54 Mr. E. B. Poulton's notes upon the colours 



by again breeding from the brown forms. If in any 

 district the green forms were especially attacked the 

 greater proportion of brown ones left would increase 

 the number of larvffi whose parents were both brown, — 

 in other words, would greatly increase the proportion of 

 brown larvte in the next generation. But this would 

 bring the brown larvae into prominence, and would 

 render it more probable that they would be especially 

 attacked, and so the disproportion would right itself. 

 According to this argument there should be a tendency 

 (apart from the effect of enemies) for the more abundant 

 form to predominate, owing to the greater chance of 

 eggs being fertilised and laid by individuals of the same 

 form. It is possible that there is such a tendency, and 

 that such a result is coming about (in this case the 

 gradual predominance of the green), but I think that 

 this is hardly likely, and there are indications that the 

 problem is very complex. There is no doubt that green 

 is the better protective colour for summer and brown for 

 autumn and winter. In the first summer brood there 

 were 30 green and 13 brown ; in the second there 

 emerged 27 from green pupae and 13 from brown (and 

 these numbers include 5 brown and 5 green, from the 

 lot of which both parents were brown, hence unduly 

 raising the average of the latter). 



There remained 34 pupse of the second brood, of which 

 15 died, while 19 will (probably all) emerge in the spring. 

 The 15 belong to the group which emerged last summer, 

 for they died just before emergence, with the colours of 

 the wings plainly visible through the pupal covering. 

 Of these 15 pupae 10 were green and 5 brown (including 



5 green and 3 brown from the lot with brown parents). 

 Of the remaining 19, which form the true winter brood, 

 12 are brown and only 7 green (including 4 green and 



6 brown from the lot with brown parents). Hence in 

 the last case only, the proportions are reversed, and there 

 are nearly twice as many brown pupae as there are green, 

 while there were more than twice as many green pupae 

 in both summer broods. I hope to make further obser- 

 vations upon this point, but there seems to be much 

 reason (from these statistics) for believing that the 

 brown forms predominate when brown is the best pro- 

 tection, and green forms when green is the best pro- 

 tection. It is noteworthy that the colours appear first 

 in the larvae (always corresponding with those of the 



