ANCESTR Y AND CLA SSIFICA TION. 231 



arranged in regular rows like the surface of the 

 ocellar riband in the butterfly chrysalis ; while in 

 the perfect stage of both these prominences have 

 crowded together, met, and therefore assumed a 

 hexagonal outline. This condition of the eye in 

 the chrysalis, immature, yet at present altogether 

 unnecessary and superficial, and partly concealed 

 by another organ, clearly points to an earlier 

 stage of usefulness, when it was in direct connec- 

 tion with the underlying parts, and when the 

 organs now concealing a portion of it were ex- 

 tended and free, that is, an active pupal stage. 

 But activity in the pupal stage implies something 

 more ; such differences as now obtain between 

 the mouth-parts of the caterpillar and of the 

 imago require a period of inactivity for their 

 alteration : but if no such period existed, there 

 must have been similarity of mouth-parts in the 

 archaic caterpillar and imago. Was it then a 

 biting or a sucking insect ? We find nothing in 

 the present structure of the group which answers 

 this question, but is it not possible that the insect 

 may have possessed a mouth, such as Lubbock 

 describes in certain low insects, which was neither 

 distinctly biting nor sucking, but of " a peculiar 

 type, capable of modification in either direction 

 without loss of utility" ? Or, as Darwin writes 

 with regard to insects in general, we have only to 

 suppose " an upper lip, mandibles, and two pair 



