STUDIES IN THE COMPOSITE. 245 
to the group, and added to it under that name not only 
some new species now known as belonging to Coreopsis, but 
also the types of Melanthera, which latter plants, it seems 
strange to say, were not excluded from Bidens until as late 
as the year 1803, by Richard, the author of Michauxs' Flora. 
In 1737 Linnæus proposed a new genus Coreopsis, mono- 
typical with him, then, and represented by the Bidens suc- 
cisæfolia of Dillenius, now called Coreopsis lanceolata. 
When, however, in 1753 Linneeus came to the distributing 
of the species, he practically abandoned the original idea of 
Coreopsis, and could only maintain the genus by relegating 
to it all the so-called Bidens species having more or less con- 
spicuous ray-flowers, retaining in Bidens those to which he 
had attributed discoid, or rayless, heads. In thus cutting 
the knot, he was undaunted by the fact that by drawing the 
line at the presence of ray-flowers, he was obliged to place 
Bidens cernua half in Bidens, the other half in Coreopsis, 
where the radiate state of the species is made to figure as 
Coreopsis Bidens. 
Before the end of the eighteenth century there were two 
proposals made to divide Bidens, one by Necker in 1790, 
and another by Moench in 1794. Necker retains the 
typical Bidens, i. e, the type having simple or tripartite 
leaves, flattened achenes and a two- to four-awned pappus, 
assigning it the new, and certainly more appropriate, name 
of Pluridens, while for the group having a more dissected 
foliage, tetragonal achenes, and an almost always four- 
awned pappus, he proposes the generic name Edwarsia. 
Moench’s idea was the same; but he left the old type under 
the name of Bidens and dong usted the new genus as Kernera. 
At about the same time, and upon a type scarcely different 
from that of Edwarsia and Kernera Cavanilles established 
Cosmos; and this last has been accepted everywhere, while 
