disappointed thinker from vitalism, if indeed his salvation depends upon this achievement.

Is then the conception of memory introduced by Hering and adopted by Dr. Semon psychologically meaningless? The answer so far reached is 'yes', if it is interpreted, as it was by Hering and by Dr. Semon, structurally. But there are other possibilities.

We noticed that Butler had trouble with Hering's account of material vestiges and preferred a theory of diminutive vibrations. Now what was at the root of this difference? It is necessary to develop Butler's view more fully. In his chapter introducing Hering's lecture Butler wrote:

'Another matter on which Professor Hering has not touched is the bearing of his theory on that view of evolution which is now commonly accepted. It is plain he accepts evolution, but it does not appear that he sees how fatal his theory is to any view of evolution except a teleological one-the purpose residing within the animal and not without it. There is, however, nothing in his lecture to indicate that he does not see this.' (ibid., pp. 95, 96.)

Butler is perhaps too optimistic, but how prone he was to read into Hering the theories which he himself held is obvious in a later passage where he declares:

'It is no strained conclusion to gather that he (Hering) holds the action of all living beings, from the moment of their conception to that of their fullest development, to be founded in volition and design, though these have been so long lost sight of that the work is now carried on, as it were, departmentally and in due course according to an official routine which can hardly now be departed from.' (ibid., p. 226.)

In the sense in which Butler means it, it would be a very strained conclusion indeed.1 For Butler there is no division between inorganic and organic matter.

The following passage from his note-books referring to his attempt to explain the physics of memory, published in the Quarterly Review, iii, No. 9, shows that Butler is not so guileless as he appears: 'I was

'We should endeavour to see the so-called inorganic as living, in respect of the qualities it has in common with the organic, rather than the organic as non-living in respect of the qualities it has in common with the inorganic.' (ibid., p. 275.) It is more coherent with our other ideas to start with every molecule as a living thing, and then deduce death as the breakup of an association or corporation, than to start with inanimate molecules and smuggle life into them; and that, alarmed by the suggestion and fathered it upon Professor Hering, who never, that I can see, meant to say anything of the kind, but I forced my view upon him, as it were, taking hold of a sentence or two in his lecture, On Memory as a Universal Function of Organised Matter, and thus connected memory with vibrations.'

The following extracts from La Marck are of interest in relation to Hering and Butler:

To try to determine how the agitations of the nervous fluid trace or engrave an idea on the organ of understanding, would be to court the risk of committing one of those numerous errors to which the imagination is liable; all that we can be sure of is that the fluid in question is the actual agent which traces and impresses the idea; that each kind of sensation gives a special agitation to this fluid and consequently causes it to impress equally special outlines upon an organ so soft and delicate and find its way into such narrow interstices and tiny cavities, that it can impress in their delicate walls traces more or less deep of every kind of movement by which it may be agitated.'

Using as illustration the memory of a building on fire, La Marck writes: This process must be due to the fact that our inner feeling sets our nervous fluid in motion, and drives it into our organ of inteiligence over the outlines impressed by the sensation of the conflagration, and that the modification acquired by our nervous fluid in its movements, as it passes over these particular outlines, is promptly transmitted to our inner feeling and thereupon restores to clear consciousness the idea that we are seeking to recall; although the idea is less vivid than when the conflagration was actually taking place before our eyes.'

In dreams the nervous fluid is said to wander uncontrolled. In perfect sleep the inner feeling undergoes no emotion' and the individual is as though he did not exist, but in imperfect sleep owing to internal irritation the free part of the nervous fluid is agitated and flows where it may over the configurations, giving rise to dream images.

Thus dreams disclose to us the mechanism of memory, just as memory teaches us the mechanism of ideas.' Zoological Philosophy, translated by H. Elliot, pp. 393, 395. The last extract would seem to show that this good office of dreams is not the discovery of the psycho-analyst.

therefore, what we call the inorganic world must be regarded as up to a certain point living and instinct, within certain limits, with consciousness, volition and power of concerted action. It is only late, however, that I have come to this opinion.' (ibid., p. 23.)

For Butler' Life is that property of matter whereby it can remember-matter which can remember is living' (p. 272). With this view of life as something instinct within each atom, the doctrine of vibrations possessed for Butler a very special meaning.

'Assimilation is nothing else than the communication of its own rhythms from the assimilating to the assimilated substance, to the effacement of the vibrations or rhythms heretofore existing in this last.' (ibid., p. 86.) I would recommend the reader to see every atom in the universe as living and able to feel and to remember, but in a humble way. He must have life eternal as well as matter eternal; and the life and the matter must be joined together inseparably as body and soul to one another.' (ibid., p. 273.)

It is in this sense that he could say:

Assimilation is nothing but the imbuing one thing with the memories of another.' (ibid., p. 87. cf. Life and Habit, p. 136.)

It would be impossible for Butler to believe that vibrations ceased to be and were subsequently revived. Their cessation would be the extinction of life and memory. It is not surprising that with these tenets in his mind Butler should try to amend his author's wording in conformity with his own meaning. It is doubtful how far Hering's paper will bear the interpretation thus put upon it. For Butler, mind and life are conterminous. Life is essentially teleological. Probably it is paragraphs such as those quoted which caused Dr. Semon to write: 'Butler's essay contains brilliant suggestions, but these are mixed with so much questionable matter, that the whole, compared with Hering's paper on the same subject, is rather a retrogression than an advance.' (ibid., p. 10.)

Butler indeed substantiates the conception of memory in biology, but he does so because, for him, life is the expression of spirit and the continuity of the past with the present is something more than the continuity of motions within the same mechanism. This pampsychist teaching will be further developed in relation to the view of M. Bergson.

In the meantime we ask whether we must say with Prof. Ward,

'The mnemic theory will work for those who can accept a monadistic or pampsychist interpretation of the beings that make up the world, who believe with Spinoza and Leibnitz that "all individual things are animated albeit in divers degrees". But quite apart from difficulties of detail, I do not see how in principle it will work otherwise.' (Heredity and Memory, p. 56.)

The need for an historical treatment is recognized by many modern writers who at the same time explicitly reject mechan. ism. For instance, Prof. J. A. Thomson writes:

Adaptation or purposiveness requires a historical explanation; it is a super-mechanical concept. . . . More important than any question of terminology is a recognition of the deep difference between the inorganic and organic processes. In the former there are no alternatives; every stage is the necessary outcome of its antecedents; all is mechanically determined. In the latter there are alternatives (for one species may split into several); the organism is a genuine agent; the mechanical categories are transcended.' (Hibbert Journal, vol. x, pp. 120, 320.)

What indications can one find of a treatment which shall not be mechanistic, yet which, in its recognition of what Prof. Thomson calls historical explanation, shall not fall into the snares of pampsychism? We will look first within the realm of biology itself.

A modern writer who has made use of historical' explanation in biology is H. S. Jennings (Behavior of the Lower Organisms, 1906). An organism's reaction to any stimulus

depends upon what Prof. Jennings terms its physiological

state.

'The same organism in different physiological states reacts differently to the same stimuli. It is evident that the anatomical structure of the organism and the different physical or chemical action of the stimulating agents are not sufficient to account for the reactions. The varying physiological states of the animal are equally important factors. . . . We shall find much occasion to realize the importance of physiological states in determining behavior. . . . The present physiological state of an organism depends on its past history.... This statement we know is markedly true for the higher organisms. What a high organism does under certain conditions depends upon its experience: that is upon its past history.' (ibid., p. 178.)

Prof. Jennings would show that all behaviour displays selection, adaptation, regulation. How does memory enter into this adaptation and regulation? To appreciate the answer we must first notice Prof. Jennings's account of response to stimulation. The role of tropisms and stereotyped reflexes in animal behaviour is regarded by Prof. Jennings as limited; he follows Bain and Spencer in believing in over-produced movements as the response to stimulation. An organism does not reply to its environment by a simple reflex which is at once relevant to the situation. On the contrary, stimulation is followed by many and varied movements from which the successful movement is selected by a process of trial and error. It will be that movement which relieves the organism with respect to the stimulation in question. Such a field of varied reactions is in Prof. Jennings's view essential for the development of relevancy and precision in behaviour. Such being the general character of reaction to stimulation, a first step towards regulation in behaviour is reaction to representative stimuli, i.e. stimuli neither beneficial nor harmful in themselves but leading to benefit or injury later. In trying to account for such reactions Prof. Jennings has recourse to the Law of the Readier Resolution of Physiological States.