CARBONIFEROUS COCKROACHES. 409 



Lithomylacris pauperata sp. iiov. 

 PI. 32, Dg. 5. 



The single specimen represents a fragment of a front wing unfortunately very imper- 

 fect, but still sufficiently preserved to show that it is a Lithomylacris, allied to Td.'penn- 

 sylvanica, but much larger than it, and certainly distinct from any known species. Not 

 one of the borders is preserved, although in several places the veins are evidently perfect 

 to their tips. 



Only oue or two of the mediastinal veins can be seen, the outer one long and straight 

 and minutely forked at the tip, with no other oifshoots, unless it be at the base, showing 

 the insect to belong to the Mylacridae ; within this the foi'ked extremity of another parallel 

 vein can be seen. The scapiilar vein is perfectly straight as far as what is probably the 

 centre of the mng, and in this basal portion runs down the middle of the wing, thus prov- 

 ing its proper reference to this genus ; beyond this it curves very gently and slightly up- 

 ward; in the fragment it emits four branches, the basal two close together, the others far 

 apart, and the last at the very tip of the piece; the first is simple, the second and third 

 simply forked at no great distance from their origin, and all run in a longitudinally ob- 

 lique direction, subparallel to the mediastinal vein. The extcrnomedian vein is exactly 

 parallel to the scapular and hardly more distant from it than two adjoining scapular 

 branches from each other; it divides far toward the base, apparently opposite the middle 

 of the anal furrow and just beyond the second branch of the scapular vein, each of its 

 branches again forking opposite the extremity of the anal furrow, and the uppermost 

 branchlet again befoi'e the end of the fragment is reached. The internomedian vein is en- 

 tirely composed of very faintly impressed lines, emits two parallel arcuate branches close 

 together a little nearer the base than the first two branches of the scapular, the first sim- 

 ple, the other forked simply in the middle of its course ; and farther apart and from these, 

 two other simple branches, the first opposite the fork of the preceding branch; all of these 

 branches on approaching the border sweep in a more longitudinal course toward it. The 

 anal fiirrow is not deeply impressed, a little roundly bent beyond the middle, otherwise 

 nearly straight; the anal veins are not preserved, but the furrow appears to strike the 

 inner margin slightly before the middle of the wing. There is no sign of any transverse 

 venation or reticulation. 



The breadth of the wing is 12.5 mm.; the length of the fragment 17 mm.; and the pre- 

 sumed length of the wing about 26 mm.; making the breadth to the length as 1 : 2.1. It 

 could hardly have been much longer than that, and woidd need be very much longer to 

 have had the same proportions as L. pennsylvanica to which it is the more nearly allied; 

 the disposition of its nervures is very similar to what we find there, and even if the propor- 

 tions given are (as they possibly may be) very much out of the way, it is a much larger 

 species than it, being half as broad again. It differs also slightly in neuration, as in the 

 forking of some of the scapular veins, and particularly in the much earlier branching and 

 greater significance of the extcrnomedian vein, which is exceedingly unimportant in L. 

 jjennsylvanica; the entire disposition of the internomedian veins is also different and they 

 play in this species a much less prominent part than in L. pennsylvanica. From the 

 other species it is as widely separated as is its ally. 



MEMOIRS BOSTON SOC. NAT. HIST., VOL. IV. 55 



