306 Mr. Vonlton sfidiJier notes upon the 



three, arrived at maturit3^ ki the same time the effect 

 produced was not so great as I expected in the case of the 

 yellowish forms, which were hardly more than intermediate 

 varieties. But there can be no doubt that some effect 

 was produced, and the subject must be considered settled 

 to that extent. The ova were bought from Mr. Davis, 

 of Dartford, and the numbers selected were too great 

 for accident to have caused the results (which were 

 uniform in each of the five lots). At the same time my 

 experience in the field this year has been different from 

 that of previous years (recorded in my last paper), in 

 that I have come across many instances of larvae upon 

 food-plants which tended towards the variety other than 

 that found. The most noteworthy instance was that of 

 a bright yellowish variety upon apple, and of two 

 opposite varieties on a tree of Salix fcrrugi7iea (?). 



Hence the question is more complicated than it 

 formerly appeared to be. It seems that the only way 

 in which the results of the breeding experiments and 

 experience in the field can be correlated is by supposing 

 that phytophagic effects are hereditary and gradually 

 accumulate until the influence of a food-plant during a 

 single larval life may not be sufficient to overcome the 

 inherited tendencies following from the effects of another 

 kind of food upon many generations of larvae. Thus in 

 my breeding experiments there was a much greater 

 tendency towards the whitish rather than the yellowish 

 variety. I was unable to ascertain if this was due to 

 the food of the parents, as Mr. Davis informed me that 

 his larvae had been kept together (having been found 

 upon various species of food-plants). As a matter of 

 fact, however, there was already evidence l)efore me 

 (last year) that the solution of the difficulty is not so 

 simple as it appeared to be. I read Mr. Meldola's notes 

 upon this subject (in Weismann's ' Studies in the Theory 

 of Descent,' part ii.), and assumed that his instances 

 pointed in the same direction as my own observations. 

 On reading them a second time I found that I had made 

 a mistake (reversing in my mind the effects of the two 

 food-plants), and that the instances are in exact opposi- 

 tion to what I should have expected. He quotes in- 

 stances of numerous yellow-green larvae being found 

 upon S. viminalis and white-green upon S. triandra. 

 Now these results are the very reverse of protective, for 

 ;S'. viminalis has leaves with very white under sides, and 



