37 



There is a curious structure that I have not been able to find any 

 references to in the numerous papers and authorities that I have 

 looked up, and so provisionally call attention to it as something 

 new, but with a strong suspicion that some reference to it exists 

 somewhere, and that my claim to novelty may be no better than 

 Mr. Cooley's in the case of the " holding area." 



If a wing be decolorised or denuded of scales, and the wing 

 margin examined as a transparent object, it will be found that where 

 each vein reaches the margin, two circles occur on its surface. 

 Sometimes these circles are within the wing margin that forms the 

 seat of the cilia, sometimes they are just clear of this, and in a few 

 instances they are distant from the wing margin several times the 

 width of the ciliary base. 



They are almost always, one nearer the base than the other, that 

 is, a line joining them is substantially the line of the vein to which 

 they belong, and the further they are from the wing margin the 

 further they are, as a rule, from each other. 



On first observing these circles they appeared to me to be circular 

 openings into the lumen of the vein, bounded by raised and thick- 

 ened edges. I have, however, concluded that the supposed opening 

 does not exist, chiefly because I can detect no passage of air or 

 fluid through them, under any experiment, in emptying the veins of 

 air or filling them with other fluids. 



The circle is, however, just like the thickened rounded margin 

 of an aperture, and in some cases very obviously is at the svmimit 

 of a conical elevation, with slightly wrinkled slopes. I have been 

 able also, in stripping off the wing membrane on the side on which 

 they are, to leave them intact beneath it, and the stripped-off mem- 

 brane shows no corresponding structure or aperture. 



This process I have only attempted in carpini and airopos, but 

 in these two cases the result appears to be as I have stated. 



These circles occur in every Lepidopteron I have examined, the 

 exceptions to there being two at the extremity of each vein, are very 

 few indeed. 



I am not quite sure, but I think I have seen one only in some 

 exceptional instances (as Megalopyge crispata), and once or twice I 

 have met with three, purely as a matter of local variation on an 

 odd vein. More frequently they appear to be wanting altogether in 

 the anal veins, and rarely on vein two or vein three. 



In Zeuzera {pyrina), however, the normal number appears to be 

 four, and in Cossus {ligniperda) five, and in these they are 

 decidedly further from the margin than usual. 



In another species of Cossus {Diiomutiis leuconohis) there are 

 only two of these circles, but their position is very variable, close 

 together, wide apart, nearer or further from the margin, transversely 

 or longitudinally placed. One might almost take them to be any 

 two taken haphazard of the five that occur in /igniperda, since these 

 variations all occur in one specimen. In the case of Macrogaster 



