44 FISHERIES OF ALASKA, 1908, 



shown by the water glass, makes it absolutely certain that the num- 

 ber of salmon passing uncounted to the spawning grounds is not 

 worth considering. The canneries ran lightly until July 1, but on 

 that day the main body of salmon encountered the fishermen at 

 the mouth of the bay. Possibly on the 7th, 8th, and 9th of July, 

 but certainly on the 10th, this run had reached the foot of the lake, 

 as shown by the sudden rise in the daily tally. Even disregarding 

 the tally, it is certain they did not arrive before these dates, for the 

 presence of fish below the web is announced before they pass the 

 gate. They are seen playing along the web, especially at its south 

 end, and breaking the surface all about the lagoon in the vicinity 

 of the rack. The total run of salmon, being a matter of millions, 

 is too large to be materially affected, even if all the salmon arriving 

 before the 10th were ignored; and since long before this the rack 

 was stopping practically all the fish, it is not necessary to correct 

 the total by estimating the error due to fish escaping unobserved. 

 While difficulties were met in making a tight rack in such deep and 

 swift water, they were definitely overcome before the arrival of 

 the redfish run. 



The whole run having passed through the gate and tunnels, the 

 count is as accurate as the method used in making it, which method 

 is perfectly adapted to the purpose, and is in a sense automatic. 

 The fish are made to pass at a uniform or slowly varying rate through 

 a given aperture, and the rate measured at regular and frequent 

 intervals of time. This measurement is merely a careful count of 

 the number passing during one minute, and requires only alert care 

 and quickness on the part of the observer. The error of the personal 

 equation of individual judgment is thus eliminated or made a mini- 

 mum, unless more fish are allowed to pass than can be actually 

 counted during one minute. During a few quarter-hour periods 

 250 fish per minute were tallied, but usually the gate when wide 

 open would not deliver more than 200 per minute, and each of the 

 tunnels alwa3^s fewer. It is rather difficult to count satisfactorily 

 more than 200 fish per minute. 



The run of salmon may be regarded as having been actually 

 counted and implicit confidence placed in the total. The round 

 number, 2,600,000, is so near the total for the redfish passing into 

 Lake Aleknagik that any difference that exists is not worth con- 

 sidering. 



Significance of the figures. — Two million six hundred thousand 

 red salmon passed up Wood River to the spawning grounds, and 

 practically all may be regarded as seed planted for the reproduction 

 of the species. These salmon were part of a large run against which 

 a heavy fishing campaign was made. It would now be of the great- 

 est interest to know just how many red salmon ascended the other 



