Nor do the gyroscopes interfere in the least with the turning of the car in passing round a curve, when the two of them are linked together, as Mr. Brennan links them, so that any lateral change in the axis of one is balanced by an opposite change of the axis of the other. With the single gyroscope, such as Mr. Brennan used when he first began his experiments, the car encounters difficulties at curves in the track. But before we can understand how the two gyroscopes balance each other in such a way as to make the Brennan car lean in while passing about a curve, we must investigate more fully the action of the individual gyroscopes. I have already said that there is another principle involved as supplementary to the principle of the fixed axis; this we must now investigate. Perhaps it would be fairer to say that what we have to consider is not a new principle but a complication as to the application of the principle of gyroscopic action already put forward. In any event there is an elementary fact about the gyroscope that I have not yet stated. It is this: in order that the gyroscope may exercise its fundamental property of holding its axis fixed, it must have that axis so adjusted that it is free to oscillate or wabble. That sounds distinctly paradoxical, but it is a very essential fact. If Mr. Brennan had merely fixed two wheels rigidly in the frame of his car, they would have had no appreciable effect in balancing it. Had nothing more than that been necessary, some one would have invented a gyrocar long ago. But very much more than that was necessary, as we shall see. The complication of which I am speaking is illustrated by the action of the simplest top, which likewise owes its stability to its wabble. Your top does not rise merely because it spins, but because it wabbles as it spins--wabbling being the familiar word for what the machinist calls "precession." A freely spinning top, if in equilibrium, has no inherency to rise up against gravitation, as your top may have led you to suppose. Your top rises because it is not spinning freely in equilibrium, its action being interfered with by the friction of the point on which it rests; it is seeking a position of equilibrium, which, owing to the location of its centre of gravity, will be found when its spindle is erect. But a top supported at both ends and properly balanced, does not tend to rise but only to maintain its position. HOW THE BRENNAN GYROSCOPES WORK