144 Comparative Classification of Cynipidae. 



much more developed. For the long and arduous work 

 which they have to accomplish, it is indispensable that 

 the terebra should be well lubricated, so that its functions 

 may be performed with ease and certainty. 



The larval state in the two generations presents also 

 some variations ; my researches in regard to these are 

 still fragmentary, and I can only note certain isolated 

 points. I will begin with the ^^^% and its evolution. 



However variable the duration of development 

 may be after the eggs are laid, as a rule embryonic 

 evolution begins at once, and it has no prolonged period 

 of egg-rest. Even in the case of eggs deposited in the 

 depth of winter, between November and February, 

 embryonic activity goes on from the first. This evolution 

 is naturally slower during the cold season, and con- 

 versely in the summer generations it requires much less 

 time, but even among these we have examples of an 

 embryo remaining a very long time in the ^%%. When 

 we find that an ^^^^ laid in December or January re- 

 quires some months for its embryonic stage, we can 

 understand it, because it is only when plant growth 

 commences, in April or May, that nourishment reaches 

 the embryo. But why should the same thing occur with, 

 for example, the eggs of Trigonaspis crustalis laid in the 

 end of May or beginning of June ? It seems difficult to 

 explain this long embryonic rest. Three whole months 

 of this latent existence pass away, and it is not until 

 September that the embryo bursts the shell of the ^g^y 

 and gall formation begins. It is impossible to suppose 

 that the conditions necessary for gall formation are more 

 favourable then than they were some weeks or months 



