594 PEOF. H. G. SEELEY Olf THE 



in front of it. Its external border appears to be straight and 

 directed forward. It cannot be determined with certainty, and 

 may be part of the scapula or the epiclavicle or clavicle ; most likely 

 the latter, for it is improbable that the epiclavicle has a separate 

 existence as a large bone in Mesosaurus, seeing that no trace of it 

 is known in European allies of the type. There is every reason 

 to look for the epiclavicle in snch a position in animals of this type, 

 but no specimen shows it. In 1865 ^ I suggested that the lateral 

 plate of the Plesiosaurian scapula was the clavicle. This was cor- 

 rected in 1874 ^ by finding the clavicles in association with the 

 interclavicle. Mr. Hulke suggested in 1883,^ that the antero- 

 internal part of the bone which articulates with the coracoid to form 

 the glenoid cavity, is the precoracoid; and that the thin lateral 

 ascending plate which extends above the head of the humerus 

 is a part of the scapula. While the body of this bone both in 

 Plesiosaurus and Mesosaurus appears to me to be the scapula, there 

 is some evidence to suggest that the ascending plate of that bone 

 in Plesiosaurus is the epiclavicle. There is no evidence of any 

 such structure in Mesosaurus, though the bone identified as a 

 clavicle in Dactylosaurus and in Stereosternum is in a not dissimilar 

 position with regard to the head of the humerus. 



After the eleventh vertebra, dorsal ribs are developed. Seventeen 

 dorsal vertebrae thus characterized are more or less perfectly pre- 

 served. The bodies of the vertebrae increase a little in depth as 

 they extend backward, owing to the transverse process being 

 inclined rather more upward. 



The attachment of tiie rib is by a transversely ovate facet, which 

 is placed beloiv the zygapophysis on the anterior face of the neur- 

 apojphysis, so as to look forward and slightly outward and downward. 

 I am not aware of a similar mode of attachment for ribs in any 

 other animal. When the ribs are in situ, their proximal ends, 

 which are compressed, have the appearance of being wedged into the 

 interspaces between the ascending processes of the vertebrae. The 

 articular faces of the contiguous centrums are seen between the 

 twenty-second and twenty- third vertebrae. The centrum contracts 

 a little to its articular ends, which are small and circular, and 

 conically cupped as in the Kimberley specimens. The ribs are 

 cylindrical, strongly curved in a bow, a little compressed proximally, 

 truncate distally, and stout like the ribs of Lariosaurus. The 

 first pair, partly covered by the shoulder-girdle, is short and con- 

 spicuously slender. The others are of approximately uniform size, 

 the earlier measuring 2-5 centimetres in length, and in the middle 

 of the specimen 3 centimetres ; they are about 0*375 centimetre in 

 diameter, but the antero-posterior width rather exceeds the thickness. 

 The interspaces between the ribs are wider than the ribs. This 

 specimen shows no trace of abdominal ribs, unless one or two hair-like 

 rods at the distal end of the right humerus should be of that nature. 



^ Ann. & Mag. Nat. Hist. ser. 3, vol. xvi. p. 358, pi. xv. 



^ Quart. Journ. Geol. Soe. vol. xxx. p. 444. 



* Ibid. vol. xxxix. (Presid. Address), Proc. p. 46. 



