behueen lepidopteovus larvai and their surroundings. 317 



hybernation is shown in either fig. 4 or fig. 9 on Plate 

 XVII. The latter appearance was a response to an en- 

 vironment of lichen-covered sticks; but after October 16th 

 these were replaced by black-barked twigs, which it is seen 

 produced no effect at all. Again, Plate XVIII, fig. 2 repre- 

 sents a later stage of Plate XVII, fig. G. Here too the 

 resemblance between older and younger larvse is very close, 

 although the former had been subjected to the same black 

 environment after October 16th. The negative result 

 of a transfer experiment in the opposite direction is seen 

 in Plate XVII, fig. 14, the representation of a larva which 

 had been exposed to lichen-covered sticks after October 

 3rd. All the larvse did not remain as uniform throughout 

 their life-history as these three. Thus Plate XVIII, fig. 

 3 represents a nearly mature larva of which the appear- 

 ance before hybernation is seen in Plate XVII, figs. 4 

 or 9. In this case the larva darkened considerably after 

 the winter, although its environment had not been shifted, 

 but consisted of lichen-covered sticks throughout. 



It is probable that the power of adjustment to environ- 

 ment possessed in so marked a degree by this species is 

 specially directed to protection during hybernation, when 

 the food -plants are leafless, and when enemies are often 

 pressed by hunger. But it is doubtless also of importance 

 later on when the larva becomes so much larger and would 

 on this account be far more conspicuous. It is probable, 

 however, that the caterpillar does not wander from its 

 food-plaut, and that complete adjustment to the old wood 

 before hybernation is an adequate defence in the following 

 spring and summer. If this be correct there would be no 

 advantage in a prolonged larval susceptibility. 



The same relationship between susceptibility and the 

 particular needs of each species is seen in the effect of 

 an environment of green leaves and shoots upon G. quer- 

 cifolia, 0. Jjidentcda, and A. hehdaria. The first-named 

 probably invariably rests by day, except for a brief period 

 after leaving the egg, upon the older wood, and the power 

 of adjustment to leaves and young shoots, being altogether 

 useless to it, has never been acquired. The last-named, 

 with its remarkable range of food-plants, including many, 

 such as broom or rose, in which green shoots are a 

 prominent feature, is frequently in a position in which 

 a green colour would best conceal its nearly smooth and 

 cylindrical form ; and we find that, as a matter of fact, 



