618 Mr. \V. P. Pycraft on the 



compressed. The synsacrum is relatively long and narrow, 

 the transverse processes shortening rapidly from the 1st sacral 

 backwards. The bone at the bottom of the pit (inverted 

 during life) is reduced to a paper-like thinness. 



Only in Synallaxis, so far, have I found these pits as well 

 developed. In Siptomis, Homorus, Cinclodes, and Ptero- 

 ptochus they are only imperfectly developed anteriorly, 

 posteriorly they are deep ; in Pipra the reverse is the case. 

 In Pseuclocolaptes and Philepitta they are wanting altogether, 

 In other birds not even remotely related to these forms I find, 

 here and there, minute traces of these pits — such occur, for 

 example, in Tardus. 



The Ribs. 



Five pairs of ribs reach the sternum. The uncinate 

 processes arc very long and .slender, and this obtains also in 

 I he Synallaxina. 



The Shoulder -yir die and Sternum. 



The shoulder-girdle of Acanthidositta is Passerine in type. 

 The coracoid and corpus sterniare subequal in length; there 

 is no procoracoid processes. 



The furcula is only slightly arched; and the expanded 

 free end does not overlap on to the acromion process of the 

 scapula. The hypocleideum is small. 



The scapula is abruptly curved at its free edge. 



The corpus sterni bears a moderately well-developed keel ; 

 the spina interna is wanting, and the spina externa is only 

 very slightly forked at its free end, thereby approaching 

 Philepitta, where the fork is wanting. In Synallaxis, Sipt- 

 omis, and Cinclodes, it is to be noted, the spina externa is 

 by no means well developed. 



The anterior lateral processes are long, slender, and 

 directed forwards ; the posterior lateral processes are wonder- 

 fully long and have slightly expanded free ends. 



The Pelvic Girdle. 



The innominate of Acanthidosilta recalls that of Pipra in 

 that the ischio-pubic fissure is extremely narrow, whereas in 



