30 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM, 
Pinnz cut more than half-way or nearly to the costa. 
Segments acuminate, usually very strongly so; 
pinne 6 to 26 cm. broad. 
Coste and costules conspicuously paleaceous, 
the scales whitish and numerous. ..... 18. H. grandifolia. 
Costze and costules devoid of scales or, in very 
immature fronds, rarely bearing an 
occasional thin brownish caducous scale. 19. H. horrida. 
Segments rounded or, if rarely subacute, at least 
never acuminate; pinne 2.5 to 5 cm. broad. 
Costee and costules bearing numerous slender 
lax white long-pointed scales........ 20. H. kohautiana, 
Costee and costules bearing fewer, larger, and 
broader, subbullate or * bullate, brown- 
ish scales... 2. ...2........2.......... 21. H. obtusa. 
1. Hemitelia speciosa (H. & B.) Kaulf. Enum. Fil. 1824, 
Cyathea speciosa H. & B.; Willd. Sp. Pl. 5:490. 1 1S10. 
Hemitelia lindeni Hook. Icon. Pl. pl. 706. 1848. 
TyPE Locauity: Caripe, Venezuela, Humboldt. 
Distrisution: Known only from Venezuela. 
ILLusTration: Hook. loc. cit. pl. 706 (as H. lindeni). 
This species has been the subject of a great amount of misunderstanding. Kaulfuss 
was the first to transfer the name to Hemitelia; and there is, so far as his diagnosis is 
concerned, no reason for supposing Hemitelia speciosa Kaulf. to be anything mofe than 
a change of name for Cyathea speciosa H. & B. 
Presl, however, in 1836, regarded H. speciosa Kaulf. as not only specifically but even 
generically different from the original speciosa of Humboldt and Bonpland. The latter 
he retained in Cyathea, as Cyathea speciosa H. & B.; the former he made the type of a 
new genus Cnemidaria, as Cnemidaria speciosa (Kaulf.) Presl [‘‘ Hemitelia speciosa 
Kaulf.nec Willd.”’]. In thus determining the identity of Hemitelia speciosa Kaulf. he 
seems to have been guided solely by herbarium material; for, as stated above, Kaul- 
fuss’s diagnosis applies well enough to the true speciosa Il. & B. In fact, Kaulfuss’s 
phrase “Sort venis simplicibus patentibus apicem versus impositi” seems almost cer- 
tainly to apply to the Humboldt and Bonpland plant, rather than to that figured by 
Presl in figures 16 and 17,! as may readily be concluded by comparing these and 
Hooker’s illustration of H. lindeni with Kaulfuss’s diagnosis. 
Kunze appears to have been reluctant to believe Kaulfuss in error, for he points out? 
with what care Kaulfuss was accustomed to compare his material with specimens in 
the Willdenow herbarium. Following a full discussion of the subject, however, he 
finally coincided in Presl’s opinion, and described a new species, Hemitelia subincisa, 
basing it in part upon ‘ Hemitelia speciosa, Kault.”” which Presl had called (as Cnemi- 
daria speciosa) a misidentification of the Humboldt and Bonpland species. Further 
notes on JT, subincisa will be given hereafter. 
The proper identification of Cyathea speciosa H. & B. was complicated further by 
Hooker, who published * under the name ‘‘Hemitelia speciosa Kaulf.,”? in 1844, an 
illustration of a wholly distinct (third) species. This, later in the same year, became 
the foundation of a new species, Hemitelia integrifolia Klotzsch,® which will be dis- 
cussed in this paper under that name.° Having thus wrongly applied the name 
speciosa to a species other than the original Hooker subsequently (in 1848) redescribed 
the true Cyathea speciosa upon Venezuela plants collected by Linden (no. 663) under 
the name Hemitelia lindeni. His illustration of H. lindeni agrees in every partic ular 
1 Presi, “Tent. Pterid. pl. 1. of 16, 17. ‘ Sp. Fil. 1: pl 18 B. 
2 Bot. Zeit. 2: 295. 1844. 5 Tinnaea 18: 539. 1844. 
3 Page 49. © Page 31. 
