Similar ideas were put forward in a more comprehensive form by J. F. McLennan (Studies in Ancient History, 1876). He starts from the wide occurrence of marriage by capture in primi tive societies, and groups the tribes of which we have definite knowledge into endogamous and exogamous societies according as they take their wives from among the kindred or outside it. Marriage by capture and by purchase are signs of exogamy, con nected with the custom in many tribes of killing female offspring. The development of marriage by capture and purchase is a power ful agent in bringing about patriarchal rule, agnatic relation ship, and the formation of clans or gentes, but the more primitive forms of relationship appear as variations of systems based on mother-right. These views are supported by ethnological observa tions and used as a clue to the history of relationship and family law in ancient Greece. In further contributions, published after his death, the peculiarities of exogamous societies are traced back to the even more primitive practice of Totemism (q.v.). McLennan's line of inquiry was taken up in a very effective man ner not only by anthropologists like E. B. Tylor or scholars like A. Lang, but also in a more special manner by students of primi tive family law. A most interesting monograph in this direction is Robertson Smith's study of Kinship and Marriage in Arabia.
But perhaps the most decisive influence was exercised by the discoveries of an American, Lewis H. Morgan. In his epoch making works on Systems of Consanguinity (1869) and on An cient Society (1877) he drew attention to the remarkable fact that in the case of a number of tribes—the Red Indians of America, the Australian black fellows, some of the polar races, and several Asiatic tribes, mostly of Turanian race—degrees of relationship are reckoned and distinguished by names, not as ties between individuals, but as ties between entire groups, classes or generations. Instead of a mother and a father a man speaks of fathers and mothers; all the individuals of a certain group are deemed husbands or wives of corresponding individuals of an other group; sisters and brothers have to be sought in entire generations, and not among the descendants of a definite and common parent, and so forth. There are variations and types in these forms of organization, and intermediate links may be traced between unions of consanguine people—brothers and sis ters of the same blood—on the one hand, and the monogamic marriage prevailing nowadays, on the other. Facts of this kind produce very peculiar and elaborate systems of relationship, which have been copiously illustrated by Morgan in his tables.
In his Ancient Society he attempted to reduce all the known forms and facts of marriage and kinship arrangements to a com prehensive view of evolution leading up to the Aryan, Semitic and Uralian family, as exhibiting the most modern type of re lationship.
To complicate the matter still further the existence of classi ficatory systems of relationship may or may not be accompanied by an ignorance of the fact of paternity. Many of the simpler peoples see no causal connection between conception and birth. This surprising fact first arose to startle the scientific world in the course of field work by Spencer and Gillen among the Arunta in Australia. It was subsequently confirmed by Rivers for three entirely unconnected peoples and elaborately tested by Malinow ski in the case of the Trobriand Islanders (see his Sex in Primitive Societies, Psyche, vols. iv. and v. and the I.R.A.I. 1916). It
may therefore now be taken as completely established. For a psychological discussion of the ignorance in question, see a valu able article by Ernest Jones, "Mother Right and Sexual Igno rance of Savages," International Journal of Psycho-Analysis, vol. vi., p. 109 (1915).
Totemism and exogamy, sexual ignorance and the classifica tion system have become one of the most difficult topics in anthropology and equally one of the most eagerly canvassed. There is a voluminous literature on the subject ; Frazer's Totem ism and Exogamy heading the list. This great work is a mine of in formation, and all the principal explanations to account for what after all remain most perplexing phenomena will be found there fully discussed. Students will also refer to his Folk lore in the Old Testament, II. vi., and to studies of special peoples that will be found in the I.R.A.I.
Among other works which should be mentioned is Freud's Totem and Taboo which offers some highly stimulating psycho logical reading. The best criticism of it is still that of A. L. Kroeber (22 American Anthropologist, 1920, p. 48-55) but un fortunately his note of warning, that henceforth Freud's thesis could never be ignored without stultification, has in some instances passed unheeded, with the result that much patient work has been entirely wasted. On the other hand, some of Freud's fol lowers have not allowed his brilliant suggestions to fructify, but, in their zeal, have reduced them to rigid formulae and spent themselves in elaboration.
Thus there has come about by itself a complete change of perspective in the comparative study of man and society. The right of ethnologists to have their say in regard to legal, political and social development was forcibly illustrated from both ends, as it were. On the one hand, classical antiquity itself proved to be a rather thin layer of human civilization hardly sufficient to conceal the long periods of barbarism and primitive evolution which had gone to its making. On the other hand, unexpected combinations in regard to family, property, social order, were discovered in every corner of the inhabited world, and our trite notions as to the character of laws and institutions were reduced to the rank of variations on themes which recur over and over again, but may be and have been treated in very different ways. There is no need to speak of the use made of ethnological material in the wider range of anthropological and sociological studies—the works of Westermarck, Tylor and Frazer are in everybody's hands—but attention must be called to the further influence of the ethnological point of view in comparative juris prudence. An interesting example of the passage from one line of investigation to another, from the historical to the anthro pological, is presented in the works of one of the founders of the Zeitschrift fur Vergleichende Rechtswissenschaft—Franz Bern hoft. He appears in his earlier books as an exponent of the com parative study of Greek and Roman antiquities, more or less in the style of Leist. Like the latter he was gradually incited to draw India into the range of his observations, but unlike Leist, he ended by fully recognizing the importance of ethnological evi dence, and the influence of Bachofen and of the ethnologists made itself felt in Bernhoft's treatment of classical antiquity itself. A somewhat similar process will occur to the minds of English readers in the case of Jane Harrison, Gilbert Murray, Sir Arthur Evans and others.