438 REPORT OF COMMISSIONER OF FISH AND FISHERIES. [12] 
weight (in per cent.) of the six groups—the weight on the Ist of Novem- 
ber being counted as 100—the average weights of different periods were 
calculated, and from these data a curved line was constructed, showing 
the average changes of weight. 
The large figures are by no means superfluous; for from June to No- 
vember the individual differences ‘are very considerable, and salmon of 
the same length, the same date, and the same state of maturity fre- 
quently differ 25 and even 30 per cent. The question is not one of de- 
grees of emaciation, but of differences in the build of the skeleton. 
There are thick-set and slender figures among salmon as well as among 
men. Not until we compare all six groups are the differences sufficiently 
equalized, and do we realize the change of weight, which in its almost 
mathematical regularity bears the undoubted impress of absolute truth. 
