the colours of certain Lepidoptem. 337 



an experiment was attempted as a comparison with the 

 above, but the 21 larvae had all died by Sept. 4. 



The chief results of these experiments, as regards 

 green and dark surroundings, have been briefly men- 

 tioned in ' Colours of Animals,' pp. 15*2, 153, where a 

 dark and light form are represented by uncoloured 

 illustrations. 



These experiments at once proved that A. hetnlaria 

 was b}' far the most suitable species for the purposes of 

 this investigation ; and I tried to obtain eggs in the 

 succeeding years. In this I was unsuccessful until the 

 present year, in which a much larger series of more 

 careful experiments have been conducted. 



The pupil; obtained from these experiments were care- 

 fully separated, and the attempt was made to breed from 

 the imagos which emerged. The great majority, however, 

 died in the pupal state, and those which emerged did not 

 pair. 



As the moths are well known to vary in darkness, 

 I noted the colours of those few which emerged, but 

 found that there was no relation between the larva and 

 imago in this respect. 



1892. 



This is by far the most extensive series of experiments 

 upon the modification of the colours of larvic by the en- 

 vironment that I have ever undertaken. The results 

 obtained in 1889 not only proved that this is the most 

 sensitive larva as yet subjected to experiment, but also 

 that it is most satisfactory to breed, and in every way 

 the most suitable for the purposes of this investigation. 

 I was therefore very anxious to repeat the experiment on 

 a larger scale, and especially to test again the effects of 

 white surroundings, which had produced such remark- 

 able results on the previous occasion ; also to make use 

 of other artificial colours, as well as the natural tints 

 of twigs of various kinds and conditions. 



A captured female laid a very large number of eggs, of 

 which probably about 200 were sent to Mr. Bateson ; 

 these unfortunately hatched during his absence from 

 home, and the larvae died. The remaining eggs began 

 to hatch in large numbers on June 29th, and all the 



