orders of sensations in one common sensorium.

It is

by it that the sleep-walker receives and accurately responds to the varied impressions which surrounding objects make on his organs, and its existence suffices to account for the simultaneous effect of sounds, sights, and smells upon an animal seeking its prey or trying to escape from pursuit.

(6) We should also take pains to understand and appreciate the distinction which exists between true "inference," which is an essentially intellectual apprehension of a truth as implicitly contained in other truths, and that mere sensuous reinstatement of past impressions which may simulate it. The latter affection, which we have distinguished as "sensuous or organic inference," manifests itself as follows: Let any group of sensations have become intimately associated with certain other sensations, then, upon the recurrence of that group, an imagination of the sensations previously associated therewith spontaneously arises in the mind, and we have an expectant feeling of their proximate actual recurrence. Thus, the sensation of a vivid flash of lightning has come, by association, to lead to an expectant feeling of the thunder-clap to follow. Such mere association of feelings, some of which when freshly experienced lead to an expectant feeling of the others, and to a feeling of satisfaction when the sense of expectation is fulfilled, may certainly exist in animals as well as in ourselves, and its presence will fully account for all those actions which are so often taken as indications of the existence in them of a truly reasoning faculty.

* See "On Truth," pp. 194, 201.

(7) We have already indicated what we deem to be the true nature of the process of abstraction; but before entering upon a consideration of the statements made by Mr. Romanes in his second chapter, it may be well, at the risk of tediousness, to repeat that so far from its being a separation and segregation of feelings, it is radically different from every sensuous process. It is the spontaneous starting forth in the mind of an intellectual cognition, or idea (upon the reception of certain sensuous experiences), like Minerva from the head of Jove. One of the earliest of our abstractions is also one of the most ultimate-namely, the idea of "being." This never was and never could have been a feeling, though the idea must have accompanied every feeling recognized by us as such. Thus abstraction is so fundamentally different from the power of forming sensuous universals, that it may be said to be a process directly contrary to it; since the latter agglutinates sense-impressions which the former discards as it emerges and escapes from amongst them.

(8) Lastly, we should be very careful to distinguish between feeling, knowing, judging, inferring, and classifying formally-i.e. when we perform this act with a distinct intention to perform them—and feeling, knowing, judging, inferring, and classifying materially—i.e. when we do so in a more or less automatic manner, without intention or advertence. This distinction takes note of the difference between direct and reflex cognition.†

* See "On Truth," pp. 205, 208, 234.

† As to this, see "On Truth,” pp. 8, 23, and also p. 189 as to the three senses of the word "know."

Thus there may be—

(a) Unconscious, merely sensuous cognition, accompanied by "consentience "—as in the actions of certain sleep-walkers and idiots.

(b) Intellectual cognition of the lowest order: where general consciousness is present, but where there is no distinct consciousness, not only as to the nature of an act performed, but even as to the fact of its performance; so that the act is far indeed from being one done with a deliberate intention of doing it. Thus, when out shooting, and in a normal state of consciousness, on firing and missing our aim, we may make some sudden gesture by which a bystander can see what has happened, though we had no intention of so indicating it, and had no distinct consciousness of the fact of our bodily movement. Such a movement is no true "sign," for the gesticulator has no intention of conveying his ideas to another by depicting any fact. If, then, a spectator exclaims, "That gesture is a sign he has missed his aim," such a spectator uses the term "sign" improperly, by a loose analogy. Similarly we may, without any intention or distinct consciousness, make a movement from which a bystander can tell in which direction an animal we have been watching may have gone.

(c) Intellectual cognition, accompanied not only with a general consciousness, but with a consciousness of an act performed, yet without special advertence to it as being a fact or to any intention we may have had on performing it. Thus we may suddenly raise our arm and point in a specially selected direction, with the

F

intention of showing which way any creature has gone. Evidently a consciousness must attend a gesture thus made to indicate direction which was absent in an aimless, altogether unintentional movement produced by vexation at having made a bad shot, however practically indicative the latter may have been. Therefore such a movement is a true "sign," being a movement made depicting a fact with the intention of conveying to other minds the ideas of the sign-maker.

(d) We may do the same thing not only with consciousness and intention, but with express advertence to the fact of our intention in the act deliberately performed.

We may know without adverting to our knowledge, and we may feel without knowing that we feel. Now, since such is the case with us, it must be, to say the least, probable that animals also may feel without knowing it.

With these premisses, we may proceed to the examination of Mr. Romanes's third chapter, entitled, "Logic of Recepts." Therein he tells us,* "The question which we have to consider is whether there is a difference of kind, or only a difference of degree, between a recept and a concept. This is really the question with which the whole of the present volume is concerned."

We call attention to this passage as an excuse for, and a justification of, what we fear some of our readers may deem too great minuteness and reiteration in this analysis of mental states. Great care is, however, necessary not to yield to the temptation of hurrying

*

over and treating incompletely a matter which is of the very essence of Mr. Romanes's whole contention.

As to his mode of procedure, he observes,* “First of all I will show, by means of illustrations, the highest levels of ideation that are attained within the domain of recepts; and, in order to do this, I will adduce my evidence from animals alone, seeing that here there can be no suspicion-as there might be in the case of infants-that the logic of recepts is assisted by any nascent growth of concepts."

Before, however, applying himself to this task, he discusses his own expression, "logic of recepts."

He tells us, in the first place, that "all mental processes of an adaptive kind are, in their last resort, processes of classification; they consist in discriminating between differences and resemblances."

Now, in this sentence much confusion of thought is indicated. In the first place, the word “discriminating" is used ambiguously, as-neglecting the distinction we have above indicated as No. (8)—it is applied to both "formal" and "material" discrimination; and yet these acts are of a radically different kind. A mere sieve materially "discriminates" between coal-dust and cinders of a certain size! It is also false to say that "all mental processes of an adaptive kind" "consist" in either a material or a formal discrimination; although, of course, like all other mental acts, they are accompanied by something of a discriminating nature. To know that any object (y) possesses any quality (x), implies that x is discriminated from the group of * p. 46. † p. 46. See above, p. 64.