21 



same way for all. As Prof. Poulton truly says, " the comparison 

 of these details is almost a demonstration of the operation of Natural 



Selection we cannot conceive of any theory not 



dependent upon the principle of selection, which could produce 

 such extraordinary superficial resemblances among numbers of 

 species by methods which are entirely unlike in all their details." 

 In 1905 I made a similar examination of the British Guiana 

 transparent groups consisting of such families as Syntomidfe, 

 Hypsidae, Geometridae, Erycinidae, and the subfamily Ithomiinfe, 

 and published the details with a plate in the " Entomologist's 

 Eecord " for that year. The differences in the methods of obtaining 

 transparency were just as varied as those obtained by Prof. 

 Poulton, and some were even different from those species and families 

 he examined, showing that natural selection could seize on any one 

 or more tendencies to become transparent. The cases of those 

 species, such as the Erycinid butterfly Stalachtis pJurdnsa, that have 

 two sets of scales, each of which has undergone an entirely different 

 modification, being particularly instructive. In this particular 

 species normally there are alternately narrow and broad scales, the 

 long narrow scales being set up on edge slightly, which gives the 

 insect under the microscope the appearance of having two layers of 

 scales. In the transparent portions the broad scales are transparent 

 and the narrow scales have become hair-like. In addition, the trans- 

 parent scales wear off. Truly a remarkable combination of processes 

 to obtain the one end of transparency. But, if objection is taken to 

 mnnicry between insects of the same ot'der, there are hundreds of 

 cases of close resemblance between members of different orders. 

 Moths may be like beetles, wasps, bugs or flies, and in nearly all such 

 cases there is an astonishing alteration in habits, altitudes and move- 

 ments to fit in with the superficial resemblance. If a Syntomid 

 moth, Correhia lycoidea, is like a Lj^cid beetle by accident, we should 

 not expect to find that the movements of the moth were abnormal 

 and like those of the beetle. We should also not expect to find that 

 it had a flattened body, which also is abnormal in moths of this 

 family, but which makes it very like in appearance to a beetle's 

 abdomen. Again, if ^lacroneine ladca, the Syntomid moth, is like a 

 fossorial wasp of the genus Salins by accident, one would not expect 

 to find that it held out its black hind legs in flight in exactlj^ the 

 saine way as the wasp does, or that it vibrated its wings when it 

 alighted on a leaf as the wasp does. If there is real resemblance it 

 must be advantageous for the moth to be disguised, or it 

 could never have come about in the precise way alluded to. How 



