446 CKITICAL NOTICES :

mind a priori." Closely connected with this divergence from Kant as regards the evidence of the causal judgment is another, which concerns its nature. This, for Prof. Knight, does not j consist in the assertion of an irreversible rule of sequence, but in the necessity that "every effect must have a cause," and that J " power is lodged within the cause adequate to produce the effect". On thefirst head. Hume is entirely in agreement with his critic^ "He acknowledges freely that " every effectTmust have a cause," 'inasmuch as this is a relation included in the nature of the ideas related. Hume's own illustration is " tTmtTTir ncT more follows from~this maxim that all events have causes, than it follows because every husband has a wife, that therefore every man must be married ". How would Prof. Knight prove that every event is, in the sense intended, an effect? How would he show that "within every atom, as its interior essence, this force or causal power resides" ? His answer would seem to be that " this link of power," this " inner tie/' is disclosed to the reason ; we discover it by an " jntuition of the^reason ". IJQJV it was just this direct intuition for which Hume searched so diligently, and Searched in vain. His conclusion was that. " if we examine this maxim, we shall find in it no mark of intuitive certainty ; but, on the contrary, shall find that 'tis of a nature quite foreign to that species of conviction". If Prof. Knight has been more successful in this quest, he ought to have explained the nature of his I intuition, and of its claims to validity. He ought to have shown that he has not mistaken an internal impression, psychologically generated, and afterwards projected by subreption, for an objective relation directly apprehended. Instead of this, he has contented himself with reiterating that the judgment is what he calls a priori, and seems not in the least aware that he is thereby handing back to Hume his own problem in place of a solution. I am far from saying that Hume has said the last word on this sub- ject ; but I utterly fail to see that Prof. Knight's criticism touches his position, or that it is even relevant to the issue raised. On the question of personal identity, Prof. Knight attacks Hume exactly where Kantian criticism has reinforced his i conclusion. The position maintained in this book is that the II existence of a series of impressions and ideas implies the 1 existence of a permanent identical substratum. The arguments I used do not seem very conclusive. Thus (p. 178) Prof. Knight triumphantly asks, " if all that I am is this series of successive and detached impressions which I subsequently recall . . how are they my impressions and my ideas ? " The obvious answer is that, if by * myself ' I mean a series of ideas, in calling an idea ' mine ' I must mean that it is part of this series. It is moreover difficult to see how a substrate can perform the work [ which Prof. Knight assigns it. If a succession of states of mind I cannot of itself yield personal identity, it is far from clear how the case is improved by substituting a succession of changes in a permanent underlying entity.