148 Mr. A. G. Butler on the 



Whilst admitting the force of many of Prof. Packard's 

 arguments and the truth of most of the facts upon which 

 they are based, it appears to me that at the outset there is 

 one great difficulty, viz. : If the arboreal habits of many 

 moth-caterpillars have tended to produce hairs and spines, 

 why is it that the greater number of butterfly-caterpillars so 

 ornamented feed upon low plants the Vanessce on nettles and 

 thistles, the Argynnides on species of Viola, the Satyridae 

 and Hesperiidae on grasses and low-growing plants ? Why, 

 again, are many of the most hairy moth-caterpillars, such as 

 that of Euprepia caja^ confined to low herbage ? 



Prof. Packard has shown that the spines and humps of 

 many larva? tend to assimilate them to their surroundings, 

 thus rendering them more liable to be overlooked by insec- 

 tivorous animals ; the existence of this resemblance, which 

 nobody will be inclined to deny, since all field-entomologists 

 have repeatedly observed it, can be explained by the action 

 of natural selection in preserving those individuals which 

 tend thus to become less conspicuous. In the case of spined 

 Geometrid larvae already assimilated to twigs it seems only 

 natural that the atrophy of the abdominal legs on the looping 

 portion of the larva should render outgrowths from other 

 portions of the same segments more probable, and when such 

 outgrowths tended more perfectly to conceal the larvae from 

 observation there can be little doubt that they would be 

 retained. 



A few of Prof. Packard's observations are open to question, 

 and to these I would now briefly advert. 



Speaking of the larva of Dryopteris at p. 490 he says it 

 " is as well fitted as that of Drepana by its protective 

 mimicry to avoid the gaze of birds and insect-enemies, while 

 its longer bizarre i tail ' renders it still more forbidding to any 

 insect assailants." Is this an ascertained fact or only a 

 supposition ? Has any bird ever whispered to Prof. Packard 

 that a caterpillar with a " tail " or with caudal appendages of 

 any kind is forbidding ? Is not all the talk about terrifying 

 colours, processes, and attitudes in caterpillars pure conjecture, 

 which experiment proves or will prove to be erroneous ? 



The larva of Centra vinula has two caudal processes with 

 exsertile tentacles, used for driving off ichneumons ; but the 

 fact that these processes are so used (as I have observed 

 personally) is not sufficient ; they are also supposed to alarm 

 birds, although they certainly do not. 



In like manner the colouring of certain caterpillars is said 

 by many able writers to serve as a danger-signal. It is quite 



