PROTECTIVE COLORATION AND DEFENSIVE STRUCTURES OF LARVAE. 87 



Poulton, however, has more recently carried out elaborate experi- 

 ments on various species, which tend to show that the response of 

 larval to their environment is due, in some instances, to phytoscopic 

 rather than to phytophagic causes, and that it is the colour of the 

 surface of the leaf, rather than its substance, that acts as the stimulus 

 in producing the different colours of larvae under varying conditions of 

 environment. As a case of extreme specialisation in larval coloration, 

 we would instance Abraxas grossulariata. The typical form of larva of 

 this species must be Avell-known, but examples may be obtained in 

 London gardens, and probably elsewhere, in which the colour is almost, 

 in others, absolutely, black, and yet there is no corresponding difference 

 in the colour of the imagines, the difference in colour being purely 

 adaptive to the needs of concealment in the larva. The larvae of the 

 Catocalids (and the allied genera, Flomoptem and Pheocampa), the tree- 

 feeding Notodonts and Geometrids, and many Lasiocampids, are 

 spotted and mottled with various tints of brown, grey, and ash, so 

 that their colours assimilate with the colours of the bark of the trees 

 on which they rest. Such larvae are also frequently provided with 

 dorsal and lateral humps and warts, so that they also resemble the 

 shape of the twigs (with their knots and leaf -buds), and thus make the 

 resemblance more complete. We have already mentioned that the 

 larvae of Amphidasys betularia are very variable — whitish grey, different 

 forms with varying shades of brown, to quite brown, whilst others take 

 another direction, and are of a distinct green hue. 



The experiments carried out by Poulton to demonstrate the 

 response of larval colours to the environment are detailed at 

 length (Trans. Ent. Soc. Lond., 1892). His experiments, based on 

 the subjection of the larvae throughout their existence to varied 

 conditions of environment, tend to show that whilst some larvae 

 are affected by the conditions of their environment, and effectively 

 respond thereto, others are quite obdurate and remain constant, whatever 

 the conditions of their environment may be. Poulton's conclusions 

 work out as follows : (1) Eegularly dimorphic forms, with inter- 

 mediate varieties rare or wanting, are never, so far as our present 

 knowledge extends, susceptible to surrounding colours, while variable 

 species tend to be so. In this respect, Geometra papilionaria is very 

 interesting, being susceptible when young, but not later, when it is 

 dimorphic. (2) The larvae of Noctuids are far less sensitive to change 

 than those of the Geometrids. The most susceptible of the former, 

 the Catocalids, are arboreal, and specialised for concealment among 

 twigs and on bark. In the Catocalids there is sometimes a most 

 extraordinary fluctuation in the amount of susceptibility within the 

 limits of the same genus. (3) Only among the Geometrids were 

 green larvae produced by the experiments. Out of eleven species 

 operated upon, all but one showed some sensitiveness to colour sur- 

 roundings. (4) There is no evidence to show that the colour acquired 

 by a larva can be transmitted to its progeny ; the susceptibility is 

 simply an adaptation to the differing environments in which the larvae 

 find themselves. 



Concerning the time necessary for the colour change to appear, 

 Poulton gives the following : — 



