198 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1914. 
cient to carry the gyrostat round when there was no spin, is now 
quite insufficient to cause any serious change of position of the 
gyrostat. Only a very small couple producing precession acted. 
This experiment ulustrates the principle of permanence of direc- 
tion of the axis of rotation, in the absence of a couple producing 
precession, the principle on which depend the gyrostatic compass and 
the self-directing torpedo. Carried within the body of the torpedo 
is a fast-spinning gyrostat, and at the instant at which the torpedo 
leaves the impulse tube this gyrostat is mounted freely with its axis 
coincident with that of the torpedo; that is, pomted, so to speak, 
exactly along the “cigar.” Any turning of the torpedo body side- 
wise brings about a relative shift between the gyrostat and torpedo 
axes, and this shift brings into operation a vertical rudder at the stern 
of the torpedo. If the nose of the torpedo turns to port, the rudder 
steers the craft to starboard, and vice versa. 
Here (fig. 2) is a skeleton frame representing a torpedo. It is 
mounted on a vertical axle, and carried on pivots within the struc- 
ture is one of our motor gyrostats. At the stern of the frame is a 
small rudder, and this is connected by means of cords to the gyro- 
stat. Iset the flywheel in rotation. When, as I do now, I turn the 
nose of the torpedo to port, the rudder steers to starboard; when I 
turn the nose to starboard the rudder steers the craft to port. 
The case of the pedestal gyrostat is provided with a hook at one 
extremity of the axis (see fig. 3). The effect of hanging a weight 
on this hook is to apply a couple tending to cause turning about the 
axis 2; that is, which would produce such turning if the flywheel 
were not spinning. But the wheel is spinning, and the visible actual 
turning is about the axis 3. Observe also that the wheel is rotating 
comparatively slowly, and that the precessional motion is great. 
I increase the speed of the flywheel and the gyrostat precesses more 
slowly. JI replace the weight by a larger one, and for the same spin 
the precessional motion is greatly increased. Thus for a given 
applied couple the faster the spin the slower the precessional motion, 
and for a given spin the greater the couple the faster the precessional 
motion. 
Now, while the weight is in position and the gyrostat precessing 
about the axle 3, I attempt to hurry the precessional motion, and 
immediately the gyrostat turns about the axis 2 so as to rise against 
eravity. I try to delay the precession, and again the gyrostat turns 
about the axis 2, but now so as to descend under gravity. 
Without being aware of it people are constantly meeting with 
examples of gyrostatic action in daily life. A child expert in trund- 
ling a hoop causes it to turn its path to the right or left, by striking 
it a blow at the top with the hoop stick, the effect of which the 
ordinary person would suppose, if he thought about it, should be to 
