88 



NORTH AMERICAN FAUNA 74 



Fig. 54. Globular dentin in the pulp cham- 

 ber of a walrus tusk from which the 

 lateral wall has been sawed awa\'. The 

 pulp was removed by maceration. Slen- 

 der vertical strands of tubular dentin 

 (slightly distorted) interconnect the 

 masses of spherical denticles. (Photo 

 bv G. C. Kellev) 



Fig. 55. Vascular canals in the central 

 core of globular dentin. Partially cleared, 

 longitudinal section from an upper pre- 

 molar of a 3-year-old Pacific walrus. 

 The central canals and radiating den- 

 tinal tubules appear \\"hitish in this view 

 under reflected light. (Photo bv G. C. 

 Kelle\-) 



some other means of filtration, not far from the pulp. I determined that they are 

 not completely blocked by filling the pulp chamber of a small tusk with carbon 

 ink, excising the tip of the tusk to expose the core, and immersing the cut tip in a 

 saline solution. The ink, drawn from the least to the greatest salinit\% flowed for 

 more than 3 cm through the core canals to the tip (Fig. 56). 



I presume that the canals become permanently blocked by calcification in the 

 more distal parts of the tusks, for I have seen no signs of blood, serum, or lymph 

 oozing from the exposed (by natural abrasion) core of the tusks of normal, free- 

 living adults. If these canals w^ere not blocked, they would be avenues for 

 disadvantageous loss of body fluids and ingress of pathogens. Walruses in 

 captivity frequently abrade their tusks to an extreme degree, exposing the core to 

 within a few centimeters of the pulp. In such instances, acute inflammation in 

 the pulp from bacterial invasion is a common pathological condition (Brown 

 1963; Brown and Asper 1966; Bartsch and Frueh 1971; Ruempler 1976). 



This central core of globular dentin apparent!)- was absent from the teeth of 

 Aivukus cedrosensis of the Pacific Miocene but appeared in the tusks of the 

 earliest known walruses of the Atlantic Pliocene, which, by stratigraphic asso- 

 ciation, are believed to have been Prorosmarus alleni (Repenning and Tedford 

 1977; C. E. Ray, personal communication). Although lesser amounts of globular 



