TIIK KV(i[.rTr<)N OF TFIK I.Fl'I noI'TEItms ni'S. L'( 



Evolution did not, however, proceed along a single line, and at some 

 point there branched off the Pterophorina, which have only three 

 abdominal segments in the thoracic mass, but differ from all the species 

 we have hitherto considered, in rarely having a cocoon, and in being 

 fixed by the tail and not possessing any longer the jiower of locomotion 

 shown in other " Micros " by forcing themselves out of their cocoons 

 when the moth emerges. 



Another line, starting off from the main trunk very early in the 

 evolutionary process, since all its forms have still only two al)dominal 

 segments involved m the thoracic mass and all force themselves out of 

 their cocoons for emergence, differs in that tlie larva? feed more or less 

 exposed, the larvaj of all the others so far (except of some Pterophcrina) 

 feeding under webs, witliin leaves, stems, seeds, A'c. ; this line, how- 

 ever, reaches a very advanced point in the Zijijnenidae. 



It is a very curious circumstance, and one that has a deeper mean- 

 ing than I have yet been able to fathom, that, with the exception of 

 Lithocolhtes and Gradlaria, all the forms in which the first four 

 abdominal segments are included in the thoracic mass, also differ in a 

 very important particular from those so far considered ; they lose the 

 7th abdominal segment as a free one in the male, the 5th and 6th alone 

 remaining so, and in some cases these also become fixed. It seems very 

 probable that this step was taken in several different lines of develop- 

 ment, or if not, that divergence began immediately after it had been 

 taken. 



'J'he Butterflies originated ver}^ low down, probably as low as the 

 Ilepialidae, probably having Castnia as a stage, or rather as a side branch 

 from a low and extinct portion of the butterfly stem. (This is usually 

 what is meant when speaking of a lower form as representing a stage 

 of a higher, and not that it is at all a lineal ancestor). The Macros cer- 

 tainly had a separate origin, and represent the highest point on another 

 hne of development ; I have already alluded to their chief charac- 

 teristics. The CIdoephon'fJae {prasiaana, chlorana, Sec), have reached a 

 high stage of development, l)ut probably along a separate line. Their 

 pupal characters are almost identical with Macros. Then there are 

 a large number of families that have a pupa largely of the Macro 

 type ; the thoracic mass is solid and includes four abdominal 

 segments, and only the oth and 6th are free ; these are to a large 

 extent no doubt side branches from the main stem supporting the 

 Macros, though some may have had a separate origin. They are all 

 separable from the true Macros, by their larvae possessing prolegs with 

 complete circles of hooks ; but they also differ as pupaa in that almost 

 every family has one or more characters which the Macros have lost, but 

 which the eaidier forms possess ; many have the dorsal head-plate, 

 others have some trace of the maxillary palpus, some dehisce after the 

 manner of the earlier forms, the head- and antenn!i3-pieces separating 

 together, whilst a few even retain the eye-cover attached to tlu' dorsal 

 head-plate. Tliese are represented by the Pvrales and their allies tlu; 

 Crambt, Phycid.*;, ttc., as well as by families hitherto jjlaced with 

 TiNEiNA, such as Hi/poiiomentidnf', Szc. 



The lowest Putterllies, the Skippers, are very parallel with this last 

 section. The mass of true butterflies differ but little essentially from 

 the Macros, but, unlike them, do not settle down in that form in a fixed 

 manner, but proceed in various ways to higher points, until what we 



