to the use even of 'somatic,' if somatic be taken to imply any experience of the distinction of the organism from external bodies. On the other hand those who prefer to speak of general feeling (Gemeingefühl) rather than of general sensation (Gemeinempfindung) or to use the two terms indifferently are in the opposite extreme, as has been already said1, if they assume that experience consists primarily of purely affective states (Zustände) without objective antecedents or consequents or if they regard the two as originally identical. What is obviously lacking at the outset, when differentiation is still inchoate, is not sensation as objective in distinction from feeling as subjective, but rather the specific objective diversity which advancing differentiation brings. But the vagueness and generality of the experience described is no reason for confusing the concepts used in its description. Again, though less definitely discriminated, the earlier, and what we call the lower, sensations are not any less concrete than the later and higher. They have been called general rather than specific; not because psychologically they lack any essential characteristic of sensation which those acquired later possess ; but simply because physiologically they are not, like these, correlated to special sense-organs.

Short, however, of resolving such sensations into combinations of one primordial modification of consciousness, if we could conceive such, there are many interesting facts which point clearly to a complexity that we can seldom directly detect. Several of our supposed sensations of taste, e.g., are complicated with sensations of touch and smell: thus the pungency of pepper and the dryness of wine are tactual sensations, and their spicy flavours are really smells. How largely smells mingle with what we ordinarily take to be simply tastes is effectively brought home to us by a severe cold in the head, as this temporarily prevents the access of exhalations to the olfactory surfaces. The difference between the smooth feel of a polished surface and the roughness of one that is unpolished, though to direct introspection an irresolvable difference of quality, probably answers to the fact that several nerve-terminations are excited in each case: where the sensation is one of smoothness all are stimulated equally; but where it is one of roughness the ridges compress the nerve-ends more, and the hollows compress them 1 Cf. ch. ii, § 3, PP. 44 f.

less, than the level parts do. Hence we infer that such sensations are really compounds of several.

The most striking instance in point, however, is furnished by the differences in musical sounds, to which the name timbre is given. To the inattentive or uninstructed ear notes or 'compound tones' appear to be only qualitatively diverse and not to be complexes of simple tones. Yet it is possible with attention and practice to distinguish these, as 'partial tones,' in a note produced on one instrument, a horn, say, and to recognise that they are different from those of the same note produced on a different instrument, for example, a violin. In like manner many persons believe that they can discriminate in certain colours, hence called 'mixed,' the elementary colours of which they are held to be composed; red and yellow, for example, in orange, or blue and red in violet; and the vocabularies of most languages seem to bear them out in the frequency of terms such as bluish green, yellowish green, and the like. It is at any rate certain that orange resembles red on the one hand and yellow on the other; it very naturally therefore reminds us of these colours, between which in the colour spectrum it invariably stands. But it is also certain that we cannot distinguish two colours in orange or purple in the sense in which we distinguish partial tones in a note or notes in a chord. Yes, it may be replied, but that only amounts to saying that the complex colour is not a plurality: it does not prove that it is not a blend or mixture of simple or primary colours-which is all that is maintained. In other words the note, like the chord, is a sensation-complex, the secondary colour is a complex sensation. If now from the fact that such a secondary colour resembles the primary colours on either side of it we could straightway infer that it must consist of both, the question would be positively settled. To many this has seemed a valid argument; nevertheless, as we shall see later on1, in the particular case of sensory continua this argument fails to apply. But we may see at once that if this argument were generally valid it would force us to conclude that a tone, since this also resembles the two between which it is intermediate, ought to be a blend of both; whereas, in fact, as Ebbinghaus pointed out, the tone d, though as regards pitch having a certain 1 Cf. below, ch. xiii, § 2.

2 H. Ebbinghaus, Grundzüge der Psychologie, 3te Auf., 1911, i. p. 201 fin.

resemblance to c and e, its neighbours on either side, differs widely from the chord c―e, which is made up of these. Nay further, so far as bare resemblance is concerned, the argument in question ought to lead us to conclude that red is complex, for it resembles purple on the one side and orange on the other. Thus even if we could argue from mutual resemblance to complexity we should still have to determine where the complexity lay; in the orange or in the red. It is not, however, enough to know that there may be two physical or two physiological processes, or both, concerned in the sensation of orange, whereas in the sensation of (saturated) red, these processes are always single. The one thing essential after all is that in the sensation of orange its components-if it be a compound-should be in some sort distinguishable. Mixture' in any literal sense of the word is not a term appropriate to psychology; and hence-so far as immediate experience is concerned-we seem driven to deny the existence of complex sensations and to recognise only sensation-complexes.

In all cases where the presence of such sensation-complexes is beyond dispute the partial sensations can be distinguished by discernible differences of extensity (and often of intensity). Thus if the skin be touched by the point of a hot or cold. bradawl, the temperature sensation has not the punctual character of the touch, but seems rather to surround this as a sort of penumbra. Similarly the ground-tone of a clang-complex has not only a greater intensity but also a greater extensity than any of the over-tones1. There is too in such cases a certain rivalry or antagonism between the complex as an unanalysed whole and the complex as analysed, and even between the several partial sensations after such analysis. Such differences are no doubt often due to differences in the distribution of attention brought about by practice, expectation, interest, and the like; but they are sometimes due to physiological variations in stimulation consequent on partial exhaustion or recuperation2: both alike however point to the underlying presentational complexity. In the absence of such evidence it is unwarrantable to infer psychical complexity from complexity in the physical stimuli or in the processes which they immediately set up. 1 Cf. Stumpf, Tonpsychologie, ii. pp. 58 f.

W. P.

2 Cf. Stumpf, op. cit., i. pp. 360 ff.

8

White light, for example, is physically the most complex of all light, whereas the sensation of white is not only simple but probably the most primitive of our visual sensations. It is difficult to give any clear meaning to the statement that two sensations become one sensation or that one sensation has two qualities. It seems best therefore to define a sensation as the simplest element in our analysis of the objective factor in experience. It is complex, indeed, inasmuch as it has a plurality of characteristics, but any further complexity would seem to be most appropriately described as due to a plurality of sensations, since the only evidence of such further complexity that is psychologically admissible is a discrimination of qualities.

We find, however, some indirect evidence of the complexity of our existing sensations in the variations in quality that in certain special cases accompany variations in intensity, extensity, and duration. With the exception of (saturated) red, all spectral colours1 give place, sooner or later, to a mere colourless grey as the intensity of the light diminishes, and all in like manner become indistinguishably white after a certain increase of intensity. A longer time is also in most cases necessary to produce a sensation of colour than to produce a sensation merely of light or brightness: the solar spectrum seen for a moment appears not of seven colours but of two only-faintly red towards one end and blue towards the other. Very small objects, again, such as coloured specks on a white ground, though still distinctly seen, appear as colourless if of less than a certain size: the relation between their intensity and extensity being such that within certain limits the intenser they are the smaller they may be without losing colour, and the larger they are the fainter in like manner. Similar facts are observable in the case of other senses, so that generally we seem justified in regarding what we now distinguish as a sensation as probably complicated in several respects. In other words, if psychical magnification were possible, we might be directly aware that sensations which we now regard as simple were really complexes-that they consisted, that is, of two or more sensational elements or changes, different in quality, of uniform or variable intensity, and occurring either simultaneously or in regular or irregular succession. So much for the

1 The light is supposed to be thrown on a perfectly black ground. Cf. v. Kries, Die Gesichts-empfindungen und ihrer Analyse, 1882, pp. 81, 82.

general nature of sensations; we have next to consider (1) their quantitative and (2) their qualitative properties in more detail.

Quantitative Continuity.

§ 4. Every sensation within the fields of consciousness has sensibly some continuous duration and seems sensibly to admit of some continuous variation in intensity and extensity. But whether this quantitative continuity of presentational change is more than apparent has been questioned. Sensations of almost liminal intensity are found to fluctuate every few seconds, and, as already remarked, when the threshold of intensity is actually reached, they seem intermittently to appear and disappear, a fact which Hume long ago did not fail to notice'. The results of numerous experiments, however, justify the conclusion that these variations are due primarily to oscillation of attention, and furnish so far no ground for the assumption that even the liminal sensation is discontinuous. Again, we can only detect a difference of intensity when this is of finite amount and bears a certain constant ratio to the initial intensity with which it is compared a fact commonly known as Weber's Law-so that, although the stimulus may be augmenting continuously, increments in the intensity are only apprehended per saltum. This imperfection in our power of discrimination is, however, no proof that our sensations vary discontinuously; and not only is there no positive evidence in favour of such discontinuity, but it is altogether improbable on general grounds. Lastly, there is always more or less distinctness in the several nerve-endings as well as isolation of the nerve-fibres and neurons themselves. The skin, for example, when carefully explored, turns out to be a complex mosaic of so-called 'spots,' severally responding to stimulation by sensations of pressure, heat, cold, and pain. But from this to argue that the extensity of a sensation is really a mere aggregate without any continuity is on a par with calling a lake a collection of pools because it is fed by separate streams. If it could be shewn that in the brain as a whole there is no functional continuity a formidable psychophysical problem would no doubt arise. Meanwhile, however, whatever the number of nerveendings or of neurons with which it is correlated, there is nothing 1 Treatise of Human Nature, Green and Grose's edn., i. p. 347 fin.