MARRIAGE. 87 MARRIAGE. by the conmiunitj'. Tlie sanction may be moral, roliyious. or legal. This definition is broader than that of legal usage, which makes marriage only a legal form or the status corresponding thereto; and it is not so broad as Westerniarck's delinition, "a more or less durable connection between male and female, lasting beyond the mere act of propagation till after the birth of the oH'spring." Pro])erly speaking, the mating of animals is not marriage, and in no community of human beings is sexual union regarded as mar- riage until it is socially sanctioned in some waj'. On the other hand, communities which can hardly be said to have a positive law- not infre- quently attach the deepest signilicance to cus- tomary and religious sanctions applied to sexual relations. There has been, however, an unbroken continuity of historical forms, some of which have fallen short of marriage in any true sense, some of which have fallen short of marriage in a legal sense, while others, emerging as civil mar- riafje, have dropped the earlier religious sanc- tions. A complete understanding of marriage as a social institution, therefore, can be arrived at only through a survey of its historical evolu- tion. Such a survey shows us that the consortings of males with females among animals and among men have not been restricted to the simple mat- ing of one individual with one of the opposite sex which becomes the basis of monogamy. There have been iniions of one woman with two or more men (polyandry) and of one man with two or more women (polygyny), and such arrangements have been socially approved. It reveals also interesting restrictions, which have had a dis- tinct evolution of their own. marking off groups or classes that might not intermarry from those that might. Finally, it discloses the origin and development of the social .sanctions themselves, whereby natural mating becomes the social in- stitution, marriage. Distinguished ethnologists have maintained that relatively permanent sexual unions have slowly developed out of an original promiscuity. There is. however, no satisfactory evidence that a state of true promiscuity ever existed among human beings, and the hypothesis is rendered in- herently improbable by our knowledge that among the lower animals a distinct progress to- ward true pairing is observed as we ascend the scale from the lower to the higher vertebrata. It must be admitted that there are few life-long unicms of one male with one female in any animal sjiecies. even among the birds, whose ten- dencies toward an exclusive mating have been the subject of some exaggeration. As a rule in the animal kingdom within the reproductive period of life the female, no less than the male, consorts at one time or another with more than one individual of the other sex, and among the relatively numerous gregariois animals many females commonly associate with one male. Chief among the fac'ts which suggested the hypothesis of a primitivf promiscuity is the widespread custom among uncivilized men of tracing names and descent through the mother instead of the father. It has been shown that the civilized races also, including the peoples of Aryan culture, in all probability passed through this matronymic stage. Furthermore. an all-sufficient explanation of descent in the female line is found in the general instabilitv of pairing arrangements among primitive men. If a mother with her infant remains with her own kindred, or returns to them, she naturally keeps, and her child lakes, her clan name ; and her brethren or other near clansmen become the child's natural protectors. it seems probable that from the first sexual mating among human beings h,as tended toward monogamic unions, but that perm.inency has been of slow growth. Among the low'cst savages, such as the Australians, the Bushmen, the Fuegians, the forest hordes of Brazil, and the Innuit, a mating of one man with one woman for an indefinite, but usually not long period, is the common arrangement. Sometimes, as in Aus- tralian tribes, it is complicated by a .system of relationships more nominal than real, svich that each man in a given class or group is theoretic- ally the husband of each woman in some other class or group, and in like manner each woman in the latter class is theoretically the wife of each man in the former. These nominal unions probably do not point to a primitive promis- cuity, but rather to an early limitation of the range of choice in the selection of consorts ; that is to say, each woman of a certain class is a possible mate for any man of some other class. Nevertheless, in tribes somewhat more ad- vanced but usually dwelling in extreme poverty, various forms of polyandry, or the luiion of one woman with two or more men, or of a group of women to a group of men, is fotmd in many parts of the world, and undoubtedly prevailed widely in the past. In Tibetan polyandry, so called, the husbands are brothers. In Nair pol- yandry, or the form which prevails among the Nairs of India, the husbands of a woman may originally have been strangers to one another. Ca>sar speaks of a polyandry like the Tibetan as practiced among the Britons. In the Hawaiian Islands before they were invaded by whites, a common form was the so-called Punaluan family, in which a number of brothers cohabited with a group of sisters, each man consorting with each woman, and each woman with each man. The men were not own brothers of their wives, but Lewis H. Morgan, from evidence which he brought together in his work on Si/stcmf! of Con- san(]iiinUij and A/Jiiiitii. drew the conclusion that Punaluan polyandry had survived from what he called a 'Consanguine Family' formed by the mat- ing of near kindred, such as own brothers and sisters and cousins. A conservative explanation of the known facts seems to be that primitive hordes, except perhaps in the most favorable en- vironments, were small, as are the hordes of the lowest savages to-day. and were therefore composed of near kindred commonly marrying in and in. Under such circumstances the cohabit- ing group may often have been a consanguine family in Morgan's sense of the term, a Punaluan family, or a family like that created by the Tibetan polyandry. Yet probably from the first a temporary consorting of one man with one woman was the more frequent arrangement. A horde thus manying in and in is called mdoij- amous. Two ways in which a group becomes cxoqamous (taking consorts from other groups) are known. Where neighboring horiles. or groups of kindred, live on friendly terms with one another, often participating in common fes- tivities or religious observances, men frequently
