NATURAL HISTORY OF THE SALMON. 75 



tcake of grilse, 43,000, or considerably more than one-lialf 

 above the average ; and in 1854, salmon reached 15,300, 

 also more than one-half above the average. In 1854 

 grilse declined to 16,739, one-third below the average; 

 and in 1855 salmon declined to 6239, also one-third 

 below the average. In a word, the proportion which the 

 grilse of any one year bore to the average number of 

 grilse, is found by these tables to have been just about 

 the proportion which the salmon of the folloiving year 

 bore to the average number of salmon. These facts seem 

 almost too neat and complete as evidence of our theory 

 that the grilse of one year are the salmon of subsequent 

 years — not, indeed, that they "prove too much," but that 

 they fit into our doctrine so exactly as almost to give 

 them the appearance of having been made to measure- 

 ment. But similar results are seen in a less regular and 

 perfect form, in the less regular and perfect returns from 

 other fisheries. Probably the Mackenzieites may hold 

 them to be only coincidences ; liut they must also admit, 

 that they are not only very remarkal)le, l)ut for them 

 exceedingly disagreeable, coincidences. 



The element of weight or size, which may be held 

 to include that of growth, is very important ; but Mr. 

 Mackenzie so deals with it that there is some difficulty 

 in getting hold of him. He says : " One simple and 

 palpable fact, which any ordinary observer might have 

 remarked, is, that grilse in May weigh from three to five 

 pounds ; in July they are met with as large as from ten 

 to twelve pounds ; and instead of finding them in August 

 and September grown to the size of sixteen or twenty 

 pounds, which would l^e but natural if they continued 



