Sellards— Structural Characters of Cockroaches. 309 



JEtoblattina mazona is now known in several stages, from a 

 young age to the adult condition* (Plate VII, figures 1-4). The 

 nymph stage of Mylacris elongata has also been recognized 

 with a fair degree of certainty. As in recent Orthoptera, the 

 young resemble the adults, growth taking place by a succession 

 of moults. The wings appear gradually, and certain other 

 changes similar to those seen in existing cockroaches, as the 

 gradual rounding out of the posterior margin and angles of the 

 pronotum can be traced. 



An important part of the data regarding the structure of the 

 body and appendages has been obtained from the nymphs, and 

 from these also it has been possible to clear up the much mis- 

 understood genus Dijpeltis. This fossil, which has been 

 variously referred to the xiphosuran and phyllopod crustaceans, 

 is in reality an immature cockroach of the tribe Mylacridse and 

 probably of the genus Mylacris.\ Through the kindness of 

 Mr. Charles Schuchert, the type of Dijpeltis together with the 

 other National Museum material has been accessible for study. 

 The contrast between fossil and matrix in the Mazon Creek 

 nodules is often not well marked, and structural characters, 

 which when once discovered are perfectly evident, are in dan- 

 ger of being at first overlooked, especially when only a few speci- 

 mens are at hand for study. This has been particularly true of 

 Dipeltis. The type specimen preserves, in fact, besides the 

 prothorax and first pair of wing pads, the head, antennae, and 

 the first and second pairs of legs (Plate VII, figure 8). About 

 half of the head projects in front of, and the remainder is seen 

 in outline through, the pronotum. The left antenna, which is 

 now more completely uncovered, is apparently joined to the 

 head by an enlarged socket-like attachment resembling that of 

 modern cockroaches. The segments are numerous. Those 

 near the base are wider than long ; farther out they become as 

 long or longer than wide. The first pair of legs is seen indis- 

 tinctly through the pronotum. The femora of the second pair 

 are well preserved. They are smooth, stout, and of moderate 

 length. The trochanters are less distinct, and only indications 

 of the coxse are to be seen. The tibiae are somewhat longer 

 than the femora. Part of the right tarsus is preserved. The 

 first segment is comparatively long. Those following are 

 indistinct. 



The true nature of Dipeltis was, indeed, suspected by Gahan, 



* The evidence for the generic and specific reference of the nymphs, in so 

 far as they have been identified, will be given in the completed paper. In 

 the case of E. mazona, the connection is quite satisfactorily established. 



f This statement refers to the type species of the genus, Dipeltis diplodiscus 

 Packard (Amer. Nat., vol. 19, p. 293, 1885; Mem. Nat. Acad. Sci., vol. iii, 

 p. 145, 1886). The second species, Dipeltis Carri Schuchert (Proc. U. S. Nat. 

 Mus., vol. 19, p. 671, 1897), from the illustration, appears to the writer to 

 belong with Etoblattina mazona (cf. Plate VII, figures 1 and 3 of this paper). 



