574 Dr. Thomas Algernon Chapman on the 



may state what I believe to be the law on this point in 

 this way. The Lepidoptera certainly cannot be arranged 

 in one line by their pupee, but the Lepidoptera of one 

 line can be arranged by their pupae. 



Two forms having different pupae, are either — 



1 . Not related (that is not nearly enough for definition). 



2. If related, then the pupae will tell which is descended 



from the other or which is furthest from the 

 common ancestor. 



A pupa with more movable segments cannot be 

 descended from one with fewer ; broadly, one that is more 

 consolidated or has lost certain parts, cannot be ancestral 

 to one with more movable parts, or that retains 

 appendages. 



I have been forced to conclude that there have been 

 several lines of evolution in the Lepidoptera all more or 

 less parallel. That the Ohtectse in fact are not a homo- 

 geneous group and were not derived from the Incom- 

 l^letx at one point only, and then diverged, bat that the 

 obtect pupal form was reached from several different 

 stirps of Incompletx independently under the pressure 

 of the problem placed from the first before the Lepi- 

 doptera, viz., what is the line of least resistance by which 

 the imago may escape from pupa case and cocoon without 

 the aid of mandibles ? I am not prepared to say how far 

 back it was that these several stirps diverged. I still 

 see many reasons for deriving them from diffeient 

 branches of the Paljeolepidoptera ; on the other hand, 

 I see that it is necessary to admit that, if there can be 

 parallel lines in the neolepidoptera, there may equally 

 be lines in the neolepidoptera parallel to others in the 

 Palaeolepidoptera, just as we have parallel families in the 

 marsupials and placentals, and even in the Reptilia. 



My arch heresy in this respect is in claiming a relation- 

 ship between Cossus and Hepialus. Hepialus (starting 

 from Micropteryx) presents many characters similar to 

 Cossus, and in some respects is even further removed 

 from Micropteryx than Cossus is. If we grant a parallel 

 progression to Cossus and Hepialus, some of my objections 

 to the orthodox view would be diminished. 



It is no doubt very diflScult to admit that neolepi- 

 doptera with the subcostal vein of the hindwing simplified 

 in precisely the same manner, should be equally derived 

 from a Hepialid and Micropterygid source, but there are 

 so many forms missing that it is not safe to form any 



I 



