146 



tropicalis), Phoca vitulina, and possibly Leptonychotes. Again, the absence of this trait in 

 other species with a juvenile skull type further weakens any supposed correlation between 

 skull type and postcanine crowding. 



148) obliqueness of postcanine implantation (upper and lower) relative to vertical: 0 = 

 straight; 1 = slanted (de Muizon 1982a; pers. obs.). 



In addition to angling away from the axis of the tooth row, as in the previous character, 

 the postcanines are also occasionally slanted, typically along the lingual-labial axis. 

 Chapskii (1955a) states that this is virtually universal in Phoca vitulina. This condition is 

 likely not associated with postcanine crowding and may, in fact, be enhanced when the 

 postcanines are widely separate, due to their reduced association with each other. However, 

 this apomorphic condition is restricted in distribution to convergent appearances in 

 Lobodon (lower postcanines #1 and #2 slanted lingually) and Mirounga angustirostris 

 (various postcanines, usually slanted labially). 



149) curvature of upper tooth row (postcanines only): 0 = sigmoidal; 1 = arched; 2 = 

 straight; 3 = kinked between PC 12 , otherwise straight; 4 = reverse arch (Ridgway 1972). 



A character dealing with the curvature of the upper tooth row was used by Ridgway (1972) 

 to distinguish between the genera Histriophoca and Pagophilus. Histriophoca is often 

 noted for its strongly curved upper tooth row (Scheffer 1958; Burns & Fay 1970), which 

 occasionally approaches a lyrate (= sigmoidal) morphology (Burns & Fay 1970). However, 

 Burns & Fay (1970) do caution that this extreme curvature is not consistent within 

 Histriophoca, and that it also occurs within other members of the Phocini, albeit to a lesser 

 extent. Howell (1928) describes an apparently straight tooth row in Zalophus. 

 Our observations revealed that a presence versus absence coding of upper tooth row 

 curvature is too simplistic and does not accurately represent the range of variation present 

 within the phocids. The major recurring patterns were ones where the tooth row arched 

 laterally (state 1), medially (state 4), or laterally anteriorly and medially posteriorly (state 

 0). The kinked condition (state 3) is obviously an extreme case of one of the other states, 

 but exactly which one is not clear a priori. Thus, this morphology was left as a distinct 

 state. 



The plesiomorphic condition is of a sigmoidal upper tooth row, which is possessed by the 

 majority of the non-lutrine fissiped outgroups. The derivation of this character in the 

 lutrines and ancestrally within the pinnipeds is unclear. Either a straight tooth row is 

 synapomorphic for Lutra plus the pinnipeds (ACCTRAN optimization), or the ancestral 

 condition for the pinnipeds is equivocal between states 2 and 4, an uncertainty which is 

 preserved into each phocid subfamily (DELTRAN optimization). In any case, the majority 

 of the pinnipeds possess the straight morphology, and, despite its possible ancestral status, 

 a reverse arch is limited to Cystophora, Mirounga spp., and Monachus monachus (where 

 it must appear independently). Odobenus obtains a laterally arched tooth row, as does 

 Enhydra among the fissipeds. Erignathus uniquely reverses to the plesiomorphic sigmoidal 

 tooth row. The kinked morphology was found only in Monachus schauinslandi plus M. 

 tropicalis, and appears to be a derivative of the straight condition. 



