TAGGING SALMON IN ALASKA, 1923 
29 
abundance. It seemed advisable to test this by a second experiment, which might 
enable us to ascertain whether the extensive banks south of the peninsula served 
constantly as feeding grounds for Bering Sea salmon, or whether their presence on 
these banks in 1922 was an exceptional occurrence. It will be noted in the present 
report that the second experiment completely verified the results of 1922, although 
the season of 1923 was marked by a red-salmon run of very different character. 
The methods employed in 1923 did not differ from those in use the previous 
year. Photographs are presented showing the character of the aluminum tags, 
the tongs that served to attach them to the upper portion of the base of the salmon’s 
tail, and the method of work. The salmon were secured from three different traps 
on Unga Island, from four in Morzhovoi Bay, and from three in Ikatan Bay and 
one in East Anchor Cove. Practically all of this material was furnished by the 
Pacific American Fisheries, from whom we received, in addition, all possible assistance 
and sympathetic cooperation. We wish also to acknowledge gratefully numerous 
courtesies extended us at the P. E. Harris plant in False Pass. 
In the prosecution of our work there was used a small seine skiff, in which we 
entered the spiller of the trap. The spiller was then sufficiently lifted to pen the 
salmon along its outer walls, where they were caught up with a dip net, tagged, and 
released. Four men ordinarily were employed in this process and, under favorable 
conditions of wind and weather, could tag and release 300 salmon in one hour. 
On two different occasions 1,000 fish were tagged in the course of a single day’s 
work. Unga Island (Squaw Harbor), Morzhovoi Bay, and Ikatan cover a span 
about 125 miles in length, and it was desired in each of these localities to tag salmon 
near the beginning, the middle, and the end of the short run, which lasts only about 
six weeks. This program was carried out with fair success, although not infre- 
quently our work was interrupted by the high winds for which this region is noted; 
but the results secured by this division of the season and the different behavior of 
the later fish of the season, compared with those running during the earlier and 
middle portions of the run, amply justified this feature of the plan. 
An interesting problem which early presented itself was the extent to which 
the homing habit of salmon controls their movements during migration. Salmon 
were found associated in the same trap on Unga Island which were shown to scatter 
to more than 20 different streams, ranging from Cook Inlet on the south to Good- 
news Bay, the Kuskokwim, and the mouth of the Yukon in the far north. The 
vital question which arises with regard to these is whether each of them is return- 
ing to the stream in which it had its origin — its home stream — or to what extent 
we may find among them strays, scattering aimlessly in search of a spawning stream. 
If the homing instinct prevails, we are compelled to accept at the other end of the 
salmon’s life an extremely wide dispersal of the young to feeding grounds hundreds 
of miles distant from the mouths of the streams in which they were reared; and at 
the close of their period of growth and development, at the beginning of the season 
in which their eggs and milt will ripen, we must figure them leaving the feeding 
grounds and independently retracing the hundreds of miles which may separate 
each of them from its natal stream. 
No landmarks have been suggested that seem in any way adequate to guide 
these fish in their prolonged journey through an apparently trackless sea. The 
44699—27 3 
