157 
Water power, if secured, as is customary, by connection with a tap in 
the laboratory, has some of the disadvantages of electric power in being 
inconstant. Every tap in the vicinity that is opened or closed varies the 
pressure, and even the initial pressure can by no means be depended 
upon. 
After many and vain efforts to obtain efficient power for my labora- 
tory the following plan was hit upon, and has proved wholly satisfactory. 
The method is very simple. It consists in providing a tank with an inlet 
valve operated by a float. As the water is withdrawn from the tank the 
float sinks and opens the valve, and the inflowing stream of water brings 
the water in the tank back to the full height. As the inlet valve permits 
a much larger stream of water to pass than the outlet valve does, the 
tank always stands essentially at the same level. 
The simplicity of my whole arrangement is one of its particular feat- 
ures. The tank consists of a vinegar barrel. It is set upon a platform in 
the story above my laboratory, and gives a fall of about fifteen feet. The 
water supply is taken from the city water that is piped throughout the 
building. The pipe conducting the water to the laboratory below is a 
small lead pipe and siphons the water over the top of the barrel. It is 
almost as easily put in place as rubber tubing, and can readily be changed 
to reach any part of the laboratory. It is closed at the lower end with a 
short piece of rubber tubing and a screw pinchcock, with which connec- 
tion can be made with a water motor or other apparatus. The float and 
valve in the barrel are such as plumbers usually furnish for dwelling 
houses. 
With this arrangement a practically uniform head of water is avail- 
able at all times, and although a fall of more than fifteen feet would be 
desirable, yet it has proven ample to run a centrifuge, using a Crowell 
water motor. 
A larger tank and more extensive plumbing might be used, but the 
arrangement as I have described it is so easily constructed and inexpen- 
sive as to be available in almost any laboratory. 
Many more uses of a small constant stream of water will come to mind 
than those mentioned. For instance, it is almost indispensable in running 
such an aspirator as is used for aerating seeds in a respirometer or for 
producing a regular alternating movement by filling a pair of buckets set 
on a pivot. The cheapness and uniformity of the power are its great 
merits. 
