it will be well in the next place, before inquiring more closely into their characteristics, to consider for a moment that persistence of preceding modifications which the principle of progressive differentiation implies. Such persistence is best spoken of as due to retentiveness. This is often confused with memory, though memory is something much more complex and special; for in that there is necessarily some contrast of past and present, whereas here there is simply the persistence of the old. But what is it that persists? On our theory we must answer, the continuum as differentiated, not the particular differentiation as an isolated unit. If psychologists have erred in regarding the presentations of one moment as merely a plurality of units, they have erred in like manner concerning the so-called 'residua' of such presentations. As we see a certain colour or a certain figure again and again, we do not go on accumulating images or representations of it, which are somewhere crowded together like shades on the banks of the Styx. Nor is such colour, or whatever it be, the same at the hundredth time of presentation as at the first, as the hundredth impression of a seal on wax might be. There is no such lifeless fixity in mind. The explanations of perception most in vogue are far too mechanical and, so to say, atomistic; but we must fall back upon the continuity of our presentation-continuum, to get a better.

Suppose, then, that in the course of a few minutes we take half a dozen glances at a strange and curious flower. We have not as many complex presentations, which we might symbolize as F1, F2, ... F. But rather, at first only the general outline is noted, next the disposition of petals, stamens, &c., then the attachment of the anthers, position of the ovary, and so on; that is to say, symbolizing the whole flower as [p' (a b) s' (c d) o' (ƒg)], we first apprehend say [p...s'... o'], then [p' (ab) s'... o'], or [p' (a.....) s' (c...) o' (ƒ.....)], and so forth. It is because the traits first attended to persist that those noticed later form an addition to them so that the complex at length may be complete. There is nothing in this instance properly answering to what are known as the reproduction and association of ideas; in the last and complete apprehension as much as in the first vague and inchoate one the flower is there as a primary presentation. There is a limit, of course, to such a procedure, but the instance taken, we may safely say, is not such as to exceed the bounds of a simultaneous

W. P.

6

field of consciousness. We assume, then, that such increase of differentiation through the persistence of preceding differentiations holds of the presentation-continuum as a whole. Next, we conclude that, in those circumstances in which we now have a specific sensation of, say, red or sweet, there would be for some more primitive experience nothing but a vague, almost 'organic,' sensation, which, however, on every repetition of the circumstances, would become somewhat further differentiated. The earlier differentiations, in short, do not disappear like the waves of yesterday in the calm of to-day, nor yet last on like old scars beside new ones; but rather the two are combined, so that the whole field of consciousness, like a continually growing picture, increases indefinitely in complexity of pattern.

Assimilation.

§ 4. This process, in which later differentiations seem to 'blend with' and thereby further restrict and specialise what is retained of earlier and less definite presentations, is thus a further implication of the principle of the progressive development of the presentational continuum. When not ignored altogether, this further process has been commonly regarded as merely a simple form of 'association,' its peculiarity being, as it was supposed, that the presentations associated-though numerically distinctwere in quality perfectly identical. In point of fact, both these assumptions seem to be erroneous and due to the so-called psychologist's fallacy'. For the experiencing subject there is apparently at this stage-as we have already urged-neither the numerical distinctness nor the qualitative identity which the words 'past impression (A,)' and 'present impression (A2)' suggest. Still the connexion between the process of association proper and the process of mere 'blending or fusion '-as it is frequently termed, though we shall call it assimilation-is so close, and the detailed analysis called for so complex, that we must needs defer further discussion till we come to treat of association as a whole. It may then be possible to shew that we have here to do with a process much simpler and more fundamental

1 As, e.g., in interpreting the conduct of children as if they were already 'grown-up' persons.

2 Cf. below, ch. vii, § 2.

than association. But it is at least clear at once that, if the term association is to be correctly used, it must imply that the presentations associated were from the first distinct, were attended to as distinct, became associated solely in consequence of such attention, and remain to the last distinguishable. Herbert Spencer seems to have been the first psychologist to appreciate the elementary character of this process, which-so far from being a form of true association-is presupposed in all association properly so called. He names it 'automatic association.' "This association," he says, "is not an act of thought [better to have said 'a result of an act of attention'] that may or may not take place, but constitutes the very recognition of each feeling [= sensation]. A feeling cannot form an element of mind at all, save on condition of being associated with predecessors more or less the same in nature.... All other phenomena of association of feelings are consequent on the union of this process with a parallel and simultaneous process to be described later." In the course of his exposition Mr Spencer frequently uses 'assimilation' as a variant for his technical term ' automatic association'; and assimilation is the term here adopted for the process1.

In view of the intimate connexion between differentiation, retentiveness and assimilation it will sometimes be convenient to refer to all three together as constituting what we may call the plasticity of the presentational continuum.

Relativity.

§ 5. This will be the most convenient place to take note of certain psychological doctrines which, though differing in some material respects, are usually included under the term Law of Relativity.

1 Principles of Psychology, §§ 115 ff. In ignorance of Mr Spencer's usage I myself proposed this term and expressed the hope that it might find general acceptance (Ency. Brit. 9th ed. art. Psychology, p. 52). I first became aware of Spencer's priority in reading Benno Erdmann's paper, "Zur Theorie der Apperception," already referred to (p. 46). He contends that Herbart's term 'apperception' is the more appropriate and also that to Herbart, as the discoverer of the process, and not to Spencer, the right to coin a name for it must be conceded. But unfortunately we shall find it needful to restrict the Herbartian term equally with the term association to much narrower limits. As to 'assimilation'-I have since come across it in Drobisch's Empirische Psychologie, 1842, p. 142, fin.

a.

Hobbes's Sentire semper idem et non sentire ad idem recidunt1 is often cited as one of the first formulations of this law. If we take this to apply to the whole field of consciousness it becomes at once true and trite; for a field of consciousness unaltered either by change of impression or of idea would certainly be a blank and a contradiction. The Law of Relativity in this sense is in fact what Hamilton called the Law of Variety: "that we are conscious only as we are conscious of difference"-i.e. of variety or change. But, though consciousness involves change, it is still possible that particular presentations may continue in the field of consciousness indefinitely. When it is said that "a constant impression is the same as a blank," what is meant sometimes turns out to be something not psychological at all, as, e.g., our insensibility to the motion. of the earth or to the pressure of the air-cases in which there is obviously no presentation, nor even any evidence of nervous changes. Sometimes this paradox proves to be but an awkward way of expressing what we may call accommodation, whether physiological or psychological. Thus the skin soon adapts itself to certain seasonal alterations of temperature, so that heat or cold ceases to be felt: the sensation ceases because the nervous change, its proximate physical counterpart, has ceased. Again, there is what James Mill called 'an acquired incapacity of attention,' such that a constant noise, for example, like the clatter of a weaver's loom, in which one has no interest, is soon unnoticed. As a rule, no doubt, impressions do not continue constant for more than a very short time; still there are sad instances enough in the history of disease, bodily and mental, to shew that such a thing can quite well happen, and that such constant impressions (and 'fixed ideas,' which are in effect tantamount to them), instead of becoming blanks, may dominate the entire consciousness, colouring or bewildering everything.

b. From the fact that the field of consciousness is continually changing it has been supposed to follow that every presentation is essentially nothing but a transition or difference. "All feeling," says Bain, the leading exponent of this view, "is two-sided....

1 Elementa philosophica, IV. xxv. 5.

2 The Works of Thos. Reid, Supplementary note, p. 932.

3 Yet these were given as 'notable examples' of this law by Bain (Senses and Intellect, 3rd ed., p. 9) but afterwards suppressed in view of the criticism in the text.

We may attend more to one member of the couple than to the other....We are more conscious of heat when passing to a higher temperature, and of cold when passing to a lower. The state we have passed to is our explicit consciousness, the state we have passed from is our implicit consciousness1." But the transition need not be from heat to cold, or vice versa: it can equally well take place from a neutral state, which is indeed the normal state, of neither heat nor cold; a new-born mammal, e.g., must experience cold, having never experienced heat. Again, suppose a sailor becalmed gazing for a whole morning upon a stretch of sea and sky, what sensations are implicit here? Shall we say yellow as the greatest contrast to blue, or darkness as the contrary of light, or both? What, again, is the implicit consciousness when the explicit is sweet; is it bitter or sour, and from what is the transition in such a case? For one thing it seems clear that the transition of attention from one presentation to another and the differences between the presentations themselves are distinct facts. It is strange that Bain, the psychologist who has laid such stress on neutral states of surprise as being akin to feeling and so distinct from special presentations, should in any way confound the two. The mistake is perhaps accounted for by the fact that, in common with the rest of his school, Bain failed adequately to distinguish between attention and the presentations that are attended to. If 'change of impression' and being conscious or mentally alive are the same thing, it is then manifestly tautologous to say that one is the indispensable condition of the other. If they are not the same thing, then the succession of shocks or surprises cannot wholly determine the impressions which successively determine them.

But we have still to consider whether the impressions themselves are nothing but differences or contrasts. "We do not know any one thing of itself but only the difference between it and another thing?," said Bain. But it is plain we cannot speak of contrast or difference between two states or things as a contrast or difference, if the states or things are not themselves presented; the so-called contrast or difference would then be itself a single presentation, and its supposed 'relativity' but an inference. Difference is not more necessary to the presentation 1 Logic, i. 1870, p. 3.

2 Senses and Intellect, 3rd ed., p. 321.