1889.] ANATOMY OF burmeister's cariama. 597 



In Chunga (fig. 3, p. 598) there are differences in all the points 

 just enumerated. The first rib is very rudimentary ; only ribs 4, 5, 

 and 6 {i. e. one less than in Cariama) are furnished with uncinate 

 processes, which are straight, directed upwards and backwards, and 

 not curved ; the sternal half of the first complete rib is slender. 

 There is a rudimentary 8th rib on each side ; on the left side of the 

 body it consists of a curved piece continuous below with the sternal 

 portion of the 7th rib ; on the right side a shorter piece lies along the 

 jiosterior border of the sternal half but not fused with it ; there is 

 also a small rudiment of a vertebral rib attached to the transverse 

 process of the 8th dorsal vertebra. 



The proportion between the length (from point to point) of the 

 lateral margin of the sternum and the length of the space occupied 

 by the attachment of the sternal ribs is : — 



in Cariama, 2'4 : 1"35, 

 m Chunga, 2*5 : 1*15, 



showing that in the latter genus the attachments of the ribs are more 

 crowded together than in the former ; at the same time the first 

 sternal rib is attached much nearer to the anterior lateral process 

 of the sternum in Chunga than it is in Cariama. The sternal ros- 

 trum is more developed in the latter type, and there is a difference 

 in the shape of the sterna on a lateral view which will be more 

 easily appreciated by an inspection of the accompanying woodcuts 

 (figs. 3, 4) than by a description. 



In the pelvis (see figs. 5, 6, pp. 600, 601) the chief differences 

 are, firstly, that the ilia extend rather further forwards in Cariama, 

 very nearly reaching the last rib but one ; in Chunga the anterior 

 extremities of the ilia only just get beyond the seventh rib. Secondlj', 

 the line of junction of the transverse processes of the lumbar ver- 

 tebrse with the border of the first acetabular ilium forms a straight 

 line ; in Chunga the corresponding line is curved. The breadth of 

 the pelvis is greater in Chunga, the proportion between length and 

 breadth in the two types being as follows : — 



Length 



of pelvis. Breadth, 



in. in. 



Chunga .... 2*95 VQ 



Cariama. . . . 325 1*6 



These measurements of breadth are taken from the extremities of 

 the post-trochanteric processes, which are well marked in both these 

 birds, but perhaps if anything rather more marked in Chunga. 

 If these measurements had been made between the antitrochanteric 

 processes, the contrast between the two types in the proportions 

 between length and breadth of the pelvis would have been greater 

 than is indicated in the above table ; these processes are more 

 strongly developed and project further out in Chunga than they do 

 in Cariatna. 



