Carinate Birds from Central Madagascar. 353 



The humerus of Sarcidiornis is at oace distiaguishablc by 

 the relatively greater length and slenderness of its shaft. 

 Measurements of a considerable number of humeri show 

 that they are separable into two groups, the average length 

 of the bone in one being 147, iu the other 132 mm. The 

 larger bones are very few in number, while the smaller are 

 very numerous; presumably the former belonged to the 

 male birds, the latter to the females. 



The radius agrees in structure with that of Cheaalopex 

 (£gypt'iacas , though some specimens seem more slender. In 

 Sarcidiornis, on the other hand, this bone is very diflferent, 

 the whole distal third being expanded, instead of there being 

 a sudden widening close to the extremity of the bone. In 

 this respect C. jubatus somewhat resembles Sarcidiornis. 

 The length of the smaller radii is about 126 mm., that of 

 the larger about 134 mm. The extreme range of variation is 

 between 125 and 135 mm. The length of this bone in Chena- 

 lopex (Bgyptiacus and Sarcidiornis is about 122 and 126 mm. 

 respectively. 



The ulna is essentially like that of Sarcidiornis and Chena- 

 lopex (Bgyptiacus, but in the smaller olecranon process and in 

 the proportions of the shaft comes nearest the latter. The 

 average length of the smaller (female) ulnae is about 130 mm,, 

 that of the larger (male) 140 ram. ; the extreme range of 

 variation observed is between 129 and 142 mm. In Chena- 

 lopex (Bgyptiacus and Sarcidiornis the length of the uluse 

 measured is about 129 and 133 mm. respectively. 



The metacarpus (Pi. IX. fig. 2) agrees precisely with those 

 of the skeletons of Chenalopex (egyptiacus with which I have 

 been able to compare it ; on the other hand, it differs widely 

 from that of Sarcidiornis. For instance, the proximal articu- 

 lation, looked at from the side, is larger and more rounded 

 in outline, and the spur on metacarpal I. is curved forward, 

 instead of pointing directly outward as in Sarcidiornis. 

 Again, in our fossil the interosseous foramen is larger than 

 in Sarcidiornis, owing to the second and third metacarpals 

 being less extensively fused at their distal extremities. 

 It may be remarked that the metacarpus figured by Newton 



