10 



the first place these ventral prolegs, in all the species I have yet 

 examined, are very small and weak when compared, for instance, 

 with Sphingid, a typical Noctuid larva. When the larva is dead it is 

 difficult to see them at all, though usually the very imperfect circles 

 of black hooks situated on the discs of the prolegs serve as a guide. 



When, however, a living larva is held in such a position that a 

 lateral view of the ventor is obtained these prolegs may be very easily 

 observed, for the larva will thrust out one pair after another in its 

 vain endeavours to free itself from restraint. 



If we carefully examine, by this and other means, certain species 

 of this genus, we shall see that they, like the (iracilariids, are without 

 prolegs on the sixth abdominal segment. They have already lost one 

 pair. 



When, in prehistoric days, the Geometrid larva began to find some 

 of its ventral prolegs more hereditary than useful, it was, no doubt, 

 as Mr. 'I'utt suggests (" Brit. Lep.," vol. i, p. 53), merely a matter of 

 speed which determined which pair of the ventral prolegs should go 

 first. The Geometrid larva, when arching the body to crawl quickly, 

 and bringing up the posterior segments rapidly, gave the anterior 

 pairs of prolegs less time in contact with the surface crawled on — that 

 is to say, made least use of these pairs, and therefore these anterior 

 pairs of ventral prolegs were the first to be lost. Then, as more work 

 was thrown on the last pair of ventral prolegs, namely, those situated 

 on the sixth abdominal segment, this pair became strengthened and 

 enlarged, and, subsequently, to such a degree that the third pair 

 situated on the fifth abdominal segment became unnecessary and 

 therefore disappeared too. As examples illustrating this idea we 

 may compare the larvfe of Gonodontis {Odo/i (opera) bidentata, Anisop- 

 teryx cesciilaria, and other Geometrids. 



The movements of the Coleophorid larva are of the methodical 

 order. For instance, even when making a new case there is none of 

 the excitement and rapid motion that we see in some other families 

 of the Tineina when making new dwellings. Therefore in the 

 Coleophorid larva there is no question of speed involved, and con- 

 sequently we should not expect the process of losing the prolegs to 

 be of the same nature as it was in the Geometrid larva. From the 

 persistent manner in which the abdomen of the Coleophorid larva is 

 held in a curved position, as previously described, we may, I think, 

 argue that the position itself offers some advantage to the larva, and 

 that, also, this advantage was enjoyed by, and this position adopted 

 by, the early ancestors of the modern Coleophorids. Now when the 

 larva is crawling by aid of its true or thoracic legs and holding the 

 abdomen in the arched position with the anal segment persistently 

 incurved and resting on the surface crawled on, the pair of ventral 

 prolegs least likely to come in contact with the surface crawled on 

 will be those nearest the incurved anal pair. My meaning may be 

 made plainer by making the sign used as a note of interrogation 

 represent the larva by writing it horizontally, thus — , the straight 



