450 Mr. 'Pouliou's further experiments upon 



and " light brown ") ; the other spun on to muslin and 

 paper was " light brown," lighter than the darkest on 

 the muslin top and sides of the cylinder with white 

 paper spills, but darker than the " very light brown " 

 ones on the roof of the same cylinder. It was in fact 

 intermediate between these two tints. Both were strik- 

 ingly different from the white one in the corresponding 

 cylinder exposed to light. 



Hence darkness produced brown and light brown 

 cocoons, under conditions which, with light, produced a 

 white one. Experiments upon other species did not 

 render it probable that darkness would produce very 

 dark cocoons, but that it should produce any effect at 

 all is inexplicable, except on the supposition of a colour 

 which can be modified by the larva as a response to 

 external conditions, and this in a normal manner, and 

 not as the result of disturbance. 



I was very anxious to apply the crucial test which 

 suggests itself after reading Mr. Bateson's criticisms. 

 He contends that the dark cocoons are normal, and are 

 always formed in nature by healthy larvae, and that the 

 light ones are produced by disturbance or the presence 

 of parasites. 



If, therefore, a much disturbed larva spun a dark 

 cocoon, or, better still, a larva, which had begun to spin 

 a white cocoon on a white surface, afterwards sjDun a dark 

 one in contact with the appropriate surroundings, it 

 would be quite impossible for Mr. Bateson's criticism to 

 be sustained. 



A larva beaten from birch spun a considerable part of 

 a perfectly white cocoon in a white chip box. A birch- 

 leaf was also spun into the cocoon, but in this case the 

 white surroundings predominated. Mr. Sidgwick has 

 also obtained a white cocoon of this species in a chip box 

 during the present year. The cocoon was opened and 

 the larva removed, and it was then found that two eggs 

 of ichneumons, probably of the genus Paniscus, were 

 fixed to one or more of the thoracic segments. In trying 

 to remove one of these with scissors, the larva was 

 rather seriously cut, and bled freely. I therefore desisted 

 from the attempt, and placed the larva in a cylinder with 

 oak shoots bearing leaves. To my surprise, it spun on 

 the glass, including the edge of an oak-leaf in its cocoon, 

 but ccery thread of silk spun after its removal was bruicn, 



