112 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1911. 
of the ship. Therefore, in order to get a good result, it was necessary 
for the compass to have a very great gyrostatic resistance, so termed, 
opposing energetically any force tending to change the direction of its 
axis of rotation, and that the friction of its bearings should be made 
as small as possible. A consequence of the latter condition, however, 
would be that the gyrostat would come to its normal position only 
after a relatively long time, oscillating to and fro about that normal 
position. Meanwhile it would be subject to new perturbations. 
Accordingly one great desideratum was to provide some device for 
lessening these oscillations; at first, Anschtitz tried for this purpose 
a second gyrostat; later he developed a much more simple and effi- 
cient method. His gyrostatic compass was tried on the steamer 
Deutschland in 1908 and has since been used in the German Navy. 
It has just been adopted by the English Navy, and other navies are 
also trying it. 
Let us consider its principles: We know that a gyroscope once 
started tends to maintain its axis in an invariable direction, and 
that if any force is applied tending to change this direction, preces- 
sional movement takes place, which displaces the axis perpendicular 
to the direction of the disturbing force. Such being the case, let us 
imagine a gyroscope, inclosed in an appropriate box, suspended from 
a float which rests in a liquid bath in such a manner that the gyro- 
scope is perfectly free to swing in any direction like a pendulum 
which is at rest; the center of gravity of the system is below the 
metacenter; the gyroscope is mounted at the lowest point possible. 
Because of its weight the axis of the gyroscope tends to maintain 
itself, as well as the whole attached mechanism, in a horizontal 
position. 
Let us set the gyroscope disk rotating. In the past such rotation 
could be effected only by rough and very unsatisfactory means; now 
we have a much more advantageous method at our disposal. We 
may, for instance, drive it by alittle three-phase motor fed by means of 
fine conducting wires so that the rotation may be kept up indefinitely. 
As soon as the gyroscope disk is in rapid rotation with its axis hori- 
zontal, then if this axis is not in the plane of the terrestrial meridian, 
the rotation of the earth will tend to alter the axis from its original 
position. The gyroscope tends to respond, but, restricted by its 
weight, which forces the axis to remain horizontal, it will undergo 
only a horizontal displacement. 
This leads it to take a north and south direction, because as long as 
its axis 1s not parallel to that of the earth, the cause of this movement 
is still effective, so that if it is sufficiently free to move, it will indicate 
true north. 
Plate 1 shows a model designed to show experimentally this action. 
It consists of a small gyroscope, driven by a small three-phase elec- 
