1898.] NATURAL SCIENCES OF PHILADELPHIA. 201 



molar ; roof of mouth gently, evenly rounded, shallow between the 

 grinders ; incisive foramina small, not indented, their distance be- 

 low the alveolus of Hii only equalling the distance between the two 

 opposing permanent incisors; nasals small, short, rectangular. 



Rosmarus virginianus (DeKay). Fossil Atlantic Walrus. 



Trichechns virginianus DeKay; Nat. Hist. N. York, I, p. 56, pi. 19, fig. 1, 

 a, b. 



Characters. — Ratio of greatest anterior maxillary width to par- 

 occipital expansion, as 2 to 2f ; permanent upper incisor as large as 

 — ; — 3 larger than 5^1, nearly as large as permanent incisor, deeply 

 rooted, persistent; superior grinders massive, crowded, the opposing 

 rows separated by a narrow incisive diastema about as wide as the 

 largest molar : root of mouth deeply furrowed between the grinders ; 

 incisive foramina large, acutely indented ; their depth below the 

 alveolus of ^ being twice as great as the space between the two 

 opposing permanent incisors : nasals large, relatively long, becom- 

 ing much wider anteriorly. 



It should be understood that the above diagnosis of the fossil wal- 

 rus rests chiefly on the Sable Island specimen, and the finer skull 

 belonging to the Academy from New Jersey which Leidy figured 

 on plates 4 and 5 of volume XI of the Philosophical Transactions. 

 Both these skulls evidently belonged to very old males. In some 

 respects, as in the relative sizes of the teeth to each other, the other 

 New Jersey specimen and the plate of DeKay's virginianus agree in 

 their closer approach to the existing walrus. Owing to their frag- 

 mentary condition, as compared with the Sable Island and Long 

 Branch specimen, and the fact that the latter two agree exactly in 

 all the characters enumerated, it is best to consider these as typical 

 of the fossil animal. As DeKay's type is destroyed and his diagno- 

 sis and plate of little value, I would recommend that if the charac- 

 ters pointed out in this paper as distinguishing the fossil from the 

 recent Atlantic walrus are sufficiently confirmed by other speci- 

 mens to warrant their separation, that DeKay's name be retained. 

 The evidence in favor of DeKay's fossil being the same as rosmarus 

 and the other fossil specimens a distinct species, to which the name 

 virginianus cannot apply, is too flimsy to merit attention. 



14 



