10 THE ENTOMOLOGIST S RECORD. 



crawled on to the bag and began to spin a network of silk on the linen 

 of which it was composed, before attempting to build the sides of its 

 little boat-shaped cocoon, and I noticed that this basement fabric (if I 

 may call it so) appeared to be invariably white. That is to say, the 

 larva nrturally spun white silk. It was suggestive, therefore, that the 

 larva had the inherent power of colouring the silk or not, according as 

 it spun a dark or light cocoon, and that the normally coloured cocoon, 

 as found in nature, was really to some extent coloured, and not simply 

 formed of the silk as originally made by the larva. The question 

 therefore arose, how was the white silk made of a different colour to 

 white, assuming of course that white was the natural and original colour, 

 unmodified by anv colouring matter whatever? Certain it was that the 

 larva did in my three bags have the power of colouring this white silk 

 within the limits I have previously suggested and pointed out. How 

 was this effected ? 



I now took a considerable number of what appeared to me to be full- 

 fed larvae from each bag, and distributed them equally, about 20 in each, 

 in three large new chip boxes, and here my troubles began. Tho'>e from 

 the dark bag (i) produced a great number of fairly dark cocoons not 

 corresponding at all to the colour of the inside of the box, although a 

 comparatively small number did so precisely. Those from the dirty 

 bag (3) were a2;ain most normal, and those from the white bag (2) spun 

 some rather pale cocoons which did not correspond at all to my idea 

 of adapting themselves to their surroundings. There were also two 

 purely white cocoons among those from the dark bag, and one in that 

 from the dirty bag. A few larvse in the various boxes showed that I 

 was also deceived occasionally in my ideas as to their being full-fed, and 

 I was afraid that the removal of some of those larvae from their 

 i^urroundings before they had made up their minds to spin was the cause 

 of the failure, and that environment had no influence unless the larvae 

 were in an active state of making or being about to make their cocoons. 



As some of these larvae had been removed when they were not full-fed, 

 I thought, as I have just mentioned, that the extent of their maturity or 

 otherwise might influence matters. Accordingly I made another 

 experiment, and was most careful to select none but those which had 

 positively commenced to spin. Here my success was most complete, 

 but in a way I had nevei anticipated. The larvae appeared utterly 

 unconscious of their new environment (in chip boxes). Those from bag 

 I spun a fairly dark cocoon as they assumedly would have done in the 

 dark bag, those from bag 3 were normal, and those from bag 2 were also 

 normal, certainly not so pale as I had expected from the behaviour of 

 the lot from bag i. Here then I had advanced a step. The colour 

 of the cocoon was already determined when they were ready to spin 

 and was only modified infinitesimally afterwards. Hence the failure of 

 my previous experiment was fully explicable by assuming that some of 

 the larviE were, and others were not ready to spin. 



I now took larvae not full-fed from the dark bag and fed them up in a 

 chip box and got either pale or normal cocoons, showing at any rate 

 that living in the dark bag throughout their existence, except the last 

 few days, did not influence them at all \ similar results were obtained 

 from the larvae in the white bag. 



I was now quite satisfied that the larva could respond to its environment 



