between the different groups, both in their common responsibility for crimes, and in their maintenance of the rights of property. But there is not a single witness which would entitle us to infer that there was originally a female line of descent. The local isolation of the clans maintained by their distribution into districts is only possible where there is a male line; we have already shown that a confusion inevitably results from the female line. We are wholly ignorant how the clans arose; no conclusions can be drawn from the fact that the gods, or images of the dead, served as a centre for each clan, since it remains doubtful whether the severance into groups was due to their common worship of the same gods, or if men worshipped the same god because they were of the same blood.

Peru, before the Incas obtained possession of the country, was apparently inhabited by a number of peoples, whose organization resembled that found in Brazil and on the Pampas. Garcilasso's description of their rude state is perhaps overdrawn, yet this much I believe to be true. Each province and nation, and frequently each village, spoke a different dialect, which was unintelligible to their neighbours. Those who spoke a common language were in close connection with each other, while they lived at enmity and in perpetual warfare with those whom they did not understand.2 This well-known classification into villages was maintained in a strongly-marked form under the Incas; no member of a given district or village could marry into another, and he was at the same time bound to the soil.3 When we are confronted with these facts, it becomes impossible to find in the tribal forms of South America the survival of an earlier and diverse tribal organization.

1 Herrera, vol. iii. pp. 321, 359, 370, 379. 2 Garcilasso de la Vega, p. 31.

Waitz, vol. iv. pp. 76, 306. 3 Ibid., p. 189.

THE COUVADE.

Before taking leave of America, we must consider some phenomena which are commonly supposed to be signs of the prior observance of the female line of descent; namely, the special customs which are included under the name of the couvade. This name is applied to the custom which obliges the father to occupy the lying-in bed instead of the mother. Lubbock attempts to explain this strange custom by saying that when the transition. from the female to the male line of descent took place, the father assumed in all respects the position of the mother. Thus it was natural that he should refrain · from everything which might hurt the child, both from certain acts and certain food. Lubbock ascribes the origin of the couvade to this abstinence, which the mother had really reason to observe, but in the father's case it was fictitious. The father was now, accordingly, regarded as the true progenitor, as the mother had formerly been so regarded.1

It is scarcely necessary to observe that the father would not submit to the couvade if he believed himself to be in no way connected with the child, and so far we agree with Lubbock. But the facts distinctly contradict his assumption that the couvade was due to the transition to the male line of descent, and a much simpler solution of the question is afforded by the later passages in the works quoted by him. Schomburgk states that the Arowaks agree with other tribes in the performance of these ceremonies of the lying-in bed by the husband together with the wife, and he says the same of the Macusis. Yet we have seen above that these two tribes observe the female line of descent, and this fact makes Lubbock's explanation untenable.

Lubbock, Orig. of Civ., p. 154. Appendix XVI. See Tylor, Early History, p. 292. Bachofen, Mutterrecht, p. 419. Ploss, Das Kind, vol. i. p. 35, and ch. v.

2 Schomburgk, vol. ii. pp. 314, 459. Brett, pp. 98, 101, states that the wife takes no part.

In order to find another explanation, an exact acquaintance with the phenomenon is necessary. I quote several accounts of this strange custom in the Appendix,1 and they all show that the husband's lying-in bed is not regarded as a sick-bed, affording rest and strength after travail. The man fasts with extreme rigour, and abstains for a long while from certain meats which might injure the child; birds and fish would give it the stomach ache, turtle would cause it to be deaf, etc. As Tylor has already asserted, the couvade only expresses the belief in a secret mystical connection between the father and the child, a belief which does not go beyond what is known in other ways of the mode of thought in primitive man. The belief that it is possible to inherit the courage of a dead man by eating his heart, or that a man may bewitched by incantations over a lock of his hair, springs from ideas which will also include the couvade.

be

"In

3יי

order that the father's courage may devolve upon the child, the former subjects himself to the most painful ordeal on the birth of a son or daughter, and a young woman also submits to cruel rites on attaining maturity.' Among the Iroquois, a mother who shrieks during her labour is forbidden to bear other children, and some of the South American Indians killed the children of mothers who have shrieked, from the belief that they will grow up to be cowards. The origin of the couvade cannot be traced to father or mother; the well-being of the child is its object; the father's powers of endurance are displayed on such occasions, and might thus be assured to the child, for no one who was deficient in courage and endurance would submit to this custom.

1 Appendix XVII.

4

2 Max Müller, strangely enough, traces the origin of the couvade to the derision of friends of both sexes (ch. ii. p. 278).

* Schomburgk, vol. ii. p. 431.

4

Lafitau, vol. i. p 592.

CHAPTER III.

AFRICA.

The regulating forces universally simple-The female line a sign of the formation of more permanent groups-Scattered groups: slave-hunts -Conquest of Africa-Twofold line of descent-Bechuana clansFatriarchal life-The kotla-Bantu and Negro villages-Their government-The village and the family-Women's prominent position-Bachofen and Giraud-Teulon-The wife and her familyNegroes-Female line-Description of West African community— Government of tribe and clan-Bechuanas the primitive type-Hottentots.

WE have endeavoured to show that the development of primitive social life lies between the Brazilian and North American forms. The social life of man begins with the partially agnatistic family, and the family group which is ruled by the father in virtue of his physical superiority; a strong man, or one who is pre-eminent in other ways, commands respect, and is obeyed by his associates, whether of the family or of the tribe. Clans are subsequently formed which, as their internal cohesion increases, gradually pass from the paternal to the maternal line of descent; the clan is then ruled by hereditary chiefs, and the family as a privileged group is lost. The organization of the community is not decided by considerations of descent, yet the same general causes also decide between the possible lines of descent. As in all cases which have to do with forces at work in wide circles, we find that here also the efficient causes are very

simple and direct. It was the influence of locality which first assigned the child to its father, and afterwards, when the mother lived apart from the other wives, to its mother; this was still more the case when the woman's clan interposed between her and the husband who belonged to another clan.

When we take a general view of these facts, the preliminary condition of the successful establishment of the female line appears to consist in the distribution of the masses into several smaller societies. Our researches show that, in accordance with the principle of this distribution, all which is found in one locality belongs to the owner of that locality. Since their inclinations and passions are social, this law is adopted by men. Whoever is drawn to a locality owned by another forfeits his own independence, and we have already seen that the family served as the nucleus of this first crystallization. The man took his wife to his own abode, and ruled both her and her children. These first groups might either be dissolved, owing to continual disputes, and then always drift further apart, or they might live together in amity, until they were by degrees so closely united that they could not be severed without losing their existence. The vitality of the group was involved in maintaining the female line of descent, since it arose from its power of retaining and protecting its members, but it destroyed the natural basis of the family by its opposition to the will of the husband. Since the natural tendency to family life can never be wholly subdued, the clan which was organized in conformity with the female line had always to contend with an enemy at home; the local limits of the clan were confused by marriage, and this confusion undermined the dominion of the female line.

We do not deny that there may have been other starting-points for groups besides that of the family. Whatever produces community among individuals, whether they have in common their name, their tattoo