64 Wheeler, The Polymorphism of Ants. [\^ol. XXIII, 



stored up in the larva were insufficient, and second, because immediately 

 after hatching, the young females had to devote themselves to the care of 

 the numerous larval colony, and especially to the function of nursing, which, 

 as we have shown experimentally, prevents the maturation of the eggs. 

 The mother being relieved of the care of feeding the larvae and having 

 over the workers the advantage of ovaries swollen with eggs, while the other 

 females at the time of hatching contained only immature germs, continued 

 to lay to the exclusion of the other females, whose nutricial functions kept 

 them in a condition of sterility. 



"At the end of the year, however, owing to the greater length of the 

 imaginal than of the larval period, and also on account of the diminution 

 in the egg laying of the queen or owing to her disappearance, the adult 

 colony came to surpass the larval colony sufficiently in numbers to permit 

 the later broods to receive and store up the reserve food which is indispensable 

 to the maturation of the eggs. Hence these later broods alone would be 

 able to hibernate and reproduce the species during the following spring. 

 Only these individuals, therefore, should be regarded as representing the 

 stirp from which all the future individuals proceed, the others being naturally 

 eliminated from the genealogical tree. 



"This single fact, namely, that only the individuals reared at the end of 

 the season participate in the direct lineage of the different generations, suf- 

 fices to account for a modification in the germ-plasm of the species; for 

 without even adducing the specific instinctive dispositions which would later 

 be acquired by natural selection, the conditions of nurture, and in particular 

 those of nourishment, which control the development of the animal till the 

 end of the year, depart from the mean of the variable conditions of nurture 

 to which the evolving presocial insect was submitted. Under the influence 

 of this modification acting as external conditions, constantly and always in 

 the same manner, a new physicochemical constitution of the germ-plasm 

 must necessarily arise and a new direction of development be opened up, 

 namely, that which leads to the queen type: 



"If the preceding theory be granted, the realization of the worker type 

 becomes equally comprehensible. As long as there is no perceptible varia- 

 tion in the constitution of the germ-plasm, the worker will differ from the 

 queen only in slight quantitative morphological variations, depending essen- 

 tially, as we have demonstrated, on the phenomena of nutrition. This is 

 the case in Polistes and less obviously in certain species of wasps (Vespa) 

 which present an uninterrupted series of forms connecting the worker and 

 queen types. 



"But as soon as a perceptible modification of the germ-plasm intervenes, 

 matters cease to be the same. The egg laid by a queen in the spring, like 



