AND OF PRODUCTS OF METABOLISM. 291 



lation between their colour and that of their food 

 seems to show that the one is directly dependent 

 on the other. Meldola * accounted for it by sup- 

 posing that the larvae had been rendered transparent 

 by Natural Selection, whereby the colour of the vege- 

 table food eaten was itself enabled to give the colour to 

 the larvae. Poulton f has shown that the colours of the 

 larvae are due partly to the pigments proper to the 

 larva, and partly to the pigments derived from the food 

 plants. These pigments undergo some modification in 

 the tissues, but Poulton states that as far as he has in- 

 vestigated the subject " all green colouration without 

 exception is due to chlorophyll; while nearly all yellows 

 are due to xanthophyll." The chlorophyll, or some 

 modification of it, tinges the blood of the larvse, the 

 green colour of which is often due to this cause alone. 



From these observations, therefore, it follows that a 

 change of food may also effect a change of colouration. 

 That this is so is strikingly shown by some other obser- 

 vations by Poulton. J Obtaining a large number of 

 larvae of Tryphcena pronuba (Common yellow under- 

 wing) from the same batch of eggs, he split them up 

 into three groups. One he fed on the white midribs of 

 the cabbage, from which the yellow blade had been 

 carefully removed with scissors. These larvae remained 

 almost white at first, and afterwards showed a moderate 

 amount of black pigmentation. The other two groups 

 of larvae he fed respectively on the yellow etiolated 

 leaves from the heart of the cabbage, and upon the deep 



* Proc. Zool. Soc. 1873, p. 155. 



f Proc. Roy. Soc., xxxviii. p. 269, 1885. 



tProc. Roy. Soc., liv. p. 417, 1893. 



