IPLE OF THE CYCLOSTAT.]

Let us consider (Fig. 1) a mirror, A B, reflecting an object, C D, and
revolving around it: when the mirror will have made a half revolution,
the image, C' D', of the object will have made an entire one. The
figure represents three successive positions of the mirror, distant by
an eighth of a revolution. The structure of the image shows that it
has made a quarter revolution in an opposite direction in each of its
positions. But if (Fig. 2) the body itself has revolved in the same
direction with an angular velocity double that of the mirror, its
image will have described a circle in remaining constantly parallel
with itself. The image will be just as insensible as the object
itself; but it is very easy to bring it back to a state of rest.

Let us suppose (Fig. 3a) the observer placed at O, the revolving
object at T, the axis of rotation being this time the line O F. Let us
place a mirror at A B and cause it to revolve around the same axis;
but, instead of looking at the image directly in the mirror, let us
receive it, before and after its reflection upon A B, upon two
mirrors, C D and D E, inclined 30 deg. upon the axis of rotation of the
system; the image, instead of being observed directly in the mirror, A
B, will always be seen in the axis, O F, and will consequently appear
immovable.

The same result may be obtained (Fig. 3b) with a rectangular isosceles
prism whose face, A B, serves as a mirror, while the faces, A C and B
D, break the ray--the first deflecting it from the axis to throw it on
the mirror, and the second throwing it back to the axis of rotation,
which is at the same time the line of direction of the sight.

The principle of the instrument, then, consists in causing the
revolution, around the axis of rotation of the object to be observed,
of a mirror parallel with such axis, and in observing it in the axis
itself after sending the image to it by two reflections or two
refractions. In reality, the entire instrument is contained in the
small prism above, properly mounted upon a wheel that may be revolved
at will; and, in this form, it may serve, for example, to determine
the rotary velocity of an inaccessible axis. For this it will suffice
to modify its velocity until the axis appears to be at rest, and to
apply the revolution counter to the wheel upon which the prism is
mounted, or to another wheel controlling the mechanism.

But Mr. Thury has constructed a completer apparat