155 



former medial side of the ilium, as well as the gluteus group extending to the femur 

 (Howell 1928; Bryden 1971: Hendey & Repenning 1972). McLaren (1975) holds that the 

 subfamilial differences arise from the retention of a more primitive muscular arrangement 

 in the monachines, a contention echoed by Hendey & Repenning (1972). The binary 

 coding employed by most authors ("strongly everted or not") disguises the unique form 

 of the ilium in all phocids by grouping the monachines with non-phocids. Therefore, we 

 have subdivided this character more finely, with the categories roughly corresponding to 

 "not everted", ^weakly everted", and '"strongly everted". 



The apomorphic eversion of the ilial wing (states 1 or 2) is largely confined to the phocids. 

 although a weakly everted ilium does occur in Enhydra and Ursus. As the phocids are 

 characterized ancestrally by state 0. independent eversions of the ilial wing must occur in 

 each subfamily, as implied by Hendey & Repenning (1972). The weakly everted ilium 

 typical of many monachines describes a synapomorphy of those species internal to 

 Hydrurga, with Monachus schauinslandi (possibly as a synapomorphy with Monachus 

 tropicalis; ACCTRAN optimization) reverting to a flat ilium. The strongly everted ilium 

 of the phocines arises ancestrally within this subfamily, with only Erignathus (states 0 and 

 1) largely reversing to the plesiomorphic caniform morphology. 



*171) gluteal fossa on wing of ilium: 0 = absent: 1 = present (King 1966; Wyss 1988a). 

 With recoding, this character was included in character #172. 



172) depth of gluteal fossa on ilium: 0 = shallow; 1 = medium; 2 = deep; 9 = absent 

 (King 1966; Wyss 1988a). 



Apparently associated with the strongly everted phocine (exclusive of Erignathus) pelvis 

 is a deep, compact gluteal fossa on the lateral side of the everted ilial wing (King 1966; 

 Wyss 1988a). The majority of this fossa in phocines serves as the origin for the gluteus 

 medius muscle (Howell 1928). The gluteal fossa is apparently absent in the remaining 

 pinnipeds (Wyss 1988a), but exists in Cards and Ursus as a shallow, elongated trough 

 (Miller 1962; Davis 1964). 



A gluteal fossa of shallow or medium depth is found in all fissiped outgroups except 

 possibly Lutra (ACCTRAN optimization). A deep fossa unites the phocines ancestrally 

 with only Erignathus (state 9) and Pusa sibirica (state 1 - possibly as a synapomorphy 

 with Pusa hispida; ACCTRAN optimization) deviating from this trend. Beyond this, 

 however, the two optimization criteria used here provide strikingly different pathways for 

 the evolution of this character. ACCTRAN optimization holds for the loss of the fossa 

 uniting Lutra plus the pinnipeds with Leptonychotes, Lobodon, and Mirounga spp. 

 independently deriving medium depth fossae. In contrast, DELTRAN optimization 

 indicates that a shallow fossa is a synapomorphy of Lutra plus the pinnipeds, with the 

 otarioids, Hydrurga, Monachus spp. and Ommatophoca losing the fossa in parallel. Of 

 these two opposing hypotheses, the latter seems the more likely, as the previous (excluded) 

 character indicates that the loss of the fossa is an apomorphic tendency occurring 

 independently in the taxa mentioned above. 



173) relationship of obturator nerve foramen to obturator foramen: 0 = distinctly separate, 

 at least unilaterally; 1 = intermediate - foramina confluent, but individually recognizable; 

 2 = confluent - obturator nerve foramen not apparent (Wyss 1988a). 



