‘ 
ROBINSON: POLYCODIUM 555 
indeed, but certainly present. Moreover, V. multiflorum was 
described as having anthers alternately awned and awnless, yet a 
recent collection from near its type locality, Pittier 1182, agreeing 
otherwise with Bentham’s description, has all the anthers of such 
flowers as were examined 2-awned, with awns as described and 
found for this and for V. consanguineum. The plate* of V. 
meridionale is wrong in this respect. 
Yet, all these can be described as awned anthers, and on other 
grounds the alliance of the species is with those possessing awned 
anthers. Much search has failed to disclose a single other excep- 
tion, not only within the limits of a species but within the limits 
of what appear to be groups of related species. 
It may be added that the presence of pubescence, in whatever 
degree, on the filaments or the connectives—and when it occurs it 
is usually on both—is equally conclusive but of secondary impor- 
tance. Thus, if a vacciniaceous plant be found in America north of 
Mexico with awnless anthers and glabrous filaments, no further 
information is needed for its determination as Gaylussacia frondosa 
(L.) T. & G., except by those who consider G. nana (A. Gray) 
Small and G. tomentosa (Pursh) Chapm. to be specifically distinct 
from that species. Each of the other three combinations of these 
two characters will describe large groups of species. 
On the other hand, investigation of the relative length and 
degree of divergence of the awns and of the relative amount of 
pubescence on the filaments seems to indicate that these do not 
afford reliable characters except possibly in rare cases. 
Up to the present the following have been ascribed to Polyco- 
dium as distinct species: P. caesium Greene, P. candicans (C. Mohr) 
Small, P. elevatum (Banks) Greene, P. floridanum (Nutt.) Greene, 
P. Langloisii Greene, P. melanocarpum (C. Mohr) Small, P. 
neglectum Small, P. oblongum Greene, P. oliganthum Greene, P. 
revolutum Greene, and P. stamineum (L.) Greene; it has already 
been stated that another species awaits transfer, Vaccinium 
Kunthianum Klotzsch. The last having been separated on the 
basis of its shorter anthers, not or barely exserted, the character 
depended on as of next importance is the relation between leaves 
or bracts and the inflorescence. 
ch csceninenemsl 
* Sw. Ic. Ind. Occ. pl. 12. 1794. 
