FLUCTUATIONS IN THE SUPPLY OF HERRING 
25 
isons with earlier years it may appear that at times the fish reached a higher level 
of abundance than formerly, whereas the obviously higher level is but an artifact. 
This will, however, tend only to minimize any fall in the trend of abundance and, far 
from invalidating any fall which may be found to have occurred, will give it additional 
significance. 
ANALYSIS OF CATCH RECORDS 
The sources of the data herein employed are the same as those given in a pre- 
vious report (Rounsefell, 1930, p. 303) and will not be repeated. In this report, 
however, the most use has been made of the daily 
catch records and not the production records, 
which were necessarily emphasized in the former 
analysis of the early years. 
In analyzing the statistics for southeastern 
Alaska the whole region has been arbitrarily di- 
vided into 33 areas. (Fig. 9.) The boundaries be- 
tween areas have been drawn as far as possible so 
as to pass through waters where little or no fishing 
occurs. This was done partly to avoid all confusion 
in assigning catches to their proper areas and partly 
so that each area would represent a natural fishing 
ground. By thus separating each natural fishing 
ground it was felt that the analysis would be more 
in confomity with what meager knowledge already 
exists concerning races (Rounsefell, 1930, p. 272), 
and any fluctuations due to the passage of dominant 
year classes might be more easily segregated and 
studied. 
As a further refinement it was found advisable 
for purposes of analysis to divide the purse-seine 
boats into three groups according to the locations 
of the plants to which they delivered their catches. 
(Fig. 10.) Group I comprises boats delivering to 
plants south of Point Ellis. The boats of this group 
fish chiefly in area 4. Group II contains boats de- 
livering to plants north of Point Ellis but south of 
Wilson Cove. These boats fish chiefly in areas 4 
and 17, but are wider ranging than the boats of 
Group I. The boats of Group III, delivering to 
plants north of Wilson Cove, fish chiefly in the 
northern areas, especially 8, 9, and 20, and in the 
central Chatham Strait areas. One company main- 
tains two plants — one located at Port Herbert in 
Group I, and the other located at Warm Springs Bay in Group II. The boats of this 
company delivered fish to either plant; therefore although they are used in studies of 
the combined groups, their data have not been used in the analysis of the individual 
groups. 
CHANGES IN AVERAGE SIZE OF CATCH 
The average delivery per boat as a record of abundance is subject to the same 
criticism as that of the total catch or the catch per week, namely, that in times of 
the statistical analysis 
