113 



whole (character #65). Doutt (1942) apparently describes a moderate case for Phoca 

 vitulina and Pusa hispida, where a large triangular notch changed the shape of the posterior 

 edge of the palate from a rounded Roman arch to a pointed Gothic arch. In such cases, 

 there is usually a slight inflection point between the arch and the notch to indicate the 

 two separate morphologies. However, in some specimens, the confluence was so complete 

 that there was no objective way to decide between a triangular posterior edge and a 

 combination of an arched posterior edge with a very large triangular notch. But, despite 

 Chapskii's (1055a) assertion that the palatal contours are generally subject to a fairly high 

 level of intraspecific variation, he still holds this character to be a useful feature for 

 subdividing the phocines. 



In keeping with the previous character, the lack of a notch is primitive among the caniforms 

 and is retained into the basal forms of both phocid subfamilies. A small notch is 

 convergently obtained in Histriophoca and the clade of Lobodon plus Monachus spp. Two 

 monk seals go on to derive larger notches - M. monachus (state 1) and M. schauinslandi 

 (state 2) - as does Mirounga angustirostris (state 2). Many species were polymorphic, 

 with the notch being equally present (i.e., one or more of the states 0, 1, or 2) and absent. 

 Yet, somewhat curiously, the "present" state was not necessarily for a small notch, as one 

 would expect if the notch was in the process of being gained, but often for a medium- 

 sized or greater one (as in Zalophus, Hydrurga, and Phoca vitulina). 

 69) relationship of bony nasal septum to posterior edge of palate: 0 = does not reach 

 posterior edge of palate; 1 = closely approaches / reaches posterior edge of palate (Chapskii 

 1955a; Ridgway 1972). 



Ridgway (1972) used this character to distinguish between the closely related genera 

 Histriophoca (state 1 - closely approaches) and Pagophilus [state 1 - reaches posterior 

 edge; also Burns & Fay (1970)]. Although useful at the level employed by Ridgway 

 (1972), the distinction between "closely approaching" and "actually reaching" seemed to 

 be fairly minor at the level employed in this study. As well, under such a coding scheme, 

 Burns & Fay (1970) observed that both Histriophoca and Phoca spp. would be 

 polymorphic for this character. One solution would be to code this character even more 

 finely using the sutures of the hard palate, especially the maxillo-palatine suture (Chapskii 

 1955a). However, this is often difficult to accurately determine in intact skulls, so we 

 instead propose a more stringent coding of Ridgway's (1972) character: either the bony 

 nasal septum extended posteriorly to approach the posterior edge of the palate, or it 

 distinctly did not. 



One exception to this dichotomy was observed fairly consistently in Mirounga 

 angustirostris. Here, state 1 was achieved through a combination of a slight posterior 

 extension of the bony nasal septum, a strong notching of the posterior end of the palate 

 (see character #67), and a dorsal arching of the palate in the midline to meet the nasal 

 septum. Together, these factors create a functionally shorter palate, allowing the otherwise 

 slightly elongated nasal septum to reach its posterior end. 



The primitive condition, where the bony nasal septum fails to reach the posterior end of 

 the palate, is shared by all the outgroups except Procyon. The apomorphic trait is typically 

 associated with the Monachinae (although two parallel reversals occur within the 



