30 
BULLETIN OF THE BUREAU OF FISHERIES 
physical and chemical conditions of their immediate environment assuredly were 
identical for all the salmon tagged and released at a given time from a single trap. 
In quick succession the numbered tags are attached to the struggling fish, fresh- 
dipped from the trap corral. Rapidly they are restored to freedom, tossed over the 
trap web into the clear waters, through which their aluminum tags shine bright. 
For a brief while they linger outside the trap to recover from their enforced sojourn 
in the upper air. A dozen or more of them can usually be seen while the marking 
is in progress, lazily fanning their gills and recuperating their strength. But they 
soon disappear, and when next they come to the attention of men, bearing their 
tags, after the lapse of a few days or a few weeks, it is found that some have journeyed 
east, some north, and some west, and some have passed into Bering Sea through an 
obscure channel, which the navigator locates with the aid of all his instruments. 
What is it that guides them in these journeys? From the same starting point, 
at the same moment of time, they pursue divergent paths to their destinations hun- 
dreds of miles away. How can we conceive that any elements of the common 
environment in which they lie immersed can so react as mechanically to lead different 
individuals in opposite directions? It is easy to demonstrate that they pursue their 
course in no haphazard fashion. Time schedules can be constructed on the basis of 
our recaptures, to which they conform with remarkable fidelity. We can not escape 
the conviction that they pass along the ocean ways with directness and speed. 
That salmon possess a "homing instinct” — whatever that term may connote — 
that they do, in fact, wherever they may have wandered, find their way back to 
their native streams, has in recent years found wide acceptance. It comes as a neces- 
sary inference from facts established by a study of separate colonies of spawning 
fish. As each colony is stamped by certain distinctive physical characteristics, it 
appears impossible to account for this fact except upon the assumption that we are 
dealing with isolated, self-perpetuating groups with the constant habit of home- 
coming at maturity. This does not signify that no individuals every stray beyond 
the boundaries of their own groups and enter streams other than those in which 
they were bred. We have never succeeded in recognizing such strays even in 
strongly marked colonies, in which it would seem an alien should be immediately 
detected; but admitting the existence of strays, the inference is unavoidable that 
they are present in very small proportion, too small to prevent, in each colony, the 
formation and the maintenance of racial pecularities. 
Whatever inferences concerning migrations in the sea we have felt warranted in 
making heretofore, based on a study of spawning colonies, we have never until now 
been in a position to take up the problem from the other end and to trace the opera- 
tion of the homing instinct in the case of salmon about to leave their feeding grounds 
in the sea and to disperse widely from a common center. While there are good a 
priori reasons for believing that this dispersal is along lines that lead to their home 
streams, the present tagging experiments have given us our first opportunity to 
verify this assumption — an opportunity too unique and too valuable to be neglected. 
This became apparent at the close of the 1922 experiment, when the tagged individ- 
uals, to our surprise, were reported from so very extended a geographic range. 
It was then too late to obtain samples of their scales, so the problem could not then 
be taken up, for our only available means of testing the return of salmon to the 
streams of their nativity is through a microscopic examination of their scales and 
