458 SAMUEL H. SCUDDER ON THE 



from living types by the complete interdopondence of two of the veins of the fore wing, 

 and by the fact that the anal veins of the same wing impinge successively uj^on the in- 

 ner margin and do not either fall ujion the anal fin-row or become clustered together at 

 the outer point of the anal area; and I have accordingly grouped them under the name 

 Palaeoblattariae to distinguish them from their descendants the Blattariae proper. Now 

 as not only all the living cockroaches belong to the Blattariae proper, but the same is 

 true of all mesozoic cockroaches later than the trias/ it becomes a matter of special inter- 

 est to inquire into the position of these Fairplay cockroaches, which number eighteen spe- 

 cies. 



Eleven of the species and five of the seven genera belong to the Palaeoblattariae. 

 These five genera are Etoblattina (2si).), Spiloblattina (4 sp.), Anthraeoblattina (1 sj).), 

 Petrablattina (2 sp.), and Poroblattina (2 sp.). Only five of the species belong to gen- 

 era previotisly known, Spiloblattina and Poroblattina having been established upon and 

 still being known only by the Fairplay fossils. But the difference is moi'e marked than 

 this; for at least one of the species referred to Etoblattina is an aberrant form with an 

 excessively long internomedian vein; and both the species of Petrablattina agree in dif- 

 fering from those heretofore known to a very considerable degree. Of the new genera, 

 Spiloblattina is very peculiar in the strongly divergent and then convergent curve of the 

 externomedian and internomedian veins around a large stigma near the middle of the 

 wing, iniknown in any other cockroach, ancient or modern, so far as I know; but other- 

 wise it is related to Etoblattina; while Poroblattina is more nearly related to Petrablat- 

 tina, and especially to the two new species of that genus from this locality. 



The average size of these Fairplay Palaeoblattariae is much less than that of the pa- 

 leozoic Palaeoblattariae in general. The average length of the front wings of the 

 paleozoic species is 26 mm.; that of these Fairplay Palaeoblattariae, 16 mm. This fact 

 has its value, for the Jurassic species are nearly all of very small size, and the wing- 

 length of the remaining species from Fairplay (i. e., those which do not belong to the Pa- 

 laeoblattai'iae) is less than 8.5 mm., ranging from 6.5-11.5 mm. This agrees completely 

 with the size of mesozoic species already known. The average of all the Fairplay cock- 

 roaches is less than 13.5 mm. 



As to the seven cockroaches from Fairplay which belong to the Blattariae, the char- 

 acteristics of their venation as well as their small size show them to be closely allied to 

 jvirassic forms, although the two genera to which they belong are peculiar so far as yet 

 known to Fairplay. One of them is distinctly allied to Rithma, as defined by me in my 

 memoir on mesozoic cockroaches (loc. citJ), while all the species have a decided mesozoic 

 aspect and would at once be considered liassic or at least Jurassic by any one at all fa- 

 miliar with the forms already known from these deposits. They have on the other hand 

 an entirely different aspect from any and all paleozoic forms, and present no points of 

 close comparison with any Palaeoblattariae excepting some of those mentioned above 

 from the same Fairplay beds, notably with the genus mentioned under the name of Poro- 

 blattina, which one of the genera not a little resembles. The manner in which the change 

 of the venation was brought about has been stated by me in my memoir on mesozoic 

 cocki'oaches. 



' See Mem. Bost. Soc. Nat. Hist., iii, 439 seq. 



