at the conclusion that among the Polynesians this practice. was not an original heirloom brought with them from their primitive homes in the Far West, but was adopted subsequently by a few of the tribes under conditions and circumstances now unknown. For various reasons, then, it is an illegitimate supposition to regard the cannibalism of modern savages as a survival from the first infancy of mankind, or, more generally, from a stage through which the whole human race has passed.

As for the moral opinions about cannibalism, we may assume that peoples who abstain from it also generally disapprove of it, or would do so if they were aware of its being practised. Aversion, as we have often noticed, leads to moral indignation, especially where the moral judgment is little influenced by reflection. Another

source of the condemnation of cannibalism may be sympathetic resentment resulting from the idea that the dead is annihilated or otherwise injured by the act, or from the feeling that it is an insult to him to use his body as an article of food; but this could certainly not be the origin of savages' disapproval of eating their foes. Among civilised races, as well as among non-anthropophagous savages, horror or disgust is undoubtedly the chief reason why cannibalism is condemned as wrong. This emotion is often so intense that the same people whose moral feelings are little affected by a conquest, with all its horrors, made for the purpose of gain, shudder at the stories of wars waged by famished savages for the purpose of procuring human flesh for food. On the other hand, where the natural aversion to such food is for some reason or other overcome, the disapproval of cannibalism is in consequence no longer felt. But an attitude of moral indifference towards this practice has also been advocated on a totally different ground, by persons whose moral emotions are too much tempered by thought to allow them to pronounce an act as wrong simply because it creates in them disgust.

1 Fornander, Account of the Polynesian Race, i. 132.

Thus, Montaigne argued that it is more barbarous to torture a man to death under colour of piety and religion than to roast and eat him after he is dead.1 And he quotes. with apparent agreement the opinion of some Stoic philosophers that there is no harm in feeding upon human carcases to avoid starvation.2

1 Montaigne, Essais, i. 30.

2 Diogenes Laertius, Vita philoso

phorum, vii. 1. 64 (121); vii. 7. 12 (188). Zeller, Stoics, p. 307.

CHAPTER XLVII

THE BELIEF IN SUPERNATURAL BEINGS

We now come to the last of those six groups of moral ideas into which we have divided our subject-ideas concerning conduct towards beings, real or imaginary, that are regarded as supernatural. But before we enter upon a discussion of human behaviour in relation to such beings, it is necessary to say some words about man's belief in their existence and the general qualities attributed to them.

Men distinguish between two classes of phenomena"natural" and "supernatural," between phenomena which they are familiar with and, in consequence, ascribe to "natural causes," and other phenomena which seem to them unfamiliar, mysterious, and are therefore supposed to spring from causes of a "supernatural" character. We meet with this distinction at the lowest stages of culture known to us, as well as at higher stages. It may be that in the mind of a savage the natural and supernatural are often confused, and that no definite limit can be drawn between the phenomena which he refers to the one class and those which he refers to the other; but he certainly sees a difference between events of everyday occurrence or ordinary objects of nature and other events or objects which fill him with mysterious awe. The germ of such a

1 I do not share the objections raised by various writers to the term " supernatural." It has the sanction of common usage; and I consider it preferable

to the word "superhuman," when applied to inanimate things or animals which are objects of worship.

distinction is found even in the lower animal world. The horse fears the whip but it does not make him shy; on the other hand, he may shy when he sees an umbrella opened before him or a paper moving on the ground. The whip is well known to the horse, whereas the moving paper or umbrella is strange and uncanny. Dogs and cats are alarmed by an unusual noise or appearance, and remain uneasy till they have by examination satisfied themselves of the nature of its cause.1 Professor Romanes frightened a dog by attaching a fine thread to a bone and surreptitiously drawing it from the animal, giving to the bone the appearance of self-movement; and the same dog was frightened by soap-bubbles.2 Even a lion is

scared by an unexpected noise or the sight of an unfamiliar object; a horse, the lion's favourite prey, has been known to wander for days in the vicinity of a troop of these animals and be left unmolested simply because it was blanketed and knee-haltered. And we are told of a tiger which stood trembling and roaring in an ecstasy of fear when a mouse tied by a string to a stick had been inserted into its cage. Little children are apt to be terrified by the strange and irregular behaviour of a feather as it glides along the floor or lifts itself into the air.5

But the primitive mind not only distinguishes between the natural and the supernatural, it makes, practically, yet a further distinction. The supernatural, like the natural, may be looked upon in the light of mechanical energy, which discharges itself without the aid of any volitional activity. This is, for instance, the case with the supernatural force inherent in a tabooed object; mere contact with such an object communicates the taboo infection. So also the baneful energy in a curse is originally conceived as a kind of supernatural miasma, which injures or destroys anybody to whom it cleaves; in fact, to

1 Morgan, Animal Life and Intelligence, p. 339.

2 Romanes, Animal Intelligence, p. 455 sq.

Gillmore, quoted by King, The Supernatural, p. 80.

4 Basil Hall, quoted ibid. p. 81. See also ibid. p. 78 sqq.; Vignioli, Myth and Science, p. 58 sqq.

sq.

5

Sully, Studies of Childhood, p. 205

STATEL TOWER, or the other tands appled in majic He vioo performs a pure & magal at uses such power of a super

It

The ILS what we generally understand by rigen and maga But in the Lann word reign: there seems to be no ina mber of such a distinction. Religio is probably related to wypark, wild means "to tie." is commonly assumed that the relationship between these words implies that in re gron man was supposed to be tied by his god. But I venture to believe that the connection between them allows of another and more natural interpretation that it was not the man who was tied by the god, but the god who was ted by the man. This interpretation was suggested to me by certain ideas and practices prevalent in Morocco. The Moors are in the habit of tying raga to objects belonging to a siyid, that is, a place where a saint has, or is supposed to have, his tomb, or where such a person is said to have sat or camped. In very many cases, at least, this tying of rags is ‘ár upon the

Though somewhat indefinite, the *pathet "regardfl” seems a necessary attribute of a religious act. We do not

call it religion when a savage flogs his fetish to make it submissive.