IV. There may be no causation either way, because the association may be due to a harmony pre-established by a superior mind.

V. There may be no causation either way, because the association may always be due to chance.

VI. There may be no causation either way, because the material order may not have any real existence at all, being merely an ideal creation of the mental order.

VII. Whether or not there be any causation either way, the association may be one which it is necessarily beyond the power of the human mind to explain.

So far as I can see, this list of possible answers to the question before us is exhaustive. I will next show why, in my opinion, the last four of them may be excluded in limine.

The suggestion of pre-established harmony (IV) merely postpones the question: it assumes a higher mind as adjusting correspondencies between known minds and animal bodies with respect to the activities of each; and, therefore, it either leaves untouched the ultimate question concerning the relation of mind (as such) to matter, or else it answers this question in terms of spiritualism (I).

The suggestion of chance (V) is effectually excluded by the doctrine of chances: even in any one individual mind, the association between mental changes and material changes is much too intimate, constant, and detailed to admit of any one reasonably supposing that it can be due only to chance.

The suggestion of pure idealism (VI) ultimately implies that the thinking Ego is itself the sole existence—a position which cannot, indeed, be turned by any assault of logic; but one which is nevertheless too obviously opposed to common sense to admit of any serious defence; its immunity from direct attack arises only from the gratuitous nature of its challenge to prove a negative (namely, that the thinking Ego is not the sole existence), and this a negative which is necessarily beyond the region of proof.

Lastly, the suggestion that the problem is necessarily insoluble (VII) does not deserve to be regarded as an hypothesis at all; for to suppose that the problem is necessarily insoluble is merely to exclude the supposition of there being any hypothesis available.

In view of these several considerations, it appears to me that, although in a formal sense we may say there are altogether seven possible answers to the question before us, in reality, or for the purposes of practical discussion, there are nowadays but three—namely those which head the above list, and which I will now proceed to consider.