MAXON—STUDIES OF TROPICAL AMERICAN FERNS. 43 
mountain by Britton and Cowell (no. 510, in 1901) are evidently not specifically dif- 
ferent from the fragment of Willdenow’s type (no. 20167) shown in plate 25, a; and there 
is, therefore, on grounds of geographic distribution a strong presumption that Presl 
was in error as to the origin of the Willdenow specimen. 
Fée’s change of name for the plant to Hemistegia willdenovii was made purely 
from a supposed nomenclatorial difficulty, arising from his desire to retain both Micro- 
stegnus grandifolius Presl and Hemistegia grandifolia Presl (‘‘Hemitelia grandtfolia 
Hook.’’) in the same genus; unfortunately he in error chose the true grandifolia for 
renaming. 
(2) Imrayana. Described from Dominica by Hooker on plants collected by Doctor 
Imray. A specimen at hand from this island differs in no essential particular from 
more copious material from Guadeloupe and Martinique. 
(3) Insignis. A name applied by Fée to specimens collected by l’Herminier in 
Guadeloupe. A considerable number of specimens from Guadeloupe and Martinique 
give us our best idea of this species. They are clearly of the same species as Will- 
denow’s type. 
Hemitelia grandifolia is sufficiently distinct from H. kohautiana and H. obtusa by 
the data given in the key. This has been made as full as possible for the purpose of 
pointing out very definitely the characters by which these species, which have so 
long been confounded by nearly every writer, may be adequately distinguished. 
EXPLANATION OF PLATE 25.—a, Fragment from the type specimen, herb. Willdenow (no. 20167), from 
Martinique; b, tip of a small upper pinna of a Martinique specimen, Duss 1605; c, tip of a large lower 
pinna of a Guadeloupe specimen, Duss 4452. 
19. Hemitelia horrida (L.) R. Br.; Spreng. Syst. Veg. 125, 1827. 
Polypodium horridum L. Sp. Pl. 1092. 1753. 
Cyathea horrida J. E. Smith, Mem. Acad. Turin 5: 416. 1793. 
Cyathea commutata Spreng. Anleit. Kennt. Gewiichse 3: 146. 1804 (excluding all 
reference to Plumier). 
Cnemidaria horrida Presl, Tent. Pterid. 57. 1836. 
Actinophlebia horrida Presl, Abh. Béhm. Ges. Wiss. V. 5: 356. 1848. 
Hemistegia horrida Fée, Gen. Fil. 351. 1850-52. 
Hemitelia hookeri Presl, Abh. Béhm. Ges. Wiss. V. 5: 350. 1848. 
Hemitelia hookeriana Schlecht. Bot. Zeit. 14: 474. 1856. 
?Hemitelia acuminata Karst.; Schlecht. Bot. Zeit. 14: 474. 1856. 
Hemitelia commutata Schlecht. Bot. Zeit. 14: 474. 1856. 
Hemistegia repanda Fée, Gen. Fil. 351. 1850-52; Mém. Foug. 11: 98. 1866. 
Typge Locauiry: Near Port de Paix, Haiti, Plumier. 
DistRIBUTION: Santo Domingo, Jamaica, Cuba, Porto Rico, Costa Rica, and doubt- 
fully Colombia. 
InLusrrations: Plumier, Traité Foug. pl. 8; Spreng. loc. cit. pl. 4. f. 82; Hook. 
Sp. Fil. 1: pl. 15; Hook. & Bauer, Gen. Fil. pl. 4. 
A redescription of this species is scarcely necessary, inasmuch as it is common in 
the Greater Antilles and is doubtless well represented in all the larger herbaria. 
Scant material of H. horrida and a consequent failure to recognize the full extent of 
its variation were, however, responsible for a large amount of speculation and critical 
comment in earlier times. Thus, Presl in 1848 founded a new species, Hemitelia 
hookeri, (without description) upon plate 15 of Hooker’s Species Filicum and plate 4 
of his Genera Filicum, supposing these to illustrate a species distinct from H. horrida, 
whereas they represent a condition not infrequently observed in particularly robust 
individuals of that species. Presl cited also Venezuelan specimens collected by 
Linden (no. 1572); but if we regard his species as typified by plate 15, as seems proper, 
it becomes a straight synonym of H. horrida, whatever may be the identity of Lin- 
den’s no, 1572. 
