of cocoons of 'El. lanosti'is and S. earpini. 19 



paper. The nature of the distinction between l)rown 

 and white cocoons is discussed below. 



V. From these experiments it appeared that light- 

 coloured cocoons were produced when the larvae were 

 confined in white substances, and also when they were 

 confined in dark substances, but that when left with 

 their food the cocoons were dark. This result suggested 

 that perhaps the alteration of colour was brought about 

 by some unhealthy condition associated with the removal 

 of the larva3 from their food. The four larviB which had 

 been disturbed whilst spinning also produced white 

 cocoons, though one of them was attached to a dark 

 object. From this it seemed likely that disturbance at 

 the time of spinning might also be sufficient to prevent 

 the cocoon from being properly coloured. It became 

 therefore necessar}^ to see what coloured cocoons would 

 be spun by larvae which of their own free will si>un upon 

 white paper. With this object the vessel in which the 

 remaining larvae were feeding was carefully tilled with 

 crumpled white paper, so that each twig of food (haw- 

 thorn) was more or less surrounded with paper. All the 

 iarva3 in this vessel chose to spin in the paper, and 15 

 cocoons were thus obtained. Of these 15 cocoons on 

 white paper — 



4 were of full colour. 



6 were lighter than tliis, but still substantially 



brown. 

 4 were light, 3 of them being white. 



Of the 3 which were white, one was spun by one of four 

 larvfe which remained at the last, and were not fed 

 owing to a mistake. 



To recapitulate: — Of 11 larvoe left with their food, all 

 spun dark cocoons on leaves. 



Of 14 larva) left with their food and white paper, 10 

 spun dark cocoons on white paper, and 4 spun light 

 cocoons on white paper. 



Of 11 larva3 which were shut up, all spun light cocoons, 

 5 being on white substances, and G being on dark sub- 

 stances. 



Sdtnrnid rarpini. — Eleven cocoons found spun in the 

 hedges in a state of nature wore all of full colour. 



Experiments made with larvte of this species agreed 

 generally with the resulls from those made on E. hincsfris, 



TRANS. ENT. SOC. LOND. 1892. PART I. (MARCH.) E 



