which are adduced to prove that it has been, and then it will become apparent that the supplement, if any, must have been added in the smelting-pot of Language -a fact, be it observed, which is conceded by all modern writers who deny the genetic continuity of mind in animal and human intelligence." The last assertion is one which is indeed remarkable. It shows that Mr. Romanes has not apprehended what is the fundamental position, on this subject, of the school to which he is opposed. The "intellectual," as opposed to the "sensational" school, energetically affirm that the supplement added was not "language," but "a distinctly rational nature,” whereof thought, language, and moral responsibility are alike results.

In concluding this chapter, its author makes an assertion which we have sincere pleasure in agreeing with and supporting. It is the assertion that children do not commence their intellectual life by special and particular perceptions from which they generalize, but that they generalize at once. Nevertheless, his misapprehension of the distinction between recepts and concepts, and his notion that a distinct intention is needed in order to form the latter, naturally make themselves manifest. As to recepts and concepts, Mr. Romanes truly says, "Classification there doubtless is in both cases; but the one order is due to the closeness of resemblances in an act of perception [i.e. senception], while in the other order it is an expression of their remoteness from merely perceptual [i.e. sensuous] associations."

The concluding sentence of this chapter is, however,

very misleading, and really once more begs the whole question which its author has to prove. He says,* "The object of this chapter has been to show, first, that the unintentional grouping which is distinctive of recepts may be carried to a wonderful pitch of perfection without any aid from the intentional grouping which is distinctive of concepts; and, second, that from the very beginning conscious ideation [which here means our consentience] has been concerned with grouping. Not only, or not even chiefly, has it had to do with. the registration in memory of particular percepts; but much more has it had to do with the spontaneous sorting of such percepts, with the spontaneous arrangement of them in ideal (or imagery) systems, and, consequently, with the spontaneous reflection in consciousness of many among the less complex relations-or the less abstruse principles—which have been uniformly encountered by the mind in its converse with an orderly world."

Certainly the world is orderly. Certainly its coexistences and sequences make manifest, objective relations and principles which pervade and govern it. Certainly, also, these objective conditions modify the sentiency of irrational organisms, and certainly, as we have elsewhere pointed out,f such objective conditions correspond, as "objective concepts," with internal perceptions or "subjective concepts in us." But this in no way even tends (as it is represented as tending) to bridge over the gulf which exists between sentiency and intellect. We might à priori expect to find a certain * p. 69. † See "On Truth," pp. 136, 137, 386, 445.

parallelism of results in the effects of one set of objective external conditions acting upon two distinct kinds of internal subjective powers-one sentient, the other rational. The wonders of vegetable life, of sentiency, and of intellect, are all parallel and similarly inexplicable. In plants we have chemical combinations organized and vivified; in animals we have vegetative, organic life raised to sentiency and receptive power; and in man we have animal, sentient life raised to perception and conceptual power.

His fourth chapter Mr. Romanes devotes to a consideration of the " Logic of Concepts." He begins it by affirming (what no reasonable person can deny) the great importance of "sign-making" and "symbols" for the growth and advance of intellectual life. But he gives us no definition or explanation as to what he means by a sign, while he makes observations, by the way, which must not be allowed to pass without criticism. Thus he says: "By the help of these symbols we climb into higher and higher regions of abstraction: by thinking in verbal signs we think, as it were, with the semblance of ideas we dispense altogether with the necessity of actual images, whether of percepts or of recepts: we quit the sphere of sense, and rise to that of thought." But so long as life, as we know it, lasts, we can never dispense with the use of mental images (phantasmata)† of some kind-whether it be of sights or of sounds or of some form of our own activity. Such images, however, are not the "semblance of ideas," but survivals and reminiscences of sensuous experiences.

* p. 71.

As to this, see On Truth," pp. 87, 88.

Mr. Romanes illustrates his contention by a reference to mathematics, which demonstrates for us with especial clearness the great value of symbols. We are told,* "Man begins by counting things, grouping them visibly [i.e. by the Logic of Recepts]. He then learns to count simply the numbers, in the absence of things, using his fingers and toes for symbols. He then substitutes abstract signs, and Arithmetic begins." But no man could begin really counting the simplest things unless he already possessed the idea of number; and, as Mr. Romanes truly says, "before the idea of number can rise at all," a distinct power of intellectual conception must be present. The very essence of "counting" is numerical distinction. To suppose that a man could voluntarily begin to count, without any idea of such distinction, is absurd. But men, like animals, may "group objects visibly" without counting. To separate objects in groups -were they in groups which accidentally had definite. numerical relations-without any regard to their number, could never be counting. To suppose that a man by "not counting" could learn to count, or that he could. acquire the idea of "number" by performing actions wherein he took no note of real numerical relations, is to add absurdity to absurdity. He could not possibly take note of any numerical relations without having the idea of numerical relation, that is, without possessing very abstract ideas and having already an intellectual We dwell on this point because it is a good

nature.

p. 72.

For what is implied in the idea of "number," see "On Truth," p. 241.

instance of that "intellectual thimble-rigging" which all men of the sensist school, from Hume downwards, must perform in order to make the innocent onlooker think he has found the pea of "intellect" under the thimble of "sense." We dwell on it the more because the sincerity and honesty which are conspicuous amongst the other merits of Mr. Romanes, show how he himself has been deceived and is all unconscious of the ways of some of his masters. It is none the less true that he is completely justified in affirming,* with Sir W. Hamilton, that signs of some kind are needed “to give stability to our intellectual progress," that "words are fortresses of thought," and that "thought and language act and react upon one another.† Not, however, that we can for a moment admit that any change in mere verbal expressions, which are not the result of a modification of thought, can improve the latter. It is thought alone which can really improve language, though verbal modifications acting with it and produced by it may greatly aid it and hasten intellectual progress.

Mr. Romanes begins the real substance of his fourth. chapter as follows: "From what I have already said, it may be gathered that the simplest concepts are merely the names of recepts." This we altogether deny. In the very simplest concepts, the ideas, “existence," "kind" or "nature," "reality," " possibility" and "impossi

Here we may ask at once, by anticipation, "If thought is thus admitted to be able to improve language, why should it be thought unable to originate it?"

+ p. 73.