326 Mr. D. M. S. Watson on 



14. The reduced number of phalanges in the hand and 



foot : 2.3.3.3.3. (Broom.) 

 35. The presence of a sacrum of more than two vertebrae. 



(Owen, Seeley.) 



16. The structure of the pelvis. (Owen, Seeley, Broom.) 



17. The structure of the tarsus. (Broom.) 



To these I now add : — 



18. The small fenestra vestibuli. 



19. The ventral position of the inner ear seen in Theromus. 



20. The presence of an extensive series of canals on the 



side of the brain-case, only comparable with those 

 which lodge a venous system in Insectivores. 



This extraordinary series of resemblances, many of them 

 of a particularly deep-seated character, can only mean that 

 the Therapsids are to be associated with the ancestry of the 

 Mammals. It is none the less improbable that the Cyuodonts 

 themselves are closely connected with the ancestral mammal. 

 They appear to be an examp'e of a group which, whilst 

 retaining a very primitive structure in some of the more 

 deep-seated regions, such as the brain-case and brain, have 

 developed specializations which are prophetic of the develop- 

 ments which will be much more slowly but securely carried 

 out by the descendants of a more conservative line of the 

 same stock. 



It is impossible with our present very slight knowledge of 

 Therapsids to decide which of the many resemblances to 

 Mammals catalogued above have descended to the Cynodouts 

 and Mammals from a common ancestor, and which are due 

 to that similarity in the course of evolution in two allied 

 stocks which is now being gradually realized as a most 

 important factor to be considered in the study of phylogeny 

 and the laws of evolution. 



The Homologies of the Zygomatic Arch of the Therapsids. 



The zygoma of Cynodonts and Therocephalians is formed 

 by the squamosal, postorbital, and jugal. As it contains a 

 squamoso-postorbital part, at any rate in the more primitive 

 types, it seems natural to regard it as representing the two 

 arches of Diapsids fused, the quad rato- jugal being lost. 



Dr. Broom has, however, made the remarkable suggestion 

 that it really corresponds to the lower arch, the single 

 temporal fossa answering to the lower fossa of Sphenodon, 

 and implying that the upper arch has fused on to the side of 



