290 Mr. 'Poulion''s further notes upon the 



Weismann argued, although it had not been observed 

 by Kleemann), the 8th stripe and horizontal stripe 

 above the subdorsal (anteriorly), the relation of the 

 large shagreen dots to the markings in the first stage, 

 and their long persistence on the thoracic segments, the 

 appearance of the purple stripes in the course of the 

 third stage, the change from a shagreened to a smooth 

 larva in the fifth stage, and the immense size of the 

 horn in the first stage, its bifid tip, and the fact that it 

 is movable. The great resemblance to a Smerinthiis 

 larva is also very interesting, the only essential diffe- 

 rences being the loss of shagreen in the last stage and 

 the acquisition of purple borders in the third, and the 

 shape of the head. It is probable that the ontogeny is 

 in some respects more primitive than that of Smerintlins 

 larvae. It is certainly more advanced in the later stages, 

 as is proved by the purple borders, the change to a 

 smooth skin, and the relatively early and complete 

 disappearance of the subdorsal. Nevertheless, in the 

 first stage the subdorsal predominates over the oblique 

 lines to a greater extent than in SmeiiiitlLiis, and the 

 horn is far larger, more distinctly bifid and it is movable. 

 It is very probable that all these are primitive characters. 

 It is certain that relative size is primitive, for the horn 

 is universally largest in the earliest stages. In this 

 respect S. Uiiitstri is probably the most primitive Sphinx 

 larva known. It is therefore likely that the other 

 characters are also primitive, and that additional know- 

 ledge concerning the horn of the ancestral Sphinx larva 

 is afforded in the early stages of this ontogeny. I shall 

 presently give reasons for the belief that the bifid 

 termination is a primitive feature, derived from a com- 

 parison of the larvae in which it has been observed. 

 It will further be rendered probable that the power of 

 movement is also primitive. 



2. FuKTHER Notes upon the Ontogeny of Smerinthus 

 ocELLATUS. — As I was rearing a number of these larv« 

 from the egg for the purpose of experimenting upon 

 phytophagic coloration, there was a good opportunity 

 for noting any additional facts in their ontogeny (which 

 is systematically given by Dr. Weismann). 



Stage I. — I have seen the young larva eating its egg- 

 shell after emergence, but it is never completely eaten, 

 and sometimes only enough for escape. Sometimes a 



