372 



REPOKT OF THE COMMISSIONER OF FISHERIES. 



The selection of water supplies for fish cultural or similar purposes 

 should include a careful scrutin}^ of their quality with respect to dis- 

 solved air. There are the two opposite faults to be guarded against. 

 When either is extreme in degree, its recognition will not be difficult. 

 But it is probable that cases will occur, and have occurred, where either 

 fault is but slight, and causes no heavy losses or marked symptoms on 

 its own account, while it at the same time is responsible for a gradual 

 and insidious lowering of condition among the fishes which makes 

 them susceptible to the sudden and rapid epidemics of bacterial or 

 protozoan infection, or to the less acute attacks of higher parasites. 

 In such cases the certain recognition of a slight excess of nitrogen, 

 with ordinary methods of gas analysis, may require the average of a 

 number of determinations. The constant ebullition of gas in bubbles 

 of moderate or large size from the water sources is sufficient to cause 

 suspicion of a nitrogen excess, but the absence of such bubbles is by no 

 means reassuring, for supersaturation may occur in the depths of the 

 spring without any of the undissolved residual gas revealing itself at 

 the surface. As for the oxygen, it is not known just what content 

 short of saturation completely supplies all the needs of fishes, but 

 since their natural abodes, and particularly" trout streams, closely 

 approach saturation (Hofer 1904, pp. 157 et seq.), it is well to lay stress 

 upon the desirability of maintaining a high oxygenation in fish cul- 

 tural waters. For trout, and particularly the brook trout, this is 

 imperative. It is probable that most spring waters are not highly 

 oxygenated. Usually they take up incidentally, in the conduits or at 

 delivery pipes, more or less oxj^gen before the}^ are actuall}^ used as a 

 fish-cultural suppl}", and sometimes means of aeration are specifically 

 provided. So important are these that it seems not too much to say 

 that devices for this express purpose should be provided in all cases 

 where spring or well waters are used for salmonoids, unless repeated 

 quantitative determinations made at different seasons show that the 

 water can not be improved. 



Table I. — Shoicing composiiion of gas delivered from the bottoms of ponds, springs, or 



ivells. 

 [All gas determinations by M. C. Marsh save where otherwise stated.] 



Nos. 1, 2, and 5 were determined by the Bureau of Chemistry. 



