10 Contributions from the Gray Herbarium 
paper. There is one phase of the situation, however, upon which 
I would comment. 
Most of the genera segregated from Brodiaea depend for their 
recognition upon characters which are virtually only modifications 
of the same phenomenon. Much importance, for instance, has 
been ascribed to the degree of development of the filament- 
appendages. Their absence, or, when present, the degree of their 
attachment to the corolla and to each other, has been used as a 
means of distinguishing genera although it is apparent that it is 
only a step from broadly winged filaments to filaments united into 
a tube by the union of the appendages. The fallacy of ascribing 
generic value to this type of characters is well shown in the results 
to which it inevitably leads. One who adopts in this group these 
or characters of similar nature for the definition of genera soon 
finds himself compelled, in order to be consistent, to go to ridicu- 
lous limits in his segregation. Rydberg himself admits as much m 
his argument for his segregate genus Dipterostemon Rydb. Bull. 
Torr. Club, xxxix. 110-111 (1912). 
The maintenance of the genus Androstephium, then, would call 
for the assignment of generic value to the sort of characters dis- 
cussed above and the consequent recognition of several other 
groups of closely related species as genera. That these plants are 
anything but Brodiaeas in aspect and fundamental character can- 
not be doubted and for the most part they have been regarded as 
species of this single and in its broader sense rather natural group. 
Brodiaea so constituted may be said to be too close to Milla. 
Baker at one time even referred many of the species of the former 
to the latter genus. Later he rejected his earlier work, no doubt 
because he realized that, considered in the light of all the species, 
two characteristic groups were concerned which on the whole 
were amply distinct. One could argue not without reason for the 
suppression on grounds of technical character of Brodiaea and 
Milla both in Allium but common sense should forbid such action 
even as it should restrain the extreme segregation of natu ele- 
ments. Rarely does it seem to make its influence felt however 2 
thwarting the carrying out of this latter tendency. a 
In this connection I would question the validity of the sevel 
segregate genera of Milla proposed to take care of certain Sow 
American plants. These genera appear to be based on the same 
} 
