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



white in the former series, such as we should expect from the parentage. 

 Series II is the one about which there is so much obscurity, but its 

 results were rather irregular when compared together and with those 

 of the other series. In Series V we should expect a greater pre- 

 dominance of the yellowish tendency, if the male parent is of equal 

 importance with the female in this respect, but the data were very 

 insufficient. 



But it must be clearly understood that the question is really settled, 

 because of the wonderfully uniform results of the comparison between 

 parents and offspring as a whole, in which comparison we are dealing 

 with strong and definite tendencies ; while in the case of the offspring 

 we are considering delicate differences between such tendencies, 

 which are obviously much more difficult to detect and need far larger 

 data for their accurate determination. 



(2.) As to the second question, I think it may be said that con- 

 clusive experimental proof has been afforded of the theory brought 

 forward in my last paper — that the colour of the leaf, and not its 

 substance when eaten, is the agent which influences the larval 

 colours. It seems to me that this is proved by the breeding experi- 

 ments in Series III, in which the larvae from the same batch of eggs 

 were whitish intermediate and white after being respectively fed upon 

 8. viminalis and upon leaves of the same plant sewn so that only the 

 under sides were visible (Nos. 6, 7, and 8). 



On the other hand an intermediate variety was produced by feeding 

 a larva from another batch of eggs upon similar leaves sewn so as 

 to expose the upper sides. (Series V, No. 1.) The same thing is 

 proved by a comparison of two sets of larvae from the same batch 

 of eggs, fed respectively upon 8. triandra and upon the leaves of the 

 same plant from the under sides of which the whitish bloom had been 

 removed. 



Concerning the food-plants, about which the evidence was con- 

 flicting, the experiments have in some cases helped to clear up the 

 difficulties. The greatest of these difficulties concerned S. viminalis 

 and 8. triandra, but in the latter case there really was no confliction 

 of evidence, as Mr. Boscher's white larvae were found upon the very 

 similar but much whiter S. alba. As to S. viminalis, the difficulty 

 does not at first sight appear to be cleared up by the breeding experi- 

 ments, but I will defer its consideration until after detailing my 

 experience with captured larvae, for what I believe to be the correct 

 solution presented itself to me from the results of this part of the 

 investigation. The experiments upon crab produced exactly the same 

 results as in 1884 : this will also be considered later. With regard 

 to other food-plants, the view I previously expressed that S. Smithiana 

 tends to produce whitish intermediate varieties, is on the whole 

 supported, and so also in the case of 8. babylonica, which as I 



