. 306 Salmonide 
various streams. (2) Return of marked salmon. (3) Intro- 
duction of salmon into new streams followed by their return. 
Under the first head it is often asserted of fishermen that 
they can distinguish the salmon of different streams. Thus the 
Lynn Canal red salmon are larger than those in most waters, 
and it is claimed that those of Chilcoot Inlet are larger than those 
of the sister stream at Chilcat. The red salmon of Red Fish Bay 
on Baranof Island are said to be much smaller than usual, and 
those of the neighboring Necker Bay are not more than one- 
third the ordinary size. Those of a small rapid stream near 
Nass River are more wiry than those of the neighboring large 
stream. The same claim is made for the different streams of 
Puget Sound, each one having its characteristic run. In all 
this there is some truth and perhaps some exaggeration. I have 
noticed that the Chilcoot fish seem deeper in body than those 
at Chilcat. The red salmon becomes compressed before spawn- 
ing, and the Chilcoot fishes having a short run spawn earlier 
than the Chilcat fishes, which have many miles to go, the water 
being perhaps warmer at the mouth of the river. Perhaps 
some localities may meet the nervous reactions of small fishes, 
while not attracting the large ones. Mr. H. S. Davis well 
observes that “until a constant difference has been demon- 
strated by a careful examination of large numbers of fish from 
each stream taken at the same time, but little weight can be 
attached to arguments of this nature.”’ 
It is doubtless true as a general proposition that nearly 
all salmon return to the region in which they were spawned. 
Most of them apparently never go far away from the mouth of 
the stream or the bay into which it flows. It is true that salmon 
are occasionally taken well out at sea, and it is certain that the 
red-salmon runs of Puget Sound come from outside the Straits 
of Fuca. There is, however, evidence that they rarely go so 
far as that. When seeking shore they do not reach the original 
channels. 
In 1880 the writer, studying the salmon of the Columbia, 
used the following words, which he has not had occasion to 
change: 
“It is the prevailing impression that the salmon have some 
special instinct which leads them to return to spawn in the 
