NO. II THORACIC MUSCLES OF COCKROACHES — CHADWICK II 



bryological data on insects (Heymons, 1895 ; Roonwal, 1937) justify 

 this inference. In the postembryonic cockroach, however, the defini- 

 tive connection in the thorax is usually with the anterior margin of 

 the succeeding episternum, and the muscles are therefore to be desig- 

 nated as isps-epsg, ssps-epsg, even though Maki (1938, p. 58) de- 

 scribes the attachment as "on the small sclerite before the [mes] epi- 

 sternum" in P. aiistralasiae. If this were all the evidence available, 

 one would conclude that in blattids the lils and 2ils had fused with 



Table 2. — Ventral intersegmental muscles of cockroaches: 

 muscles witJiout spinasternal attachments 



Symbols such as M6, C55 refer to the numbers given the corresponding mus- 

 cles in P. australasiae by Maki (1938) and in P. americana by Carbonell (1947). 

 For muscles with spinasternal attachments, see table i. 



^ The 3 cruciate muscles are not serially homologous. For discussion of shifts in origin 

 of eps,-cXjX and eps,-fUjX, see text sections 2, a and 2,b. The three muscles or their equiva- 

 lents are present in all species examined. 



* The three postcoxal ligaments are serially homologous. For variations in their occurrence, 

 see text section 2,c. 



* The furcal muscles are probably all serially homologous, at least in a broad sense. For 

 variations in the abdominal insertion of fus-s,j^, sec text section 3. Carbonell (1947) records 

 the abdominal attachment of muscles number 192 as on SJ^^ in P. americana. All the furcal 

 muscles listed are present in all species investigated. 



epS2 and epSs, respectively. The arrangement of the cruciate muscles 

 of the profurca and procoxa in P. americana and in Cryptocercus 

 (see below) is also consistent with this interpretation. 



However, the full story is not that simple, for Cryptocercus pos- 

 esses not only muscles isps-epSz, 2sps-epS3 that are clearly homol- 

 ogous with those of other blattids, but has in addition fibrous liga- 

 ments (figs. 10, 13: 4, 17) that run from the spinae to small sclerites 

 in the intersegmental membrane well in advance of the episternal 

 margin. These transverse ligaments have evidently been derived from 

 former muscles, and their lateral attachment sites not only occupy the 



