244 STUDIES IN AUSTRALIAN NEUROPTKRA, iii., 



with a central "disc" and a broad rim all round it; in both, So, 

 K and Rs unite before the tip of the wing, and the disc is closed 

 distally by a gradate series of cross-veins. All these characters 

 are almost certainly the result of convergence. The only archaic 

 character common to the two groups is the basal ly-placed origin 

 of Rs in both wings. It seems to me that the Apochnjsidoi lie 

 far off the main line of descent of the Chrysopidoi, and that it is 

 useless to attempt to derive the latter from Apochrysid-like 

 ancestors. 



The stages by which the original plan of a true media and 

 cubitus, with branches of Rs running above them, and all more 

 or less parallel (the archaic Osmylid pattern), gave place to the 

 Chrysopid pattern, are quite easily conceived, when it is remem- 

 bered that the pseudo-media and pseudo-cubitus are, in reality, 

 for all except the most basal portions of their lengths, true 

 Banksian lines. It will not be necessary to repeat the diagrams 

 given in No.l of this series of Studies {I.e., p. 748, Text fig. 10), 

 in which the formation of a Banksian Line in Myrmthontidcc 

 was contrasted with the formation of a Gradate Series in Chry- 

 sopid(P, and both were developed from the archaic Osmylid forma- 

 tion. It is only necessary to understand that what we have 

 discovered in this paper is the fact that only the distal branches 

 of Rs, in Chrysopa, develop gradate cross-veins, whereas the 

 proximal branches of Rs develop Banksian Lines (M' and (Ju') 

 and thus behave in the same manner as tlie distal branches of 

 Rs in the Myrmeleontidm. A combination, then, of the two 

 series of diagrams, for the two portions of tlie Chrysopid wing, 

 will bring about the required result. In this connection, we 

 might also study with profit the irregular series of cross-veins 

 between Rs and M' in Oliyochrysa. We must also, of course, 

 allow that, from the very beginning, the true media and cubitus 

 became encroached upon by the branches of Rs, until they finally 

 underwent sucli reduction and displacement as we now see in 

 Chrysopa. 



We may postulate, then, for the Chryso2nd(e, an ancesti-al 

 wing-form (Text-fig.7) in which the radial sector arose close to 

 the base in both wings, and ran evenly parallel to and under the 



