From the ensemble of the facts as above set forth, it will be seen that rhythm results wherever there is a conflict of forces not in equilibrium. If the antagonist forces at any point are balanced, there is rest; and in the absence of motion there can of course be no rhythm. But if instead of a balance there is an excess of force in one direction—if, as necessarily follows, motion is set up in that direction; then for that motion to continue uniformly in that direction, it is requisite that the moving matter should, notwithstanding its unceasing change of place, present unchanging relations to the sources of force by which its motion is produced and opposed. This however is impossible. Every further transfer through space must alter the ratio between the forces concerned—must increase or decrease the predominance of one force over the other—must prevent uniformity of movement. And if the movement cannot be uniform, then, in the absence of acceleration or retardation continued through infinite time and space, (results which cannot be conceived) the only alternative is rhythm.
A secondary conclusion must not be omitted. In the last chapter we saw that motion is never absolutely rectilinear; and here it remains to be added that, as a consequence, rhythm is necessarily incomplete. A truly rectilinear rhythm can arise only when the opposing forces are in exactly the same line; and the probabilities against this are infinitely great. To generate a perfectly circular rhythm, the two forces concerned must be exactly at right angles to each other, and must have exactly a certain ratio; and against this the probabilities are likewise infinitely great. All other proportions and directions of the two forces will produce an ellipse of greater or less eccentricity. And when, as indeed always happens, above two forces are engaged, the curve described must be more complex; and cannot exactly repeat itself. So that in fact throughout nature, this action and re-action of forces never brings about a complete return to a previous state. Where the movement is very involved, and especially where it is that of some aggregate whose units are partially independent, anything like a regular curve is no longer traceable; we see nothing more than a general oscillation. And on the completion of any periodic movement, the degree in which the state arrived at differs from the state departed from, is usually marked in proportion as the influences at work are numerous.