550 Mr. R. W. Hooley on the 



diagonally from the preaxial distal border. This is not 

 exhibited in the Cambridge examples. The specimen no. 2 

 is not as compressed as nos. 1 and 3, or as Ornithodesmus 

 latidens, and the dorsal surface is more concave distally. 



The Ulna. 

 Proximal end. 



The six specimens on tablet J. a. 9 are the proximal ends 

 of ulnse. Nos. 1, 2, 4, and 5 are figured in ' Ornithosauria/ 

 plate iii. figs. 4<, 5, 6, 7, 8, 9. J. a. 9, \, belonging to the 

 left ulna, differs from Ornithodesmus latidens in the absence 

 of the longitudinal ridge on the ventral surface of the shaft, 

 in lieu of which there is a raised and roughened surface, 

 preaxial to the radius instead of postaxial, for the attachment 

 of the biceps tendon. This feature is also seen in nos. 2 

 and 4. The dorsal surface is strongly convex, and the 

 ventral slightly, and. free from any pit or ridge. The 

 pneumatic foramen is near the articulation in the centre of 

 the ventral surface. The articulation is much worn. This 

 specimen is interesting, because from it Seeley obtained the 

 suggestion of an olecranon *. There is a well-defined line 

 around the upper dorsal half, which might be accidental. 

 The surfaces in all the other examples appear to be articu- 

 latory, and the roughened edges the effect of wear, and not 

 caused by the tearing away of an epiphysis. The main 

 articulatory surface is an oblique oval-shaped basin, looking 

 upwards, in the centre of which in specimens nos. 4, 5, and 

 6 is a circular opening into the bhaft, as is seen in the 

 liumerus of Orniihudesinus lutidtns and in J. a. 6, nos. 20, 

 21, and 32. Moreover, the general form of the bone is not 

 very different from the distal extremities of the humeri, 

 exhibiting the circular opening into the shaft — for example, 

 J. a. 6, 20. In those examples wheie the supposed olecranon 

 has conic away the dorsal half of the articulatory surface is con- 

 cave. It looks upwards and is divided from tlie ventral half 

 by a convex ridge. The ventral surface looks downwards and 

 is feebly convex dorso-ventrally and concave pre-postaxially. 

 In no. 1 the articulation has two feebly concave surfaces, 

 with a raised ridge for the trochlea of the distal end of the 

 humerus. In no. 2 the dorsal half of the articulation is 

 destroyed. The postaxial concave surface is more oblicpie 

 and carried further on to the shaft of the bone, thus looking 

 more outward than in the other specimens. This example 



* 11. G. Seeley, ' OrnJLliosauii;i/ 1870, pp. 45 & 46. 



