40 (5 



Mr. G. A. K. Mnrsliall en 



1)0 (jiutc unrelated to dryness or cold (see pp. 443-8), But 

 it would probably be necessary to apply artificial conditions 

 to the larval stage. Indeed, the fact that the winter 

 phases of certain species of frccis are so very much larger 

 than the summer phases seems to require the conclusion 

 that the change is pre-determined during or previously to 

 the stage in which material is accumulated. 



The differences in weight are well shown in captured 

 individuals of two species in the following list: the method 

 of proceckire has been already described on pp. 452-3. 

 It is seen that the dry phase always weighs more and 

 sometimes over twice as much as the wet one. 



It may be argued that the results from captured speci- 

 mens are untrustworthy because some females will have 

 laid their eggs, some males will have paired, and others 

 not. The five series of specimens of sesdnii's an<l nataliynHia 

 bred by Mr. Marshall an- not open to this objection and 



