360 PROF. OWEN ON EVIDENCES OF THEEIODONTS IN PERMIAN 



its superiority of size and length of crown to be regarded as a 

 " canine " (Cut, fig. 9, c). It is followed by five much smaller but 

 similarly shaped teeth, to be classed by position as "molars " (ib. ib.m). 

 In the premaxillary bone but one incisor has been preserved (ib. i) : 

 " its anterior border is obtuse and not crenulated, while the posterior 

 border is acute, but is too much broken to judge whether it is cre- 

 nulated ; the enamelled crown has been about 9 lines long by 4 lines 

 in breadth at its base. The fang of the tooth, like that of the other 

 ones, is oval in section " *. This incisor is separated by a diastema 

 an inch in extent from the canine. The length of the crown of 

 the tooth is 1 inch 9 lines, with a breadth of base of 7 lines. " Its 

 fang can be seen in the wide fissure of the jaw extending two inches 

 from the alveolar border " f . 



The foremost molar, with a visible crown 6 lines in length, pro- 

 trudes at a distance equal to its own antero-posterior breadth from 

 the canine. The crown in the succeeding molars, which are com- 

 pletely protruded, is 9 lines in length. Dr. Leidy describes a minute 

 crenation of the anterior as well as posterior trenchant borders of 

 the crown J. 



On the hypothesis that a portion of the upper jaw is here de- 

 scribed, I see in a semicircular notch more than an inch across the 

 base, at the upper anterior angle of the fossil, evidence of an external 

 nostril (Cut, fig. 9, n). Dr. Leidy, however, has described and figured 

 this fossil as " the right dental bone " of the lower jaw. The grounds 

 of the contrary homology, which I here submit, are afforded by an 

 almost complete skull of a Lycosaurus, from the Karoo series of S. 

 Africa (Cut, fig. 8). 



In this, as in other Theriodonts, a well-marked canine (ib. ib. c) 

 divides a series of small incisors (ib. ib. i) from a similar series of 

 small molars (ib. ib. m). The diastema between the canine and in- 

 cisors is rather longer than between the canine and the molars. The 

 crowns of all the teeth have the same general shape, even to the 

 " minutely crenulated " hinder border, as in Bathygnathus, Leidy ; 

 the incisors also show the distinction of the " more obtuse anterior 

 border." 



Both fossils are remarkable for the depth or vertical extent of the 

 upper jaw, 21 ; and, in both, the alveolar border of the maxillary de- 

 scribes a moderate convex curve. This seems to me to be fatal to 

 the mandibular hypothesis. I exhibit an impression of the plate 

 (lxviii.) of my ' Catalogue of S. African Eeptilia,' giving a side view 

 of the skull of Lycosaurus curvimola, also an impression of Leidy's 

 plate xxxiii. of his Bathygnathus borealis (reduced in Cuts 8 & 9), 

 from which may be better appreciated the comparisons leading 

 me to refer Bathygnathus to an allied genus of the Theriodont order 

 of Eeptilia. 



If the proposed homology and affinity be accepted, then the students 

 of the South- African fossils have a new interest in the determination of 



* Op. cii. p. 328. t lb. ib. _ 



\ This character is repeated in certain teeth of Cynodracojserridcns, ' Catal. 

 of S.-African Eeptilia,' 4to, 1876, pi. xvii. fig. 7. 



