80 
off from a common stock, or more, as in Astragalus and most of 
the larger genera. But this can be allowed for by primary and 
and species in the families on the basis of the highest first! This 
ing a new key and arranging the families in order from the low- 
est up, but puts at least the species in the reverse order, which 
is also inconsistent. The new Manual is the only one that does 
not follow the unwise method of having only skeleton keys and at 
the beginning. 
Taking up the conception of genus as worked out by Torrey, 
Gray, Watson, Bentham and Hooker, the Pflanzenfamilien, and 
most modern botanists we find them grouping the species in genera 
and subgenera in the natural way, the scientific and sensible way, 
and avoiding the creation of new generic names as far as possible. 
n some cases they went too far and in others not far enough as 
some put certain species in one genus and some in another such 1s 
shown by thorough research. In scattered cases, debatable ones, 
) 
exclusion of others and other systematists following a different 
method and getting other results, but in the main they agreed, the 
policy followed by most of them being to keep all problematical 
species and subgenera confined to the groups to which they showed 
the nearest approach. This was carried to the extreme limit in 
Aplopappus, it being considered far better to hold these loosely 
ageregated groups in a composite whole than to let them loose in 
the family to wander unplaced among the Solidaginoid genera 
piles anything to indicate their poe as Greene and oth- 
ers have done. Under the Grayan od we know where to find 
them, under Gieche’s all is chaos. oie same is true of changes 
in other genera such as Astragalus, Oenothera, Ranunculus, the 
Umbellifere, and many others. Most of these changes were 
made with a great hlare of trumpets and claims of superior wis- 
