100 W. H. NUKNEY ON SCALE EVOLUTION. 



of butterflies and a group of curious and evil-smelling forms 

 (Acreince) almost confined to the African continent. Bates and 

 some others have argued, and perhaps justly, that these insects, 

 being well protected from the attacks of their greatest enemies, 

 and being at the same time very frequently imitated by members 

 of other groups without this means of defence, should be placed at 

 the head of all the butterflies as possessing the highest organiza- 

 tion. Such view, however, is by no means generally accepted, 

 and the swallow-tailed species still hold their old place at the head 

 of the entire order. It is to the somewhat isolated genus Ithomia 

 to which the occurrence of the features mentioned in Mr. Ingpen's 

 paper are almost entirely restricted. I have sought long and 

 carefully for a parallel amongst the clear-winged species of all 

 groups of macro-lepidoptera, and so far have been quite unable to 

 find such. Each group — in fact, almost each species — has its own 

 peculiarities of wing-covering, and the whole forms a most interest- 

 ing and instructive study, but nowhere, even amongst Ithomia' 's 

 direct congeners, do we see the numerous variations of scale-shape 

 to which Mr. Ingpen has drawn attention. 



Commencing our comparison with the typical family of swallow- 

 tailed butterflies, the clear-winged species of which are but few, 

 we find in the curious little Leptocircus (said when at rest to 

 mimic a dragon-fly) a peculiar disposition of the wing-covering. 

 The fore-wings alone are provided with clear spaces. Over the 

 whole upper surface of these long-stalked, bifid, wedge-shaped 

 scales are sparsely arranged in rows, whilst the lower surface 

 apparently has absolutely no scales, their place being occupied by 

 long and curved finely-pointed hairs. The attachments of both to 

 the wing membrane are so nearly coincident (in some cases exactly 

 so) as to render it extremely difficult to convince oneself, even by 

 the use of a good microscope, that loth hair and scale do not 

 actually proceed from an identical point of origin on the upper 

 surface of the wing. The accompanying Figure 1 shows this 

 appearance well. 



Fig. 1. Fig. 



