COLLECTION AND DISTRIBUTION OF SCIIOODIC SALMON EGGS. 821 



no special pains was taken to avoid the dripping of water from the fish 

 into the pan while the process of taking the eggs was going on. I have 

 thus minutely described the method pursued in order to explain the 

 occasional reference to the "usual method" in the tabular statement of 

 spawning operations (Table II). Occasional variations of the method 

 were made for experiments' sake, and such variations are noted in the 

 table. Subsequently, as soon as the eggs had reached the proper stage 

 for examination, the date of impregnation of each lot was ascertained by 

 taking 50 to 200 and examining each one critically. This number was, 

 I am convinced, too small for the purpose, and the results obtained and 

 indicated in the table can only be considered as rough approximations to 

 the actual rate of imin-egnation. The proper estimation of the impreg- 

 nation of a large lot of eggs, say 100,000, would, I think, require the 

 careful examination of as many as 500 eggs, or, better still, 1,000, taken 

 from different parts of the lot, which, for a large number of such lots, 

 would be quite a task. Such as it was the examination showed the ratio 

 of impregnation to have ranged from 87 to 100 per cent. Probably the 

 true average ratio was not much less than 95 per cent. No advantage 

 is indicated as resulting from the pains taken on several occasions to 

 give the eggs an extra quantity of milt, nor from unusual care in select- 

 ing ripe and rejecting unripe fish. I am not, however, willing to accept 

 these indications as conclusive. 



It is a matter of some importance to ascertain in what way, if any, the 

 percentage of eggs lost through non-impregnation may be still further 

 reduced, and many experiments were tried, for this as well as other ob- 

 jects, but they need to be supplemented by repetitions another year, 

 and can best be described in a future report. 



When we were not too much pressed with work, all the fish were 

 weighed and measured immediately after leaving the hands of the 

 spawn- takers. A siunmary of the results is given iu Table IX. From 

 this it appears that the average length of the males was 16.8 inches ( = 

 43 centimeters); longest, 22 inches (= 56 centimeters); shortest, 14 

 inches (= 35 centimeters). The males weighed, on the average, 1.8 

 pounds (= .8 kilogram); the heaviest, 3.7 pounds (= 1.7 kilograms); 

 the lightest, 1.1 pounds (= .5 kilogram). These are the residts of the 

 measurement of 235 fish. Of gravid females, 343 were weighed and 

 measured, with the following results: heaviest female, 3.6 pounds ( = 

 1.6 kilograms) ; hghtest, 1.2 pounds (= .5 kilogram) ; average, 1.9 pounds 

 (=: .9 kilogram) ; length of same females, average, 16.1 inches (= 41 cen- 

 timeters) ; longest, 20 inches (= 51 centimeters); shortest, 13 inches 

 (= 33 centimeters). 



At the close of the season's work the fish were taken in cars specially 

 arranged for such work, and towed about two miles up the lake, where 

 they were turned loose. Some of the males found their way back to the 

 stream within a few hours of being set at liberty, but the most of them, 

 aud nearly all the females, staid in the lake. 



