THE SEA OTTER IN THE EASTERN PACIFIC OCEAN 243 



three were liberated together, she ignored one and took on her 

 chest the one she habitually groomed. It was our conclusion that 

 the deserted pup was an orphan toward which this mother 

 exhibited tolerance. 



Large young, still dependent on their mothers, will frequently 

 play together for long periods while their mothers are diving for 

 food. If one mother should desert her young at such a time, it is 

 possible that an association such as we observed could occur. 

 Although it is possible that a mother may bear twins, I believe it is 

 doubtful that a mother could feed and groom and successfully 

 bring two young to the stage of independence. 



In the northern fur seal (Callorhinus ur sinus) which, like the 

 sea otter, normally bears one young at parturition and is highly 

 specialized to marine existence, twinning is rare. Peterson and 

 Reeder (1966) summarize data derived from Fiscus, Baines, and 

 Wilke (1964) demonstrating the incidence of twin implantation 

 in the fur seal at 0.14 percent among 4,223 pregnant females 

 examined. In spite of the observation that the young fur seal 

 appears less burdensome to its mother than the young sea otter, 

 no evidence exists to date that a fur seal can successfully raise 

 twins to weaning age (Peterson and Reeder, 1966). 



Length of Stages of the Reproductive Cycle 



Dr. D. G. Chapman, Chairman, Biomathematics Group, Labora- 

 tory of Statistical Research, Department of Mathematics, Univer- 

 sity of Washington, studied the data obtained from female repro- 

 ductive tracts and prepared the following analysis : 



Analysis of fetal weight data; The procedure of Huggett and Widdas (1951) 

 is used to estimate the fetal growth velocity and hence the length of the 

 growth period. Table 53 shows the data of table 50: fetuses classified by 

 weight classes according to time of collection reduced to a percentage basis. 



The last row in the table is obtained by (1) calculating an unweighted 

 mean of the mean weights in the three periods; (2) finding the cube root of 

 this mean. 



The use of percentages in the table and the unweighted means in the 



Table 53. — Percent of fetuses in each weight class 



[This table is derived from data in table 50] 



Time of collec- 

 tion after July 



Class 

 1 



Class 

 2 



Class 

 3 



Class 

 4 



Class 

 5 



1 month . 



31 



15 



15 



38 



0 



3 months. 



18 



12 



22 



35 



12 



7 months 



0 



0 



22 



33 



44 



Mean time in months 













after July 



1.7 



1.9 



4.0 



3.5 



6.1 



Mean of cube root of weight 



0.74 



2.21 



3.27 



7.18 



11.45 



