314 MR. H. BOLTON ON INSECT-EEMAINS FROM [Sept. I912, 



the wing must clearly have possessed an acute tip. IS"© evidence 

 is forthcoming as to the character of the anterior portion of the 

 wing, nor as to that of the costal and subcostal veins. 



Description. — The radius is represented by a single forked 

 vein, the stem of which has not been preserved. The radial sector 

 gives oif four (possibly five) branches backwardly, which occupy 

 the distal sixth. 



The median is close to the origin of the radial sector, and divides 

 low down into two principal branches, the outer of which divides 

 into two and the inner into three rami before they reach the inner 

 margin. The marginal area occupied by the median is a little 

 longer than that occupied by the radial sector, and together with 

 the latter forms the outer half of the inner wing-margin. 



The cubitus is a relatively short and stout vein. It divides into 

 two principal branches, the foremost having two rami and the 

 hinder breaking up into four by a double bifurcation. Owing to 

 the wide divergence of its branches, the cubital area is larger than 

 that of the median. 



The anal area was evidently large, but the whole surface has 

 been pitted by attempts to clear the matrix, and so the presence 

 and character of veins cannot be distinguished with certainty. 

 It is possible that certain surface-indications are evidence of two 

 simple widely-separated anal veins. 



The wing-surface exhibits a strong plication, the area between 

 any two veins dipping down straightly and smoothly on each side 

 towards a clearly-defined furrow. Viewed obliquely, the wing 

 shows a series of ridges formed by the veins with V-shaped inter- 

 vening sulci, which only flatten out close to the wing-margin. 

 Springing from the principal veins are a close series of stout cross- 

 veins, which follow an oblique course in the direction of the wing- 

 margin. The portion of wing preserved is sufficient to indicate 

 that the wing was strongly triangular in outline, the proximal 

 portion being probably over 20 millimetres wide, and the width 

 of the wing near the tip measuring not more than 8 mm. 



Affinities. — The attempt to classify this wing has been a task 

 of considerable difficulty. I finally came to the conclusion that it 

 had affinities with the genus Polycreagra, but was of a more simple 

 and more Dictyoneurid type. With the latter view Dr. Handlirsch 

 agrees, and has pointed out that it is not possible to bring it within 

 the family Polycreagridae, but that it must be taken as the type of 

 a new family of the Palaeodictyoptera. 



To this new family, which is not remote from the Heolidse, 

 I propose to assign the name of Pteronidise. 



Family Pteronidiae. 



Characters. — Wing three times as long as wide. Costal border 

 gently arcuate. Subcosta and radius closely approximate to the 

 costa, and reaching the latter near the tip of the wing. 



