1886]. Larva of Smerintlms ocellatus and its Food-plants. 137 



been given a different tint artificially by removing the " bloom " from 

 the under surface, and to test whether the ocelli formed the impres- 

 sionable part of the larvae by investigating the effect upon colour of 

 covering these organs with some innocuous opaque pigment. 



I also wished to further investigate the effect of certain food-plan! s, 

 about which the evidence was conflicting, and to carefully w T atch for 

 instances of individual variation among the larvae from the same batch 

 of eggs and fed upon the same food, to look out* for any indications 

 which would throw light upon the red-spotted varieties, and also to 

 further enquire into the periods during which. the larvae are most 

 susceptible to the influence of the food- plant. 



As far as these questions could be answered by work in the field, I 

 was very successful, for the larvae were even more abundant than in 

 the summer of 1884, and I was able to extend my area of observation 

 to Switzerland. I have also been exceedingly glad to be able to 

 reconcile the conflicting evidence given^in my last paper. But the 

 breeding experiments did not yield adequate results, considering the 

 immense amount of labour bestowed upon them. In the first place 

 the moths emerged in an unfortunate c^der — a great many males, and 

 later a great many females. Then of. those which emerged together, 

 there was great difficulty in obtaining such a system of pairing as I 

 was desirous of instituting, the result, being that I could get no larvae 

 with a strong hereditary tendency towards the.- yellowish variety ; and 

 these I was most anxious to obtain, because all ray bred larvae in 

 1884 tended very strongly in the opposite- direction. The eggs I 

 obtained in nearly all cases resulted from, pairing the moths which 

 came from these bred larvae (1884). Although disappointed in the 

 pairing of the moths, it seemed likely that the experiments would 

 yield sufficiently convincing results from the very comprehensive 

 scale on which they were conducted, for in July, 1885, I had many 

 hundreds of young larvae belonging to five different families. After 

 the great labour of bringing this large number through the most 

 delicate period of their lives, and just before the results appeared, the 

 larvae began to die off in hundreds, so that only seventy-five lived to 

 an age at which trustworthy observations could be made. I can only 

 suggest that this altogether exceptional mortality may have been due 

 to the excessive heat and dryness which prevailed at the time. I do 

 not think that it can have been due to the fact that both parents of 

 the majority of the 1885 larvae resulted from larvae belonging to the 

 same batch in 1884, because such interbreeding among moths does 

 not produce injurious effects, at any rate until after it has been con- 

 tinued for many generations. Besides, in one instance, the larvae 

 were the offspring of parents which came from quite different 

 localities, and these did not succeed auy better than the others. But 

 although a very small proportion of the larvae survived, they were 



