68 BULLETIN OF THE BUREAU OF FISHERIES. 



accuracy the age of any particular seal, to test the judgment of the natives, and to 

 learn the range of variation in size among seals of a given class. Thus, if a supposed 

 3-year-old showed a small body measurement it was possible to examine its skull and 

 compare it with that of a known 2-year-old and so leain whether it was in reality a small 

 3-year-old or a 2-year-old mistaken for a 3-year-old. As the combined result of field 

 observations and measurements and the study of skulls and teeth, it was found that 

 the total length of 3-year-old bachelors is subject to but little variation, and that the 

 natives are able to distinguish seals of this class with a very small percentage of error. 



During a food killing on July i, 1914, at which the native clubbers were instructed 

 to proceed as usual and kill only 3-year-olds, 37 seals taken at random as they were killed 

 were caiefully measured with a steel tape and their skulls tagged and preserved. Thirty- 

 five of these proved to be undoubted 3-year-olds, one was a 2-year-old, and one was 

 larger than the others and may have been a small 4-year-old. The total length vaiied 

 from 45 to 52 inches and in 71 per cent of cases it was from 46 to 48 inches. At a later 

 killing on August 10, 61 seals were measured in similar manner, using calipers instead 

 of tape, which gave a slightly smaller result in each case but the same lelative uniformity 

 prevailed. One of the 61 proved to be a 2-year-old and the remaining 60 were un- 

 doubted 3-year-olds. The length measurement varied from 42 >^ to 51 inches and in 

 93 per cent of cases was from 44 to 49K inches. Such uniformity is not found in the 

 weight of the animals, which may be fat or lean, nor in the weights of the skins, which 

 vary according to the amount of blubber removed. It is found in the skulls, however, 

 and these serve to corroborate the accuracy and significance of the length measure- 

 ments. The seals of any two generations differ from each other as a class by not less 

 than 10 months in age, and since in the males there is a rapid growth from 2 until 6 

 years of age, it is evident that differences due to age are likely to be more pronounced 

 than those due to individual variation. A study of skulls proves this to be the case, 

 and with rare exceptions the age of any given skull may be determined upon the basis 

 of growth characters familiar to students of osteology. With the skull, as with the 

 animal, the length measuiement is the principal reliance, although other characters 

 are considered. The skull of a newly bom pup is short with a broad flattened brain- 

 case having no bony ridges or prominences ; the facial part of the skull is relatively 

 undeveloped and the teeth are just beginning to appear. In the 2-year male these 

 conditions in general still prevail, although the bone has thickened and the skull become 

 more lengthened. In the 3-year-old a more definite lengthening has taken place, the 

 braincase is higher and relatively narrower, and ridges and prominences begin to show. 

 This process is carried farther in the 4-y€ar-old and in succeeding years until in the old 

 male the skull which began smooth and flat becomes relatively high with various prom- 

 inences and a high bony ridge extending lengthwise over the top of the braincase. (See 

 pi. IX and X.) 



The data obtained from the 3-year-olds measured on July i, 1914, of which all 

 the skulls were preserved, are as follows: 



