of blood-covenanting is reckoned in the East even a closer tie than that of natural descent," and the same was the case among the ancient Scandinavians.2 I do not see how Mr. Hartland's theory can account for this. Mingling of blood is sometimes supposed to be a direct cause of mutual sympathy and agreement, in accordance with the principle of transmission of properties by contact; even in Europe there are traces of the belief that a few drops of blood transferred from one person to another inspire the recipient with friendly feelings towards him with whose blood he is inoculated. But the genuine blood-covenant imposes duties on both parties, and also contains the potential punishment for their transgression. It involves a promise, and the transference of blood is vaguely or distinctly supposed to convey to the person who drinks it, or who is inoculated with it, a conditional curse which will injure or destroy him should he break his promise. That this is the main idea underlying the blood-covenant appears from the fact that it is regularly accompanied by curses or self-imprecations. In Madagascar, for instance, when two or more persons have agreed on forming the bond of fraternity, a fowl is procured, its head is nearly cut off, and it is left in this state to continue bleeding during the ceremony. The parties then pronounce a long imprecation and mutual vow over the blood, saying, inter alia, "O this miserable fowl weltering in its blood! thy liver do we eat, thy liver do we eat; and should either of us retract from the terms of this oath, let him instantly become a fool, let him instantly become blind, let this covenant prove a curse to him." A small portion of blood is then drawn from each individual and drunk by the covenanting parties with execrations of vengeance on each other in case of either violating the sacred oath. According to another description the parties, after they have drunk each other's blood, drink a mixture from the same bowl, praying that it may turn into 1 Trumbull, Blood Covenant, p. 10. 2 Maurer, Bekehrung des Norwegischen Stammes, ii. 171. 3 Cf. Crawley, Mystic Rose, p. 236 sq. 4 von Wlislocki, 'Menschenblut im Glauben der Zigeuner,' in Am UrQuell, iii. 64. Dörfler, 'Das Blut im magyarischen Volkglauben,' ibid. iii. 269 sq. 5 Forbes, A Naturalist's Wanderings in the Eastern Archipelago, p. 452 (natives of Timor). Burns, Kayans of the North-West of Borneo,' in Jour. of the Indian Archipelago, iii. 146 sq. New, Life, Wanderings, and Labours in Eastern Africa, p. 364 (Taveta). Decle, Three Years in Savage Africa, p. 494 (Wakamba). Trumbull, op. cit. pp. 9, 20, 31, 42, 45-47, 53, 61 sq. For the practice of sealing an agree ment by transference of blood accompanied by an oath, see also Partridge, Cross River Natives, p. 191 (pagans of Obubura Hill district in Southern Nigeria). Ellis, History of Madagascar, i. 187 599. 2 poison for him who fails to keep the oath.1 As we have seen before, blood is commonly regarded as a particularly efficient conductor of curses, and what could in this respect be more excellent than the blood of the very person who utters the curse? But the blood of a victim sacrificed on the occasion may serve the same purpose, or some other suitable vehicle may be chosen to transfer the imprecation. The Masai in the old days "spat at a man with whom they swore eternal friendship"; and the meaning of this seems clear when we hear that they spit copiously when cursing, and that "if a man while cursing spits in his enemy's eyes blindness is supposed to follow."3 The ancient Arabs, besides swearing alliance and protection by dipping their hands in a pan of blood and tasting the contents, had a covenant known as the hilf al-fodûl, which was made by taking Zemzem water and washing the corners of the Ka'ba with it, whereafter it was drunk by the parties concerned.* The blood-covenant is essentially based on the same idea as underlies the Moorish custom of sealing a compact of friendship by a common meal at the tomb of some saint, the meaning of which is obvious from the phrase that "the food will repay' him who breaks the compact.5 Besides marriage, local proximity, and a common descent, a common worship may tie people together into 1 Dumont d'Urville, Voyage pittoresque autour du monde, i. 81. Hinde, Last of the Masai, p. 47. See also Johnston, Uganda, ii. 833. Hinde, op. cit. p. 48. Robertson Smith, Marriage and Kinship in Early Arabia, p. 56 sqq. Cf. Herodotus, iii. 8. See supra, i. 587. According to another theory the inoculated blood is regarded as a pledge or deposit, which compels the person from whom it was drawn to be faithful to the person to whom it was transferred. Suppose that two individuals, A and B, become "blood-brothers" by mutual inoculation. Each, then, Mr. Crawley argues (Mystic Rose, p. 236 sq.), has a part of the other in his keeping, each has "given himself away to the other in a very real sense; and the possibility of mutual treachery or wrong is prevented both by the fact that injury done to B by A is considered equivalent to injury done by A to himself, and also VOL. II by the belief that if B is wronged he may work vengeance by injuring the part of A which he possesses. To this explanation, however, serious objections may be raised. The belief in sym pathetic magic does not imply that injury done to B by A is eo ipso supposed to affect A himself through that part of him which has been deposited in B; it does not imply that two things which have once been conjoined remain, when quite dissevered from each other, in such a relation that "whatever is done to the one must similarly affect the other" (Frazer, Golden Bough, i. 49), unless there is an intention to this effect in the agent. The severed part then serves as a medium by which magic influence is transferred to the whole. Again, it is difficult to see how B could injure A through the part of him which he possesses when that part has been absorbed into his own system, as must be the case with those few drops of A's blood with which he was inoculated. Р social union. But among savages a religious community generally coincides with a community of some other kind. There are tutelary gods of families, clans, and tribes;1 and a purely local group may also form a religious community by itself. Major Ellis observes that with some two or three exceptions all the gods worshipped by the Tshi-speaking tribes on the Gold Coast are exclusively local and have a limited area of worship. If they are nature-gods they are bound up with the natural objects they animate, if they are ghost-gods they are localised by the place of sepulture, and if they are tutelary deities whose origin has been forgotten their position is necessarily fixed by that of the town, village, or family they protect; in any case they are worshipped only by those who live in the neighbourhood, the only exceptions being the sky-god, the earthquake-god, and the goddess of the silkcotton trees, who are worshipped everywhere.2 When the religious community is thus at the same time a family, clan, village, or tribe, it is of course impossible exactly to distinguish the social influence of the common religion from that exercised by marriage, local proximity, or a common descent. It seems, however, that the importance of the religious bond, or at least of the totem bond, has been somewhat exaggerated by a certain school of anthropologists. We are told that in early society "each member of the kin testifies and renews his union with the rest" by taking part in a sacrificial meal in which the totem god is eaten by his worshippers. But no satisfactory evidence has ever been given in support of this theory. Dr. Frazer knows only one certain case of a totem sacrament, namely, that prevalent among the Arunta and some other tribes in Central Australia, who at the time of Intichiuma are in the habit of killing and eating totem animals; and this practice has nothing what 1 See infra, on Gods as Guardians of Morality. 2 Ellis, Yoruba-speaking Peoples of the Slave Coast, p. 284 sq. For various instances of village gods see Turner, Samoa, p. 18; Crozet, Voyage to Tas mania, &c. p. 45 (Maoris); Christian, 3 Hartland, op. cit. ii. 236. Frazer Golden Bough, i. p. xix. ever to do with the mutual relations between kindred. Its object is only to multiply in a magic manner the animals of certain species for the purpose of increasing the food-supply for other totemic groups. In his book on Totemism Dr. Frazer writes :-" The totem bond is stronger than the bond of blood or family in the modern sense. This is expressly stated of the clans of western Australia and of north-western America, and is probably true of all societies where totemism exists in full force. Hence in totem tribes every local group, being necessarily composed (owing to exogamy) of members of at least two totem clans, is liable to be dissolved at any moment into its totem elements by the outbreak of a blood feud, in which husband and wife must always (if the feud is between their clans) be arrayed on opposite sides, and in which the children will be arrayed against either their father or their mother, according as descent is traced through the mother, or through the father." In the two or three cases which Dr. Frazer quotes in support of his statement the totemic group is identical with the clan; hence it is impossible to decide whether the strength of the tie which unites its members is due to the totem relationship or to the common descent. But even the combined clan and totem systems seem at most only in exceptional cases to lead to such consequences as are indicated by Dr. Frazer's authorities. With reference to the Australian aborigines Mr. Curr observes :-"Of the children of one father being at war with him, or with each other, on the ground of maternal relationship, or any other ground, my inquiries and experience supply no instances. To Captain Grey's statements, indeed, there are several objections." 1 Spencer and Gillen, Native Tribes of Central Australia, ch. vi. Iidem, Northern Tribes of Central Australia, ch. ix. sq. 2 Frazer, Totemism, p. 57. 3 Grey, Journals of Expeditions in North-West and Western Australia, ii. 230. Petroff, Report on Alaska, p. 165. Hardisty, Loucheux Indians,' in Smithsonian Report, 1866, p. 315. 4 Curr, The Australian Race, i. 67. In Hardisty's statement, referring to the Loucheux Indians, there is a conspicuous lack of definiteness. He says:-"In war it was not tribe against tribe, but division against division, and Among the Arunta and some other Central Australian tribes we have fortunately an opportunity of studying the social influence of totemism apart from that of clanship, the division into totems being quite independent of the clan system. The whole district of a tribe may be mapped out into a large number of areas of various sizes, each of which centres in one or more spots where, in the dim past, certain mythical ancestors are said to have originated or camped during their wanderings, and where their spirits are still supposed to remain, associated with sacred stones, which the ancestors used to carry about with them. From these spirits have sprung, and still continue to spring, actual men and women, the members of the various totems being their reincarnations. At the spots where they remained, the ancestral spirits enter the bodies of women, and in consequence a child must belong to the totem of the spot at which the mother believes that it was conceived. A result of this is that no one totem is confined to the members of a particular clan or sub-clan,' and that though most members of a given horde or local group belong to the same totemic group, there is no absolute coincidence between these two kinds of organisation." How, then, does the fact that two persons belong to the same totem influence their social relationships? "In these tribes," say Messrs. Spencer and Gillen, "there is no such thing as the members of one totem being bound together in such a way that they must combine to fight on behalf of a member of the totem to which they belong. . . . The men to assist a particular man in a quarrel are those of his locality, and not of necessity those of the same totem as himself, indeed the latter consideration does not enter into account and in this as in other matters we see the strong as the children were never of the same caste (clan) as the father, the children would, of course, be against the father and the father against the children. . . . This, however, was not likely to occur very often, as the worst of parents would have naturally preferred peace to war with his own children." Petroff's passage concerning the Thlinkets, referred to by Dr. Frazer, simply runs :-"The ties of the totem or clanship are considered far stronger than those of blood relationship." 1 Spencer and Gillen, Native Tribes of Central Australia, ch. iv. 2 Ibid. pp. 9, 32, 34. |