582 BR. EIXAR LONNBERG ON THE [Mar. 15, 



South Georgia, but one of them, that regarded as the type of 

 " M. I. crosetensis," is somewhat different, showing a percentage 

 larger than that of any of the adult skulls from South Georgia. 

 Although a direct comparison between this Crozet Island specimen 

 measuring 412 mm. and the quite young skull from South Georgia 

 measuring 193 mm. (conf. the table) cannot be made, though the 

 former is .young also, it is of interest to find that the percentage 

 expressing the relation between basal length and zygomatic width 

 is almost the same in both, viz. 84*8 and 84 - 9. This depends in 

 both cases on the shortness of the snout, which thus is a retained 

 juvenile characteristic in the Crozet Sea-Elephant, if constant. 



If we then turn our attention to the length of the palate, this 

 measurement will be found to vary from 49*5 to 56 - 5 per cent, of the 

 basal length of the skull in the seven South Georgia male skulls. 

 The figures expressing the same relative dimension in Lydekker's 

 specimens all fall within these limits and have thus no value as 

 racial characteristics. 



The width of the palate varies in the South Georgia skulls from 

 32*2 to 37*1 per cent, of the basal length. Even in this feature 

 Lydekker's specimen from the Crozet Islands is somewhat aberrant 

 as its corresponding percentage amounts to 39 - 3, thus indicating 

 a very broad palate. In this respect the Crozet specimen does not 

 exhibit a juvenile characteristic, for the corresponding percentages 

 of three quite young males from South Georgia are from 34 - 5 

 to 357. 



Among other characteristics used by Lydekker for distinguishing 

 his races of Sea-Elephants, he mentions the length of the inter- 

 palatine suture, which he says is "long"' m. ii fqlclandicus" "shorter" 

 in " macquariensis," and " longer" again in "crosetensis." It must 

 be said, however, that the shape and size of the palatine bones in 

 the Sea-Elephants are so extremely variable that no value whatever 

 can be laid on any character derived from them. To prove this 

 it may be mentioned that although, as a rule, the palate ends 

 mesially behind in a projecting tip, there are other specimens 

 also adult in which there is quite a deep palatine notch mesially. 

 Not only depending upon this difference but because the bones 

 themselves vary in size, the length of the interpalatine suture 

 becomes very variable, as direct measurements indicate. Thus in 

 the largest of the 8outh Georgia skulls (conf. table of measure- 

 ments) the interpalatine suture was only 30 mm. but in the next 

 81, in the next 57, then 62, 44, 61, and 68 mm. respectively. It 

 was thus more than 2| times longer in the next largest skull than 

 in the largest. The percentages expressing the relation between 

 the basal length of the skull and the length of the interpalatine 

 suture vary from 5 - 6 to 16 "5. and not two of them agree (conf. 

 the table). It is of interest to see that such a great variation in 

 the size of the palatine is not only due to modifications during the 

 growth at a somewhat advanced stage, but that already in quite 

 young individuals a similar difference makes itself apparent to the 

 extent that the length of the interpalatine suture in one of two 



