THE EVOLUTION OF INSTINCT 127 
has been emphasized by Weismann and others; but, as 
Spencer has ably shown, the case is not so easily disposed of. 
According to Spencer the fact that certain structures and 
instincts occur in the workers which are not found in the 
fertile insects is not because the workers have acquired them 
since their separation as a distinct caste, but because the 
fertile insects have lost them. The worker caste is produced 
by lack of sufficient nutrition. There is a stunting of growth, 
a failure on the part of the reproductive organs to develop, 
and in ants an atrophy of the wings and wing muscles. In 
the early stages of the evolution of the social state the fertile 
female possessed all the instincts for making the nest, 
gathering food and caring for the young, as is now done in the 
more primitive social communities of bees and wasps. In so- 
cial wasps, as a rule, the occupants of cellsrichly supplied with 
food emerge as fully developed females; where the food supply 
is limited the females are smaller and sterile, and with varied 
amounts of food various intermediate gradations of size are 
produced from the largest females to thesmallest workers. The 
smaller individuals which on account of the partial atrophy 
of their reproductive systems take little or no part in repro- 
duction busy themselves with building cells and storing them 
with food, and in taking’ care of the young. The fertile 
females live with the rest of the community and share its 
labors, but in the following year they scatter and form the 
nuclei of new colonies. At first the queen starts the nest, 
rears and cares for the young, and performs all the tasks in- 
cidental to her small household economy. When the young 
emerge they codperate in the work of the queen. As the 
community grows the queen gradually withdraws herself 
from the labors shared by the workers and devotes herself 
more and more to laying eggs. 
Among bees there are numerous gradations between the 
