NO. II THORACIC MUSCLES OF COCKROACHES — CHADWICK I7 



consists of the infolded furcabasis and the laterally extended furcal 

 arms. In cockroaches, as in other primitive forms, right and left 

 apophyses remain separate. For this reason, purists avoid applying 

 the term "furca" to them, but for convenience we shall continue to 

 refer to them as the furcae or furcal arms, with which they are 

 homologous. Despite the seemingly incontestable segmental nature 

 of these apophyses, they nevertheless carry a large fraction of the 

 surviving longitudinal ventral intersegmental musculature in all ptery- 

 gote insects. This situation poses a contradiction, long recognized and 

 accepted by students of thoracic structure, that has never been satis- 

 factorily resolved (cf. Snodgrass, 1929). 



Weber (1928) surmised that the present furcal intersegmental mus- 

 cles had been derived from spinasternal muscles. He proposed that, 

 as the furcal arms were gradually elevated in phylogeny, they inter- 

 cepted the spinasternal muscles, which thereupon acquired furcal at- 

 tachments and lost their primary connections with the spinae. This 

 hypothesis, which regards the furcal muscles as replacements for the 

 spinasternal muscles, is clearly untenable in the face of the presence 

 of the usual complement of furcal intersegmental muscles in all those 

 primitive forms, such as larval Dytiscus (Speyer, 1922) or Cyhister, 

 larval Corydalus, and the cockroaches, which still retain an extensive 

 spinasternal musculature, including (in the blattids) both spinaspinal 

 muscles isps-2sps, 2sps-"jsps." Conceivably, the rising furcal arms 

 could have intercepted some of the more lateral bands of the primary 

 ventral longitudinal intersegmental muscles, for instance those at- 

 tached on or below the transverse ligaments, isps-iils, etc. ; but even 

 this modification of Weber's hypothesis is unconvincing in the absence 

 of any known situation in insects where interception of a muscle by 

 a skeletal element has led to the development of an attachment be- 

 tween the two. Moreover, the data of Carpentier, Barlet, and others 

 (see Barlet, 1954, for references) show that the essential features of 

 the furcal complex exist in the Apterygota, which also possess an 

 extensive array of muscles homologous with the spinasternal muscles 

 of higher forms. Any notion that the furcal longitudinal muscles have 

 arisen in the Pterygota through transfer of muscles from some other 

 category must therefore be abandoned. How then can they be ac- 

 counted for? 



An answer may be approached, we believe, through realization that 

 the principal endoskeletal structures of insects and other arthropods 

 have all developed as the result of sclerotization along the course of 

 former muscles, and that the present sternal arms are of this nature. 

 Although the genesis of certain endoskeletal structures lies so far in 



