ECOLOGY AND BIOLOGY OF THE PACIFIC WALRUS 



85 



data), and presumably in all other pinnipeds. 



Each of the three primary premolars in the lower tooth row develops labial to 

 a permanent premolar and is a well-formed, calcified tooth. These teeth are 

 similar to p2-3 '4 of the upper row and comparable to P2-3-4 of the lower row in 

 other pinnipeds (see Laws 1953Z?; Scheffer and Kraus 1964; Scheffer 1967a). The 

 three secondary lower premolars of the walrus also resemble p2-3-4 of the upper 

 tooth row and are comparable to P2-3-4 of the lower row in other pinnipeds, in 

 that the first is about as large as or larger than the second and is at least as deeply 

 rooted. This resemblance is indicated also by the occlusion of these teeth with the 

 upper premolars. Whereas in all other pinnipeds and other carnivores, Pj is 

 situated slightly ahead of Pl, P2 slightly ahead of P2, etc., the first lower 

 premolar of the walrus usually is situated behind P^, the second behind P2, and 

 so on, in the "normal" occlusal positions of P2-3-4 (see Cobb 1933, plates IV, V). 



These findings pose the question of homologies of the three lower postcanines. 

 Has the first or the fourth been lost from the primitive dental condition? Are the 

 "survivors" Pi.2-3, as suggested for example by Berry and Gregory (1906) and 

 Cobb (1933), or are they P2-3-4? The available evidence suggests the latter. 

 Although I found no traces of either a primary or secondary tooth germ of the 

 missing first premolar in my fetal series, I suspect that these eventually will be 

 identified, probably in fetuses younger than those available to me. In the 

 meantime, I feel compelled to accept the circumstantial evidence, which 

 indicates that the three lower premolars of Odobenus are 2-3-4, rather than 

 1-2-3. Absence of P^ is not an unusual situation in carnivores (e.g., in many mus- 

 telids, some procyonids, hyaenids). Although its absence is unusual in pinnipeds, 

 Pj in other recent and fossil otarioids ordinarily is (or was) the smallest of the 

 lower premolars (see Mitchell 1966, 1968; Barnes 1972; Repenning and Tedford 

 1977). In at least one of the modern forms, Eumetopias juhatus, P^ is by far the 

 smallest of the lower premolars and occasionally is absent from the tooth row 

 (personal observation). 



Two late fossil walruses believed to have been on the ancestral line to 

 Odobenus had four well-developed lower premolars. The Pi of Aivukus cedro- 

 sensis of the late Miocene North Pacific was slightly larger in diameter than P2, 

 but it was the less deeply rooted (Repenning and Tedford 1977) . In the somewhat 

 later Prorosmarus alleni of the early Pliocene North Atlantic, the P^ was slightly 

 smaller in diameter than the other postcanines (Berry and Gregory 1906) and, in 

 my interpretation of a cast of the specimen (kindly provided by C. A. 

 Repenning), it also was less deeply rooted than either P2 or P3. In Alachtherium 

 cretsii, another fossil walrus of the Pliocene North Atlantic, four lower premolars 

 also were present, the first of which was appreciably smaller than the second 

 though still separated from Ci by a wide diastema (DuBus 1867). In the cast of 

 the mandible of Prorosmarus, P2 is in precisely the same location as is the first 

 lower premolar of Odobenus, relative to the mental foramen and coronoid 

 process of the jaw. 



Gingival Eruption of Secondary Teeth 



The sequence of gingival eruption of the secondary teeth was determined in 60 

 free-living walruses, 0 to 6 years old, and in 3 captives when they were 2 to 15 



