"For the sake of brevity I call relations of position and motions recurring at regular periods, that is after like intervals, in the particles forming a material system or in the centres of whole masses conceived as forming a larger system, 'stable relations.' Among such relations is to be reckoned the condition of rest of the particles or masses in relation to each other, as the extreme case, which we may call the state of 'absolute stability'; while a dissipation of the particles or masses, to infinity, in different directions, constitutes the other extreme of absolute instability. "We do not speak of 'absolute stability,' but of 'full stability,' in cases where motion still takes place, but this brings continually, in exactly the same periods of time, the same relations of particles or masses, not only as regards position, but also as regards velocity and direction of the motion and change of velocity and direction.... "To absolute and full stability may be added, as third case, that of greater or less approach to full stability, which we may term briefly 'aproximate stability'... and of which we have an example in the chief bodies of our solar system. "It may serve as a simplification of the consideration of stable relations of motion to remark... that, in an isolated system or one under constant outer conditions, exactly or very nearly the same relations of velocity and direction recur when exactly or very nearly the same relations in the position of the particles or masses return. As regards the velocity, this follows directly from the principle of the conservation of energy; as regards the direction, it is indisputably possible to assume the connection of its recurrence with that of the other relations, although I cannot remember that a direct general proof of this has been found.