Why Bessler’s Wheel started spontaneously.
The rotation of the wheel depends on the swing of each primary weight acting like the child on a swing to amplify the swing and thus the rotation. This action can only take place when the length of the swing is shortened at or just after the six o’clock point. This shortening of the pendulum is effected by the fall of the shifter weight which lifts the primary weight at the six o’clock position. Several successive pendulum swings can be amplified to generate continuous. accelerating rotation.
It is recorded that Bessler’s first two wheels would spontaneously begin to turn as soon as the brake was released. They were only able to turn in one direction, unlike his later wheels. 
It has been suggested elsewhere that the inventor might have stopped his wheel in a position in which it was inherently out of balance just so that he could restart it without giving it a push. It was thought that this would be an attractive feature of his exhibition. It might be thought that with the current explanation the same factors apply and it is certainly true that even when the swing has been started, the parameters have to be altered to add energy to the swing, or to put it another way, the pendulum has to be shortened, and therefore no amplification of any swing can occur before a swing has begun, and that is why it might be thought that the wheel would require a push in order for this to happen. However the fact that the primary weights follow a path for three-quarters of one orbit, closer to the centre of rotation (six to three o’clock) and a more distant path for a quarter of a turn (three to six o’clock) means that the wheel is always out of balance anyway so it will begin to turn spontaneously, and once a swing has been started it will amplify as its bob is raised or its pendulum shortened, leading to accelerated rotation of the wheel (see fig 6 again). This exactly what was reported by witnesses; the wheel began to rotate slowly at first and then within three turns had accelerated to its maximum speed.
Later Bessler produced wheels which could turn in either direction, but they did not start spontaneously. It seems clear that these later wheels contained two identical systems, one the mirror image of the other, so that each had the ability to turn the wheel in a different direction. From the above explanation it is clear that the two competing systems cancelled out the overbalancing tendency, so a slight push was required to set the wheel in motion. Clearly this brief push allowed the shifter weight to come into a position from which it could fall and in doing so, lifted the primary weight which had begun to swing, due to the rotation of the wheel, into its overbalancing position, thus amplifying the swing leading to continuous rotation. The sequence was dependent on which way the wheel was initially pushed. Those mechanisms which were rotating in a reverse direction to normal, were unable to provide any mechanical advantage, so were in effect neutralised, allowing those operating in the other direction to provide the rotational force. However there was no overbalancing component because both sets of mechanisms balanced each other, so there would only be the acceleration from the swinging to drive the wheel around, and its final speed would be less, resulting in a slower acceleration and less power available, which concords with the facts reported by numerous witnesses.
