172 DR. R. BROOM ON THE STRUCTURE AND [June 4, 



anterior angle is notched, probably for the lodgment of the inter- 

 claviele; the outrr angles are rounded and the posterior angle 

 truncated as if to give attachment to a cartilaginous meso- or 

 xiphisternum. For its lower two-thirds there is a well-marked 

 median ridge for the attachment of the pectoral muscles. The 

 bone is a little longer than broad. 



Humerus. 



In the little skeleton of Udenodon gracilis (PL XVI.) one of the 

 humeri (hu.) is perfectly preserved and beautifully displayed ; the 

 other though less perfect has the opposite side showing. In speci- 

 men 4 one humerus is fairly well preserved, and of the other the 

 lower half is almost perfect ; while in specimen 5 one humerus is 

 shown, but in bad condition. The difference between the humeri 

 in U gracilis and U. bairn is very striking, and much greater than 

 one could have expected to find in two species apparently so closely 

 allied. 



In Udenodon gracilis (PI. XVI.), the humerus, while constructed 

 on the well-known Dicynodont type, is characterized by a number 

 of peculiarities. The delto-pectoral crest (c.d.) is greatly developed, 

 its border curving forwards and downwards from the articular 

 surface and ending, as in " Platypodosaurus robustus" in a somewhat 

 hooked process. The ento-tuberosity is developed to a greater 

 extent than is met with in the humerus of any S. African reptile 

 hitherto discovered, and forms a long flattened tapering process 

 which ends in a rather sharp point. Near the middle of the bone on 

 its inner side is developed a very prominent tricipital ridge resem- 

 bling more that seen in Echidna than the rounded prominence on the 

 humerus of "Platypodosaurus." The articular surface of the head 

 of the bone resembles greatly that in Omithorhynchus and Echidna 

 in being considerably developed transversely, while it is but very 

 narrow. As in the Monotremes, a sharp ridge runs down from 

 the articular head to the external condyle ; and this is very 

 peculiarly developed, in that while in the humeri hitherto dis- 

 covered though it may be very prominent it is generally slender, it 

 is here a markedly rounded boss (c.e.). The internal condyle is 

 not very large. The entepicondylar foramen is only of moderate 

 size, and is situated a little more distally than is usually the case 

 in Dicynodonts. The ridge of bone which forms the bridge 

 over the foramen runs up to the base of the delto-pectoral crest. 



In Udenoilon baini the humerus varies much less from the 

 normal dicynodont type. The delto-pectoral crest is not developed 

 downwards to form a hooked process in front ; the tricipital ridge 

 appears to have been small ; while the external condyle is not more 

 greatly developed than is the case in Omithorhymhus. 



Radius and Ulna. 



The radius and ulna are very mammal-like in form (cf. Plate 

 XVI.). 



Tiie radius (rd.) is considerably shorter than the humerus. Its 



