104 Transactions of the South African Philosophical Society. 



everything of its osteology is now known except the structure of the 

 skull and the shoulder girdle. 



A very fine specimen of a moderately large Mesosaurus has 

 recently been discovered in Bushmanland by the Eev. C. A. 

 Neethling, of Nieuwoudtville, and Mr. Neethling has been good 

 enough to forward the specimen to me for examination. 



This specimen consists of the greater part of the thoracic and 

 abdominal regions, both hind limbs, and a very considerable part of 

 the tail of a moderate-sized individual. Unfortunately the matrix is 

 an exceedingly soft shale, and the specimen has been considerably 

 weathered, especially in the thoracic region and the posterior part of 

 the tail. The pelvis and both feet are well preserved, and there is 

 the impression of a portion of the skull. 



Owing to crushing it is difficult to make very much of the 

 impression of the skull bones. They appear to me to represent 

 the right maxillary bone, parts of both premaxillaries, the nasals, 

 and in a much less satisfactory condition some of the bones in the 

 orbital region. The premaxillaries, as in many other reptiles with 

 long snouts, form a large part of the rostrum, probably as much as 

 two-thirds of the upper surface between the point of the snout and the 

 region of the orbits. Here they meet what are apparently the nasals. 

 The maxillary is long and appears to extend backwards below the 

 orbit. The nostril appears to be between the nasal, the maxillary, 

 and the premaxillary bones. The other bones are not sufficiently 

 well preserved to admit of description. 



The impressions of 16 presacral vertebrae can be counted, but 

 most of them are badly weathered. A few, however, show the 

 structure of the under surface fairly satisfactorily. So far as can be 

 made out, they agree closely with those of the Cape Town specimen. 

 The ridge for the articulation of the rib runs from the small 

 transverse process inwards and slightly forwards to the upper 

 and outer corner of the cupped articular surface. On the left side 

 16 presacral ribs are seen in almost natural position, and two others 

 much displaced. On the right side 15 ribs can be counted. A 

 small, very imperfect impression may be the inner part of the 

 right humerus, but this is very doubtful. In the case of the last 

 two presacral vertebrae the ribs are not attached as in the more 

 anterior vertebrae. In the last vertebra the short rib is attached to 

 the transverse process, and in the second last the rib appears to be 

 attached immediately below the transverse process. 



The sacrum consists, as in Stcreosternum, of two vertebrae, but 

 their structure is not very satisfactorily, shown. 



Of the pelvis, the pubis and ischium of each side are fairly well 



