62 



BULLETIN OF THE BUREAU OF FISHERIES 



back. 10 On account of the time spent in passing through the estuary, between the 

 time the fish first reach the beach and the time they pass the weir, we are not justified 

 in adding the catch and the escapement of the same dates and calling the result 

 the "run" for those dates. Several systems have been tried, and the best evidence 

 we have indicates that an average time of approximately one week elapses while 

 the fish are passing through tbe estuary. For the purpose of securing the "run," 

 therefore, we have added the catch for each week with the escapement for the following 

 week. The dates given in Table 26 are those on which the catches were made. We 



1921 1924 (CALCULATED) I 



1922 1925 - 



1923 1926 uJ 



• • AVERAGE OF 192l/2.-3,-5,AND6. ^ 



-> 



Fio. 34.— Accumulated percentages of the total run for each week, 1921 to 1926 



have found that our data, when aggregated into weekly totals, serve every purpose, 

 and we have adopted this system throughout. 



The years for which we have detailed data on the runs (1921 to 1926) include 

 completely the last cycle mentioned in the section dealing with the statistical his- 

 tory of the fishery; and it is worthy of special note that, with the exception of 

 1922, this cycle follows the previous cycles in so far as the relative size of the corre- 

 sponding years is concerned. This fact indicates to us that here we are dealing 

 with a series of years that, in all probability, are typical of the years of the other 



•i Marking experiments conducted during the summer of 1926 showed that the time required for the migration between the 

 lower weir and the upper weir, situated at the Larsen Bay portage, was only four or five days. It is believed that the entire 

 journey to the lake requires only about a week. 



