FAUNA OF THE GOSATJ FOEMATION. 



683 



face was moderately rough, with the margin slightly rounded ; and 

 it was concave from front to back. Just above the posterior rounded 

 termination of the articulation the bone is thickened so as to form a 

 very slight tuberosity, most conspicuous on the internal surface, 

 where a groove divides it from the articular head. This inferior 

 surface of the bone is convex from side to side, but soon becomes 

 compressed, so that the blade, where fractured, at 4 centim. from the 

 articulation, is less than 2 centim. thick. On the external surface 

 the specimen is concave above the articulation, partly because there 

 is a slight rounded ridge on the posterior edge, not, however, separable 

 from the rounded posterior border, and partly because the bone is 

 becoming thickened anteriorly, evidently in relation to the develop- 

 ment of a strong spine or crest at about 2 or 3 centim. above the 

 articular surface. The articular margin is marked with longitudinal 

 lines of ligamentous attachment. It is of course impossible to speak 

 of the affinities presented by this form of scapula ; but the bone ap- 

 pears to me to be distinct from the scapula of Cratceomus, Anoplo- 

 saurus, Iguanodon, Scelidosawrus, and other types with which it 

 might be supposed to be allied. 



A flat expanded bone, very imperfectly preserved, appears to be 

 a portion of a coracoid. It has but one margin remaining, which is 

 straight for a length of upwards of 5 centim. and indicates an arti- 

 cular surface at right angles to the bone, such as may have adjoined 

 the scapula. The specimen is upwards of 7 centim. broad and 

 10 centim. wide, as preserved, but is too imperfect for description. 



Vertebrae. 



The vertebral fragments have shared in the general fracturing. 

 There are two portions of sacrum which had thoroughly decomposed 

 before fossilization. One fragment indicates two vertebrae anky- 

 losed together ; the other fragment I regard as showing the base of 

 the centrum of another vertebra, and on this evidence should infer 

 that there were at least three, and probably four, sacral vertebrae, 

 though the second specimen is imperfect in every direction, so that 

 it has to be regarded carefully to be accepted as a vertebra at 

 all. This fragment, as preserved, is 5 centim. long, broad, rounded 

 on the under side and somewhat flattened, with a shallow median 

 depression which is fully 1 centim. wide. 



The same characters are shown in one of the ankylosed vertebrae. 

 These also are imperfect at both ends, so that it is impossible to judge 

 what the length of each centrum may have been ; but, as preserved, 

 the length of the two together is a little over 5 centim. ; and 3 centim. 

 of this belong to what I regard as the more anterior of the two. This 

 vertebra is very slightly concave from front to back, and convex from 

 side to side, with the aspect of being flattened on the under side and 

 with a very slight median depression, which becomes more marked in 

 the second vertebra. The sides are rounded and concavely impressed 

 between the transverse processes. This appears to have been given 

 off at the suture between the two vertebrae, which is somewhat ele- 



