208 
BULLETIN OF THE UNITED STATES FISH COMMISSION. 
These eggs were hatched in two McDonald jars with gratifying success. I would not 
be understood as advocating muddy water in preference to that naturally clear, but 
to prove that clear water is not absolutely necessary to successfully hatch the eggs of 
fishes spawning in fresh water. 
The prime requisite is an unfailing water supply of a pressure as nearly constant 
as is attainable. For hatching semi-buoyant eggs the pressure should not be less 
than 6 pounds per square inch at the point where the water is to be drawn from the 
pipe ; a variation of 2 pounds will not materially affect the results, as it can, to a large 
measure, be corrected by raising or lowering the central tube of the jar. For eggs 
of the Salmouidee less pressure will answer ; three or even two pounds per square 
inch can be used. If the water is taken from a spring, race, or other abundant or 
cheap source, and carried into a supply tank by means of a flume, the tank may be 
kept always full and an overflow maintained, (Plate 30.) 
When the water is taken from the main of a city water- works it should be dis- 
charged into an open tank through a ball-cock, which, working automatically, will pre- 
serve a nearly uniform level or pressure. (Plate 31.) The bottom of the tank for 
semi-buoyant eggs should be not less than 6 feet above the top of the hatching-table. 
The water is taken by iron pipes from the bottom of the tank and distributed to the 
hatching-tables. (Plate 32.) 
A point especially to be guarded against is the presence of bubbles of air in the 
pipe. If the regulating tank is of sufficient dimensions the bubbles coming in with 
the supply will rise and escape from the surface. But if the level of the water is 
allowed to get low in the tank the water will drag air down with it into the supply pipe. 
To guard against the possibility of this I have devised the following apparatus : A 
float is placed in the tank from which a line, passing over pulleys, extends down into 
the hatching- room. On the lower end of the line is a weighted rod, carrying a metallic 
index sliding in a groove; to either side of the groove are aflixed contact points of an 
electric circuit. Two contacts are placed where the index would mark 8 inches, which 
amount is regarded as extreme low water in the tank. If desirable, other contacts 
may be arranged to indicate high water. As the metal index passes up and down, 
consequent upon the fluctuation of the water in the tank, it meets a contact, closes 
the electric circuit and thereby rings a vibrating alarm bell. (Plate 33.) This bell 
can be placed at any convenient point. As the greatest degree of fluctuation of 
pressure occurs in city water mains during the night, it has been found desirable, in 
hatcheries where no night watchman is maintained, to run an auxiliary alarm to the 
sleeping quarters. The ringing of the alarm bell calls attention to the fact that more 
or less water is needed, which can be easily remedied if a proper system of valves has 
been provided. 
The quantity of water necessary for hatching a given number of eggs must neces- 
sarily vary, not only with the supporting quality of the water, that is the relative 
amount of oxygen it contains, but to a considerable extent it depends upon its muddi- 
ness. In hatching semi-buoyant eggs it will be found necessary to employ considerably 
more water if it be very muddy ; a part to support the eggs and an additional quantity 
to wash away the mud. Allowing for air vents and other wastes it will be safe to esti- 
mate for 7,000 gallons of water each twenty-four hours for every million eggs of shad, 
and 4,000 gallons per twenty-four hours for each million whitefish eggs. Thirty gal- 
lons an hour through one jar will easily hatch 25,000 eggs of any of the Salmonidae. 
