93 



As well, our observations reveal that Odobenus does possess a very subtle divergent 

 morphology, partially disguised by a rounding of the posterior edge of the nasals (see 

 character #17). In any case, the convergent morphology has been repeatedly noted for 

 phocids (Burns & Fay 1970; King 1983: Wozencraft 1989; Wyss & Flynn 1993). 

 The divergent morphology is present only in a number of the outgroup taxa: Ursus, the 

 otarioids, and polymorphically in Lutra. Although the phocids share the plesiomorphic 

 convergent morphology with Cams, they differ subtly in that the nasals intrude far more 

 deeply between the frontals than in Canis, extending posteriorly past the anterior orbital 

 rim. This represents yet another phocid synapomorphy (Wyss & Flynn 1993). 

 17) shape of posterior edge of nasals, H: 0 = pointed; 1 = rounded (Wozencraft 1989; 

 pers. obs.). 



Inclusion of this character follows from observations of the previous character dealing 

 with the nasofrontal suture. Seemingly independent from the nature of the nasofrontal 

 suture was the observation that the nasals either terminated in sharp points or were 

 rounded. A rounded termination is plesiomorphic, characterizing most outgroup taxa except 

 Martes and Zalophus. Most phocids display the apomorphic pointed morphology, but this 

 condition arises independently within each subfamily: internal to Cystophora in the 

 phocines, and in Mirounga leonina and internal to Hydrurga in the monachines. Among 

 these more terminal taxa, only Erignathus and Monachus schauinslandi deviate from the 

 typical pattern of their respective subfamilies. 



*18) shape of posterior edge of nasals, HE: 0 = pointed v-shape; 1 = rounded v-shape; 2 

 = rounded w-shape; 3 = pointed w-shape (Wozencraft 1989; pers. obs.). 

 This character is a combination of the previous two characters dealing with the shape of 

 the nasals. However, it is inferior to the previous two in that it is too particular and thus 

 reduces the number of potential synapomorphies in favour of autapomorphies (e.g., only 

 Zalophus obtains state 3). It was, therefore, abandoned. 



19) distinct caninus fossa: 0 = absent; 1 = present (de Muizon 1982a) (Fig. 19). 

 A well-marked fossa running anteriorly along the alveolar edge of the maxilla from below 

 the infraorbital foramen has been described as a synapomorphy of the phocines, with a 

 convergent appearance in Mirounga spp. (de Muizon 1982a). However, there appears to 

 be some confusion centred around the muscle this caninus fossa receives. De Muizon 

 [pers. comm., citing Howell (1928)] states that the caninus muscle is lacking in phocids. 

 so that the fossa serves instead as the origin for the maxillo-naso-labialis. However, such 

 an interpretation is not at all clear from Howell (1928), which, together with other sources 

 (e.g.. Miller 1962; Crouch 1969; Bryden 1971), would seem to indicate that the two 

 muscles are merely synonyms for one another. Only Pierard (1971) indicates the two 

 muscles to be distinct entities, thereby voicing an opinion in agreement with de Muizon. 

 Although de Muizon (pers. comm.) indicates that maxillo-naso-labialis fossa would be a 

 more appropriate name for this structure in phocids, we will continue to use the term 

 caninus fossa. 



Among outgroup taxa, a distinct fossa was only present for Canis rendering the polarity 

 of this character equivocal at the level of the Caniformia. Although the distribution of this 

 character closely matches that of de Muizon (1982a) (we additionally noted the presence 



