1915] SHERFF—BIDENS 307 
because the achenes lack wings and the general characters coincide 
closely with those of certain unquestioned species of Bidens (e.g., 
B. humilis H. B. K., with awns retrorsely barbed). 
FERNALD, in a recent discussion of the awn characters of Bidens 
(Rhodora 15:74-78. 1913), lists no less than six American species 
in which occurs a more or less pronounced form having the awns 
barbed in the reverse from the normal direction. Thus, for ex- 
ample, Bidens connata Muhl. has awns retrorsely barbed, while 
var. anomala Farwell has awns antrorsely barbed. Again, Bidens 
aristosa (Michx.) Britton has antrorsely barbed awns, while the 
probably valid var. Fritcheyi Fernald has retrorsely barbed awns. 
Hence it is obvious that, were the old artificial method of distin- 
guishing between Coreopsis and Bidens (namely by the direction 
of the barbs on the awns) to be retained, an anomalous situation 
would result. We should be compelled either to regard each of 
these varieties as a hybrid between two species of distinct genera, 
a course certainly unwarranted in several cases (cf. FERNALD, /. c., 
and WIEGAND, Bull. Torr. Bot. Club 26:401. 1899), or to refer 
each variety to the other genus, an entirely indefensible alterna- 
tive. We are compelled, then, to view these varieties, in at least 
the majority of cases, as merely more or less distinct and pronounced 
forms of their respective species. This being true, the awn char- 
acter method of separating Coreopsis from Bidens proves utterly 
worthless, and must be permanently abandoned, as it indeed has 
een by American botanists. 
On reference to recent descriptions and types of African species 
of Bidens, we find that in several cases the awns, even on achenes 
on the same head, are barbed both antrorsely and retrorsely. Thus, 
for example, Moore (Jour. Linn. Soc. Bot. 37: 322. 1906) created 
the name Bidens ambigua for Gossweiler 1189, precisely for the 
reason that some of the awns are smooth, others antrorsely barbed, 
and others retrorsely barbed (‘‘achaeniis . . . . aristis 2 quam se 
ipsa brevioribus dentibus perpaucis nunc erectis nunc recurvis 
onustis vel etiam omnino calvis . . . . , hence the trivial name’’). 
Yet in the same year (Jour. Bot. 44:22. i906) he likewise some- 
what arbitrarily created the name Coreopsis Taylori for a plant 
Showing the same variation (coll. W. E. Taylor, Jan. 5, 1886; 
