conceivable. After the pumps were stopped, on this occasion,
tests were made with each in succession as to the power of
the ejectors with which each is fitted to recharge the pumps.

"The valves in the discharge pipe were closed and steam
admitted to the ejector, the pump being still and no water in
the gauge glass on the pump casing, which must be full before
the pumps will work. The suction pipe of the ejector is only
two and a half inches in diameter, the steam pipe one inch in
diameter. To fully charge the pumps at this point required
filling the pump casing and the suction pipe containing about
2,000 gallons; this was accomplished in four minutes, and
when the gauge glass was full the pump operated instantly and
with certainty, discharging its full volume of water.

"I went on several occasions down in the valve pits on the
ladder of the casing, and to all accessible parts while in
motion at its highest speed, and there was no undue
vibration, only a uniform murmur of well-balanced parts, and
the peculiar clash of water against the sides of the casing
as its velocity was checked by the blank spaces in the
runner.

"The pumps are noisy while at work, due to the clashing of
the water just mentioned, but it affords a means of detecting
any faulty arrangements of the runner or unequal discharge
from any of its openings. While moving at a uniform speed,
this clashing has a tone whose pitch corresponds with that
velocity of discharge, and if this tone is lacking in
quality, or at all confused, there is want of equality of