886 THE AMERICAN NATURALIST. 
considerable importance may indeed be attributed to the ini- 
tiative of the workers in determining the character of the 
adult ants which they rear, the macroérgates of Ph. commu- 
tata prove, nevertheless, that we must also attribute a certain 
amount of initiative to the larvae themselves. If this be 
granted, it is but a short step to the admission that the initia- 
tive of the larva, even under normal circumstances, — 7.e., when 
not infested with internal parasites, — may be considerable. 
It is not altogether improbable that further investigation with 
this possibility in mind may lead to some alteration or emen- 
dation of the various hypotheses that have been framed for the 
purpose of explaining the complicated phenomena of sexual 
polymorphism. Thus we may find eventually that the tend- 
ency to develop abortive ovaries is really inherited (through 
the fertile queens of course), and that differences in the chem- 
ical nature of the internal secretions, perhaps analogous to 
those which are supposed to obtain between castrated and 
non-castrated animals, may furnish the different stimuli that 
induce the larvae to demand of their own accord more or less 
food, or food of a different quality, and to develop accordingly 
into queens or workers. 
COLEBROOK, CONN., August Io, rgor. 
