of kind between being and being consciously perceived'. (Matter and Memory, p. 30.) According to M. Bergson we have not to deduce consciousness, because we have assumed it in a world of objects acting and reacting upon each other. It is, however, selectiveness and indetermination that give rise to 'discernment'. Consciousness is the light which plays around the zone of possible actions or potential activity which surrounds the action really performed by the living being. It signifies hesitation, choice.' (Creative Evolution, p. 152.) They' (living beings) 'allow to pass through them, so to speak, those external influences which are indifferent to them; the others, isolated, become "perceptions" by their very isolation.' (Matter and Memory, p. 29.) Such is the doctrine of pure perception.

Such perception, however, is merely a convenient fiction to illustrate the relation of the percipients' body to other objects in the world of change. It is a fiction, because in the first place it neglects duration. Even the act 'whereby we place ourselves in the very heart of things' is not instantaneous. In order to understand how an uninterrupted series of instantaneous visions which would be a part of things rather than ourselves' can become continuous duration, we have to invoke memory. In the second place, the choice of response which constitutes discernment is guided by memory. It is past experience which determines the actual response made. Thus perception without memory is a fiction.

M. Bergson proceeds to consider recognition where perception and memory are interlaced. Recognition is often an affair of action. It means knowing what to do, and does not necessarily involve the representation of past experience. If images occur their function is to interpret the present with a view to action. Failure of recognition is often merely a failure in motor habits, a failure of habit memory. None the less there is a type of recognition which does involve the spontaneous recall of images, and it has been held that failure

of such recognition is due to the loss of memory-images. The different varieties of aphasia were regarded as demonstrating this Sensory Aphasia, wherein the patient fails to recognize spoken or written words, illustrating the loss of auditory or visual images; Motor Aphasia, wherein the patient is unable to speak though free from any defect in the requisite muscles, illustrating the loss of the images of articulated words. It was on this interpretation of aphasia that M. Bergson made his great attack. Attentive recognition, under which rubric the recognition of words may fall, involves, according to M. Bergson, an attitude of the body, an attitude of inhibition or strain, and so far M. Ribot was right in describing attention as a muscular experience. But this attitude is only preparatory. It prepares for the positive movements initiating perception, and for the invocation of the interpretative memory-images which finally determine perception. Every attentive perception truly involves a reflexion, in the etymological sense of the word, that is to say the projection, outside ourselves, of an actively created image, identical with, or similar to, the object on which it comes to mould itself.' (ibid., p. 124.) M. Bergson would attribute the failure in spontaneous speech, or in understanding spoken or written language, to inability to adopt, under the influence of the external stimulus, the precise attitude by means of which a choice could automatically be made among our memories'. (ibid., p. 132.)

Recognition can thus break down at two different ends of the motor process. In attentive recognition it breaks down through failure in the fine bodily adjustments requisite for the initiation of imagery; in automatic recognition, through failure in the bodily movements requisite for the appropriate response. In neither case is there any loss of, or impairment in, the store of relevant images. M. Bergson works out the absurdities consequent on any hypothesis of images as something stored in brain cells, and of brain lesions as destructive of images. The nearest approach to a theory of an image centre in

the brain which M. Bergson makes is the suggestion that there may be something, metaphorically called a keyboard, which may be played upon by memory in the same way as the sense organs are played upon by sense stimuli, and which may affect the same nerves and brain centres as the sense organ itself.

From the consideration of memory interwoven with perception M. Bergson passes to his theory of Pure Memory. 'Coextensive with consciousness, it' (true memory) 'retains and ranges alongside of each other all our states in the order in which they occur, leaving to each fact its place and consequently marking its date.' (ibid., p. 195.) The whole of the individual's past experience exists in pure memory. But as in pure perception only a fragment of the activities of the whole universe is selected by our body, so in recollection only a fragment of the past is represented in memory. As the fragment selected in perception is continuous with the whole universe, so the fragment represented in memory is continuous with the whole past. We have no difficulty in conceiving of the unperceived existence of the sense-world, and we ought to find none in conceiving of the unconscious existence of the past, since the unconscious plays in each case a similar part.' (ibid., p. 187.)

Pure memory of the past is, as such, powerless, it is detached from the actual. To be of use it must become potential action and share to some degree the character of sensation. Our present is sensori-motor; it is our attitude towards the impending future. To insert itself into the present, the past preserved in pure memory must become an image. M. Bergson makes use of the figure of an inverted cone to illustrate his conception of the relation of pure memory to memory-images, and of both to perception. The apex of the cone resting on the plane of change stands for the response of the body in perception. Pressing into the apex are the images of memory which guide such response. Behind them at the base of the cone lies pure memory. Sections across the cone between

base and apex at different levels would represent different degrees of what M. Bergson calls 'psychic tension'. At the base all the events of the individual's past life are set out in their smallest detail. Only the man who dreams, detached from the needs of present action, can live at this level. At the apex the responses to the world of change are immediate and automatic. Between the two extremes lie varying degrees of imagery and memory.

The so-called laws of association are for M. Bergson the different ways in which memory supplies the needs of the present. Perception gives knowledge neither of the particular nor of the general. For action the discernment of individualizing features is of no importance. The perception only becomes particular' when it is interpreted by a memoryimage. The records of pure memory, however, are individual, and it is from these that the class concept is built up by analysis. Such concepts, M. Bergson says, move continually between the plane of memory and the plane of action. They tend either to become the individuals which they epitomize or to give expression to their basic common feature through action.

M. Bergson insists upon the distinction between pure memory and pure perception, the former as the work of the soul, the latter as the work of the body. It is true that it is merely convenience of method which leads him to consider pure perception apart from memory, and pure memory apart from perception, but the abstractions adopted for purposes of exposition bring out the fundamental duality of body and soul.

Let us compare this account of memory with Butler's. If we start from the position that all action is evidence of life, and that life is in its nature psychic, then in the action and reaction, which our intellect sketches as a world of things, we have an interaction of psychic forces. Life as continuous has the whole of the past ever behind it. We have seen the sense in which for M. Bergson the present moment can be regarded as

an effect created by the past. Pampsychism, as Dr. Ward allowed, will justify in principle the claim for the survival of the past. So far as the memory which enacts the past is concerned, the memory of instinct and the memory of habit, M. Bergson's doctrine presents a fair parallel to that of Butler. There is the important difference already noticed; viz., for M. Bergson such memory is expressive of the continuity of life, for Butler it is expressive of the stores of an individual mind.

In considering the relation of M. Bergson's two memories to each other we may notice the two faculties of knowledge which this thinker distinguishes in Creative Evolution', Instinct and Intelligence. The two forms of memory would appear to be adapted respectively to the two forms of knowledge. Instinct, as the power by which an organism adapts itself to the requirements of life, can be sustained by a memory which enacts the past, but intellect as the power which decomposes and reconstructs, which aims at predicting movements and manufacturing tools, requires a memory which can form representations. Intellect, in carving out of the flux of the world of change material things, is master of the inanimate. For its work it needs stable images. Indeed we might regard the very detachment of images from the web of pure memory as itself an effort of intellect. In virtue of his instinctive endowment man is one with the rest of organic nature, but in virtue of his ability to use representations he has won a special place in evolution. Instinct and intelligence remain unreconciled in M. Bergson's philosophy. He tells us that they are complementary, ' and they are complementary only because they are different, what is instinctive in instinct being opposed to what is intelligent in intelligence.' (Creative Evolution, p. 143.) Advance from one to the other is therefore impossible. Intuition, which might at first sight seem the parallel to Butler's perfect knowledge, is not the outcome of an intellect saturated with memory. It is in its nature more akin to instinct. It is 'instinct that has