119 



phocids clearly separates them from the non-inflated bulla of the otarioids (Howell 1928; 

 Repenning 1972; Wyss 1987), a difference that is manifested even during fetal 

 development (Howell 1928). We attempted to elucidate the pattern of bulla inflation using 

 three characters from Wozencraft (1989) that identify different bullar elements. 

 Unfortunately, differentiation between the ectotympanic and entotympanic portions of the 

 auditory bulla is difficult among carnivores due to their high degree of fusion in the adult 

 bulla (Repenning 1972; Hunt 1974). Differentiating between the two regions is aided in 

 some phocids by the presence of a distinct sulcus between them (van der Klaauw 1931: 

 Burns & Fay 1970; see character #83); however, the distinction between this character 

 and the following two remains somewhat arbitrary. Instead, we will use these three 

 characters to represent inflation of respective regions of the bulla (lateral, middle, and 

 medial respectively) rather than of the elements themselves. A further problem is that the 

 inflation of the auditory bulla is best judged internally (see Repenning 1972), but we were 

 limited to an external examination of the bulla in almost all cases. 

 Despite noting that the ectotympanic forms a large percentage of the bulla in otarioids, 

 Repenning (1972) singles phocids out from the pinnipeds for their enlarged ectotympanic 

 portion. This is corroborated by Wozencraft (1989), who indicates that only the ursids and 

 the otarioids lack an inflated ectotympanic among the carnivores. Again, the plesiomorphic 

 condition for the Caniformia is uncertain; however, it is likely an ectotympanic that is not 

 inflated, as this is the condition shared by all the outgroups except Corns and Martes (both 

 state 1). These observations are largely in conflict with those of Wozencraft (1989). As 

 expected, an inflated ectotympanic (state 2) is reasonably common among phocids, but 

 more so among phocines where it is ancestral and retained by all members except 

 Pagophilus (state 0), and possibly Histriophoca (states 0, 1, and 2) and Phoca vitulina 

 (states 0 and 1). Among monachines, state 2 characterizes only Hydrurga and the clade 

 of Lobodon plus Monachus spp. In this last clade, a slightly inflated ectotympanic is a 

 synapomorphy of Monachus schauinslandi and Monachus tropicalis. 

 81) inflation of caudal entotympanic along anteroposterior axis: 0 = not inflated; 1 = slight 

 / moderate inflation; 2 = inflated (Wozencraft 1989) (Fig. 22). 



Hunt (1974) indicates that inflation of the caudal entotympanic is primarily responsible 

 for the overall inflation of the carnivoran bulla. Yet, the contribution of the entotympanic 

 to the auditory bulla among the arctoids is quite variable. It is the greatest in some 

 phocines, comprising two-thirds to three-quarters of the bulla (Burns & Fay 1970), but 

 generally comprises about two-thirds of the bulla in most phocids (Repenning 1972). 

 Mustelids display an intermediate ratio, usually comprising more than one-third of the 

 bulla (King 1983), while the entotympanic contributes little more than the formation of 

 the carotid canal in otarioids (Repenning 1972). Among the carnivores, Wozencraft (1989) 

 lists canids, procyonids, mustelids (excluding lutrines and mephitines), and phocids as 

 possessing a caudal entotympanic inflated along the anteroposterior axis. Our observations 

 agree with this distribution, although the polarity is reversed for the caniforms. Here, an 

 inflated entotympanic is plesiomorphic, and its presence in the phocids (except for 

 Hydrurga and Monachus schauinslandi, which independently obtain state 1) represents a 

 reversal. 



