The seals wintering off Japan 
have always been aseumed to breed on Rob- 
RECOVERIES IN JAPANESE WATERS pas felaay off Bakimii stand obthe Cake 
OF FUR SEALS mander Islands off Kamchatka. Thie as- 
BANDED ON THE PRIBILOF ISLANDS sumption is based on geographical grounds 
2 alone, without definite proof. Unfortu- 
nately very little tagging has bean done 
on the Commanders, and none of the few 
eceals tagged and branded on Robben Ialand 
has ever been reported from the wintering 
grounds. Nevertheless, geographical prox- 
imity suggests that the small migration in 
the Japan Sea probably is composed largely, 
if not entirely, of seale from Robben Ie- 
land, which is only a short distance fron 
oya-kaikyo, their apparent point of en- 
trance and egress. It is also possible, 
though again positive evidence is lacking, . 
that some of the Robben Island seals win- 
ter off the Pacific coast of Japan and 
return northward in spring to their birth- 
place via the Kurile and the Okhotsk Sea, 
But most of the seale wintering off the 
eastern side of Japan probably nove north- 
eastward, past the long-defunct rookeries 
in the Kurils, and head straight for the 
Commander Islands and the Pribilofs. 
The fact must now be accepted 
that a considerable, but as yet undeter— 
mined, number of Pribilof seals visit 

Figure 11 Japanese waters every winter. Seven seals 
tagged on the Pribilofs, all under two 
years of age, have been picked up in the 
waters off Japan (Figure 11). These re- 
coveries were! 

March 1929, east of Shiriya-saki, Honshu 
21 May 1929, off Muroran, Hokkaido 
March 1930, 10 miles off Kinkazan, Honshu 
May 1943, off Onagawa, Honshu 
15 May 1949, off Iwate Prefecture, Honehu 
March (t) 1949, off Iwate Prefecture, Honshu 
25 April 1949, off Miyagi Prefecture, Honshu 





fhe three tags attached in 1928 have already been discussed (pp 24-25). Three of 
the remaining four tage were taken between February and May 1949 from seale marked in the 
Pribilofs in 1947 (Figure 12), That as many as three seals in the same age group should 
twice be reported from Japan in one season is of considerable significance and cannot be 
regarded as accidental. The three 1947 tags recovered in 1949 were from 19,138 seals 
marked that year, representing 0.5 percent of the 1947 census of 3,600,000 seals in the 
Pribilof herd. Therefore, in a random sampling one of these 1947 tages should appear for 
every 200 Pribilof seale taken, and, following normal expectancy, about 600 Pribilof seals 
would have been killed off Japan to obtain these three tags. 
