Now the

tion, for instance, as that Socrates and Plato are mortal, therefore Aristotle is mortal, it is because Socrates, Plato, and Aristotle resemble each other in being men that we can infer that they also resemble each other in being mortal. They also resemble each other in other points, e.g. in being Greeks and philosophers, etc., and differ from each other, e.g. in size and weight; but these points of resemblance and difference do not affect the question: it is not because they were Greeks that they died, and their differences in physical characteristics did not exempt any of them from the common doom. These irrelevant points, therefore, have to be set aside, or, in technical language, "abstracted," and the result of the abstraction is that we are enabled to assert the coexistence of the two qualities of humanity and mortality. savage also is capable of abstract ideas and of asserting their coexistence. He recognises the hardness of some substances and the scent of others, and he wears a ring of iron in order that it may impart its quality of hardness to his body, as he might wear a flower for the sake of its scent; or when he is bargaining for a cow or asking a woman for wife, he chews a piece of wood to soften the heart of the person he is dealing with. In the same way, having discovered in the lion the quality of courage, or in the deer that of swiftness, he eats the former that he may become bold and the latter that he may run well. So also he will eat an enemy to acquire his boldness, or a kinsman to prevent his virtues from going out of the family. The points of resemblance between what he does and what he wishes to effect seem to the savage to be the essential points for his purpose: the man of science deems otherwise. Doubtless the man of science is right; but the savage is not therefore superstitious in this matter. applies a principle of logic-to the wrong things perhaps, but still the process is one of logic, savage if you like, but not superstitious.

He

The savage theory of causation, again, is not fundamentally different from the scientific: it is only incomplete and exaggerated. The effect is the offspring of the cause, and resembles its parent; to produce motion in a body you must impart motion, to moisten a thing you must communicate moisture to it. Hence the savage makes the generalisation

that like produces like; and then he is provided with the means of bringing about anything he wishes, for to produce an effect he has only to imitate it. To cause a wind to blow, he flaps a blanket, as the sailor still whistles to bring a whistling gale. Before going on the warpath or the chase, a mimetic dance, in which the quarry or the foe are represented as falling before his weapon, will secure him success. If the vegetation requires rain, all that is needed is to dip a branch in water and with it to sprinkle the ground. Or a spray of water squirted from the mouth will produce a mist sufficiently like the mist required to produce the desired effect; or black clouds of smoke will be followed by black clouds of rain. If the moon's light threatens to fail, firetipped arrows are shot up to it by the Hottentots; and the same remedy is applied by the Ojibways to the sun when eclipsed.

To complete these outlines of savage logic, it is only necessary to point out that hypothesis is an instrument of thought which is of great service in primitive speculation. A hypothesis is any assumption made for the purpose of explaining a fact or facts already known to be true. But whereas the assumptions of the savant are hypotheses, those of the savage are called myths. Thus, when it is sought to account for the observed fact that the moon periodically decreases in size and that her appearance in the sky is the signal for the departure of the sun, a savage hypothesis accounting for the facts is that sun and moon are husband and wife who have quarrelled and separated; periodically the moon makes overtures of reconciliation and periodically wastes away before our eyes in grief at their rejection. Or the observed facts of thunderstorms are accounted for on the supposition that a jar of rain is carried by one spirit and is smashed by the mace of another; whence the crash of the thunder and the descent of the rain. The importance of hypothesis as a savage instrument of thought may be judged by the fact that it is a quite tenable position that all the countless myths in the world were originally explanatory (ætiological) myths, primitive hypotheses.

It should now be clear that there is no fundamental difference between savage and scientific logic, but that, on the

contrary, they are fundamentally identical. The uniformity of nature, the principle of induction, the theory of causation, the inductive methods, form the common framework of both logics: the savage would probably be able to give his assent to all the principles of Mill's logic. In other words, the differences are not formal but material. The errors of the early logician were extra-logical, and therefore were such as could be remedied by no process of logic but only by wider experience. The problem of induction is to ascertain the cause (or effect) of a given phenomenon; and the cause (or effect) is to be looked for amongst the immediate antecedents (or consequents) of that phenomenon. But the antecedents (or consequents) comprise every single one of the countless changes which take place in any part of the universe the moment before (or after) the occurrence of the phenomenon under investigation: any one of these antecedents (or consequents) may be the cause (or effect), and there is nothing à priori or in logic to make us select one rather than another. It is plain, therefore, that as long as man is turned loose as it were amongst these innumerable possible causes with nothing to guide his choice, the chances against his making the right selection are considerable, and that to speak of the savage's choice as haphazard and illogical is to misconceive the nature of logic. It should also be clear that no progress could be made in science until man had distinguished, at anyrate roughly, possible from absolutely impossible effects (or causes), and had learned to dismiss from consideration the impossible. Now it might be expected that, as it was only experience which could show what was impossible, so experience would suffice of itself to teach man this essential distinction. But, as a matter of fact, experience by itself has done no such thing, as is shown by the simple fact that great as is the age and long as is the experience of the human race, the vast majority of its members have not yet learnt from experience that like does not necessarily produce like: four-fifths of mankind, probably, believe in sympathetic magic, and therefore neither need nor can make any intellectual progress, whilst the progressive minority are precisely those from amongst whom magic has been uprooted by its relentless foe, religion. The reason why the real order and sequence of natural events

does not mechanically impress itself in its correct form upon the human mind, is that the mind is not the passive recipient of external impressions, but reacts upon them and remodels them, so that the ultimate shape taken by them depends as much on the form of the mental mould, so to speak, into which they are poured, as it does upon their own nature. In other words, the mind does not pay equal attention to everything which is presented to it: it only sees what it is prepared to see. Thus the preconception that things causally related to one another must be similar and vice versa-a preconception due to the mental law by which similar ideas suggest one another is so strong as to prevent the savage from seeing facts which are at variance with it, and thus the experience which might be expected automatically to correct the error serves but to strengthen it. But when the consequences of that error came in conflict with the religious sentiment, that hostility between magic and religion was aroused of which the existence is universally admitted though differently explained.

Now the fallacy that things causally related must be similar to one another, is one that the human mind, from its very constitution, must have fallen into in its very first attempts to interpret the complex manifold of nature. It is also a fallacy from which most savages, who in this may be taken as representing primitive man, have not yet escaped. But the fallacy, though primeval, has nothing to do with magic or the supernatural: it requires for its existence no belief in supernatural powers or even in spirits, it might perfectly well flourish in a region where neither religion nor magic had been heard of. Thus the fact of a man's using this fallacious mode of procedure to produce or forecast certain desired results does not in the least tend to show that he considers the process itself to be magical or supernatural; the savage who wears an iron ring to give strength to his body has not advanced so far in science as the man who takes iron in a tonic, but he no more believes himself to be dealing in magic and spells than the educated persons of to-day do who forecast the weather by the changes of the moon.

This will perhaps be made clearer if it be pointed out that it is not merely the fallacy of "like produces like," but

the inductive methods themselves which the savage uses in order to work his wonders. Most of the examples of savage logic already given in this chapter are instances of "sympathetic magic "; but as the means which the savage employs for this purpose are precisely those used for the ordinary commonplace purposes of life both by him and by civilised. man, it cannot be argued that those means are in themselves considered magical or supernatural.

These, then, are the grounds on which it is here maintained that sympathetic magic, which is the germ of all magic, does not involve in itself the idea of the supernatural, but was simply the applied science of the savage. Yet out of the theory of causation and the methods of induction, which under certain rare, favouring conditions, and with the assistance of the religious sentiment, developed into modern science, elsewhere the process of evolution produced "one of the most pernicious delusions that ever vexed mankind, the belief in magic." It remains for us to inquire how this came about.

Art magic is the exercise by man of powers which are supernatural, i.e. of powers which by their definition it is beyond man to exercise. Thus the very conception of magic is one which is essentially inconsistent with itself; and, being such, the belief in it seems to be thought by many writers to require no further explanation. Now, doubtless it is the conception's very inconsistency with itself which gives it its fascination; the prospect of being able to do the impossible is singularly attractive. At anyrate the hold which the idea, when once introduced, has over the mind of man is so familiar a fact that it does not need to be proved. But all this does not show how the idea ever could have occurred to the human mind in the first instance; it only proves what a very suitable nidus was ready for the germ when it should come. To read some writers, who derive the powers of priests (and even of the gods) from those of the magician, and who consider apparently that magic requires no explanation, one would imagine that the savage, surrounded by supernatural powers and a prey to supernatural terrors, one day conceived the happy idea that he too would himself exercise supernatural power-and the thing was done;