1886.] Larva of Smermthus ocellatus and its Food-plants. 153 



direction of yellow, while the hereditary tendency was presumably 

 weaker. At the same time the effect was more marked than in the 

 set of larvae to which the parents belonged ; and there was nothing at 

 all unusual in the results themselves. 



Populus nigra (No. 3). — There is little to be said about this result. 

 The larvae were too young to warrant any conclusion, but they were 

 whitish when they died. At the same time the larvae of Series III 

 (No. 14), which were fed upon this food-plant were also whitish 

 when young, while those that lived progressed in the direction of 

 yellow ; so that the most mature was an intermediate variety at the 

 time of its death. It is probable that the larvae of Series Y may have 

 also changed in the same direction if they had lived. 



It is noteworthy that the strong hereditary influence in the direc- 

 tion of yellow, which we should suppose existed in Series V (because 

 of the colour of the parent larvae), depends chiefly upon the male 

 parent ; and how far this element asserts iteelf in opposition to the 

 other sex is quite unknown in this class of experiment. Indeed, a 

 large number of data of this kind might be valuable in gauging the 

 relative strengths of the sexes in this form of heredity, but the 

 present data are far too limited to be regarded as a serious contri- 

 bution to this aspect of the subject. 



3. The General Results of the Breeding Experiments. 



It is now necessary to consider how far the questions suggested at 

 the beginning of this paper have received answers from the experi- 

 ments which have been detailed above. 



(1.) With regard to the first question, it is, I think, certain that 

 the larval tendencies towards certain colours are transmitted, as was 

 proved by the fact that the parent larvae had very strong tendencies 

 towards the whitish variety, while in the next generation only a 

 single yellowish form appeared out of seventy-five larvae. On the 

 other hand, there was conclusive evidence of the modified tendency 

 towards white in the offspring following the change wrought in the 

 parents by food-plants with strong tendencies. Thus, although food- 

 plants such as 8. rubra (tending strongly towards yellow) did not 

 produce yellow varieties, yet the larvae were, as a rule, yellower than 

 in the case of the parents. There was no difference between the 

 parents and offspring in the results of food-plants which tended 

 strongly towards white, these being strong enough to overcome any 

 ordinary hereditary tendencies. The results obtained by comparing 

 the different series together are less conclusive, but it is unfortunate 

 that a really satisfactory number of larvae was only obtained in one 

 case (Series III), the others being insufficient to afford any very con- 

 vincing comparison. The comparison between Series I and III was 

 certainly, as far as it went, in favour of a stronger tendency towards 



