68 



fresh leaves then remaining in the bags. They were, in the one 

 case, surrounded by the leaves on the other branches of the trees, 

 but fully exposed to the sunlight, the bag being attached to one of 

 the topmost branches of the tree ; but in the other case, the bag 

 was partially shaded by a wall. On taking the bags down I found 

 that the cocoons were similarly placed in each, the majority among 

 the old nests and debris that had fallen to the bottom of the bag, 

 but some few were constructed between the folds of the leno. The 

 cocoons varied greatly in colour, some being of an ashy bone colour, 

 others very dark brown, and various shades between these two 

 extremes ; but I could trace no connection between the colour of 

 the cocoon and its immediate surroundings; both light and dark 

 cocoons were in the folds of the leno, and among the debris, and, 

 indeed, in some instances, I found light and dark cocoons almost 

 touching each other. Nor can I suppose that the larvae were sub- 

 jected to any undue amount of disturbance beyond what was occa- 

 sioned by the removal of the bags from one branch to another, and 

 this was reduced to a minimum by the old branch being gently cut 

 off and the bag slipped over a fresh one, an operation that did not 

 take long to perform at the time when the cocoons were being con- 

 structed ; the cocoons were therefore made under conditions which 

 in this respect would pertain to larvse in a wild state, yet, so far as 

 I am aware, cocoons of this species so light as some of those exhi- 

 bited are of exceeding rare occurrence, if not altogether unknown, 

 in a state of nature. 



The remaining hypothesis, that light-coloured cocoons result from 

 the larvse having become sickly, is not so easily answered ; that in 

 this case they had an abundant supply of food during the latter part 

 of their existence is certain, but that they may have been on short 

 commons in some of their earlier stages is probable, as the bags 

 were not moved to fresh branches until the old ones were stripped 

 of their leaves, and this may have allowed the larvae to remain with- 

 out food for some hours. Then again, it will be remembered that, 

 the summer of last year, during which these larvee were feeding, 

 was abnormally wet, and larvae enclosed in a bag of ever so light a 

 material, would be likely to feel the effects of wet more, and to 

 derive less advantage from the invigorating effects of the sunshine, 

 than those fully exposed, and, for this reason, they may have been 

 to some extent sickly. No doubt such a state of affairs would affect 

 some individuals to a greater extent than others ; and thus, while 

 the most robust would be able to form normal cocoons, the more 

 sickly would be unable to perform their full functions, and would 

 produce imperfect (or diseased) cocoons, the particular form of im- 

 perfection in this case being in point of colour ; and 1 am supported 

 in this view by the fact that some of the lightest coloured of them 

 were of slighter construction than any of the dark ones. 



These results therefore appear to support Mr. Bateson's theory ; 

 at the same time I fail to see that they are antagonistic to the views 



