the colours of certain Lepidoptera. 449 



larvfe of one cylinder were more disturbed than those of 

 the other. 



Wliite surround iufis. — 3 cocoons were spun upon the 

 muslin roof, which in this case was a considerable dis- 

 tance from the oak-leaves ; 1 was light brown ; 2 very 

 light brown and almost white. 3 were spun on the glass 

 side near paper spills and oak-leaves. Of these 1 was 

 rather dark brown and two were light brown. 



Dark surroundinfis. — 1 was spun on the muslin top, 

 in this case near the oak-leaves but away from the 

 sticks ; it was intermediate between the tints last de- 

 scribed. 3 were spun on the undersides of oak-leaves in 

 close proximity to the dark sticks, and all these were 

 dark brown, far darker than any in the other cylinder. 

 One of them was built low down, and some of the white 

 cotton wool wrapped round the oak twig, where it passed 

 through the hole in the i^late. was spun into one end of 

 it, while the ends of 3 dark twigs were also fixed to it. 

 As in many other cases among larvae and pupae, where 

 dark and light surroundings contend, the former proved 

 to be more powerful, for the cocoon is as dark as the 

 other two. Another 5th cocoon was fixed to the glass 

 cylinder near the leaves, and closely surrounded by dark 

 twigs. It is very dark brown, just like those upon leaves. 

 Mr. Sidgwick tells me that he has always obtained such 

 dark cocoons upon leaves. 



Another kind of experiment was then begun. A very 

 small twig of oak with few leaves was placed in each 

 of two similar cylinders, which were then filled up with 

 fragments of white tissue paper, slightly crumpled. 

 One of these was then covered with a larger darkened 

 cylinder so as to be in nearly complete darkness ; the 

 other was left in strong light. 



In the latter a single cocoon was spun on the white 

 muslin roof far above the oak-leaves, but closely sur- 

 rounded by white paper, the corner of one piece being 

 spun into it. It was the whitest cocoon obtained, with 

 hardly a trace of brown. In the other darkened cylinder 

 2 cocoons were obtained, both spun, 4 centimetres apart, 

 on the muslin roof, the white paper close to them, and 

 abundantly spun into one, the leaves far away. The 

 cocoon on the muslin alone was a slightly deeper shade 

 than that on the muslin roof of the cylinder with dark 

 sticks {riz., intermediate between " rather dark brown " 



