terms of this form. But this is not the case. Let us consider electrified bodies. The electro-static energy due to their mutual action will evidently depend on their charge—i.e., on their state; but it will equally depend on their position. If these bodies are in motion, they will act electro-dynamically on one another, and the electro-dynamic energy will depend not only on their state and their position but on their velocities. We have therefore no means of making the selection of the terms which should form part of T, and U, and Q, and of separating the three parts of the energy. If T + U + Q is constant, the same is true of any function whatever, φ (T + U + Q).
If T + U + Q were of the particular form that I have suggested above, no ambiguity would ensue. Among the functions φ (T + U + Q) which remain constant, there is only one that would be of this particular form, namely the one which I would agree to call energy. But I have said this is not rigorously the case. Among the functions that remain constant there is not one which can rigorously be placed in this particular form. How then can we choose from among them that which should be called energy? We have no longer any guide in our choice.
Of the principle of the conservation of energy there is nothing left then but an enunciation:—There is something which remains constant. In this form it, in its turn, is outside the bounds of experiment and reduced to a kind of tautology. It
