C. W. Andreics — ^pyornis frofn Madagascar. 249 



borne on vertebras 21-23, are not connected with the sternum, to 

 which the next four are joined by sternal elements. The last dorsal 

 bears a free rib, as also do the first two pelvic vertebree. On none 

 of the ribs are there any uncinate processes. 



The ratite sternum is of very peculiar form, and has undergone 

 great reduction. It is a thin plate of bone, very wide from side to 

 side, and extremely short from before backward, especially in the 

 middle line. It has well-developed coracoid grooves and prominent 

 antero-lateral processes, at the base of which pneumatic foramina 

 open. It is very unlike the sternum of any other ratite bird except 

 Apteryx, but this similarity is probably due to the fact that in both 

 cases the bone has undergone extreme retrogression. 



The sternum has been figured on pL ix of the above-mentioned 

 paper. 



The coraco-scapula is not unlike that of Casuarius. It consists of 

 a broad ventral plate, the coracoid, and a dorsal curved rod-like 

 region, the scapula, the two making a very obtuse angle with one 

 another, and being completely fused in the adult. The coracoid is 

 perforated in its upper portion by a large supracoracoid foramen, 

 and articulates with the coracoid groove of the sternum by a long 

 straight ventral border. There is a blunt precoracoid process. 

 The upper extremity of the scapula is flattened and somewhat 

 expanded. On the outer border of the compound bone, at the 

 junction of the two elements, there is a small but very clearly- 

 marked glenoid cavity. The coraco-scapula has been figured loc. cit., 

 p. 386, fig. A. 



The wing in these birds was reduced to a mere rudiment, and 

 probably scarcely any trace of it was visible externally. The 

 humerus is a short rod of bone, the upper end of which bears 

 a rounded articular head and a mere trace of the pectoral crest. 

 The lower end has not the usual surfaces for articulation with the 

 radius and ulna, but only a single one. The remainder of the wing 

 is not certainly known, but in Dr. Forsyth Major's collection there 

 is a bone which I believe to represent the radius, ulna, metacarpus, 

 and phalanges, greatly reduced and fused into a common mass. 

 This remarkable structure bears on its thickened proximal end 

 an articular cup, into which the distal articulation of several of 

 the rudimentary humeri fits very well. Just beyond the articular 

 end the outer surface bears a deep depression in the middle 

 line, indicating, as I believe, the position of the division between 

 the radius and ulna. Distally the bone becomes thin and 

 plate-like, and terminates in an irregular bifid extremity ; this 

 portion probably consists of the last remnants of the metacarpus 

 and phalanges. 



The complete bones of the hind limb from Sirabe have never been 

 heretofore described and figured, with the exception of the metatarsus, 

 but I can only refer briefly to them on this occasion. The femur 

 is a stout massive bone, of which the head is nearly spherical 

 and the neck short and thick. The trochanter is very large ; its 

 upper surface slopes steeply up from the neck, and rises high above 



