S. H. SCUDDER ON PALAEOZOIC COCKROACHES. 55 



extreme apex ; it emits four branches at unequal distances apart, all of them nearly longi- 

 tudinal, the first being compound and dividing only at the middle of the wing, 1 the second 

 doubly forked, and the third simply forked, both at a long distance from the origin, while 

 the last, arising opposite the fork of the third, is simple. The externomedian vein is very 

 broadly sinuous, being rather strongly arcuate at the base, then runs in a nearly straight 

 line a little divergent from the costal margin, and, finally, in the apical third of the wing, 

 becomes more longitudinal, and terminates just before the apical margin ; it first divides 

 opposite the second branch of the scapular vein, or at the end of the basal two-fifths of the 

 wing, and emits at subequal intervals, the last a little beyond the middle third of the wing, 

 four superior longitudinal branches, the first of which runs clown the middle line of the 

 wing, forks at a little before the end of the middle third of the wing, its upper fork again 

 dividing ; the second forks in the middle of its course, and the others are simple ; all are 

 closely crowded together, and occupy upon the border the lower part of the apical margin. 

 The internomedian vein follows nearly the direction of the preceding, being strongly 

 arcuate at the base, straight and considerably oblique in the second quarter of the wing, 

 beyond this subparallel to the costal margin ; at its change of direction, almost exactly in 

 the middle of the wing, it emits a branch, which runs close to the main stem, and, except- 

 ing for an apical shoot, emits all the regular branches beyond its origin ; including these 

 secondary branches there are about ten simple slightly arcuate oblique veins, whose direc- 

 tion, especially that of the basal ones, is rather at variance, from their regular obliquity, 

 with that of the branches of all the other veins; the basal branches are more closely 

 approximated than the apical. The anal area being broadly tumid, the anal furrow is 

 very deeply impressed, and is very strongly arcuate on the basal half, nearly straight on 

 the apical half, and terminates a little before the middle of the wing ; the anal vein next 

 the furrow is straight and nearly longitudinal at base, curved gently downward beyond, and 

 emits three or four arcuate distant branches; the other veins are very numerous and 

 crowded, generally simple, nearly straight or arcuate, in an opposite sense to the first, and 

 about as longitudinal as the mediastinal branches. 



The wing is of extreme size, the largest of the American species, and only exceeded by 

 Anthracobl. spectabilis of Europe; it is 48 mm. long and 18 mm. broad, or the breadth is 

 to the length as 1 : 2.7; the specimen is almost absolutely perfect, and represents the under 

 surface of a right wing; the principal veins and the main branches of the mediastinal 

 scapular, and externomedian areas before they fork are all distinctly pronounced ; the forks 

 of the same are delicately elevated, while the branches of the internomedian and anal areas 

 are very delicately impressed, — all as seen on the under surface ; the surface is flat, except- 

 ing where the principal veins are most pronounced, and here the interspaces are a little 

 and broadly sulcate ; all the interspaces, even in the anal area, but especially those which 

 are sulcate, exhibit a minute tracery of nearly straight, very closely approximated, exces- 

 sively delicate, scarcely impressed cross lines; those of the anal area are not sufficiently 

 distinct in the plate. 



This insect, from its extreme size alone, cannot possibly be confounded with any other 

 American species, nor from the peculiar distribution of the mediastinal branches, in longi- 

 tudinal bunches depending from the main vein, with any palaeozoic species. This peculi- 



1 The plate represents the first offshoot of this first branch ion; it forks near the tip or directly opposite the extremity 

 as simple, but this is inaccurate, and was overlooked in revis- of the first branch itself. 



