general terms, the conclusion which physiological psychology tends more and more confidently to endorse. But the third proposition, that consciousness is a collateral product of brain action, introduces a bit of theory which appears to me neither satisfactory nor necessary. I, for one, find as much difficulty in imagining or conceiving how matter in motion can produce consciousness, which, "as such, has not the slightest community of nature with matter in motion," as in conceiving how "volition, which is a state of consciousness, can act upon the moving matter of which the body is composed." The difficulty in each case appears to me to be precisely the same. Moreover, it would seem that each one of us has at least as good reason for believing that one state of consciousness directly suggests another in a chain of psychical causation, as that these conscious states are merely collateral products which symbolise occurrences in a chain of physical causation. Furthermore, the introduction of this piece of theory is unnecessary so far as the present discussion is concerned. The facts, or what we believe to be the facts, are just as well expressed by saying that, from one point of view, certain physical occurrences have conscious concomitants, and that, from another point of view, certain conscious occurrences have physical concomitants; and that, from either point of view, these occurrences are links in a causation chain. This leaves Huxley's main contention exactly where it was. It merely strikes out a redundant hypothesis.

We come now to the fourth proposition. Professor Huxley does not, indeed, anywhere define the terms "automaton" and "automatism"; but the definition above given, that the term "automaton" is applicable to any piece of mechanism all the workings of which at any given time are explicable in terms of physical causation, may, I think, be fairly inferred from what is explicitly or implicitly contained in the essay. And this is no doubt in accord with such a definition as that given in the Encyclopædia Britannica, where an automaton is described as "a self-moving machine, or one in which the principle of motion is contained within the mechanism itself." It is true that we are told that "the word is generally applied to contrivances which simulate for a time the motions of ani

mal life." But if we apply it to any of the motions of animal life, there would appear to be no logical grounds for rejecting its application to all these motions. And if we accept these definitions as they stand, Huxley's position, as summarised in propositions 5 and 6, follow in logical sequence, and we must hold with him that in the life of animals and man automatism reigns supreme.

We may fairly ask, however, first, whether the definition, so applied, is in accordance with general usage; secondly, whether it is helpful in the study of animal life; and, thirdly, whether it preserves the spirit of the teaching of that acute thinker, René Descartes, whose thought Professor Huxley interpreted in terms of modern science.

It certainly does not appear to be in accordance with common usage. When I receive a telegram from a friend, who has recently returned to England, begging me to come and see him, and deliberate whether, in view of certain engagements into which I have entered, I can accede to his request, it would seem to be scarcely in accordance with established usage to say that I fill in the reply-telegram automatically. Nor would most persons, I imagine, describe my action as instinctive, as they should do if Huxley's view be accepted in its entirety, and if, "when we speak of the actions of the lower animals being guided by instinct, . . . what we really mean is that, though they feel as we do, yet their actions are the results of their physical organisation." For the words which Professor Huxley inserts after instinct-"are guided by instinct and not by reason"-may be omitted if reason, too, like volition, be no less than reflex action, "one of the results of our physical organisation." Nor, again, is it in accordance with established usage to call a being which profits by experience and which is susceptible of progressive education an automaton.

To the second question, as to the first, I am disposed to give a negative answer. Distinctive terms are of service just in so far as they help us to draw the distinctions which are necessary for clearness of thought and expression. If we universalise the term automatism so as to comprise the whole active life of man and animals, it loses all its distinctive value. The term as applied to animal life

is useful just in so far as it serves to distinguish actions which are automatic from others which are not automatic. On these grounds, I am prepared to advocate a more restricted definition, according to which an automatic action is one that we have reason to suppose is not performed under the immediate guidance of consciousness, this phrase being understood to be a shortened expression for "with the intervention of certain controlling physical occurrences which are accompanied by states of consciousness." Of the exact nature and sequence of these physical occurrences, we are at present profoundly ignorant; but of the nature and sequence of the states of consciousness as they occur in ourselves, we do, at any rate, know something. And we may fairly infer the existence of somewhat similar states from the observable behavior of animals.

But does Professor Huxley's position preserve the spirit of the teaching of Descartes? I venture to think not. Huxley himself, in an earlier essay-that on "Descartes's Discourse on Method "— thus briefly indicates the Cartesian conception of the rôle of consciousness:

"According to Descartes all the functions which are common to man and animals are performed by the body as a mere mechanism, and he looks upon consciousness as the peculiar distinction of the 'chose pensante,' of the 'rational soul,' which in man (and in man only in Descartes's opinion) is superadded to the body. This rational soul he conceived to be lodged in the pineal gland as in a sort of central office; and here, by the intermediation of the animal spirits, it became aware of what was going on in the body, or influenced the operations of the body. Modern physiologists do not ascribe so exalted a function to the little pineal gland, but, in a vague sort of way, they adopt Descartes's principle, and suppose that the soul is lodged in the cortical part of the brain-at least this is commonly regarded as the seat and instrument of consciousness."1

Now what is the essential feature of Descartes's conception of the part played by consciousness? Is it not that that which controls, stands apart from the automatic mechanism over which its control is exercised? It is true that his enthronement of consciousness in the pineal body was about as wide of the mark as was his

1 Collected Essays, Vol. I., Essay IV., pp. 188-189.

conception of the nerves as conduit pipes through which the animal spirits, pumped from the heart to the brain, are emptied into the muscles. But if in the latter case his principles were sound, though his facts were conjectural, so, too, in the former case, his conception was valid in essence though his pineal gland took no share in its elaboration. And one may be permitted to wonder in what manner, "unwearied dissector and observer" as he was, Descartes regarded the pineal body in the animals he dissected and observed. Was it an empty throne awaiting its royal occupant? This, however, by the way. The essential feature of his teaching, as I understand it, is that when, as in the actions of man, we have evidence of guidance and control, in view of certain data afforded to consciousness, that which guides and controls stands apart from the bodily mechanism concerned in merely automatic response. Descartes himself believed that the soul, enthroned in the pineal gland, performed this function. Later thinkers have believed that the soul used the cerebral cortex as the instrument through which its control was exercised. Professor Huxley, wielding the sword of logic, forces the soul to abdicate its throne, and by his extended hypothesis of automatism does away altogether with the conception of guidance and control. But if we dethrone the soul, and deny its divine right to rule our actions, that is no reason why we should leave the body politic without any form of government. The truer inference is that the cerebral cortex is the organ of control not as the instrument of the soul, (which may or may not exist,1 so far as the matters we are discussing are concerned,) but in its own right. For the cortex itself stands apart from the lower brain-centres which are concerned in automatism in the more restricted sense.2 The cortex is not the instrument of that which controls, but is, from the physical point of view, that which controls. The molecular changes therein, evoked by bodily conditions, are such as to augment, or inhibit (and by augmenting here

1 This I conceive to be the rigidly agnostic position.

"The cerebral hemispheres, as we have more than once insisted, seem to stand apart from the rest of the brain." Professor M. Foster, Text-book of Physiology, 5th Edition, Part III., p. 999.

and inhibiting there to modify1) the action of the lower automatic centres; and these molecular changes are accompanied by consciousness. The physiology of the future may be able to indicate the physical conditions under which control is effected; but as matters now stand, we know far more about the accompaniments in consciousness than we do about the concomitant molecular changes. In describing therefore what we believe to occur, we may say, if we desire to be somewhat pedantically accurate, that the actions which we term voluntary are the effects of those molecular changes in the cortex which are accompanied by consciousness; or we may say in brief and to avoid circumlocution, that they are the results of conscious guidance and control. Thus we preserve the essence of Descartes's teaching but interpret it in terms of modern scientific thought.

On all grounds, then, a more restricted definition of the term "automaton" than that which Professor Huxley adopted in his later Essay? seems advisable; on the ground of general usage, on the ground of scientific utility, and on the ground of historical precedent. And our consideration of Descartes's teaching helps us to reach a further definition of animal automatism, in the more restricted sense. Automatic action is that which is performed without the immediate and effective intervention of those molecular changes in the cerebral cortex which are accompanied by consciousness (such intervention being rendered possible by association); or, in brief, automatic action is that which is performed without conscious guidance and control. Consciousness as an adjunct there may be; but it takes no share in the direction of active response. Professor Huxley returns to the subject of animal automatism in a subsequent essay-that on The Connection of the Biological Sci

1 Descartes used similar expressions when he likened the rational soul to the engineer amidst the automatic figures of a grotto "when he wishes to increase, or to slacken, or in some way to alter their movements." Quoted by Huxley, Collected Essays, Vol. I., Essay IV., p. 183.

2 In his earlier essay on Descartes's Discourse he seems to accept the more restricted usage. See his remarks on the effects of education by which acts become mechanical. Loc. cit., p. 188. See also the sentences at the top of p. 187.