123 



petromastoid ridge is, like the paroccipital processes (see character #85), pressed against 

 the auditory bulla. For our purposes, we required the petromastoid ridge to be separate 

 from the auditory bulla, although not necessarily a prominent, obvious structure. 

 Our analysis supports Wozencraft (1989), with an apomorphic petromastoid ridge 

 occurring only in Ursus and the otarioids, although this is due here to parallel evolution 

 rather than uniting the two groups as a synapomorphy. However, it should be noted that 

 the acceptance of the analogous structure in Canis and Procyon as a proper petromastoid 

 ridge supports a scenario in keeping with that of de Muizon (1982b), with the loss of the 

 ridge being a synapomorphy linking the mustelines, lutrines, and the pinnipeds as a whole, 

 with the otarioids reversing to re-obtain the plesiomorphic condition. 

 *90) source of "paroccipital" process: 0 = occipital; 1 = occipital and mastoid; 2 = mastoid 

 (Burns & Fay 1970). 



This character reflects the confusion created by having numerous synonyms for a given 

 structure present in the literature. Burns & Fay (1970:375) are entirely correct in saying 

 that the paroccipital processes of phocids should more properly be referred to as the 

 paramastoid processes, as they are "of the occipital and near the mastoid." However, this 

 is also true for all caniforms we have so far examined. It would appear that this process 

 has been historically misnamed (e.g., Turner 1848; Flower 1869; Mivart 1885; Howell 

 1928; Davis 1964; DeBlase & Martin 1981). Perhaps the least confusing alternative would 

 be to use the non-origin specific synonym "jugular process" (e.g., Miller 1962; Crouch 

 1969). However, Burns & Fay (1970) suggest that the use of this term be restricted to 

 those instances when the paroccipital and paramastoid processes are confluent or 

 connected by a petromastoid ridge (see character #89). Other than being an unnecessary 

 restriction, this suggestion confuses matters further as it is unclear exactly what the term 

 "paroccipital process" refers to in such a definition (but likely the mastoid process). An 

 overriding complicating factor in all this is that the processes in this general region appear 

 to be distinguished on the basis of which muscles originate from them, rather than on their 

 specific bone of origin. As the paroccipital process serves at least as a partial origin for 

 the digastric muscle (Howell 1928; Miller 1962; Davis 1964; Crouch 1969; Bryden 1971; 

 King 1972), perhaps in one of the early descriptions of this muscle, its process of origin 

 did, in fact, originate on the mastoid, but near the occipital. This character was excluded 

 because of all of this confusion. 



91) morphology of paroccipital processes: 0 = absent; 1 = elongated ridges; 2 = bumps / 

 pillars (pers. obs.). 



Flower (1869) comments that the general form of the arctoid paroccipital process is one 

 of a roughly triangular bony prominence projecting posterolaterally from the skull which 

 is generally separate from the auditory bulla (but see character #85). Yet, the morphology 

 of the paroccipital process, and its relationships with other structures of the basicranial 

 region, does appear to possess phylogenetically useful variation within the arctoids. Some 

 of this variation is summarized in this and the following three characters. 

 One immediate observation of the general form of the paroccipital processes (other than 

 simply their presence versus absence) is that they are laterally compressed in some taxa 

 to take on the form of elongated ridges. Among the fissiped caniforms, Turner (1848) 



