( 3 ) 



the limit of variation of the recent animals. It will appear that in 

 most respects the fragment approaches the old walruses, hut that 

 the tusks deviate considerably from those of the living animals. 



Looking at the hind side of the fossil skull, one is at once struck 

 by its relatively great height. This is caused by the strong develop- 

 ment of the mastoid process and by a high ridge on the lambda 

 suture (crista lambdoidea). 



Dividing the height of the skull and that of the mastoid process by 

 the breadth of the skull, we obtain two quotients, which for the fossil 

 are greater than for the walrus. In determining these quotients the 

 height of the mastoid process was measured by the vertical distance 

 between the lower edge of the foramen magnum and the base of I he 

 mastoid process. The diiferences found are small, however; for this 

 characteristic the fossil stands consequently at the end of the variation 

 series of the walrus. The strong relief on the mastoid process and 

 the extraordinary size of the crista occipitalis externa as well as of 

 the crista lambdoidea are characteristics which the fossil skull has 

 in common with some old walruses. 



Small deviations are also found in the vicinity of the foramen 

 magnum. The canalis hypoglossi always opens with the walrus into 

 the inside of the condyli of the cranium with two openings at each 

 side, whereas the fossil only shows a single small opening. Rut this 

 characteristic has not much value, since the aperture of the canalis 

 hypoglossi always varies strongly. The foramen magnum is much 

 more flattened dorsoventrally with the fossil than with the walrus, 

 but this too is a very variable characteristic. Comparable numbers 

 are here obtained again when the breadth of the foramen magnum 

 is divided by its height. The condyli occipitales are in the fossil less 

 strong than normally and present a shuttle-dike appearance, while 

 in the walrus they project more and more when we proceed 

 upwards; also they here project above the upper edge of the foramen 

 magnum, while there they remain below its upper edge. 



Of all these small differences the shape of the condyli and of the 

 foramen magnum have the greatest importance, while the height- 

 ratios and the aperture of the canalis hypoglossi are of less value. 



The base of the skull shows no more differences with the walrus 

 than the back part. The length of the two skulls compared with 

 their breadth, agrees entirely. The first difference concerns the position 

 of the foramen lacerum and of the canalis alisphenoideus. These 

 namely lie close behind each other with the walrus, while with 

 the fossil they are separated by a rather massive bony lamella, 

 extending from the os petrosum towards the pterygoid process. Also 



1* 



