38 



casianece there are two circles, with little more than ordinary 

 variability as to position. 



Their size is by no means proportional to that of the insect or 

 of the vein they are on. They are, for instance, small in the 

 Sphingidcv. and Saturnida. The longest, I think, occur in a species 

 of Cryptophaga i^XyloryctincE). In some P\ rales they are also large, 

 as in .S'. pruualis. 



I found also that in at least a majority of species similar circles 

 existed on veins in the area of the wing, most frequently near the 

 transverse vein and on branches of the radial vein, and occasion- 

 ally on the surface of the wings, where, or close to where, veins 

 had disappeared or dwindled. When they so occur they have 

 a tolerably constant position in the same species ; thus in 

 H. prasinana there are on the fore-wing three on the radial vein 

 before its first branch, one beyond this, and two on the combined 

 stem of 7 — 8. On the transverse vein one between three and four, 

 and one between four and five, and one on the base of four. 



In Vanessa tirticce they are numerous along a number of veins. 

 They are less abundant, but still frequent in other Vanessids 

 that I have examined. 



When once the wing is mounted transparently, it is practically 

 impossible, in most instances, to say which side is uppermost, and 

 in many slender wings it is equally difficult to distinguish one 

 side from the other when under the microscope. W^ith a very 

 wide margin, therefore, for error, I regard them as being always 

 on the under surface of the wing, since that has been their position 

 whenever I have been quite certain as to which side of a wing 

 I was looking at. 



I have already alluded to their structure; they seem to exist in 

 the thickened wall of the vein, as distinct from the included trachece 

 or the superficial hypodermis. 



I have not met with them at all in the few Trichoptera and 

 Neuroptera I have examined. Though in some He?nerobiidie, there 

 are some circles that I am not absolutely positive are the sites 

 of bristles that have been rubbed off, but I believe they are. Nor 

 have I seen them in any Diptera I have examined. 



On the question as to what these circles really are, what they 

 represent, or what are their functions, I am up to the present time 

 unable to make more than a suggestion. 



That they have some useful function seems highly probable from 

 the simple fact that they occur throughout the whole order, since 

 there can be little doubt but that, if useless, they would have 

 been eliminated, at least in the higher families, even if they have, as 

 I have not ascertained, a long genealogy somewhere amongst the 

 more ancient insects. 



The only guess I can make is that they are perhaps affected, as 

 diaphragms, by variations of pressure, and so by communicating 

 such variations of pressure along the fluid or air in the veins, give 



