98 THE SALMON. 



and there is great danger of being deluded as to the 

 changes in the amount of produce, by taking figures 

 for localities too small and periods too short to insure 

 any approach to accuracy in the deducing of a total. It 

 is worse than useless to draw inferences from the yield 

 of single fisheries, or even an entire river in a single 

 season, because it often happens that, from the accidents 

 of flood and atmosphere, a good fishery fails in the same 

 year, and sometimes even in the same river, in which 

 even the bad fisheries are doing well, and that sometimes 

 from the same causes the prime of the season will be 

 half-lost to an entire river or district, not from the 

 absence of fish, but from the presence of natural or 

 accidental obstructions to catching them, such as a de- 

 ficiency of water. The nearest approach to an authentic 

 statement of the amount of the general decline during a 

 more or less well-defined period, is that given in 1860 

 by the English Commissioners of Inquiry, who state that 

 the evidence as to those rivers of England and Wales 

 where the fish had not been quite extinguished, showed 

 a decline, ranging from nine-tenths to A^tlis, within the 

 memory of living witnesses. This sort of evidence, how- 

 ever, it will be seen, is not much more to be relied on 

 for accuracy than the deductions to be drawn from facts 

 better specified, but too petty, and possibly exceptional. 

 The best plan seems to be to avoid on the one hand 

 data too narrow, and on the other any attempt to grasp 

 a larger mass of facts than can be accurately obtained 

 or easily handled, by taking one or two large rivers or 

 districts, having nothing exceptional in their circum- 

 stances, and a period of time sufficiently long to prevent 



