130 THE EVOLUTION OF INSTINCT 
social life as among the workers of the hive bee. If so con- 
siderable a portion of this remarkable evolution has occurred 
since the separation of castes excluded the Lamarckian factor 
from playing any helpful part, we might conclude that the 
Lamarckian theory is not necessary in order to account for 
the earlier stages of social progress. 
In the hive bee we not only have the instincts of the more 
primitive forms carried to a high degree of perfection and 
specialization, but we have new instincts which must have 
arisen since the separation of the worker caste, since they 
have a direct relation to this very separation. Such, for 
instance, are shown in the remarkable behavior of bees 
when deprived of their queen. When the queen is removed 
from the hive and there is no other soon to hatch, the workers 
destroy a number of old cells and construct a new queen 
cell, and then proceed to feed'the young grub that is en- 
closed on liberal quantities of royal jelly by means of which 
it is caused to develop into a new queen. In this case of 
the instincts which lead the bees to regulate the supply 
of queens the theory of inherited effects of experience, 
as Spencer has advocated it, cannot apply. The instinct 
cannot have antedated the separation of the castes because 
it is based upon the existence of caste differences. 
Other difficulties are presented by the differentiations 
within the neuter class among ants and termites where there 
occur two or more kinds of sterile insects adapted by struc- 
ture and instinctive endowment for different functions in 
the community. In many cases the evidence clearly 
points to the evolution of new adaptive characters in the 
worker caste, instead of merely the degeneration of the 
fertile insects. And where this has occurred the Lamarckian 
theory fails to explain the facts. 
In the chapter on Instinct in the Origin of Species Darwin 
