of cocoons o/E. lanestris and S. carpini. 47 



of food at least as much on the sense of smell as on that 

 of sight. 



It may be remarked in passing that there is abundant 

 evidence that the larvae of these insects are infested by 

 Tachina, and by hymenopterous parasites, and, as in 

 other cases, probably these are really their most for- 

 midable enemies. 



As to their enemies in the pupal state, there is no 

 evidence. In the absence of such evidence it may be 

 contended that any disquisition on the modes by which 

 they may be protected from hypothetical enemies is 

 premature. This, however, is a line of argument of 

 which Mr. Poulton and the apologists of Adaptation 

 are well aware, and to which they expose themselves 

 avowedly. 



The fact, however, that the colour of these cocoons 

 varies in accordance with that of adjacent substances 

 did not seem, in my judgment, to be established beyond 

 possibility of question, and it was in the belief that 

 some simple sources of error were not excluded that the 

 following experiments were undertaken. 



Eriogaster lanestris. — I. A large colony of these cater- 

 pillars were brought home, they being then about three- 

 quarters grown, and fed in a large plain glass vessel till 

 Aug. 15th. On that date the whole was examined, and 

 11 cocoons were found spun on leaves. Of these 6 were 

 of full colour — 



4 were dark, but not quite so full in colour. 

 1 was a good deal lighter, but still brown. 



The dark colour is about the tint of black coffee, and the 

 lighter specimen may be described as having the colour 

 of strong tea with some milk it. It will be convenient 

 to refer to this specimen for comparison, and its tint 

 may be spoken of as "half-colour." 



II. From this colony a number were chosen which 

 seemed to be ready to spin These were shut up in a 

 white muslin bag full of torn, crumpled strips of white 

 paper. Of these larvae several died, but live survived, 

 and all spun cocoons attached to the muslin, or to the 

 white paper, or to both. Of these live — 



3 were quite white. 



2 were very pale cream -colour. 



