of cocoons o/'E. lanestris and S. carpini. 61 



large vessel in which the Safurnia larvae were kept 

 without disturbance. The lluid itself is viscous, and of 

 a dark coffee-brown colour, closely resembhng that of 

 the cocoons. It generally contained some foecal matter 

 and particles of semi-digested food. From this, there- 

 fore, it may be concluded that the fluid is voided from the 

 intestines, but I never saw a larva in the act of evacuating 

 it. If this should be found to be the origin of the fluid, 

 it may probably be looked on as being of the nature of 

 " meconium." 



The presumption that it is with this fluid that the 

 cocoons are coloured rests on the following observations : 

 A considerable number of larvas, which were known to 

 have voided the brown fluid, spun white cocoons. On 

 the other hand, many spun white cocoons which were 

 not known to have voided any fluid, though nevertheless 

 they may have done so. Next, it was observed that 

 some of the dark Saturnia cocoons, after they were just 

 finished, were wet, as though drenched with brown fluid. 

 Several also of the pale Saturnia cocoons had a darker 

 patch in one part, generaUy upon the neck of the cocoon, 

 though in one case there was a dark patch on the side. 

 The appearance of these patches was exactly as if a 

 quantity of brown fluid had been ejected upon the inside 

 of the cocoon. In one case a brown cocoon of Saturnia, 

 which was spun against a piece of white paper, lay on a 

 large stain of the brown fluid ; and there could be little 

 doubt that the Huid had soaked through the cocoon on 

 to the paper. 



There is, then, good evidence that a brown meconial 

 fluid is voided by caterpillars which are removed and 

 shut up before they spin, and if it were to be established 

 that the colouring matter of the cocoons is due, or 

 largely due, to this fluid, the phenomenon of the 

 colour-variation of cocoons becomes much simpler; for 

 the cocoons of secluded larva3 are, on this hypothesis, 

 white by reason of the previous voiding of the brown 

 fluid, and the consequent absence of a supply of colouring 

 matter. 



It should be mentioned, as making against this view, 

 that in the case of three Eringaster larva), which were 

 disturbed whilst spinning, and which afterwards spun 

 white cocoons, it was almost certain that no brown fluid 

 was previously voided. It is, of course, possible that the 



E 2 



