Let us consider again the condition of things when rounding a curve, as in the sketch on page 32. As long as the machine is moving forward in a straight line, the operator sits upright. When it begins to tip, he will unconsciously tip himself the other way, as represented by the line _xy_ in the rear view. Any bicyclist will recognize this as plausible. Why not take advantage of this involuntary movement to provide a stabilizing force? If operating wires are attached to the aviator's belt and from thence connected with ailerons or wing-warping devices, then by a proper proportioning of levers and surfaces to the probable swaying of the man, the control may become automatic. The idea is not new; it has even been made the subject of a patent. The Gyroscope [Illustration: THE GYROSCOPE] This device for automatic control is being steadily developed and may ultimately supersede all others. It uses the inertia of a fast-moving fly wheel for control, in a manner not unlike that contemplated in proposed methods of automatic balancing by the action of a suspended pendulum. Every one has seen the toy gyroscope and perhaps has wondered at its mysterious ways. The mathematical analysis of its action fills volumes: but some idea of what it does, and why, may perhaps be gathered at the expense of a very small amount of careful attention. The wheel _acbd_, a thin disc, is spinning rapidly about the axle _o_. In the side view, _ab_ shows the edge of the wheel, and _oo´_ the axle. This axle is not fixed, but may be conceived as held in some one's fingers. Now suppose the right-hand end of the axle (_o´_) to be suddenly moved toward us (away from the paper) and the left-hand (_o_) to be moved away. The wheel will now appear in both views as an ellipse, and it has been so represented, as _afbe_. Now, any particle, like _x_, on the rim of the wheel, will have been regularly moving in the circular orbit _cb_. The tendency of any body in motion is to move indefinitely in a straight line. The cohesion of the metal of the disc prevents the particle _x_ from flying off at a straight line tangent, _xy_, and it is constrained, therefore, to move in a circular orbit. Unless some additional constraint is imposed, it will at least remain in this orbit and will try to remain in its plane of rotation. When the disc is tipped, the plane of rotation is changed, and the particle is required, instead of (so to speak) remaining in the plane of the paper--in the side view--to approach and pass through that plane at _b_ and afterward to continue receding from us. Under ordinary circumstances, this is just what it would do: but if, as in the gyroscope, the axle _oo´_ is perfectly free to move in any direction, the particle _x_ will refuse to change its direction of rotation. Its position has been shifted: it no longer lies in the plane of the paper: but it will at least persist in rotating in a parallel plane: