l8 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. I3I 



the past that it will probably be impossible forever to document the 

 details of the process in these instances, there are many other cases 

 where the course of evolution can be deduced with reasonable cer- 

 tainty from comparisons of existing forms. The complex endosternal 

 structures of the Apterygota furnish a number of examples, for, as 

 described by the Belgian authors cited above, many endosternal ele- 

 ments that are ligamentous in one species or group are still repre- 

 sented by functional muscles or by apparently degenerate muscles in 

 others. In the opinion of the present writer, yet other parts of the 

 endosternum that are invariably ligamentous in the apterygote species 

 so far studied are homologous with muscles, such as the transverse 

 muscles, that persist as contractile elements in some primitive Ptery- 

 gota and as ligaments in others. Another clearcut set of examples of 

 the replacement of muscles by endoskeletal structures is found in the 

 later history of the sternal arms themselves, for instance in the de- 

 velopment of the furcopleural fusion, which has occurred inde- 

 pendently in numerous lines of descent. Here the process can be fol- 

 lowed in some detail through several series of intermediate stages 

 provided by existing forms. 



As a generalization we offer the hypothesis that all such endoskele- 

 tal developments owe their inception to other structural or functional 

 changes that have limited freedom of movement at the insertions of 

 certain muscles. These muscles, deprived of their original effective- 

 ness as contractile organs, are doomed to disappear unless they happen 

 to retain some value in the role of static supports or braces. Further- 

 more, the organism evidently finds it more economical to construct 

 the braces it requires from other than contractile tissue, which cannot 

 resist compression and which can maintain tension only through a 

 continuous expenditure of energy, so that replacement of bracing mus- 

 cles or tensors by noncontractile ligaments or by stiffer sclerotized 

 apodemes is the usual evolutionary pattern. In our view, structures 

 originating in this manner constitute the primary endoskeletal rudi- 

 ments. Once established, these may be variously molded in later evo- 

 lution in accordance with the mechanical requirements they are called 

 upon to fill ; and in the course of such modification their original 

 derivation from muscles may be almost wholly obscured. 



Returning to the narrower problem of the nature of the sternal arms 

 and their longitudinal musculature, we may point out that the arms 

 are represented in the Apterygota by ligamentous straps that con- 

 nect the thoracic endosterna, which are mainly intersegmental in char- 

 acter, with the respective preceding segmental sternal regions (refer- 

 ences in Barlet, 1954). In these insects the endosterna provide the 



