16 Contributions from the Gray Herbarium 
G. canescens Robinson has the involucre and habit of the series 
Dentatae of Viguiera, from which it differs, aside from specific 
characters of foliage and pubescence, solely in the lack of pappus. 
G. ovata Gray, with G. Liebmannii Klatt, which I cannot distin- 
guish specifically, is very nearly related in habit, inflorescence, 
foliage, and involucre to V. sessilifolia of the series Grammatoglos- 
sae, but differs in its lack of pappus. Nos. 18-20 (G. tenella, G. 
hondensis, and G. triplinervia HBK.) are three of the original 
species of Gymnolomia and as already noted form a genus remotely 
allied to the other species of Gymnolomia as heretofore taken, but 
closely related to Aspilia. G. cruciata Klatt proves, on examina- 
tion of the good sketch and fragments of the type in the Gray 
Herbarium, to be identical with Wulffia baccata (L. f.) Ktze. (W. 
stenoglossa (Cass.) DC.). G. quitensis (Benth.) B. & H. is in habit 
and involucre a close analogue of the section Diplostichis of 
Viguiera, no pappus-bearing member of which is known south 0 
Colombia. G. Kunthiana (Gardn.) Baker, the only Brazilian 
species of the genus, is unfortunately not now available for exami- 
nation, but my recollection of the specimens seen at Kew and the 
British Museum is one of great likeness, aside from the difference 
in pappus, to some species of Oyedaea from the same region. G. 
Goebelii Klatt is as previously indicated a close relative of the 
somewhat isolated Hymenostephium angustifolium Benth. of Co- 
lombia. G. hirsuta Klatt, the type of which is in a very young state, 
possesses a coroniform lacerate pappus and belongs either to true 
Gymnolomia or to the closely related genus Aspilia, which it sug- 
gests in habit, but is best left in Gymnolomia until the fruit becomes 
known. G. Parkinsonii Hemsl. of Mexico, somewhat suggesting 
the remote series Bracteatae (Viguiera) of Brazil in general appeal- 
ance, is easily homologized with the series Grammatoglossae of 
Viguiera, which occur in the same region. G. serrata Rob. & 
Greenm. is known only from a single collection with immature 4 
achenes, but these prove, when closely examined, to have a dis: 
tinctly winged margin. The species must accordingly be referred 
to Verbesina, where it finds a very close ally in V. longifolia Gray, 
likewise an epappose species. There being already a valid Ve — 
serrata Cav., the species may be called Verpesina scotiodonta, — 
from the dark glandular teeth of the leaves. 

