MIGRATIONS OF PINK SALMON 
665 
was very small. The percentage of recoveries in area 1, although by no means as 
great as those in areas 3 and 4, did show some indication of a seasonal decrease. It 
is not improbable that the majority of the Sumner Strait pink salmon migrating as 
far south as area 1 in Clarence Strait come from the early, rather than the late, part 
of the run. 
SUMMARY OF POINT COLPOYS EXPERIMENTS 
In reviewing the distribution of the pink salmon tagged at Point Colpoys (see 
tables 7 and 8), it will be noted that the majority of the tagged salmon recovered 
during the early part of the season were captured in the Ernest Sound region of area 
3. This region supports a large number of excellent spawning streams whose pink- 
salmon populations enter them during the early part of the season. The majority of 
the tagged salmon recovered during the latter part of the season, on the other hand, 
were captured in area 4. This region also supports a large number of excellent spawn- 
ing streams whose pink-salmon populations are known to migrate into them during 
the latter part of the season. Hence it may be assumed that most of the pink salmon 
migrating through Sumner Strait as far as Point Colpoys are bound for the localities 
in Ernest Sound and its contiguous channels, and the localities in the northern region 
of Clarence Strait. Furthermore, the pink salmon destined for localities in Ernest 
Sound and its adjoining channels, which are the farthest from the sea, pass through 
Sumner Strait early in the season, whereas those bound for the localities in Snow Pas- 
sage and the northern region of Clarence Strait, which are closer to the sea, migrate 
later in the season. 
The results from both the Clarence and Sumner Straits experiments indicate that 
most of the pink salmon migrating through these channels in the early part of the 
season are bound for the inside localities farthest from the sea, and as the season 
progresses they tend more and more to migrate into the localities which are closer to 
the sea. In view of these results the authors do not feel that the contention of the 
salmon packers, that their catches of pink salmon in area 2 during the latter part of 
the season are made from the runs migrating through Sumner Strait, is well founded. 
CONCLUSIONS 
Cape Fox region . — The pink salmon migrating through Dixon Entrance to the 
shores near Cape Fox are either bound for localities east of the cape or those in 
Revillagigedo Channel and the south arm of Behm Canal. 
Cape Chacon region . — Most of the pink salmon entering Clarence Strait by way of 
Cape Chacon, during the early part of the fishing season, are destined for localities 
in Revillagigedo Channel, the north arm of Behm Canal, and the more distant localities 
in the northern region of Clarence Strait. Most of those migrating by the same 
route later in the season are bound for localities on the east shore of Prince of Wales 
Island south of Approach Point. Thus, the early migrants are native to the streams 
farthest distant from the sea, whereas the later migrants are native to those in the 
more proximate localities. 
Cape Muzon region .- — The tagging experiments carried on near the entrance to 
Cordova Bay were all made during the latter part of the season. The results indicate 
that at this time the incoming pink salmon are bound for localities in Cordova Bay, 
with a small percentage continuing around Cape Chacon to the southeast shore of 
Prince of Wales Island. 
