BY U. J. TILLYARD. 739 



because of the veiy typical formation of the brandies of M, 

 which are, as far as they are preserved, practically identical, in 

 number, position, and amount of divergence from their common 

 stalk, with those of T. pallida. The branches of the radius also 

 agree in the two forms; and it should be noticed that the differ- 

 ence in level between the bifurcation of R4+5 and R2+3 is greater 

 in these two wings than in the hindwing. We are justified, 

 therefore, in concluding that P. mitchelli represents a forewing, 

 and quite likely the forewing of the species of which P. aiislraHca 

 is the hindwing; though, as this cannot be proved, I think we 

 should retain the two names for the two separate specimens. 

 As, however, the venation of different species of the genus 

 Permochorista probably did not vary more than it does at the 

 present day in species of Tcuniochorista and Chorista^ I have 

 shown the restoration given in Text-fig. 6 as that of a typical 

 pair of ivingH for the genus Permochorista^ the forewing being 

 restored from P. mitchelli^ the hindwing from P. australica. A 

 comparison of this restoration with the wings of 7\ pallida will 

 show us the small extent of the evolutionary process which has 

 occurred in this very stable wing-type between Permian and 

 recent times. It may be briefly summarised as follows : — 



(1) Reduction in the number of branches of M, from six to 

 four in the hindwing, and from six* to five in the forewing. 



(2) A tendency for the cross-veins to arrange themselves more 

 advantageously as supports for the main veins, alternately above 

 and below any given main vein; and hence 



(3) The breaking of the original regular course of the main 

 veins into a very slight irregularity or zigzagging. 



(4) Greater specialisation of the basal portion of Cii. In the 

 Permian genus, Cuj has already fused with the main stem of M 

 basally, and probably Cu., with lA. In Tceniorhorista and other 



* Although only frt branches are to he seen in the part of the wing of 

 P. mitchdll preserved to us, it will be quite evident that at hast one extra 

 branch is required more distally along Mn in restoring this wing, unless 

 we are going to allow a much greater divergence in the direction of these 

 branches of M than is warranted by our knowledge of Mecopterous vena- 

 tion. 



