332 Piof. D. M. S. Watson on 



deep and remarkably narrow, differing considerably from tlie 

 much more usual shape of Wiiaania and Alexia. 



General Discussion of the Coslacanth Skull. 



Although in general I accept Dr. Stensio's interpretation of 

 the Coelacanth skull, I am unable to concur in some of the 

 identitications he makes. 



A teim like supratemporo-extrascapularis implies tliat 

 there has been an actual fusion of bones, and that we should 

 expect to find that the bone so called develops from two 

 centres. That this is so there is no evidence in the case of 

 any bone which Stensio calls by a name of this type. 



The evidence on which he relies is simply that the bone in 

 question covers an area which in a more primitive form is 

 covered by two or more bones. 



If this mode of interpretation were carried to its logical 

 conclusion, one would have to call the parietal of such a 

 mammal as lemur a parieto-post-fronto-postorbital, because it 

 occupies an area which in a Gorgonopsid is filled by these 

 three independent bones. It is, however, quite certain from 

 its development that the mammalian parietd is single, and 

 we h;ive series of Anomodont reptiles which show a steady 

 reduction in size of the postfrontal until that bone is repre- 

 sented by an extremely minute strip partially se[)arating the 

 parietal and postorbital. 



In reptiles, at any rate, there can be no doubt that the 

 normal way in which the number of membrane-bones in the 

 skull is reduced is not by the fusion of neighbouring bones, 

 but by a gradual reduction in size and final loss of one of 

 them. 



It seems to me most probable that this method of loss is 

 that which usually occurs in bony vertebrates, and as a 

 general policy I have always of recent years regarded a bone 

 as single and corresponding to a single bone in an ancestor, 

 unless there is very good direct evidence that it is formed by 

 fusion of two or more bones. 



That such fusions do sometimes occur is certain ; in this 

 })aper 1 record a fusion taking place quite late in develop- 

 ment between the parietal and supratemporal of Macropoma 

 and the formation of the interparietal in mammals by the 

 fusion of a pair of bones is familiar to everyone. 



Thus, untd he adduces direct evidence that they are formed 

 by fusion, I am unable to accept any of Dr. Stensio's com- 

 pound names of bones. 



