107 



jugal has a wider distribution, characterizing the monachines as a whole (DELTRAN 

 optimization), or possibly all phocids ancestrally (ACCTRAN optimization). This situation 

 arises from the convergent possession of a flat jugal in Cystophora, Erignathus, and 

 Mirounga spp. from the subset of monachines mentioned above. In these three genera, 

 state 1 results from the increased robustness of the body of the jugal. In no case is the 

 maxillo-jugal suture elongated, nor does the suture contribute to the ventral margin of the 

 jugal. 



53) degree of overlap of maxillary and squamosal processes of zygomatic arch on medial 

 surface of zygomatic arch: 0 = little or none; 1 = approach closely - maxilla and squamosal 

 almost or in contact (pers. obs.). 



The phocids, and especially the monachines, seem to be characterized by modifications 

 to the sutures in the zygomatic arch (see characters #52 and 56). Generally, these 

 modifications take the form of an elongation of the sutures, and, in the case of the maxillo- 

 jugal suture, a rotation to a more horizontal position. In some cases, this elongation and 

 rotation is sufficient to bring the maxillary and squamosal processes of the zygomatic arch 

 into contact with each other, or at least in very close proximity. Either situation is 

 uncommon in the carnivores. This derived condition (state 1) is not widespread throughout 

 the phocids either, and is found independently in only Erignathus and Eeptonychotes. 



54) approach of jugal to lacrimal region: 0 = does not approach lacrimal region; 1 = 

 reaches lacrimal region / almost touches or does touch anterior wall of orbit (Wozencraft 

 1989; Wyss & Flynn 1993). 



Wozencraft (1989) originally employed a more restrictive coding than that used here, 

 examining whether the jugal contacted the lacrimal or not. This resulted in the derived 

 condition ("does not contact lacrimal") occurring in a wide variety of carnivores, including 

 ursids, mustelines, lutrines, and pinnipeds. However, as pointed out by Wyss & Flynn 

 (1993), the separation between the two bones in many cases is merely due to the 

 intervention of a thin sliver of the maxilla. Recognition of this as a trivial variation of the 

 primitive condition ("jugal and lacrimal in contact") reduces the distribution of the derived 

 condition (state 0 here) to the pinnipeds alone (and possibly the lutrines; see below). Such 

 a coding also reflects the peculiar contribution of the maxilla to the anterior orbital wall 

 in pinnipeds (Wyss 1987; Wyss & Flynn 1993; see character #40). Although Wyss (1987) 

 indicates that lutrines also possess the derived condition, he discounts this as the 

 configuration of the bones of the anterior orbital wall approximates that of the remaining 

 mustelids more so than that of the pinnipeds. In any case, we too have opted for a less 

 severe coding, partially reflecting our agreement with Wyss & Flynn (1993), and also 

 reflecting the problems caused by the reduced nature of the lacrimal in the pinnipeds. 

 Yet, despite the failure of the jugal to contact the lacrimal region apparently being a 

 pinniped synapomorphy, an obvious transition sequence for this character occurs within 

 the monachines. In the supposedly primitive Monachus spp., the jugal terminates relatively 

 medially [also Allen (1887) for Monachus tropicalis], above the centre of the infraorbital 

 foramen. This termination point moves progressively laterally through the intermediate 

 fossil taxon Homiphoca capensis to the more derived lobodontines, where it occurs lateral 

 to the infraorbital foramen (Hendey & Repenning 1972; de Muizon & Hendey 1980). 



