object is nothing more nor less than our knowledge of certain forms of our own consciousness. There is no observation, strictly speaking, of any external object. We must either say, with Prof. Bain and his school, that "the belief in the existence of any portion of matter outside and independent of our consciousness is a most anomalous fiction," or else we must rest for its truth on an intuitive conviction of the veracity of the senses and the existence of an external world such as would give rise within us to our felt states of consciousness.

Again, you say that you saw the rose last week. Still more intuitive beliefs must you, then, lean upon. For, how can you testify with certainty to what occurred last week? You have no present sensation such as you describe. You have only in your present consciousness an image or recollection of it, and how do you know that this present image is a truthful copy of the past sensation? There is no reason for it except that intuitive conviction of the veracity of memory which John Stuart Mill himself is forced to acknowledge as an "ultimate belief."' No past experience can prove this trustworthiness of memory. For it must in each case be taken for granted before you can have any cognizance whatever of your past experience.

But still another intuition belongs to the chain. You said that it was you who observed the rose last

Mill's "Examination of Hamilton," p. 216, vol. i., American edition.

week. You remember it as an experience of yourself, and you imply, and the worth of your testimony depends upon the fact, that you who a week ago had a certain sensation, and now have it not, are yet one and the self-same person. Now, how do you know this personal identity? Again, you must admit you know it only by an intuitive conviction.

Thus, to be able to trust the simplest past observation of a natural object, we must accept these four intuitive beliefs: 1. In the veracity of the senses; 2. In the reality of an external world; 3. In the veracity of memory; 4. In our continuing personal identity.

Among the fundamental principles on which science depends are the three doctrines of the indestructibility of matter, the continuity of motion, and the persistence of force. Were it possible for matter to become non-existent, or for motion or force to lapse into nothing, there would exist in science incalculable elements, fatal to all positive knowledge or scientific interpretation of phenomena. What warrant have we, then, for the truth of these great principles? Inductive experiment? This has certainly contributed much to establish it. Delicate tests with balance and retort have shown that when matter, motion, or force, seemed to disappear, they simply changed their form, place, or direction. Solids changed to gases, molar motion to molecular motion, force of heat passed into magnetic or chemical force. Track the cunning Proteus into his new

haunt, and you will find him there undiminished in quantity. This is what experience has suggested and approximately proved. But it has only done so approximately, never absolutely. It has shown that the more delicate were its means of measurement, the more closely it could follow every diverging motion or escaping matter, the more nearly equivalent was the quantity accounted for at the end with that with which the experiment began. But it has never proved this with any absolute exactness, nor for any larger field than the narrow circles which have been specially investigated.' Moreover, all through the so-called process of inductive proof, the truth to be demonstrated has been continually assumed, as Herbert Spencer has admirably shown in the fourth, fifth, and sixth chapters of his "First Principles." Whatever inductive experiments are made depend for their validity upon the continual assumption that the gravitation of the weights, or whatever unit of force is taken as the measure, remains constant, and of this, says Herbert Spencer (p. 187), no proof is assigned, nor can be assigned. "Nor is it only in their concrete data," he continues, "that the reasonings of terrestrial and celestial physics assume the persistence of force. They equally assume it in the

1 Prof. Joseph Lovering, in his address as president, before the American Association at Hartford, 1874, says of the doctrine of the conservation of energy: "The most that physical science can assert is, that it possesses no evidence of the destructibility of matter or force." See also Lewes's "Problems of Life and Mind," vol. ii., p. 262.

abstract principle with which they set out, and which they repeat in justification of every step. The equality of action and reaction is taken for granted from beginning to end of the argument; and to assert that action and reaction are equal and opposite, is to assert that force is persistent. . . . Clearly, then, the persistence of force is an ultimate truth of which no inductive proof is possible."

Or take the other great basic principle of science—the uniformity of Nature, embracing in that term both the uniformities of coexistence, or accompanying qualities of things, and the uniformities of succession, or, as Mill calls it, the universality of causation. The validity of all induction, of all reasonings as to matter of fact in past, present, or future, depends upon the assumption of this uniformity of Nature. What scientific foundation, then, does science present for this universal basis of its knowledge? It has none. John Stuart Mill, to be sure, sought to rest this general basis of induction upon induction itself, even upon an induction by simple enumeration.' But all his logical skill could not cover up the fact that he was thus proving the universal by a limited number of particulars, the greater by the less, the stronger by the weaker. It hardly needs to be pointed out that any particular experience, short of universal extent, cannot prove

1 "I hold it to be itself an instance of induction, and induction by no means of the most obvious kind."-("Logic," Book III., Chapter III., § 1.)

a universal law. However the evidence be manipulated, a general and absolute conclusion cannot be established upon a limited and uncertain premise. Experience can only testify as to what has been, not as to what will be. It can testify to what has come within its field, not as to what is outside. Although two events have accompanied each other a hundred thousand times under our observation, that is no proof that they will do so the next time. It is possible that they may not. As one whose predispositions all lead the other way says, "Water has quenched our thirst in the past; by what assumption do we affirm that the same will happen in the future? Experience does not teach this; experience is only of what has actually been; and, after never so many repetitions of a thing, there still remains the peril of venturing upon the untrodden land of future possibility. The fact, generally expressed as Nature's uniformity, is the guarantee, the ultimate major premise of all induction. What has been will be,' justifies the inference that water will assuage thirst in after-times. We can give no reason, no evidence, for this uniformity; and therefore the course seems to be to adopt this as the finishing postulate." (Bain's "Logic," I., 273.)

This is, indeed, the only logical course, to admit the belief in the uniformity of Nature as a fundamental postulate, a primary intuition, an ultimate law of the mind. As Bain says in another work ("Emotions and Will," second edition, page 537),