152 MR. W. p. PYCRAFT OX TliE OSTEOLOGY [Feb. 20, 



enable me to say much concerning the form of the pelvis in these 

 somewhat hetei'ogeneoiis groups, but those genera which I have so 

 far examined in this connection show considerable specialisation in 

 adaption to climbing habits. In the smaller Synallaxine species 

 the pelvis resembles that of the Philepittidie, having the ischia 

 widely exi^anded posteriorly smd produced into a long rod-like 

 foot bent sharply on its long axis, but the pre-acetabular ilivim is 

 I'elatively shorter. In Siptornis the pre-acetabular ilia ai-e 

 widely separated, being divided by a deep trough on either side 

 of the synsacral crest ; but in Synallaxis and Homorus, by the 

 shortening of the transverse process of the supporting vertebi'a?, 

 the innominate bones almost touch one another above the synsacral 

 crest. In Pseudocola-pies the pre-acetabular ilia are subconical in 

 shape and rise to the level of the synsacral crest, but are sepai'ated 

 therefrom by a narrow space. The form of the ischium agrees 

 very closely with that which obtains among the more specialised 

 Dendrocolaptines : since it turns downwards instead of running 

 backwa,rds beyond the level of the post-acetabular ilium ; by this 

 means the depth of the hinder region of the pelvis is greatly 

 increased. In Xiphorliynchus and Dendrocolaptes this deepening 

 is especially noticeable ; furthermore, this is associated with a 

 tendency to close up the obturator fissure by the ossification of 

 the teiulinous fascia stretched between the infeiior border of the 

 ischivim and the pubis. In XiphorJiynchxis this newly ossified 

 matter hangs down from the ischium in the form of a delicate 

 cui'tain of bone and is continued forwai-ds and downwards to form 

 a broad bony plate enclosing the obturator foi'amen. 



The pelvis of Batara, one of the Formicariidse, presents some 

 extremely interesting features. In many respects resembling the 

 pelvis of Dendrocolaptes, it differs therefrom in having the dorsal 

 border of the pre-acetabular ilium strongly arched, and this curve 

 is followed by the synsacral crest which lies between. The ischium , 

 in its general shape and characters, also closely resembles that of 

 Dendrocolaptes, but it differs therefrom mainly in that it fuses 

 completely with the pubis, which is unusually broad and also very 

 short. 



In all these pelves t\\e fovea lumhalis is extremely small, and the 

 fovea ischiadica and j^udsndalis are confluent. 



vii. The Pectoral Limb. 



The pectoral limb in the groups here described presents a some- 

 what remarkable uniformity even in structural details : so much 

 so that it would be difficult, on the evidence of the wing alone, to 

 determine to which of these families the skeleton belonged. 



With but few exceptions the forearm is the longest segment of 

 the wing, but then is never mai'kediy longer ; the arm and manus 

 are subequal. There is no coraco-humeral groove, and the 

 chai'actei'istic pit foi- the hrachialts anticus is situated below and 

 proximad of the ulnar condyle. 



