1886.] Larva of Smerinthus ocellatus and its Food-plants. 157 



in other ways which are suggested in the larval ontogeny. (See 

 " Proc. Ent. Soc. London," 1885, Part II, August, pp. 290—296.) 

 The newly hatched larvae are always brightly yellowish even when fed 

 upon apple. The particular form of protection now gained by the 

 larva, by a resemblance to the foliage of its food-plant, has involved 

 the laying aside of this ornamentation, but some of its features 

 occasionally appear (by reversion), and when this is the case they are 

 associated in nearly all cases with the ancestral ground-colour. It is 

 possible that the differences of ground-colour which are now de- 

 pendent on the food-plant arose independently, and persisted for a 

 long time as ordinary cases of dimorphism or polymorphism, and 

 that their relation to the colour of the food-plant was determined by 

 natural selection at a much later date. But although the differences 

 may have commenced in this way, they did not probably reach any- 

 thing like their present condition until they came to depend on the 

 food-plant, for without such a relation the colours would often render 

 the larva conspicuous instead of protecting it. 



4. Observations in the Field upon Larvai of S. ocellatus during 1885. 



The larvae were very abundant last year and the results were more 

 uniform than in 1884. An account of all the captured larvae is given 

 below. 



August 2nd. — Upon Salix rubra in some fields by the River Cher- 

 well, near Oxford, seven nearly full-grown larvae and one small in 

 the fourth stage, which was a bright yellowish variety. Of the 

 former number four were bright yellowish varieties, and three were 

 well on the yellow side of intermediate, almost good yellow varieties. 

 Also upon 8. cinerea in the same locality, one nearly adult larva 

 which was a good yellow variety. Also upon a small tree of 8. baby- 

 lonica in a garden at Oxford (the same tree upon which seven larvae 

 were found in 1884; see "Proc. Roy. Soc," No. 237, 1885, pp. 301 

 and 302), one nearly full-grown larva decidedly on the yellow side of 

 an intermediate variety but not strongly yellowish. 



August 3rd. — Upon 8. viminalis on the River Cherwell, near Ox- 

 ford, three larvae in the last stage (upon the same tree) of which one 

 was nearly adult and rather yellowish, but not more than an inter- 

 mediate variety ; while the other two were much less advanced in the 

 last stage, and were whitish varieties. 



August 4tth. — Upon 8. rubra by the Cherwell (as above, August 2), 

 two nearly adult bright yellowish larvae. 



August 9th. — Upon 8. rubra by the Cherwell (as above), one nearly 

 adult bright yellowish larva. Also upon S. cinerea in the same 

 locality, two almost full-fed larvae which were good yellow varieties. 

 Also upon . 8. linearis in the University Parks, one larva which was 

 oerhaps slightly on the yellow side of an intermediate variety. 



