LIBRARY OF THE UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN cc f c 580.5 FB 33-35 BIOLOGY The person charging this material is re- sponsible for its return to the library from which it was withdrawn on or before the Latest Date stamped below. Theft and underilnlna of book. or. . To renew call Telephone Center, 333-S4OO To renew ca UN,VERS,TY OF .Ll.NO.S L.BRARY AT URBANA-CHAMFA.GN L161— 0-1096 0 FIELDIANA Botany Published by Field Museum of Natural History / I - Volume 34, No. 8 October 23. 1972 Tropical American Plants, XII Louis O. WILLIAMS CHAIRMAN, DEPARTMENT OF BOTANY FIELD MUSEUM OF NATURAL HISTORY The manuscript for the second half of part IX and the first half of part X of the "Flora of Guatemala," the Labiatae through the Acanthaceae, will soon be ready for press. A portion of this pre- liminary paper has to do with two large families which I have pre- pared for the "Flora," while the rest has to do with the study of miscellaneous groups of plants which I saw during the year. In addition to our own collections of the Labiatae and Scrophu- lariaceae, I have seen those in the Smithsonian Institution, in the Texas Research Foundation, in the Gray Herbarium, and Arnold Arboretum. Special collections have been borrowed from the Uni- versity of California, the University of Michigan, Museum d'His- toire Naturelle, Institute of Jamaica, and the University of Wiscon- sin, and have made possible some of these notes and observations as well as the floristic work to follow. Special collections sent by Antonio Molina R., Thomas MacDou- gall, Rolando Tun Ortiz, Eizi Matuda, George Proctor, and Roy Lent have been helpful. The illustrations were prepared by Miss Marion Pahl. The continuing support of the National Science Foundation has been important to us at Field Museum who are working on the flora of Guatemala and of adjacent tropical areas. It is greatly apprecia- ted. ARALIACEAE Oreopanax ripicola L. Wms. sp. nov. Arbores debiles usque ad 12 m. vel ultrae, glabrae. Folia simplices petiolis longis, laminae ovatae vel late ovatae, breviter acuminatae, subcoriaceae; inflores- Library of Congress Catalog Card Number: 61-17730 Publication 1154 101 ..« o «nio 102 FIELDIANA: BOTANY, VOLUME 34 c?ntiae terminales paniculatae usque ad 25 cm. longae, capitula 5-15-flora, sessiles; styli plerumque 10. Small weak trees to 12 m. or more, glabrous throughout, the branches terete, smooth or rugose when dry. Leaves simple, long petiolate, the petioles to 20 cm. long, the lamina ovate to very broadly ovate, short acuminate, subcoriaceous, entire, 15-25 cm. long and 9-15 cm. broad, primary nerves 3-5 from the base, secondary nerves spreading at a narrow angle, base of the lamina obtuse or some- what acute; inflorescences terminal on the branches, paniculate, 15-25 cm. long, the lateral branches 15 and about 5-15 cm. long, the heads of pistillate flowers densely to somewhat openly arranged on the lateral branches of the inflorescence, each 5-15-flowered and sessile, flowers unknown; the fruits densely crowded into heads, each subglobose, fleshy but rugose when dry, white when mature, 3-4 mm. in diameter; the styles usually 10, free to the base and borne from the apical cavity of the fruit, reflexed. Guatemala: tree 12 m. or more, at edge of small river, cut-over mixed forest in hills about 3 km. west of San Julian, municipio of Tactic, Alta Verapaz, at about 15° 20' N, 90° 15' W, alt. 1,600 m., February 1, 1969, Williams, Molina, Williams & Molina 40395 (type F; EAP, US). Allied to Oreopanax capitatus (Jacq.) Dene. & Planch., 0. san- derianus Hemsl., and perhaps to the Mexican 0. platyphyllus Mar- chal which I have not seen. Dr. A. C. Smith indicates (Araliaceae in No. Am. Fl. 28B: 29-41. 1944) that the last of these is the only one in which the heads of flowers are sessile, but that species has tridentate leaves and apparently differs also in other details. ASCLEPIADACEAE Cyanchum liesneriana L. Wms. sp. nov. Lianae parvae usque ad 1 m.; caules graciles leviter flexuosi glabrescentes. Folia lineari-lanceolata vel lanceolata acuminata ciliata, petioli graciles breves; inflorescentia axillaris subumbellata, pedunculi 1-2 mm. longi, pedicelli 2-2. 5 mm. longi; calyx 5-lobatus, lobi angusti-ovati obtusi cucullati glabri; corolla cam- panulata, lobi lati-lanceolati extus glabri intus dense barbellati; coronae lobi lineari leviter complanati gynostegio excedentes. Small twining vines to 1 m. long, the stems slender, somewhat flexuose, up to 1 mm. in diameter, the nodes about 3 cm. long, puberulent and soon glabrous. Leaves equal, bicolored, linear-lanceolate to lanceolate, acuminate, ciliate but otherwi.ee glabrous, with 10 or more pairs of inconspicuous nerves, blades mostly •J •'!.") cm. long and 0.3-0.8 cm. broad, petioles slender, short, mostly 3-5 mm. long, usually with a single digitiform gland at the apex; inflorescence axillary, sub- umbellate, 2-6-flowered, the peduncles 1-2 mm. long, bracteoles minute, pedicels •J :2..~> mm. long; calyx deeply 5-lobate, the lobes narrowly ovate, obtuse, cucullate, glabrous, about 1 mm. long; corolla 5-lobate to below the middle, campanulate, about 2.5 3 mm. long, the lobes broadly lanceolate, acute, reflexed, about 1.5 mm. WILLIAMS: TROPICAL AMERICAN PLANTS, XII 103 FIG. 1. Cynanchum liesneriana. A, subumbellate inflorescence, X 6; B, flower partially dissected, X 12; C, inner face of petal to indicate direction of pubescence, X ± 8. Drawn from the holotype. long and 1 mm. broad, glabrous outside, densely barbellate on the inner face; gynostegium about 1.2 mm. long, very short stipitate; anthers about 0.5 mm. long; corona scales linear, somewhat flattened, acute, about 1.5 mm. long, slightly longer than the gynostegium; follicles unknown. Costa Rica: liana with white flowers and white latex, climbing, to 1 m.; on tuffa outcrop, 11 miles south of Liberia, vicinity of Canas, Prov. Guanacaste, Dec. 12, 1969, Daubenmire 406 (type, F). Closely allied to Cynanchum stenomeres Standl. & Steyerm., an endemic of British Honduras and one of the three lowland species now known of the "Metastelma group" to be found in the Central America-Panama region. It is distinguished by the broader calyx lobes, by the smaller corolla with relatively broad lobes, and the shorter and coarser corona lobes. The species is named for my colleague Ronald Liesner, who has named most of Mr. Daubenmire's large and important collections from Costa Rica. Mr. Liesner called the plant to my attention when he was unable to match it in the herbarium. Marsdenia agnota L. Wms. sp. nov. Lianae. Folia suborbiculari-ovata acuminata cordata subtus sparse tomen- tulosa super sparse hirsuta ciliata; inflorescencia cyma paniculata vel racemiformis multiflora; calyx 2.5-3 mm. longus, lobi elliptico-ovati acuti hirsuti; corolla cam- 104 FIELDIANA: BOTANY, VOLUME 34 panulata subcarnosa, lobi ovati aut oblongo-ovati reflexi; folliculi dense tomentosi agnoti (licit ur. Vines of unknown size, the stems about 5-6 mm. in diameter, densely puberu- lent and sparsely hirsute with erect segmented hairs. Leaves of a pair unequal, suborbicular-ovate, acuminate, deeply cordate, the sinus 3-5 cm. deep, the blade 10-24 cm. long and 8-20 cm. broad, short and sparsely tomentulose below, sparsely short hirsute on the margins and upper surface, petioles 3.5-8 cm. long, pubescence same as on stems; inflorescence axillary, a paniculate or racemiform cyme, about 4 cm. long (one seen) with about 30 flowers; calyx about 2.5-3 mm. long, divided to the base into 5 lobes, the lobes elliptic-ovate, acute, appressed hirsute pubescent with segmented hairs, about 2.5 mm. long and to 2 mm. broad, corolla somewhat fleshy, campanulate and constricted at the base around the ovary, about 5 mm. long, the tube about 3 mm. long, provided with a ring of inflexed hairs in the throat, the ovate or oblong-ovate lobes reflexed and about 2 mm. long and 1.5 mm. broad; gynostegium about 2 mm. high and 1 mm. broad, provided with erect diaphanous stigmatic processes about 1 mm. high, pollinia about 0.6 mm. long and 0.3 mm. broad; follicles 8-10 cm. long, densely covered with dense white tomentum said to resemble a lamb's wool, about 2-3 mm. deep; seeds not seen but in photograph with an apical coma. Mexico: Nom. local: borreguito, pastorcito en Guerrero. Hoja y flores [vienen del] Canon de Zopilote, Guerrero, Agosto 30 de 1970, alt. 500-700 m.; fruto [viene de] Quechultenango, Guerrero, enero 1969, Erich Halbinger y E. Matuda 38052 (type F; isotype MEXU.) ; vine climbing on small trees, swampy, open woods, Hacienda Cal- ifornia, Municipality Apazinga, Tancitaro region, Michoacan, alt. 400 m., August 1, 1941, Leavemvorth & Hoogstraal 14.26 (F). Closely allied to Marsdenia tubularis L. Wms. also of Central Mexico, but quickly distinguished by the campanulate corolla with broader calyx lobes as long as the corolla tube; the leaves relatively much larger (twice as large) ; the follicles are densely tomentose, not simply pubescent. The Leavenworth & Hoogstraal specimen cited has an immature follicle but no flowers. The leaves are almost identical to those of the type and the follicle is provided with the deep tomentum that I know on no other species. Dr. Standley had attached an unpub- lished name to the specimen. The specific name is derived by Latinizing the local name, bor- reguito. Professor Matuda advises that the specimens were tagged in the field, by Mr. Halbinger, to assure that flowering and fruiting speci- mens were the same. x- "8 X 105 106 FIELDIANA: BOTANY, VOLUME 34 LABIATAE CATOPHERIA Bentham. The type species. The genus Catopheria Bentham was described in Bentham & Hooker, Genera Plantarum 2: 1173. 1876. Bentham mentioned that there were three species in the genus, natives of Colombia, Central America, and Mexico. No specific name is mentioned as is usually the case with genera described in the Genera Plantarum. A refer- ence is made to De Candolle's Prodromus where Catopheria is given as a section of Orthosiphon. The next mention of the genus Catopheria by Bentham was in Hooker's /cones Plantarum 13: 13, t. 1215. 1877 where Orthosiphon spicata Benth. is properly transferred to Catopheria, described and illustrated. The next reference to Catopheria, of importance, is Carl Epling's note in Feddes Rep. Sp. Nov. Beih. 85: 320-321. 1937 where he in- dicates that Catopheria capitata should be generic type. This desig- nation has been followed in Index genericorum. There is a possibility that Catopheria capitata Benth. ex Hemsl. and C. spicata Benth. do not belong in the same genus and conse- quently the selection of a type is important in deciding with which species the generic name will remain. Epling chose Catopheria capitata Benth. ex Hemsl. as the type species. However, Bentham, so far as I am able to find, never men- tioned this name in his own writings. He described the genus Catopheria in 1876 and a few months later in 1877 actually made a combination as C. spicata Benth. It seems reasonable to me to assume that Bentham selected C. spicata as representative of Catopheria and I believe that it should be considered to be the type species of Catopheria. The generic description of Catopheria in the next part of the "Flora of Guatemala" will be based on C. spicata Benth. Hyptis americana (Aubl.) Urban, Fedde Rep. Sp. Nov. 15: 322. 1918. Nepeta americana Aubl. Hist. PL Guian. Fr. 2: 623. 1775. Hyptis spicigera Lam. Encycl. 3: 185. 1789. Non Hyptis americana sensu Briquet vel Epling. This combination for a common tropical American plant has been rejected by Dr. Epling (Rev. Museo de la Plata Bot. 7: 255.1949) because of a presumably older use of the name "Hyptis americana." WILLIAMS: TROPICAL AMERICAN PLANTS, XII 107 Dr. Briquet (in Engler & Prantl, Nat. Pflanzenfam. IV. 3a: 338. 1897) in the account of Hyptis wrote "H. americana (Poit.) Briq. (H. scoparia Poit.)." So far as I am able to find there is no plant described as "americana" by Poiteau that might be assumed to be the basionym of "H. americana (Poit.) Briq." and Dr. Briquet gave none. The name "H. americana (Poit.) Briq." is illegitimate since it has no basis. If Dr. Briquet intended to make a new name "H. americana Briq." then this also would be illegitimate for there was given as a presumed synonym H. scoparia Poit. If Dr. Briquet in- tended to write "H. americana (Poir.) Briq.," as Dr. Epling(supra) apparently assumes he did, then there is a possible basionym— Satureja americana of Poitet. Since Hyptis americana (Poit.) Briq. cannot be shown to be based on any species described by Poiteau, who wrote a monograph of Hyptis (1806) ; since a new name by Briquet would be illegitimate; and, since the substitution of an entirely different parenthetical authority, Poitet, in the combination seems not justified, I am con- sidering Hyptis americana sensu Briquet and later of Epling to be a nomen illegitimum et confusum. Specimens annotated by Dr. Epling as H. americana should be called Hyptis scoparia Poit. while those annotated as H. spicigera Lam. should be called Hyptis americana (Aubl.) Urban. Hyptis scandens Epling, Carnegie Inst. Wash. Publ. 522: 239. 1940. The type, Lundell 4421 from El Paso, Pete"n, which Dr. McVaugh has been kind enough to loan to me, is inadequate but I presume that the plant is a Hyptis and is like no other one known to me. Another inadequate specimen from near the type locality is doubtless the same, — Steyermark 46305. OCIMUM. The genus Ocimum contains perhaps as many as 40 or 50 species, principally of Asia and Africa with relatively few of the species in the American tropics. At least one species, 0. basilicum L. the common basil, is widely cultivated. Whether basil is actually grown in Central America, I do not know. We have a number of specimens from Mexico, presumably not from cultivation, that have been considered to be 0. basilicum. It seems likely that they may not be this species, but perhaps are a native species. Until someone can find the time to revise this confusing group of plants perhaps it is just as well to leave them as 0. basilicum. The situation today is 108 FIELDIANA: BOTANY, VOLUME 34 not unlike that described by Bentham 140 years ago (Benth. Labiat. Gen. et Sp. 5-6. 1832). The common Ocimum of Mexico and Central America is 0. nn'cranthum Willd. described in 1809. It is exceedingly abundant from Mexico and Guatemala to Honduras, less common on to Panama and in South America. It is often found in over grazed places and old cultivated fields and might well be called a weed. Ocimum selloi Benth. (Labiat. 6. 1832) was described from Brazil and Mexico. Since Sello collected one of these specimens it is logical to consider the Brazilian one as type. The Sesse" & Mocifio specimen cited was from Berlin Herbarium and undoubtedly is lost. We have specimens of four Sesse" & Mocifio collections that seem to be this species but, whether one is the same collection as Bentham saw, I do not know. There are a number of collections in our herbarium from Mexico that I cannot distinguish from material from southern Brazil, Paraguay, and Argentina. If all the material is, in fact, one single species, the range is a most curious one. Ocimum selloi is closely related to 0. micranthum but can be distinguished easily by the peculiar small round glands subtending the segments of the ver- ticels. It is curious that Bentham did not mention these glands for they are quite clear in the phototype or photoisotype (F 17813) from Berlin. Salvia alariformis L. Wms. sp. nov. Frutex altitudine usque ad 2 m. ramulis pilosis, internodiis 2-5 cm. longis. Folia lanceolata vel lanceolato-ovata acuminata serrata sessilia vel breviter petiolata subtus tomentosa vel dense pilosa super sparse pilosa; inflorescentiae alares vel alariformes pauciflorae; calyx venosus piloso-hirsutus; corolla caerulea bilabiata. Shrubs to 2 m. tall, the branches terete, densely pilose with short erect seg- mented hairs, the internodes 2-5 cm. long. Leaves lanceolate to lanceolate-ovate, acuminate, serrate, sessile to short-petiolate, tomentose or densely pilose below, sparsely pilose above, 3-9 cm. long, 1.5-3 cm. broad; inflorescences axillary or borne on short branches from axils of leaves, 1-4-flowered, shorter than or subequal to the subtending leaf; flowers blue, few; corolla bilabiate, tube about 15 mm. long, the lips subequal or the posterior one slightly longer, 9-10 mm. long, posterior lip galeate, pubescent outside, anterior lip trilobate, broad, slightly shorter than the posterior; calyx at an thesis about 7-8 mm. long, becoming to 12 mm. long post- anthesis, 11-nerved, the nerves sparsely pilose-hirsute with flattened hairs; anthers included in the galea; style bifurcate, one lobe much reduced. Guatemala: corolla dark purple, Cerro Tixixi, 3-5 miles north of Jocotan, Dept. Chiquimula, alt. 500-1,500 m., Nov. 10, 1939, Steyermark 31644. WILLIAMS: TROPICAL AMERICAN PLANTS, XII 109 Honduras: woody herb 6-8 ft., flowers dark blue, common, Barranco de Trincheras, Dept. Comayagua, alt. 1,050 m., April 17, 1951, Allen 6210; flores azules, arbusto 1-2 m., poco frequente, bosque mixto y humido cerca de Dulce Nombre, Dept. Copan, alt. 1,200 m., Marzo 30, 1963, Molina 11735; flores azul-violaceo, planta 1-2 m., Barranco Trincheras 3 kms. a Montanuela, Dept. Comaya- gua, alt. 1,200 m., Marzo 28, 1964, Molina 13641 y 13656; flowers blue, woody herb 1-1.5 m., common, cut-over pine-oak forest, La Montanita, Cordillera Merend6n, Dept. Ocotepeque, alt. 1,600 m., Aug. 31, 1968, Molina 22562; flowers dark blue, shrub 1-2 m., common in ravine, cut-over mixed forest of Barranco Trincheras, 16 km. to Siguatepeque, Dept. Comayagua, alt. 1,300 m., March 11, 1970, Molina & Molina 25483; flowers royal blue, herb 1-1.5 m., common in the forest, vicinity of El Carrizal 14 km. northwest of Siguatepeque, Dept. Comayagua, alt. 900 m., June 27, 1971, Molina & Molina 26056; flowers blue, shrub 1 m., edge of mesophytic forest 10 miles N. W. of Siguatepeque, alt. 800 m., Dec. 26, 1946, Williams & Molina 11459 (type, F; EAP); flowers blue, shrub 2 m., in broad leaf forest in Barranco Trincheras, Dept. Comayagua, alt. 1,200 m., April 30, 1947, Williams & Molina 12549. Two additional collec- tions are in EAP according to Ing. Molina, they are Williams and Molina 14649 and 17995. The species is unusual in that the inflorescences are borne in the axils of stem leaves or are borne on short axillary shoots. No other blue-flowered Salvia of this region with this kind of inflorescence is known to me. Most all specimens have been determined as Salvia comayaguana Standl. The range from Guatemala to Honduras is one shared by many species of plants. Salvia areolata Epling, Bull. Torr. Bot. Club 71: 493. 1944. This species is locally common in the tops of the Sierra de los Cuchumatanes north of Huehuetenango, Department of Huehue- tenango, at some 3,000-3,300 m. The type specimen, Steyermark 50338, is immature and consequently the original description is not good. The description in "Flora of Guatemala" will be based on several mature collections: Molina 21249; Molina, Burger & Wal- lenta 16506; Williams, Molina & Williams 21915, 22003. Salvia cacaliaefolia Benth. in DC. Prodr. 12: 348. 1848. S. hempsteadeana Blake, Proc. Biol. Soc. Wash. 32: 197. 1919. S. mendax Epling, Fedde Rep. Sp. Nov. Beih. 110: 96. 1938. 110 FIELDIANA: BOTANY, VOLUME 34 A beautiful large flowered Salvia which was suggested as possibly useful as a bedding plant for gardens in southern Florida and southern California by Dr. Wilson Popenoe long ago. Dr. Epling has placed Salvia cacaliaefolia and S. mendax in dif- ferent sections in his "Revision of Salvia, subgenus Calosphace," but I find no difference in these plants. The geographic distribution in the highlands of Chiapas (Mexico) across Guatemala to western Honduras is a natural one. Salvia excelsa Benth. in Lindl. Bot. Reg. 27: 90. 1841 (type from seed collected in Guatemala by Hartweg). S. nervata Mart & Gal. Bull. Acad. Brux. 11, pt. 2: 77. 1844. S. monocheila Donn.-Sm. Bot. Bot. Gaz. 23: 13. 1897 (type from Huehuetenango, Nelson 3635}. This is a common plant in the highland forest area of western Guatemala and is occasional in southern Mexico. Dr. Epling has maintained both Salvia nervata and S. excelsa in his revision of Salvia (Fedde Rep. Sp. Nov. Beih. 110. 1939). Several dozen recent collec- tions from Guatemala indicate that the two cannot be separated. The specimens are usually found in herbaria as S. nervata. Salvia fracta L. Wms. sp. nov. Herbae fractae fragiles usque ad 60 cm. altae. Folia ovata vel leviter cordata pilosa acuminata aut acuta; inflorescentiae spicatae, verticillastri compacti aut laxi 8-12 flori; bracteae subreniformes venosae serrulatae; calyx campanulato-tubularis sparse pilosus trilobulatus, lobi angusti triangulates perbreves; corolla bilabiata, labia superiore perbrevis, labia inferiore trilobulata lobi laterales dentiformes lobus terminate ovatus retusus. Small, weak, and fragile perennial herbs 30-60 cm. tall, stems simple or sparsely branched, quadrate and prominently angled, densely pilose with retrorse crisped hairs above, soon glabrate. Leaves ovate to shallowly cordate, shallowly serrulate, acute or acuminate, sparsely pilose on both faces with appressed crisped hairs, the blade 1.5-3.5 cm. long and 1-3 cm. broad, petiole 1-1.5 cm. long, pilose; inflores- cence a spike with 3-several verticels, compact or open, the verticels each with 8-12 flowers, subtended with opposite persistent bracts, the bracts subreniform, short acuminate, sessile, venose, serrulate, sparsely pilose outside, glabrous within, about 8 mm. long and as broad or slightly broader; flowers blue, very small; calyx campanulate-tubular, sparsely pilose along the nerves, trilobate, the lobes narrowly triangular, the dorsal about 0.7 mm. long, the laterals slightly longer, the tube 3-3.5 mm. long; corolla bilabiate, the tube 3-3.5 mm. long, the superior lip nar- rowly triangular, about 1 mm. long, the inferior lip about 2-2.5 mm. long, trilo- bulate, the lateral lobules minute and dentiform, the terminal lobule transversely oval, retuse, about 2-2.5 mm. broad; stamens 2, included, the anthers about 0.5 mm. long; style bifid; seeds not known. WILLIAMS: TROPICAL AMERICAN PLANTS, XII 111 Guatemala: flowers blue, on moist bank, wet mountain forest near Aldea Fraternidad between San Rafael Pie de la Cuesta and Palo Gordo, west facing slope of the Sierra Madre Mountains, de- partment of San Marcos, alt. 1,800-2,400 m., December 10-18, 1963, Williams, Molina & Williams 26069 (type, F). A species belonging in the section Membranaceae of Dr. Epling's revision of the subgenus Calosphace of Salvia and related to the complex surrounding Salvia mocinnoi Benth. It is a smaller, more delicate plant than 5. mocinnoi, with leaves rounded at the base or even shallowly cordate, the flowers about half as large, the bracts subtending the verticels serrulate. Only the collection cited is known. It was distributed as S. mocinnoi. Salvia gracilis Benth. Lab. Gen. & Sp. 258. 1833. S. mem- branacea Benth. I.e. 259. S. pseudogracilis Epling, Fedde Rep. Sp. Nov. Beih. 110: 229. 1939. S. myriantha Epling, I.e. 248. S. natalis Epling, I.e. 265. The five specific names given above were all maintained by Dr. Epling in his "Revision of Salvia, subgenus Calosphace" and placed in three different sections. I find no satisfactory differences to dis- tinguish them. In addition to these names, which have been used with Guatemalan specimens, there are others which may be found in Epling's revision under S. gracilis and S. membranacea. Salvia hyptoides Mart. & Gal. Bull. Acad. Brux. 11, 2: 74. 1844. A semi-weedy plant often to be found in fields. It is distributed from central and southern Mexico through Central America and Panama to northern South America. The plant is a variable one, especially with relation to pubescence of the calyces and leaves. Dr. Epling in his revision (Fedde Rep. Sp. Nov. Beih. 110: 143-146. 1938) has built a key to the related annual species that is based on pubescence of the calyces and the leaves and whether the verticels of flowers are close together or remote. It is doubtful if these char- acters are of specific value and S. hyptoides M. & G. (1844) is atbesta variety of Salvia lasiocephala Hook. & Arn. (1841). The Mexican Salvia galinsogifolia Fernald, is a form of S. lasiocephala with almost or quite glabrous leaves. Salvia misella HBK. Nov. Gen. & Sp. PI. 2: 290. 1817; Epling, Fedde Rep. Sp. Nov. Beih. 110: 16. 1938. S. riparia HBK. I.e. 300, syn. nov.; Epling, I.e. S. privioides Benth. Bot. Voy. Sulph. 150. 1844. 112 FIELDIANA: BOTANY, VOLUME 34 Salvia misella, as I interpret it, is a weedy species that is dis- tributed from Mexico and the West Indies south to Peru. It is found more often than not in door yards and in disturbed and cultivated places. Dr. Epling in his "Revision of Salvia, subgenus Calo- sphace," cited above, has maintained S. misella and S. riparia and distinguished them by length of the calyces, those of S. misella being 3.5-5 mm. long while those of S. riparia are 5.5-7 mm. long. I find that this character overlaps, sometimes on a single specimen, and I have found no other way to distinguish the species. Salvia misella is closely related to S. occidentalis Sw. and perhaps hardly distinct, as Bentham has pointed out well more than a century ago (Lab. Gen. & Sp. 720. 1835). It is a weedy species occurring in the same kinds of habitats as does S. misella. There are other species of Salvia that belong here as synonyms but no other names have been used with Guatemalan material. Salvia mocinnoi Benth. Lab. Gen. & Sp. 271. 1833. S. mocinoi Benth. in DC. Prodr. 12: 300. 1848; Epling in Fedde Rep. Sp. Nov. Beih. 110: 151. 1938. S. lophantha Benth. in DC. Prodr. 12: 301. 1848; Epling I.e. 150. Epling has kept these two species separate in his "Revision of Salvia, subgenus Calosphace." The large mass of specimens collected since his revision seem to indicate that there is no way that these species can be separated for the guide lines given by Epling do not hold. Salvia rubiginosa I am keeping in "Flora of Guatemala" with some reluctance. The single specimen from Guatemala, Skutch 1728, which I believe to represent this species might almost as well be placed in S. mocinnoi. The specific name originally given by Bentham was "mocinni," an attempt to retain the "n" in Mocino's name. In DeCandolle's Prodromus Bentham changed the name to "mocinoi" and this has been mostly followed since that time. We will use "mocinnoi" to retain the sound of the "n" and using the full Latin name Mocinno adding an "i" as suggested by article 73 of the rules. Salvia polystachya Ortega, Hort. Matr. Dec. 55. 1797. S. compacta Kuntze, Rev. Gen. 530. 1891. S. polystachya var. albicans Fernald, Proc. Am. Acad. 35: 512. 1900. S. reducta Epling, Fedde Rep. Sp. Nov. Beih. 110: 168. 1939. Salvia polystachya is an attractive and often abundant plant which ranges from central Mexico through Central America to Panama. WILLIAMS: TROPICAL AMERICAN PLANTS, XII 113 Dr. Epling in his "Revision of Sal via, subgenus Calosphace," 1939, has maintained S. compacta and S. reducta Epling, the first with a range which includes Guatemala and the second described from Gua- temala. I find no reason to segregate them from S. polystachya and have reduced them in the manuscript for "Flora of Guatemala." The synonomy of S. polystachya, through its range, is more complex than that indicated above. SALVIA TILIAEFOLIA AND ITS VARIATIONS IN CENTRAL AMERICA AND MEXICO have been unduely complicated by maintaining three of these plants as distinct species, — Salvia tiliaefolia, S. albiflora, and S. alvajaca. I believe that they represent in Guatemala, at best, three not very different varieties of a single species as indicated below. Salvia tiliaefolia var. tiliaefolia. Salvia tiliaefolia Vahl, Symb. Bot. 3: 7. 1794; Epling, Fedde Rep. Sp. Nov. Beih. 110: 241. 1939. The common form of this small flowered Salvia in Mexico and Central America — distributed as far south as Ecuador. The dis- tinguishing characteristic of this variety is the hispid pubescence on the veins of the calyx. The variety as I understand it is almost exactly as Dr. Epling accounted for it in his revision cited above. Salvia tiliaefolia var. albiflora (M. & G.) L. Wms. comb. nov. Salvia albiflora Mart. & Gal. Bull. Acad. Brux. 11, pt. 2: 76. 1844; Epling, I.e. 254. S. jurgensenii Briq. Ann. Conserv. Jard. Bot. Geneve 2: 144. 1898. This variety is common in Mexico and less so in Central America and reaches to Panama. The only distinguishing character of this rather weak variety is the lack of hispid pubescence on the calyx— and the calyx rarely more than 5 mm. long. Dr. Epling in his revi- sion places this variety in a different "series" from S. tiliaefolia. The flowers may be white or blue. Salvia tiliaefolia var. alvajaca (Oersted) L. Wms. comb. nov. Salvia alvajaca Oersted, Dansk. Vid. Medd. 1853: 38. 1853. This variety, and two additional names proposed by Dr. Briquet (S. ourophylla Briq. and S. permixta Briq.) were described on material from Costa Rica. It is known in Guatemala and Panama as well. It is distinguished from var. tiliaefolia in having the calyx glabrous or slightly hirtellous (not hispid on the veins). The calyx is 7-8 mm. long or perhaps slightly more in fruit and by this greater size is dis- tinguished from var. albiflora. 114 FIELDIANA: BOTANY, VOLUME 34 Scutellaria dumetorum Schlecht. Linnaea 7: 400. Before Feb. 1832. S. caerulea Moc. & Sesse" ex Benth. in Lindl. Bot. Reg. 18: sub t. 1493. May 1, 1832; Epling, Univ. Cal. Publ. Bot. 20: 106. 1942. S. chalicophila Loesener, Bull. Herb. Boiss. 7: 569. 1899. S. affinis Leonard, Contr. U. S. Nat. Herb. 22: 741. 1927. Dr. Stafleu, Reg. Veg. 52: 420. 1967, states that volume 7 of Linnaea through page 688 was published "ante Feb. 1832." There- fore Scutellaria dumetorum must be used, instead of the better known 5. caerulea, for the plants of Mexico and Guatemala. TEUCRIUM Dr. George R. Proctor made two collections of Teucrium near the falls of the Rio de las Violetas about 2.5 miles north of Nebaj, Department of Quiche". One of these, no. 25499, was the widespread Teucrium vesicarium Mill, which ranges from southern Mexico and the West Indies to Central America and southward to Argentina. The other collection, no. 25498, was collected at the same place but is a very different species, although more closely allied to T. vesicarium than to other species. The vesicular calyx, inflated after anthesis, is less prominent than in T. vesicarium; the corolla tube is much longer than the calyx, in anthesis nearly twice as long; the inflorescence is much more open; the pubescence on all parts of the plant is much less; the plant seems to be much larger. Teucrium proctori L. Wms. sp. nov. Herbae altae, caules quadrati et subalati leviter pubescentes. Folia ovata acuminata basi obtusa vel truncata serrata petioli graciles puberuli; inflorescencia spicata terminalis; flores subsessiles breviter pedicellati; calyx tubularis, 5-den- tatis, puberulus; corolla tubularis bilabiata quinquelobata; stamina 4, didynami; stylus usque ad 4 mm. longus; stigma arcuatum verrucosum. Tall, herbaceous plants of unknown duration, the stems prominently quadrate and subalate at the angles, sparsely pubescent, especially on angles, the internodes 7-9 cm. long, somewhat constricted at the nodes; leaves ovate, acuminate, the base obtuse or truncate, serrate, strigillose above, puberulent below, the blade 7-10 cm. long, 3-5.5 cm. broad, petiole 1-2.5 cm. long, puberulent; inflorescences terminal and terminal from axillary branchlets, spicate, rather open, to about 15 cm. long; flowers rose-lilac, subsessile or borne on pedicels to 2 mm. long, subtend- ing bracts linear-lanceolate, acute, about 2 mm. long; calyx 2.5-4.5 mm. long, tubular, 5-dentate, puberulent with short often crisped hairs, the teeth triangular, acuminate, 0.5-0.7 mm. long, the posterior lobe shortest, in fruit vescicular and subglobose, 4-5 mm. long; corolla rose-lilac, the tube about 5-6 mm. long, bilabiate, slightly pilose in the throat, the anterior lip trilobate, the lobes broadly ovate to WILLIAMS: TROPICAL AMERICAN PLANTS, XII 115 rotund, about 1.5 mm. long, the posterior lip bilobate, the lobes rounded, about 0.75 mm. long; stamens 4, didynamous, inserted near middle of tube, the filaments about 1 mm. long, the anthers about 0.5 mm. long; style about 4 mm. long, the stigma arched, verrucose, somewhat thickened; seeds (immature) about 3 mm. long. Guatemala: erect weedy herb, 2 m. high; corolla rose-lilac; near falls of Rio de las Violetas, 2.5 miles north of Nebaj, Dept. Quiche", ca. 5,500 ft., Aug. 11, 1964, Proctor 25^98 (LL, type; IJ). Named for the collector, Dr. George R. Proctor, long a resident of Jamaica and student of the flora of that and other islands of the Caribbean. ORCHIDACEAE Ponthieva trilobata L. Wms. comb. nov. Cranichis trilobata L. Wms. Field Mus. Bot. 31: 258. 1967. Mexico: Breedlove 11856 (type); Roe, Roe & Mori 1088, 1279. This obvious species of Ponthieva was originally placed in Cranichis by me for no obvious reason. RANUNCULACEAE Clematis caleoides Standl. & Steyerm. Field Mus. Bot. 23: 52. 1944. Mexico: heavily wooded slopes with Abies, Podocarpus and Quercus in the Colonia 'Ach'lum, municipio de Tenejapa, Chiapas, alt. 9,100 ft., December 12, 1966, Alush Shilom Ton 1759 (F, DS). This species, originally described from Guatemala, is new to the Mexican flora. ROSACEAE PHOTINIA. There are two species of Photinia in central and western Mexico. The oldest of these was described as Chamaemeles mexicana by Baillon, based on Galeotti 1660. Galeotti says "Fleurs blanches en Aout. Bois pres Vera-Cruz a 3500 pt." The type, which was loaned to me, is in Paris (phototype, F). I have seen only two other collections that belong to this species — Liebmann 1756 and Purpus 2968. The three specimens mentioned are all from the state of Vera Cruz. Photinia parviflora, described below, has been called P. mexicana without exception. It is a species of central Mexico — the states of 116 FIELDIANA: BOTANY, VOLUME 34 Michoacan, Jalisco, and Mexico. It is distinguished from P. mexi- cana in having flowers half as large; smaller leaves that are 2-4 cm. broad and 6-11 cm. long, the margins are entire or at most obscurely serrulate, not crenate, the apex is acute and usually shortly acumi- nate. The fruits of this species are ovoid and 8-10 mm. long while the fruits of P. mexicana are unknown to me. South of the Isthmus of Tehuantepec in Chiapas and on into Central America there are perhaps two or three additional species of Photinia. Photinia microcarpa Standl. is the best represented of these, having been found in Guatemala, British Honduras, and Honduras. Photinia matudae Lundell described from Chiapas is closely related to P. microcarpa. Standley wrote (Flora of Guate- mala, 1946) that more material was needed to determine whether it is really distinct from P. microcarpa. There is another species from Chiapas (Breedlove 9689) with large globose fruits to 20 mm. in diameter and with coriaceous, crenate, oblanceolate leaves (to 13 X 5 cm.). I determined this specimen a few years ago as P. mexicana but it is not. I do not know what it is but surely not one of those previously known. Photinia parviflora L. Wms. sp. nov. Frutices vel arbores usque ad 12 m., ramuli nigrescentes. Folia lucentes, glabra, elliptica, elliptico-ovata vel oblanceolata, acuta vel acuta et breviter acuminata; inflorescentia pauciflora, brevis, ferruginia; calyx campanulatus, 5- lobatus, lobi triangulares, acuti; petala suborbiculares; stamina ±20; stylus bi-vel trifidus; fructus ovoideus, 8-10 mm. longus. Shrubs or small trees to 12 m. tall. The branches at first ferruginous pubescent and brownish, becoming glabrous and black (at least in dry specimens). Leaves clustered near apex of the new growth, 2-10 on each branchlet, glabrous or slightly ferruginous pubescent at the base, shining, the petiole 0.8-20 mm. long, blade elliptic to elliptic-ovate or oblanceolate, acute and apex shortly acuminate, lighter in color below than above, with 10-13 pairs of widely divergent lateral nerves, entire or obscurely serrulate, 6-11 cm. long and (1.5-) 2-4 cm. broad; inflorescence a few-flowered terminal corymb to about 5 cm. long, shorter than the leaves, ferruginous; puberulent flowers small, about 8 mm. broad at an thesis; calyx campanulate, 5-lobate, ferruginous pubescent, from base of ovary about 3.5 mm. long, the lobes triangular, acute, about 1 mm. long; petals 5, suborbicular, lobulate, about 3 mm. long; stamens about 20, attached at the throat of the hypanthium and with bases of filaments joined; style bifid nearly to the base with one branch bicapitate, about 2 mm. long; ovary inferior, arachnoid-pubescent above; fruits red, ovoid, 8-10 mm. long. Mexico: shrub or small tree 6 ft. tall, 3 miles south of Tancitaro, Municipio of Tancitaro, Michoacan, alt. 6,000 (feet), July 19, 1941, Wm. C. Leavemvorth & H. Hoogstraal 1024 (type F); berries red WILLIAMS: TROPICAL AMERICAN PLANTS, XII 117 (edible), shrub or small tree to 15 ft. high, pedregal 3 miles south of Tancitaro, Municipio of Tancitaro, Michoacan, alt. 6,000 ft., July 19, 1941, Leavemvorth & Hoogstraal 104.1 (F); Carniceria, Tomas- caltepec, Mexico, Feb. 20, 1932, Hinton 327 (US) ;Nanchititla, Temas- caltepec, Mexico, 9-1-1933, Hinton 3117 (US); Mina de Agua, Temascaltepec, Mexico, Feb. 26, 1935, Hinton et al. 7233 (US, F) ; Telpintla, Temascaltepec, Mexico, Feb. 16, 1936, Hinton 8918 (US); summit of pass 9 miles from Autlan, Jalisco, alt. 1,200-1,500 m., April 9, 1949, McVaugh 10233 (US). RUBIACEAE Hoffmannia pusilla L. Wms. sp. nov. Herbae pusillae repentes aut subacaulescentes usque ad 6 cm. altae. Folia oblongo-oblanceolata vel oblanceolata acuta leviter bullata nerviis pilosis; in- florescentiae axillares 1-3-florae pedunculi folia subaequilongi; corolla rotata lobi lanceolati; calyx perbrevis lobi lanceolato-triangulares; capsula usque ad 4 mm. longa. Small herbs, repent or subacaulescent. The stems almost none up to about 6 cm. long, pilose with reddish crisped pubescence, becoming glabrous; leaves crowded on the short stem, oblong-oblanceolate to oblanceolate, acute, obscurely bullate, pilose on the nerves and ciliate, raphids abundant on lower leaf surface, upper surface prismatic-cellular, cuneate to the base or with short winged petiole, blade 2-6 cm. long and 0.6-2.5 cm. broad; inflorescences axillary, with a single flower or with two or more flowers, the peduncles usually shorter than the leaves; flowers yellow, small and inconspicuous; corolla rotate, 4-4.5 mm. long, divided almost to the base, the lobes 4, lanceolate, 3.5-4 mm. long and about 1-1.2 mm. broad; calyx lobes lanceolate-triangular, about 0.5-0.6 mm. long, somewhat glandular on either side at the base; anthers 4, attached on tube of corolla, oblong, about 1.4 mm. long; capsule about 4 mm. long. Guatemala: between Sepacuite" and Secoyocte", Alta Verapaz, alt. 1,100 m., May 24, 1905, Pittier 348 (type, US); herb, flowers yellow, Chapultepec Farm on Coban-Sebol road along arroyo in forest, Alta Verapaz, May 25, 1964, Contreras 4793 (LL, fragment F). Related to Hoffmannia refulgens (Hook.) Hemsl., a species of Mexico and adjacent Guatemala, from which it is distinguished in being a much smaller plant — perhaps the smallest of the genus— with smaller yellow flowers instead of red ones. These specimens were received in loans of Napeanthus (Ges- neriaceae) and one, Contreras 4793, may have been distributed as N. apodemus Donn.-Sm. 118 FIELDIANA: BOTANY, VOLUME 34 SCROPHULARIACEAE Alectra aspera (Cham. & Schlecht.) L. Wms. comb. nov. Pedicularis melampyroides L. C. Rich. Act. Soc. Hist. Nat. Par. 1: 111. 1792, non Alectra melampyroides Benth. 1846; Glossostylis as- pera Cham. & Schlecht. Linnaea 3: 23. 1828; Alectra brasiliensis Benth. in DC. Proclr. 10: 339. 1846; Melasma melampyroides Pen- nell, N.Y. Acad. Sci. 6: 188. 1925. Originally described from material collected at low elevations in Brazil. Presumed to be native in the West Indies and southward to Brazil. Not known to be common anywhere in Central America and suspected of being introduced as a weed. The genus is closely related to Melasma. Bacopa parviflora Standl. sp. nov. Herbae annuae parvae erectae usque ad 20 cm. altae ramosae. Folia sessiles, lineari-lanceolata, acuta, 2-3 cm. longa, leviter serrulata vel subintegra, glabra, punctato-glandulosa; inflorescentiae axillares, uni-pluriflorae; flores perparvi, ses- siles aut breviter pedicellati; calyx viridis, 1.5-2 mm. longus, lobi elliptici vel oblongo-ovati, obtusi vel subacuti, glabri, dense punctato-glandulosi, unicostati; corolla perbrevis, 1.5-2 mm. longa. Guatemala: swampy meadows and wet pockets on ridge in pine forest below Finca San Jose1, southeast of Concepcion de las Minas, Dept. Chiquimula, alt. 1,500 m., Nov. 4, 1939, Steyermark 31169 (F). Honduras: open grassy bog, region of Las Mesas, Dept. Mor- azan, alt. 800-900 m., Oct. 30, 1948, Standley 13865 (F); near Suyate in drying bog, Dept. El Paraiso, alt. 650 m., Dec. 31, 1950, Standley 28053 (F) ; moist field, vicinity of Siguatepeque, Dept. Comayagua, alt. 1,080-1,400 m., Feb. 14-27, 1928, Standley 56046 (F) ; edge of pool on savanna near Gal eras, Dept. El Paraiso, alt. 850 m., Nov. 24, 1946, Standley & Williams 83 (F) ; Santa Inez, Dept. Morazan, alt. 850 m., agosto 1943, Valerio 409, 437 (F). Nicaragua: flores blancas, matorrales humidos-pedregosos, bos- ques pantanosos de San Miguelito, drenaje del Lago Granada, Depto. Chontales, alt. 30 m., Nov. 13, 1951, Shank & Molina 4572 (F). Panama: Tecumen River, vicinity of Juan Diaz, Prov. Panama, Nov. 9, 1917, Killip 3240 (type, F). The name Bacopa parviflora was used by Standley in "Flora of the Panama Canal Zone" and, although the name has been in use since that date, no description has been published nor was a specimen WILLIAMS: TROPICAL AMERICAN PLANTS, XII 119 cited. The specimen here selected as type must have been one of those seen by Standley and it was originally determined by him. Bacopa sessiliflora (Benth.) Pulle, Enum. PI. Surinam 415. 1906. Herpestis sessiliftora Benth. Comp. Bot. Mag. 2: 58. 1836. British Honduras: small plant, flowers white, Coronel English Pine Ridge, Belize— Cayo Road, Gentle 9439 (LL). This species originally described from French Guiana, is known in our region from Nicaragua and Panama and now from a single col- lection from British Honduras. BUCHNERA L. — The genus Buchnera is a reasonably distinct and homogeneous one in which the New World species, at least, have been much confused. Mr. Philcox' "Revision of the New World species of Buchnera L.," in Kew Bulletin 18: 275-315. 1965 has helped the situation but does not make the Buchneras much easier to deter- mine. While Mr. Philcox acknowledges having seen the material from Field Museum there are very few annotation labels by him in the collection and fully half of the 300 or more collections here are not to be found in the index of collectors at the end of the work. Mr. Philcox has reduced B. leiantha Standley, a species from Pan- ama, to B. weberbaueri Diels, a native of Peru. We cannot find differences between these species and B. palustris (Aubl.) Spreng. and shall use this older name in "Flora of Guatemala." A more logical distribution results by this reduction. The type of Buchnera weberbaueri was presumably destroyed at Berlin. Mr. Philcox has selected the type of the Panamanian B. leiantha as the neotype of the Peruvian B. weberbaueri. Buchnera floridana Gandoger, Bull. Soc. Bot. France 66: 217. 1919, is reported from British Honduras by Philcox (Lundell 4866) and we now have it from the Pete"n in Guatemala (Tun 918). Castilleja integrifolia L. f. var. alpigena L. Wms. var. nov. Herbae parvae subalpinae foliis dentatis aut pectinatis vel subintegris differt. Guatemala: on wet cliffs, pine-juniper forest area, limestone re- gion 3-15 km. north of Chemal, Sierra de los Cuchumatanes, Depart- ment of Huehuetenango, alt. 3,400 m., December 2, 1962, Williams, Molina & Williams 22186 (type, F; EAP). Additional specimens in Field Museum, all from Guatemalan highlands: Roe, Roe & Mori 636; Standley 61861, 61912, 67725; 120 FIELDIANA: BOTANY, VOLUME 34 Steyermark 34183, 34813, 35507, 36130, 48252, 50285; Williams, Molina & Williams 22215. Additional specimens in U. S. National Herbarium from Guatemalan highlands: Beaman 2912, 2958, 3073, 3238, 4028, 4073, 4120; Maxon & Hays 3704; Nelson 3708; Pittier 38; Shannon 318, 3647; Donnell-Smith 2146. This variation of Castilleja integrifolia is perhaps an altitudinal one usually much reduced in size and with the leaves either dentate, pectinately divided or sometimes subentire. The plants vary from quite glabrous and even glandular on the veins to sparsely pilose. The variety is on most of the high mountains of Guatemala at an elevation of 3,400 m. or above while the typical variety of the species is to be found on the lowers slopes of the same mountains. The specimens cited had mostly been determined as Castilleja pectinata Mart. & Gal., a species originally described from Mt. Orizaba in Mexico. Authentic material of C. pectinata has not been available for study and, while it is possible that the Guatemalan material is the same, I believe that ours represents nothing more than a high country and usually dentate or pectinate leafed variety of C. integrifolia. Escobedia guatemalensis Loesner, Verk. Bot. Ver. Brandenb. 53:83. 1911. The type specimen was photographed in Berlin by J. Francis Macbride (F 12368) and, it is presumed, has since been destroyed. Loesner cited two collections by Seler and Seler, 2663 and 2810, and Macbride photographed number 2663. There is a specimen of Seler & Seler 2663 in Gray Herbarium which is obviously the same species as the specimen photographed by Macbride at Berlin; I designate the specimen in the Gray Herbarium as lectotype. Gibsoniothamnus epiphyticus (Standl.) L. Wms. comb. nov. Clerodendron epiphyticum Standl. Field Mus. Bot. 22: 168. 1940. Costa Rica: flowers purple, large woody epiphyte in upper level of branches, April 3, 1931, Lankester 1296 (type, F); flowers red-vio- let, fruit green, small epiphytic shrub pasture near Rio Naranjo 3 km. east of Cachi, prov. Cartago, alt. 1,350 m., 11 June 1969, Lent 1731 (F, 2 sheets; EAP; NY; US). The specimens, received too late to include in the account of the new genus Gibsoniothamnus (Field Mus. Bot. 32: 213. 1970), prove that Clerodendron epiphyticum Standl. is, in fact, scrophulariaceous, although I thought this unlikely from the nearly sterile type sped- WILLIAMS: TROPICAL AMERICAN PLANTS, XII 121 men. Dr. Lent's recent collections match so closely the type, col- lected by Mr. Lankester, in all vegetative characters that there is no question that the two are the same. The new collection, which bears both flowers and fruits, strengthens the belief that the most closely allied genus Schlegelia Miq., most often placed in the Big- noniaceae, is best placed in the Scrophulariaceae. LAMOUROUXIA. The genus Lamourouxia was revised by B. L. Robinson and J. M. Greenman (Am. Jour. Sci. 50: 169-174. 1895) and W. R. Ernst currently has a revision of the genus in progress. He has annotated specimens in herbaria and I have seen some of these. The treatment in the "Flora of Guatemala" will not follow Dr. Ernst's in some of the species complexes. There are five species of Lamourouxia in Guatemala, one of these divided into three var- ieties, as follows: Lamourouxia longiflora Benth. var. longiflora. L. longiflora Benth. PI. Hartw. 22. 1839; in DC. Prodr. 10: 540. 1846. To be found in south Mexico and Guatemala. Lamourouxia longiflora var. integerrima (Donn.-Sm.) L. Wms. comb. nov. L. integerrima Donn.-Sm. Bot. Gaz. 13: 189. 1888. Differs from the typical variety in having broader and more tri- angular calyx lobes, but probably not really distinct. Known only from Guatemala. Lamourouxia longiflora var. lanceolata (Benth.) L. Wms. comb. nov. L. lanceolata Benth. in DC. Prodr. 10: 542. 1846. Distributed from southern Mexico, through Guatemala to El Sal- vador and Costa Rica. LIMOSELLA. In our region the genus Limosella is distributed in in Mexico and Guatemala. We have no record to indicate that the genus is known in other Central American countries, although it is to be expected on the high volcanoes of Costa Rica and Panama. We then have collections from scattered localities in Venezuela, Peru, and Bolivia. The material from the high volcanoes of Mexico and Guatemala and that of Venezuela, Peru, and Bolivia seems to be the same spe- cies. There are two possible names for this plant, neither of which is based on a specimen that I have seen. Limosella acaulis Sesse" & Mocifio is described in Flora Mexicana ed 2. 143. 1894 and there is a 122 FIELDIANA: BOTANY, VOLUME 34 a Sess4 & Mocino illustration of this in the herbarium at Geneva (F Photo 30813). The illustration was apparently based on Sess£ & Mocino number 963. There is no duplicate of this number in Field Museum. The other name available is Limosella americana Gliick, published in Notizbl. Bot. Gart. Berlin 12: 75. 1934. The species and three growth forms are described and the range for all of them is given as "Siid-Amerika (Bolivien, Peru, Argentien); Mexiko (San Luis Po- tosi); New Mexico (Mogollo Mountains)." No specimens are cited by Gliick. We have Schaffner 353 from San Luis Potosi, collected in 1879, and Bang 1968 and 2606 from Bolivia without date. These collections could have been among those seen by Gliick. These spe- cimens seem to me to be the same species and most probably to be the same as L. acaulis, the older of the two names and the one which will be used in "Flora of Guatemala." I have seen the following specimens which seem to be the same and referable to Limosella acaulis Sess4 & Mocino. Mexico: Leavenworth & Hoogstraal 1221a; Liebmann 1585; Pur- pus 1662; Rzedowski 22847; Schaffner 353. Guatemala: Standley 81094, 81095; Steyermark 50271. Venezuela: Steyermark 55899. Peru: Iltis et al. 203, 439, 1339, 1439, 1448c; Ranti—Hirsch P1132. There is another specimen of Limosella from Ecuador, Steyer- mark 83137, which must represent an undescribed species. Lindernia rotundifolia (L.) Standl. & L. Wms. comb. nov. Gratiola rotundifolia L. Mantissa PI. 174. 1767. Ilysanthes rotundi- folia Benth. in DC. Prodr. 10: 420. 1846. This species was discovered as a weed in the lawn at Bananeras, Guatemala (Steyermark 38796). It is not otherwise knowTn on con- tinental North America but is naturalized in the West Indies and in South America. Native of the Old World. MICRANTHEMUM Michaux The generic name Micranthemum Michaux, 1803, is conserved over the name Globifera J. F. Gmelin, 1791. The type species of the genus is Micranthemum orbiculatum Michx. which is an illegitimate name, and a synonym of Micranthemum umbrosum (Walt.) Blake (Anonymos umbro. Walter, Fl. Caroliniana 63. 1788). In the Index Genericorum the name Micranthemum umbros (J. F. Gmelin) Blake is WILLIAMS: TROPICAL AMERICAN PLANTS, XII 123 given as the correct name for the species while Rickett & Stafleu (Taxon 9: 111. 1960) give M. umbrosum (J. F. Gmelin) Blake. It appears to me that Dr. Blake was correct in writing Micranthemum umbrosum (Walt.) Blake and that the insertion of J. F. Gmelin as parenthetical authority has no basis in the literature. De la Torre and Harms, in Genera Siphonogamarum give the genus as "Micranthemum (L. C. Rich, in:) Michx." but there would seem to be no basis for using the parenthetical authority since Richard's name does not appear in connection with the description in Flora Boreali- Americana. Two species of Micranthemum are described below, one from Gua- temala, the other from Mexico. Micranthemum procerorum L. Wms. sp. nov. Herbae perparvae. Folia opposita, ovato-orbiculares, indistincte quinque- nervia (vel fere plinervia) ; inflorescentia axillaris, flos solitarius e folii axilla prodiens, perparva; calyx 4-lobatus, lobi similes, anguste oblongo-oblanceolati, obtusi; corolla usque ad basim divisa, bilabiata; capsula globosa; semina oblonga. Very small, prostrate, much branched herbs, stems rooting at the nodes, glabrous, up to 5 cm. long or perhaps more. Leaves opposite, ovate-orbicular to orbicular or transversely ovate-orbicular, indistinctly 5-nerved (or nearly pli- nerved) from the base, 2.5-5 mm. long and somewhat narrower than long to broader than long; inflorescences consisting of a solitary nearly sessile flower in the axils of leaves; flower very small and inconspicuous; calyx lobate, the lobes similar, narrowly oblong-oblanceolate, obtuse, about 1.6 mm. long and 0.4 mm. broad; corolla divided almost to the base and bilabiate, about 1-1.3 mm. long, posterior lip entire, oblong or oblong-lanceolate, the apex rounded, about 1 mm. long and half as broad, the anterior lip trilobate almost to the base, the lateral lobes spreading, oblong, apices rounded, the mid-lobe similar with the filaments attached along the sides near middle of the center lobe, about 0.2 mm. long, issuing from thicker or subglandular tissue of the corolla; anthers bilocular, the locules parallel, about 0.15-0.2 mm. long; style 0.3-0.4 mm. long, the stigma obscurely bilobate, about 0.1 mm. in diameter; capsule globose, about 1 mm. in diameter; seeds oblong, 4-ridged with the ridges roughened, about 0.3 mm. long and 0.17 mm. broad. Mexico: "Jalapae in paludosia" 1787, 1795, 1804, Sesse, Mocino, Castillo et Maldonado 150 (type, F; US). There are, in addition, two unlocalized and unnumbered Mexican collections by the same collectors. The specific name is intended to honor the illustrious persons who collected these specimens some 175 years ago. The specimens were marked by Sess£ & Mocino "grati- olis affinis" so they have been incorrectly placed in the folders of Gratiola in our herbarium for some 35 years! 124 FIELDIANA: BOTANY, VOLUME 34 Allied to M . standleyi of Guatemala and somewhat less so to M . umbrosum (Walt.) Blake of the eastern United States and possibly Mexico. Specimens of Micranthemum often have abundant flowers. It would be an interesting small genus for monographic study. Micranthemum standleyi L. Wms. sp. nov. Herbae perparvae procumbentes ramosissimae. Folia obovata, obtusa, tri- nervia, glabra; inflorescentia axillaris, uniflora; flos breviter pedicellatus; calyx 4-lobatus, lobi oblanceolati acuti ciliati; corolla bilabiata tubus brevis, labium dorsale dentiforme, labium anterior trilobulatum; stamina 2; stylus brevis, apice leviter incrassatus bilobatus geniculatus; capsula globosa, verosimiliter unilo- cularis. Small, prostrate, much branched herbs. Stems subfiliform, the nodes 4-5 mm. apart; leaves opposite, obovate, obtuse, trinerved, glabrous, 1.5-3.5 mm. long and 1-2 mm. broad, sessile; inflorescence a single short pedicellate flower in the axils of leaves, one at each node and apparently alternating at these nodes, pedicel about 0.5 mm. long; calyx lobed to the base, the lobes 4, oblanceolate, acute, ciliate, about 0.7 mm. long and 0.2-0.3 mm. broad; corolla about 1 mm. long, bilabiate, the tube short, upper lip simple, much reduced, dentiform, lower lobe trilobulate, the laterals oblong, obtuse, about 0.2-0.3 mm. long, the mid-lobe oblong, obtuse, about 0.4 mm. long; stamens 2, inserted on the throat of the corolla and reaching the apex of the corolla, inserted on the exterior lobes, anthers about 0.1 mm. long, bilocular, the loculi subparallel and somewhat explanate; style bilobate or bilamellar and geniculate at the apex, 0.2-0.3 mm. long; capsule globose, appearing unilocular at maturity, about 1-1.3 mm. in diameter; seeds many, oblong, angular, about 0.2-0.3 mm. long and half as broad. Guatemala: common, creeping, moist thickets on plains near Ayutla, Dept. San Marcos, alt. 45 m., March 14-15, 1939, Standley 68810 (type, F; US). Related most closely to the Mexican M. procerorum described above but a slightly more delicate plant with smaller somewhat nar- rower leaves, the flowers smaller (!), and the calyx lobes ciliate or dorsally pubescent. The type specimen was collected by Standley and the generic name placed on the label but the genus was not included in the pre- liminary manuscript of the "Flora of Guatemala." It is a pleasure to name this species in memory of Dr. Standley, my colleague of years past, who was the most discriminating of all botanists who have worked with the Central American flora. Russelia chiapensis Lundell, Field and Lab. 13: 12. 1945; Carlson, Fieldiana, Bot. 29: 257. 1957. \ u ^ -j , X / S / I ^ ,,, ... / FIG. 3. Schlegelia silvicola. Branch J^ natural size; a bud shown from the anterior side to show aestivation, X 1%; calyx dissected to show ovary and style, Xl%; apex of style much enlarged; corolla dissected to show stamens and stamen- ode, X 1 Yz ; anthers, two views, X ± 5. 125 126 FIELDIANA: BOTANY, VOLUME 34 Guatemala: in secondary growth on Temple 1, Tikal National Park, Tikal, Peten, March 13, 1966, Contreras 5579 (F). This Mexican species, to be expected in Guatemala has been found at Tikal. A photograph of the type is available, F41083, and it is assumed that the type is now in Lundell Herbarium, Renner, Texas. Schlegelia silvicola L. Wms. sp. nov. Frutex epiphyticus usque ad 2 m. altus, vel verosimiliter liana silvicola, ramulis teretibus; folia coriacea, elliptico-ovata, acuta vel obtusa; inflorescentia axillaris racema 1-5-flora, sessilis aut breviter pedunculata; calyx campanulatus, glabrus, vel parce lepidoto-glandulosus; corolla bilabiata, tubulosi-campanulata, lobi posteriori ovato-orbiculari, obtusi, lobi anteriori transversi-ovales; stamina 4, didynami; staminodium praesens; stylus perbrevis; fructus ignotus. Epiphytic shrubs to 2 m. tall or possibly elongated lianas in the cloud forest. Stems terete, obscurely angulate, grayish, glabrous; leaves coriaceous; glabrous but obscurely glandular-punctate below at the junction of petiole, broadly elliptic- ovate, acute or obtuse, 7-14 cm. long and 4-9 cm. broad, petioles thick and cori- aceous, about 1 cm. long, leaves of a pair subequal in size; inflorescence axillary, a sessile or short pedunculate 1-5-flowered raceme, borne in the axil of leaves, the pedicels slender, puberulent, bracteolate at about the middle, mostly about 1 cm. long; calyx broadly campanulate, about 10 mm. long and 7-8 mm. in diameter, glabrous or very sparsely lepidote-glandular, obscurely bilabiate with shallow and broad lobes; corolla tubular-campanulate, bilabiate, glabrous outside, about 4 cm. long, narrowed to a short tube about as long as the calyx at the base, narrowly campanulate above, the lips spreading, lobes of the posterior lip ovate-orbicular, 10-12 mm. long and nearly as broad, obtuse, glabrous, anterior lip trilobate, farinaceous-puberulent on inner surface, the lateral lobes transversely oval, about 10 mm. long, mid-lobe smaller, ovate, about 10 mm. long; stamens 4, didynamous, pair on the sini of the mid-lobe of the anterior lip longest, 22-23 mm. long, the others about 15-17 mm. long, the filaments broadened and barbate at the base; staminode present, fleshy, barbate at base, about 4 mm. long, situated opposite the sinus of the posterior lip ; anthers attached at the middle, the cells explanate, each cell about 3 mm. long; style short, about 5 mm. long, stigmas 3, ovate-lanceolate, about 1 mm. long; capsule unknown. Nicaragua: lobes pink, tube cream, epiphytic shrub 2 m. tall in old and cut-over cloud forest between Disparate de Potter and Aranjuez, Cordillera Central de Nicaragua, Department of Mata- galpa, alt. 1,300 m., January 12, 1963, Williams, Molina & Williams 23659 (type, F; EAP). Guatemala: flowers rose, woody vine climbing to top of tall forest trees, Finca Moca, Dept. Suchitepe"quez, January 12, 1935, alt. 3,800 ft., Skutch2116 (F). Both specimens cited have been determined by us as Schlegelia nicaraguensis and the Guatemalan one was also seen by Dr. Schery WILLIAMS: TROPICAL AMERICAN PLANTS, XII 127 who concurred in the determination. In preparation of the account of the species for "Flora of Guatemala" it became apparent that the two specimens from the highlands offered several differences from the lowland S. nicaraguensis. Schlegelia silvicola belongs in a complex about which we know much too little for there are few specimens in herbaria. Schlegelia parasitica from Jamaica is closely allied but apparently has smaller flowers than any of the Central American species with a much smaller calyx at anthesis, the leaves are less coriaceous and longer in rela- tion to width. Schlegelia nicaraguensis which is known from lowland Nicaragua has much smaller, relatively thin leaves that are abruptly acuminate, the calyx 8-11 mm. long and obscurely undulate-lobed. This I believe to be the nearest ally of S. silvicola. Schlegelia cos- taricensis Standl. is a highland species, again with smaller less cori- aceous leaves, deeply bilabiate corollas, prominently lobate calyx with the lobes acuminate and tomentulose on the margins within. SIBTHORPIA L. The genus Sibthorpia is represented in Mexico and Central America by four names based on specimens from the area and by two additional names based on material from Andean South America. It appears to me that all of the species described from and credited to our area are perhaps only one species. There is, however, a collection made by Pringle in Mexico, number 4681, that seems not to be the same as the other Sibthorpias known from North America. Sibthorpia repens (Mutis ex L.) 0. Kuntze is the oldest of the names applied to North American Sibthorpias, it is based upon Mexican material, and will be used in the "Flora of Guatemala." It seems quite probable that the other names based on Mexican and Central American material, 5. americana Sess£ & Mocino, S. par- vifolia Mart. & Gal., and S. triandra Suesseng., are synonymous. Two names based on South American material, Sibthorpia pichin- chensis HBK. and S. rotundifolia (Ruiz & Pavon) Edwin, have been applied to Central American material. These two names may be referable to S. repens also. Tetranema bicolor L. Wms., sp. nov. Herbae terrestres parvae, caules repentes aut erecti. Folia inaequalia bi- colorata elliptico-ovata acuta vel acuminata, paulo pubescentia, margine leviter repanda; inflorescentiae breves pauciflorae axillares, pedunculi breves; calyx anguste campanulatus, lobi angusti lanceolato-triangulares acuminati; corolla FIG. 4. Tetranema bicolor. A, habit of the plant, X J-3; B, calyx with ovary and style, X 4; C, corolla dissected (one anther deformed), X 4; D, stamen, X 13; E, two buds and calyx with immature capsule, X 3, and immature seed much enlarged. Illustrated from the type by Marion Pahl. 128 WILLIAMS: TROPICAL AMERICAN PLANTS, XII 129 campanulata, labellum posticum oblongum obtusum et retusum, labellum anticum trilobatum, lobi obtusi oblongo-ovati; stamina didynama; stylus leviter incras- satus; stigma rotatum integrum. Small terrestrial herbs, the stems repent and rooting with the apex erect, arachnoid pubescent, glabrescent. Leaves of a pair very unequal in size, opposite, bicolored, elliptic-ovate, acute or acuminate, narrowed to the subsessile or sessile base, joined with that of the opposite leaf, margins slightly repand, somewhat pubescent along the veins below and on the very young leaves, otherwise glabrous, mature leaves 10-20 cm. long and 4-6.5 cm. broad; inflorescences short, few- flowered axillary cymes, or single axillary flowers, up to about 4 cm. long, peduncle to about 1 cm. long with one or two linear or linear-lanceolate bracts at the apex, arachnoid-pilose, deciduous in age, pedicels to about 10-12 mm. long, slender; calyx narrowly campanulate, about 1 cm. long, glabrous or nearly so, the tube short, about 2-3 mm. long, the lobes narrowly lanceolate-triangular, acuminate, 8-9 mm. long and 2-3 mm. broad at the base; corolla campanulate, the posterior lip obtuse, somewhat retuse, free portion oblong, about 4 mm. long and nearly as broad, the anterior lip trilobate, the lobes obtuse, oblong-oval; stamens 4, didyna- mous, the anthers 2-celled; ovary 2-celled, about 2 mm. long; the style rather thick, terminated by a entire rotate stigma; mature capsule and seeds not seen. Nicaragua: leaves dark green above, paler beneath, flowers lavender, in tall forest near Rio San Juan at "el Relos," about mid- point between El Castillo and Delta de San Juan, alt. 0-50 m., March 23, 1961, Bunting & Licht 748 (type, F). This species differs in several important ways from the few others in the genus but vegetatively it is similar to the other known species and to the one described below. The inflorescence is much shorter than in any of the others, the peduncle being not more than a centi- meter long. The calyx is much larger and more foliaceous. In the single good flower available for dissection there was no staminode. In the flower dissected one anther seems to have failed to develop properly and is so illustrated. This like other species of the genus is a plant of the forest floors in the wet lowlands and at first glance appears to be similar to plants of the Gesneriaceae and Rubiaceae that inhabit similar localities. The range of the genus is extended southeastward from southern Mexico, Guatemala, British Honduras, and Honduras to Nicaragua. Tetranema cymosum L. Wms. sp. nov. Herbae parvae usque ad 20 cm., caules perbreves. Folia anguste obovata, obtusa, basi cuneata et alata, ciliata; inflorescentia cymosa, longe pedunculata; calyx fere ad basem divisus, 5-lobatus, lobi lanceolati, acuti; corolla bilabiata, campanulata, alba; stylus filiformis; capsula ovoidea. Small herbs to 20 cm. tall, the stems short and perhaps suffrutescent at the base, the opposite leaves crowded at the apex of the stem. Leaves narrowly obovate, obtuse, narrowed to a broad-winged petiole at the base, 12-25 cm. long FIG. 5. Tetranema cymosum. A, habit from the type, X 1A', B, inflorescence with nearly mature capsules, X 2; C, calyx and pistil, X 5; D, corolla dissected to show interior, X 4; E, anthers, X 10; F, stamenodium, X 25; G, hairs from in- terior surface of corolla, much enlarged; H, stigma, much enlarged. Del. Marion Pahl 1970. 130 WILLIAMS: TROPICAL AMERICAN PLANTS, XII 131 and 4-8 cm. broad, glabrous except on the margins and on the veins below pubes- cent with weak several-celled hairs about 0.5 mm. long, veins 6-9 pairs; inflores- cence axillary, an open 2-8 flowered cyme borne on a long, slender peduncle somewhat shorter than the subtending leaves, sparsely pubescent, bracts narrowly lanceolate, cucullate, 2-3 mm. long, pedicels slender, about 5 mm. long; calyx divided almost to the base, 5-lobate, the lobes lanceolate, acute, 2.5-3 mm. long; corolla bilabiate, white, about 1.5 cm. long, campanulate with the lips spreading, posterior lip as broad as long or broader, notched at the center, subquadrate, about 4 mm. long, anterior lip deeply trilobate, the mid-lobe elliptic-ovate, about 5-6 mm. long, lateral lobes subquadrate-ovate, about 5-6 mm. long, tube campanulate, narrowed at the base, about 9 mm. long; style filiform, about 10 mm. long; anthers 4, didynamous, inserted near base of corolla tube, filaments of short pair on the sini of posterior lip, about 5 mm. long, filaments of longer pair inserted on sini of the mid-lobe of anterior lip, anther cells explanate, each cell about 0.5 mm. long; staminode vestigial, inserted at base of corolla tube on middle sinus of posterior lip, about 0.4 mm. long; capsule near maturity ovoid, 5 mm. long; seeds irregular, subquadrate, about 0.4 mm. long. Guatemala: crevices of moist shaded limestone bluffs along water- fall, corolla white, dull purple at base of tube with dull purplish spots on upper side; leaves subcoriaceous, dull green above, silvery pale green beneath, corolla lobes spreading outward, down and up; jungle between Escobas and waterfall, across bay from Puerto Barrios, Dept. Izabal, alt. 20-50 m., April 23, 1940, Steyermark39858 (type,F). There are in Mexico and Central America five species of Tetra- nema, two including the present one with open cymose inflorescences, two with compact, almost subumbellate, cymes. Tetranema cymosum with an open cymose inflorescence seems more closely related to those species with compact inflorescences, T. roseum and T. megaphyllum, than to T. evolutnm which has an open inflorescence. Tetranema megaphyllum (Brandegee) L. Wms. comb. nov. Allophyton megaphyllum Brandegee, Univ. Cal. Publ. Bot. 6: 62. 1914; Pennell, Proc. Acad. Nat. Sci. Phila. 77: 272. 1925. Dr. Pennell in 1925 took up the generic name Allophyton Brande- gee for this small group of plants because Tetranema Benth. was a later homonym of Tetranema Sweet. However, Tetranema Benth. of the Scrophulariaceae has been conserved. There are five reasonably distinct species of this rare genus from south Mexico and adjacent Central America; T. cymosum L. Wms. and T. bicolor L. Wms., described above; T. evolutum Donn.-Sm. known only from the department of Alta Verapaz in Guatemala; T. roseum (Mart. & Gal.) Standl. & Steyerm., which includes T. mexicanum Benth., is the most widely distributed of the species being found in the Mexican states of Vera Cruz and Chiapas, as well as in 132 FIELDIANA: BOTANY, VOLUME 34 Guatemala, British Honduras, and Honduras; T. megaphyllum ( Brand egee) L. Wms. is known only from the state of Chiapas in Mexico; T. bicolor L. Wms. is known from Nicaragua. TILIACEAE Belotia mexicana (DC.) Schum. in Engler & Prantl, Pflan- zenfam. 3 (6): 28. 1890. Grewia mexicana DC. Prodr. 1: 510. 1824. Belotia galeottii Turcz. Bull. Soc. Nat. Mosc. 19 (2): 504. 1846. Trichospermum mexicanum Baill. Hist. PI. 4: 179. 1873. Belotia caribaea Sprague, Kew Bull. 276. 1921. B. campbellii Sprague, I.e. 277. B. tabascana Sprague, I.e. 278. Trichospermum galeottii Kosterm. Reinwartia 6: 278. 1962. T. caribaeum Kosterm. I.e. T. tabascana Kosterm. I.e. 279. Belotia and Trichospermum have been reviewed by Dr. Sprague (Kew Bull. 270-278. 1921), by Dr. Bullock (Kew Bull. 517-521. 1939), and more recently by Dr. Kostermans (Reinwartia 6: 277-279. 1962), who has written that he believes Belotia Rich, is a synonym of the Old World Trichospermum and gave his reasons, made combina- tions from Belotia to Trichospermum for most of the names of American origin. I prefer to retain the name Belotia for the American plants for it seems that characters of the inflorescence, possibly of the capsule and of the position of the stamens on a raised disc in the New World, circum-Caribbean Belotias is sufficient to distinguish them from the Old World, principally Asian, Trichospermums until more critical work can be done on the group. There are perhaps no more than two or three species of Belotia as Standley has pointed out (Fieldiana, Bot. 24 (6): 305-306. 1949) in spite of a dozen or so names; I think perhaps only Belotia mexicana in North America. The synonomy given above is that which has been used on Guatemalan plants. f- UNIVERSITY OF ILLINOIS-URBANA