90 



origins (DELTRAN optimization), or as a complex series of reversals (ACCTRAX 

 optimization). In this latter case, the straight morphology is a synapomorphy linking 

 Procyon through to the lutrines which is lost ancestrally in the pinnipeds, before being 

 regained several times therein. Again, the convex morphology was never consistently 

 present at the species level. State 9 was unique to Mirounga leonina. 



10) shape of dorsal portion of nasal processes of premaxilla along maxilla: 0 = concave; 

 1 = straight; 2 = convex; 9 = n/a - dorsal portion not present (Burns & Fay 1970). 

 Burns & Fay's (1970) initial coding for this group of characters (see character #8) points 

 to the dorsal portion of the nasal processes as perhaps being the key region of variance. 

 However, there does not appear to be a distinct pattern for the distribution of the various 

 states present. The plesiomorphic condition of a convex shape of the dorsal portion of the 

 nasal processes is generally retained throughout the Caniformia (and especially within the 

 monachines). with numerous independent derivations of the remaining apomorphic states. 

 (Again, the distribution of state 9 is more of a consequence of character #12.) Even within 

 the phocines, there does not appear to be a readily discernible pattern of shared derived 

 states: however, the distribution for the more generalized character used by Burns & Fay 

 (1970) is tentatively supported here (but only tentatively; many of the taxa in question are 

 polymorphic for this character and are differentially affected by the optimization criterion 

 used). 



11) contact between nasal processes of premaxilla and nasals: 0 = none: 1 = little (less 

 than width of nasal processes); 2 = broad (greater than or equal to width of nasal processes) 

 (Ridgway 1972: Wozencraft 1989). 



The plesiomorphic condition among carnivores is for a broad contact between the 

 premaxillar}' nasal processes and the nasals, with a reduction occurring in the phocids (de 

 Muizon & Hendey 1980: Wozencraft 1989: Wyss & Flynn 1993). Wozencraft ( 1989) also 

 indicates a reduced contact in lutrines. but Wyss & Flynn (1993) report that this is only 

 true of Enhydra. Among phocids. this reduction is generally partial, if any. for the 

 phocines, and full (or virtually so) for the monachines. but with Monachus spp. re- 

 obtaining the primitive condition (Mivart 1885; Hendey & Repenning 1972: Ridgway 

 1972; de Muizon & Hendey 1980). A total lack of contact has been held to be diagnostic 

 for the lobodontines (de Muizon & Hendey 1980). Additionally, Phoca vitulina may be 

 naturally dimorphic for this character, with state 0 characterizing Atlantic forms, and state 

 1 or 2 being typical of Pacific individuals (Allen 1902: Doutt 1942: Chapskii 1955a. 1967). 

 Here, reduced contact is diagnostic of the phocids alone, with the family primitively 

 characterized by state 0. This extreme situation for the hypothetical phocid ancestor could 

 be an artifact associated with character #12. as two of the genera lacking the dorsal portion 

 of the nasal processes (Cystophora and Mirounga) hold the basal positions within each 

 phocid subfamily. However, it is likely an accurate portrayal, as the monachines generally 

 retain a state of no contact between the nasals and the premaxillary nasal processes. The 

 distribution reported above is generally upheld here. Among the monachines. the 

 lobodontines. exclusive of Leptonychotes (state 2). are characterized by state 0. with 

 Monachus spp. reverting to re-obtain a broad contact (although M. monachus only does 

 so under ACCTRAN optimization). The phocines internal to Halichoerus likewise revert 



