in harmony with facts and with each other, not fearing to provisionally adopt, as favourite candidates for belief, hypotheses which, although otherwise well recommended, do not capture certitude. If the speculation achieve a system of hypotheses perfectly explanatory of a vast multitude of facts and in harmony with one another, the system, owing to a well-known mental law, would compel certitude of its truth. The explorer starts on the voyage equipped with a system of beliefs and with common sense which serves him, not only as ballast, but, in connection with his beliefs, as compass; for, besides saving him from dangerous careening, it indicates the direction he should take, viz., along the line of consistent hypotheses that most accord with his beliefs and in the least degree innovate upon the system of those beliefs. For example, if two data be inconsistent with one another, he is to prefer that the elimination of which would cause the greater change in the system of his beliefs. Of course common sense is tenacious of all data that are not discredited by inconsistency, but, above all, of those that serve as foundations of morality and religion, e.g., that there is a soul, that we are free agents. When such data become doubtful, the moral and religious faculties unite in a challenge to Will to prevent doubt from causing the moral paralysis and decay that might be inevitable if certitude of the falseness of the datum were in place of doubt. They suggest to Will to apply what was known to the Latins under the name arbitrium,-an act which founds resolve on mere opinion, an act indispensable to those who have to navigate a sea of conjecture. Decree, they exclaim, the truth of the questionable datum, and, as regards conduct, rely on it as though it were the certitude it substitutes. Man

liness, it seems to me, concurs with morality, religion, and common sense, in this challenge. How should it indolently "gape on while doubt is undermining human dignity? Considering the fallibility of the human mind, its dependence on data, the necessity it is under to proceed upon conjecture, the superiority of a limited conservatism to an unballasted proneness to novelty in the interpretation of nature, and the probable degradation of the race if it lose faith in free agency and responsibility, it seems to me that the foregoing method is recommended by transcendent credentials.

The method repudiates the doctrine that virtue is an impediment to research—an impediment as indisposing the mind to beliefs that are hostile to it : the method proceeds on faith that virtue or wisdom is a faculty as needful to research as that of vision, though also as fallible. It is true that, if men be no better than maggots, the discovery of that truth by research under the tutelage of Wisdom risks postponement; but is the postponement a respite or a loss? Morals at least would not be the worse for it. The tendency of the method to prevent research from bolting is elucidated by the extravagance of the doctrine, that human behaviour is exclusively automatic-that consciousness has no more to do with it than the whistle of the locomotive with its motion. All that can be said for this doctrine is that it is not inconsistent, and that it is competent to molecular change to cause behaviour which seems to be intentional. infer from the facts which indicate this competence that man is a mere automaton, is a non sequitur. The method puts the doctrine out of court.

Το

III.

It is new, even to philosophy, that exploration and discovery are possible to the faculty of Definition. It is taken for granted that the office of the faculty is confined to the humble work of making knowledge ship-shape, and explaining the meaning of words,that he, for example, who achieves a definition of Induction has not augmented-has merely arrangedknowledge. The obvious agreement of definitions with the known tends, when they augment knowledge, to hide the appearance of increase. The detection of a differentia is an increase of knowledge, and often an increase of the greatest importance; but, though this must be manifest to the discoverer, it tends to elude those to whom he imparts his discovery: they think that they have profited only by having their knowledge put for them in a clearer light. An analogous error disputed Bacon's title to be the originator of an intellectual epoch. Forsooth, people had inferred inductively prior to Bacon, and therefore the Novum Organum contained nothing new. It has escaped philosophers that the faculty of definition was the supreme faculty of Socrates, and that his dialectic was a method of driving people to the border of definition which was to enrich the world with new knowledge. Now in this Essay error is sapped and truth put in its place by a noiseless process of definition that tends to exclude an appearance of addition to knowledge. I might easily seem to have done no more than decant the known into another form. It is important, no less to the reader than to myself, that this error be avoided.

IV.

The treatise consists of three books. The First consists of Definitions demanded by a new classification of mental events and faculties-not the less new that the classes are denoted by familiar names. The Second treats of Reasoning. The Third consists of expositions which concur in showing the dependence of personal agency on Self-Denial. The first chapter of the third book shows that science is unconscious knowledge. The second deduces from familiar mental event the existence of an unconscious part of the mind and of unconscious mental event. The third proves that the unconscious part of the mind is corporal, consisting of the encephalon, etc. The fourth is an exposition of Wisdom. The fifth proves that man has been for the most part puppet, dupe, and victim of unconscious forces, and that self-denying conduct is a sine qua non of escape. It may be asked, at this hour of the day, so long after Leibnitz had called attention to unconscious mental event, and Dr. Carpenter had popularized knowledge of unconscious cerebration,-what need was there of a deduction of an unconscious part of the mind and of unconscious mental event? I answer that, except as regards the insignificant species of unconscious event noticed by Leibnitz, it has never been shown that there are unconscious mental events. It has been abundantly shown that certain unconscious events are conditions sine qua non, and otherwise accessories, of mental action, but never hitherto that mental events include other unconscious events than those indicated by Leibnitz. No one will suspect

Professor Bain of overlooking the bearing of corporal upon mental event, yet his definition of Mind supposes mental event to exclude unconscious event. According to Professor Bain Mind is a sum of operations and appearances that are either feelings, volitions, or thoughts. Even Mr. Lewes, who held that event of which the obverse aspect belongs to the kind, mental events, has a reverse aspect which correctly ranks it, as being a neural tremor, in the kind, corporal events,

1 "The operations and appearances that constitute Mind are indicated by such terms as Feeling, Thought, Memory, Reason, Conscience, Imagination, Will, Passions, Affections, Taste. But the Definition of Mind aspires to comprehend in few words, by some apt generalisation, the whole kindred of mental facts, and to exclude everything of a foreign character."

"Mind is commonly opposed to Matter, but more correctly to the External World. These two opposites define each other. To know. one is to know both. The External, or Object, World is distinguished by the property called Extension, which pertains both to resisting Matter, and to unresisting, or empty Space. The Internal, or the Subject, world is our experience of everything not extended; it is neither Matter nor Space. A tree, which possesses extension, is a part of the object world; a pleasure, a volition, a thought, are facts of the subject world, or of mind proper.

"Thus Mind is defined, in the first instance, by the method of contrast, or as a remainder arising from subtracting the External World from the totality of existence. It happens that the External World

is easily defined or circumscribed; the one well-understood property, Extension, serves for this purpose. Hence the alternative, or the correlative, Mind, can be circumscribed with equal exactness. But it is desirable to possess, in addition to this negative definition, however precise it may be, a positive definition, or a specification of the quality or qualities that appertain to the phenomena designated mind. Now, we have not here the good fortune to be able to refer to a single precise quality, like Extension for the object world; we must refer to several qualities that conspire to make up our mental framework. Hence our positive definition, instead of being a unity, is a plurality, and is not only a Definition, but also a Division of the Mind."

"The phenomena of the Inextended Mind are usually comprehended under three heads :

"I. Feeling, which includes, but is not exhausted by, our pleasures