tJ(e colours of certain Lepidoptera. 357 



pigment, and so reveal the more deeply placed green, lie 

 between a wave-length of jouSoo and touuoo nam. ; in 

 other words, about the D line of the solar spectrum. 



It will be of the greatest value to now test these con- 

 clusions by the use of coloured glass or gelatine screens. 

 White spills or painted sticks might be employed in a 

 large number of experiments with screens of various 

 colours. This method has been already tried to some 

 extent with the pupae (see Vanessa io and the Pieridce), 

 but the larvffi of A . hetidarui are far more suited for the 

 investigation, being so highly sensitive, and possessing 

 such a wide range of possible coloui's and combinations 

 of colours. 



The fact that each of these artificial colours produces 

 nothing peculiar, but only some one out of the well-known 

 api^earances which are liable to occur in the surroundings, 

 is strongly in favour of the essentially protective signi- 

 ficance of the change, which is thus only possible when 

 it leads to harmony with some natural environment. 

 The same fact holds universally throughout the species 

 which have been proved to be susceptible, unless an 

 exception is to be made in favour of the golden puj)8e of 

 Vanessa urticce. These, however, are discussed in a 

 later part of the paper (see Conclusions). 



The Structural Cause of the varied Colours of 

 THE Larv^ of Amphidasis betularia. 



This was partially investigated in 1889 (see p. 336), 

 and was proved to be due to colour in the skin or just 

 below it. In the present year the following method was 

 adopted, and found to work well. The larva was stretched 

 with its ventral line uppermost across a glass slide 

 covering a window cut in a sheet of cork. The anterior 

 and posterior ends of the larva extended beyond the 

 glass, and were pinned to the cork. The body walls 

 were then divided along the median ventral line and 

 pinned out flat at each end, so that the section of the 

 body passing across the glass was flat also. The latter 

 part could be examined from above or below with the 

 lens or a compound microscope, and the effect of 

 removing any coloured layer was at once seen. 



In such stretched and flattened larvae the loss of the 

 green blood made the colour rather less deep, and the 

 same effect followed the removal of a section of the 



