In this connection let me call particular attention to a cardinal distinction with reference to the action in rolling upon and from an obstruction. If the wheels in descending hold the man back in order to remain in contact and thus roll off, it will, of course, result in a check of momentum exactly equal to that which would occur in such advance upon an obstacle, as would be shown by a similar curve in the opposite direction; but, as a matter of fact, the momentum being a certain amount, the effect is to cause the wheel to leave the obstruction entirely and not roll, but jump off, which result causes a great loss of energy and is sure to occur in rapid running. In this case the forward momentum gets no benefit from the potential energy acquired in mounting the obstacle, which shows the great necessity of proper springs such as will enable a man to swing forward slightly without rigidly drawing the machine after him. The object of the springs in this connection should be to hold the wheel in contact and permit it to roll instead of forcing it to jump off; if it rolls and is not carried off by the force of momentum, the energy will be given out in driving the machine forward instead of being lost in the vibration caused by impact when the machine strikes the common level. That is to say, the machine should roll off, but not hold the man back in order to do so; by proper springs the wheels remain in contact, while the man goes on at the regular pace of momentum. The liability of the rear wheel to jump off is a serious difficulty in the present Rover type of rear-driver; there is no reversion of the momentum, nor such a tendency to drop perpendicularly, as in the Ordinary, yet it drops a greater distance and is charged with more weight. This objection cannot be entirely remedied by any springs we now have in use; it requires a lively vertical as well as a horizontal amplitude in the motion of the springs, and they should be placed at the hub of the rear wheel in a manner similar to those used of late in connection with the front wheel. It will be seen from the diagrams that the curves shown by the front wheels leaving the obstructions are never such as would show any liability to jump off; advancing _upon_ the obstruction must, in them, be mostly provided for. In Fig. 8 we have a machine provided with a thirty-inch front and twenty-four-inch rear driving-wheel. This is a modification of the Rover type recently favored by some English makers. The drop of the rear wheel is more radical than that of a full thirty-inch. In Fig. 9 appears a Dennis Johnson machine, with two wheels of the same size, having the seat low down and exactly midway between them. This is perhaps the easiest riding contrivance in so far as vibration, jolt, and shock are concerned. Observe the equable motion it displays. This machine was patented in England, as spoken of in an early chapter, seventy years ago.