1906.] OF THE TRACIIEOPHOXE PASSHUES. 147 



V. Sternum and >Shoulder-Girdle. 



The sternum of the Tracheophone Passeres throws an interesting 

 light on the evolution of the typical Passerine sternum, in wliich 

 the corpus sterni is quadrangular in shape, with a pair of notches 

 in its posterior border, long anterior lateral processes, and a 

 forked spina externa, the spina interna being wanting. 



In the Conopophagidse the sternal plate, however, differs con- 

 spicuously from that of all other Passeres, in that there are 

 four posterior notclies. Yarying in the depth of the notches and 

 other small points in the different genera, these peculiar sterna,, 

 together with the sternum of Pitta, have hitherto been regarded as 

 modifications independently acquired — those who held this view 

 not regarding the genera which I have grouped together as 

 closely allied forms. Thus, Forbes, for example, in his paper on 

 the genus Conopophaga (2), wrote, " As regards the possession of 

 a four-notched sternum b}'' these birds and the Pteroptochina^, I 

 am not inclined to consider it in any way a primitive character', 

 but rather as an instance of a simple modification having been 

 independently acquired in diiferent groups of birds." Nevertheless, 

 I venture to think that the brigading of these types with four 

 notches to the sternum is justified, not on this character alone, of 

 course, but because of the numerous characters which all share 

 in common. 



It is to be noted that among the Ooraciiform birds, from which 

 the Passeres are probably derived, a four-notched sternum is the 

 rule. Hence the sternum of the Conopophagidse may be a 

 survival of this older order of things ; on the other hand, it may 

 be that this family has developed this character anew. 



I incline to the opinion that the number of these notches is to 

 be regarded as a primitive character', and that while in Hylactes, for 

 example, the notches have tended to deepen (text-fig. 51 a, p. 148), 

 so as to acquire a superficial likeness to those of the Bucconidae, 

 in Conopophaga they are on their way to reduction to the normal 

 Passerine number, by the closing of the inner pair, which are 

 now represented apparently by fenestras only (text-fig. 51 b). 

 Such a method of reduction is common in the Accipitres, for 

 example. In the Menuridje the posterior border of the sternum 

 is entire ; in the Pittidee but a single pair ai-e left, which, 

 however, are quite unlike the notches in typical Passerine sterna 

 and recall those of the Turnicidfe (text-fig. 51 d)! This form 

 of sternal plate seems to be, however, nothing more than an 

 exaggeration of that met with in Cinclodes, for example, as may 

 be seen by comparing text- fig. 51 c. It would seem, indeed, 

 that the Eurylaemid and Formicariid Passeres represent a state of 

 flux in this matter, as well as in the form of the spina externa, 

 which varies in difterent genera of the same family, as in the 

 Eurylsemidee and Cotingidse. In the higher Passeres the bifurcate 

 spina externa and single pair of notches to the posterior border 

 of the sternum are universal. 



10* 



