity on the development of fungi was noticed for a number of years. Marantonio (93) found that a more acid medium favored a greater quantity of mycelium. These observations were confirmed by Concetti (’00) and were extended to include a large number of media. Strains of an organism isolated from cases of thrush were used. Fineman (21), working with Monilia albicans apparently, found that mycelium grew better in media under low surface tension and oxy- gen tension while the yeast form of the organism predominated on solid media, simple carbohydrates, and a high pH. — Milochevitch (29) found that varying the hydrogen-ion concentrations of media had no effect on the same sort of a fungus. Talice (30) found the best development of hyphae on dilute potato decoction, the hydro- gen-ion concentration being perhaps fairly high. several workers have pointed out that fungi possess a certain buffer action, that is, they have the ability to regulate their growth according to the reaction of the medium, and it is found to exist only under external changes. This means, then, that acid media become less acid during the course of growth and alkaline media become less alkaline. What has been found to be true for the effect of various tem- peratures on the growth of E. capsulatus may also be said to hold with regard to hydrogen-ion concentration. For testing this, [Vor. 20 506 ANNALS OF THE MISSOURI BOTANICAL GARDEN beef extract agar made up with Liebig’s extract of beef but adjusted to the various pH’s shown in table rr, was inoculated as described previously. The same procedure was carried out as in the above experiments, namely, daily diameter measurements and occasional microscopic examinations. Here, as before, TABLE II MEAN DIAMETERS IN CENTIMETERS OF COLONIES GROWN IN VARIOUS HYDROGEN-ION CONCENTRATIONS, AT A TEMPERATURE OF 26° Beef extract agar (Liebig’s extract of beef) Days iue pH values inocula- tion 2.1 | 3.3 | 42 | 5.2 | 5.5 | 6.1 | 6.4 | 7.0 | 7.4 | 7.7 | 8.2 | 9.3 [10.4 1 0.2 | 0.8 | 0.3 | 0.3 | 0.3 | 0.3 | 0.3 | 0.4 | 0.3 | 0.4 | 0.3 | 0.4 | 0.4 2 g» 0 0.3 | 0.3 | 0.3 | 0.3 | 0.3 | 0.4 | 0.3 | 0.4 | 0.3 | 0.4 0 3 0 0 0.3 | 0.4 | 0.4 | 0.4 | 0.4 | 0.6 | 0.5 | 0.6 | 0.4 | 0.4 0 4 0 0 0.3 | 0.5 | 0.5 | 0.6 | 0.7 | 0.9 | 0.8 | 0.9 | 0.6 | 0.5 0 5 0 0 | 03] 0.7 | 0.8 | 0.9 | 0.9 | 1.2 ]1.2] 1.2108 10.7] 0 6 0 0 0.3 | 1.0] 1.1 | 1.2 ]1.2 | 1.4116] 1.6] 1.1 | 0.9 0 7 0 0 0.31}1.8/1811.6118) 1.7 | 20,19] 14] 1.1 0 8 0 0 0.311.411. | 1.18]1.9|2.1|]2.3|22]|1.8]| 13 0 9 0 0.|0,88.1171108]|3.151 22:1 3.8 |.3/77 | 3812.0) 1.8) 0 10 0 0 0.4 | 2.0 | 2.2 | 2.4 | 2.4 | 2.6 | 3.0 | 2.8 | 2.4 | 1.8 0 11 0 0 0.5 | 2.3 | 2.5 | 2.7 | 2.7 | 2.9 | 3.3 | 3.2 | 2.7 | 2.0 0 12 0 0 0.6 | 2.6 | 2.8 | 3.0 | 8.1 | 3.2 | 3.7 | 3.5 | 2.9 | 2.2 0 13 0 0 0.8 | 2.9 | 3.1 | 3.3 | 3.4 | 3.6 | 4.0 | 3.8 | 3.2 | 2.8 0 14 0 0 | 10] 381 | 3.4] 3.5 | 3.7 | 38148 | 411]356)124/ 0 15 0 0 1.1 | 3.3 | 3.6 | 3.8 | 4.0 | 4.1 | 4.6 | 4.4 | 3.8 | 2.4 0 16 0 0 13|3.7|3.99|42|4.2|4.5|4.9|4.7| 41 0 0 17 0 0.4 | 1.6 | 4.0 | 4.3 | 4.5 | 4.6 | 48 | 5.3 | 5.0 | 44 0 0 18 0 |} 06] 1.8]43 | 45148 |50 | 5.2 |56] 54)471] 0 0 19 0 | 0.8] 2.1 | 46) 4.8] 51153 | 5.5 )]59] 5.7] 5.1] 0 0 20 0 | 0.9 | 2.3 | 4.9 | 5.1 | 5.4 | 5.6 | 5.9 | 63 | 6.0 | 54) 0 0 21 0/|[09/|[25|51/|53/|57/|60/|62/|606/|6.4|56]| 0 0 22 0 0.9 | 2.7 | 5.3 | 5.5 | 6.0 | 6.3 | 6.5 | 6.9 | 68 | 5.9 0 0 23 0.|090|2,8154|56|62]8.5|068|72|7.1|02]| 0 0 24 0 1.0 | 3.0 | 5.7 | 5.9 | 6.4 | 6.7 | 7.0 | 7.4 | 7.2 | 6.4 0 0 25 0 1.0 | 3.1 | 5.8 | 6.1 | 6.6] 6.9 | 7.2 | 7.6 | 7.4 | 6.5 0 0 26 0 0132/|58/|63/|068|7.0|7.4|]7.8|7.0|6.7| 0 0 27 0 0 0 5.9 | 6.4 | 6.9 | 7.1 | 7.5 | 8.0 | 7.8 | 6.9 0 0 28 0 0 0 6.1 | 6.5 | 7.0 | 7.2 | 7.6 | 8.2 | 8.0 | 7.0 0 0 29 0 0 0 6.2 | 6.6 | 7.1 | 7.8 | 7.7 | 83 | 8.1 | 7.1 0 0 30 0 0 0 568-167 | L1] T4 | 7.1] 84 182 71 0 0 *0 indicates no growth 1933] MOORE—ENDOMYCES CAPSULATUS 507 contaminations were present, so that only three plate measure- ments were incorporated in the table. The mean results in table 1 represent a typical set of data in this work. An analysis of the data shows that no growth took place at a pH of 2.1 and 10.4. At pH 3.3, growth was not evident until the seventeenth day, when the diameter was 0.4. "The total growth was 0.7 cm. on the twenty-fourth day, after which it stopped. A transfer of the culture to a medium of pH 7.2 gave normal good growth. With an increase in pH to 4.2, growth was at first slow, but the colony attained a diameter of 3.2 em. on the twenty-sixth day and then growth stopped, showing a total increase of 2.9 em. In like fashion, by subtracting the initial inoculum measurement from the final colony diameter, we find that at pH 5.2 total growth for thirty days was 6.0 cm.; for pH 5.5, 6.4 em.; for pH 6.1, 6.8 em.; for pH 6.4, 7.1 cm.; for pH 7.0, 7.3 em.; for pH 7.4, we find the optimum pH with a total growth of 8.1 em.; for pH 7.7, 7.8 cm.; for pH 8.2, 6.8 cm., which is the same as that for pH 6.1; for pH 9.3, total growth of 2 em. was acquired on the fifteenth day, with the colony showing no additional growth. By plotting the diameter of the colony against the time in days, fig. 3, we obtain the rate of growth for the optimum pH which shows, except for the initial lag and a final lag, the latter being due perhaps to external factors as slight drying and probable utilization of most of the nutrients, an almost straight line. This varies with the pH, showing a decreased growth rate from the optimum, with an increased lag and decreased period of activity. By plotting the maxima points of growth or total growth for each pH against the pH values, we obtain a clear relationship of the growth at various hydrogen-ion concentrations, with the point of optimum growth forming a peak, as seen in fig. 4. In addition, the pH of several of the media after growth was determined. No definitely accurate results were obtainable due to the dried-up condition of several of the media, but indications seemed to be that there was an increased pH where growth was best, with no change in the strongly alkaline and strongly acid. The results may be interpreted by considering the amount of growth as an index of the production of alkali due to the utiliza- tion and decomposition of the protein substances in the medium. [Vor. 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN 508 'Suorsrjugsouoo uor-uoZo1pÁq SNOVA Y} 1929 j991jxo joeq uo einjwreduroe) juvjsuoo 4B YMOIZ Jo ovp. "e “BLY 92 +z skep Uf 9euTl zz T oz 81 91 v! 2t ot + (Zt nd) © (zs pd) o (s.s Hd) c Sud (à Ka) o 2H?) o joe Hd) o (22 Wd) o rz Hd) © (ce " m IM" 3 * z ‘wo uj Soajuo]oo Jo s193euvT] w 9 1933] 509 MOORE—ENDOMYCES CAPSULATUS 'suonerjuoouoo uor-uogo1pKq snorreA YIM IVBV j9urjxo Jood uo e1njuredure, JUVJSUOI 4B YYMOIS umuIxvpN ‘P "Sur səna Hd ol 6 e E 9 S v € 2 1 + to "uo ug GEeTUOTOD Jo s1oqeuv[q Lu v. [Vor. 20 510 ANNALS OF THE MISSOURI BOTANICAL GARDEN SUMMARY OF CULTURAL WORK The results of the cultural studies described above for the three organisms, Hndomyces capsulatus, E. capsulatus var. isabellinus, and E. dermatitidis, but more especially for E. capsulatus, may be summarized as follows: 1. Ability of the organisms to grow on a fairly wide variety of media. 2. No growth or very little growth on media lacking or having very little protein or carbohydrate. 3. An increase in the number of conidia and decrease in diam- eter of hyphae on acid media as compared with fewer conidia and increased thickness of hyphae on media with a high pH. The number of conidia and size of cells may be an index of the hydrogen-ion concentration of the medium. 4. E. capsulatus and its variety isabellinus show an almost direct change from a yeast-like to mycelium growth macroscopi- cally, as evidenced by the change from a moist colony to a filamentous colony, whereas E. dermatitidis shows a macro- scopically evident secondary stage in the form of a definite, prickly or coremium-like growth. The secondary stage is identical microscopically, for all the organisms. 5. E. capsulatus has a white color, isabellinus a light chamois or isabella, while E. dermatitidis has a dark brown or cinnamon color. 6. Size and number of various morphological characteristics differ on various media. 7. Reversion to the yeast-like form by growth on serum agar. 8. No growth on litmus milk. 9. The organism shows an optimum temperature of 25° C. with no growth occurring at temperatures above 37° C. and below 8.0? C. 10. The optimum pH is shown to be approximately 7.4 with deviations from the mean, and no growth occurring in a pH higher than 9.3 or lower than 3.3. BIOCHEMICAL REACTIONS Carbohydrate Reactions.—There have been many conflicting results in the past with regard to the carbohydrate reactions of yeasts and yeast-like organisms. Systems of classification have 1933] MOORE—ENDOMYCES CAPSULATUS 511 been established whereby an organism was placed in a definite taxonomie position because of its production of either acid or gas, or both, or whether or not it had an effect on a carbohydrate. Such a system has been established by Castellani for Monilia. Whether any confidence can be placed in a method of that sort is doubtful, first, because it is problematical, at least in some cases, whether absolutely pure sugars are obtainable, and small amounts of impure substances may bring about an altered reaction which may not coincide with the specified phenomenon. That is, a reaction may be indicated as slight or weak, which actually should have been negative. Dekker in 1931 outlined the various methods of sugar fermentation and the faults to be found with each, and for further information, the reader is referred to her work. In the second place, it has been evidenced in the past and frequently noticed at present, that pathogenic organisms of the sort mentioned above have often lost their fermentative abilities on standing in artificial media. The constant sub- culturing has resulted, as in the case of many fungi, in a change from a pathogenic and virulent parasite to a non-harmful sapro- phyte, and while this may not necessarily be an index of the sugar reactions of an organism, still we must consider such occurrences as of some significance. It would seem, therefore, that not much reliance must be placed on a system of that sort. In the work carried out here, stress was laid not so much on the desire to organize and establish a new method of classification, such being altogether too numerous, but to determine whether a yeast-like organism, such as E. capsulatus, could produce any acid or gas. For this purpose, Pfanstiehl sugars, which have been found to be considerably above the average, were used. The sugar was added, at the rate of one per cent, to a beef extract broth as prepared previously, to which had been added phenol red as an indicator. The phenol red was made up as a 0.02 per cent solution in distilled water and added in the amount of 1.5 cc. of indicator to 10 cc. of medium. This gave a fairly deep color which proved satisfactory in the work. This particular dye was chosen because preliminary experimentation had shown that the reaction tended more towards alkalinity. Litmus was also used in several series, but the results were not so clear and definite as with phenol red. [Vor. 20 512 ANNALS OF THE MISSOURI BOTANICAL GARDEN A 12-day-old culture on beef extract agar was mashed up with a sterile needle and 5 cc. beef extract broth of the same pH were added to form a suspension. A loopful of this supension was inoculated into Smith fermentation tubes which contained the above medium. All the tubes were grown at 25? C. The carbohydrates used were those listed in table rri. For the sake of clarity, they may be outlined as follows: Monosaecharides. Pentoses: Aldoses; 1-arabinose, 1-xylose Hexoses: Aldoses; rhamnose, dextrose, d-galactose, d-mannose Disaccharides. lactose, maltose, saccharose Trisaccharides. raffinose Colloidal polysaccharides. dextrin, starch (soluble), inulin Glucosides. amygdalin, salicin These carbohydrates represent a fairly wide range and include those sugars which are used in the regular routine sugar reactions, as xylose, dextrose, maltose, and lactose. There will be no attempt at present to explain the work in detail or to offer any complex chemical formulae for possible decompositions, but the reactions and a possible empirical reason for the observations will simply be pointed out. An examination of table mr in which the carbohydrates are arranged in the order of their increasing groups, shows that when compared with the control (the same medium minus the carbo- hydrate) in the first column there was a changed reaction due to the presence of a carbohydrate. It will be noticed that the control tube showed a change on the eleventh day and a period of reactivity for six days, reaching its maximum on the seven- teenth day after inoculation. It is further seen that 1-arabinose, l-xylose, dextrose, d-levulose, lactose, maltose, saccharose, and soluble starch showed the same initial period of definitely de- 513 y MOORE—ENDOMYCES CAPSULATUS rg i i re re T t re r9 1 1 ! Te 1 t eds OPUS oe E er EV S Eee ceo PN rm t e ee ae es Ba Y r ---— ---- --—- ---- ---— ---- ----| | |---- [7 ---- ---- ---- 62 62 -€—— t re ---- ---- ---- i ——-— ---— ---- ---- 94 [y] 1 o9 o9 ——— | —— — —— — o 34 I -| =--> ipli = — lei BL 62 r akg csa 6L SL] ==- [) ---- 1 97 gL SL BL tg 9L eZ 62 t ---- eL £z| —- 8 ---- GL NE E&L -—— -—— |e2 o T€ EX ---—|üu| ~ jrg 34 —— —— {ez} —-— ---|e1|---- --- --— --- |62|---- ----|||---- --- o |jez|----|s4| --— ---- -— -— --- ---|ez 5L --— |e 19 ---— ---j|oi1| o |94|---- =- rei 7---— 5:201 L3 EON ER EL m- n^ ---|S21|--— == |94 62 aL o M e =- |&z|---- == ò Pij -== isi -— —— |»z2|---]|s2 BL Sz or MI ae ee ez| —— |o4| o —-- |s4|-—-— -— —-— |s4|--— —— |e2| --— Ira|---- -— o [|sz|---|£2| -= joz| --— o - o -— —— jez] -— |[s2| == fez] —— jez] == |s2| =-= |e2|-7-7-7] | | 7— o -- - zh use o = o E -— | — - - - --— je4|----|e2| -— o jez] -- - |e4| -— o7 TE = o SSL = a = a E n oe eee ee te Ce me Nip HIP — o o o -— = | ò o o o = [|se24| —- o o o o o ot - o |o4| 9 o -— [r£ — joz| o |jo4 o Jaz} o joZ| ò Jide — -— pz) o o {92} o jo, © oJ o 6 o [7 o o jez] -- o [e] o o o o -- o o 9 o o [ o o o o = Jož o o o o o |o2| o == o o [*] o o 2 o o o o - o © o o o o |e4| -- [s] o o o o E] o o o x = o o o o o o | = o o o o o 8 o o o o u - 'o o o o o 9 iii -= o o o o o * o o o o |jo2| o o o o o o o j|o4| o o o o o o £ o o o o o o 'o o o o o o o o o o o 2 63] o o |e9| o les! o Isa] o ləèə| o Jee o les} o jeə| o {es} o |s»| o |sv} o |s9} o {aa} o jej o || o [89] © 9, 9 9, ©, % rele, e Si, P e el eti, e Pg | ey, ele, |e e abd LEN FENFU D [4] nd %, e E “On % ^ + “bs B > ^ ^ F 2 u^ 7 * % 1933] SNOILOVEY ALVHOAHOHNVO III @IaVL (Vor. 20 514 ANNALS OF THE MISSOURI BOTANICAL GARDEN monstrable reactivity, by the color changing to red, and that rhamnose showed a lag and hindered the reaction greatly before the color change, this being demonstrable on the nineteenth day. Amygdalin had a somewhat less harmful effect than rhamnose, with the reaction occurring on the sixteenth day, and salicin even less than that, having the initial color change on the fourteenth day. In contrast with the normal, we find that d-galactose and dextrin were very favorable towards color pro- duction, inducing a reaction on the fourth day, whereas d- mannose, raffinose, and inulin took effect in perceptible color changes on the ninth day. It might be added here that three series of tubes were used in each experiment, two series being used to obtain the changes in pH, which were measured by a potentiometer and also a set of colorimetric standards. A set of tubes was prepared with phenol red at various hydrogen-ion concentrations. 'This method proved very satisfactory for further work, being much more rapid and fairly accurate since readings were made only to the first decimal place. It was found in practically every case that & change in hydrogen-ion concentration occurred two days before a color change, except in the case of d-galactose and dextrin where the reaction was more rapid, taking place on the third day. This latter reaction may be explained somewhat by the fact that the two carbohydrates are very favorable to growth, since the color change is due to the growth of the organism which brings about a reaction in the medium as a result of the utilization of the nutrients. A striking feature of these reactions is the varied range of reactivity as may be seen in fig. 5. No explanation, outside of the rate of utilization of the carbohydrate, can be given at the present time for this phenomenon. As compared with the control, the range for dextrin is very large. By the range of reactivity is meant the period over which the color changes and the decrease in hydrogen-ion concentration take place, coming to an end point of an approximate pH of 8.1. The various carbo- hydrates present various ranges which are clearly evident in fig. 5. Soluble starch, added to the solution at the rate of 0.2 per cent, showed a partial hydrolysis on the nineteenth day as evidenced 1933] 515 1 MOORE—ENDOMYCES CAPSULATUS 'SsejurpAqoqr99 JO APIATZOVaI JO oguv] o “BLT [Vor. 20 516 ANNALS OF THE MISSOURI BOTANICAL GARDEN by a reddish-brown color on addition of a dilute iodine solution. Fehling's solution reduced to cuprous oxide as shown by a faint reddish-violet color. There was no complete hydrolysis at the end of the experiment. It may also be noted that in its production of alkalinity the reaction proceeded in every case from the aerobic portion of the Smith tube, or rather from the bulb, to the arm of the tube, or the closed portion. In other words, there was a diffusion of ions from the aerobic to the anaerobic. This of course was to be expected, since the organism is strictly aerobic, as will be shown later, and utilized all of the little material available, before growing towards the anaerobic portion. In every case, however, no matter how marked the reaction, the color change in the arm was compara- tively slow. Proteins are broken down very easily by reacting substances, and the cultural experiments have shown that E. capsulatus requires some protein, its derivatives, or carbohydrate, for growth. It is also known that these proteins and their derivatives may be further reduced or broken down to amino acids, peptones, pep- tides, and hexone bases and other substances, whereas higher carbohydrates may be reduced to simpler sugars. Ammonia is often found to be a product in the hydrolysis of proteins. In the presence of sugars, amino acids have characteristic reactions, particularly in an alkaline solution, and these have been dealt with by several authors, recently by Watanabe (732). This author lists many reactions and covers the literature of the field rather well, and the reader is referred to his paper for detailed information. Euler and his associates, in some work carried out in 1926 and 1927, found that the reaction between amino acids and sugars which were reversible, took place at an equimo- lecular ratio and that their velocity became greater in an alkaline medium. ‘This may account for the rate of activity as noticed in table rm. The work of several others seems to point to the conclusion that a reaction between amino acid and sugar requires a neutral reaction and a fairly high temperature, whereas Wald- schmitt-Leitz and Rauchalles observed that the optimum hydro- gen-ion concentration for the reaction between dipeptides and glucose was pH 8.1. This seems to coincide with the results obtained here. 1933] MOORE—ENDOMYCES CAPSULATUS 517 It is clearly evident from the experimental data that, instead of acid or gas or both, the carbohydrates here used induced the production of an alkaline condition. In addition to the protein decomposition substances as mentioned above, ammonia may be an end product. This substance, where present in fairly large amounts, brings about a strongly alkaline condition. To test for it, several cc. of the culture broth were poured into a test-tube, and a pinch of anhydrous sodium carbonate (Na,CO;) was added. 'The tube was shaken and a rather strong odor of ammonia was given off. It is possible that, besides the evident ammonia, hexone bases, as arginine, lysine, and histidine, which are alkaline in solution, are formed as products of the protein derivatives in the medium and that in addition to the sugar reaction they cause a decrease in the hydrogen-ion concentration. Indol and Skatol.—Reactions which have long been used in bacteriological technique but which have not received much attention until lately, due to the additional knowledge and better methods which are applied, are those of indol and skatol. The work of such investigators as Pittaluga (08), Goré (21), Holman and Gonzales (23), Morelli (09), Sasaki (10), Zipfel (12), Krumwiede and Pratt (113), Fellers and Clough (25), Koser and Galt (26), and several others has greatly added to the methods of technique in investigations. Tests have been advocated from time to time, the importance of which lay in their specificity for the reaction. Fellers and Clough list about thirteen tests which have been applied, some successful some not. Indol and skatol are protein derivatives and their determination rests on the presence of tryptophane. From it they are broken off by a deamidization to form first indol proprionie acid, then indol acetic acid, and finally indol or skatol. The medium to be used in such a reaction must therefore contain a supply of tryp- tophane sufficient to determine whether the organism would actually be capable of breaking it down to indol or skatol or both. For this purpose, it was necessary to use peptone plus 0.1 per cent casein in water to insure good growth and also to comply with the requirements as stated. The broth was inoculated and the organism was allowed to grow at 25°C. In the first test the Gnezda (799) oxalic acid principle was [Vor. 20 518 ANNALS OF THE MISSOURI BOTANICAL GARDEN applied. Strips of filter paper were placed in a warm solution (saturated) of oxalic acid. On cooling, crystals of oxalic acid re- mained on the strips. The paper was then dried well and inserted in the mouth of the flask under aseptic conditions, so that it was pressed against the side of the flask and just above the surface of the broth. Ifindolis formed it volatilizes and colors the oxalic acid pink. In these tests for E. capsulatus, a very faint pink was barely discernible. This test has been advocated rather strongly by the Society of American Bacteriologists for regular routine work in bacteriology. A test that was applied successfully is the Ehrlich-Bóhme technique which consists of two solutions. Solution 1 contains 1 gm. para-dimethyl-amino-benzaldehyde, 95 ec. ethyl alcohol (95 per cent), and 20 cc. hydrochloric acid, concentrated. Solu- tion 2 is a saturated solution of potassium persulphate (K;S,0;). In the experiment, 5 ec. of solution 1 are added to 10 cc. of the culture fluid, then 5 cc. of solution 2, and the mixture is shaken. A faint red color appearing about five minutes later indicated a weakly positive reaction. This method has proven rather successful for indol. Steensma (06, 068) substituted vanillin for the para-dimethyl-amino-l lehyde, but this was not applied here. The Salkowski ('83) nitroso-indol test was negative for indol. Skatol was tested for by the Sasaki (10) methyl alcohol test. This method consisted of adding to the broth to be tested about four drops methyl alcohol and an amount of concentrated sul- phuric acid equal to the broth. The acid contained a trace of ferric salt. A reddish-violet color indicated a positive skatol reaction. Under the proper conditions, indol and skatol may be produced by E. capsulatus. Relation to Free Oxygen.—Aerobiosis has often been considered important in experimental work. Organisms are classified as strict aerobes, facultative aerobes, or facultative anaerobes and strict anaerobes. In order to find the position of E. cap- sulatus in this classification, tubes of broth were prepared and inoculated by the Liborius method (Zinsser, ’29) and kept at 25? C. Cultures were kept at the same temperature and in the 1933] MOORE—ENDOMYCES CAPSULATUS 519 same broth, but were plugged only, allowing oxygen to enter. No growth occurred in the former while a luxuriant growth occurred in the latter. Agar stab cultures were negative for growth in the stab, but positive on the surface. The relationship to free oxygen was also demonstrated in the carbohydrate re- actions as mentioned previously. We may conclude, therefore, that E. capsulatus is a strict aerobe, at least in so far as its growth on agar after a period of time has demonstrated. When in the yeast form, anaerobic cultures showed slight growth, a condition of facultative anaero- biosis. Continued growth on an artificial substrate has con- verted it into a strict aerobe. Reduction of Nitrates.—The tests for nitrate reduction as advocated by the American Society of Bacteriologists were inconclusive and unsatisfactory. Ammonia was found to be present, but that was apparently due to the breakdown of the peptone and other protein decomposition products. ‘Tromms- dorf's method was also indefinite, being negative in general, but at times gave a slightly colored reaction. Production of Hydrogen Sulphide.—The formation of hydrogen sulphide in culture has often been concerned with a toxic action of the lead salt or lead acetate in the medium toward certain bacteria. Whether hydrogen sulphide would be produced here was of course problematical. However, a lead acetate agar put up by the Digestive Ferments Co., ‘‘Bacto lead acetate agar," which is claimed to contain too little lead salt to be toxic and yet was favorable for the demonstration of the hydrogen sulphide, was used. Making stab cultures was of course illogical, for reasons pointed out before, and so surface inoculations were made and the cultures incubated at 25? C. A slight brownish coloration was found to occur directly around the colony after thirty to forty days, and increased somewhat on standing. No definite method has been advocated for the above experi- ment, and since the results were slightly evident even though the period was a great deal longer than for bacteria, the reaction might be considered positive. Gelatine Liquefaction.—Tubes of 15 per cent meat extract gelatine were inoculated by stabbing and incubated at 25°C. A [Vor. 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN 520 slow liquefaetion took place between 30 and 40 days, and in some cases even longer than that. and proceeded downward. growth of the fungus. Liquefaction began at the surface This may be correlated with the REACTION To LIGHT A great deal of experimental work has been done in the past, and much is being attempted at the present, with respect to the AALL LTT LLL. UP RARA AARAA LR 2 CLL LL LLL SLA LAA ho rr 8 A AFF Bm SS ff WAR ma SAA AeA A AAAA ART O DSL AR AARAA ITTI anon SIF TD A A T Fig. 6. Explanation in text. It is action of light on higher plants and also microbiology. accepted as a fact by many that plant growth and vitality can be greatly altered, either for better or worse, by continued treat- ment under variously colored lights. No great amount of work, however, has been done on the action of the primary colors on Several workers have found a destructive fungi or bacteria. 1933] MOORE—ENDOMYCES CAPSULATUS 521 action of light on the lower organisms, and Rahn has found that red and green light have no effect while blue light has a very harmful action, with the degree of destruction decreasing in the violet and ultra-violet portion of the spectrum. Effect of White Light on Growth.—To determine the differential action of light and darkness on EF. capsulatus, series of plates were e (| Fig. 7. Explanation in text. covered with opaque black paper of the sort used in covering photographic plates, as illustrated in fig. 6, the portions with the crossing lines representing the covered areas. The plates were treated as follows: plate A was inoculated in the center and completely exposed to the light; plate B was covered as shown and was inoculated at a point about 1.5 em. from the rim of the plate, or about 5 em. from the uncovered area. "This plate was used to determine whether the organism would grow towards the [Vor. 20 522 ANNALS OF THE MISSOURI BOTANICAL GARDEN light or not. The plates covered as in C were inoculated in the center of the portions covered, and plate D was inoculated as in plate A and served as the complete darkness plate. Nutrient agar was used. ‘These plates were placed in a light-proof con- tainer illustrated in fig. 7, at room temperature. A General Electric mazda bulb of 50 watts and 115 volts was used as a source of light. A flow of water was kept constant beneath the source of light to prevent the heat from reaching the interior of the container. A sheet of colorless glass (A) was kept between the flowing water and the cultures. Several cultures of each series were kept in the container and the experiment was per- formed on several different occasions. Plates prepared as above were kept at room temperature and were subject to diffuse day- light and darkness. Growth at the end of 24 days showed no cultural or microscopical differences in any case, except for a slight increase in diameter of three plates kept in the freely circulating air. This small number of attempts seemed to indicate that light had no effect on growth. Effect of Red Light on Growth.—It has been pointed out that red light stimulated growth in higher plants but, according to Rahn, had no effect on lower organisms. Shrewsbury found that when cultures of Willia were placed at a distance of 4 inches from a Wratten safelight No. 1 screen, illuminated by a 60-watt gas-filled lamp, with the temperature of the dark room ranging from 23 to 28° C., growth was more rapid and more luxuriant than in the dark or diffuse daylight, as expressed by a hastening of spore formation. He attributed this increased activity to a reflex of increased growth energy, with red light acting as a stimulant. However, he adds that much the same growth was obtained in cultures grown at 30° C. as in those grown at room temperature, so that it was very likely that the heat and not the red light was the activating agent. To avoid this condition, the light-proof container as described above (fig. 7) was again used. This time a red bulb of the same intensity as before was used and a plate of red glass replaced the colorless glass. The same medium of the same pH was used here too. Cultures were grown at room temperature in the dark and diffuse daylight as controls. After 27 days it was found that 1933] MOORE—ENDOMYCES CAPSULATUS 523 those cultures under red light displayed no cultural or micro- scopic differences from those grown in the dark and daylight, and showed no change from those in white light and total darkness, except for a small increase in diameter which may be ascribed to the extra time of 3 days. REACTION TO DYES The róle of certain dyes in medical work has received much critical and experimental attention for a number of years. Their application for therapeutic purposes has resulted in clinical cures in various cases and has had no effect in others. Particular dyes, as methylene blue, gentian violet, and crystal violet, have been used locally or intravenously for skin lesions in cases of blasto- mycosis, coccidioidal granuloma, sporotrichosis, monilial infec- tions, and other diseases both of mycological and bacteriological etiology. Some of these have yielded favorable results as mentioned, but the value of the dyes applied is empirical since dye therapy rests on no definite scientific basis, according to Spring (’29). Faber and Dickey (’25) treated 15 infants afflicted with thrush, with local applications of a 1 per cent aqueous solution of gentian violet and found that 50 per cent showed an apparent cure in one day or less; 36 per cent in from 2 to 3 days; and 14 per cent in from 4 to 5 days. In several cases, however, the lesions returned at various intervals, and again some dis- appeared after one application while others resisted for several applications. In general, the lesions seemed to become smaller after continuous applications and to disappear. As a result, Faber and Dickey advocate a trial of the dye for prophylactic measures and also for therapy. Churchman (’20d), working on chemotherapy with gentian violet in vitro, advised careful considerations of the amount of bacteria affected, before any conclusion as to the therapeutic value of the dye be evaluated. Sanderson and Smith (’27) postulated the possibility of the utilization of gentian violet dye for the treatment of blastomy- cosis. It has been known and recorded since the days of Koch that certain aniline dyes have a ''bacteriostatic" effect on the cultur- ability of various bacteria. The viability of the organisms is [Vor. 20 524 ANNALS OF THE MISSOURI BOTANICAL GARDEN diminished to such an extent that they lose their power to multi- ply. This selective action of the dyes is very remarkable, certain organisms growing very abundantly and luxuriously on a medium containing gentian violet in a rather high concentration, whereas other organisms which are very resistant to harmful factors do not grow at all. This high selectivity or specificity was attributed not only to gentian violet, but also to others, as erystal violet and various members of the tri-phenyl methane series of dyes. In early work on dyes, smears and broth cultures were used, but it was found that the bacteria ceased growing or were killed in the smears and that various concentrations of dye in the broth had practically the same effect. Later work (Churchman, 712) was performed on the divided agar plate to determine whether the reaction would be the same. This method consists of in- serting a metal strip (Halsted’s aneurysm metal) across the diameter of the petri dish and then pouring the agar on one side and allowing it to cool. The agar containing the dye is poured on the other side ofthe strip. When theagar hardens, the metal strip is lifted out with sterile forceps. Thus one half of the plate con- tains plain agar and the other half the dye-containing agar. "This method has been used by several investigators, including Sander- son and Smith. Churchman (’12,’ 20, '20a, '20b, ’20c, '20d, '20e, '21) found that & parallelism with the reaction of the Gram stain could be made. That is, considering as violet-positive those organisms which were inhibited in growth by a certain concentration of a dye, gentian violet for instance, and as violet negative those upon whose growth the dye had no effect, then it was found that a great number of Gram negative organisms were also violet negative. This is a general rule, with a few outstanding organisms being exceptions. The same author also found that the selective action of the dye may be expressed in one of two ways. The dye may be toxic to the violet-positive organisms when a direct application is made, or a very strong inhibitory action is expressed when it is incorporated in the medium. On the other hand, the dye is not toxie to violet-negative organisms, and further, no inhibitory 1933] MOORE—ENDOMYCES CAPSULATUS 52b action is expressed except in very strong concentrations when it is incorporated in the medium. His experimental evidence pointed to a dilution of 1 : 1,000,000 as capable of stopping growth, and 1: 2,000,000 as possibly retarding it. This work was confirmed by Krumwiede and Pratt (13), who used dahlia agar, and also they confirmed the belief that the action was quantitative. They further observed that several dyes, in addition to those already mentioned, had this specific action. This work then called forth several applications of this principle to culture media, as that of Petroff (15) which was used by that author to isolate tubercle bacilli from the sputum and feces, and that of Farley (’20) as a restrainer in the isolation of pathogenic molds. Farley found that a dilution of 1:500,000 inhibited the growth of Gram positive bacteria. Later work on this problem showed that there might exist two types of organisms within a single strain, one which might grow vigorously on gentian violet media, and the other not at all, yet both showing the same stain and cultural characteristics. Also, heavy and repeated inoculations of violet-positive organisms would give fair and even good growth, and on this point Church- man (’20e) made the following statement: “This would indicate that bacteria do not, as is commonly supposed, act as isolated individuals; they possess the power, in numbers, of accomplishing effects which, alone, they are incapable of. The nature of this community of action it is at present impossible to guess at.” The same author (Churchman, ’21) placed dead bacterial bodies between living bacteria and gentian violet media and found that the Gram positive organisms grew, the dye having no effect. He attributed this occurrence either to a filtration or a stimulation of growth. The same year Churchman and Kahn (’21) found that a number of cells could accomplish more than a single cell, hence the communal action belief was further emphasized. Work on dyes was not confined to bacteria alone, for Lewis (’30) found that a certain number of dyes failed to inactivate the virus of chicken tumor while many had a harmful effect in certain concen- trations. Several papers have been published showing the action of certain of these dyes on various fungi, but only in a qualitative way; Greenbaum and Klauder (22) tested the action of gentian violet [Vor. 20 526 ANNALS OF THE MISSOURI BOTANICAL GARDEN as advocated by Farley (20a); Sanderson and Smith (27) found that dilutions up to 1:500,000 inhibited E. dermatitidis in its yeast-like form; Clark (727) showed that no growth occurred in a yeast-like organism and its mycelial form in dilutions from 1: 100 to 1 : 25,000 and that it gradually increased from 1 : 50,000 to 1:1,000,000, which gave good growth. No quantitative determinations of the colony growth were made. Stearn and Stearn (29) found that several fungi, including E. dermatitidis, were inhibited at various concentrations of gentian violet, with the organism being affected at 1 : 500,000. The later work of Churchman indicated that the ‘“bacterio- static properties” of dyes ascribe to them the ability to inter- fere with the reproductive mechanism, which he terms “genesis- tasis," without killing or in any way interfering with their other properties. This same phenomenon has been found bot qualitatively and quantitatively to hold for E. capsulatus. Since the author-could find that no particularly quantitative re- sults had been obtained for ascomycetous fungi, at least for the hyphal form, it was decided to investigate the problem of the dye re- action. For this purpose, several dyes, picked at random, were used. ‘These are as follows: Dyes of the nitro, - "— oxyquinone groups. Dyes of the azo orange G acid (N TT Aniline and Chemieal Co.) Sudan III—Nweakly acid—(E. Merck) poids d = oxyquinone group: aliz —acid—(E. Merck) The quinone-imide dyes. The thiazins: mn blue—basic— (Coleman and Bell Co.) The a Kaioak or eurhodin hakai — Psi dias and Bell Co.) Safra aie A—basic— (Coleman and Bell Co.) The phenyl-methane dyes. Tri-phenyl methane spurte Di-amino tri-phenyl methan light green S. F. Ee A E and Bell Co.) Tri-amino tri-phenyl methanes (rosanilins); basic fuchsin—basic— (Coleman and Bell Co.) crystal violet—basic— (Coleman and Bell Co.) aniline blue—acid—(National Aniline and Chemical Co.) 1933] MOORE—ENDOMYCES CAPSULATUS 527 The xanthene dyes. Fluorane derivatives: eosine B—acid—(Coleman and Bell Co.) Phenolphthalein and the sulphonthaleins: phenol red or phenol-sulphonphthalein—acid—(Dr. 'T. Schuchardt, G. m. b. H. Chemische Fabrik) The natural dyes Brazilin and haematoxylin: haematoxylin (Coleman and Bell Co.) These dyes were incorporated in the media, nutrient agar (produet of Digestive Ferments Co.) at a pH of 6.8, in the following concentrations: .00001; .000025; .00005; .000075; .0001; .00025; .0005. Several plates were used for each concen- tration of the dye, but only three plates for each are recorded in the tables for reasons stated previously. These plates were inoculated in the same manner as were those in the temperature and pH experiments, a broth of pH 6.8 being used here. The plates were kept at room temperature, which showed slight changes from day to day but not enough to affect the results seriously. The diameters of the colonies were measured daily at approximately the same hour each day. Inasmuch as the colonies may show slight irregularities in peripheral growth, these measurements may not be considered absolutely correct. However, if the shape of the colony was irregular, several diam- eters were measured and the mean taken as representative of that colony. Where contaminations occurred, as has been explained previously, the plates were discarded. It is to be noted that these experiments were carried out on several different occasions and that the figures in the tables denote the mean of a representative series. Since preliminary experiments with these dyes had given some seemingly strange results, microscopic examinations were made daily, in the case of the crystal violet and methylene blue series, and then, some time after growth had started, in the case of eosine B. Several of the dyes produced no effect on the growth of the organism. These dyes were orange G, Sudan III, alizarin (table 1v), safranine A, light green, basic fuchsin, aniline blue, phenol red, and haematoxylin. Neutral red was also used, but, as was expected, the results showed no difference from the normal control, the same being true for the other dyes. [Vor. 20 528 ANNALS OF THE MISSOURI BOTANICAL GARDEN If normal conditions are maintained for the plates, that is, if there is no drying out and no change in temperature the colonies grow at a rate of speed which is fairly constant as long as there is sufficient nutrient material for the organism. However, when the plates are kept in an atmosphere which allows drying out of the agar and when the amount of nutrient material is relatively small for the particular organism, the colony will show a decrease Fig. 8. Mycelium on .00025 per cent crystal violet agar. in growth rate. Thus, in the former case, the growth curve would be a straight line for a considerable portion, depending of course on the rapidity of growth of the organism compared from day to day, which in this case was sufficient to cover the plate almost completely before drying of the substrate. An initial lag on artificial substrates, as agar, must be taken into consideration. In that case the curve would tend to be sigmoid, that is, there would be an initial lag due to an acclimatization of the organism to the medium, a portion of increased activity which would be a sharp rise, and then a portion of decreased activity due to factors mentioned previously. The data of the growth on the several non-affecting dyes 529 .0005 0.4 «o oo N =) xn i er 5 2- . 00025 . 0001 6 6.7 .000075 5 6.6 TABLE IV Percentage concentration of dye .00005 . 000025 MOORE—ENDOMYCES CAPSULATUS TEMPERATURE OF APPROXIMATELY 22? C. VARIOUS CONCENTRATIONS OF ALIZARIN, AT A Nutrient agar (product of Digestive Ferments Co.) pH 6.8 .00001 6 6.6 MEAN DIAMETERS IN CENTIMETERS OF COLONIES GROWN IN tion 1933] inocula- [Vor. 20 530 ANNALS OF THE MISSOURI BOTANICAL GARDEN d E E E A o m " E E 8 5 2 S : B " g - > ©. E “4 50 c * g S 1 T 2 R : & E [^] eB g e § v J m = o t 3 g & A " 4 E o + = g 3 o g oO + 3 * g 3 + & te o " à M m * "m ~ e" eo *uo uj S3[uoi[oo jo suszouerq bb EA showed that the rate was practically the same in each case, and if the diameters of the colonies in em. were plotted against the time in days, for each concentration of the dye, one would find this to be the case. This is illustrated in fig. 9, the growth rate for 1933] MOORE—ENDOMYCES CAPSULATUS 531 alizarin, results representative of these dyes. The differences in the curves are within the range of error, so that in general we may .9005 000075 +0001 Percentage concentration of dye Oa .00005 -00001 Fig. 10. Maximum growth at constant temperature and same pH on nutrient agar with various concentrations + ™ *uo uj sejuo[oo Jo StozPouetd say that the growth rate is the same. Except for the initial lag and a slight curve due perhaps to utilization of nutrient materials of alizarin. [Vor. 20 532 ANNALS OF THE MISSOURI BOTANICAL GARDEN or drying of the agar, the line is about straight. When the maximum points of growth for each concentration is plotted, that is, diameter of colony against percentage concentration of dye (fig. 10), it is found that a straight line results, with perhaps some excusable error. In other words, growth is identical for all the concentrations of dye, at least in the observations made here. In dealing with the other dyes, as crystal violet (table v), methylene blue (table vr), and eosine B (table viz), a striking difference in growth was observed. The cultures of crystal violet grown under the same conditions as those above presented a growth which was proportional, within a range of limits, to the concentration of the dye. Like results were found to be true for methylene blue, but not quite so well marked. Colony measurements were made daily, and, as may be seen in table v, the growth was inhibited in the higher concentrations to the point of complete cessation. Along with the measure- ments, microscopic examinations were also made, and it was found that on the media of higher concentration of dye there was a corresponding decrease in production of reproductive parts. In other words, the number of asci with the ascospores gradually diminished on the increasing concentration of dye media as compared with the normal growth. A well-marked feature was the development of thick walls in the hyphae, as may be seen in fig. 8 which illustrates variously thick-walled swellings, chlamydo- spores and hyphae, taken from a culture of .00025 concentration, forty days after inoculation. The cellular material stained heavily, and showed granular cytoplasm and a reduction in the number of nuclear divisions. An increase in the number of swellings, cellular contents, and thickness of walls, as well as the number of chlamydospores and terminal hypnospores, is a condition which is usually associated with unfavorable phenom- ena. Similar characteristics were present approximately three months later. Macroscopically, the colonies on the higher concentration of crystal violet showed a growth upwards on the inoculum and not until 15 days after inoculation did any growth appear on the agar and then very little. In the case of methylene blue, however, growth was flat and took on a blue color very soon after inoculation. 1933] 533 MOORE—ENDOMYCES CAPSULATUS TABLE V MEAN DIAMETERS IN CENTIMETERS OF COLONIES GROWN IN VARIOUS CONCENTRATIONS OF CRYSTAL VIOLET, AT A TEMPERATURE OF APPROXIMATELY 22? C. Nutrient agar (product of Digestive Ferments Co.) pH 6.8 Percentage concentration of dye . 0005 eo et ES eo OD c» co e$ OD OD ~H i A M )w vir Rise id . 1*5 00o0o0o0o0ooooooooo e Co oo oc O OOO O O OQO Lm] e *H BOBO oH wB H a6 ooo o iy p. rp. © 00 CO 2 e eoo occ OOOO oo oo © @ ha S © o00 oo © CG --2 — & eo OO OO OO OO Ke m n Te] S eo C» c9 =H oo © ort by 00 OQ»: OQ» OQ» O Q r4 O1 «t i Or-AAOAOCr Oo «f 2 D D O © DOO OOOO SOC CH HH HHH S aa aa an aa EN EN | NN a S © OOO O m m - —- N AN ANNMAN CÓ CÓ SH H H BO H oH 10 10 — e Co x» rr NNOD - 09150 «D O0 OC» r4 C$ «M (D b. C» OQ Q HAW A n <+ «o o OOO m m on NN NAN OD C9 OD Hi oo © c coc - e cm om om ST 00000 [e] . eo o 2 |a * = eo co co co OD e N MaN ob cr-couoOncuoowococdua S S ooo c e — — tien! — ON C OD cO cO (CO H H oH 5 e - ES S | one a wD Omn oo C C co «M «B r- OO» — 01 00 «as Z|occc e = ANN e d d fæ 4 o. 9 & Q O N c o R ; bi V 4 HH A . : n 1 Jog o zo B Q Q odd oO E E LJ " Y à Q "3 " [| b Q O5 d s "n " tr Md Jy 4 e 3 3 Q Q OQ 4 | =] b à |o 5 Y o wm" 2 it e o Nf I] a AM E Lhe 3 A. & A S O99 Q + 34 le 3 ed = E] wo + D»] N = = ‘wo UT so]uoioo jo sLelewetd = author for eosine-methylene-blue agar where the mycelium had taken on a pink to blue color and ceased growing. ‘These cultures when transplanted to fresh agar grew just as luxuriantly as [Vor. 20 538 ANNALS OF THE MISSOURI BOTANICAL GARDEN though they had never been affected. What may have happened is that the dye, being acid, may have been absorbed by the fungus, .o00075 +0001 -90025 .0005 Percentage concentration of dye .00005 .000025 -00001 1 Y) N *uo ug Sojuo[oo Jo 3434əƏwLe ad Fig. 12. Maximum growth at constant temperature and same pH on nutrient agar with various concentrations o crystal violet. as evidenced by the color of the mycelium, and was strong enough to counteract the metabolic activities of the cell. It would seem 1933] ion MOORE—ENDOMYCES CAPSULATUS ouo[4q3our JO suorjurjuoouoo SNOBA qjr« 1e29 juotrjnu uo Hd eures pus o1njwroduro] 31uvjsuoo 49 q14013 jo NLY “ET “BI "eni — 8z əz tz zZz o2 SI 9l Yi rad or Q 9 * zZ $0000° $20900* TO De wwe SR e CN ae em > G SSS SO o y O - o + rz ‘wo UT Se[uo[oo Jo sileQeuvIT O Ko from a study of the pH of the medium, which changes from a slightly acid to alkaline in normal growth, that this could be detected. However, the medium was already covered with [Vor. 20 540 ANNALS OF THE MISSOURI BOTANICAL GARDEN «000075 00025 Percentage concentration of dye 000025 uw t re) N *"u0 UT Ssojuoioo Jo S4ə}əwvyd Fig. 14. Maximum growth at constant temperature and same pH on nutrient agar with various concentrations of sufficient growth to cause an alkaline condition, so that no direct evidence could be obtained from that direction. There are no definite facts to uphold this belief, and the statement should be considered more as a conjecture for the present. "There have methylene blue. 1933] c NN E MOORE—ENDOMYCES CAPSULATUS Time in days is "n M N *wo ug Sojuoloo jo s4Iəwerd been many theories proposed, but none have had any final acceptance. The action of eosine B on the organism is illustrated in figs. 15-16. The former shows the rate of growth of the cultures on Fig. 15. Rate of growth at constant temperature and same pH on nutrient agar with various concentrations of eosine B. [Vor. 20 542 ANNALS OF THE MISSOURI BOTANICAL GARDEN a o mn dá S + g 5 3 $ 8 9 S mM Ss S E g ^ > N Fel o * $ E g Bas fas] b0 e + 2 5 27 È E i 5 o o a & g 2 o 9 Hn gS u 5s. E I ? 2 z v g 3 g $ g A ga g Q + — =| e 1 JE E -| p E o ER b0 3 H $ d : B E os] = 5 — $ rz T n N - a0 "uo UT saTuo[oo Jo s.a3euvrtd [e various concentrations of dye. Growth on the concentration .0005 was inhibited on the nineteenth day, and that on the con- centration .00025, on the thirty-fourth day. Fig. 16 represents the maximal points of growth corrected for the inoculum. "The inhibition of growth is clearly distinguishable here. eosine B. 1933] MOORE—ENDOMYCES CAPSULATUS 543 A further phenomenon is noted here, not because it was abso- lutely specific for this experiment but because it had appeared most frequently on crystal violet media of concentrations .0001 and .00025. This is the formation of crystals in the media, appearing to emanate from the growing zone or peripheral region of the colony and found in conditions which have not been very favorable towards growth. They have occurred on beef extract agar of pH 7.2, 7.0, and 6.8. No crystals of this sort were found on media where there was no growth or where growth was very luxuriant. They appear to resemble very closely those of ammonium magnesium phosphate (NH,MgPQ,.6 H;O) as was reported also by Scudder (26). They were insoluble in cold water, completely and immediately soluble in hydrochloric acid, and were reprecipitated with sodium hydroxide, tests which are applied to crystals of the above composition. They were apparently the result of fungal activity, and Scudder has asso- ciated this condition with bacteria which “do not ferment the ordinary carbohydrates and intensify the alkalinity of the medium rapidly," a fact which has been clearly demonstrated in the carbohydrate reactions carried out in this paper. ANIMAL ÍNOCULATIONS From time to time papers appear in the literature pointing out that some particular organism which had been kept in culture over a long period and then inoculated in experimental animals, had retained its pathogenic properties. On the other hand, it is quite well known that this property may be lost on subculturing. It is also known that by growth on certain media at a specific temperature, toxicity may be retained over a long time. It has also been found that certain microorganisms may be made to regain lost toxie powers by repeated inoculations. In order to determine whether E. capsulatus had undergone any such changes since it was isolated in September, 1928, and kept in culture at different temperatures, on different media, and at various hydrogen-ion concentrations, several animals were inoculated with a twelve-day-old suspension of the organism in 10-ce. physiological saline solution. The culture was stirred up thoroughly with a sterile needle to eliminate the clumps. [Vor. 20 544 ANNALS OF THE MISSOURI BOTANICAL GARDEN The animals and inoculations were as follows: Rabbits: Two females—intracerebral (0.2 ec.) and intravenous (0.5 cc.). Two males—subcutaneous (2 ec.) and intratesticular (1 cc.). Guinea-pigs: Two, one male and one female—intracerebral (0.2 cc.). Two males—subcutaneous (1.5 ec.) and intratestic- ular (1 co.). Mice: Two, one male and one female—intracerebral (0.2 co.). One male—subcutaneous (1 cc.) and intratesticular (0.5 cc.). The results of these inoculations were that one of the rabbits receiving à subcutaneous and intratesticular inoculation died twenty-three days later. Autopsy revealed no infection or lesion, except an intestinal obstruction. "Three mice, healthy, with no loss of weight or appetite, were killed fifty days after inoculation. The organs were completely negative. Cultures of the heart’s blood also negative. The rabbits and guinea-pig were healthy, with no loss of weight, three months after inoculation. The remaining mouse, with an intracerebral injection, died sixty-two days after inoculation as a result of pneumonia. Au- topsy showed no complication of the other organs. Cultures of infected lung tissue and brain negative for E. capsulatus, the lung in a hemorrhagic condition. It may be concluded, therefore, that the organism had lost its power to infect after four years’ growth on artificial substrates. That it failed to produce a toxic action in an animal may be attributed to the changed environment and habitat of the fungus. It has been pointed out previously that the optimum temperature has been reduced from 37.5° C. to approximately 25° C. In addition, experimental data has shown that growth at the former temperature was very slow and slight, with complete cessation and death after a short time. It is very possible that the organ- ism was killed here, as in the in vitro experiments, as a result of the body temperature of the experimental animals. fructose> glucose > sucrose. A year later, Quednow (30) published his observations on a more extended list of sugars. He found the order of excel- lence to be glucose > fructose > sucrose > maltose > mannite > galactose > lactose. Smith (’32) used sucrose, glucose, and maltose, singly and in all sorts of combinations, and observed no apparent difference in the growth of the seedlings. Unusual facilities for research have enabled the writer to extend the list of sugars which has been used for orchid germination to include the rarer and more expensive forms. ‘This opportunity, together with the discrepancies between the work of La Garde, Quednow, and Smith, led to the present work. ‘The sugars were added to three different complete mineral nutrient solutions in amounts to give 7 gms. of carbon per liter. The compositions of these solutions were as follows: [Vor. 20 572 ANNALS OF THE MISSOURI BOTANICAL GARDEN Knudson's (722) Solution Shive's (15) Solution La Garde's ('29) Solution Mg80,.7H;0 .250 gm. MgSO,.7H;O 4.930 gm. MgSO,7H;O 1.00 gm. Ca(NOs3)2.4H20 1.000 gm. Ca(NOj,4H;O 1.228 gm. Ca(NOj,4H;0 1.00 gm. (NH480, .500 gm. KH;PO, 1.960 gm. KH;PO, 1.00 gm. K;HPO, .250 gm. CaCl, 1.00 m NH4NO; .50 (NH4)CO;. H:O .500 n Iron was added in all cases as 10 cc. of à M/200 suspension of FePO, prepared according to Livingston (19) in a liter of nutrient solution, Both Knudson and La Garde added iron in such quantities as to cause a heavy precipitate of iron phosphate. La Garde states that this precipitate was filtered off before the final sterilization and contained, besides iron and phosphate ions, calcium and potassium. It seems unwise to cause this bulky precipitate because it tends to adsorb other ions which are removed with it in filtration. The quantity of FePO, added by the author does not cause appreciable precipitation at the hydro- gen-ion concentration used. This reduced amount is undoubtedly sufficient in quantity, since it is ten times that originally rec- ommended by Livingston. La Garde designates the iron com- pound used by him as Fe;(PO4),.8H;,O. Knudson (722), Qued- now (730), and Smith (732) also added iron as ferrous phosphate. Since it is well known that the ferrous ion in the presence of oxygen reduces nitrate to nitrite, the ferric ion was used in the present work. The solutions were made up in liter flasks, and adjusted to pH 4.00 with HCl. All precipitate dissolved at this acidity, but there was a slight opalescence due to ferric phosphate. The solutions were then titrated by means of the quinhydrone elec- trode to such pH (see tables) that the values after sterilization were 4.8-5.1. Aliquots of 100-ec. portions were placed in 200-cc. Erlenmeyer flasks and 1.75 grams of Merck's Reagent Powdered Agar added. Sterilization was by autoclaving at twenty pounds pressure for twenty minutes. The medium was allowed to solidify in a slanting position. It is extraordinarily difficult to maintain sterile cultures in warm moist atmospheres over long periods of time, and after many preliminary failures, the following plan was adopted. "The solutions were added to the culture flasks through a funnel, care 1933] WYND—CARBOHYDRATE FOR ORCHID SEEDLINGS 573 being taken not to moisten the necks of the flasks. The flasks were then closed with cotton plugs. A duplicate set of cotton stoppers was carefully and tightly rolled, sterilized in empty flasks in the autoclave, and then immediately transferred to the dry-air oven until thoroughly dry. The seeds were shaken vigorously for thirty minutes in a small vial of calcium hypo- chlorite prepared as recommended by Wilson (15). "This vial was then clamped in a sloping position so that contaminating sub- stances might not fall in from the air. A culture flask and an empty flask containing the especially prepared cotton plug were held in a horizontal position in the left hand. The temporary cotton plug was withdrawn from the culture flask and dropped. A platinum-loop inoculating needle, held in the right hand, was quickly flamed and a loopful of seeds transferred directly from the hypochlorite solution to the drop of moisture that al- ways exudes from the solidified agar. The neck of the flask was then flamed and the especially prepared stopper quickly drawn from the blank flask and inserted in the culture flask. This procedure is advisable as it involves a minimum of exposure of the cotton plug that is finally used in the culture flask, and insures its perfect dryness. The drop of moisture containing the seeds on the edge of the agar was then distributed around the entire margin by rotating the flask carefully. This even distribution of seeds was main- tained by placing the flask in a rack so constructed that the agar surface was perfectly level, thus preventing the liquid drop from draining to one side and carrying the seeds with it. After a convenient number of flasks had been inoculated, the necks of the flasks were again flamed and the cotton plug well charred on the surface. The plugs and the outer surface of the necks were then moistened with saturated HgCl, solution. Heavy waxed paper was then dipped in the HgCl, solution, tightly wrapped around the stoppers and the upper part of the flask, and held firmly in position by rubber bands. The writer has found that unless these precautions are taken, fungi will frequently grow along the surface of the flask and penetrate tbe stopper, con- tamination usually not appearing until three or four months after inoculation. Bernard, certainly a well-trained and experienced [Vor. 20 574 ANNALS OF THE MISSOURI BOTANICAL GARDEN mycologist, has commented on the extraordinary difficulty of maintaining orchid cultures sterile in the moist warm atmosphere desirable for germination. By taking the above precautions, the writer has maintained sterile cultures as long as three years in moisture-saturated atmosphere at 25-35? C. In every case, the cultures were prepared in triplicate. "The seeds for the entire series were from a single pod of Cattleya Trianae Linden & Rchb. f. The flower was pollinated November TABLE I DATA OF KNUDSON'S SOLUTION pH adjusted pH at time of | pH after support- Sugar before steri- planting ing growth lization 8 months d-glucose.............. 4.12 5.1 4.2 d-fructose............. 4.12 4.9 3.8 d-galactose............ 4.12 5.0 No growth d-mannose............. 4.13 5.0 4.2 l-xylose. i ois dao 4.15 5.0 No growth l-arabinose . 4.10 5.0 No growth Maltose............... 4.15 5.1 4.4 l-rhamnose 4.15 5.0 No growth oo LL io intr 4.15 5.1 4.4 BD. 665.666 64048445 4.12 4.8 4.4 TABLE II DATA OF SHIVE’S SOLUTION pH adjusted pH at time of | pH after support- Sugar before steri- planting ing growth ization 8 months d-glucose. ............. 4.12 5.1 4.4 d-fructose............. 4.12 4.9 4.0 d-galactose............ 4.12 5.0 No growth d-mannose 4.13 5.0 4.5 Ixylose. .............. 4.15 5.0 No growth l-arabinose............ 4.10 5.0 No growth Maltose............... 4.15 5.1 4.6 ]-rhamnose 4.15 5.0 No growth NONE Lco ed dex, 4.15 5.1 4.5 Raffinose............-. 4.12 4.8 4.4 1933] WYND—CARBOHYDRATE FOR ORCHID SEEDLINGS 575 TABLE III DATA OF LA GARDE'S SOLUTION pH adjusted pH at time of | pH after support- Sugar before steri- planting ing growth lization 8 months OAEPUCON. Lco oa Lese 4.25 4.9 3.5 d-fructose............. 4.24 4.8 3.8 d-galactose............ 4.22 4.8 No growth d-miBnnose. er 4.24 4.9 3.8 LX WIGNN oco dor dai purs 4.24 4.9 No growth l-arabinose............ 4.24 4.9 No growth alto8e.. Ie eean tn eas 4.25 4.9 4.2 l-rhamnose. ........... 4.22 4.9 No growth IUCTOSO avi oe ees 4.24 4.9 4.1 Rafünose e o e: 4.51 4.9 4.2 27, 1930, and the matured pod harvested March 11, 1932, after a developmental period of over sixteen months. The inoculation of the culture media was made June 15, 1932, and the observa- tions were recorded February 25, 1933, after a growth period of about eight months. In the younger stages of development the diameter of the protocorm is an accurate basis for comparative determinations of growth, but after the seedling has developed leaves the growth is largely vertical rather than mere enlargement of the nearly round protocorm. Accurate measurements of the height of the young plant are difficult to obtain, since it is too large to measure by a microscope micrometer and too small for any less accurate means. For these reasons, seedlings as old as eight months can best be rated with the eye by comparing different culture flasks and sorting them into a few groups. The results of such a comparison after the growth period of eight months are shown in table rv. The relative excellence of the cultures is designated by the number of X's. 'The results show that d-mannose produced definitely the best growth. This is followed by the group glucose-maltose-fructose, and this in turn by the third group sucrose-raffinose, and then l-xylose. No growth was obtained on d-galactose, arabinose, or rhamnose. A comparison of these results with the molecular [Vor. 20 576 ANNALS OF THE MISSOURI BOTANICAL GARDEN TABLE IV GROWTH DATA , Kn Description Mineral Solution Mineral Solution Mineral Solution of plants Sugar |Growth| Sugar | Growth| Sugar | Growth Group I Exceptionally good growth — |d-mannose| xxxx* |d-mannose| xxxxx |d-mannose | xxxxx and chlorophyll development roup II Moderately d-glucose | xxx d-glucose | xxx good growth ^ |maltose XXX maltose XXX maltose XXX and chlorophyll |d-fructose | xxx d-fructose | xxx d-fructose | xxx development sucrose XXX Group III Poor growth ^ |sucrose XX sucrose XX an chlorophyll raffinose XX raffinose XX raffinose XX development l-xylose x l-xylose x Group IV d-galactose| 0 d-galactose| 0 d-galactose| 0 No growth l-arabinose | 0 l-arabinose | 0 l-arabinose | 0 l-rhamnose | 0 l-rhamnose | 0 l-rhamnose | 0 *The number of x's denotes the relative excellence of the cultures. configuration of the respective sugars indicates that in but one physical characteristic do they show any consistent physiological action—and that is that the pentoses do not allow the germination and growth of orchid seedlings. It is not impossible that their quality of being levo-rotatory is related to their physiological reaction. ‘The two instances in which l-xylose allowed growth are of doubtful authenticity. Only two or three seeds germinated in each flask, and these never developed beyond a very rudi- mentary protocorm. Such rudimentary development was oc- casionally noted even on sugar-free agar cultures, but in those cases also development never proceeded beyond a rudimentary protocorm. It is interesting to note that rhamnose, although it 1933] WYND— CARBOHYDRATE FOR ORCHID SEEDLINGS 577 has six carbon atoms, is structurally a methylated pentose and reacts physiologically to orchid seedlings as a pentose. Galac- tose, although an aldo-hexose, as is also d-mannose, supported no growth. These results are in approximate agreement with those of Quednow, cited above. From the following considerations the author believes that the conspicuous superiority of d-mannose is of especial signifi- cance. Mannose, in the form of mannan, is known to be widely distributed as a constituent of the cell wall of many plants. This is particularly true of seeds (Onslow, ’23). For example, various complex mannans have been found in the seeds of palms, aspara- gus, clover, coffee bean, onion, and various Leguminosae, Coni- ferae, and Umbelliferae. Mucilages are particularly rich in mannans, as, for instance, those obtained from lily bulbs (Parkin, 701) and tubers of various genera of Orchidaceae. Pringsheim and his coworkers (’24, '28) succeeded in isolating and studying mannan from orchid tubers. In this instance, it was water soluble and was precipitated by alcohol as a white powder. Klein (’32) gives a procedure for isolating mannan from salep itself. Salep, a product of orchid tubers, has been used as a substrate for the orchid fungi and for germinating orchid seeds in the presence of some symbiotic fungus from the earliest days of orchid research. It therefore appears fairly certain that the fungal element of the orchid mycorrhiza is able to hydrolyze mannan to soluble mannose. In this way the symbiotic fungi could make available any mannose which might be present in the woody and mossy substrate of epiphytic orchids in their natural habitat. The extraordinary slowness of the develop- ment of orchid seedlings would be supported by the presence of only very small quantities of mannan. The Missouri Botanical Garden has produced seedlings by the symbiotic method that surpassed in quality anything that the author has seen obtained by asymbiotic methods. In many instances the agar substrate consisted of the usual mineral substances and finely shredded cocoanut fiber. An appropriate fungus was inoculated on this medium some time before the seeds were sown. Seedlings grown on fungus-inoculated cocoanut- [Vor. 20 578 ANNALS OF THE MISSOURI BOTANICAL GARDEN fiber substrate always surpassed those grown on La Garde’s maltose media. Bultel (25) also found that symbiotic cultures gave superior results with all genera tested by him except Pha- laenopsis. The nature of his substrate was not indicated. The writer has already reported (33) on the superior value of the La Garde mineral solution over other solutions, and it seems scarcely probable that symbiotic cultures could owe their superiority to any inorganic constituent. It is much more probable that the carbohydrate relationship is the important feature. Knudson has shown the effectiveness of orchid fungi in hydrolyzing starch to available sugar. Might not the superiority of the symbiotic cultures be due to the mannose produced by hydrolysis from the cocoanut fiber? In order to determine the actual presence of mannans in cocoanut fiber, an analysis was carried out according to the method of Haegglund and Klingstedt (’24, '27). Ten gms. of cocoanut fiber were allowed to stand in 150 ce. of 72 per cent H.SO, for 244 days. The mixture was then diluted with water and the acid neutralized with CaCO;. The precipitate was filtered off, the residue washed on a suction filter, and the com- bined filtrate and washings were evaporated on the steam bath to 150 cc. Then HSO, was added to give a 2 per cent solution and the mixture boiled 2 hours. The acid was again neutralized with CaCO;, then weakly acidified with HC,H;O., and evaporated to. 100 ec. After cooling, 10 cc. of phenylhydrazine plus 20 ce. water were added and allowed to stand several days. A very definite precipitate of the insoluble phenylhydrazone was obtained, indicating the presence of mannose. The above method is scarcely quantitative because of the difficulty of obtaining complete hydrolysis of mannans without causing their oxidation at the same time. The very heavy precipitate of CaSO, which is produced when the acid is neutral- ized with CaCO; is bulky and difficult to wash thoroughly. The seedlings grown on La Garde’s maltose solution equalled those of the same age originally obtained by La Garde. Hence the mannose cultures might be regarded as distinctly superior to those yet obtained by the use of a purely synthetic medium and approach in quality the best of the symbiotic cultures. It 1933] WYND—CARBOHYDRATE FOR ORCHID SEEDLINGS 579 is to be regretted, however, that this could not be verified posi- tively by comparison, since seedlings of the same age grown symbiotically were not at hand. The glucose cultures were also of high quality, particularly as to their green color. No trace of yellowness or of inferior chlorophyll development was observed in seedlings grown on this sugar. It must be noted that superiority of symbiotic cultures may depend not only on a carbohydrate relation, but also on a favor- able degree of acidity. Knudson has pointed out that satis- factory fungal symbionts maintain a favorable pH of the media for the germination of the seeds. The pH of asymbiotic media can, of course, be adjusted artificially, but the initial favorable acidity is difficult to maintain over long periods of time. The author hopes that some worker equipped to carry both symbiotic and asymbiotic cultures simultaneously will test further the suggestion that symbiotic cultures owe their superi- ority to mannose and to the constantly favorable pH relation- ship. SUMMARY 1. The growth of orchid seedlings over a period of eight months was observed on a series of sugars, each added to three different inorganic media in amounts to give seven grams of carbon per liter. The order of excellence of growth on the different sugars was: d-mannose > d-glucose > maltose > d- fructose > sucrose > raffinose. 2. No growth was obtained on d-galactose, and the pentoses, larabinose, l-rhamnose, and l-xylose. This inability of the pentose sugars to support growth may be related to their levo- rotating property. Galactose occupies an anomalous position. 3. Mannose gave conspicuously the better growth. This may be related to the fact that symbiotic cultures containing cocoanut fiber, an effective source of mannan, produce seedlings superior to any that the author has seen produced asymbiotically. 4. The mineral nutrient medium of La Garde plus d-mannose is regarded as the best asymbiotic culture medium for orchids, as shown by the reaction of Cattleya Trianae Linden and Rchb. f. seedlings. [Vor. 20 580 ANNALS OF THE MISSOURI BOTANICAL GARDEN ACKNOWLEDGMENTS The author wishes to thank Dr. E. S. Reynolds, Physiologist to the Missouri Botanical Garden, for helpful suggestions through- out the work, and Dr. G. T. Moore, Director of the Missouri Botanical Garden, for the supply of orchid seeds. BIBLIOGRAPHY Ballion, G., and Ballion, M. (24). The non-symbiotie germination of orchid seed in Belgium. Orchid Rev. 32: 305-309. 24. ——,——— —,(328). Asymbiotie germination of orchid seed. Ibid. 36: 103- 2. 1928. Bernard, N. (00). Sur quelques germinations difficiles. Rev. Gén. Bot. 12: 108- 120. 1900. —— — ——, (02). Études sur la tubérisation. Ibid. 14: 5-25, 58-71. 1902. — — ——, (08). La germination des orchidées. Comp. Rend. Acad. Sci. Paris 137: 483- 485. 1903. — — ——, (04). Recherches expérimentales sur les orchidées. Rev. Gén. Bot. 16: 405-451, 458-476. 1904. '00). Symbiosis d’orchidées et de eia champignons endophytes. Com ips Rend. Acad. Sci. Paris 142: 52-54. 1906. Bultel, G. (24-25). Germinations aseptiques Fac Cultures symbiotique et asymbiotique. Rev. Hort. Paris 96: 208-271. 1924 ; 97: 318-321, 359-363. 1925. — — ——,(26). Lesorchidées germé hampig t des plantes normales. Ibid., 98:155. 1926. Burgeff, H. ('09). Die Wurzelpilze der Orchideen, ihre Kultur und ihr Leben in der Pflanze. Jena, 1 — E. (24a). Germination of Mm and other seed without fungal Orchid Rev. 32: 233-238. 192 — — ——, (24b). The non-symbiotic ue of orchid seeds. bid. 359-365. — — ——, (26). The non-symbiotic and symbiotic germination of orchid seeds. Ibid. 34: 165-169. 1926. — ———, (29). Non-symbiotic and symbiotic germination of orchid seed. Ibid. 37: 68-75. 1929. — ————, (32). Raising orchid seedlings. Ibid. 40: 195-206. Costantin, J. ('26). La vie asymbiotique des orchidées. Ann. ‘ai “Nat. Bot. 8: TK = H. (55). Recherches sur les tubercules de l'Himantoglossum hircinum. id. IV. 3: 253-291. 55. — — ——, (56). Dela germination des ophrydées et de la nature de leurs tubercules. Ibid. IV. 5: 163-186. 1856. Haegglund, E., und Klingstedt, F. W. (24). Cellulose-Chemie 5: 58. 1924. ————-, ——, (27). Zur Charakterisierung von Celluloseprüparaten mittels der Pri eina Ann. der Chem. 459: 26-38. 27. Klein, G Ar ). Organische Stoffe. Membranstoffe. Handb. d. Pflanzenanalyse 21: 48. 1932. 1933] WYND—CARBOHYDRATE FOR ORCHID SEEDLINGS 581 Knudson, L. (16). Influence of certain carbohydrates on green plants. Cornell Agr. Exo. Sta. Mem. 9: 1-75. Non symbiotic suininétion of orchid seeds. Bot. Gaz. 73: 1-25. a~ Ta Y NI MU Mee — o . Further observations on non-symbiotie germination of orchid seeds. Ibid. 77: 212-220. 1924. , 25). Physiological study of the symbiotic germination of orchid seeds. , (26). Pissiclogisul investigations on orchid seed germination. Inter- nat. Cong. Pl. Sci. Ithaca, Proc. 1183-1189. ; —, (27). Symbiosis and asymbiosis relative to orchids. New Phytol. 26: 328-336. 1927. La Garde, R. V. (29). Pi Rod germination of orchids. Ann. Mo. Bot. Gard. 16: 499-514. Livingston, B. E. (19). A a for cooperative research on the salt requirements of representative agricultural plants. 1-54. 2nd ed. Baltimore, 19 Onslow, Muriel Wheldale (’23). Practical plant biochemistry. Camibridge, 1923. Parkin, J. (01). On a reserve carbohydrate which produces mannose, from the bulb of Lilium. Cambridge Phil. Soc., Proc. 11: 139-142. 01. Prillieux, E. (56). De la structure anatomique et du mode de végétation du Neottia nidus avis. Ann. Sci. Nat. Bot. IV. 5: 267-282. (60). Observations sur la germination du Miltonia spectabilis et de mud: autres orchidées. bid. 13: 288-296. 1860. , et Rivière, A. (56). Observations sur la germination et le développement d'une orchidée. Ibid. 5: 119-136. 1856. Pringsheim, H., und Genin, A. (24). Uber die fermentative Spaltung des Salep- ns. VI. Mitteilung über Hemicellulosen. Zeitschr. f. physiol. Chem. 140: 299-304. 1924. — — ——., und Liss, G. (27). Über das Salep-Mannan. Liebig’s Ann. der Chem. 460: 32-42. 1927 Quednow, K. G. (32). Beitrüge zur Frage der Aufnahme rd Kohlenstoffver- bindungen und andere Pflanzen. Bot. Archiv. 30: 51-108. 1930. Ramsbottom, J. (22). The germination of orchid seed. Orchid Rev. 30: 197-202. 1922. Shive, J. W. (15). A three salt nutrient solution for plants. Am. Jour. Bot. 2: 157-160. 1915. Smith, F. E. V. (82). i: Meer seedlings asymbiotically under tropical conditions. Gard. Chron 932 ares W. (’86). Batas. zur v h dn Orchideenwurzelpilze. Bot. Zeit. 1886. 81-488, 497—505. "d $ K. (15). Calcium hypochlorite as a seed sterilizer. Am. Jour. Bot. 2: 420—427. 1915. Wynd, F. L. (33). Nutrient solutions for orchids. Ann. Mo. Bot. Gard. 20: 363- 372. 1933. SUPER OPTIMAL AND THERMAL DEATH TEMPERATURES OF THE COTTON PLANT AS AFFECTED BY VARIATIONS IN RELATIVE HUMIDITY: DOROTHY MEGOWEN BERKLEY FormerlyJ essieR. Barr Fellow inthe HenryShaw School of Botany of Washington University AND EARL E. BERKLEY Research Fellow 4n the Henry Shaw School of Botany of Washington University? HISTORICAL REVIEW In 1863 Sachs reported the results of an attempt to determine the effects of high temperatures on the sensitivity of Mimosa pudica. 'Transitory insensitivity, he found, was caused by an exposure of one hour to a temperature of 40° C., and at 45° C. for a half hour and 49? C. for a very brief time the same effect was produced. When permanent insensitivity was attained, at higher temperatures, death invariably followed. Sachs also reported (64) on the effects of high temperatures on tobacco, pumpkin, corn, nasturtium, and rape, exposed for various periods of time. All tbe plants were able to withstand tem- peratures of 49-51? C., but none survived 51? C. for more than 10 minutes without injury. The power of resistance to high temperatures was found to vary at different ages. Developing leaves, stems, and roots were more easily killed than older ones. Ewart ('03) noted a decrease in the rate of protoplasmic streaming in Elodea, Tradescantia, Chara, Spirogyra, root hairs, pollen tubes, etc., depending upon the height of the temper- ature above the optimum and upon the length of exposure. Pfeffer (03) made the generalization that all turgid plants ultimately die when the temperature reaches from 1° to 2° C. above the maximum where the plant will grow indefinitely, although growth may continue for a time, and that at temper- atures of 10? C. above this maximum all flowering plants seem to be rapidly killed. He noted that plants, which at first 1 Portions of this work, together with certain microscopical studies of the treated plants, were submitted by Mrs. Dorothy Megowen Berkley as a thesis in partial fulfillment of the requirements for the degree of master of science in the Henry Shaw School of Botany of Washington University. ? A fellowship established by the American Creosoting Co. Ann. Mo. Bor. GARD., Vor. 20, 1933. (583) [Vor. 20 584 ANNALS OF THE MISSOURI BOTANICAL GARDEN appeared fresh and unharmed after a short exposure to fatally high temperatures, frequently died later as an after-effect, even under the best external conditions. Fung ('11) emphasized the necessity of considering the relative humidity in determining the effects of high temperature on plants. He found the maximum temperature for growth of cotton to be 113° F. in a relative humidity of 90 per cent and the optimum 85-90^ F. in a relative humidity of 70.6—72.2 per cent. Cotton plants treated for four hours in a saturated atmosphere at a variety of high temperatures also gave interesting results. At 42-45" C. the stems and leaves were badly wilted but re- covered; at 44-48? C. the stems and leaves were browned but new leaves appeared after one week (probably secondary growth from uninjured nodes); at 49—55? C. the plants were killed. The degree of injury was determined after the plants had been trans- ferred to ‘‘proper conditions," which were not described. The value of Fung's results is limited by the fact that he worked with an insignificant number of plants and used only very young seedlings. Bose (013), who gave 60? C. as the average fatal temperature for plants in general, found the death point to be lower in young plants, which confirmed the earlier statement of Sachs. Collander (24) determined the temperatures at which death occurred in individual cells of various plants. T'radescantia discolor was killed at 65? C. within an average of 1.8 minutes, Brassica oleracea at 60? C. within an average of 0.8 minute, Beta vulgaris at 60° C. in 0.7 minute, Draparnaldia glomerata at 55? C. in 0.32 minute, and Pisum sativum at 55? C. in 0.095 minute. He found that these plants could live at slightly lower temper- atures for some time, thus demonstrating that the thermal death point is suddenly reached. Lepeschkin (725) discussed the effects of optimum, maximum, and thermal death temperatures on bacteria, fungi, and higher plants. He stated that most plants died at 60-70? C. in one minute, although some died at 40-45? C. in that time. Gilbert (’26) found that cotton grew better at 80? F. in a relative humidity of 50 per cent than it did in a relative humidity of 85 per cent. Wallace (31) tested the effect of 1-24-hour 1933] BERKLEY AND BERKLEY— TEMPERATURE AND HUMIDITY 585 exposures to temperatures ranging from 15 to 60? C. on the sensi- tivity of Mimosa pudica. Above 45? C. injury or death resulted, depending upon the length of the exposure. He reported that the relative humidity had little or no effect upon the sensitivity of the plant. Baker (29), in studying the effects of excessively high tem- peratures on conifer seedlings 1-3 months old, also emphasized the suddenness at which the thermal death point was reached. The living tissues were quickly killed at 54° C. but were un- injured after à prolonged exposure to a temperature only a few degrees lower. Just below the thermal death point he noted a region of no growth where photosynthesis was apparently unable to keep up with catabolic changes, the chlorophyll decomposing faster than it was made and the leaves becoming yellow or withered. On prolonged exposure the plants died. He also found the age of the plant to be a factor in resistance, due to the development of protective tissues as the plant grew older. The indications were that there was no increased protoplasmic resistance with age. APPARATUS Four glass cases of identical design and size (60 in. long by 32 in. wide by 40 in. high) were used for the experimental work. Three of these cases were variously used for growing plants, for germinating seeds, and for plants under observation following treatment. Additional plants were grown on benches in the experimental room. The fourth case was converted into the electrically controlled temperature chamber (fig. 1). The experimental plants were placed in a flat of sand which rested on a lattice rack supported 12 inches above the bottom of the case. The heating units consisted of three heaters, two of which were controlled by ordinary switches, and were placed on the floor of the case, one in either half about midway between the center line and the end. They were composed of resistance wires wrapped on racks which extended practically the width of the case. With these two heaters there were three possible combinations: when the two were in parallel maximum heat was obtained, when one was used alone approximately one-half of the maximum could be had, and [Vor. 20 586 ANNALS OF THE MISSOURI BOTANICAL GARDEN when the two were in series one-fourth of the maximum. A further reduction in temperature was obtained by turning off these heaters altogether, leaving only the third heater which was composed of resistance wire wrapped on the guard-wires of an electric fan. This heater was controlled by a Thyratron! and relay combination which in turn was controlled by a thermostat with à mercury make-and-break contact. EM uu M ~ "4 Thyratron ms ec 7 " — F Ea eee e a ee | l Ay relay | » s ' C \ T M L — 3 N toluol / \ mercury H VLA Fig. 1. Diagrammatic cross-section through the center of the controlled tem- perature chamber in which the plants were treated. C, capillary; E, electric fan r The fan, which was located on the floor at one end of the chamber, was hooked directly to the current so that it ran continuously, blowing the warm air over the heaters and through ! Behmitt, F. O., and Schmitt, O. H. A. A vacuum tube method of temperature control. Science N. S. 73: 289-290. 1931. 1933] BERKLEY AND BERKLEY—TEMPERATURE AND HUMIDITY 587 humidifying cloths made of cheese-cloth. In this way the heat and moisture were uniformly distributed throughout the com- partment. Good ventilation was assured by boring small holes in the frame of the compartment on either side of and in back of the fan. The thermostat consisted of a glass tube extending the full length of the chamber, one end of which was shaped into a U terminating in a capillary. The main body of the tube had a capacity of about 200 cc. and was filled with toluol; the U tube, with a capacity of about 20 cc., contained mercury. The contact was made in the capillary in the usual manner. Due to the large capacity of this tube and the fact that it extended across the case at a level with the tops of the plants, it was possible to control the temperature within 0.5 of a degree. The temperature of the chamber was determined by means of an incubator thermometer inserted through the top of the case, with its bulb reaching down to the level of the plants in the center. For experiments run at high humidity, a practically saturated atmosphere was obtained by hanging from four pans at the top of the chamber strips of cloth which were kept supplied with water by tubes running from a tank above. These cloths drained into three pans below, one of which was placed in the flat of moistened sand and the others located on the floor. Addi- tional cloths, through which the warm air from the fan was blown, extended the entire width of the temperature chamber below the rack and were moistened by constant sprays of water. An overflow in the bottom pans, which received most of the run- off water, was avoided by the use of constant-level siphons which carried off the surplus through drains in the floor. Still other cloths were hung above either end of the pan in which the plants were placed, so that the pots of plants were not only standing in water but were more or less surrounded by moist cloths. Under these conditions the relative humidity was so high that the entire inner surface of the temperature chamber was covered with a film of water which dripped from the top and ran down the glass sides in streams. For the low-moisture experiments, all humidifying devices were removed from the chamber. [Vor. 20 588 ANNALS OF THE MISSOURI BOTANICAL GARDEN Since there is no method devised at the present time for measuring the relative humidity at temperatures above 50? C., it was decided to measure the evaporating power of the atmos- phere. This was done by the use of atmometers which consisted of two porous cups of the cylinder type connected to burettes on the outside of the chamber, all connections being made under water to eliminate air bubbles. These atmometers were care- fully filled with boiled water, and placed on either side of the flat on which the plants were to be set. Notched corks were placed in the top of each burette to prevent evaporation at that point. After the chamber had been adjusted to a desired temperature, readings were made on the burettes at regular intervals. ‘lhe results of these readings are shown in fig. 2. At the high relative humidity, after the humidifying apparatus had been thoroughly adjusted, there was very little evaporation from the atmometers. Even at the higher temperatures, with no plants in the temperature chamber, there was less than 0.001 cc. evaporation per minute. "These results give some indication of the conditions existing within the chamber, although it is not intended to imply that the plants lost water at exactly the same rate as the atmometers. At temperatures below 50? C. an approximately saturated relative humidity for the higher moisture experiments and an average of 69 per cent for the lower moisture experiments were calculated from the readings of wet and dry bulb thermometers. When readings were made with the plants in the chamber, the relative humidity varied with the time of day, being as low as 55 per cent at night when transpiration was cut to a minimum and as high as 78 per cent at mid-day. Accordingly, most of the plants at the lower humidity were treated during the day. Since transpiration raised the relative humidity of the chamber, care was taken, when watering the plants in preparation for treatment, to keep the foliage dry and to allow the water to soak into the soil before placing them in the temperature chamber. PROCEDURE These experiments were made in order to determine the effects of humidity of super optimal and thermal death temperatures on 1933] BERKLEY AND BERKLEY—TEMPERATURE AND HUMIDITY 589 the cotton plant. They were suggested by the previous work of Berkley (31) and the unpublished studies of Fung (11). Cotton plants (variety Upland Big Ball) were grown in pots containing a mixture of sand and loam in a room of the greenhouse having a usual temperature of 25-30? C. Since individual variations made it necessary to use large numbers of plants of various ages, seeds were planted at frequent intervals from September, 1931, until the following September. Two distinct series were run, one at a low, and the other at a high, relative humidity. The low-humidity experiments were made at temperatures between 42 and 84? C. inclusive, and those at the high humidity, between 40 and 65° C. Plants of various ages (their ages being computed from the day the seeds were planted) were exposed to a particular temper- ature and humidity combination. After the chamber had been adjusted to the desired temperature, the plants were quickly passed through one of the glass doors, the slight drop in temper- ature occasioned by their entrance being quickly adjusted. They were removed after treatment to a compartment at room temperature which had a relative humidity similar to that of the treating chamber. 'The subsequent behavior of all plants treated was watched and the results noted. A minimum of 24 hours was allowed to elapse before the plants were pronounced dead. When the leaves and the growing tip of a plant were killed, it was called “dead” but was kept watered for some days to allow secondary growth to take place at the nodes in case the entire plant had not been killed. Such plants as did put forth secondary growth are listed in the tables. It will be noted that this was characteristic of the higher humidity experiments only. SERIES I Low humidity—The plants treated at the lower relative humidity wilted to some extent before they were removed from the temperature chamber. When not too severely injured by excessive treatment, they revived and became turgid again im- mediately after removal. When treated for a longer time than was necessary to kill them, the wilted leaves and cotyledons temporarily regained their turgidity and to all appearances were [Vor. 20 590 ANNALS OF THE MISSOURI BOTANICAL GARDEN unharmed, but the petioles of the younger leaves, the stems just below the growing tips, and the hypocotyls of the seedlings were withered beyond recovery. After some hours, the time de- pending upon the intensity of the treatment, the injured portions of all plants, including those that temporarily recovered, dried up, only the cotyledons and leaves remaining green and turgid. These living organs, connected to the main stem of the plant only by dead tissue, remained in the green condition from 3 to 14 days before showing signs of death. A microscopic study revealed no living tissue in the withered portions of the petioles and hypocotyls, but in the leaves and growing tips, which had recovered, the tissue was found to contain a number of mitotic figures. In the meristematic regions of the plants treated at times and temperatures just sufficient to kill them, the cell contents were disorganized. Very few of the plants whose growing tips were killed put forth secondary growth at the nodes. Tables 1 and m and fig. 3a give the detailed results of the lower humidity experiments. The thermal death point for the older plants treated at the lower relative humidity is in the neighbor- hood of 63? C., as shown by a sudden break in the line when the temperature is plotted against the time (fig. 3a). The time required to kill all of the plants at this temperature was about 7 minutes. The evaporation rate of the atmometer at 63° C. was 0.26 cc. per minute (fig. 2 and table 11). If the plants were evaporating at a similar rate the cooling effect of the evaporation would be appreciable and would account for the greater length of time necessary to kill the plants at this temperature than at slightly higher temperatures. The seedlings 10 days old or less had a slightly lower and more variable critical temperature (table 1). The cotyledons and plumules showed about the same resistance as the older plants but the hypocotyls were permanently injured when treated for 8 minutes at 60° C. This injury to the hypocotyls ultimately caused the death of the entire plant. In the older plants the petioles of the succulent young leaves were similarly affected but the plants were not killed unless the treatment was prolonged until the younger portion of the stem 1933] BERKLEY AND BERKLEY— TEMPERATURE AND HUMIDITY 591 85 Temperature in Centigrade Degrees malo .15 .20 .25 30 35 .40 Cubic Centimeters of Water Evaporated per Minute Fig. 2. The number of cubic centimeters of water evaporated per minute from an atmometer, plotted against the temperature in Centigrade degrees. The deter- minations were made at 10° intervals. The figures in table 1 relating to the evap- oration rate were taken from this graph [Vor. 20 592 ANNALS OF THE MISSOURI BOTANICAL GARDEN and growing tip was injured. ‘Therefore, the lower critical temperature of the seedlings can be attributed to the lack of resistance of the hypocotyl and not to the injuring of meriste- matic regions. TABLE I COTTON SEEDLINGS TREATED AT THE LOWER RELATIVE HUMIDITY Temp. in ; Age of | Number | Number | Number | Evap. rate degrees C. | Time ants | of plants | living dead (cc. per days) min.)* 42 72 hr. 10 12 12 .14 45 45 hr. 24 6 .15 47 26 hr. 9 1 15 .16 2 hr. 10 1 12 .19 : 10 min. ; 3 32 21 J 15 min. 1 1 5 20 min. ; 1 4 6 45 min. ; 2 20 215 min. , 1 11 .22 5 min. > 1 10 7 V$ min. ) 1 10 d 10 min. 8 min. 4 1 39 22 7 20 min. 1 2 8 12 7 35 min. 6 66 min. 1 15 25 ( min. i 55 ( 5 min. ( .26 ( min. 1 17 .27 ( 2V$ min. 1 ) 6 9 ( 5 min. 1 29 29 ( 1 min. y 18 3 .28 68 1 min. ) 10 .29 68 2 min. 40 40 69 1 min. 10 2 .29 ) 2 min. ) ) 9 1 min. ) ) 1 5 .30 2 min. ) ) 4 2 7 min. ) 8 74 1 min 1( 2 8 32 74 1% min ) 2 21 5 1 min ) 4 4 .93 [7 1 min. ) 11 1 10 .94 18 1 min 6 6 .95 80 1 min. ] 11 3 8 .96 *This column shows the evaporation rate of water in cubic centimeters per minute from an atmometer, as described under “Apparatus,” for the temperatures indicated. In order to determine the cause of this differential resistance of the hypocotyls and cotyledons an experiment was made to find the amount of water lost during treatment and during various periods of time following treatment. This was done by deter- 1933] BERKLEY AND BERKLEY—TEMPERATURE AND HUMIDITY 593 TABLE II COTTON xin FROM 16 TO 180 DAYS OLD TREATED THE LOWER HUMIDITY Temp. in A Age of Number | Number | Number | Évap. rate degrees C. | Time plants | of plants | living dead (cc. per (days) . 42 24 hr. 25 € .14 42 48 hr. 16 16 12 6 42 72 hr. 53 € 2 45 2 hr. 76 € 6 .15 47 18 hr. 77 3€ 36 .16 47 26 hr. 1 11 3 47 26 hr. 7 6 1% hr. 1 10 2 18 i 21í hr. ‘ ] 3 9 ? 334 hr. [ 4 48 .19 6 hr. , ] 9 3 1 hr. 28 28 .20 hr. 12 E 30 min ) 19 3 16 .21 30 min l , 2 45 min ( 10 8 z hr. , 18 18 ( 3 min 1 4 7 24 ) 15 min. 25 5 20 ) 20 min. 27 2 25 ) 25 min. 55 55 ( min. 16 4 12 .25 ( 20 min. 128 128 ( | min 7 3 4 .25 ( ; min 10 2 8 ( 10 min 5 50 63 ) min 1 2 12 .26 ( / min. 1 18 ( 5 min. 8 82 .26 ( 1 min. 4 8 197 2 min. 40 4 4 min. 4 11 1138 3 min. 40 T rt .29 2 min 70 16 2 1 .90 2V$ min. € 20 2 3 min. ^ 1 1 E 1 min. E 1 6 Ax 74 1 min ' T 2 : /4 14% min 4 1 1 10 4 2 min 4 10 10 1 min 4 1 2 ) .94 J 1 min ^ 3 ] 1 min 120 2 ; 1% min 40 2 2t ) 1 min d 3 2 ( .96 ) 1 min í 10 5 i ) 1% min ( 32 3: 4 1 min ' 3 2 6 .97 t l6 min f f 5 2 .88 | 1 min. 7 11 1 10 t 114 min. 30 10 10 *This column shows the evaporation rate of water in cubic centimeters per minute from an atmometer, as described under “Apparatus,” for the temperatures indicated. [Vor. 20 594 ANNALS OF THE MISSOURI BOTANICAL GARDEN mining the average percentage of moisture in the hypocotyls and cotyledons of treated and untreated seedlings (table rr). The treated seedlings had been exposed for 15 minutes to a temperature of 60? C., a treatment known to leave the cotyledons of most of the seedlings in a living condition for a week or ten days. In all, five sets were run: in one, which was used as a control, the plants were weighed before treatment, in one they were weighed immediately after treatment, in one, after 2 hours, in one, after 24 hours, and in another, 96 hours after treatment. Before weighing, the seedlings were cut about one-half inch above the surface of the soil and again immediately below the junction of the cotyledons and the hypocotyls. The hypocotyls and cotyledons were weighed separately, dried, and reweighed. The results (table 111) show that the hypocotyls and cotyledons had approximately the same water content before treatment, but the cotyledons lost considerably more during the treatment (3.3 per cent loss in the cotyledons as compared with 0.2 per cent in the hypocotyls). During the first 2 hours the cotyledons still showed a slightly more rapid rate of loss than the hypocotyls, but after 24 hours they had regained more than 3 per cent, while the hypocotyls still continued to lose. After 96 hours the hypocotyls were very much withered, having an average of only 78.3 per cent moisture, while the cotyledons still retained an average of 87.4 per cent and were more or less turgid. TABLE III LOSS OF MOISTURE IN SEEDLINGS DURING AND AFTER TREATMENT AT 60° C. FOR 15 MINUTES AT LOW RELATIVE HUMIDITY oo of Per cent loss Per cent loss l of moisture in | of moisture in plants hypocotyls cotyledons 400 po BE. osos RERERI 92.2 92.1 200 Treated & weighed immediately. . 92.0 88.8 200 Weighed 2 hours after treatment. 91.1 87.7 200 Weighed 24 hours after treatment 88.2 90.8 200 Weighed 96 hours after treatment 78.3 87.4 1933 BERKLEY AND BERKLEY— TEMPERATURE AND HUMIDITY 595 SERIES II High humidity.—Under the high humidity conditions also the seedlings showed a somewhat lower and more variable critical temperature than did the older plants. When first removed from the temperature chamber, neither the seedlings nor the RS o2 Temperature in Centigrade Degrees L L 1 ie] lo 20 30 40 50 60 TO Time in Minutes Fig.3. The time in minutes necessary to kill all plants, plotted against the temperature in Centigrade degrees. The line (a) shows the results of the lower relative humidity experiments and (b) the results of the higher. Note the sharp break in (a). The lines in this graph are not drawn through any definite set of points but are plotted from tables m and v and are based upon the results obtained from the treatment of more than 4000 plants. [Vor. 20 596 ANNALS OF THE MISSOURI BOTANICAL GARDEN older plants showed any evidences of injury unless they had been left in the chamber for a period of time considerably longer than was necessary to kill them. In such a case, the young leaves and cotyledons appeared as if scalded. In the seedlings, injurious effects were later evidenced by the wilting or withering of the TABLE IV aiii p iege TREATED AT APPROXIMATELY ER CENT RELATIVE HUMIDITY Temp. in ] Age of Number Number Number degrees C. Time plants of plants living dead (days) 41 24 hr. 12 22 4 18 41 48 hr. 13 22 22 42 9 hr. 10 6 6 45 2 hr. 10 6 6 50 30 min. 14 12 2 10 50 45 min. 14 21 4 17 50 60 min 14 24 8 16 52 10 min 8 11 9 53 5 min 15 6 6 53 10 min 9 6 6 54 3 mi 8 21 20 1 54 5 min 8 15 6 9 55 5 min 8 48 25 23 55 7 min 8 23 23 57 2 min 8 14 14 57 3 min 8 15 8 7 57 4 min 8 15 2 13 58 2 min 8 25 15 10 58 3 min 8 16 3 13 58 5 min 8 20 20 59 3 min 8 23 1 22 59 5 min 8 14 14 60 1 min 8 8 8 60 2 min 8 22 12 10 60 21% min 8 40 40 61 2 min 8 79 79 62 Ló min 8 9 9 62 34 min 8 14 14 62 1 mi 8 37 3 34 62 2 min 8 24 24 63 34 min 8 45 45 64 4% min 8 54 54 65 14 min. 8 28 28 1933] BERKLEY AND BERKLEY—TEMPERATURE AND HUMIDITY 597 cotyledons, followed either by their abscission or by the drying up of the whole plant, the amount of injury and the time of its appearance depending upon the severity of the treatment. In older plants, injury appeared in the form of flaccidly wilted leaves and blackened growing tips which proved to be dead. In all cases, the growing tips were more resistant than the leaves, and even when the growing tips were killed secondary growth often appeared at the nodes. There was no withering of the hypo- cotyls and petioles (the cotyledons and leaves remaining green) as described for the lower humidity experiments. Tables rv and v and fig. 3b give the detailed results of the higher humidity experiments. It will be noted from the tables and particularly from the figure that plants of any given age were much less resistant to high temperatures in the saturated atmosphere than at the lower humidity. The thermal death temperature evidently lies in the neighborhood of 55° C., if a definite point can be determined from this curve. This is 8° C. below that shown by the plants treated at the lower humidity. TABLE V COTTON PLANTS FROM 18 TO 120 DAYS OLD TREATED AT APPROXIMATELY 100 PER CENT RELATIVE HUMIDITY Temp. in ] Age of Number | Number | Number |Secondary degrees C. Time lants | of plants living dead growth* days) 40 7 hr. 18 8 5 3 40 10 hr. ] 30 7 23 40 12 hr. ) 6 6 4] 24 hr. 50 30 24 6 4 72 hr t 6 2 4 4 9 hr 3 26 26 4 2 hr. 4 18 18 4 2 hr. 12 12 4 20 min. 6 6 4 40 min 6 5 1 48 1 hr. 6 2 4 435 114 hr. 6 3 3 48 134 hr. 38 38 ) 30 min. 14 11 a ) 45 min. 77 19 13 6 6 hr. hae 38 17 19 19 10 min 12 11 11 min 26 6 1 5 min 36 3 3 10 min 35 102 102 min 72 1 11 min 70 4 2 2 [Vor. 20 598 ANNALS OF THE MISSOURI BOTANICAL GARDEN TABLE V—Continued Temp. in x Age of Number | Number | Number |Secondary degrees C. | Time P E of plants | living dead growth” 5 3 min. 1 2 16 1 2 — © = = to O e en GU is i PRE 00 t2 DIDI CSO:-i bt z = 5 — = M pi h ^ C.b2b20D-—-——Z L L 114 min. 60 Angra ih RR DAD BEEBE PRPPPRPE SES PRIUS, cc a a a *This column shows the number of plants that put forth secondary growth at the nodes after the growing tips had been killed. DISCUSSION When the data of the two sets of plants were compared, it was found that the cotton plant was much more resistant to high temperatures under the lower-humidity conditions. Further- more, the nature of the injury and the parts of the plant first to be affected were entirely different in the two sets. At the lower relative humidity, the petioles, young stems, and the hypocotyls were the first to be killed, whereas at the higher relative humidity they were the last to be affected, the leaves and cotyledons 1933] BERKLEY AND BERKLEY— TEMPERATURE AND HUMIDITY 599 dying first. "The seedlings treated at the lower relative humidity were invariably killed by injury to the hypocotyls. Experiments showed that the cotyledons of the seedlings treated at the lower relative humidity lost moisture at à much greater rate than the hypocotyls during the treatment. Im- mediately following their removal from the temperature chamber the hypocotyls lost moisture rapidly while the cotyledons re- gained a large percentage of that lost during treatment. This would indicate that the evaporation of the water immediately utilized enough of the heat energy around the leaves and cotyle- dons to protect them until the other portions became injured and even killed. This assumption is further substantiated by the fact that all plants were so readily killed in the saturated atmosphere where transpiration was reduced to à minimum. These experiments tend to prove the statement first made by Sachs (764) and later by Ewart ('03) and Baker (729) that reduced transpiration allows a more rapid concentration of heat in the plant. This phenomenon has not been emphasized by sufficient experimental data to show its true significance, and apparently not in any case has it been previously demonstrated under the conditions of thermal death temperatures. Clum (’26) claimed that this cooling effect was greatly over-estimated, but his experiments on thermal death temperatures were not sufficiently well controlled to justify his statement. In the light of the facts shown by the present experiments it is evident that the thermal death point has not as yet been clearly defined. It will be necessary to limit the term to the death of the protoplasm alone and not to the entire plant. The death of the plant might be caused by the injuring or killing of some portion or organ which would prevent normal functioning, as shown by the seedlings treated at the lower relative humidity. Local injury which is not outwardly evidenced may occur at temperatures below that designated as the thermal death tem- perature. Litardiére (25), working with onion root tips, found that the nuclei were affected at 48? C. after 24 hours and that the injury increased with higher temperatures even for shorter periods of time. Yamaha (’27) gave 38? C. as a critical temper- ature for the root tips of the bean, and Milovidov (32), working [Vor. 20 600 ANNALS OF THE MISSOURI BOTANICAL GARDEN with a number of different plants, found that temperatures around 45-47? C. were injurious to the plants studied if the treatment were prolonged. The cursory microscopic studies made on the cotton plants in the present work showed that temperatures of 55° C. and above, even for very short periods of time, were injurious to the protoplasm, causing plasmolysis and deformation of the nuclei. Since these were not uniformly distributed in the plant it would be difficult to estimate their ultimate effects. Blackman’s idea of the “extinction temperature"! (’05) necessitates the setting of an arbitrary time limit which, in the opinion of the writers, is not justified since the time required for the protoplasm itself to assume the thermal death temperature would vary with atmospheric conditions under which the plants were treated. At present no entirely satisfactory definition of the thermal death point can be given, but the following may serve until more data is available: The thermal death point is that temperature which, at a given relative humidity, will kill the protoplasm im- mediately upon its assumption. Were it not for the indication that humidity has an effect in addition to preventing cooling caused by the retardation of the transpiration rate, it would be possible to eliminate from the definition the qualifying phrase dealing with humidity. If the humidity had no other effects it would merely vary the time required for the plant cells to assume the temperature necessary to kill the protoplasm. This does not appear to be the fact, however, since there is so much difference between the low- and high-humidity experiments. It is likely that the extremely high relative humidity has an addi- tional effect, that of smothering the plants. When plants from an atmospheric temperature of 25-30? C. were abruptly trans- ferred into a temperature of 50-60? C. and a practically saturated atmosphere, condensation immediately took place on the surface of the leaves and stems. By reducing the transpiration stream 1 . we ought to find a temperature at which the earliest estimation that could be actually made would give no measurable assimilation. The lowest temperature to give this result might be called the ‘extinction temperature’ (say in 100 seconds, for the accepted specific extinction temperature would an course have to be arbitrarily defined in time units).’’ 1933] BERKLEY AND BERKLEY— TEMPERATURE AND HUMIDITY 601 and the usual exchange of gases, this film of water would naturally limit the oxygen supply of the plant. Since plants in the higher temperature would demand more oxygen than could be obtained, it is possible that the protoplasm was killed by a combination of factors. The evaluation of the extent and full importance of this smothering effect will necessitate further experimentation. In regard to the time limit as a criterion of the thermal death point, we must consider the factors influencing the rapid change of temperature of the plants from that of the atmosphere in which they were grown to that of the treating chamber. Some of these factors are transpiration in the plant, condensation of moisture on its surface, and temperature gradients immediately around the plants. Temperature gradients can be prevented by keeping the air thoroughly mixed, but it is very difficult to determine the exact relations of transpiration or condensation and the change in temperature of the protoplasm. An examination of fig. 3a shows a sharp break in the line between 60 and 65° C., which marks the critical temperature for the cotton plant at the low relative humidity. A thermal death temper- ature of 63° C. may then be chosen, but that is the temperature of the atmosphere. It will be noted that it took from 5 to 7 minutes to kill the plants at 63° C. and progressively shorter intervals of time as the temperature increased, which would indicate that the protoplasm died immediately upon assuming the temperature of the atmosphere. The time lag, caused primarily by the cooling effect of transpiration, makes it neces- sary either to determine the exact temperature of the protoplasm or, more simply, to plot the temperature against the time and determine from the graph thus obtained the approximate thermal death point. It is true that prolonged temperatures below 63° C. will kill the plants, but it is also evident from the slope of the upper portion of the line that it did not take, for example, 25 minutes for the protoplasm to assume a temperature of 60° C. when treated at this temperature. Then they were killed gradually, and in such a case the reactions set up by the high temperatures may have led to the death of the protoplasm. It may be concluded that no definite temperature can be set [Vor. 20 602 ANNALS OF THE MISSOURI BOTANICAL GARDEN as the thermal death point of the cotton plant without stating first the atmospherie conditions under which the plants were treated and giving the ages of the plants. Transpiration plays a definite role in the cooling of the leaves and cotyledons under low relative humidity conditions. SUMMARY 1. T'wo series of cotton plants 5 to 180 days old were exposed to super optimal and thermal death temperatures for periods of time ranging from 1% minute to 72 hours. In Series I the plants were exposed to temperatures of 42-84? C. at an average relative humidity of 69 per cent at temperatures below 50? C. The evaporation rate was substituted for relative humidity in this series. In series II the plants were exposed to temperatures of 40-65* C. at an approximately saturated atmosphere. 2. Seedlings were less resistant to high temperatures at any given relative humidity than older plants. 3. The plants were less resistant to high temperatures at the higher relative humidity. 4. At the higher relative humidity, the leaves and cotyledons were the first parts of the plant to be affected. 9. At the lower relative humidity, the hypocotyls of the seed- lings and the petioles and young stems of the older plants were the first to be affected. 6. During treatment at the lower humidity, the cotyledons lost water rapidly whereas the hypocotyls lost very little. After removal from the treating chamber the cotyledons regained a large percentage of the water lost and became turgid again, whereas the hypocotyls continued to lose moisture until they were completely withered. 7. The more rapid rate of transpiration in the cotyledons apparently utilized sufficient heat energy to protect them from being noticeably injured until the hypocotyls were killed. "These data tend to substantiate the theory that the cooling effect of transpiration is of great value to plants in preventing an accu- mulation of heat energy. 8. The saturated atmosphere of Series II appeared to have an additional effect, that of smothering the plants. This was 1933] BERKLEY AND BERKLEY— TEMPERATURE AND HUMIDITY 603 apparently due to the retardation of the gas exchange caused by the condensation of moisture on the surface of the plants, thus reducing the oxygen supply. 9. Due to the two-fold effects of humidity the following is given as a tentative definition of the thermal death point: The thermal death point is that temperature which, at a given relative humidity, will kill the protoplasm immediately upon its assumption. 10. No definite temperature can be given as the thermal death point of the cotton plant without stating the humidity of the atmosphere and the age of the plant. The writers wish to express their appreciation to Dr. E. S. Reynolds, Physiologist to the Missouri Botanical Garden, for suggestions and criticisms given during the progress of the work. BIBLIOGRAPH Y Baker, F. S. 29). Effects of excessively high temperatures on coniferous reproduc- tion. Jour. Forest. 27: 192 Berkley, E. E. (31). Studies of the effects of different lengths of day, with varia- tions in temperature, on vegetative growth and reproduction in cotton. Ann. Mo. Bot. Gard. 18: 573-604. : Blackman, V. H. (05). Optima and limiting factors. Ann. Bot. 19: 281-295. 1905. Bose, J. C. (13). Researches on irritability of plants. Longmans, Green and Co., 192 Clum, H. H. (26). The effect of transpiration and environmental factors on leaf temperatures. I. Transpiration. Am. Jour. Bot. 13: 194—216. —— ———, (26). The effect of transpiration ie environmental faoltas rs on leaf temi pow bns II. Light intensity and the relation of ins nici to the thermal death point. bid. 13: 217-230. Collander, R. (24). Beobachtungen über die quantitativen Beziehungen zwischen Tótungsgeschwindigkeit und Temperatur beim Würmetod pflanzlicher Zellen. Soc. Sci. Fennica, Comment. Biolog. 1: 1-12. 1924. Ewart, A. J. (03). On the physics and Mas as of protoplasmic streaming in plants. pp. 59-68. Oxford, 1903. Fung, H. K. (’11). An ecological study of the American cotton plant with incidental reference to its possible adaptability in China. An unpublished doctor's thesis of Cornell University. June, Gilbert, B. E. (26). The response of certain photoperiodic plants to differing emperature and humidity conditions. Ann. Bot. 40: 315-320. 1926. Lepeschkin, W. (25). Pflanzenphysiologie. Julius Springer, Berlin, 1925. Litardiére, R. de (25). Les diverses étapes de l’ “agonie” des noyaux sous l'in- fluence d'une température élevée. Soc. Biol, Compt. Rend. 92: 796-798. 1925. [Vor. 20, 1933] 604 ANNALS OF THE MISSOURI BOTANICAL GARDEN Milovidov, P. F. (32). Einfluss von Wasser hoher Temperatur auf den Kern der Pflanzenzellen im Lichte der Nuklealreaktion. Protoplasma 17: 32-88. 1932. Pfeffer, W. F. P. ('03). Physiology of plants 2: 224-231. Eng. ed. i Sachs, J. (63). Die vorübergehenden Starre-Zustünde periodisch beweglicher und reizbarer Pflanzenorgane. Flora 46: 449—459. 1863 ————, » 64). Über die obere Temperatur-grünze der Vegetation. Ibid. 47: 5-12. 1864. Wallace, R. H. (31). Studies on the sensitivity of Mimosa pudica. III. The effect of temperature, humidity and certain other factors upon seismonic sensitivity. Am. Jour. Bot. 18: 288-307. 1931. Yamaha, G. (27). Experimentelle zytologische Beiträge. II. Mitteilung. Uber die Wirkung des destillierten Wassers auf die Wurzelspitzenzellen von Vicia Faba bei verschiedenen Temperaturen. Imp. Univ. Tokyo, Jour. Fac. Sci. 2: 215-290. 27. STUDIES IN THE APOCYNACEAE. IV THE AMERICAN GENERA OF ÉCHITOIDEAE ROBERT E. WOODSON, JR. Research Assistant, Missouri Botanical Garden Instructor in Botany, Henry Shaw School of Botany of Washington University INTRODUCTION Because of their diverse and highly complicated floral mech- anism, the American genera of the subfamily Echitoideae are at present perhaps the most imperfectly understood of Apocynaceae. As in pre-Linnean and early post-Linnean times laticiferous herbs of several distinct affinities were grouped indiscriminately under the name ‘‘Apocynum,” so even to-day practically any echitoid liana indigenous to the western hemisphere may pass as a species of the inclusive genus Echites P. Br., although several excellent genera have been segregated from that amorphous and hetero- geneous aggregate by Bentham, Robert Brown, Alphonse de Candolle, Mueller-Argoviensis, and other eminent systematists. To the less exacting botanical publie, however, Echites has re- mained a convenient catch-all and species of very dubious congenericity have been described and redescribed under that name until the literature has become so involved that it is a dangerous task to attempt routine determinative work, and much more so to essay the description of novelties. The Apocynaceae of the western hemisphere attain their greatest complexity in tropical South America. Although mono- graphic work hinting of any degree of finality upon most groups of tropical American plants had best be deferred for the future because of the incomplete state of our knowledge concerning the flora of that region, constant exploration in behalf of varied interests is bringing to the attention of science increasing multi- tudes of plants which must be critically identified by the system- atic botanist. In order to facilitate that activity and to render the results more sure with respect to the numerous representatives of Apocynaceae, it has been considered highly desirable to under- take at this time a tentative revision of the troublesome Echit- oideae. Issued December 19, 1933. Ann. Mo. Bor. Garp. Vor. 20, 1933. (1) (605) [Vor. 20 606 ANNALS OF THE MISSOURI BOTANICAL GARDEN The most perplexing problem, as well as that first encountered in the course of such a revision, as has already been intimated, concerns generic relationships. The first requirement of a re- vision of the American Echitoideae should be the careful delimita- tion of the genera most deserving of recognition, and their epitomization in the form of an identificatory key. The second problem, of necessity wholly dependent upon the first, is that of a separate revision of each genus maintained within the sub- family. In order to judge and to correlate more competently the numerous genera which have been proposed from time to time, the morphology of certain critical vegetative and floral organs as been investigated and interpreted in separate reports which are in preparation for publication in the near future. Only a brief account of the structural features employed as classificatory criteria is included here as an aid in the use of both generic and specifie keys. It may be well to point out at this time that the limits of the subfamily Echitoideae as accepted in this instance are those established in the first paper of this series! to the exclusion of the genera of Apocynoideae characterized by pollen maintained within persistent tetrads. For the purely taxonomie portion of the revision, recourse has been had to the extensive collections of most of the principal herbaria of America and Europe. Since it is believed that such inclusions generally enhance the value of taxonomie studies, ex- siccatae have been freely cited, particularly in the instance of difficult or little-known species. The herbaria where specimens have been examined or obtained for study, together with the symbols employed in their citation, are as follows: Arnold Arbo- retum of Harvard University, Jamaica Plain (AA); Botanischer Garten zu Berlin-Dahlem (B); Museo Nacional de Buenos Aires (BA); Herbier Boissier, Institut Botanique de l'Université, Geneva (BB); British Museum (Natural History), London (BM); Jardin Botanique de l'État, Brussels (Bx); Botanisk Museum, Copenhagen (C); California Academy of Sciences, San Francisco (CA); Lindley Herbarium, Botany School, Cambridge, Engl. (Camb.); Delessert (D) and de Candolle (DC) herbaria, Con- ! Woodson, R. E., Jr. Ann. Mo. Bot. Gard. 17: 9. 1930. (2) 1933] WOODSON—STUDIES IN THE APOCYNACEAE. IV 607 servatoire Botanique, Geneva; Field Museum of Natural History, Chieago (FM); Gray Herbarium of Harvard University, Cam- bridge, Mass. (G); Royal Botanie Gardens, Kew (K); Linnaean Society of London (Linn.); Dudley Herbarium, Leland Stanford Junior University, Palo Alto (LS); Botanisches Museum, Munich (M); Missouri Botanical Garden, St. Louis (MBG); Museo Com- ercial de Venezuela, Caracas (MC); Muséum National del’ Histoire Naturelle, Paris (MP); New York Botanical Garden, New York City (NY); Academy of Natural Sciences of Philadelphia (PA); Pomona College, Claremont (PC); Naturhistoriska Riksmuseum, Stockholm (S); Botanisch Museum, Rijks Universiteit, Utrecht (U); University of California, Berkeley (UC); United States National Herbarium, Washington (US); Naturhistorisches Mu- seum, Vienna (V). To the curators and members of the staff of these institutions, as well as to many other friends, the writer would express his gratitude for innumerable instances of aid and good will. He is also greatly indebted to the Board of Trustees and to the Director of the Missouri Botanical Garden for the privilege of a trip to Central America in 1930 for observation and study. HisroricaL REVIEW None of the species which are now comprised within the American genera of Echitoideae was known to Linnaeus at the time of publication of the first edition of the ‘Species Plantarum’ in 1753 or of the fifth edition of the ‘Genera Plantarum’ the year following. In 1756, however, Patrick Browne? inaugurated the genus Echiles, citing rather full distinguishing characters, but failed to assign a binomial species. The single polynomial referred to the genus by Browne, ''Echiles foliis ovatis nitidis venosis; floribus herbaceis," was founded upon a plant well illus- trated by Sir Hans Sloane; and quite clearly establishes the common species of the Greater Antille, the Bahamas, the peninsula of Yucatan, and southern peninsular Florida, Æ. umbellata Jacq., as the type of the genus as we shall deal with it here. 2 P. Br. Hist. Jam. 182. 1756. 3 ** Apocynum scandens majus, folio subrotundo," Sloane, Nat. Hist. Jam. 1: 207. pl. 181, fig. 2. rfe (3) [Vor. 20 608 ANNALS OF THE MISSOURI BOTANICAL GARDEN The monotypic genus Echites as defined by Browne was inter- preted by Linnaeus‘ in the tenth edition of the ‘Systema’ merely as constituting a species of T'abernaemontana, and the former generic name was employed as a substantive adjective. Four years after its first publication, the genus Echites was expanded by Jacquin® to include ten species: E. biflora, E. umbellata, E. agglutinata, E. trifida, E. quinquangularis, E. suberecta, E. torosa, E. repens, E. spicata, and E. corymbosa. In 1763, Jacquin® maintained the species which he had described three years previously, illustrating each in an admirable manner. Of the ten enumerated, which are now distributed among nine well-known genera, all but three were original. For E. umbellata, Jacquin’ was able to cite four pre-existing polynomials, two for E. suberecta,’ and one each for E. torosa,? and E. biflora.1 In the second edition of the ‘Species Plantarum’ Linnaeus" included Zchites within the ‘“Pentandria Monogynia," and appended the ten species of Jacquin. Two years subsequently the genus also appeared in the sixth edition of the ‘Genera Plantarum.’! It would be a weary task to record in detail the great enlarge- ment of the genus Echites, which almost immediately followed its original publication, to include scores of Apocynaceous species of the subfamily Echitoideae native to nearly all parts of the * L. Syst. Nat. ed. 10. 945. 1759. 5 Jacq. Enum. Syst. Pl. Carib. 13. 1760. ë Ibid. Select. Stirp. Am. Hist. 1: 30-35; 2: pls. 21-30. 1763. 7“ Echites scandens foliis ovatis nitidis venosis; floribus herbaceis,” P. Br. Hist. Jam. 182. 1750; sprit scandens folio cordato, Tesi albo," Catesb. Nat. Hist. Carol. 1: 58. pl. 68. 1754; “Apocynum scandens majus, folio subrotundo," Sloane, Nat. Hist. Jam. 1: 207. pl. 131, fig. 2. 1707; ^ Periploca alia, floribus ug god cir- cinatis & crispis, seu Nerium scandens, radice Bryoniae tuberosae," Plum. Pl. A 210. pl. 216, fig. 2. 1759. 8 “Apocynum scandens, amplo flore villoso, luteo, siliquis angustissimis," Plum. Cat. Pl. Am. 2. 1703; “Apocynum erectum, fruticosum, flore luteo maximo specio- sissimo,” Sloane, Nat. Hist. Jam. 1: 206. pl. 130, fi edd . 1707.—The former appears to apply somewhat more obviously to E. biflora Ja “Nerium sarmentosum, scandens, ramulis fepe P folliculis gracilibus torosis,"' P. Th Hist. Jam. 181. pl. 16, fig. 2. 10 ** Apocynum scandens, ips nerit albo,” Plum. Descr. Pl. Am. 82. pl. 96. 1693. u L. Sp. Pl. ed. 2. 307. 1762. 2 L. Gen. Pl. ed. 6. 117. 1764. (4) 1933] WOODSON—STUDIES IN THE APOCYNACEAE. IV 609 world.? The result was soon so obviously conglomerate that in 1811, in his treatise ‘‘On the Apocineae," Robert Brown" rec- ommended the limitation of the generic name to the species of the western hemisphere, distinctly citing E. umbellata Jacq. as theoriginalspecies. Although, unfortunately, he did not attempt to redefine the characters of Echites, Brown described in the same work" the genus Prestonia, which is therefore the second of the American genera of Echitoideae in point of age. Since 1811, the publication of American genera of the sub- family Echitoideae has constantly augmented. Summarized chronologically in tabular form, the appearance of genera from 1756 until the present has been as follows: 1756 Echites P. Br. Hist. Jam. 182. 1756; Jacq. Enum. Syst. Pl. Carib. 13. 1760. 1811 Prestonia R. Br. Mem. Wern. Soc. 1:69. 1811. 1818 Forsteronia G. F. W. Mey. Fl. Esseq. 133. 1818. 1819 Thenardia HBK. Nov. Gen. 3: 209. 1819. 1825 Haemadictyon Lindl. Trans. Hort. Soc. Lond. 6: 70. 1825. 1828 Syringosma Mart. ex Rchb. Consp. 134. 1828. 1838 Exothostemon G. Don, Gen. Hist. Dichlam. Pl. 4: 82. 1838. 1840 Mandevilla Lindl. Bot. Reg. N. S. 3: pl. 7. 1840. 1841 Odontadenia Benth. in Hook. Jour. Bot. 3: 242. 1841. Thyrsanthus Benth. loc. cit. 245. 1841. 1844 Malouetia A. DC. in DC. Prodr. 8: 378. 1844. Anisolobus A. DC. loc. cit. 395. 1844. Robbia A. DC. loc. cit. 444. 1844. Secondatia A. DC. loc. cit. 445. 1844. Laseguea A. DC. loc. cit. 481. 1844. Dipladenia A. DC. loc. cit. 1844. Laubertia A. DC. loc. cit. 486. 1844. 1849 Cycladenia Benth. Pl. Hartw. 322. 1849. 1855 Cylicadenia Lem. Illustr. Hort. 2: Misc. 9. 1855. 1860 Heterothrix Muell.-Arg. in Martius, Fl. Bras. 6': 133. 1860. 13 Of. Stadelm. Flora 24!: Beibl. 1-13. 1841. M R. Br. Mem. Wern. Soc. 1: 59. 1811. 1$ [bid. loc. cit. 69. 1811. (5) 610 1878 1897 1905 1909 1917 1920 1924 1927 1931 1932 [Vor. 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN Macrosiphonia Muell.-Arg. loc. cit. 137. 1860. Amblyanthera Muell.-Arg. loc. cit. 141. 1860, not Blume. Mesechites Muell.-Arg. loc. cit. 150. 1860. Rhodocalyx Muell.-Arg. loc. cit. 172. 1860. Rhabdadenia Muell.-Arg. loc. cit. 173. 1860. Stipecoma Muell.-Arg. loc. cit. 175. 1860. Elytropus Muell.-Arg. Bot. Zeit. 18: 21. 1860. Prestoniopsis Muell.-Arg. loc. cit. 22. 1860. Urechites Muell.-Arg. loc. cit. 22. 1800. Chariomma Miers, Apoc. So. Am. 110. 1878. Eriadenia Miers, loc. cit. 117. 1878. A ptotheca Miers, loc. cit. 150. 1878. Rhaptocarpus Miers, loc. cit. 151. 1878. M icradenia Miers, loc. cit. 158. 1878. Homaladenia Miers, loc. cit. 163. 1878. Angadenia Miers, loc. cit. 173. 1878. Perictenia Miers, loc. cit. 182. 1878. Temnadenia Miers, loc. cit. 207. 1878. Mitozus Miers, loc. cit. 217. 1878. Streptotrachelus Greenm. Proc. Am. Acad. 32: 298. 1897. Bracea Britton, Bull. N. Y. Bot. Gard. 3: 448. 1905, not King. Orthechites Urb. Symb. Ant. 6: 36. 1909. Belandra 8. F. Blake, Contr. Gray Herb. N. S. 52: 78. 1917. Neobracea Britton, in Britton & Millsp. Bahama Fl. 335. 1920. Codonechites Mgf. Notizblatt 9: 80. 1924. Macropharynx Rusby, Mem. N. Y. Bot. Gard. 7: 327. 1927. Salpinctes Woodson, in Gleason, Bull. Torrey Bot. Club 58:453. 1931. Allomarkgrafia Woodson, Ann. Mo. Bot. Gard. 19: 45. 1932. Asketanthera Woodson, loc. cit. 46. 1932. Fernaldia Woodson, loc. cit. 48. 1932. (6) 1933] WOODSON—STUDIES IN THE APOCYNACEAE. IV 611 Galactophora Woodson, loc. cit. 49. 1932. Peltastes Woodson, loc. cit. 375. 1932. An interesting if frequently perplexing feature shared in common by many species of the majority of genera enumerated is their persistent popular association with the parent genus Echites within which they were once included. Perhaps the highest attainment of the research upon the Apocynaceae of Alphonse de Candolle and particularly of Mueller-Argoviensis was the distillation of the American representation of the in- clusive genus Echites into numerous more natural entities. Un- fortunately, the painstaking studies of those eminent systematists was discredited to a certain degree by the contemporaneous treatment of John Miers which has discouraged the employment of the valid segregates by confounding with them a miscellaneous assortment of incongruous species and genera. The revision of the Apocynaceae under the authorship of K. Schumann" in Engler & Prantl’s ‘Natiirlichen Pflanzenfamilien’ has helped to reclarify the status of the American Echitoideae, but has suffered as a result of the acceptance of Miers's mistaken interpretations in several instances. Schumann unfortunately profited little by the excellent natural order given the genera of the subfamily by Mueller, and as a consequence in his enumer- ation one finds Eriadenia Miers and Mandevilla Lindl., considered as synonymous by Markgraf!” and in the present revision, sepa- rated by such discrepant genera as Macrosiphonia Muell.-Arg., Rhodocalyx Muell.-Arg., Cycladenia Benth., Dipladenia A. DC., Odontadenia Benth., Elytropus Muell.-Arg., Rhabdadenia Muell.- Arg., and Laubertia A. DC., in the order named. This situation is the result of the key characters employed by Schumann, particularly that of gross habit, which not only separate closely related genera but even exclude generically certain species from others which are manifestly their congeners. MORPHOLOGY OF THE TAXONOMIC CRITERIA The danger of basing generic distinctions within the Echit- oideae upon habit already has been observed. As a group, the 16 K, Sch. in Engl. & Prantl, Nat. Pflanzenfam. 4?: 109-189. 1895. 17 Mgf. Notizblatt 9: 85. 1924. (7) [Vor. 20 612 ANNALS OF THE MISSOURI BOTANICAL GARDEN subfamily is predominantly composed of extensive lianas so characteristic of the tropical American rain-forests. Forsteronia, Prestonia, Odontadenia, and Echites (sensu stricto) are familiar examples of genera with exclusively volubile habit among all known species. Neobracea is fruticose, while Malouetia is the only known genus attaining arboreal proportions. The species comprising Macrosiphonia, Salpinctes, Galactophora, and Rhodo- calyx might with almost equal aplomb be called suffrutescent herbs or subherbaceous undershrubs. Cycladenia is unique with- in the entire family because of its low, subsucculent, herbaceous habit, and its subalpine habitat. Various combinations of habit render that character an extremely fallible guide in particular instances. Mandevilla is probably the outstanding example in this regard, including among its many species lianas, suffruticose shrubs, and suffrutescent herbs in a wide range of intergradation. Although predominantly terete, the stems of certain American Echitoideae display a conspicuous alate development, as in several shrubby species of Mandevilla subgen. Exothostemon (§ Eriadenia Mgf.). Such development of the stem appears to be caused by the activity of the phellogen rather than of the vascular cambium, however, and thus bears only a superficial resemblance to the polydesmic stems of certain other tropical lianas. Phyllotaxy is relatively uniform throughout the subfamily in America, the distichous arrangement being almost invariable. Mandevilla Benthamii (A. DC.) K. Sch. constitutes a striking exception in its ternate or quaternate foliage. The leaves of species of Laubertia and Macrosiphonia are occasionally ternate as well. The foliar glands of several genera are easy clues for identi- fication. "These organs are small, occasionally quite inconspicu- ous, multicellular, aculeolate emergences borne upon the ventral surface of the midrib. In Allomarkgrafia, Mesechites, Macro- siphonia, and most species of Forsteronia, they occur at the base of the midrib in shapes, positions, and numbers generally char- acteristic of genera or subgenera and to a somewhat less extent of species. In Allomarkgrafia, Macrosiphonia, and Forsteronia the glands are fusiform and indistinctly gathered into groups of few to several. In Mesechites, however, the two component sub- (8) 1933] WOODSON—STUDIES IN THE APOCYNACEAE. IV 613 genera are rather nicely distinguishable by means of the glands, those of subgen. Humesechites assuming a rather flat or laminate shape and occurring in groups of two to five clustered concentri- cally along the base of the midrib, while those of subgen. Antille- chites are less conspicuous, fusiform bodies which are predomi- nantly found in pairs radially placed at the very base of the midrib. n Mandevilla subgen. Eumandevilla, the foliar glands are fusiform and are grouped at the base of the midrib as in Allo- markgrafia, but in subgen. Exothostemon the same structures occur in variable numbers along the entire length of the rib. Were it not for three common species supplying an indubitable link between the two subgenera, which also have technical differences in the reproductive organs, the characteristic position of the foliar glands of subgen. Exothostemon would render it one of the most easily recognized genera of the subfamily. Species of the latter subgenus, however, have long been popularly associated with those of Humandevilla, and in view of this consideration, together with that of the intermediate characteristies of M. funiformis (Vell.) K. Sch., M. callista Woodson, and M. congesta (HBK.) Woodson, it has been thought desirable to maintain their unity for the present. The Apocynaceae are popularly known as an exstipulate group. However, immediately subtending the petiole of many species of that family, usually forming a definitely interpetiolar girdle when the phyllotaxy is opposite or verticillate, are more or less conspicuous appendages, variously arranged, which should probably be interpreted as stipular vestiges. In the Echitoideae of the western hemisphere, these structures are usually glandular in function, although in Odontadenia anomala (Heurck & Muell.- Arg.) Macbr. they are somewhat foliaceous, this departure from the predominant condition prompting Miers to establish for its inclusion the genus Perictenia. 3 ERE ns e 214; parallela, 157, 218; parvi 140, var. latifolia, 142; paucifolia, 140; avoniana, nduliflora, 194, 216; p hyteumoides , 194; inc nifolia, 185; plantaginea, diflora, '176, var. lat ma, A , var. congesta, 200, 216 nC 174; pusilla, 182, 216; 811 recurva, 161, 214; Hage a ie. rhyaco- phila, 168, 214, V. oda 2,170, var. pr rocumben n 170; M ion 190, 216; Rothrockii, 161; Rusbyi, 9, 216; scapiformi. Schiedeana, 172, 214; Schultzei, 151; Shannonii, 177, 214, f. compa p. 178; sain 3, 4 195, 216; spatulata, 183, 216; s cola, 201; , stellarioides, 1152; Stuebelii, in So E heimii, 145; : valerianoides, 182; Ente, 153, d ys cetoxieoides, 1 viridis, eberbaueri, 189, Worldellizna, 202, 216 Helleranthus quadr angulati us, 313 Heterothrix, 645; pycnantha, 668; Van- heurckit 771 Hieracium carneum, 798 | tei a 70 Holoc s Leprieurii, 397 seus dumosus 800 , 645; "br. evifolia, 680; lin- ariaefol ia, 6 » pastoru m, 680; pedunc ularis Jhon puberula, 680; tonuifolia, 680; 8 Gr Frente, Super optimal and thermal death temperatures of the gg plant as affected by variations S omi ion Mcr ne of nutrient solutions in relat o growth of i pelle 27, 579; various, reaction of myces capsulatus to, myces capsulatus, Hymenopappus ie 798 Hypopitys sanguinea, 799 I Indol, reaction of, to Endomyces capsu- latus, 517 Iron in nutrient solutions for orchids, 226, 365, 572 Isoetes, 74; Butleri, 75; Engelmanni, 74; melanopoda, 75 J Juniperus flaccida, 796 K Keerlia, 37; bellidifolia, 37; effusa, 38; linearifolia, 33; mexicana, 31; ramosa, 4, 25; skirrhobasis, Knudson's solution for growing orchids, 365, 572, 574 Kujala, Dr. Viljo, lichens collected by, 7 812 ANNALS OF THE L La Garde's solution for growing orchids, 226, 234, 363, 572, 578 sther L. Astranthium and re- kan ap 23 645; en 673 poe ie d 6 f E Jrisco-fusca, 674, . — E aa. OU "n 'aniennacea, 741; bracieta, 671; emarginata, "d erecta, 673, Guille eminiana, 673, olivacea, Mie 2. €o-f obliquinervi A id. griseo- aa a, obovata, 673, » 0 tr psi 673; ô. acutifolia, 673, 673; scabri 673, ‘ foliosa, 658; glabra, 673; s Quilleminiana, ; ookeri, 6 Mandoni, 671; obliquinervia, 673; Pentlandiana, 671; subspicata, 740; inuidia. 740; villosa, 74 4 Lecanora yg pa, "e ‘gy timate 14; symmicta, 14; umbr cidea goniopl r^ 19;' lavicide: 19; E: 3 E opsis os Lempholemma, 404; compactum, 404; — , 406; Dussii, 405; obli ique- peltatum, 40 5 Lepidoc na carassense, 440 reir vt americanum, 404; Finkii, Leptogidium byssoides, 403 Leptogium, 404, 416; E bi 418 e "0 O o'a 3EE UL coralloideum, 419, 430; corrugatulum, 131; denticulatum, 419, 432; diapha- 4. diaphanum f. cor 0; 8 ; foveolatum, 418, 495; fusisporum, 418; granulare, 419; Hide ii var. osum, 2; ynum, 419; inflexum, 418, a var. isidiosulum, 418, 423; laevius, 420; Lafayetteanum, 420; lepto phyl- linum, 432; subg. Mallotium, 418, ^92; —— um, 408, 419, = matto- ; mega 18; centem di , 419; se Aamar 418, 426; moluccanum, 436; olivaceum, 418, [Vor. 20 MISSOURI BOTANICAL GARDEN 424, var. granulosum, 424; pachy- cheilum, 420; papillosum, S 422; phyllocarpum, 419, 428 cam- pestre, , 43 reg 420; Bos hellum latum, 421; resupina oides, 419, ; Standleyi, 490, 435; allan, 420; jstipitatum, 418, 428; tremelloides 17, 416; t melloides, , Var. azureum . isidios Min "132, it pressopunctatum, 421, var. leptophyl- lim , var. minor, 436; T Tieng: Mar a 11, f. compacta, No rth niin "Contri- n to de. 7 — ,asureun 434; MN be , 408; diaphanum, 421; Fe 431; melanocarpus , 394; Loa: sus, 453; velut linus, 403; vesiculosus, 6 a sr he of wae Rica, The foliose and F Light: white, effect of, on growth of En- omyces capsulatus, 521; of red, 522 Linum Schiedeanum, ' 79 Lithospermum viride, 797 Lobaria linita, 17; oregana, 17; pul- Lobarina verrucosa, 17 Loeselia reggii, icera arizonica, 797 eria areolata, 69 Bo ^W 12; complanatum flabelli- orme, 73; lucidulum, 73, var. poro- philum, 73; paeen v A 73 M MODUS, TEL lansae, 788; Berlandieri, 783; Éraohysipbon, 780; 784; Guar- v 788, var. typica, 787; pinifolia, 788, f glabr 88, var. i ia, . f pedu aris, 788, f. b : prostrata, 787; subg. Telosiphonia, 778; Velame, 788; Velame B. goyazen- sis, 788; verticillata, 786, a. petraea 787, B , 787, y. zr var. Missionum, 785; Wrightit, 81 1933] INDEX 813 Magnesium, relation of growth of orchid 22 alme lla, 440, 447; albida, 448, caesiocinerea, 448, 452; die 448, 451; Randii, 448; Santamariae, 448, 450 Maltose, growth of orchid seedlings in 5 ace nsis, 663; ca İsta, 700; cau- rensis, 770; cercophylla, 669; cereola, 12; coccinea, 734; congesta, 675; convolvulacea, 687; crassinoda, 703; cuspidata, 730; denticulata, 764; Di jel- J "D mithii Exothostemon, 737; Fendleri, 758; fluminensis, 764; foliosa, 658; fragilis, . 13; funiformis, 697; fun niform peduncularis, 697; 750; laxa, 695; sect. laxae, 682; lepto- phylla, 750; linearis, 738; linearis, 756; Loesneriana, a luc ida, 718; Lue tzel- burgii, 698; oni, 693; Mar a n 702, var. S bei. "203, var. typica, 702; mexicana, 660; minor, 711; mollissima, na, 665; sect monta e, 661 ri- can a, 705, var. bahiensis, ded var typiea, 705; Moritziana, 768; Muel- um, 68 pei pital 644, 86; oblongifolia, 708; palus ris, 764; parvifolia, 753; Pavonii, 749; pend la, 725; Pentla ndiana, 671; permix 71 5; pla tydactyla, be polyantha, 769; polyantha, 769; sina, 777; pycnantha, 668; Rai Schlimii, 759; Sc humanniana, 6 scutifolia, 663; Sellowii, 719; sert ae gera, 649, 654; spigeliaeflora, 736; splendens, 707; Spruceana, 776; suav- colens, 695; subcarnosa, 644, 772; subcarnosa, 772; subcordata, 693; subpaniculata, 678; subsagittata, 644, 645, 743; subscorpioidea, 653; sub- Vahleana, 764; torosa, 644, 656; Meri torosae, 655; Trianae, 760; tubiflora, 648; sect. tubiflorae, 647; iana, 695; Ulei, 771; urophylla, 722; Van- heurckii, 771; velutina, 731, var. glabra, 732, var. typica, 732; velutina, Wie venulosa, 723; a asensis, 690; veraguensis, 690; vi villos Td sulphate in eara Aih solu- 226 Mannose, growth of ongie seedlings in solutions containing, 574 Maua nodulosa, 65; Struthiopteris, Media, growth of Endomyces capsulatus in various, 494 Meningitis, fatal cerebrospinal, Endo- of 471 capsulatus as a causative agent o Mentzelia asperula, 799. Mesechites, 629; acumi Deer An- drieuxti, 645, 688; ongkeaix, 644, 775; angustifolia, 640; bicorniculata, 634; brevipes, 638; rownei, , 656; citri- folia, 637; dichotoma, 644; subg dymadenia, 638; dis , 4081; subg rore 30; Guayaqui- lensis, 6 4; nsis, 645, 7 hast ata, 644, Jui “hirtella, 644, 744: hirtellula, 6 4, 686; japurensis, 630: jeter lg 644, 744; lanceolata, 639; linearifo lia, "EUH 636; minima, 643; myrtifolia, axa- e 644 686; repens, 639 nll 641; Micradenia, 645; atroviolacea, 724; cras- sinoda, 703; fragrans 714; oo. rliana, 7 oricandia 706; Riedelii, 740; Sellowii, 719. Missouri, Ferns and fern allies of, 45; A neutral (?) strain of Mucor sphaero- sporus Mitozus brachysta achyus, 753; brevipes, 638; ds "owed 753; exilis, 697; funi- formis, 697; Guanabaricus, '697; epto- phyllus , 756; micr ronila. € rugosus, 755; symp looarpus 755; tenuicaulis, 753; versicolor, 752 814 Monilia, 82, 103 Moore, John Adam, Steyermark, Julian and. Report of a botanical can tion into the mountains of western Texas, 791 Moore, Morris. prt p qe report of a case, with a s of a etiologic yers: a causa- tive agent of fatal bend men- ingitis, Mucor sphaerosporus m Missouri, A neutral (?) strain of, 46 Mycoblastus melinus, 19; sanguinarius, , var. Dodgeanus, 19 Mycoderma, 474 My pait nim dermatitis, 103 Nama ea 799; xylopodum, 799 Nephrod Heskeuntun "Tasting se, 17; laeviga- var. papyraceum, 17; 'subparile, Nephromopsis ciliaris, 13; platyphylla, Nerium sarmentosum, 642 Nitrates, reduction of, in Endomyces capsulatus, — in re to growth of orchid seedlings, , 368. N we pen E 6 Notholaena, 60; de albata 60; Greggii, 796; nivea « lealbat Nutrient solutions for silo 225, 363, Nutritional ions, The sensitivity of or- chid seedlings to, 22: Nine am A new ‘yellow, from tropical A eel iy (§ Ne yy erat "im 1, 6; Stuhlmannii, 3; sulphur Obleti Ockrolckia upsaliensis, 14 Odontadenia cuspidata, 730; glandulosa, Oidium, 82, 102, 474 On oclea, 61; sensibilis, 61; Struthiopteris, Oosporaceae, 104 Ophioglossum, 46; vulgatum, 46 Orbilia, 396 Orehid seedlings, Nutrient solutions for ; The sensitivity of, to nutritional ions, 223; Sources of carbo ydrate for germination and growth of, 5 Engelmanni, 47; or. 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN Osmotic concentration, panty of, to growth of orchid seedling - Osmunda, 71; cinna siad y- toniana, 72; regalis NET p? Struthiopteris, 64 Ostrya Baileyi "79 7 Oxygen, free, relation of Endomyces capsulatus to, 518 P Pannaria, 440, 441; euler 441; bras- iliensis, 442; uleo by m ans, 442; imbricatu la, 442; isid var. 2, f. cinerascens, 452; 8 tlopory, "rm 445; Vainii, 442, 445 Pannariaceae 13; cronia, 462; tula ana, 14; enteromor- pha, 13; exasperatula, 13; hyperopta, i 463; lophyrea, 13; ata, vagans, = vitta , 13 Parmeliella, 440, 452; mierophylloides, o 61; atropurpurea, 62, va ushii, 62, var. cristata, 62; Tiao. 6l; pe a Peltigera puoi 16; canina, 16, va membranacea, 16, var. oce 10; NR NN 16; le phora, 16; malacea, 16; Maurizii, 16; microphylla, 16; Nylasi i, 16; perfida, 16; rufes- cens, 16; suomemsis, 16; variolosa, 16; enosa, TN growth of orchid seedlings in, Pentstemon baccharifolius, 801 i, 798 Por, Lil rat Perry . A revision of the North n species of Verbena, 239 ye oat r amara, 14; faginea, 14; multi- puncta, 14; pertusa, 1 xanthostoma, argh 66; heragonoptera, 68 Phlyctis argena, 14 Phosphorus, growth of orchids in solu- tob containing, 233, 367 Phryma caroliniensis, Physcia, 15; adglutinata var. pyrithro- 1933] INDEX cardia, 15; astroidea, 15; caesia, 15; a, 15; pyri- rele na 15; stellaris, 15, var. rosu- lata, 15 Physma, 404; Boryanum, 407; byrsinum, 08; pruinosum, 408 Pilophoron aciculare, 17 Pinaropappus parvus, 798 e M. Elizabeth. Ferns and allies of icy Placodium chrysodetum, 15; citrinum, td aak Sinn a. "15; ferrugineum, 15; ostigmoideum, 15 Platy, tysma seplentrionale, 12 Polychidium, 403 Po pen Timulicola, 800 Polyp m, 62; incanum, 63; polypodi- oldda "63; "virginianum, 63; vulgare, 63 Polystichum, 63, 66; acrostichoides, 63, incisum, 64, Hen Schweinitzii, 64 sel ohoi. otamogeton clystocarpus, 802; diversi- folius, 802 Potassium i in relation to growth of orchid 234, pe est = hirsuta, 749; ana, 692 psis Fendleri, 670; y cese 676; red venosa 7 H. and Robert E. Wood- on, Jr. new yellow Nymphaea opical Africa, 1 Pteretis, 64; nodulosa, 64 er borg 65; pese aa 65; latiuscu- var. pseudocaudatum, 65 Perlini té, new or rare, in Texas, 796 Pterigyopsis atra, 403 Pieri ilina, Pterospora Andromeda, 799 Quercus hypoleuca, 798 R Raffinose, growth of a seedlings in solutions containing, Ramalina calicaris, E geniculata 11; Roesleri, ori thrausta, ä A B Contribution to ains of western Texas, 791 Rhabdadenia campestris, Rhamnose, of orchid seedlings in solutions containing, 574, 576 Rhamnus fasciculata Rhodocalyx coccineus, 734; hypoleucus, 815 781; lanuginosus, 782; ovatus, 735; 35 81; Tweedianus, 7 Rosa mirifica, 800 S Saccharomycetaceae, 104 Saccha: rem 82, 103, 474; hominis, 103 EAM tpe cent, relation of growth of d id seedli ings to, 371 jum, 408 59 : apoda, 74; apus, 74; Pringlei, 796; E estris, 74 rie ed x. chid seedlings to nu- tional io Shive's solution for growing orchids, 572, Shuttleworthia, 251 iphu E = Sisymb , 798 Skatol, empet e of, to ee cap- Sodium borate i in ovde solutions, 226 98 Sorghastrum nutans, 801 ea aa new or rare, in Texas, 9 Sphaerophoraceae, 393 Sphaerophorus globosus var. lacunosus, 20; melanocarpus, 394; Tuckermanii, Spheno Pp obtusata, 801 ripe Spiranthes s saltensis, 802 Sporotrichum, 8 quamari pum placa, 14 Stau retails circinata, Stephanophorus Stereocaulon tomentosu um, 14 Steyermark, Julian A., John Adam oore port of a botanical expedition into the mountains western Texas, 79 Sticta scrobiculata, Styleurodon carolinianum, 309 Mgr rw in Giu"; grow lutions oleh ted Tikica, 226, Sulphate, relation of growth of orchid Mu ngs to, 368 816 Super optimal and thermal death tem- peratures of the cotton plant as affected by variations in relative humidity, Swertia, 138; — 162; asclepiadea, wnis, 140; cornioulata, 162; 140; deflexa, 162; acilis, M ichaun ciana, 14, y angusti- ] t lali 142; pingfolia, 185; plantaginea qu adricornis, urva, 161; R4 LA ln 191 Symp oriearpos rotundifolius, 797 Synechoblastus, 413 e T — 646; cordata, 686, glauces- 686; lasiocarpa, 750; Lo bbiana, 750; poliidifora, 764; palustris, 764; ndi or 744; tomentosa, 764; redi 1, uds chamber for treating cotton plants, 585, 586 ny st Ar bay reaction of Endomyces capsulatus to, 500; Super optimal and thermal dead: of the cotton plant as affected by variations in relative hu- midity, vana. 138; alatus, 139; Bren- tonianus, 167; brevicornis, 140; Peta -" " 178; pi dn 163, var. Brentoni- var. r heterantherus ayt p 163; longicorn 178; meri, 171; rein florus, Lu ps hc m 140; gineus, 175; DEN 161; Sc Phiede- anus, 172; Wh T! Report of ^ botanical bei o the moun "^ ' of western, 7 mess new or r NUN + of, 795, additional MN. of interest, 802 Thamnolia — 14 Thelypteris, 66; onoptera, 68; marginalis, 67; E pubescens, 66; spinulosa, 67, var. intermedia, 67; Thel- ris, 66 Thermutis, 403; velutina, 403 Tillandsia recurvata, 801 Trachelopsermum stans, 659 Tradescantia Wrightii, 801 Trentepohlia, 396 U Usnea wr z comosa, 9, var. stuppea, 9; dasypoga, 9, var. subsca- rede hirta p a e, 9; M in cu: a, 9; pro ostrata, 9 rugulosa milis, 9; o tain sa, 9 ule, 251 (Vou. 20 ANNALS OF THE MISSOURI BOTANICAL GARDEN V UM texana, 801 uelinia angustifolia, 800 Ventures sinensis, Verbena, A revision of the North Ameri- can species. of, Verbena, 251; affinis, 257; Alopecurus, 285; brosse 326, i eglandulosa, amoena, 341; An t Aubletia var. Lamber 342; bipinnatifida, 323, v 325; eti da ta, 268; bonariensis, 254; onariensis var. litt ar. DING 253; bracteata, "304, $58; 304, var. albiflora se: Hearn PRA 304; raat lands 255; brevibracteata, 338; anadensis, 315, nadensis var. Zone rgii, 318, subsp. elegans, 318, var. Lamberti, 316; canescens, 301, var. Roemeri riana, 302; canescens, 304, var 296; caracasana, 257; polystachya, 268, f. or E 33 e flor ora, 342; 207, ed punica 278, f. mite 278, suta, 268; imbricata, 304; 2; i wa, 338; UE 342; Lamberti, 316, var. rosea, 316; lastostachys, 290; d 205; lilacina, 340; litoralis, 257 rosella, 288; macrodonta, 289, 358, 360; maritima, 320; Ma ines 342; menithsefalla, 263; eras 208, 280; mexicana, 296, v . 298, rt xylopoda, 297; T 315; Obletia, 315; officinalis, 262; officinalis var. hirsuta, 296; 9; tifida, ak UP ot 268; prostrata, 290; 1933] INDEX 817 bera, AE puce 332; pumila, 338; pumila albiflora, 318; quadrangu- laris, i: quadrangulata, 313; race- repens, 342; rigens, 286; rigida, 252; riparia, 267; ' robusta, 292; Roemeriana, 302; rubra, 316; rudis, 304; rugosa, 263; simplex, Ran. sphaerocarpa, 262; squarrosa, ad str iste, a 5; Wrightii, 328, 358; Xo. W Woodsia, 68; obtusa Woodson, Robert E., Pu Studies in the Apocynaceae. IV, 605; George H. Pring, and. A new nor Nymphaea f tropiodl Africa Woodwardia, 69; pea, 69; areo- Sources of carbohydrate for germina- tion and growth of orchid Lbs. 569 Xanthocephalum, i linearifolium, 25 Xanthoria polycarp Xylose, growth of i chid seedlings in solutions containing, 574, 576 Y Yeast-like organisms, 471; cytology of, 477 Z Zapania bracteosa, 304 Zygadenus elegans, 801 MORE wA Cw Ai Ar reco 3 è = tes EE e Em T x " x Un i DET. me Sy wa A zs XA HERE TU EN f 1 d K3ifiitwo wee tr. (m a A Yes " E ~ J rari Die eer P "n TE E VIE uU XH CMM I IUÉONE zs IE T t n PP pe RS RE EES ne — PR ESS — Jj T NEM SUES UT EE EON Il A? PRT? EE er oA QM 8 Pee), 2 _ OF THE MISSOURI BOTANICAL GARDEN . = È à ; : L ^ : : .. GEORGE T. Moogs. ' Assistant to the Director, : Karnerine H. LE:GH. | TIR VON Bonmewx, di Carnot W. Dover, : pagine. j i4 57 "oMyeologist. © ~ JESSE M. es aie. Rowann V. La Ganon, Curator of the Herbarium, — Research Assistant. ge S, Revi, | 1t. RoBERTE. WooDsoN, JR., Physiologist: ce oe eee Assistant. = Newt C. Horner, - Librarian me Editor of Publications, J^ BOARD OF TRUSTEES $ E THE SECUR BOTANICAL GARDEN President, GxoncE C. HrrCHCOCK. - Vice-President, Samuzt C. Davis. SSS o ded : Second. Vice-President, ue eee E BON E Case 3 L. Ray ore o E -— ALBERT T. Paniers. | | THowAs B. Mayers. ~~ EUGENBPETPUM i GroRaE T. MOORE. . Ersan A. H. Buspust. Ande P TORNA Frep G. ZEIBIG. SS es BX-OFEICIO MEMBERS: 3 Gana Paucar, Bernard F, Dickmann, SS Chancellor of Washington University. ZE Mayor of the Clty of Bt: Louis. Eo Me Fumpzatcx F. JOHNSON, - SH SCC Be ALBERT Kuntz, . 1:5... Bishop of of the Diocese of Missouri. | President of The Academy of sd ric eor Se. ‘Louis. - X m Davi» C. Tob itn oto nds tvm of bobus : Gron F. Barrmnan, Secretary,