151 



As indicated by Wyss (1988a), the monachine pattern (state 0) represents the plesiomorphic 

 condition, and this extends back to the level of the Caniformia. The phocine pattern, 

 whereby the deltopectoral crest ends abruptly at a virtual right angle to the shaft, is a 

 synapomorphy uniting all phocines, with a parallel appearance in Leptonychotes. Again, 

 a distinct crest was uniquely absent in Martes. 



161) entepicondylar foramen of humerus: 0 = absent; 1 = present (King 1966; Wyss 

 1988a). 



One obvious distinguishing characteristic between phocines and monachines is the 

 presence of an entepicondylar foramen in the former and its absence in the latter (King 

 1966; Wyss 1988a). However, it should be noted that this is a generalization applying 

 primarily to extant forms. Early Pliocene monachines (e.g., Homiphoca capensis, various 

 species of Monotherium) do possess an entepicondylar foramen (Hendey & Repenning 

 1972; Ray 1976b; de Muizon & Hendey 1980; de Muizon 1982a; Repenning 1990). 

 Various authors take this to be evidence of a recent loss of the foramen by the monachines 

 (Hendey & Repenning 1972; de Muizon & Hendey 1980; Repenning 1990). As well, we 

 observed phocine specimens in which the foramen was unilaterally (Pagophilus, AMNH 

 180016) or bilaterally absent (Halichoerus, USNM 446408), something also infrequently 

 observed in other phocines (King 1966). A polymorphic distribution for this character is 

 also indicated for the Ursidae, where the foramen is generally absent, except for the genera 

 Ailuropoda and Tremarctos (Davis 1964). 



As mentioned by Wyss (1988a), the interpretation of the distribution of this character 

 depends on the outgroup relationships assumed for the phocids. A diphyletic pinniped 

 origin with lutrine affinities for the phocids, as advanced by de Muizon (1982a), yields 

 the monachine pattern as being apomorphic. A monophyletic Pinnipedia with ursid 

 affinities, as postulated by Wyss (1987, 1988a), instead holds the phocine pattern to be 

 apomorphic. Our analysis indicates that the polarity of this character is equivocal at the 

 level of the Pinnipedia. At the level of the Caniformia, however, the possession of an 

 entepicondylar foramen is apomorphic, being found in Procyon, Martes, Enhydra, and 

 Lutra, in addition to all phocines. The connection between these fissipeds and the phocines 

 is dependent on the optimization criterion used. With ACCTRAN optimization, the lack 

 of the foramen is a synapomorphy of the pinnipeds, with the phocines homoplastically re- 

 deriving the foramen. Under DELTRAN optimization, the possession of the foramen is 

 synapomorphic for these taxa, with the otarioids and monachines independently reversing 

 to the plesiomorphic condition. Only this latter pattern can account for the proposal 

 whereby the foramen is present in monachines primitively before being lost (see above). 



162) distally projecting ledge (palmar process) on cuneiform of carpus: 0 = absent; 1 = 

 present (King 1966; Wyss 1988a). 



King (1966) used this character to group Cystophora with the phocines (state 1) and 

 Mirounga spp. with the monachines (state 0). In noting the absence of the ridge in 

 otarioids, Wyss (1988a) presumed the presence of the ridge to be a phocine synapomorphy 

 at the level of the Pinnipedia. However, such an interpretation is upheld here only under 

 ACCTRAN optimization. Under DELTRAN optimization, the otarioids and monachines 

 independently lose the ridge to match the state found in Canis. Otherwise, the palmar 



