501 



quantity of yolk within the eggs would have to decrease, or 

 the number of eggs would have to diminish. While the latter 

 may have occurred, there is no doubt that the former process 

 has predominated, till in the chalcid wasps we find eggs in 

 which yolk may be almost absent. 



Under the influence of the first factor (a), with increase 

 in specialization it became increasingly necessary that feeding 

 should occur earlier, and ever earlier in the free living period 

 of the insect. Under the influences of the second factor (h), 

 it became necessary that the larva should hatch in a more 

 and more incompletely developed condition. The result of 

 the co-operation of these two processes has been very marked. 

 In the less specialized insects as growth became more con- 

 centrated in the earlier part of the free living period, the 

 imaginal cells had to adapt themselves temporarily to rapid 

 feeding conditions; only after a considerable time did the 

 delayed differentiation occur. The metamorphosis here is very 

 simple (Odonata), and the cells which have already begun to 

 specialize in a certain direction (that of rapid feeding and 

 growing) can, nevertheless, evidently attain the imaginal con- 

 dition with comparative ease. The Coleoptera emerge from 

 the egg in a more primitive condition, and while some tissues 

 can still rejuvenate, others are unable to do so; they die and 

 the adult organs are formed from imaginal discs. In the 

 Muscidae this process has gone much further. But in the 

 chalcid wasps the divergence from the imaginal condition is 

 most marked of all. It it customary to regard the Muscids 

 as showing the most marked of metamorphoses, but in this 

 they are far surpassed, it seems to me, by the chalcid wasps. 

 The state of embryonic development in which their larvae 

 hatch has been pushed so far back that the head is still in a 

 bisegmentated condition, appendages are almost entirely 

 absent, malpighian tubes are not yet formed, the mandible 

 may still exhibit the crustacean palp, and the proctodaeal 

 ingrowth has not yet opened into the archenteron. So 

 specialized, moreover, has the larva become to absorbing food 

 rapidly, that it does not apparently even take time, as far as 

 I could observe, to excrete nitrogen. All its food is stored 

 in the form of hypertrophied larval cells within its body, and 

 not till the cells attain their critical volume does transforma- 

 tion occur. 



It is scarcely necessary to point out that as the insects 

 gradually developed metamorphoses, the instincts of the 

 parents had to undergo marked modification. As feeding had 

 to become concentrated at the beginning of life, they had to 

 deposit their ova in places where such food was to be obtained. 



