colour of cocoons, jnipce, and larvce. 207 



With one exception all these cocoons are of the full dark 

 colour. The exception is also a brown cocoon, but it is 

 very thin and deficient in substance, and consequently 

 of rather a lighter colour. It is one of the 19 named 

 above. 



B. Forty-four larvae were enclosed in a sleeve of black 

 muslin, and placed on another branch of the same bush. 

 Into this sleeve I put a quantity of crumpled brown 

 paper, of the darkest colour I could get. On opening 

 this sleeve in September, I found 48 cocoons, namely — 



2 in brown paper. 



1 between paper and leaves. 



4 on the black sleeve. 



31 in the leaves, or massed against each other. 



All these were of the full dark colour. I should say that 

 the brown paper had become so much bleached by ex- 

 posure to weather that it could scarcely be called brown. 



This experiment must, I think, be considered to show 

 conclusively that there is no relation between the colour 

 of the cocoons of S. carpini and that of the substances 

 to which they are attached. We need not therefore, in 

 this case, consider the difficult problem whether, if such 

 a relation did exist, it might or might not be properly 

 considered a protective device. 



(2). As to the origin of the colouring substance, I 

 have satisfied myself that it is obtained from the con- 

 tents of the alimentary canal. This conclusion is made 

 for the following reasons : — 



{a). The white cocoons are thin and papery, while the 

 dark cocoons are stiff and very shiny, on the inside 

 especially, looking as if they had been stiffened with 

 brown size. 



(b). In the case of some brown cocoons spun against 

 white paper, there was a brown stain on the paper, as 

 though a brown fluid had oozed through. 



(c). In the case of a majority of larvfe, which, in 1891, 

 spun white cocoons, there was evidence to show that an 

 evacuation of the contents of the alimentary canal had 

 taken place. 



{d). This evacuation is, when still wet, of a reddish 

 brown colour, of a viscous consistency, and contains 

 small pieces of chewed leaves, and sometimes half- 

 formed faeces. 



TRANS. ENT. SOC. LOND. 1892. — PART III. (NOV.) R 



