538 W. L. TOWEK, 



this is perhaps what one might expect, although they have not put 

 forward any conclusive evidence upon this point. Their evidence was 

 the venation, which might have arisen secondarily from mechanical 

 necessity so that the resemblance between the wings of insects in this 

 respect would have been purely accidental. I have, 1 believe, presented 

 here for the first time observations showing that the development of 

 the wings in insects with complete and incomplete metamorphosis is 

 essentially the same, but is obscured or complicated by conditions 

 prevailing in the more specialized orders, undoubtedly the type of 

 wing development found in most of the Coleoptera examined is one 

 derived directly from Heterometabolic ancestors and has persisted in 

 nearly its original condition as regards development, although retarded 

 or postponed in its activities until the final larval stage. 



In those forms which have the wings in the larva lodged in a 

 shallow pocket of the hypodermis we encounter the first specialization. 

 In the Scarahaeidae the condition as regards wing development is 

 peculiar. In some species as in 0. scahra the wings arise late in 

 larval life or shortly before pupation, but in L. fusca et al. the wings 

 are present in the larvae of the first year as a typical recessed imaginai 

 disc. It remains dormant, however, until just before pupation, when 

 growth begins. The diversity in wing development shown by these 

 two species is difficult if not impossible of explanation at the present 

 time. We may of course say, that in L. fusca the wing appears at 

 the normal time, and that in 0. scahra^ because of the long larval life 

 the wings have been retarded from generation to generation until now 

 it does not begin to develop until late in the larval period. This, 

 however, is not an explanation, but a statement of the stages through 

 which we suppose it passed in reaching its present condition. At 

 present there is not enough known of insect metamorphosis from the 

 standpoint of experiment to permit of attempting to explain phenomena 

 like these. 



Both of the preceding types of wing development have been 

 found in beetles having a long larval period, extending over several 

 months or even years, but in the families having short larval periods, 

 a diflerent and more complicated type of wing development is found. 

 In beetles with long larval periods the wing requires from ten weeks 

 (Phymatodes) in which to complete its development to fifteen weeks 

 {Osmoderma)^ and the type of wing development found there is obviously 

 adapted to the larval conditions, since no ec^yses occur where the wing 

 may become prematurely an external structure, and the slow wing 



