80 MR. D. M. S. WATSON ON THE 



IVhaitsko is a I'euiarkable form with a somewhat primitive 

 basicranial region, and a palate which in general agrees with 

 Gorgonojjs, but differs in the development of a special process 

 passing out on each side of the internarial bar so as to divide 

 each posterior nostril into two. [The meaning, morphology, and 

 function of this arrangement are quite uncertain, the anterior 

 vacuities are not exactly homologous with the anterior palatine 

 incisions of mammals and Cynodonts, because the posterior 

 border of these incisions is always formed by the anteiior edge 

 of the secondary phite of the maxilla.] Whaitsia is veiy advanced 

 in the reduction of the wide parietal region to a narrow sagittal 

 crest, which characterizes it, and in the extreme reduction of the 

 dentition. 



The preceding discussion shows that the Gorgonopsids include 

 a series of forms which exhibit in their skulls a gradual series of 

 changes by which so primitive an animal as Arctops passes in- 

 sensibly into a Cynognathid. It establishes clearly the existence 

 of a series of evolutionary trends, wliich pei'sist without change 

 from the beginning of the Anomodonts in Varanosaurus to their 

 end in Diademodon, and indeed to lead on to mammals. It 

 remains to discuss the other primitive Theriodonts included in 

 Broom's order Therocephalia and the Deinocephalia, to see how 

 far these evolutionary trends apply also to them, and to consider 

 the relation of these forms to the Gorgonopsids, which are 

 plainly the central group of the Theriodonts. 



No Therocephalian is at all well known, despite the very large 

 number of forms which have been described. We know the 

 dorsal and lateral surfaces of the skull in a good many foriBS 

 (^Scylacosaurus, Lycosaui-us^ Sccdoposcmras, etc.), the palate is 

 known more or less completely in others [Scylacosaitnts, Scym- 

 nosaurus, Scaloposaarus, etc.). The basici-anial region is known 

 in no Therocephalian, neither has any occiput been figured. 

 Haughton has described the brain-case of Alopecognathus, but 

 his figure is not in all points (e.^., the character of the supra- 

 occipital and the relations of the interparietal and parietal) very 

 convincing. 



The most important materials of Therocephalia in the British 

 Museum aie the more or less complete skulls of Sccdoj^osaurus 

 from the Oistecephalns-'/jOue and Hcylocosaurus and Scymno- 

 scmrus loatsoni from the Tajyinocephcdus-zowQ. 



ScYMNOSAURUS WATSONI Broom, Proc. Zool. iSoc. 1915, p. 169, 

 fig. 6. 



Lycosuchiisl Watson, Proc. Zool. Soc. 1914, p. 1U36, fig. 7, 



Type : a, skull with seven cervical vertel)rpe in natural articu- 

 lation, other vertebra^ and fragmentary limbs doubtfully 

 associated. Tapinocephcd'iis-'Aone, TJitkyk, Dist. Prince Albert, 

 Cape Province. 



The skull of the type is curiously preserved: it is embedded in 

 a calcareous nodule, which breaks with a conchuidal fracture and 



