146 BR. K- J- TILLYAED ON THE 
It will be noticed that no less than five branches of Il2a are indicated in 
the new notation. An explanation o£ this will be fonnd in the argument on 
page 148 concerning the nature of triads. 
The results obtained by the three methods of study, already mentioned, 
support one another, except that in the case of the study of the larval 
tracheation a certain amount of variation is normally present. I have there- 
fore presented the argument based on the Convexity and Concavity of the 
veins first, followed by that deduced from the study of these conditions as 
noted in the fossil Protereisma, and have kept for the last place the results 
obtained from the larval trachese. In dealing with these last, the alternative 
interpretations given in the table on p. 145 will be considered, and reasons 
given for rejecting them. 
Text-fig. lb. 
Fore-winj? of Ameletus ornatus (Eaton), Recent, for comparison with text-fig. 1 a. in, tornus. 
Lettering as on p. 162. 
Convex and Concave Wings. 
In all generalised insects, we are able to distinguish the presence of two 
kinds of veins on the wings, viz. those which occupy tlie summits of ridges, 
commonly called convex veins (indicated by a plus sign), and those which lie 
in the bottoms of grooves or hollows, commonly called concave weMi.s (indicated 
by a minus sign). In the ideal archetypic wing, convex and .concave veins 
follow one another alternately across the main portion of the wing, 
So being — , Rj and its sector -h, M — , the vein commonly called Cuj +, 
and Cu2 — . This last is alwaj^s to be distinguished by lying either in or 
very closely anterior to the deep anal furroio, which separates off the davits 
or anal area from the rest of the wing. The clavus itself is a loholly convex 
area, and carries only convex veins, viz. the three anal veins 1A,2A, and 3 A. 
In an ideal forked vein, both branches of the fork keep the same condition 
of convexity or concavity as the main stem ; but the allei-nation of ridge and 
