80 ODOBvENUS EOSMARUS ATLANTIC WALRUS. 



NOMENCLATURE. Several specific names have been in more 

 or less current use for the Atlantic Walrus, or rather for the 

 Atlantic and Pacific species collectively. Accepting Odobamus 

 as the proper generic name of the group, there is nothing to 

 prevent the adoption of rosmarus for the specific name of the 

 Atlantic species. It was used for this species exclusively by 

 Linne, Erxlebeu, and other early systematic writers, the Pacific 

 Walrus being at that time unknown tq the sy stematist s. If Ros 

 marus be used as the generic name of the group, as it has been 

 by a few late writers, as a substitute for the wholly untenable 

 one of Trichechus, it will be, of course, necessary to adopt some 

 other name for the species. Dr. Gill has used obesus of Illiger ; 

 but as this was applied by Illiger exclusively to the Pacific Wal- 

 rus, it cannot properly be used for the Atlantic species. It 

 would be difficult to select a subsequent name that would not 

 be open to objection, if one should stop short of trichechus, used 

 (inadvertently"?) in a specific sense ("Rosmarus triclieclius") by 

 Lamont in 1861. The name longidem of Fremery, 1831, was 

 based on what subsequent writers have considered as probably 

 the female, but the name is highly inappropriate, inasmuch as 

 it is the Pacific species, and not the Atlantic, that has the longer 

 tusks. There are left virginianus of DeKay and dubius of 

 Stannius : the first is objectionable on account of its geograph- 

 ical significance ; the other is only doubtfully referable to the 

 Atlantic species. Adopting Odobccnus for the genus, leaves 

 rosmarus available for the species, thus settling the whole diffi- 

 culty. 



As already noticed (antea, p. 20), two species besides virgi- 

 nianus have been based on fossil remains, and have been made 

 the basis of new genera. The first of these is the Odobenothe- 

 rium lartetianum of Gratiolet, since referred by Defrance to the 

 existing species ; the other is the Trichecodon liuxleyi of Lan- 

 kester, which there is perhaps reason for regarding as the large 

 extinct progenitor of the existing Walruses. 



ETYMOLOGY. The term rosmarus was originally used by 

 Olaus Magnus, about the middle of the sixteenth century, in a 

 vernacular sense, interchangeably with morsus, the Latinized 

 form of the Eussian word morsz (or morss). It was used in the 

 same way by Gesner a few years later, as well as by numerous 

 other pre-Linna3an authors. Respecting the etymology of the 

 word, von Baer gives the following : " In dem historisch-topo- 



