If objection is taken to the interpretation of matter as consisting in certain relations, if its character as substance is insisted on, it remains to ask what is meant by substance. It is not denied that there are material substances, but their qualification both as substances and as material will be found to depend on relations. By a substance we mean that which is persistent throughout certain appearances. It represents that identical element throughout the appearances, that permanent element throughout the times of their appearance, in virtue of which they are not merely so many different appearances, but connected changes. A material substance is that which remains the same with itself in respect of some of the qualities which we include in our definition of matter--qualities all consisting in some kind of relation--while in other respects it changes. Its character as a substance depends on that relation of appearances to each other in a single order which renders them changes. It is not that first there is a substance, and that then certain changes of it ensue. The substance is the implication of the changes, and has no existence otherwise. Apart from the changes no substance, any more than apart from effects a cause. If we choose to say then that matter exists as a substance, we merely substitute for the designation of it as consisting in relations, a designation of it as a certain correlatum of a certain kind of relation. Its existence as a substance depends on the action of the same self-consciousness upon which the connexion of phenomena by means of that relation depends. (§ 53 ¶ 1)

And the subject, of which the action is implied in the connexion of phenomena in one system of nature by means of this correlatum of change, is one that can itself be as little identified with that correlatum--with any kind of substance--as with the change to which substance is relative. It has already been pointed out that a consciousness, to which events are to appear as changes, cannot itself consist in those events. Its self-distinction from them all is necessary to its holding them all together as related to each other in the way of change. And, for the same reason, that connexion of all phenomena as changes of one world which is implied in the unity of intelligent experience, cannot be the work of anything which is the substance qualified by those changes. Its self-distinction from them, which is the condition of their appearance to it under the relation of change, is incompatible with being so qualified. Even if we allow it to be possible that a subject, which connects certain appearances as changes, should itself be qualified by--should be the substance persistent in--certain other changes, it is plainly impossible that a subject which so connects all the appearances of nature should be related in the way of substance to any or all of them. (§ 53 ¶ 2)