162 BB. E. BEOOM Olf THE STETJCTUEE AND [June 4, 



1° 21' S. lat, except for the depth of the bod}^ which is 4^ to 4|- 

 times in the total length, the maxillary not extending quite to below 

 the anterior border of the eye, the diameter of which is 4|- times 

 in the length of the head. 9 gill-rakers on lower part of anterior 

 arch. D. XV- XVI 9; A. Ill 8-9; Sq. 30-32 ^; Lat. 1. 20-21/ 

 13-14.' 6 or 7 scales between the first dorsal spine and the lateral 

 line. 



Smaller specimens (85-95 millim.) differ in the smaller head, 

 the larger eye (3^4 times in length of head), and the lower jaw 

 not projecting beyond the upper. The dark longitudinal bands 

 are very indistinct and are traversed by 7 or 8 ill-defined dark 

 cross-bars. 



In all the specimens the ventral fins are of a bright yellow. 



3. On the Structure and Affinities of Udenodon. 

 By R. Broom, M.D., B.Sc' 



[Eeceivecl May 21, 1901.] 



(Plates XVI.-XVIII.) ^ 



(Text-figures 10 & 11.) 



A considerable number of skulls of Udenodon and of the closely 

 allied genus Dicynodon have long been known, and there have also 

 been found many other bones of the skeletons ; but as in almost all 

 the specimens the association of the skull and other bones has 

 been quite lost, it is at present impossible to refer limb-bones to 

 their proper species of which the skulls are the types, and it is 

 only with some doubt that they can be referred even to their 

 proper genei*a. In a few cases some bones of the skeleton have 

 been found in association with Dicynodon-sknWs, but in the case 

 of Udenodon the post-cranial skeleton is quite unknown. 



The most important specimen in which the Dicynodont skull is 

 in association with a considerable portion of this is the little form 

 which has been described by Seeley (1) as " Keirognathus cordylus." 

 In this specimen the skull, upper vertebrae and ribs, front limbs, 

 shoulder-girdle, aud sternum are shown, but all in a very bad 

 state of preservation. There is scarcely a doubt that the skeleton is 

 that of a young Dicynodon, and it is specially valuable as showing 

 the relations of the shoulder-girdle, sternum, and interclavicle. 

 Seeley's restoration is unsatisfactory. 



In the Lower Karroo beds of Pearston, S. Africa, while the 

 remains of various species of Bicynodon are met with, the genus 

 which most commonly occurs is Udenodon and from the specimens 

 which I have recently discovered I am now in a position to give 

 an almost complete account of its skeleton. 



1 Communicated by Prof. G. B. Howes, LL.D., F.E.S., F.Z.S. 

 - * For an explanation of the Plates, see p. 190. 



