1901.] AFFINITIES OP tTDBNODON. 183 



homologue in the mammal in the interarticular cartilage of the 

 lower jaw. If recent advances in palaeontology have not fully 

 confirmed this view, they have at least shown that practically all 

 the alternative theories are untenable. 



In the higher Thecodonts, e. g. GynognatTvus, the zygomatic 

 arch, though composed of the same elements as in Udenodon, 

 differs very greatly in the relative proportions of the parts ; but 

 in the lower Theriodonts, i. e. Ictidosuchus (11), the zygomatic 

 arch bears much resemblance to that in the Anomodonts. 



The palate in Dicynodon and Udenodon differs considerably from 

 the normal reptilian type, and agrees essentially with that in the 

 Theriodonts and Mammals. 



All known Reptiles have palates more or less modified from a 

 primitive type such as is found in Pariotichus or Procolophon, and 

 which consists of the following elements : — an anterior pair of 

 bones wbich meet in the middle line, aud which are usually 

 regarded as " vomers " ; a pair of large pterygoids, which have 

 each a posterior and outer branch to the quadrate, an outer and 

 anterior branch joining the transpalatine, and an anterior and 

 inner branch which usually meets the so-called " vomer '* ; a pair 

 of palatines lying on the outer side of the anterior and inner 

 braucb.es of the pterygoids ; a pair of transpalatines ; and a 

 median element passing forwards from the basisphenoid between 

 the two pterygoids, and usually referred to as the " parasphenoid." 

 This arrangement we find with only slight modifications in all 

 the early groups, including types so dissimilar as Pareiasaurus, 

 Procolophon, Ichthyosaurus, Plesiosaurus, Dimetrodon, aud /S^/te- 

 nodon. 



The palates of the Theriodonts and Anomodonts are very con- 

 siderably modified derivatives of the same type. Unfortunately 

 in none of the primitive Theriodonts is the palate at all well 

 known. We know, however, that a secondary palate was not 

 formed in either Qorgonops or Ictidosuchus, nor apparently in 

 JSlurosaurus. In Udenodon we have an imperfect secondary 

 palate, and in the higher Theriodonts a secondary palate as well- 

 developed as that in mammals, and essentially similar to that in 

 mammals. As the general structure of the skeleton of Udenodon 

 is essentially Theriodont, and as the palate only differs from that 

 of the typical Theriodont in the specialization of the beak and in 

 the secondary palate being imperfect, one is justified in concluding 

 that the Anomodonts are a specialized offshoot from the earliei 

 Theriodonts. In the evolution of the Theriodont palate from the 

 primitive Reptilian type, the changes which have taken place 

 appear to have been the following: — The anterior and internal 

 branch of the pterygoid has become greatly reduced and finally 



lost, and its place and function to a large extent has been taken 

 up by a greal increase, in the so-called " parasphenoid " ; while the 

 anterior and outer branch has become more developed, and with 



its increase the palatine has conic to lie rather internal than 



external to the pterygoid. With the developmenl of a secondary 



