Ttie wide variation in vitamin A content of fur seal liver oil 

 poses an interesting problem. In outward appearance and "behavior the 

 a n i m als are alike. They are juvenile males, alx>ut 3 years of age, appear- 

 iog on land at the same time. They are in a resting or sani-resting stage 

 after a sojoiim of many months at sea. They are, as has "been shown, fairly 

 unifonn in Ixidy length. Why, then, the vitamin A potency in their livers 

 should vary so greatly is difficult to explain at the present time. That 

 the jrariation is caused ty great differences in the diet of the individaal 

 seal, is one of the more plausihle explanations.!/ 



Periiaps the variation depends upon the length of time "between the 

 most recent meal of the seal and the time of sampling. That is, perhaps 

 the vitamin A reserve tends to change as the male seal continues to fast on 

 the "breeding grounds. It is known that the adult male seals may gp two 

 months or more without food. How long the juvenile males, like those in 

 our sanple, habitually fast in the summer season has not been determined. 

 It can "be deduced, however, from study of the annuli or growth rings on the 

 teeth that the deposition of calcium is interrupted, hence there is proTsably 

 an annua l fasting period for the subadult as well as the adult males. 



Theoretical Considerations 



The data on vitamin A potency (Table 1) do not resemble a normal 

 distri"bution. When the o"bserved range is divided into ten equal classes, 

 62 percent of the values fall within the lowest class. When the individual 

 values are arrayed in order, equidistant on the x-axis, while their loga- 

 rithms are plotted on the y-axis, the plotted points fall nearly on a 

 strai^t line. Dr. Z. W. Bimhaum, of the Laboratory of Statistical Research, 

 University of Washington, has kindly examined our data. He states that he 

 does not know of any model to which tiiey conform, although they can be quite 

 well fitted by a distribution of the type 



f (X) = M 



L+ X 

 where X stands for vitamin A potency and f (X) for the number of individuals 

 having potency X. The parameters M and L will be positive constants which 

 can be fitted by the method of least squares. The two constants can then be 

 used to characterize the distribution. (Miss Elizabeth "^aughan, Statisti- 

 cian, Alaska Fisheries Investigations, Fish and Wildlife Service, has kindly 

 conpited the parameters for our data. She finds that L = k,150 and 

 M = 23. U6.) Dr. Bimbaum suggests, further, that the distribution is bi- 

 modal and the superinposition of two normal distributions. All evidence ■ 

 from the field, however, points to a homogenous sanple. 



1/ A tabulation by Victor B. Scheffer (in press) of the contents 

 of 1,300 seal stomachs from the North Pacific Ocean and Bering Sea shows 

 that the following items are of most frequent occurrence; squid, pollack, 

 herring, Crustacea (chiefly from the fish stomachs), seal fish (Bathylasus ) . 

 sa.1mon, ailachon, and rockfish. 



