35 



what structures in Panorpa they represent, we find nothing like 

 them in that genus ; and if we say then they are no doubt the 

 bristles modified, we look again at the Micropteryx wing to see 

 if any traces of bristles exist \ we at once find along the margin of 

 the wing amongst the cilia a certain number of unmistakable 

 bristles typical in every respect, and placed so that they are not in 

 range at all with the scales, but evidently another structure alto- 

 gether. On looking at the nervures we find also a series of bristles, 

 placed at intervals solitarily and not in line with the rows of scales, 

 but following a separate order of distribution quite independent of 

 that of the scales. A close examination of them shows that they are 

 bristles in all respects but one, and in that they are slightly modified 

 in the direction of scales, that is, in their attachment to the wing. 

 They do not possess the ordinary scale-cups such as scales possess, 

 and their attachments look extremely different to that of the sur- 

 rounding scales ; still the bristles are slightly contracted at their 

 bases, and penetrate a little below the surface into their circle of 

 attachment, which has not, however, the hour-glass constriction of a 

 typical scale-cup. On returning to the marginal bristles these do 

 not present quite the same modification, but looking at them with 

 one's suspicions aroused in this way, they do present a slight modi- 

 fication of the ordinary bristle attachment. Such bristles do not 

 occur anywhere between the veins on the general wing surface ; 

 towards the base of the costa of the hind wing they become larger 

 and more evident, and culminate in the little row of bristles that 

 represents the frenulum. Hair-scales that occur freely towards the 

 wing bases are seen to be true scales, by their pigmentation, by their 

 cylindrical form, their length, and especially by their attachment to 

 typical scale-cups. 



Seeing, then, that we find these bristles persisting along the ner- 

 vures, along which they have a remote ancestry in the earlier insects, 

 and that on the wing membrane, where hairs have been less frequent, 

 we find a new structure, the scales, it seems more probable that 

 scales are a new development from the hypodermic cells, and that 

 the bristles have resisted (instead of originating) the new movement, 

 except in having been compelled to accept some trace of the scale- 

 cup structure. 



When we examine other Lepidoptera, we do not always find our- 

 selves able to say so easily as in Micropteryx what is a veritable 

 scale, what is a modified scale, and what is a slightly modified bristle, 

 since the variations that occur are so numerous and various. 



What we may actually observe, however, is, in the first place, 

 that almost everywhere we meet with bristles scattered along the 

 ciliary margins, and in many of the highest Lepidoptera these are 

 as distinct from any form of scales and have their attachments 

 as little modified as in Micropteryx. I have observed these in 

 almost all families I have examined, and when in some of the 

 smaller Incompletcc (micros) I have failed to observe them, I have 



