MIGRATION OF ADULT SOCKEYE SALMON. a 
The work at these stations was in charge of the following men: 
A. Sooke, British Columbia— 
W. H. Rich, field assistant, U. S. Bureau of Fisheries. (July 14 to 23.) 
Dr. C. McLean Fraser, director of the biological station, Nanaimo, British Co- 
lumbia. (July 23 to the end of the season.) 
Alex. Robertson, superintendent, Harrison Lake hatchery, Dominion fisheries 
department. 
. Salmon Banks— 
Walter C. Buckmaster, apprentice fish-culturist, U. S. Bureau of Fisheries. 
‘. Point Partridge— 
Clive L. Henry, apprentice fish-culturist, U. 8. Bureau of Fisheries. 
D. Lummi Island— 
Don E. Courser, apprentice fish-culturist, U. S. Bureau of Fisheries. 
E. Point Roberts— 
Joseph Kemmerich, foreman, U. S. Bureau of Fisheries. 
The last four of these stations were under the general supervision 
of Dennis Winn, field superintendent, U.S. Bureau of Fisheries. 
ey |s2, 
COLLECTION AND ORGANIZATION OF DATA, 
In the collection of data reliance was necessarily placed in fisher- 
men, trap tenders, and cannery men. A reward of 25 cents was 
offered for the return of each button accompanied by mformation 
giving the date and place of capture. Specimens taken i American - 
waters were taken care of by Dennis Winn, at the Seattle office of the 
U. S. Bureau of Fisheries, and those taken in Canadian waters by 
Col. Cunningham, at the office of the Dominion fisheries department 
at New Westminster, British Columbia. One of the tags reported 
from the upper regions of the Fraser River was secured by J. P. 
Babcock, assistant to the commissioner of fisheries for the Province 
of British Columbia. 
While the majority of these records are considered approximately 
correct, data secured in this manner are necessarily subject to some 
inaccuracies, both as to time and place of capture. In the authors’ 
opinion, however, these inaccuracies will be balanced so that with 
reasonably large series the averages should be reliable. In cases 
where the data were obviously wrong the records have been omitted. 
This has been done in several instances where the date given for the 
capture was earlier than that on which the fish was recorded as hav- 
ing been marked. A number of tags were returned with incomplete 
data, either the date or the place of capture, or both, wanting. All 
such cases have been emitted entirely from consideration. 
For the convenient organization of the data the American waters 
of the Sound from which returns were reported have been divided 
into nine regions. Several factors guided in establishing the boun- 
daries of these regions: (1) The general geography of the district, (2) 
the grouping of traps, and (3) the nature of the descriptions giving 
the locality where the tags were recovered. These last frequently 
indicated merely the general region in which the tag was taken, 1. e., 
Rosario Strait. The boundaries of these regions, as well as the loca- 
tion of the marking stations, are shown on the accompanying map. 
All of the records reported from the Canadian waters which are open 
to commercial fishing have been treated together. For this reason 
that part of the Strait of Georgia just outside the mouths of the 
Fraser River and the river itself, from the mouth to Mission Bridge, 
together constitute the tenth region. The reasons for so treating the 
Canadian returns are given on pages 26 and 27. 
