Communal marriage. Reference has already been made to the question of communal marriage in connection with the origin of exogamy. It has been observed that the evidence is insufficient to justify the belief that among any of the tribes or castes of this part of India the women are at the common service of all the men of the group. On the authority of a compilation entitled, “The People of India,”51 it has been regarded as established that “the Teehurs of Oudh live together almost indiscriminately in large communities, and even when two people are regarded as married the tie is but nominal.” This has been since quoted as one of the stock examples of communal marriage in India.52 Now of the Tiyars we have fairly complete accounts. The Oudh people of that name are a sept of Râjputs in the Sultânpur District, who do not appear in the enumeration of the last census. There is another body of Tiyars who are a sub-caste of the Mallâh, or boatman class, found to the number of 1,865 souls in the Ghâzipur District. They are numerous in Behâr and Bengal, and Mr. Risley has given a full account of them.53 There is no evidence whatever that anything like communal marriage [clxxxiv]prevails among them. The fact seems to be that by the necessities of their occupation the husbands leave their wives for long periods at a time and go on voyages as far as Calcutta. That a high standard of female morality is maintained during their absence it would be rash to assert: but this is very different from communal marriage. A rather better example comes from the Beriyas, one of the nomadic and criminal gypsy tribes. The girls of the tribe are reserved, in the Central Ganges-Jumna-Duâb, for prostitution, and if any member of the tribe marries a girl devoted to this occupation, he has to pay a fine to the tribal council. This is what Sir John Lubbock would term “expiation for marriage,” the annexation of the woman by one individual man of the group being regarded as improper.54 Dr. Westermarck, it may be remarked, disputes the connection of this custom with communal marriage.55

Laxity of female morality. 2. It is true that among many of the Dravidian tribes and those of the lower Himâlayas, like the Thârus, the standard of female morality is very low. Intrigues of unmarried girls, or even of married women, are very lightly regarded, provided the paramour is a clansman. Numerous instances of customs of this kind will be found in the following pages. The penalty on the relatives of the offenders is usually a fine in the shape of a compulsory feast to the tribesmen. On the other hand, the penalty is much more [clxxxv]severe if the woman’s lover belongs to a strange tribe. If he belongs to one of the higher tribes, the punishment is much less than if he belongs to one of the degraded menial races, such as the Dom, Dharkâr, or Bhangi. In such cases the woman is almost invariably permanently excommunicated. The tolerance of intertribal immorality, while significant is, however, far from actually legalised community of women.

The jus primæ noctis. 3. The custom of the jus primæ noctis has been also adduced as a proof of the existence of communal marriage. Of this the examples collected in the present survey are slight and inconclusive. The Ahîrs and many similar tribes have a custom of paying a fee to the village landlord at a marriage. This is known as mandwâna from mândo, the hut or pavilion in which the marriage is performed. This is hardly more than one of the common village manorial dues, and it is pressing the custom to an illegitimate extent to regard it as a commutation for the jus primæ noctis. There is reason to believe that in comparatively modern times some of the Râjas of Rîwa, a native state bordering on these Provinces, in their annual progresses, insisted on a supply of girls from the lower tribes, and there are still villages which are said to have been presented to the ancestors of women honoured in this way. But this is far from sufficient evidence for anything like the general prevalence of the custom, which is regarded with abhorrence by the public opinion of the country side. [clxxxvi]

Polyandry. 4. The same feeling prevails as regards polyandry which, according to Mr. McLennan, formed one of the regular stages in the evolution of marriage. There is certainly no ground for believing that at any time polyandry flourished as a permanent domestic institution. At the same time it seems quite certain that it has prevailed and does still prevail in Northern India, but usually among isolated communities and under exceptional circumstances.

5. To begin with the evidence from history or myth. The legend of the five Pândavas who took Draupadî as a joint wife, has been generally accepted as a proof that it existed among the people whom, for the sake of convenience, we call the early Aryans. It is true that the compilers of the Mahâbhârata clearly wish to refer to it as an exceptional case, and to whittle away its significance by representing it as a result of their misconception of their mother’s order. But there is reason to believe that it was not so exceptional as they endeavour to make out. In the discussion which followed, one of the princes quoted as a precedent the case of Jatilâ, “that most excellent of moral women who dwelt with seven saints, and Varkshî, the daughter of a Muni, who cohabited with ten brothers, all of them Prachetas, whose souls had been purified by penance.” We have next the case of the Aswins who had between them one woman, Sûryâ, the daughter of the sun. Even in the Râmâyana the giant Viradha imputes that Râma and [clxxxvii]Lakshmana jointly share the favours of Sîtâ.56 Professor Lassen’s theory that the whole story of Draupadî and her five lovers is only the symbolical indication of an alliance between the king of Panchâla and the five tribes represented by the five Pândavas has met with little support.

For the fraternal form of polyandry practised by some of the Himalayan races, there is ample evidence. According to Mr. Drew, a very careful observer, it originated in the smallness of the amount of land which could be tilled and the general inelasticity of the country’s resources: while the isolation from the rest of the world, isolation of manners, language and religions, as well as geographical isolation, hindered emigration.57 According to Dr. Wilson, polyandry in Tibet is not due to the scarcity of women, as a number of surplus women are provided for in the Lama nunneries.58

6. As regards the plains, we know that the prevalence of polyandry was noticed by the Greeks in the Panjâb.59 Of the Gakkars Farishta60 tells us that “it was the custom as soon as a female child was born to [clxxxviii]carry her to the door of the house and there proclaim aloud, holding the child with one hand, that any person who wanted a wife might now take her, otherwise she was immediately put to death. By this means they had more men than women, which occasioned the custom of several husbands to one wife. When the wife was visited by one of her husbands she left a mark at the door, which, being observed by any of the other husbands, he withdrew till the signal was taken away.” Similar customs prevailed among the Khokars of the Panjâb,61 and the Panjâb Jâts.62

7. In all these cases it would seem that polyandry is associated with, and in fact dependent on, female infanticide. In the course of the present survey, it has been ascertained that the custom prevails among some of the pastoral tribes, such as Ahîrs, Gûjars and Jâts, chiefly in the upper valleys of the Ganges and Jumna. It has even been embodied in the current proverb:—Do khasam kî joru, Chausar ka khel,—“The wife with two lords is like a game of backgammon.” The arrangement suits these pastoral people, who graze their herds in the river valleys. The brothers take it in turn to attend the cattle, and one remains at home in charge of the house-wife.

Niyoga and the levirate. 8. Whether the customs known as niyoga and the levirate are or are not connected with polyandry has been the subject of [clxxxix]much controversy. Mr. McLennan63 asserted that the levirate, that is the practice of marrying the widow of a deceased brother, was derived from polyandry. The niyoga, or the custom of a widow cohabiting with the brother of her deceased husband, seems to be referred to in the Veda.64 Manu65 allows such unions of a widow with a brother-in-law or other relative of the deceased husband to continue only till one or at the most two sons have been begotten, and declares that they must then cease. In the verses which follow he restricts such temporary unions to classes below the twice-born, or (in contradistinction to what proceeds) condemns them altogether. By the law, as stated by Gautama,66 a woman whose husband is dead, and who desires offspring, may bear a son to her brother-in-law. “Let her obtain the permission of her gurus (husband’s relatives under whose protection she lives), and let her have intercourse during the proper season only. On failure of a brother-in-law she may obtain offspring by cohabiting with a sapinda, or sagotra, or samân-pravara, or one who belongs to the same caste. Some declare that she shall cohabit with none but her brother-in-law. She shall not bear more than two sons. The child belongs to him who begot it, except if an agreement to the contrary have been made, and the child begotten at a living husband’s request on his wife [cxc]belongs to the husband, but if it was begotten by a stranger, it belongs to the latter, or to both the natural father and the husband of the mother, but being reared by the husband belongs to him.”

9. The best recent opinion is in opposition to the theory that the levirate or niyoga is a survival of polyandry. “The levir,” says Mr. Mayne, “did not take his brother’s widow as his wife. He simply did for his brother or other near relation, when deceased, what the latter might have authorised him, or any other person to do during his lifetime. And this, of course, explains why the issue so raised belonged to the deceased and not to the begetter. If it were a relic of polyandry, the issue would belong to the surviving polyandrous husband, and the wife would pass over to him as his wife.”67

10. In modern times, in this part of India, practically all the tribes which permit widow marriage allow the levirate in the restricted form that it is only the younger son of the late husband who is allowed or expected to take the widow to wife. Whatever may have been the idea connected with this practice in early times, the fiction that the son was supposed “to raise up seed unto his brother” seems to have altogether disappeared, and no survival of this rule of affiliation has been discovered. In fact, according to common custom, the widow is regarded as a kind of property which has been purchased into the family by the payment of the bride-price; [cxci]and among some of the Dravidian tribes there is a rule of tribal law that if the widow goes to live with a stranger to the family, he is bound to repay the bride-price, and in some cases the costs incurred in her first marriage, to her younger brother-in-law or his father. It is noticeable that in this form of the levirate alliance with the elder brother of her late husband is rigidly prohibited: in fact all through the Hindu caste system any intercourse, even to the extent of speaking to, touching, or appearing unveiled in the presence of, her husband’s Jeth, or elder brother, is strictly guarded by a special taboo. There is a Behâr proverb—Latul bhainsur dewar barâbar—“a weak elder brother-in-law is like a younger brother-in-law, with whom you may take liberties.”

Prevalence of widow marriage. 11. The statistics of the last Census fully illustrate the prevalence of widow marriage. To use Mr. Baillie’s summary of the figures68 “of 10,000 of the total Hindu population, 331 males and 817 females are widowed, 306 males and 747 females among Muhammadans, and no less than 639 males and 1,054 females among Jains.69 It is clear, therefore, that both males and females, but particularly the latter, re-marry more extensively amongst Muhammadans than Hindus, and very much more frequently [cxcii]among Hindus than amongst Jains. As regards females this is exactly what might have been expected from what is known of the social circumstances of the three religions. Muhammadans permit re-marriage alike amongst males and females, and the excess of female widowed is due to the same reasons as the excess in England. The higher proportion of widowed of both sexes as compared with England is, of course, mainly due to the higher proportion of marriages. The somewhat higher proportion of excess among Muhammadan widows over Muhammadan widowers, as compared with English figures, is probably due to the greater facilities an English widow enjoys for re-marriage. Amongst Hindus, as is well known, re-marriage is in the higher castes permitted only for males. The castes which do not permit widow marriage are roughly one-fourth of the whole,70 so that Hindus as regards female re-marriage occupy a position between Muhammadans and Jains, but nearer the former than the latter. The latter are practically, as regards such matters, Hindus of high caste, and permit no widow re-marriage: hence the high proportion of widows.” [cxciii]

12. This marriage of widows, known to the east of the Province as sagâi and to the west as karâo and dharewa, is a perfectly legal form of marriage, and when recognised by the tribal council the children are regarded as legitimate and succeed to their father’s estate. In subsequent pages will be found numerous details of the ritual in widow marriages. Among many of the lower castes the general rule appears to be that the widow is married to a widower: but this rule is subject to exceptions. The prohibited degrees for the widow are the same as for the virgin bride, with the additional limitation, as already explained, that she cannot marry her elder brother-in-law or her senior cousin. Though the marriage is quite legitimate, there is a certain amount of secrecy connected with it. It is performed at night. The bridegroom after eating with the woman’s friends invests her with a new robe and some jewelry, and withdraws with her to a private room. Next day he brings her home and procures the recognition of the union by feasting his clansmen. The rules as regards the custody of children by the first marriage are not very clearly defined. The usual course seems to be that if she has an infant she takes it with her to her new home, where it is practically adopted by its step-father. Children who have passed the stage of helplessness fall under the guardianship of their uncles, who manage their estate until they attain years of discretion, or, in the case of girls, arrange their marriages.

Age for marriage. 13. As regards the age for marriage the following table taken from the last Census Report71 deserves re-production. [cxciv]

[cxcv]

Thus 1,971 persons are shown as married in the first year of life. What is known as the petmanganiya or “womb betrothal,” that is the engagement of unborn children should they turn out to be of different sexes, is noted in the case of Kanjars. It is remarkable that the returns show that the proportion of children married below the age of 4 is as high among Muhammadans as Hindus. Mr. Baillie believes that the custom prevails mainly among Muhammadan sweepers; but this is not quite certain. Assuming 9 to be about the age of puberty, about 2½ per cent. of boys and 5 per cent. of girls enter the state of matrimony below that age. But it must be noted that this does not imply premature consummation: these infant marriages are probably nearly all in the families of persons of some wealth and social importance, and in such cases cohabitation is practically always postponed till puberty, when the gauna or bringing home of the bride takes place. Mr. Baillie goes on to remark:—“Between 10 and 14 nearly nine-tenths of the female population pass into the married state; but considerably more than one-half of the males remain unmarried. Between 15 and 19 there are 15 married females for each one unmarried, whilst at the end of the period only 60 per cent. of the males have been married. By 24 practically the whole of the female population have been married, almost the whole of those unmarried at this and later ages being women whose avocations preclude marriage, or whose physical or mental health forbids it. Of men considerably more than a fourth are unmarried up to 24, whilst an appreciable but diminishing number [cxcvi]remains unmarried through all subsequent age periods.”72

Bachelors and old maids. 14. The census figures show, as might have been expected, that “the largest proportion of males who remain permanently unmarried is among Jâts, Râjputs, Brâhmans, Kâyasths, Khatris, and to a less extent among Banyas. It shows that marriage is latest for men in these castes also, while it is earliest for the low-caste cultivators, forest and hill tribes, Julâhas, Kumhârs, Telis, Dhobis, fishing castes, Chamârs, Pâsis and vagrant castes, the highest figure of all being for Kumhârs. The figures for women are in certain respects both more pronounced and more important than for men. For women, the largest numbers permanently unmarried among respectable Hindus are amongst Râjputs and Khatris. The high proportion among the former may have to do with the claim made by many of the dancing castes to be [cxcvii]Râjputs. Why it should be so high among Khatris I have been unable to understand or imagine.73 Banjâras and vagrant Hindu castes show proportionately much higher numbers. Amongst the Muhammadans, the higher the caste, the higher the proportion of women not married at all. Female infant marriage is most extensive amongst cultivating castes, grazing castes, forest and hill tribes, Koris, Julâhas, Kumhârs, Telis, Dhobis, Chamârs, Pâsis, sweepers, and vagrant castes. Of the whole Pâsis are easily first, Kumhârs following a close second. Widows are most numerous among Brâhmans, Râjputs, Kâyasths, Banyas, Khatris and Sayyids easily, the highest proportion being among Khatris and Brâhmans. The lowest proportion of widows is among the forest and hill tribes, and after them amongst sweepers, Pâsis, Julâhas and Chamârs, in all of which castes woman is peculiarly a helpmate to man.”74 The prenubial laxity of Dravidian girls enables the men to avoid marriage till they are well advanced in life, and desire to found homes for their old age.

Polygamy. 15. Polygamy is permitted both among Hindus and Muhammadans. As Mr. Mayne remarks75:—“One text of Manu seems to indicate that there was a time when a second marriage [cxcviii]was only allowed to a man after the death of his former wife (V., 168; IX., 101, 102). Another set of texts lays down special grounds, which justify a husband in taking a second wife, and except for such causes it appears she could not be superseded without her consent (Manu, IX., 72–82). Other passages provide for a plurality of wives, even of different classes, without any restriction (Manu, III., 12; VIII., 204; IX., 85–87). A peculiar sanctity, however, seems to have been attributed to the first marriage.… It is now quite settled that a Hindu is absolutely without restriction as to the number of his wives, and may marry again without his wife’s consent, or any justification except his own wish.” There seems no doubt that a Muhammadan may marry as many as four wives: but the question is debated by the authorities.76 In spite of this polygamy is most infrequent. The last Census shows 11,820,598 married males to 11,873,838 married females. Similarly in the Panjâb there are 101·2 wives to 100 husbands. The proportion of husbands who have more than one wife is probably under 1 per cent.

Marriage by capture. 16. Something has already been said on the subject of marriage by capture. It may be well to consider if there are any facts which indicate that the people of Upper India in early times procured brides by force. Mr. McLennan, as we have seen, in his theory of marriage, starts with the stage of communal marriage next to polyandry, merging in the [cxcix]levirate. This stage attained, some tribes branched off into endogamy, some to exogamy. Exogamy was based on infanticide, and led to marriage by capture.77 We have already seen the weakness of the evidence for the existence of a general stage of polyandry or communal marriage.

17. In describing the various forms of marriage Manu speaks of that known as Râkshasa:—“The seizure of a maiden by force from her house, while she weeps and calls for assistance, after her kinsmen and friends have been slain in the battle, or wounded, and their houses broken open, is the marriage called Râkshasa”.78

18. The difficulty in examining the apparent survivals of marriage by capture lies in determining which are indications of the usual maiden modesty of the bride, her grief at leaving home and her dread at entering a new family, and which are signs of violence on the part of the bridegroom and his friends.

19. From the early literature, beyond the reference in Manu, to which reference has already been made, the traces of the custom in myth are not very numerous or clear. The myth of Urvasî probably indicates the existence of some ancient rule or taboo which prevented ordinary unrestrained intercourse between husband and wife, with the inference that possibly from capture their relations were strained.79 In the Mahâbhârata the followers of Kîchika attempted to burn Draupadî with [cc]his corpse, apparently because from the fact of her capture she was assumed to have been his wife. In the same epic Bhîshma declares that the Swayamvara is the best of all modes of marriage for a Kshatriya, except one, that of carrying away the bride by force. He acquired in this way the beautiful daughters of the Râja of Kâshi as wives for his brother Vichitra Vîrya. In the Sûtras it was provided that at a certain vital stage in the marriage ceremony a strong man and the bridegroom should forcibly draw the bride and make her sit down on a red ox skin.80

20. There are numerous examples of feigned resistance to the bridegroom. Thus among the Korwas the bridegroom and his party “halt at a short distance from the bride’s house, and there await her party. Presently emerges a troop of girls all singing, headed by the mother of the bride, bearing on her head a vessel of water surmounted by a lighted lamp. When they get near enough to the cavaliers they pelt them with balls of boiled rice, then coyly retreat, followed, of course, by the young men, but the girls make a stand at the door of the bride’s house and suffer none to enter until they have paid toll in presents to the bridesmaid.”81 In a Gond marriage “all may be agreed between the parties beforehand, nevertheless the bride must be abducted for the fun of the thing: but the bridegroom has only to overcome the opposition of the young lady’s female friends—it is not [cci]etiquette for the men of her village to take any notice of the affair.”82

21. Numerous instances of similar practices have been recorded at the present survey. Thus, among the Ghasiyas, the bride hides in a corner of the house, and the youth goes in and drags her out into the presence of the assembled clansmen. It is etiquette that she makes some resistance. Much the same custom prevails among the Bhuiyas and Bhuiyârs. The Kanjar bridegroom comes armed to the bride’s house after the negotiations have been settled, and demands delivery of the girl in threatening tones. Similarly the bridegroom is armed with a bow and arrow.

22. There are numerous other customs which seem to be based on the same form of symbolism. Thus, the members of the bridegroom’s party are mounted on horses and armed: they, on arriving at the bride’s village, do not enter her house, but halt outside; the bridegroom on reaching her door makes a feint of cutting at the arch (toran) with a sword: there is the invariable fiction, no matter how near the houses of the bride and bridegroom are, that she must be carried in some sort of equipage. This the Mânjhis and some other Dravidian tribes call “a boat,” or jahâz; possibly a survival of the time when the bride was taken away by water.

23. We have then the etiquette by which the bride screams and wails as she is being carried away. When she reaches her new home she is lifted across the threshold [ccii]by her husband, or carried inside in a basket. This was an old custom on the Scotch border,83 and may be as much a survival of the respect paid to the threshold as a reminiscence of marriage by capture. As she enters the door is barred by her husband’s sister, who will not allow her to enter until she is propitiated with a gift.

24. We have just noticed the fiction by which a bride is supposed to be brought from a distance. This is a standing rule among the Orâons and Kurmis of Bengal,84 and more than one example of it may be found in the present survey, as among the Nâis and Pankas. This repugnance to marriage among people residing in close communities has been taken by Dr. Westermarck to be one of the causes which have led to exogamy.85 In this connection, the system of gang exogamy, prevalent among the gypsy Kanjars and Sânsiyas, with whom it is a rule that the bride must be selected from an encampment different from that of the bridegroom, is most significant. It is possible that here we are very close to exogamy in its most primitive form.86

25. In the same category are the numerous taboos of intercourse between a man and his wife and her relations. We have already noticed the legend of Urvasî. The wife must not mention her husband by name, and if he addresses her, it is in the indirect form of mother [cciii]of his children. Mr. Frazer has directed attention to the rule by which silence is imposed on women for some time after marriage as a relic of the custom of marrying women of a different tongue. Hence the familiar incident of the Silent Bride which runs through the whole range of folklore.87 On the same lines is the taboo of intercourse between a man and his mother-in-law, of which Dr. Tylor, though he gives numerous instances, is unable to suggest an explanation.88 This, also, perhaps accounts for the use of the terms “brother-in-law” (sâla), “father-in-law” (sasur), as abusive epithets.

Runaway marriages. 26. The next form of marriage is the runaway marriage, which was dignified by the early Hindu lawgivers with the name of Gandharva, “the reciprocal connection of a youth and a damsel, with mutual desire, contracted for the purpose of amorous embraces, and proceeding from sensual inclination.”89 This prevails largely among the Dravidian tribes of the Central Indian plateau. At the periodical autumn feast the Ghasiya damsel has only to kick the youth, of whom she approves, on the ankle, and this is a signal to her relatives that the sooner the connection is legalised the better. We have the same custom in another form in the well known institution of the Bachelors’ Hall among the Orâons and Bhuiyas.90 This merges [cciv]into the Mutʼah marriage, which is legalised among Muhammadans.

Marriage by exchange. 27. Next comes marriage by exchange, known commonly as adala badala, where two fathers exchange daughters in marriage between their sons. This is the simplest form of marriage by purchase.91 The present survey has disclosed instances of this among Barhais, Bhuiyas, Dharkârs, Ghasiyas, Kanaujiyas, Meos, Musahars and Tarkihârs. It thus is in a great measure confined to the lower castes, and Mr. Ibbetson remarks92 that in the East of the Panjâb “exchange of betrothal is thought disgraceful, and, if desired, is effected by a triangular exchange,—A betrothing with B, B with C, and C with A: in the West, on the contrary, among all classes, in the Hills and Submontane Districts, apparently among all but the highest classes, and among the Jâts, almost everywhere, except in the Jumna District, the betrothal by exchange is the commonest form.”

Beena marriage. 28. The next stage is what has been called by ethnologists Beena marriage,93 in which the bridegroom goes to the house of the bride and wins her after a period of probation as Jacob wins Rachel. In these Provinces the custom seems to be confined to the Dravidian tribes of the [ccv]Vindhyan plateau, Bhuiyârs, Cheros, Ghasiyas, Gonds, Kharwârs, Majhwârs, and Parahiyas. Among them it bears the name of gharjanwai, which means “the son-in-law residing in the house of the bride.”

Bride purchase. 29. Immediately arising out of this is the more common form of bride purchase which prevails among most of the inferior tribes. In many cases, as will be seen by the examples which have been collected, the bride-price is fixed by tribal custom, and it marks a progressive stage in the evolution of marriage, where the purchase of the bride is veiled under the fiction of a contribution given by the relatives of the youth to cover the expenses of the marriage feast, which is, except in the dola or inferior form of marriage, provided by the relatives of the bride. “Let no father,” says Manu,94 “who knows the law, receive a gratuity, however small, for giving his daughter in marriage: since the man who, through avarice, takes a gratuity for that purpose, is a seller of his offspring.”

Marriage with dowry. 30. The last stage is when the relatives of the bride provide a dowry for the bride, which is the subject of careful negotiation, and is paid over in the presence of the tribesmen when the wife lives with her husband. [ccvi]

Confarreatio. 31. In all these forms of marriage the ceremony of Confarreatio, or the feeding of the married pair by the relatives on both sides, takes an important place. We have seen that it is the main rite in widow marriage. It is regulated by rigid rules of etiquette, one of the chief of which is that both bride and bridegroom must at first refuse the proffered food, and accept it only after much pressure and conciliation by gifts.

The Matriarchate. 32. According to Baudhayana “there is a dispute regarding five practices both in the South and in the North. Those peculiar to the South are to eat in the company of an uninitiated person, to eat in the company of one’s wife, to eat stale food, to marry the daughter of a maternal uncle or paternal aunt. He who follows these in any other country than the one where they prevail commits sin.”95 There is some want of moral perspective in the classification of these prohibitions: but they chiefly concern us in connection with the matriarchal theory. The prohibition of marriage with a cousin on the mother’s side has been accepted as an indication of the uncertainty of male parentage. There can be no doubt that in Northern India there is some special connection between a boy and his maternal uncle, as is shown by many instances drawn from the usages of the inferior tribes, such as the Agariya, Majhwâr and other Dravidian races. We also find among the Doms and Dharkârs that it is the [ccvii]sister’s son who performs the duties of priest at the cremation and worship of the sainted dead, which follows it. He is not, however, regarded as an heir to the deceased to the exclusion of his sons. Similarly though a foster-child has no rights to succeed,96 the relationship is universally recognised as a bar to intermarriage. There is thus some evidence for some of the tests of female kinship as laid down by Professor Robertson Smith.97 [ccviii]