lition of 1) to the certain sense dewever, be written i.e. that it becomes must not, of course, service. There is compels us of in

The effect is asesult of past direct perhaps the matter reometrical analogy. follows of inherent rcumference on any conception flows not Cogical necessity from pressions involves in nce might be chaotic it has become for us The noteworthy not so much in the

because it was buried in the individuality of another person. We have further noticed that as will and motion are more carefully analysed, the conception that will originates motion ceases to have any consistency. But with will as first cause falls to the ground any possible experience of first causes on our part. We can no longer infer even the possibility of the existence of first causes, for there is nothing like them in our experience, and we cannot by the second canon of logical inference (p. 60) pass from the known to something totally unlike it in the unknown. Science knows nothing of first causes. They cannot, as Stanley Jevons has supposed,' be inferred from any branch of scientific investigation, and where we see them asserted we may be quite sure they mark a permanent or temporary limit to knowledge. We are either inferring something in the beyond of sense-impression, where knowledge and inference are meaningless words, or we are implying ignorance within the sphere of knowledge, in which case it is more honest to say: "Here, for the present, our ignorance begins," than, " Here is a first cause."

§ 8.-Cause and Effect as the Routine of Experience

We are now in a position, I think, to appreciate the scientific value of the word cause. Scientifically, cause, as originating or enforcing a particular sequence of perceptions, is meaningless—we have no experience of anything which originates or enforces something else. Cause, however, used to mark a stage in a routine, is a clear and

1 In the remarkably unscientific chapter entitled "Reflections on the Results and Limits of Scientific Method," with which his, in so many respects, excellent Principles of Science concludes.

2 The latter alternative-the temporary limit in ignorance—has been the chief source of "first causes.". So long as the routine of history cannot be traced back more than a few centuries, we find no difficulty in asserting that the world began 6000 years ago. So long as we do not grasp the evolution of life from its most primitive types, we postulate a first cause creating each type (Paley). So long as we do not observe the various grades of animal intelligence and consciousness, we suppose a soul implanted in every human being at birth. So long as we do not see that the mutual motion of two atoms is as mysterious as the life changes in a cell, we postulate a total difference between the two kinds of motion and a separate creation of life.

valuable conception, which throws the idea of cause entirely into the field of sense-impressions, into the sphere where we can reason and can reach knowledge. Cause, in this sense, is a stage in a routine of experience, and not one in a routine of inherent necessity. The distinction. is, perhaps, a difficult one, but it is all the more needful that the reader should fully grasp it. If I write down a hundred numbers at chance-say by opening carelessly the pages of a book-there results a sequence of numbers beginning, say

141, 253, 73, 477, 187, 585, 57, 353, . etc.,

in which I cannot predict from any two or three or more numbers those which will follow. The number 477 does not enable me to say that 187 will follow it, the numbers which precede 187 in no way enforce or determine those which follow it. On the other hand, if I take the series

1, 2, 3, 4, 5, 6, 7, 8,

each individual number leads (by addition of 1) to the immediately following number, or in a certain sense determines it. The first series can, however, be written down so often that we learn it by rote, ie. that it becomes a routine of experience. The analogy must not, of course, be pressed far, but it may still be of service. There is nothing in any scientific cause which compels us of inherent necessity to predict the effect. The effect is associated with the cause simply as a result of past direct or indirect experience. Or again, perhaps the matter may be grasped more clearly from a geometrical analogy. If I form the conception of a circle, it follows of inherent necessity that the angle at the circumference on any diameter is a right-angle. The one conception flows not as a result of experience but as a logical necessity from the other. No sequence of sense-impressions involves in itself a logical necessity. The sequence might be chaotic. like our first series of numbers; it has become for us a routine by repeated experience. The noteworthy fact in a routine of perceptions lies not so much in the

particular order of the stages in the sequence as in the result of experience that this order can exactly repeat itself.

The reader may perhaps wonder how, if the sequences of sense-impressions are really of the chaotic nature represented by our first series of numbers, it is possible to describe such sequences apart from their repetition by those brief formulæ we term scientific laws. As the per

ceptive faculty presents us, indeed, with the sequence, it is undeniably more like the second than the first series of numbers, for natural phenomena can without doubt be largely described by certain brief laws. We must rather put the actual case in the following form. We observe a person whose motives are quite unknown to us writing down the series—

1, 2, 4, 8, 16, 32,

and at present he has reached the number 32. A law describing the series is obvious-each number is twice the preceding one. With a great degree of probability we infer that he will now write down 64, especially if we have seen him write the series up to and beyond 32 before. But there is nothing of logical necessity about his writing 64 after the preceding numbers. Those numbers, when we know the law, suggest his doing so, but do not enforce it.

We are now in a position to scientifically define cause. Whenever a sequence of perceptions D, E, F, G is invariably preceded by the perception C, or the perceptions C, D, E, F, G always occur in this order, that is, form a routine of experience, C is said to be a cause of D, E, F, G, which are then described as its effects. No phenomenon or stage in a sequence has only one cause, all antecedent stages are successive causes, and, as science has no reason to infer a first cause, the succession of causes can be carried back to the limit of existing knowledge, and beyond that ad infinitum in the field of conceivable know ledge. When we scientifically state causes we are really describing the successive stages of a routine of experience.

Causation, says John Stuart Mill, is uniform1 antecedence, and this definition is perfectly in accord with the scientific concept.

$9.-Width of the Term Cause

The word cause, even in its scientific sense, is somewhat clastic. It has been used to mark uniform conjunction in space as well as uniform antecedence in time; while if we take an actually existing group of perceptions, say the particular ash-tree in my garden, the causes of its growth might be widened out into a description of the various past stages of the universe. One of the causes of its growth is the existence of my garden, which is conditioned by the existence of the metropolis; another cause is the nature of the soil, gravel approaching the edge of the clay, which again is conditioned by the geological structure and past history of the earth. The causes of any individual thing thus widen out into the unmanage able history of the universe. The ash-tree is like Tennyson's "flower in the crannied wall": to know all its causes would be to know the universe. To trace causes in this sense is like tracing back all the lines of ancestry which converge in one individual; we soon reach a point where we can go no further owing to the bulk of the material. Obviously science in tracing causes attempts no task of this character, but at the same time it is useful to remember how essentially the causes of any finite portions. of the universe lead us irresistibly to the history of the universe as a whole. This thought suggests how closely knit together are in reality the most diverse branches of our positive knowledge. It shows us how difficult it is for the great building of science to advance rapidly and surely unless its various parts keep pace with each other (p. 13). Practically science has to content itself with tracing one line of ancestry, one range of causes at a time, and this not for a special and individual object like the ash-tree in my garden, but for ash-trees or even trees in

1 "Uniformity" and "sameness" are, in the perceptual world, however, only relative terms (see Chapter V. § 6).