FLUCTUATIONS IN THE SUPPLY OF HERRING 
29 
Table 5. — Comparisons of average number of deliveries per week of each year with a standard 
delivery curve 
This method of analysis gives some- 
what the same results as that in which the 
average size of the deliveries was used, as 
can readily be observed by comparing Fig- 
ures 13 and 14. One marked difference 
occurs in the month of June in which the 
number of deliveries is highest in 1929. 
The two figures correspond quite closely in 
July, except for Group II in 1929, which has 
a high average size of catch and a low number 
of deliveries. The same is true in August 
for both the I and II groups. During 
August all of the groups have a high average 
size of catch in 1926, but the number of de- 
liveries shows no rise in Groups I and II and 
only a moderate rise in Group III. The 
total curves for the two methods of analysis 
agree rather well, except that the number of 
deliveries in 1926 is not so high relatively 
as the average size of catch. This may be 
an artifact, however, caused by the failure 
of the factor 0.786, by which the 1926 data 
was multiplied, to give a true valuation of the 
change in number of deliveries caused by 
the lack of a weekly closed season in 1926. 
CHANGES IN AVERAGE WEEKLY CATCH 
As explained above, the average size of catch and the average number of deliv- 
eries per week are both subject to certain errors. Thus when the fishing is being con- 
42456 0 — 31 -2 
Figure 14. — Showing the geometric means (expressed as 
percentages of the 1928 mean) of the percentages that the 
average number of deliveries per boat per week is of a 
standard curve (see text). Solid line, Group I; dotted 
line, Group II; broken line, Group III 
