ROSS' SEAL. 47 



from the normal number and arrangement of incisors in Ommatophoca, I think that, 

 however satisfactorily a seal may fall into a certain sub-family in this respect, one is 

 bound to ask first whether the number of incisors is a point of such great importance 

 as is thus implied, and secondly, whether a special taste in foodstuffs may not have 

 led to something like a reversion in the number as well as in the character of the teeth. 

 Presumably, the less differentiated types of seal have the greater number of incisors, 

 and so presumably the Cystophorince, having reduced them to a minimum, have 

 departed farthest in this respect from the ancestral type. But having done so, it 

 would be as easy at any rate, if not easier, for a seal to revert again to the greater 

 number of incisors than to make a still greater reduction as in the Cystophorince. 

 This seems to me to have been the case with Ommatophoca, and quite a considerable 

 number of points can be adduced in which it shows close affinities with Cystophora 

 and Macrorhinus, far more than can be adduced to connect it with the Steno- 

 rhinchince. 



For example, if we examine the post-canine series we find in Ommatophoca a 

 strong tendency to reduce their number, giving the following abundant variations : 

 p p. 6 6 p p 5 5 p p 6 6 p p 5 5 pn pn 5 5 pn 5 5 

 " 6-6' * 5^5' RC> 5=6' RC> 6 = 4' RC> = 0' ^ 6-6' RC ' 4^ 



In Machrorhinus, also a strong tendency to reduce, P.O. , P.O. - 



5 5 5 4 



3 3 5 5 



, P.C. . and almost the same extraordinary variability in their number, 



pointing to a similar functional worthlessness such as we have seen holds good in 

 Ommatophoca, while in Cystophora, though there is no tendency to reduce the number 

 of post-canines, the reason is probably to be found in the fact that its food consists not 

 wholly of Cephalopocls, but also to a considerable extent of some food which necessitates 

 a more useful set of molars for purposes of mastication. 



There is therefore upon consideration of the post-canine teeth, good reason for 

 thinking Ommatophoca to be more closely related to the Cystophorince than to the 

 Stenorhinchince, not a member of which shows any tendency whatever to vary from the 



g c 



normal type P.C. -, save Leptonychotes, in which the only aberrant examples dis- 



o o 



coverable in about forty skulls, have P.C. 6 , and P.C. C , showing, if anvthing, 



5 5 5 5 



not a tendency to reduce the cheek teeth, as in Ommatophoca, but to multiply them. 

 Both in the incisors and in the cheek teeth therefore the affinity of Ommatophoca to 

 the Cystophorince seems to be upheld. 



If we now turn to other points, and examine the skulls of the species under 

 consideration, we find quite as many and as important points in which Ommatophoca 

 resembles Cystophora and Macrorhinus, as we find points in which it resembles any 

 members of the Stenorhinchinse. For example, to quote from Sir William Turner, the 

 skull of Ommatophoca approaches that of Cystophora not only in the vertical inclina- 



