day’s self, which I construct, did, as such, have an actual existence in the past. The concrete conditions, into which my ideal construction must be launched, may alter its character. They may, in fact, unite with it so that, if I knew this unknown fact, I should no longer care to call it my self. Thus my past self, assuredly, is not demonstrated. We can but say of it that, like other selves, it is practically certain. And in each case the result, and our way to it, is in principle the same. Both other selves and my own self are intellectual constructions, each as secure as we can expect special facts to be. But, if any one stands out for demonstration, then neither is demonstrated. And, if this demand is pressed, you must remain with a feeling about which you can say nothing, and which is, emphatically, not the self of any one at all. On the other hand, if you are willing to accept a result which is not strictly proved, both results must be accepted. For the process, which conducts you to other selves, is not weaker sensibly, if at all, than the construction by which your own self is gained. On either alternative the conclusion of Solipsism is ruined.
And if memory, or some other faculty, is appealed to, and is invoked to secure the pre-eminent reality of my self, I must decline to be persuaded. For I am convinced that such convenient wonders do not exist, and that no one has any sufficient excuse for accepting them. Memory is plainly a construction from the ground of the present. It is throughout inferential, and is certainly fallible; and its gross mistakes as to past personal existence should be very well known (pp. 84, 213). I prefer, in passing, to notice that confusion as to the present limits of self, which is so familiar a feature in hypnotic experiments. The assumption of a suggested foreign personality is, I think, strong evidence for the secondary nature of our own. Both, in short, are
