him would be but inventions, and would deserve a much lower place in the history of belief. When the gods have become like men, and have lost all memory of the phenomena out of which they sprang, they have laid aside the individuality of their characters; henceforward they will tend more and more towards uniformity of nature; and this uniform nature will more and more adapt itself to a godlike ideal. Thus the influence of moral ideas will become paramount while the influence of the experience of outward nature fades away.

Of the growth of morality in belief, and of the way in which it may develope along with the contemplation of mere external phenomena, we have an excellent example in one among those mythic beings which Sidney enumerates. All the three--the Cyclops, the Chimæra, the Furies—are fearful creations; but the first two draw all their terror directly from the things which they personify; they are fearful because the storm itself is fearful. No natural dread surrounds Erinys, who is the Dawn; her terribleness arises solely from a moral character which the Dawn is led to take upon her. She is the detector of crimes; at first in the merely passive way in which we say that all crimes will some day come to light, afterwards in a more active sense. In time the Erinyes become altogether moral beings, and purely abstract ones, the honoured ancient deities, supporters of the throne of Justice, dear to Zeus,' whom Eschylus knows. Yet all this moral character springs out of their natural character. They become the detectors of crimes solely because the daylight must be a detector of crimes.

These three examples are fairly typical of the whole range of beings who play the mythic dramas of a people. Though all must have had a beginning in outward nature,

This is Max Müller's explanation of the origin of the Erinyes (Chips, ii. 153); and it seems to me a valid one, despite the criticisms of Welcker (Griech. Götterl. iii. p. 75, &c.) and the different origin found for Saranyû by Kuhn (Zeitsch. für verg. Sp. i. 439). Gubernatis makes some suggestions which tend to reconcile these discrepancies (Mitol. Ved. p. 156).

EARLY PHASES OF BELIEF.

29

some (as the Greek furies do) will have strayed far, others less far from it. Some will keep the whole nature which belongs to outward things, some will half clothe themselves with a human personality. But never in early times shall we have a god unlinked to external phenomena. Wherefore if we read of some primitive race retiring to worship in its rocky fastnesses or woody solitudes, as Tacitus says the Germans retired to their forest haunts and worshipped an Unseen Presence there, we must not think of them going to meditate upon the riches and goodness, nor yet upon the power and wonder, of God. The presence made known to them may be an unseen, it is certainly not an unfelt one; it is in the breath of the wind or in the murmuring of the stream; it is in the storm or in the whirlwind, but it is not yet in the voice of the heart. The sensations of this external nature stir man's imagination, they raise his awe; and this stirring of the inner senses constitutes his worship. And let those doubt that religion may have had such beginnings who have never listened to the voices which arise from the solitudes of nature; those who have never known the brightness of sunny fields and streams, the sad solemnity of forests, and the majesty of mountains or of the sea.

§ 2. Early Phases of Belief.

Thus much to show the mere existence and the essential character of this faculty of Belief. We have now to say something concerning the phases of it. Here the history of language will still be our guide. What we have at present learned of the parallel histories of religion and of language is this: That, as at first all words expressed only the ideas of definite material objects, but many of these words which had once a purely material significance came in time to have a purely moral or metaphysical significance, so throughout all the natural world, though men at first gained from it only ideas of outward sensation, these in time changed, and metaphysical and moral ideas came to

take their place. In the case of words the change from the physical to the metaphysical use was not, we may be sure, made at a bound. Stretched did not suddenly come to mean right, nor heap to mean truth.

Now one stage in that slow process of change we can certainly detect. The first step was made when the name for an individual thing had expanded its meaning to take in a class of things. When words, from being individual, or what we now call proper names, had grown to be generic terms, they had already become half abstractions, for they had become names for aggregates of qualities and not for individual things. I took just now stretched as the example of a word in its most material form; but in reality a word was in its most material form only so long as it was not an adjective, but expressed some single object. If we could imagine for a moment the word straight or stretched as the name, not of any string, but of some particular string, then we should have a word in its most primitive possible condition. The next stage would be when the same word was used to express a class of objects-in this case all strings which had been stretched. The stage which would immediately follow would be that the word should come to be an adjective (an attribute), and no longer an individual name. We have every reason to suppose that the process of the human thought, exemplified by the history of words, is traceable equally well in the development of belief; whence it would follow that belief too has passed from individual objects to groups of things, and thence has fastened upon some attribute, still physical, but no longer apprehensible by all the senses, which belonged to the whole class. In a word, religion began with fetichism, with the worship, we will suppose, of an individual tree; it passed on to the worship of many trees, of the grove of trees, and it soon proceeded thence to a worship of some invisible belonging of the grove. This might be the sacred silence which seems to reign in the wood, or the storm which rushes through it, or any of the dim, mysterious

MEANING OF FETICH.'

31

forest sounds. From the visible and tangible things of earth religion looked farther away to the heavenly bodies, or to the sky itself. And then at last it emerged from the nature-worshipping stage, and the voice of God, which was heard once in the whirlwind, was now heard only in the still small voice within.

With the last phase of all we shall in these chapters have nothing to do; nothing directly, at all events. It scarcely needs to be said that no one of the three phases of belief which I have described is to be found in its purity among any of the peoples whose religious career we are going to study. Each phase is found mingled with some other. All the Indo-European races have arrived at some point in the third condition of development; that is to say, all have achieved some idea of an abstract god, who is separate from phenomena. But few or none of them have completely left behind any of the other two conditions of belief. Wherefore it lies in our hands which phase we choose to study. The strata of belief are like the geological strata; primitive ones may be discovered sometimes quite near the surface; the nature of the former are no more to be told by measuring their distance from us in time than that of the latter by any measurement from the surface of the earth. It is the character and not the actual time of the formation which allows us to call it primitive; and both the first two phases of belief, both pure fetichism and that which, to distinguish it from fetichism, we may call nature worship, both, wherever they are encountered, may fairly be called phases of primitive belief.

The same kind of difficulty over the meaning of a word which has obscured discussion upon the nature of religion itself has been stirred up, in a minor degree, about the word fetich; and here with less excuse, for this word carries with it no strength of old association. It was never during the days of its early use applied with scientific ex

actness, and it was first employed at a time when the study of belief had, in any effective way, hardly begun. If, therefore, we were to wrest the word a little from its first application, in order to make it serve us in a scientific sense, there would be no great harm.

Mr. Max Müller has, with many strong arguments, called in question the very general assumption-systematised somewhat in the hands of Comte-that fetichism lies at the root of all religion. His arguments have certainly been sufficient to make us reconsider our use of the word fetichism, and in future to define it more exactly; but I do not think they have really shaken the position which Comte has taken up on this point. It is one thing to show that the great positive philosopher has not used 'fetichism' in its etymological significance, or even that he has not always attached to the term the same meaning, and that others who followed him have been yet more vague in the use of the word; it is another thing to show that there has been no primitive belief clinging to the worship of visible external things.

Fetich (feitiço) was, it is known, the general name by which the Portuguese sailors in African seas called the charms and talismans they wore their beads, or crosses, or images in lead or wood. Seeing that the native Africans likewise had their cherished amulets (their grigris), deemed by them sacred and magically powerful, the Portuguese called these by the same name of fetich. Then, in 1760, came De Brosses, with his book on 'Les Dieux fétiches,' proposing this condition of belief as an initial state of religion. His term as well as his views were adopted, and fetichism assumed a fixed place in the history of religion.

Neither the feitiço of the Portuguese mariner, nor any Christian amulet or relic, is distinctive of a primitive phase of belief; and if it were a mere question of etymology this would be enough to show that 'fetichism' did not correctly describe the phase of belief which we do intend to