97 



This condition is retained for the otarioids (but see character #26 for Odobenus) and largely 

 throughout the monachines. Mirounga spp. show a tendency to develop a large process, 

 while the process is lost entirely in the clade of Monachus schauinslandi and Monachus 

 tropicalis. A distinct process is largely lost in the phocines. although many taxa are 

 characterized by a slight ridge or roughening of the maxilla in its former location. Only 

 Cystophora (state 1). Erignathus. and Histriophoca (both state 0) possess distinct 

 preorbital processes. 



*25) distinct pbstorbital process of maxilla: 0 = absent: 1 = present (pers. obs.). 

 With recoding, this character was included in character #26. 



26) size of postorbital process of maxilla: 0 = small: 1 = medium: 2 = large; 9 = absent 

 (pers. obs.). 



To date, this feature has been used primarily to elucidate the higher level relationships 

 within the Caniformia. Among the pinnipeds, the presence of the postorbital process has 

 been used to distinguish the otariids from Odobenus and the phocids, where it is absent, 

 or, at best, rudimentary (Mivart 1885; Ridgway 1972: King 1983; Wyss 1987). Howell 

 (1928) attributes this absence in the phocids to the lack of an interorbital extension of the 

 temporalis, combined with the larger, more dorsally positioned eyes of this group. 

 However, it may develop with age in phocids as Allen (1887) mentions a distinct frontal 

 (= postorbital?) process in very old individuals of Monachus tropicalis. Wozencraft (1989) 

 lists a large postorbital process as being unique for the otariids among caniforms, although 

 Wyss (1987) indicates that it is also present in ursids. 



Although Odobenus is typically listed as lacking a postorbital process [only Mivart (1885: 

 497) indicates the presence of one, but he earlier (page 493) contradicts himself], there is 

 cause to doubt these reports. Wozencraft (1989) notes that the strong facial compression 

 characteristic of this taxon has resulted in the confluence of the postorbital process with 

 the lacrimal flange (= preorbital process?) into a single process of questionable homology. 

 Our observations support this finding, although the preorbital process is implicated in the 

 place of the lacrimal flange. However, both the pre- and postorbital processes can be 

 individually differentiated (primarily due to their different bones of origin), and we have 

 chosen to recognize each as being medium in size (state 1 ). 



The general trend within the Caniformia is for a reduction in the size of the postorbital 

 process to its eventual loss ancestrally in the phocids. However, the primitive state for this 

 character is uncertain due to both Canis and Ursus being polymorphic for this trait (both 

 states 1 and 2). Beyond this, ACCTRAN optimization indicates an initial derivation of a 

 small process for Procyon, Martes, and Enhydra, followed by a medium process uniting 

 Lutra and the pinnipeds ancestrally. In contrast, DELTRAN optimization holds for state 

 1 being a synapomorphy extending from Martes to the pinnipeds ancestrally, with Procyon 

 and Enhydra independently evolving state 0. Zalophus always uniquely obtains a large 

 process. The postorbital process is lost ancestrally in the phocids and is never regained 

 within this group. At best, various species possessed a slight ridge, or a roughening of the 

 frontal bone, but nothing that we would term a distinct process. 

 *27) nasolacrimal (= lacrimal) foramen: 0 = absent; 1 = present (Wozencraft 1989). 

 With recoding, this character was included in character #28. 



