33 



The formation of the cells appears to differ in very important 

 points from anything known in the Lepidoptera of to-day. Comparing 

 the wing with those of certain archaic Lepidoptera, such as one of 

 the Hepialidae or Limacodidse, we see that in the latter the cell d 1 

 is of considerably greater length, while there is also another closed 

 cell below it. between M and Cu. It is by fusion of these two cells 

 that the wide cell of the butterfly-wing has been evolved. In 

 Dunstania, however, the upper cell of the two (d^) alone is completed, 

 and even that is only weakly closed distally. The cubitus remains 

 quite distinct, and its upper branch shows no signs of any attempt 

 at approaching the media. This arrangement is evidently in cor- 

 relation with the extreme breadth of the wing, with which also we may 

 perhaps correlate the presence of the perpendicular : ' support " on 

 the lower branch of the cubitus. 



When we come to discuss the second cell d. 2 , comparisons fail 

 us, and we can only see in it an instance of specialisation along a line 

 quite extinct at the present day. It would be natural to suppose that 

 this cell is enclosed between the radial sector above and the upper 

 branch of the media beneath, and that the two continue in a fused 

 condition beyond the cell. This, however, is not borne out by the 

 structure of the distal end of rf T . It appears definitely as if both tin- 

 veins enclosing d., were branches of the radius, and that the media 

 only maintains a weak connection with the lower vein by means of 

 the weak distal side of d^. As this problem can never be definitely 

 solved in the absence of the possibility of referring to the precedent 

 wing-tracheatioii, it is preferred to leave these veins unnamed. 



MEASUREMENTS : Total length of wing to apex 20 mm., 

 greatest breadth of fragment 11.5 mm. 



TYPE : Spec. 2a (Plate 3, fig. 6). TYPE-COUNTERPART : 

 Spec. 26. (B. D. Coll.) 



Order PROTOHEMIPTERA. 

 Genus MESOGEREOX, gen. nov. 



Venation, so far as preserved, resembling that of Eugereon as 

 regards the main veins. Cross-veins excessively numerous, but in 

 process of reduction, being represented by numerous short transverse 

 ridges crossing the main, veins, but not proceeding far into the inter- 

 mediate spaces. Spaces between the main veins, beyond the ends 

 of the cross-ridges, filled with moderately numerous and very distinct 

 raised hemispherical tubercles and hollow pits. 



Costa, subcosta, and radius uiibranched in the fragment ; media 

 branched, and probably cubitus also. Cross-ridges absent from the 

 three first-mentioned, veins. 



TYPE : Mesogereon neuropunctatum, sp. nov. 



