The Inheritance of Colour in Carausius morosus. 115 



last in April, 1915. In all, over 3000 insects were reared. About half the 

 families were reserved for experiments on the effect of reduced light and 

 darkness (v. infra) ; the remaining 1676 insects were reared under normal 

 conditions, and of these 1612 were green and 62 brown, whilst 2 were 

 classified as yellow. Leaving these two out of account, the proportion of 

 brown to green is not quite 4 per cent, or 1 in 25. If we examine the case 

 of individual insects this conclusion is amply borne out. Of 91 young of a 

 green insect whose mother was brown 89 were green and 2 brown. Of 

 105 offspring of a brown insect' whose mother was green 103 were green 

 and 2 brown. It must be remarked that the mortality in these later 

 experiments was very much reduced. Whereas in the 1913 experiments 

 from 2056 eggs 174 insects were reared, in 1914 and 1915 from 4746 eggs 

 1676 insects were raised. The proportion of browns amongst adults was 

 in the first case 15"5 per cent., in the second case 3'6 per cent. As the 

 mortality is reduced therefore the proportion of browns diminishes. 



It is obvious that we have to do with a type of segregation of varieties, but 

 a type which has not hitherto been described. In the first place the insects are 

 almost certainly the offspring of three or four generations of parthenogenetic 

 development. If the process of segregation were the result of an originally 

 hybrid constitution of the females which laid the eggs which we received 

 from India, then, even if an initial segregation had taken place, since the 

 only reproduction is parthenogenetic this segregation ought to cease, after 

 the first generation, since no crossing amongst the Fi generation takes place. 

 Then the proportions of brown to green do not bear any relation to any 

 Mendelian ratio, and a brown parent produces just the same proportion of 

 browns to greens as a green parent. At the same time, whilst the factor for 

 brown is present in all insects, the factor for green is definitely eliminated in 

 a certain proportion as growth proceeds. 



It .is an obvious reflection that possibly the production of a brown rather 

 than a green adult may be clue to a colour reaction — those larvee, it might be 

 supposed, which pass a critical period of their growth in a shaded environ- 

 ment turn brown, whilst the rest exposed to full illumination turn green. 



In order to test this^ hypothesis the experiment was made of taking the 

 eggs laid by a certain number of insects and dividing the progeny of each 

 into two equal lots, one lot being reared as in other experiments in glass 

 jars closed at the top by muslin, the other in glass jars of equal 

 size closed at the top by two layers of muslin and pasted round the 

 sides by white paper so that only light of reduced intensity could reach 

 them. The two lots were compared in order to see if a greater proportion of 

 browns were produced by the insect which were exposed to diffused light. 



