AN EXPERI-MENTAL DETERMINATION OF THE 
SPEED OF MIGRATION OF SALMON 
IN THE (^OLUMBIA RIVERA 
CHARLES WILSOX GREENE 
Professor of Physiology, University of Missouri 
TWO FIGURES 
Certain fishes, like some birds, carry out extensive migrations, 
but unlike birds, their movements are hidden from direct observa- 
tion. Commercial fishermen dip their nets into the waters and 
learn to know many of the movements of fishes by the character 
of the catch. Certain fishes travel in great schools, and this 
tendency to herd together furnishes an easy metljod for following, 
especially those school-s that swim near the surface. Presum- 
ably the migratory mo\'ement serves one of two purposes; either 
it is a means for providing food, or it brings fishes to the spawning 
ground. 
The method of following the movement of fishes by the quan- 
tity of the catch is at best crude. One has no assurance that the 
school from which the catch is made in one locality at a given time 
is identical with that from w^hich another catch is taken, even in 
approximately the same locality and at a nearly related time. For 
this reason it is almost imperative that one shall identify the indi- 
vidual fishes observed in order to determine their movements from 
one locality to another. Following this line of reasoning, Rut- 
ter^ undertook to brand migrating Pacific salmon in the Sacre- 
mento River in 1902. This seezr.s to have been the first effort at 
marking individual salmon with a view to determining their 
migrations in our North American waters. 
^ Published by permission of the Commissioner of Fish and Fisheries. 
2 Rutter: Bull. U. S. Bureau of Fisheries, 22, 122, 1902. 
