170 THE entomologist's record. 



general humidity coupled with an area in which large quantities 

 of smoke are being produced, e.g., Leucania impura and L. 

 pallens produce much darker specimens in the London marshes 

 than in the open fields near Strood, in Kent, and Scotch spe- 

 cimens are occasionally darker still. Again, Agrotis Jiigricans 

 is blacker from the Greenwich marshes than from the fields and 

 marshes in the neighbourhood of Rochester, where they are 

 generally much marbled with yellow, and the same (Greenwich) 

 marshes give a much larger percentage of dark specimens than 

 any locality I know of a drier character. Apamea fibrosa has 

 a larger percentage of pale variegated forms in its more open 

 localities. Many examples of this kind could easily be adduced 

 in the same direction. But it must be remembered that the 

 Fen District of England is probably the driest part of the 

 British Isles, so far as actual rainfall is concerned, and this 

 must be carefully borne in mind when considering the species 

 obtained in this district, or those of a similar environment. 



Another query based on the same line of argument is occa- 

 sionally heard. " Why do we get white butterflies and black 

 moths on the same geological strata ? " Such a question always 

 appears to me to show entire ignorance of the subject of 

 "natural selection." The colours of butterflies assimilate 

 generally to the flowers or other objects on which they rest. 

 Their wings are drawn up over their bodies when at rest, and 

 we only see the undersides. It is, therefore, to the undersides 

 we must especially look for variation, and here we chiefly find 

 it. The variation of the undersides of our most common 

 butterflies is a most interesting study, and opens up a wide 

 field to those who have never looked for variation, where, in 

 this group, variation must show itself. Of course the variation 

 of butterflies on the underside is often reproduced in a mild 

 form on the upper side, but from a scientific point of view, the 

 variation of the upper sides is practically valueless, compared 

 with the variation of their undersides. Incidental cases of 

 variation occasionally occur on the upper side of butterflies, 

 but as a general rule such varieties are, in my opinion, consti- 

 tutional, and each must be considered on its merits as to its 

 probable mode of causation. 



These undersides, serving as they do for a protection to the 

 insect, must be studied then, if we are to find the changes 

 which " natural selection " plays in this group. As an example 

 of the influence of environment on a butterfly, I will instance 

 Pieris napi. In our southern districts of England it is generally 



