322 Mr. D. M. S. Watson on 



to the rim to which the membrane is attached in Chehjs. 

 As, however, it is continuous with a wide groove on the side 

 of tlie zygomatic arch in Cynoynathus and is of a curious 

 sliape, so that any membrane attached to it would be twisted 

 in two directions and wouhl not have a developable surface, 

 it can scarcely be the actual point of attachment of the 

 membrane, whose situation must be sought at the inner end 

 of the groove. 



Here there is a small roughened ridge lying just external 

 to the paroccipital process, which could well have given it 

 attachment. 



The presence of a membrane in this position is consistent 

 with the very small fenestra ovalis, which implies a small 

 stapes. 



If the positions of the digastric muscle and membrana 

 tympani here adopted be correct, the smooth groove on the 

 back of the zygoma can only be an external auditory meatus, 

 a suggestion of its function which has already been tenta- 

 tively made by Gregory. 



It is now of importance to compare the ear we have thus 

 arrived at with that of a mammal. 



We have seen that the tympanic cavity of a Cynodont is 

 bounded b}'^ the paroccipital process of the periotic above and 

 behind, that in front it is covered by the posterior ramus of 

 the pterygoid, and that the squamosal may probably enter 

 very slightly into its roof externally. 



The tympanic cavity in a marsupial is bounded by the 

 periotic above, and behind by the glenoid ramus of the 

 alisphcnoid, which usually forms a bulla in front, and by the 

 squamosal externally above. 



It is seen that except for the replacement of the i)ostcrior 

 ramus of the pterygoid by the glenoid part of the alisphcnoid, 

 these bones are tlie same, and this suggests a comparison of 

 the pterygoid of Reptiles with the alisphcnoid of Alammals. 



The pterygoid in a Cynodont has the following relations : — 



1. It articulates with the basisphcnoid. 



2. It has a large ascending process, forming part of the 

 wall of the brain-cavity and articulating with the parietal. 

 This separates the exits of the II., 111., 1\',, \., and W. 

 nerves from those of the second and third branches of the V. 



3. It has an anterior ramus articidating with the palatine, 

 the cctopterygoid, which lies entirely ventral to it, and the 

 jugal. Internally it is separated from its fellow by the vomor 

 anteriorly and the basisphcnoid posteriorly. 



