642 U. S. BUREAU OF FISHERIES 
spawning return. During the sea life of the salmon the fluctuations 
in mortality are not sufficient (in the case, at least, of the Fraser 
River) to destroy the correlation. 
That last statement regarding the general correlation and the fact 
that there is very little destruction in the sea, is another interesting 
aspect. If a very small number of fish returned in any particular 
year there would be two possibilities—either that they had not been 
successful in their early natural history, or else that there had been 
some destruction out in the open sea. That makes possible, I should 
think, a very interesting study of the needs of the young fish. 
Doctor Gitpert. As regards this matter of destruction in the sea, 
we must not minimize the amount that occurs. There is always a very 
great destruction. The best evidence we have comes from our experi- 
ments with marked fish. Doctor Rich has been foremost in this 
investigation, and for many years we have marked large numbers of 
fish so that they could be recognized on their return, and the ratio 
of return to the number marked (assuming that the marking has 
not caused any injurious effect on the young fish) will give a fair 
evidence of the destruction in the sea. 
Doctor Ricu. I think the work of marking will be handled by Mr. 
Holmes. He has been taking charge of that work on the Columbia 
River, and I think he has that all in hand. 
Doctor Ginpert. With reference to the amount of destruction that 
occurs to fish on entering the sea from the Karluk River, we believe 
we have some very favorable results in that respect. The majority of 
fish remain until their third year, when they proceed to sea. They 
are somewhere between 5 and 8 inches long, and we believe are fairly 
well able to take care of themselves. At the same time, I shall be very 
agreeably surprised if we find, on the completion of our investigation 
in the Karluk River (which Doctor Rich now has in progress, he 
having marked some 47,000 fish on their way to sea last year), 
we have a return of more than 10 or 15 per cent of these fish. So that 
destruction in the sea is not negligible, and the fact that it has not, 
in all cases, succeeded in destroying the correlation between the 
spawning and the final returns, merely shows that there are fluctua- 
tions in spawning which may not easily be overcome by destruction 
in the sea. 
I may emphasize, at this point, our experience in the Fraser River. 
In this river we had had, from time immemorial, a very extensive run 
every fourth year and small runs in the intervening three years. In 
the fourth year there was very heavy spawning, for some unknown 
reason. Far back, beyond historical times, the four-year cycle 
became a fixed feature in the fishery of that stream, but by a very 
unfortunate occurrence in one of those fourth years the run was 
checked. The fish were not permitted to reach their spawning 
grounds, and the peak run in the fourth year was lost entirely and 
never was recovered. 
I must call your attention to the fact that the pink salmon, or 
the humpbacked salmon, is a 2-year fish, as Doctor Rich has said, 
and we have alternate cycles almost universally throughout the range 
of the pink salmon. Every other year is a successful year. There 
is no overlapping of groups. We do not have 3, 4, 5, 6, or 7 year 
fish to help breach over the results of unsuccessful spawnings; so 
