83 
work consists almost entirely in raising Grayan subdivisions to 
generic rank, Jn my review of the treatment of the Umbellifere 
in Contributions No. 12 and above in this paper J] have shown the 
fallacy of splitting up Cymopterus and Cogswellia and the re- 
sultant confusion therefrom. In the borages Gray began the 
work himself by splitting off Plagiobothrys on the one end and 
Omphalodes on the other from the old Eritrichium, and he fol- 
keeping together of all the parts in a composite whole, about as 
Gray first put it. Another case in point is the treatment of Clay- 
tonia. No-one ever seems to make any stir about putting annuals 
dberg of course carries it to an absurdity by making Crunocallis 
and Erocallis in addition. The treatment of Ranunculus and Cle- 
matis is another case in point, which is adopted by Greene, Ryd- 
berg, etc. What for example is gained by making three or four 
genera out of Clematis by putting the climbers with small white 
flowers in one genus, the large-flowered lavender-colored ones in 
another (semiclimbers), and the nonclimbers in another when by 
keeping them all in one genus we have a well defined unit clear- 
ly separable from all others except Pulsatilla which along with 
Anemone occidentalis might just about as well be put in Clematis, 
connecting the Douglasii group with Anemone, or up as 
a connecting link. Every one knows that there is a question as to 
what we shall do with Pulsatilla, but it only befogs the whole is- 
sue and multiplies the confusion to disintegrate Clematis, just as 
it does only harm to disintegrate Cymopterus and Cogswellia. In 
Ranunculus the buttercups are well defined, and even a novice 
