150 Mr. E. B. Poulton. Colour-relation between the [Feb. 4, 



by a male moth bred from a larva which had been fed for its whole 

 life upon Salix rubra (mentioned at p. 300), and which became a 

 yellowish intermediate variety. The tendencies were thus presumably 

 towards intermediate, or slightly on the yellowish side. A large 

 number of eggs were laid, and nearly all of them hatched successfully, 

 yielding apparently healthy young larvae, but the most extraordinary 

 mortality prevailed, so that no single larva arrived at maturity, or 

 indeed at an age which would render any conclusions possible (except 

 in the case of very marked colours which were not manifested). This 

 is all the more to be regretted because I had reserved by far the 

 greater part of this lot of larvae for some experiments wmich would 

 have conclusively decided some of the points in this difficult problem. 

 A few larvae were fed upon some of the same food-plants as in the other 

 instances, in order to gain further evidence as to their effect. Thus 

 larvae were fed upon Salix cinerea (in this case it would have been 

 interesting to ascertain the effect of the food upon two generations of 

 larvae, although the female parent only had been so fed upon 8. cinerea) 

 and upon 8. Smithiana (the leaves sewn together so as to expose the 

 under sides only). A few of the larvae were blinded before they had 

 seen the food-plant, by carefully painting over the ocelli with lamp- 

 black, a lens being used to make certain that all of the ocelli were 

 covered. This was a task of considerable delicacy and difficulty in 

 the small and restless larva, but when once accomplished the larvae 

 did not seem any the worse, and behaved in every way as the others 

 which were not blinded. As the lamp-black only formed an opaque 

 film over the transparent cuticular covering of the ocelli, and as the 

 former is thrown off at ecdysis, the pigment had to be renewed at the 

 beginning of each fresh stage, and the greatest care was necessary to 

 prevent the larvae from changing their skins at unexpected times, 

 and thus having the opportunity of seeing the food-plant. Hence 

 any larva which had ceased feeding before ecdysis was isolated and 

 only put back upon the food after it had changed its skin and the 

 pigment had been renewed. Larvae treated in this way were fed 

 upon the two food-plants which tend most strongly in opposite 

 directions — ordinary apple and Salix rubra, and at the same time 

 large numbers of unblinded larvae from the same batch of eggs were 

 fed upon the same plants. Had the larvae lived there must have been 

 conclusive evidence as to one obvious theory of the origin of afferent 

 impulses which determine the selection and the use of certain 

 proportions of the mixed vegetal pigments, and the deposition of 

 certain amounts and kinds of true larval pigments — the theory that 

 such impulses are caused by the colour of one or both sides of the leaf 

 acting as a stimulus by means of the ocelli. I am indebted to Pro- 

 fessor Gr. J. Romanes for the suggestion that experiments should be 

 made upon blinded larvae, while Professor E. Ray Lankester advised 



