# 1911 Encyclopædia Britannica/Horse-power

**HORSE-POWER**. The device, frequently seen in farmyards,
by which the power of a horse is utilized to drive threshing or
other machinery, is sometimes described as a “horse-power,”
but this term usually denotes the unit in which the performance
of steam and other engines is expressed, and which is defined as
the rate at which work is done when 33,000 ℔ are raised one
foot in one minute. This value was adopted by James Watt
as the result of experiments with strong dray-horses, but, as he
was aware, it is in excess of what can be done by an average
horse over a full day’s work. It is equal to 746 watts. On the
metric system it is reckoned as 4500 kilogram-metres a minute,
and the French *cheval-vapeur* is thus equal to 32,549 foot-pounds
a minute, or 0.9863 of an English horse-power, or 736 watts.
The “nominal horse-power” by which engines are sometimes
rated is an arbitrary and obsolescent term of indefinite significance.
An ordinary formula for obtaining it is 115.6D^{2} ^{3}√S for high-pressure
engines, and 147D^{2} ^{3}√S for condensing engines, where
D is the diameter of the piston in inches and S the length of the
stroke in feet, though varying numbers are used for the divisor.
The “indicated horse-power” of a reciprocating engine is
given by ASPN/33,000, where A is the area of the piston in
square inches, S the length of the stroke in feet, P the mean
pressure on the piston in ℔ per sq. in., and N the number of
effective strokes per minute, namely, one for each revolution of
the crank shaft if the engine is single-acting, but twice as many
if it is double-acting. The mean pressure P is ascertained from
the diagram or “card” given by an indicator (see Steam-Engine).
In turbine engines this method is inapplicable.
A statement of indicated horse-power supplies a measure of the
force acting in the cylinder of an engine, but the power available
for doing external work off the crank-shaft is less than this by
the amount absorbed in driving the engine itself. The useful
residue, known as the “actual,” “effective” or “brake”
horse-power, can be directly measured by a dynamometer (*q.v.*);
it amounts to about 80% of the indicated horse-power for good
condensing engines and about 85% for non-condensing engines,
or perhaps a little more when the engines are of the largest sizes.
When turbines, as often happens in land practice, are directly
coupled to electrical generators, their horse-power can be
deduced from the electrical output. When they are used for the
propulsion of ships recourse is had to “torsion meters” which
measure the amount of twist undergone by the propeller shafts
while transmitting power. Two points are selected on the surface
of the shaft at different positions along it, and the relative displacement
which occurs between them round the shaft when
power is being transmitted is determined either by electrical
means, as in the Denny-Johnson torsion-meter, or optically,
as in the Hopkinson-Thring and Bevis-Gibson instruments.
The twist or surface-shear being proportional to the torque, the
horse-power can be calculated if the modulus of rigidity of the
steel employed is known or if the amount of twist corresponding
to a given power has previously been ascertained by direct
experiment on the shaft before it has been put in place.