38 FISH CULTURE. 



Let us suppose it is desired to hatch one or two 

 traysfull, a stream which may be described as a 

 strong drip or dribble will be ample. If pure, cool, 

 and well aerated, you may pass such a stream through 

 three, but beyond that it is advisable to increase 

 the stream. I am supposing now that the water is 

 supplied from a cistern or some reservoir over which 

 the pisciculturist has fuU control. As this is the 

 most general, and perhaps the easiest way of con- 

 ducting operations, I shall speak of fish hatching 

 in streams by means of boxes filled with gravel, &c., 

 presently. To run such a stream as I have described 

 for twenty-four hours without failing, it would need 

 a reservoir or cistern capable of holding from 1 to 

 200 gallons; a very little visible increase in the 

 stream will be found to make a considerable increase 

 In the volume of water that runs away. If the 

 cistern can be filled twice in the twenty-four hours, 

 of course a considerable reduction in the size of it 

 may be made, but it should always be fuH large 

 for the purpose, in case of accidents, which at times 

 no amount of foresight can altogether prevent. All 

 cisterns should have a float on the inside, and com- 

 pensating plumb on the outside, ascending and de- 

 scending over a graduated scale, similar to that which 

 I use for the Acclimatisation Society, which will be 



