But even with regard to the mechanical affections of matter themselves, though all the changes which take place in nature were truly reducible to them, we should still have ultimately the same difficulty in attempting to predict, without experience, the changes that would ensue from them. The mechanical properties are indeed the most familiar to our thought, because they are those which we are constantly witnessing in the great displays of human power that are most striking to our senses. The house, the bridge, the carriage, the vessel, every implement which we use, and the whole wide surface of the cultivated earth, present to us, as it were, one universal trophy of the victories of the great mechanist, man. We cannot look back to the time when we were ignorant of the mechanical properties of matter; but still there was a time when they first became known to us, and became known by experience of the motions that resulted from them. What can be simpler than the phenomena of impulse? That a ball in motion, when it meets another at rest, should force this to quit its place, appears now to be something which it required no skill or experience to predict; and yet, though our faculties were, in every respect, as vigorous as now,--if we could imagine this most common of all phenomena to be wholly unknown to us,--what reason should we be able to discover in the circumstances that immediately precede the shock, for inferring the effect that truly results, rather than any other effect whatever? Were the laws of motion previously unknown, it would be in itself as presumable, that the moving ball should simply stop when it reached the other, or that it should merely rebound from it, as that the quiescent ball should be forced by it to quit its state of rest, and move forward in the same direction. We know, indeed, that the effect is different, but it is because we have witnessed it that we know it; not because the laws of motion, or any of the mechanical affections of matter whatever are qualities that might be inferred independently of observation. Experience, then, is necessary in every case, for discovering the mutual tendencies of the _elements_ of bodies, as much as for determining the reciprocal affections of the _masses_. But experience teaches us the past only, not the future, and the object of physical inquiry is, not the mere solitary fact of a change which has taken place, but the similar changes which will continually take place as often as the objects are again in the same circumstances,--not the _phenomena_ only, but the _powers_ by which the phenomena are produced.