154 



HINTS ON THE ANCESTRY OF INSECTS. 



remarkable fact, as K we have previously noticed, that when it 

 begins to differ from the Caddis fly embryo, it begins to assume 

 the Poduran characters, and its development consequently in 

 some degree retrogrades, just as 

 in the lice previous to hatching, 

 as we have shown in a previous 

 chapter, so that I think we are 

 warranted at present in regarding 

 the Thysanura, and especially the 

 family of Podurids as degraded 

 neuropters. Consequently the Po- 

 duras did not have ah independent 18G - Embryo of Diplax. 

 origin and do not, perhaps, represent a distinct branch of the 

 genealogical tree of articulates. While the Poduras may be said 

 to form a specialized type, the Bristle-tails (Lepisma, Machilis, 



Nicoletia and Campodea) are, as 

 we have seen, much more highly 

 organized, and form a generalized 

 or comprehensive type. They re- 

 semble in their general form the 

 larva of Ephemerids, and perhaps 

 more closely the immature Perla, 

 and also the wingless cockroaches. 

 Now such forms as these Thysa- 

 nura, together with the mites and 

 the singular Pauropus, we cannot 

 avoid suspecting to have been 

 among the earliest to appear upon 

 ]-.yl the earth, and putting together the 

 facts, first, of their low organiza- 

 tion; secondly, of their compre- 

 hensive structure, resembling the 

 larvae of other insects ; and thirdly, 

 'am o f their probable great antiquity, 

 we naturally look to them as being 

 related in form to what we may 

 187. Embryo of Louse. conceive to have been the ancestor 

 of the class of insects. Not that the animals mentioned above 

 were the actual ancestors, but that certain insects bearing a 

 greater resemblance to them than any others with which We are 

 acquainted, and belonging possibly to families and orders now 



dZL 



