culties. Here I would add merely that I have presupposed nothing except that which I take to be present in principle at the very lowest level of mind.
Now, so much being assumed, it is no great step to advance from it to serial connexions. Wherever A tends to call up B and B to bring in C, A being present will tend to produce the series A-B-C. The means and the condition of this mediate connexion is the identity of B. There is here a common link which is one and the same, or which at least somehow behaves as if it were so, and which also again on examination seems so. Without this identical link there is certainly no series at all, but how far its identity must be perfect is a further question to be considered later. And at this point there arises the difficult and most important problem about the unity of the whole series, a problem at which I shall be able to do no more than glance.
But when once we have such series joined by common links, it seems easy from this point to proceed to the future and past and to transcend the present. For given the disposition to an ideal series such as c-d-e, and given on the other side a present qualified as A(b-c), there is, through the identity of c, a transition from A to e through b-c-d. And with this transition memory, it might be said, is at once explained. Now in principle I think memory is so explained, and the explanation is correct, but it on the other hand is insufficient, and takes no account of serious differences. For in the first place memory has perforce to go backwards if it is to reach the past, while our series, it seems, run all the other way, and we can only think forwards. And in the second place memory is certainly not the mere extension of the present. It gives us rather something which is not the present, something which is known as different and incompatible. I will proceed briefly to discuss these two difficulties, beginning with the second.
To know the past or future as such is a hard and late achievement of the mind, for it implies an enormous degradation of the present. We do not properly represent the past or future until we have gained an order of things in which the present has become but one thing among others. These other things, not the present, are not presented, and, if by a miracle they were so while the present itself still remained untransformed, the result would be chaos. But past and future do not and cannot exist for us until reality appears as a series in which the present has sunk and has become but one member among others. Such an order is an array into the ranks of which the present is cashiered, it is an order
