Vol. 67.] EAUNAL HORIZONS IN THE BRISTOL COALFIELD. 335 



surface is crossed by one vein forking three times, and by five other 

 veins, two of which bifurcate twice. All the veins are narrow and 

 deep, while the interspaces are crossed at right angles by trans- 

 verse veins which divide them into somewhat regular rectangular 

 areas. Where forking of the veins takes place, the resultant 

 branches first diverge rapidly, and then pass outwards at right 

 angles to the wing-margin in parallel lines. The upper and lower 

 branches of the third and fourth veins appear to run together, and 

 must ultimately fork a little beyond the present broken edge. The 

 hinder edge is somewhat sinuated. There can be no doubt, how- 

 ever, that the margin is broken away, and still hidden under the 

 matrix of the counterpart block of shale. 



Notwithstanding the fragmentary character of the wing, it is, 

 I believe, possible to determine that the veins described are 

 portions of the radius, the trifurcate upper vein forming the apex 

 of the radius, and extending somewhat forward of the tip of 

 the wing. A little of the distal outer margin can be traced in 

 front of the tip of the wing. The small area enclosed between it 

 and what I suppose to be the main axis of the radius (that is r 

 the trifurcate vein) is crossed by fine, short, deeply-incised veins. 

 The lower wing-fragment has a well-defined outer (?) margin 

 evidently bounded by the costal vein. It is regularly and broadly 

 convex, and shows signs of passing into a straight or incurved line 

 proximally. The tip of the wing is narrow and bluntly rounded. 

 Five longitudinal veins are present, the third forking just in front 

 of the fractured proximal edge. The fifth vein forks about 2*5 

 millimetres below the tip of the wing. All the veins run out into 

 the costal margin. 



These wings, fragmentary though they be, present features of 

 unusual interest, as they are wholly unlike blattoid wings in every 

 particular ; and their relationship must be sought for in other groups. 

 The resemblance in venation and in structure to fragments of 

 locustid wings is remarkably close. There are the small, thin, 

 sharp longitudinal veins in each; the division of the interspaces by 

 transverse branches into rectangular areas; and the fragments show 

 a texture which was quite filmy, apart from the marks left by the 

 veins : this again is a locustid feature. 



"Wings of this character were classed by Scudder in a genus 

 Genentomum, referred by him to the family Homothelidas. Brauer 

 found an affinity in Scudder's form to the Sialids, while Bron- 

 gniart recognized its relationship to the Orthopteres sensu stricto. 1 

 Dr. Handlirsch refers Scudder's genus to a new family (Edischiidae, 

 in which the forms are characterized by the coalescence of the 

 superior branch of the median vein of the front wing with the 

 radial sector, and its bifurcation farther out as an apparent offshoot 

 of the latter vein. 



So far as determination of such fragments is possible, I should 

 assign the above-described specimens to Scudder's genus Genentomum 



1 A. Handlirsch, ' Revision of American Palaeozoic Insects ' Proc. U.S. Nat. 

 Mus. vol. xxix (1906) pp. 661-820, especially pp. 700-701. 



