In all cases of physical change produced by one body in another, it is generally possible to assume such a meaning of action, that the reaction shall be of the same nature as the action; and when this is done, the third axiom of causation, that reaction is equal to action, is universally true. Thus if a hot body heat a cold one, the change may be conceived as the transfer of a certain substance, _heat_ or _caloric_, from the first body to the second. On this supposition, the first body _loses_ just as much heat as the other _gains_; action and reaction are equal. But if the reaction be of a different kind to the action we can no longer apply the axiom. If a hot body _melt_ a cold one, the latter _cools_ the former: here, then, is reaction; but so long as the action and reaction are stated in this form, we cannot assert any equality between them. In treating of the secondary mechanical sciences, we {191} shall see further in what way we may conceive the physical action of one body upon another, so that the same axioms which are the basis of the science of Mechanics shall apply to changes not at first sight manifestly mechanical. The three axioms of causation which we have now stated are the fundamental maxims of all reasoning concerning causes as to their quantities; and it will be shown in the sequel that these axioms form the basis of the science of Mechanics, determining its form, extent, and certainty. We must, however, in the first place, consider how we acquire those conceptions upon which the axioms now established are to be employed.