lO SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I3I 



The principal primary muscular relationships of the Us are tolera- 

 bly well preserved in the typical abdominal intersegment, granted that 

 a secondary extension of sclerotization has here merged the in- 

 tersegment indistinguishably with the following segmental sternal 

 plate, of which the former intersegment now forms the antecosta 

 (Snodgrass, 1929). The Us are here represented in the anterolateral 

 angles of the definitive abdominal sterna, which in many insects dis- 

 play the muscular relationships outlined above (Ford, 1923; Maki, 

 1938). 



In the intersegments that follow each of the three thoracic seg- 

 ments, the situation is rarely so transparent. One gains the impres- 

 sion that, even in the most primitive forms that have come to hand, 

 the musculature of the Us has already been subject to extensive shift- 

 ing and reduction, while in more recent insects only a few scattered 

 remnants suggest the original role of the Us as attachment sites for 

 part of the longitudinal body muscles. Moreover, where the trans- 

 verse muscles have been preserved, their lateral attachments now 

 usually appear to be on segmental parts. In addition, we find a few 

 muscles that originate on the thoracic Us or on their present equiva- 

 lents, whose insertions are segmental and which have no counterparts 

 in the legless abdomen. 



Equally difficult to analyze, because of the extremely varied skeletal 

 and muscular relationships that exist in different groups, is the situa- 

 tion in the cervical intersegment. Here one must be content for the 

 present with the assurance that the former OUs are usually somehow 

 represented, most often as part of one or more of the definitive cervi- 

 cal sclerites. 



These problems are well illustrated in the cockroaches, in which 

 the musculature of the Us, though rich in comparison with that of 

 more recently differentiated orders, can only be considered vestigial 

 in relation to the inferred ancestral condition. 



Ventral muscles of cockroaches that appear to belong to the Us 

 complex include (a) transverse muscles ; (b) cruciate coxal and furcal 

 muscles; (c) certain other furcal muscles; and (d) spinasternal 

 muscles of the abdominal Us. (See tables i and 2.) 



a. Transverse muscles. — The transverse muscles of the thorax ordi- 

 narily have a median attachment on the spina, and have therefore 

 been discussed under section i,a, above. The nature of their lateral 

 attachments remains to be considered. As already noted, abdominal 

 relationships support the view that the lateral attachments of the 

 transverse muscles are morphologically intersegmental, i.e., on the 

 Us, Comparative evidence from other arthropods and the scanty em- 



