Matnuialia. 9 



views of relationship and descent, it becomes, I think, impossible to 

 refuse recognition to a character such as that presented by the 

 number of the teeth, until it can be shown to be useless or misleading. 

 Thus, although I am aware that, by doing so, I am forming a kind 

 of hotch-potch, in bringing together in the case of the Stenorhyncli- 

 inae some, in other respects, highly distinct animals, I find myself 

 unable to depart from the arrangement of sub-families already laid 

 down by previous writers. 



Phocinae. Stenorhyxchinae. Cystophorisae. 



Phoca (including all north- Mouachus^, Cystophoray 



ern sub-genera, such as Leptonycliotea, Macrorhinus. 



Erignatus and Histri- Ogmorhimis, 



cophoca), J.ohodotty 



IlaJicJuerus. Ot/imafophoca. 



Of these Sub-families, one, viz., of the Cystopliorinae, is quite 

 satisfactory. Further, I have no doubt about the propriety of 

 including, in a distinct group, all the species represented by Phoca 

 in its broad sense. Halichmrus I regard as the most distinct form 

 of northern Seal. Yet, except in its teeth, it shows no characters not 

 easily attributable to its isolated existence, and, if we are to make 

 sub-families for every Seal whose teeth show remarkable features, 

 we shall find ourselves instituting almost as many Sub-families as 

 there are species at present. In the strength and shape of its teeth 

 Halichoerus approaches most Dearly to Monachus, but, in the bulbous 

 character of their roots, it resembles to a certain extent Cydophorct 

 and Macrorhinus. 



The greatest difficulties beset us when we come to consider the 

 Stenorhynchinae, a sub-family in which every genus possesses teeth 

 which are at first sight unparalleled elsewhere. I think, however, 

 that these dental characters must be regarded as instances of high 

 specialisation to suit the particular needs of each species. They can 

 afford no aid in the search for relationship. We are, therefore, I 

 think, justified in neglecting them for the present purpose, and we 

 may turn to more generalised characters, such as the appearance 

 and structure of the skull. Apart from the question of dentition, 

 there is a very close general resemblance between Monachus and 

 Lcptoiiychotes. At the other end of the group and nearer the Gysto- 

 Ijhorinac lies Ommatophoca, but even this, the most aberrant member 

 of the sub-family, is connected by the form of its nasals with the 

 remainder through Lobodon, in which the peculiar formation of the 

 pre-maxillae and the inclination of the external nares is paralleled, 

 although to a lesser degree. Lohudun, again, is related to Oymo- 



