2. Among 64 seals collected in June, 5 are 

 just beginning to molt. (The five include 1-, 

 2-, and 3-year males; 2- and 4-year females^) 

 The mean date of collection is 24 June. Among 

 49 seals collected in October and 36 in 

 November, 5 have passed through the molt and 

 have returned to resting stage. (The five include 

 1-, 2-, 3-, and 4- year males and a 1-year 

 female.) The mean date of collection is 11 

 November. The difference between 24 June and 

 11 November is 140 days, or 20 weeks. 



3. From the report of Bartholomew and 

 Hoel (1953, fig. 1) the peak of the pupping 

 season is mid-July. At birth the silver pelage 

 has already started to form. By the end of 

 September nearly all pups have replaced the 

 black birthcoat with a silvery one. It is reason- 

 able to deduce that by mid-October the pelage 

 of nearly all pups has returned to resting stage, 

 and that 3 months or 13 weeks have elapsed 

 during the growth of the first adult-type coat. 



4. Scant evidence on individual molt had been 

 obtained in an earlier study (Scheffer, 1962, 

 p. 27). In the Seattle Zoo, a male and a 

 female silver pup in prime new pelage were 

 partly shorn on 24 November 1952; regrowth 

 of the pelage was observed up to 26 October 

 of the following year. The findings of that 

 experiment are compatible with those of the 

 present study. The old and the new findings 

 have been reviewed and a reconstruction of 

 events in the molt has been prepared (table 3). 

 The evidence points to a duration of about 15 

 weeks. 



The molt of the underfur precedes that of the 

 guard hair. The time difference is certainly 

 days, rather than weeks, and is assigned a 

 value of 1 week (table 3). A more exact value 

 could be obtained by taking biopsy plugs from a 

 captive seal at intervals of a few days. 



There is no evidence from study of individuals 

 on duration of molt in older, as against younger, 

 seals. In the next section (page 15) it will be 

 shown that the overall length of the molting 

 period increases with age of seal. This could 

 mean that, in older individuals, either molt is 

 more erratic or its progress is slower, or both. 



At this point it will be helpful to compare 

 duration of molt in the seal with that in three 

 other carnivores: 



1. In the ferret Putorius vulgaris, "the new 

 winter coat is completed within 7 or 8 weeks" 

 (Bissonnette, 1935, p. 166). There is a second 

 molt of unspecified length in the spring. 



2. The mink Mustela vison, grows a summer 

 coat between mid-April and mid-July; a winter 

 coat between early September and late Novem- 

 ber (Oolnick, 1959, p. 6). During each molt 

 the duration of growth is 13 weeks. 



3. The silver fox Vulpes fulva, grows a single 

 coat between late April and late November 

 (Bassett and Llewellyn, 1948, p. 599-601). 

 While the total duration is 7 months, the growth 

 of new fibers on the back alone requires less 

 time. Bassett has estimated (in letter, 1962) 

 that the duration of molt on the back is from 

 1 August to 10 November, or 101 days (14.5 

 weeks), with great variation. 



Variation in Molt by Age, Sex, and Month 



Excluding pups, six age groups of each sex 

 are involved in the present study. An attempt 

 has been made to estimate the mid-date of 

 molt for each group. From the estimates, 

 certain conclusions have been drawn with re- 

 gard to variation in time of molt by age and 

 sex. In order to prepare the estimates of mid- 

 date, the following steps were taken: 



1. For each stage of molt except resting a 

 code number was assigned, representing its 

 distance (progression in weeks) from resting 

 condition (table 3). The code number is, for 

 beginning stage J., early 6, late U, and ending 

 14. In addition, code number 15 was assigned 

 to the "resting post-molt" stage in order to take 

 advantage of 6 specimens observed in this con- 

 dition in October-December and 13 in January. 



2. For each monthly sample by age and sex, 

 a mean code number was calculated. For 

 example, for the yearling male in September 

 (table 2) this is [(1 X 6) + (3X 11) + (2X14)] ^ 

 6= 11.2, The calculated results are given in 

 table 4. 



3. For each group (e.g., yearling males) a 

 regression line was plotted on the basis of the 

 monthly mean code numbers. The X-axis for all 

 groups was the same, scaled as June = 1, 



