PROTECTIVE COLORATION AND DEFENSIVE STRUCTURES OF L&KVM. 85 



of this description in the larvae of Eutricha quercifolia and Poecilocampa 

 populi, which enable them to rest more securely on branches and 

 twigs without throwing a sharp shadow. Poulton shows that in 

 Geometric! larvae such protection as is afforded by growths of this kind, 

 necessitates their development only at those parts between the two 

 posterior pairs of prolegs, where the lasva is in contact with the twig, 

 because, the -bark of a twig and its branch being continuous, anything 

 that suggested a deep furrow between them would destroy the pro- 

 tective resemblance. On these parts of the Geometrid larvae, therefore, 

 we find such minute fleshy outgrowths frequently developed. In the 

 larva of K. luteolata, the fleshy processes exactly correspond with that 

 part of the body which would otherwise come as a dark shadow in the 

 deep cleft between itself and the branch. The processes appear to 

 soften the contact between the larva and its food-plant, not only by 

 partially filling up the cleft, but also by neutralising the shadow in the 

 groove which remains. These lateral growths are very noticeable in 

 the larvte of Catocalids, Lasiocampids, etc., and are also well developed 

 in the larvae of Metrocampa manjaritaria and Aventia flexula. 



It is difficult to deal with the broad question of the influence that food 

 has on the colours of larvae, but some short account appears to become 

 absolutely necessary at this stage of our enquiry. In the first place, 

 it appears advisable to point out that, in the larval stage of insects, 

 it is of the utmost importance that the storage of reserve material on 

 which the organism can draw, to enable it to undergo its later trans- 

 formations, should be effectually carried out. We find in lepidopterous 

 larvae that this need completely overwhelms the necessity of perfect 

 assimilation, and, hence, material assimilated in a more or less un- 

 altered condition, may frequently carry with it the cruder constituents 

 of which the food is made up. In this way only is it possible to 

 imagine a slightly modified form of chlorophyll becoming transferred 

 to the animal tissues, and hence affecting directly the colour of the 

 larva. 



McLachlan noticed, in 1874, that flower-feeding lame often assume 

 (in the same species) the colour of their food. Meldola considered 

 that this might be due to the colouring matter of the food being assimi- 

 lated in an unaltered state (E. M. M., xi., p. 162). Later, Meldola 

 pointed out that it was probable that the food-plants directly influenced 

 the variation found in the ground-colour of the larva of Smerinthiis 

 ocellatus, and Poulton described at length experiments made upon this 

 species, by feeding the larvae ab ovo on various food-plants. In these 

 experiments, apple (cultivated and crab) gave whitish-green larvae ; 

 Salix cinerea and S. rubra, larvae inclining to the yellowish form ; 

 <S. viniinalis, intermediate forms. In the field, although this general 

 tendency is observable, there are many striking exceptions, and the 

 latter, probably, may be explained by supposing that they are due to 

 hereditary influences, and that such larvae are not, therefore, able to 

 take full advantage of their food as a means of protection. Poulton 

 further notices the darker coloration of larvae of Sphinx liyustri, when fed 

 on ash: or lilac, compared with that of those fed on privet. The former 

 are greyish-green and the purple stripes duller. In 1885, Poulton 

 gave (Proc. Roy, Soc. Land., p. 269) a number of details relative to 

 the way in which altered plant pigments are utilised in larval colora- 

 tion. He considered that the green ground colour of many lepidop- 



