GYROSTATS AND GYROSTATIC ACTION—GRAY. 205 
which, when put in the proper way, is really very simple. Excessive 
rolling of a ship is due to the cumulative action of the waves, and 
such cumulative action is only possible where the period of the ship 
and that of the waves are of about the same order. A large ship has 
a very long period, and synchronism of the ship and the waves is 
impossible. The effect of introducing a gyrostatic control, operated 
in the manner just described, is to endow the small ship with the 
period of a very large one. 
In the second mode of operating the gyrostat, friction is intro- 
duced at the bearings on which the frame of the gyrostat is mounted. 
With this addition the ship is forcibly prevented from excessive 
rolling. In the trials of the device it was found that, with the con- 
trol in operation, the angle of roll of the ship did not exceed 1° in a 
cross sea which produced a total swing of 35° when the control was 
out of action. It is interesting to notice that, contrary to the opin- 
ions which were expressed when the device was first suggested, the 
preventing of the rolling of a ship does not result in the waves break- 
ing over her; a ship controlled by a gyrostat is, I believe, a dry one. 
I have here a motor-gyrostat fitted within a skeleton frame repre- 
senting a ship. (Fig. 12.) The frame is mounted on two bearings 
arranged on wooden uprights, and may be made to oscillate on 
these bearings, so as to imitate the rolling of a ship in a cross sea. 
The frame of the gyrostat is mounted on two bearings placed athwart 
the frame, and a weight is attached to the outside of the case in a 
position in line with the axis of the flywheel. The center of gravity 
of the gyrostat is in line with the bearings. A clip device is provided 
which allows the gyrostat to be clamped to the skeleton frame, and 
provision is made whereby a graded amount of friction may be 
applied at one of the bearings. 
I now set the skeleton frame vibrating with the flywheel at rest. 
You observe the period. I start the motor gyrostat, and repeat the 
vibrations, with the gyrostat clipped to the frame. The ship rolls 
precisely as before. I free the gyrostat from the frame, and again 
set the ship rolling, when you see that not only is the period vastly 
increased, but the rolling motion is quickly wiped out. 
When the gyrostat is clipped to the frame it produces no effect 
upon the rolling motion. The couples opposing the rolling motion 
arise from the precessional motion, and hence the gyrostat must be 
given freedom to precess. In this connection it is interesting to 
observe that in 1870 it was proposed by Sir Henry Bessemer to 
obtain a steady cabin for a cross-channel steamer by placing it on a 
gyrostat with its axis vertical and supported on fore and aft trunnions. 
This plan was bound to fail. The dependence of the effect on freedom 
of the axis to precess, in a direction which is not that of rolling, was 
not understood. We now see that the object would have been 
