Chapter IX. Marriage

Volume II: Part I pp. 427–541

Chapter IX. Marriage

This is the most important of all saṁskāraṣ Throughout the ages for which literary tradition is available in India marriage has been highly thought of. The several words that are employed to denote the idea of marriage indicate one or more of the elements of the saṁskāra of marriage. Such words are udvāha (taking the girl out of her parental home), vivāha (taking the girl away in a special way or for a special purpose i. e. for making her one's wife), pariṇaya or pariṇayana (going round i. e. making a pradakṣiṇā to fire), upāyama (to bring near and make one's own), and pāṇigrahaṇa (taking the hand of the girl). Though these words express only one component element of the rite of marriage they are all used in the śāstras¹ to indicate the totality of the several acts that go to make up the ceremony of marriage. The word 'vivāha' occurs in the Tai. S. VII. 2. 87 and Ait. Br. (27. 5). In the Tāṇḍya² Mahābrāhmaṇa VII. 10. 1 it is said that "heaven and earth were once together but they became separate"; then they said 'let us bring about a marriage, let there be a co-operation between us'."

Before dwelling upon the various aspects of marriage, the question, whether our authorities point to a state of society when there was no institution of marriage, but there was only promiscuity, requires to be considereḍ The Vedic works contain no indications about a society in which the relations of the sexes were promiscuous and unregulateḍ In the Mahā-bhārata,³ however, Pāṇḍu is made to state to his queen Kuntī that women in former ages were under no control, indulged themselves as they liked and that they left off one man and went after another, that this state of things continued to his ( Pāṇdu’s ) day in the country of Uttara Kuru, that it was Śvetaketu, son of Uddālaka, who for the first time stopped all this license and laid down the rule that if a woman proved false to her husband or if a husband was false to a chaste wife, very grave sin would be incurreḍ In the Sabhāparva ( 31. 37-38 ) it is said that through the favour of Agni women in Māhiṣmatī did what they liked and could not be restraineḍ These passages cannot be relied upon for proving promiscuity of intercourse. In the first place, the country of Uttara Kuru is more or less mythical. This passage rather gives expression to what the poet imagined about remotest ages and not what he knew was the real state of society thousands of years before hiṃ The theory of an original state of promiscuity once advanced by several sociologists has now ceased to be respectable ( vide Mrṣ M. Cole in ‘Marriage, past and present’ p. 10.

The purpose of marriage, even according to the Ṛgveda, was to enable a man, by becoming a householder, to perform sacrifices to the gods and to procreate sonṣ The verse in Ṛg. X. 85. 36 shows that the husband took a woman as a wife for ‘gārhapatya’. Ṛg. V. 3. 2., V. 28. 3 speak of the co-operation of husband and wife in the worship of godṣ Ṛg. III. 53. 4 contains the emphatic assertion ‘the wife herself is the home’ ( jāyed-astam ). In later literature also the same statement occurṣ A wife was called ‘jāyā’, because the husband was born in the wife as a son ( Ait. Br. 33. 1 ). The Śat. Br. V. 2. 1. 10 says ‘the wife⁴ is indeed half of one’s self; therefore as long as a man does not secure a wife so long he does not beget a son and so he is till then not complete ( or whole ); but when he secures a wife he gets progeny and then he becomes complete’. The Ait. Ār. ( I. 2. 4 ) says ‘therefore a man, after securing a wife, regards himself as more complete’. When Āp. Dh. S. II. 5. 11. 12 forbids taking a second wife if the first is endowed with progeny and the performance of religious rites, it indicates that the main purposes of marriage are two,⁵ viz. the wife enables a man to perform religious rites and is the mother of a son or sons who were supposed to save a man from hell. Manu (IX. 28) states that on the wife depend the procreation of sons, the performance of religious rites, service, highest pleasure, heaven for oneself and for one's ancestorṣ So these three viz. dharmasampatti, prajā (and consequent freedom from falling into hell) and rati (sexual and other pleasures) are the principal purposes of marriage according to the smṛtis and nibandhas. Yāj. I. 78 is to the same effect. Jaimini (VI. 1. 17 ff) establishes that husband and wife have to perform sacrifices together and not separately and Āp. Dh. S.⁶ II. 6. 13. 16-17 emphatically says that there can be no separation between husband and wife, for since marriage they have to perform religious acts jointly.

Marriage is a composite rite comprising several subordinate elements which have to be done in a certain order and the last of which is seeing the constellation of the seven sages; it (rite) brings about the status of a woman as a person's wife.⁷

The first consideration is: how to choose a bridegroom and what qualifications make a person a very desirable bridegrooṃ Āśv. gr. (I. 5. 2) says 'one should give a maiden⁸ (in marriage) to a man endowed with intelligence'. The Āp. gr. (3. 20) remarks “the accomplishments of a bridegroom are that he must be endowed with good family, a good character, auspicious characteristics, learning and good health”. Bauḍ Dh. S. IV. 1. 12 states 'a maiden should be given in marriage to one who is endowed with good qualities and who is a celibate (till then)'. Even the Śakuntala (IV) echoes the words of Baudhāyana 'the best idea is to give one's daughter to a man endowed with good qualities'. Yama quoted in the Sṃ C. (I. p. 78) says 'one should seek for seven qualities in a bridegroom viz. good family, good character, bodily appearance, fame, learning, wealth and support ( of relatives and friends ); the other matters need not be considered'. Bṛhat-parāśara ( p. 118 Jiv. eḍ ) enumerates eight qualities in a bridegroom viz. caste, learning, youth, strength, health, support of many ( friends &c. ), ambitions ( arthitva ), and possession of wealth. Aśv. gr. (I. 5. 1) and others place kula ( a good family ) in the forefront in the case of both the bride and the bridegrooṃ ⁹ 'One should first examine the family ( of the intended bride or bridegroom ) as it has been said above ( in the Aśv. Śrauta sūtra IX. 3 ).' The Aśv. Śrauta reads 'those who on their mother's as well as their father's side through ten generations are endowed with learning, austerity and meritorious works or whose pure brāhmaṇa lineage can be traced on both sides ( for ten generations ) or according to some on the father's side.' Manu regarded a good family as the most important from the eugenic point of view. In IV. 244 he says 'one who desires to raise his family to excellence and eminence should always enter into marriage alliances with the best and avoid alliance with the low'; and adds (III. 6-7) that ten kinds of families even though richly endowed with cattle, wealth &c. should be avoided in marriage viz. in which the saṁskāras are not performed, in which there is no male progeny, which are devoid of Veda (study), which have hairy members, that suffer from piles or consumption or indigestion or epilepsy, white or black leprosy. Manu ( III. 63-65 ) explains under what circumstances good families are reduced to a bad state. Hārīta states that the offspring is in accordance with the (qualities of the) family of the parentṣ The Harṣacarita ( IV ) gives expression to the view that generally the wise look to good family first even though there may be other qualities in the bridegrooṃ Manu II. 238, however, allows a man to marry a girl even from an inferior or bad family provided she is a jewel among womeṇ Yāj. (I. 54-55) emphasizes the importance of a good family famed for ten generations and freedom from diseases that are hereditarily transmitted and adds that the bride-groom must be young, intelligent, a favourite among the people and his virility must be carefully testeḍ The Mit. on Yāj. 1. 55 quotes Nārada ( stripumsa ) who mentions some characteristics of the virile man and enumerates fourteen varieties of impotent persons ( verses 11-13 ). Kātyāyana¹⁰ mentions what defective bridegrooms are to be avoided viz. the lunatic, one guilty of grave sins, leprous, impotent, one of the same gotra, one bereft of eyesight or hearing, an epileptic and adds that these defects are to be avoided even among brideṣ The Mahābhārata observes ‘friendship and marriage should take place between those alone whose wealth is similar and whose learning ( i. e. the learning in whose families ) is of equal status, not between opulent ( well-off ) and the poverty-stricken’.¹¹

Though Manu and Yāj. indicate that impotent persons are not eligible for marriage, yet such persons rarely marrieḍ Their marriages were held valid by Manu, Yāj. and others and their sons ( by niyoga ) took property as if they were aurasa sonṣ Vide Manu IX. 203 and Yāj. II. 141-142.

The Saṁskāra-prakāśa ( pp. 752-754 ) contains a long discussion on the question how from bodily characteristics one can find whether a man will live long or what his prospects would be. These are passed over for want of space.

Rules for the selection of the bride are far more elaborate than those for selecting a bridegroom, though in some respects they are the same ( such as about the necessity of good family, about the absence of diseases &c. ). Vide Vas I. 38, Viṣṇu Dh. S. 24. 11, Kāmasūtra III. 1. 2.¹² Even the Sāt. Br. ( I. 2. 5. 16 ) gives expression to the then view that broad hips and slender waists make women attractive. Āśv. gr. I. 5. 3 says ¹³ one should marry a girl who is endowed with intelligence, beauty, a good character and auspicious characteristics and who is healthy'. Śāṇ gr. I. 5. 6 and Manu III. 4 and Yāj. I. 52 also require that the girl should be possessed of auspicious characteristics (or indications). These (lakṣaṇas) are of two kinds, bāhya (visible or bodily characteristics) and ābhyantara (invisible). Āp. gr. ¹⁴ (III. 21) states a commonsense rule: 'a girl on whom his mind and eyes are riveted will bring him happiness (or prosperity), he should pay no heed to other things; this is the view of some'. The Kāmasūtra quotes the view of Ghoṭakamukha 'he should proceed to marry a girl on taking whom as his wife he would regard himself as blessed and would not be blamed by his friends (or persons in a similar station in life )' ¹⁵. Manu III. 8 and 10, Viṣṇu Dh. S. 24. 12–16 say that one should not marry a girl having tawny hair or having an excessive limb (such as a sixth finger) or a deficient limb, who is hairless or very hairy, who is talkative and has yellowish eyes; but should marry a girl who has limbs void of any defect, whose gait is like that of a swan or an elephant, the hair on whose head or body is of slight growth and whose teeth are small, whose body is delicate. The Viṣṇu-purāṇa (III. 10. 18–22) adds that the girl must not have a marked growth of hair on her chin or lip, her voice must not be hoarse or like that of a crow; her legs and ankles must not be very hairy, there should be no dimples on her cheeks when she laughs, she should not be very dwarfish or very tall &c. Manu III. 9 and Āp. gr. (III. 13) say that the girl to be married must not bear names of the lunar mansions (such as Revatī, Ārdrā &c.), trees or rivers, she must not bear a mleccha name or that of a mountain, of a bird, of a snake or of a slave or a name that is terrific. Āp. gr. (III. 14) and Kāmasūtra III. 1. 13 mention that a girl should not be chosen, the penultimate letter of whose name is r or l (such as Gaurī, Śālī, Kamalū). Nārada ¹⁶ (stri- purṇsayoga 36) says that defects of girls are as follows:—when they suffer from long-standing or disgusting diseases, when they are devoid of a limb or have already had connection with an- other man, when they are wicked or have their minds fixed on another; and Āp. gr. (III. 11-12) states other defects of girls viz. one should not choose a girl who is asleep or weeps or has left the house when persons come to see whether she can be choseṇ Vide Mārkaṇḍeya-purāṇa 34. 76-77 for guṇas and defects of girls as ‘brides’.

Bhār. gr. I.11 says that there are four inducing reasons ¹⁷ for marrying a particular girl viz. wealth, beauty, intelligence and family. If all four cannot be secured, wealth may be neglected (as the least important of all); then beauty may be neglected if there is intelligence and good family, but there is a difference of view as to the latter two, some preferring intelligence to family and others family to intelligence. Mānava gr. (I. 7. 6-7) adds a fifth inducement for marriage, viz. vidyā (learning) after beauty and before prajñā. Vide also Varāha gr. 10.

Some of the gṛhyasūtras propose a peculiar mystical method of selecting a bride. The Āśv. gr. (I. 5. 3) after stating that one should select a girl endowed with good characteristics (lakṣaṇas) proceeds ‘lakṣaṇas are very difficult to discern’ and therefore prescribes (I. 5. 5-6) that eight lumps of different kinds of earth should be taken respectively from a field that yields two crops a year, from a cow-stable, from a vedi (sacrifi- cial altar, after sacrifice is performed), from a pool of water that does not dry up, from a gambling place, from a place where four roads meet, from a barren spot, and from a burial ground; then he should recite over the lumps the formula ‘ṛta (right) has been born first in the beginning; truth is founded (or fixed) in ṛta; may this girl attain here that for which she is born; may what is true be visible;’ then he says to the girl 'take one of these'. According as she chooses the lump, it may respectively be deemed that she will have offspring rich in food ( if she chooses the lump of the earth taken from the field of two crops a year &c. ), or rich in cattle, or rich in spiritual lustre or rich in everything, or addicted to gambling, or wandering in different directions or poor, or that she will bring death to her husband ( if she takes the lump of the earth from the burial ground ). The Gobhila gr. II. 1. 4-9 speaks of these lumps and adds that a ninth lump may be formed by mixing up the earths of all eight varieties and that if she takes up any one of the four lumps of earth from an altar, furrow, a pool or a cow-stable or ( according to some ) the ninth lump, she may be selecteḍ The Laugākṣi gr. 14. 4-7 contains the same rules as in Gobhila. Āp. gr.¹⁸ (III. 15–18) prescribes a somewhat different methoḍ If both sides agree, the bridegroom (or his friends) should place in one lump of earth several kinds of seeds (such as rice,barley &c.); he should take (a lump having in it) the dust from an altar, a third having a clod from a ploughed field, a lump having cowdung inside and (a fifth having) a clod of earth from a cemetery and keep- ing them before the girl ask her to touch one of them ( the five ). If she touches any one of the first four, that is an indication of future prosperity ( of the nature of the object touched ), but the last ( viz. clod of earth from cemetery ) is objectionable. The Vārāha gr. 10 and Bhār. gr. I. 11 speak of only four lumps of earth viz. from a field, from an altar, from a cow-stable, and from a cemetery and say that one should not marry a girl who takes up the lump of cemetery earth. Mānava gr. ( I. 7. 9-10 ) speaks of eight lumps but substitutes a lump of earth where dūrvā grows and a lump from under a tree filled with fruit for earth from a pool of water and one from a gambling place and adds that the eight lumps should be placed in a temple and if the girl takes up the lump from a cemetery or from a barren spot or from where four roads meet she should not be marrieḍ Many digests like the Gr. R. ( pp. 13-22 ) contain long quotations which dilate upon the indications about the auspiciousness or otherwise of girls from their several physical featureṣ

Gaut. IV. 1, Vaṣ 8. 1, Mānava gr. I. 7. 8, Yāj. I. 52 and several others say that the girl must be younger ( yavīyasī ) than the bridegroom and the Kāmasūtra (III. 1. 2) recommends that she must be younger than the bridegroom by at least three yearṣ The Mit. on Yāj. I. 52 explains 'yavīyasī' as meaning 'younger in age and smaller in stature' (than the bridegroom). What the age of marriage was will be discussed a little below. Gaut. IV. 1, ¹⁹ Vaṣ 8. 1, Yāj. I. 52, Manu (III. 4 and 12) and others say that one should marry only a girl who is a virgin and of the same caste. How far widow-marriages and intercaste marriages were allowed would be discussed later oṇ The Mānava gr. I.7.8, Manu III. 11 and Yāj. I. 53 require that the girl to be chosen must not be brotherlesṣ This requirement which has been not in force for centuries has a long history behind it. In Rg. I. 124. 7 it is said ²⁰ 'as a brotherless maiden comes back towards her male relations (her father's family)...so the dawn reveals objects (or her beauty).' In the Atharvaveda I. 17. 1 we read 'like brotherless women let them sit still with their splendour gone'. Both those passages are quoted and explained in the Nirukta III. 4–5. In ancient times when a man had no son, he could make his daughter do for a son (i. e. she herself became putrikā) and stipulate with the person marrying her that the son born of her would be his (i. e. the girl's father's) son and would offer piṇḍas as a son to his maternal grandfather. The result would be that the son of such a girl would not offer piṇḍas to his father and would not continue the line of his father. Rg. III. 31. 1 (a very obscure and difficult verse) has been explained by the Nirukta (III. 4) as referring to the practice of declaring a daughter to be one's son ²¹. Therefore, brotherless maidens were not chosen as brides and the Rg. speaks of spinsters growing old in their parental home (Rg. II. 17. 7). Vaṣ Dh. S. 17. 16 refers ²² to Rg. I. 124. 7 and quotes a verse about a brotherless girl. Such a daughter at whose marriage the father made the stipulation stated above was called putrikā and according to Gaut. 28. 17 some ²³ teachers went so far as to say that a daughter became a putrikā by the mental resolve of the father alone (without there being an express agreement with the bridegroom). It is therefore that Manu (3. 11) ordains that one should not marry a girl that has no brother, as there is the danger of her being a putrikā. The Nirukta (III. 5) quotes another ²⁴ Vedic passage 'one should not marry a brotherless woman, for she becomes the son of him (her father)' and remarks that in this passage there is a direct prohibition against marrying a brotherless maiden (while in the passage of the Atharvaveda it is indirect, being involved in a simile) and it is also expressly said that she becomes the son of her father. In medieval times this prohibi- tion against marrying a brotherless girl gradually was ignored and in modern times the pendulum has swung the other way, a brotherless girl being a coveted prize if her father be rich. In course of time popular feeling changed and no girl could remain unmarried if she wanted heavenly worldṣ There is an interesting story in the Śalyaparva chap. 52 of the daughter of Kunigarġa who, when told by Nārada that an unmarried woman could not secure heaven, married for one day Śṛngavat and then went to heaveṇ

There are further restrictions about choosing a girl for marriage. The rule was that a man should marry a girl of the same caste. So far there is what is called endogamy (i. e. rule requiring marriage within a certain large community). But within this large community there were certain groups which were prohibited for marriage to a person belonging to another group of the same caste or community (i. e. the principle of exogamy. operated within the large community itself). The Hir. gr. I. 19. 2, Gobhila gr. III. 4. 4, Āp. Dh. S. II. 5. 11. 15 require that the bride to be chosen must not belong to the same gotra as that of the bridegrooṃ They are all silent about the sameness of pravara. Gaut. IV. 2., Vaṣ Dh. S. VIII, 1, Mānava gr. I. 7. 8, Vārāha gr. 9, Śankha¹⁰³⁶ Dh. S. forbid marriage with a girl whose pravara (or ārṣeya) is the same as that of the bridegroom; but they say nothing about the prohibition against the sameness of gotra. It is somewhat remarkable that some of the gṛhyasūtras like Aśv. and Pār. say not a word about sameness of gotra and pravara. Viṣṇu Dh. S. 24. 9, Vaik. III. 2, Yāj. I. 53, Nārada (stripuṁsa, verse 7), Vedavyāsa II. 2 and many others prohibit sameness of gotra as well as of pravara. Gobhila gr. III. 4. 5, Manu III. 5, Vaik. III. 2 and Ap. Dh. S. II. 5. 11. 16 require that the bride must not be a sapiṇḍa or blood relation of the mother of the bridegroom; while Gaut. IV. 2, Vaṣ VIII. 2, Viṣṇu Dh. S. 24. 10, Vārāha gr. 9, Śaṅkha Dh. (quoted above), Yāj. I. 53 and others restrict the prohibi- tion against marrying a sapiṇḍa girl to seven degrees on the father’s side and five degrees on the mother’s side. There were others like the Vedavyāsa-smṛti which not only prohibited marriage with a girl who had the same gotra as the bride- groom’s, but prohibited marriage with a girl whose mother’s gotra was the same as the bridegroom’ṣ

All these prohibitions against marrying a sagotra, sapravara or sapiṇḍa girl are extremely important, as the following considerations will show. It is a canon of the Pūrvamīmāṁsā that if there is a seen (dṛṣṭa) or easily perceptible reason for a rule stated in the sacred texts, it is only recommendatory and a breach of such a rule does not nullify the principal act. But if there is an unseen (adṛṣṭa) reason for a rule and there is a breach of such rule, the principal act itself is rendered invalid and nugatory thereby. The rule about not marrying a woman who is diseased or who has superfluous or deficient limbs has a seen reason viz. marriage with such a girl causes unhappiness (if she is diseased) or comment (if she has defi- cient limbs). Therefore, if a person marries such a girl the marriage is perfectly valiḍ But there is no seen or easily perceptible reason for the prohibition against marrying a sagotra or sapravara girl. Therefore, such rules go to the root of the matter and are obligatory and, if there is a breach of them, the marriage is no marriage, it is null and voiḍ So even if a person goes through a ceremony of marriage with a girl who is a sagotra or sapravara or sapinda (within prohibited degrees) she does not in law become his wife at all. These principles are very lucidly set ²⁶ forth by Medhātithi on Manu III. 11, by the Mit. on Yāj. I. 53, the Madanapārijāta and other workṣ Therefore, these questions of sagotra, sapravara and sapinda will be dealt with at some length later oṇ

The question of the age of marriage for men and women will now be considereḍ This is a very interesting and instructive study. The age of marriage for both sexes has varied considerably from age to age, from province to province and also from caste to caste even at the same time. ²⁷ As regards men there is no special rule as to the age before which a man was obliged to marry. A man could remain celibate all his life, while at least in medieval and modern times marriage has been absolutely necessary for every girl. A man was to marry after he had finished his Vedic studies; but the period of Vedic study was fluctuating (i. e. it could be 12, 24, 36, 48 years or as much time as was necessary to master one Veda or a portion of it). Usually twelve years were devoted to brahmacarya in ancient times and as upanayana ordinarily took place in the 8th year (for brāhmaṇas) a man would ordinarily be 20 years old or more at the time of marriage. It is therefore that Manu (IX. 94) remarks that a man of thirty may marry a girl of twelve years or a man of 24 who is in a hurry to become a householder may marry a girl of eight. Basing on this the Viṣṇu purāṇa ²⁸ (III. 10. 16) says that the ages of the bride and bridegroom should be in the ratio of 1 to 3. Aṅgiras says that the bride should be two, three, five or more years younger. In the Mahābhārata ²⁹ Gautama is prepared to give his daughter in marriage to Uttanka if he could be a youth of 16 yearṣ In another place the Mahābhārata (Sabha 64. 14 and Vanaparva 5. 15) uses the simile of a maiden not liking a husband of 60 yearṣ That shows that girls were sometimes married to even old men of 60 in those days as rarely in modern times also. In the Mahābhārata (Anusāsana 44. 14) the respective ages of the bridegroom and bride are recommended as 30 and 10 or 21 and 7; while the Udvähatattva (p. 123) and S. Pr. p. 766 quote a verso from the Mahābhārata that a man of 30 should marry a girl of 16 (but from the metre and context it appears that ‘șodașa-varșām’ as printed should be ‘daśa-varșām’)'.

In the Ṛgveda there are no clear statements about the exact age when girls were marrieḍ But there are indications that many girls were married at a sufficiently mature age (at least they were not married at the tender age of eight). We have already seen (p. 435) that brotherless girls often remained unmarried till old age. One of the benevolent deeds³⁰ of the Āśvins is that they bestowed a husband on Ghoṣā who was growing old in her father's house. Vide also Ṛg. I. 124. 7, II. 17. 7, and Atharva-veda I. 17. 1. In Ṛg. X. 27. 12 it is said ‘when a bride is fine looking and well adorned, she by herself seeks her friend from among men’. That shows that girls were grown-up enough to select their husbandṣ Some of the verses in the marriage hymn (such as Ṛg. X. 85. 26-27, 46) indicate ³¹ that married girls could not have been child-wives, but must have been grown-up. On the other hand in Ṛg. I. 116. 1 it is said that the Nāșatyas (Āśvins) bestowed a wife on Vimada who was an arbhaga (of tender age).³² But all that is meant seems to be that Vimada was a mere boy as compared with his rival kings who had come as suitorṣ The two verses Ṛg. I. 126. 6-7 (which are rather too naive) indicate that girls were married before they had attained puberty. In Ṛg. I. 51. 13 Indra is said to have given to old Kukşivat a wife named Vṛcayā who is styled ‘arbhā’ (young). But that word is only used in contradistinction to the word ‘mahate’ (grown old) applied to Kakşivat and conveys no idea of her exact age.³³ On the whole one may conclude that in the period of the Ṛgveda girls were probably married at any age (either before puberty or after puberty ) and sometimes remained spinsters all their life. The other Saṁhitās and the Brāhmaṇas do not shed much light on the question of the age of marriage for girlṣ In the Chāṇ Up. Uśasti Cākrāyaṇa is described as dwelling in the Kuru country with his wife, who is said to have been ‘āṭikl’ which Śamkara explains as ‘a mere undeveloped girl’.³⁴

Coming to the ancient gṛhya and dharma sūtras, it will be seen from several considerations that girls were married just before the time of puberty or immediately after it. Among the several requirements of the bride to be chosen several gṛhya sūtras lay down that she must be a ‘nagnikā’. Vide Hir. gṛ. I. 19. 2, Gobhila gṛ. III. 4. 6, Mānava gṛ. I. 7. 8, Vaik. VI. 12.³⁵ The word nagnikā is variously explained by the commentatorṣ Mātṛdatta on Hir. gṛ. explains that nagnikā means ‘one whose menstrual period is near’ i. e. one who is fit for intercourse. Aṣṭāvakra, the commentator of the Mānava gṛ., explains nagnikā as ‘one who has not yet experienced the impulses and emotions of youth’ or ‘one who looks pretty even without clothes’ (taking the word along with ‘śreṣṭhām’ that follows). The Gṛhyāsaṁgraha ³⁶ (which is much later than the Gobhila gṛ.) explains ‘nagnikā’ as one who has not yet reached puberty. These varying explanations of the word ‘nagnikā’ are due to the fact that when some smṛtis and commentators wrote child marriages were not in vogue, while they were in vogue when and where others flourisheḍ Vaṣ Dh. S. 17. 70 says ‘the father should give away his daughter while she is still nagnikā through fear of her attaining puberty ( while unmarried ).' This shows that according to Vaṣ nagnikā must have been a girl of tender years or one that had not attained puberty.

There is another very important indicatioṇ Most of the grhyasūtras prescribe that the married couple should be celibate for at least three nights after marriage ( if not for a longer period ). For example, the Pār. gr. ( I. 8 ) says³⁷ ' the married pair should (after marriage) not partake of kṣāra and lavaṇa for three nights, should sleep on the ground (and not on a cot, for the same period) and should not have intercourse for a year, twelve nights, six nights, three nights in the last resort '. Vide Āśv. gr. I. 8. 10, Āp. gr. 8. 3-9, Śāṅ. gr. I. 17. 5, Mānava gr. I. 14. 14, Kāṭhaka gr. 30. 1, Khādira gr. I. 4. 9 and otherṣ Such an injunction against intercourse for three nights after marriage would have been uncalled for and extremely inappropriate if girls had been usually married when they were only eight to ten years olḍ Such injunctions by so many authors presuppose that girls must have generally been very near puberty or past puberty at the time of marriage. Haradatta who belongs to about the 12th century A. D. expressly says³⁸ that ( in his days ) in certain countries intercourse was commenced immediately after marriage and that such a usage being opposed to Āśv. gr. I. 8. 10 should not be followed ( but continence should be observed at least for three days after marriage ). This shows that so late as 1200 A. D. in several countries the marriageable age of girls must have been at least about 14.

In most of the grhyasūtras there is a rite called ‘caturthī- karma’ (rite on 4th day after marriage). Vide Gobhila II, 5, Śāṅ. gr. I. 18-19, Khādira gr. I. 4. 12-16, Pār. gr. I. 11, Āp. gr. 8. 10-11, Hir. gr. I. 23-24 &c. This rite has been described above (pp. 203-204) and corresponds to the garbhādhāna of later writers (e. g. Yāj. I. 11). As cohabitation is expressly mentioned in connection with this rite performed on the fourth day after marriage it follows that girls must have generally been quite grown-up at the time of marriage.

In some of the gṛhyasūtras and smṛtis a slight prāyaścitta (purificatory ceremony) is prescribed if the bride has menstruation during the progress of the marriage ceremonieṣ Vide Bauḍ gr. IV. 1. 10, Kauśika sūtra 79. 16, Vaik. VI. 13, Atri (eḍ by Jiv. part I. p. 11).

Gaut.³⁹ (18. 20–23) says ‘a girl should be given in marriage before she attains the age of puberty. He who neglects it commits siṇ Some declare that she should be given in marriage before she begins to wear clotheṣ A marriageable girl (who is not given in marriage by her father &c.) should allow three monthly periods to pass and after- wards unite herself to a blameless man of her own will and give up the ornaments received from her father (or her family)'. This passage establishes that even before Gautama (i. e. 500 or 600 B. C.) there were people who advocated marriages of infant girls that did not care to put on clothes (i. e. who had no sense of shame if they went about without clothes), that Gautama disapproved of this view, that he only prescribed that the marriage of girls should take place before puberty and he was not horrified if girls got themselves married sometime after puberty. But it is clear that no blame attached to the girl married after puberty or to the husband; the idea, however, had arisen even then that the father or guardian incurred blame or sin by not getting a girl married before puberty. Manu (IX. 89–90) goes so far as to say ‘A maiden may rather stay in her father’s house even till her death, though she may have attained puberty; but the father should never give her to one who is devoid of good qualitieṣ A maiden after attaining puberty may wait for three years (to see if she is given away by her father or brother &c.) but after this period she should seek a husband who is similar to her.' Anuśāsanaparva 44. 16 is to the same effect. Bauḍ Dh. S. IV. 1. 14 and Vaṣ V. Dh. S. 17. 67–68 give the same rule as in Manu IX. 90. But both add (Vaṣ 17. 70–71 and Bauḍ Dh. S. IV.1.12) that the father or guardian incurs the sin of destroying an embryo at each appearance of menses as long as the girl is unmarrieḍ Yāj. I. 64 and Nārada (strīpumsa, verses 25-27) state the same rule.

On account of this change in popular beliefs and senti- ments it came to be recommended that a girl must be married early and may be given even to one ⁴⁰ who is devoid of qualities (in spite of Manu IX 89). Vide Bauḍ Dh. S. IV. 1. 12 and 15.

So we see that from about 600 B. C. to about the beginning of the Christian Era it did not matter at all if a girl was married a few months or few years after puberty. But by about 200 A. D. (when the Yāj. smṛti was composed) popular feeling had become insistent on pre-puberty marriageṣ The reasons for this change are not quite clear. But it is possible to hold that it was due to the following circumstanceṣ Buddhism had spread far and wide during these centuries with its encouragement of the institution of monks and nunṣ There was laxity of morals among nunṣ A further reason may be adduced viz. that girls had generally ceased to study anything, though some of them certainly did study in the times of Pāṇini and Patañjali, and so society did not like girls to remain doing nothing. From the times of the Ṛgveda (X. 85. 40-41) there was a mystical belief that Soma, Gandharva and Agni were the divine guardians of a girl and the Gṛhyāsaṁgraha (quoted in the coṃ on Gobhila gr. III. 4. 6) says that Soma enjoys a girl first, then Gandharva enjoys her when her breasts develop and fire when she menstruateṣ Therefore, a feeling arose that a girl must be married even before she develops any distinct signs of femininity. Saṁvarta ⁴¹ (verses 64 and 67) gives expression to this idea. Further, since marriage came to be looked upon as the upanayana in the case of women, naturally the age for upanayana (8th year) came to be looked upon as the proper age for marriage. Vide Saṁskārakaustubha quoted below. It came to be believed that there was no hope of heaven for a woman who died unmarrieḍ In the Śalyaparva⁴² chap. 52 we have the story of a girl, daughter of Kuni Garga, who practised severe penance till she reached old age and yet was told by Nārada that if she died unmarried she would not go to heaveṇ The woman induced Śṛngavat of the Gālava family for a day previous to her death to marry her by the promise of giving him half of her merit (puṇya). The Vaik.⁴³ (V. 9), while describing the ceremony of funeral rites in cases of distress, mentions the curious practice of find-ing out a male of the same caste for a girl, who dies unmarried though of the age of puberty, with whom a sort of marriage is gone through and the girl is then cremateḍ Whatever the reasons may be, this tendency to bring down the age of marriage for girls was accentuated in the first five or six centuries of the Christian era. The Laugākṣi⁴⁴ gr. (19. 2) says 'brahmacarya for girls lasts till the 10th or 12th year'. The Vaik. (VI. 12) says that a brāhmaṇa should marry a brāhmaṇa girl who is a nagnikā or gaurī and defines nagnikā as a girl over eight years but less than ten and gaurī as one who is between ten and twelve and has not had menstruatioṇ Aparārka (p. 85) quotes the Bhaviṣyapurāṇa to the effect that nagnikā is one who is ten years olḍ Vide Sṃ C. quoting Samgraha. Parāśara, Yāj. and Saṁvarta go even beyond thiṣ Parāśara⁴⁵ (VII. 6-9) says 'a girl of eight is called gaurī, but one who is nine years old is a rohiṇī; one who is ten years old is a kanyā; beyond this (i. e. after 10 years) she is a rajasvalā. If a person does not give away a maiden when she has reached her 12th year, his pitṛs have to drink every month her menstrual discharge. The parents and also the eldest brother go to hell on seeing (an unmarried) girl reaching the state of a rajasvalā. Parāśara adds that the brāhmaṇa who marries such a girl should not be spoken to or admitted to dinner in the same row (as other brāhmaṇas) and that he becomes the husband of a vṛṣalī. On the other hand the Vāyupurāṇa (83. 44) extols the marriage of a gaurī by remarking that her son purifies 21 ancestors on his father's side and 6 male ancestors on his mother's side. Saṁvarta (verse 65-66) has the same two verses as Parāśara (VII. 6 and 8) and winds up (verse 67) by saying that the marriage of a girl of eight is highly commendeḍ The same four verses (Parāśara VII. 6-9) occur in Bṛhad-Yama (Ānaṇ eḍ) chap. 3. 19-22, but the order is different. Aṅgiras (vv. 126-128, Ānaṇ eḍ) has the same verseṣ Kāśyapa as quoted in the Gṛ. R. (p. 46) says that a girl is called gaurī when she is seven, a kanyakā when she is ten and kumārī when she is twelve. It will be noticed that the smṛtis of Vaikhānasa and Kāśyapa differed from Parāśara as to the definition of gaurī. Further the three smṛtis make marriage of a girl after puberty an extremely sinful act and condemn not only the parents but also the husband, while Bauḍ prescribed only a slight prāyaścitta for the father alone even if there was menstruation during the marriage riteṣ Marīci⁴⁶ (quoted in Par. M. I. 2. p. 177) said that choosing a bride who was five years old was best. Even Manu (9. 88) recommended an early marriage if a very desirable bridegroom was available. Rāma and Sītā are said to have been respectively 13 and 6 at the time of marriage (Araṇyakāṇḍa 47. 10-11). But this passage appears to be an interpolatioṇ In the Bālakāṇḍa⁴⁷ it is expressly stated that Sītā and her sisters enjoyed in private dalliance with their respective husbands immediately after marriage. If this is true then Sītā could not have been only 6 at the time of marriage.

The rule that brāhmaṇa girls were to be married between 8 to 10 years became general from about the 6th or 7th century and continued down to modern timeṣ During the last two or three decades the marriageable age of girls owing to several causes, particularly the ravages of plague and the economic condition of the middle classes, has risen very high and marriages of girls even of poor brāhmaṇas hardly ever take place before 16 and have to be postponed to the age of 20 and beyonḍ Besides the Child Marriage Restraint Act (XIX of 1929, as amended by Act 19 of 1938) has fixed the minimum age limit for girls at 14 and persons getting their daughters married before the completion of 14 years are liable to be punished in a criminal court under the Act. There is no reason to laugh at ancient Indian writers for the low age of marriage they advocateḍ Child marriages were common in all countries of Europe. Even in England the age limits for boys and girls were recognised by the law only at 14 and 12 respectively until 1929, when the lowest marriageable age for both was fixed at 16 (19 and 20 Geo. V. Chap. 36). Besides it must be remembered that when marriages of girls of tender years took place they were purely a sacrament. There was no question of consummation which took place only after puberty. Sages condemned intercourse even with one's wife before she attained puberty.⁴⁸ Winternitz in ‘Die Frau.’ p. 135 refers to a work of F. J. Furnivall on ‘Child Marriages, Divorces &c.’ between 1560-66 A. D. in England from which it is clear that marriages of children of 9 or 10 (and rarely of even 2 or 3 years) took place in England only about 300 years ago.

These rules about the proper age for the marriage of girls affected only brāhmanaṣ Sanskrit poets and dramatists always depict that the heroines in their works are grown-up at the time of marriage and even Bhavabhūti imagines that Mālatī, the heroine of his romantic drama of Mālatīmādhava, was so grown-up that it was practically a case of love at first sight. Vaik. (VI. 12) requires a brāhmaṇa bride to be a nāgnikā or gauri, but does not prescribe the same qualification for a kṣatriya or vaiśya bride. In the Harṣacarita princess Rājayāśrī is described as quite grown-up at the time of her marriage which was consummated on the very day of marriage (vide the last para. of the 4th Ucchvāsa ). The Saṁskāra-prakāśa⁴⁹ expressly says that there is no prohibition against marrying a girl who has passed the age of puberty for kṣatriyas and otherṣ Even in Pauranic times grown-up unmarried girls must not have been unheard of. The Gr. R. p. 83 quotes a passage from the Brahmapurāṇa where the injunction to observe complete celi- bacy for one year, 12 days, 6 days or 3 days is construed as respectively referring to girls whose ages were 8, 12, 16 or 20 at the time of marriage.⁵⁰ Another important question is inter-caste marriageṣ It has already been shown (pp. 50-58) how anuloma marriages were allowed, while pratiloma unions were condemned and how it was supposed that numerous sub-castes arose from such mixed marriageṣ It is the contention of several eminent scholars (e. g. Senart in his ‘Caste in India’ translated by Ross p. 124) that caste as implied in modern usage and in its strict sense did not exist in the times of the Ṛgveda and the other Vedic Samhitāṣ We have seen, however, that the four varņas had been recognised in the times of the Samhitās and that ideas of superiority and inferiority due to being of a particular varṇa had become quite prevalent. But the practices as to marrying outside one's varṇa and taking food had not become as rigid and cast-iron as they became in medieval timeṣ A few clear examples of inter-caste marriages may be cited from the Vedic Literature. The Śat. Br. (IV. 1. 5, S. B. E. vol. 26, pp. 272–275) narrates the story of the old and decrepit sage Cyavana who was a Bhārgava (descendant of Bhrgu) or Āṅgirasa to Sukanyā, the daughter of king Śaryāta, a descendant of Manu. The Śat. Br. (XIII. 2. 9. 8, ⁵¹ S. B. E. vol. 44, p. 326) quotes a half verse from the Vāj. S. (26. 30) and then remarks 'therefore he does not anoint (as king) the son of the vaiśya womaṇ' This suggests that a king might marry a vaiśya girl, but her son would not have the Vedic coronation ceremony performed on hiṃ Verses 17–19 of the Ṛgveda V. 61 are interpreted by the Brhad-devatā (V. 50) as referring to the marriage of the brāhmaṇa sage Śyāvāśva to the daughter of king Rathavitī Dārbhya. The story of Kavaşa Ailūşa who was styled ‘dāsyāḥ putra’ (either meaning ‘the son of a dāsī or used as a term of abuse) by the sages has been mentioned above (at p. 36 ). Turning to the dharma and grhya sūtras, the matter stands thuṣ Some of the grhya sūtras (like Āśv., Āp.) do not expressly say anything about the varṇa of the bride. The Āp. Dh. S.⁵² (II. 6. 13. 1 and 3) requires that one should marry a girl of the same varṇa, who was not given before to another and marriage in accordance with śāstra and says that by contravening these rules sin is incurreḍ So he condemns mixed marriageṣ The Māṇava gr. I. 7. 8 and Gaut. IV. 1 require that one should marry a girl of the same varṇa and are silent about marriages with a bride of another varṇa. But Gautama knew of such marriages and enumerates the names of several sub-castes due to mixed marriages (IV. 14-17) and he includes among those who are unfit to be invited at a śrāddha dinner a brāhmaṇa who is the husband of a woman of the śūdra caste (15. 18). Manu (III. 12), Śaṅkha and Nārada⁵³ (stripuṁsa, v. 4) say that the best course is to marry a girl of one's own caste. This is said to be pūrva kalpa (the foremost or the best procedure). Many also speak of a less advisable course (anukalpa, as it is called) viz. that a brāhmaṇa may take as wife a woman of any caste, a kṣatriya may marry a woman of his own caste or a vaiśya or a śūdra woman, a vaiśya may marry a vaiśya or śūdra woman and a śūdra only a śūdra womaṇ This is stated by Bauḍ Dh. S. I. 8. 2, Śaṅkha, Manu III. 13, Viṣṇu Dh. S. 24. 1-4, Pār. gr. I. 4 and Vaṣ Dh. S.⁵⁴ I. 25 inform us that some teachers allowed a dvija to marry a śūdra woman but without Vedic mantraṣ But Vasiṣṭha himself condemns it severely by saying ‘one should not do so (i. e. marry a śūdra girl), for by doing so the degradation of the family is certain and loss of heaven after death’. Though the Viṣṇu Dh. S. and Manusmṛti state that a dvijāti may marry a śūdra girl, that is not their own view, but they simply voice practices and sentiments prevalent in their time, since both denounce the marriage of a brāhmaṇa with a śūdra girl in the most un-measured termṣ Viṣṇu Dh. S. 26. 5-6 (S. B. E. vol. 7 p. 112) says that the union of a dvijāti with a śūdra woman can never produce religious merit ; it is from carnal desire alone that he marries her being blinded by lust ; and that dvijāti men who thus marry śūdra women quickly degrade their families and progeny to the status of śūdras'. The ancient editor of the Manusmṛti did not tolerate such marriages ( though he gave expression to the common trend of view in III. 12–13 ) and condemns a dvijāti's marriage with a śūdra woman in the strongest language ( III. 15–19, Manu III. 15 being the same as Viṣṇu Dh. S. 26. 5 ) ' a brāhmaṇa by having intercourse with a śūdra woman goes to hell and by procreating a son on her he loses his status as a brāhmaṇa'. Yāj. I. 57 allowed a brāhmaṇa to marry a girl of his own varṇa or of the varṇas next in order, but laid it down as his emphatic opinion that ' a dvijāti should not marry a śūdra girl'. It appears however that prevailing public opinion and practice was too much for both Manu and Yāj. ; for, both declare ( Yāj. II. 125 and Manu IX. 152–153 ) that when a brāhmaṇa has sons from wives of the four varṇas, the son of a brāhmaṇa wife takes four shares ( out of 10 in which his wealth is to be divided ), the son of a kṣatriya wife takes three, of a vaiśya wife two and of a śūdra wife one. Yāj. ( I. 91–92 ) recognises the marriage of a brāhmaṇa with a śūdra and says that the son of such a union is a pārāśava'. Manu III. 44 also recognises the marriage of a śūdra girl with one of a higher class by saying that she holds the hem of the garment of a bridegroom of a higher clasṣ

This shows that the ancient smṛtis ungrudgingly recognised marriages between a brāhmaṇa and a kṣatriya or a vaiśya girl. But opinion was not unanimous about the marriage of a dvijāti with a śūdra womaṇ Such marriages took place, but were looked upon with disfavour and often condemned with severity. Anuloma marriages were frequent enough till the 9th or 10th century A. D., but became rare later and for the last several hundred years they hardly ever took place or they were not at all recognized as valid by the communities concerneḍ The epigraphic records furnish well authenticated instances of inter-caste marriageṣ The Vākāṭakas were brāhmaṇas (their gotra being Viṣṇuvṛddha). The Poona plates of Prabhāvatīguptā (E. I. vol. 15 p. 39) establish that she was a daughter of Candragupta II. (first quarter of 5th century A. D.) of the Imperial Gupta dynasty and became the chief queen of the king Rudrasena II of the Vākāṭaka line. The Talagunda pillar inscription of the Kadamba king Kakusthavarman (E. I. vol. 8 p. 24) recites that . the founder of the Kadamba family was Mayūraśarman, a brāhmaṇa, who being exasperated with the Pallavas of Kāñcī took up the sworḍ His descendants have names ending in varman (as kṣatriyas should have, according to Manu II. 32 ). The inscription mentions that Kakusthavarman (who was 4th in descent from Mayūraśarman) got his daughters married to Gupta and other kingṣ In the Ghaṭotkaca Inscription of Yaśodharman and Viṣṇuvardhana (A. S. W. I. vol. 4, p. 140 ) we are told that Soma, a brāhmaṇa and ancestor of Hastibhoja, minister of the Vākāṭaka king Devasena, gave his heart to wives born in brāhmaṇa and kṣatriya familieṣ The Tippperūḷ copper-plate grant of a chieftain called Lokanātha (E. I. vol. 15, p. 301 ff) dated in the 44th year (probably of the Harṣa era i. e. about 650 A. D.) says that Lokanātha's ancestors belonged to the Bharadvāja gotra (p. 306) and that the maternal grand-father of Lokanātha was Keśava who is said to have been a pārāśava (p. 307), while Keśava's father Vira was a brāhmaṇa (dvijasattama). Virūpādevī, a daughter of the famous Vijayanagara king Bukka I (1268–1298 A. D.), was married to a brāhmaṇa named Brahma or Bomanna Woḍeya, who was the Governor of the Āraga province (E. I. vol. 15, p. 12). Vide also E. I. vol. 18, p. 87 (dated 894 of the Vikrama era ) for Pratihāra kings being descended from a brāhmaṇa Hari- candra and the latter's kṣatriya wife; the Ātpur Inscription of Śaktikumāra dated in saṁvat 1034 (977 A. D.) which states that the founder of the Guhila dynasty was a brāhmaṇa Guhadatta, whose descendant Bhartṛpaṭṭa married a Rāṣṭrakūṭa princesṣ

Classical Sanskrit Literature also yields certain well-known instances of inter-caste marriageṣ The Mālavikāgnimitra of Kālidāsa shows that Agnimitra, a son of Senāpati Puṣya- mitra of the Śuṅga dynasty and a brāhmaṇa, married Mālavikā who was a kṣatriya princesṣ In the Harṣacarita of Bāṇa (Ucchvāsa I towards end) we are told by Bāṇa himself that among the friends and companions of his wanderings there were his two pārāśava brothers Candrasena and Mātrṣena (i.e. step-brothers from a śūdra wife of his father). Rājaśekhara, teacher of king Mahendrapāla of Kanoj, says in his Karpūra- mañjarī (I. 11) that his accomplished wife Avantisundarī was descended from the Cāhuāṇa (modern Cavan) family (i. e. from a kṣatriya family).

It is extremely difficult to say when exactly inter-caste marriages even between dvijātis came to be prohibited by the smṛtis and writers of digestṣ Viśvarūpa⁵⁵ on Yāj. III. 283 (not later than first half of 9th century) clearly suggests that in his day a brāhmaṇa could marry a kṣatriya girl. Medhātithi on Manu III. 14 suggests that about 900 A. D. at the latest marriages of brāhmaṇas with kṣatriya and vaiśya girls took place rarely in his day but not with śūdra women; and the Mit. on Yāj. does not at all say that though marriages among dvijātis (anuloma ones) were allowed by Manu and Yāj. they had entirely ceased to be regarded as valid by its time. Many of the medieval digests and writers like the Sṃ C. and Hemādri quote verses stated to be from the Āditya-purāṇa or Brahma-purāṇa on matters forbidden in the Kali age ( and so called Kalivarjya) among which inter-caste marriages are includeḍ

The Āpastamba⁵⁶ smṛti (in verse eḍ by Jīvānanda I, p. 549, chap. IX ) says that by marrying a girl of another caste a man incurred the sin of mahāpātaka and had to undergo the penance of 24 kṛcchraṣ The Mārkaṇḍeya-purāṇa (113. 34–36 ) narrates the story of king Nābhāga who married a vaiśya girl by the rākṣasa form of marriage and who consequently incurred siṇ

The state of the law as to inter-caste marriages in British Indian courts may be briefly indicated here. By the Special Marriage Act (III of 1872 as amended by Act XXX of 1923 ) both anuloma and pratiloma marriages are validated if they are registered and solemnized according to the procedure prescribed by the Act. But if a marriage is not so solemnized under that Act, but under the general rules of Hindu Law, then all pratiloma marriages are invalid throughout British India. But anuloma⁵⁷ marriages have been recognised as valid by some High Courts in India; the Allahabad High Court, however, regards all anuloma marriages even as invaliḍ

The next restriction that the girl to be married must not be a sapinda of the bride-groom has now to be considereḍ Sapinda relationship is of special importance in three matters, viz. marriage, inheritance and āśauca⁵⁸ (impurity on birth and death). The prohibition against marriage with a sapinda girl applies to all varṇas including the śūdra⁵⁹. There are two schools about the meaning of sapinda, one represented by the Mitākṣarā and the other by Jīmūtavāhana (author of the Dāyabhāga). Both are agreed that a sapinda girl cannot be married, but they differ as to the meaning of the worḍ We shall first understand one interpretation of sapinda. The locus classicus is the commentary of Vijñāneśvara on Yāj. I. 52-53 a substantial portion of which is translated here: “‘asapinḍam’ (in Yāj. I. 52) means a woman who is not a sapinda and sapinda means one who has the same piṇḍa i. e. body (or particles of the body). Sapinda relationship (between two persons) arises from (their) being connected by having particles of the same body. Thus, the son has sapinda relationship with his father by reason of the (fact that) particles of the father's body continue in him (the son). Similarly (there is sapinda relationship) with the paternal grandfather and the like (of the grandson and the like) because through his (the grandson's) father particles of the body of the grandfather continue in (or are connected with) him (the grandson). Similarly (the son has sapiṇḍa relationship) with his mother because particles of the mother's body continue in hiṃ Thus (a person has sapiṇḍa relationship) with his mother's father through his mother; so (one has sapiṇḍatā) with one's mother's sister or mother's brother also by reason of the connection with (or continuity of) particles of the same body (viz. the body of the maternal grandfather). So also (a person has sapiṇḍatā) with the paternal uncle, the paternal aunt and the like. Similarly the wife has sapiṇḍa relationship with the husband because she produces (with the husband) one body (viz. their son). ¹⁰⁷¹ In the same way brother's wives (are sapiṇḍas) of each other, because they produce one body (viz. their respective sons) with their respective husbands who are produced from one body (viz. their husbands' father). In this manner wherever the word sapiṇḍa occurs, there one has to understand connection with (or continuity of) particles of the same body either directly or mediately.⁶² On the word ‘asapiṇḍām’ it was explained that sapiṇḍa relationship arises immediately or mediately by reason of the connec- tion with particles of the same body; this may prove to be too wide a statement, since in this beginningless saṁsāra, such a relationship might exist in some way or other between all men;⁶² therefore (the sage Yājñavalkya) states ‘after the fifth on the mother’s side and after the 7th on the father’s side’. After the fifth on the mother’s side i. e. in the mother’s line and after the 7th on the father’s side i. e. in the father’s line sapiṇḍa relationship ceaseṣ These words (viz. sapiṇḍa.......ceases) are to be taken as understood (in the text of Yāj.). Therefore the word ‘sapiṇḍa’ though it applies everywhere (i. e. to a very wide circle of persons) by the expressive (i. e. literal) power of its component parts, is restricted to a certain definite sphere, like the words ‘nirmanthya’ ⁶³ and ‘paṅkaja’. And so the six (ascendants) beginning with the father and the six descendants beginning with the son and the man himself as the seventh (in each case) are sapiṇḍas; wherever there is a divergence of the line, the counting shall be made until the seventh in descent is reached including him (i. e. the ancestor) from whom the line diverges; in this way the computation is to be made everywhere. And so the fifth (if a girl), who stands in the fifth generation when a computation is made beginning with the mother and going up to her (mother’s) father, mother’s paternal grandfather &c., is styled in an indirect way ‘fifth from the mother's side' (by Yāj.). In the same way 'the seventh from the father's side' (in Yāj.) is she who is the seventh in degree (from an ancestor) when computation is made beginning from the father and proceeding up to the father's father and so on ". According to this explanation of the Mitākṣarā⁶⁴ the following rules about prohibition based on sāpiṇḍya follow : ( 1 ) In computing degrees the common ancestor is to be included ; ( 2 ) regard is to be had to the father and mother of the bride and bridegroom both ; ( 3 ) if computation is made from the mother's side of either the proposed bride or bridegroom, they must be beyond the fifth degree ( i. e. they must be 6th or further on ) from the common ancestor and if it is made through the father of either, they must be beyond the seventh from the common ancestor ( i. e. they must be eighth &c. ). This last postulates four different classes of cases, viz. descent from a common ancestor may be traced through the fathers of both bride and bridegroom or through the mother of both, or through the mother of the bridegroom and the father of the bride or through the father of the bridegroom and the mother of the bride. The method of computing degrees is entirely different from the English method as is made clear by the remarks of the Mit. translated above. The rules require that both must be beyond sāpiṇḍa relationship as defined and limited above. For example, tracing descent through the fathers of both bride and bridegroom, if the bride is 8th from the common ancestor ( called kūṭastha in dharmaśāstra works ) and the bridegroom is 6th, there can be no marriage, as though the bride is beyond the limits of sāpiṇḍa relationship to the common ancestor, the bridegroom being 6th from the kūṭastha has sāpiṇḍa relationship with the kūṭastha. These rules apply according to the Mit. only where the ancestors married women of their own varṇa. But if an ancestor married a brāhmaṇa girl and also a kṣatriya girl and a question arose about the eligibility of marriage among descendants of these two wives, then prohibition based on sāpiṇḍa relationship extends up to only three degrees ( and not seven or five ).⁶⁵ It should not be supposed that these rules of the Mit. are universally observeḍ The texts themselves are in conflict. Further, the customs and usages as to prohibited degrees for marriage are so divergent in the various parts of India and among the hundreds of castes and sub-castes that it is impossible to state any rule as universally or even generally applicable. A few examples of conflict among the smṛtis are cited by the Mit. itself. Vaṣ Dh. S. (8. 2) says ‘(he may marry) the fifth on the mother's side and 7th on the father's side', while Yaj. as interpreted by the Mit. makes it necessary that the girl to be eligible must be 6th from the common ancestor traced through the mother and 8th when traced through the father. Paithīnasi says ‘the girl must be beyond three degrees on the mother's ⁶⁶ side and five degrees on the father's side '. These two are explained away by the Mit. by saying that these texts do not authorise a marriage with a girl who is distant from the common ancestor by a lesser number of degrees than those propounded by Yaj. but they only prohibit marriage with girls who are nearer in degrees than those stated in those textṣ But this reconciliation of the conflict among the smṛtis is not at all satisfactory.

The following diagrams illustrate the application of the rules of sapinda relationship according to the Mitākṣarā. In all of them A represents the common ancestor and the letters S and D represent respectively sons and daughterṣ According to the Mit. computation has to be made from ( and inclusive of ) the common ancestor in all cases and both the bride and bride-groom have to be beyond the prohibited degreeṣ ⁶⁷ No. 1 A D (2) D (2) S (3) S (3) S (4) S (4) S (5) D (5) S (6) S (6) S (7) S (7) D (8) S (8) Here a valid marriage might take place between S(8) and D(8) because sapinda relationship for both is traced through their fathers and both are removed from the common ancestor A by more than seven degrees or generationṣ

No. 2 A S (2) S (2) S (3) S (3) S (4) S (4) D (5) D (5) S (6) D (6) Here a valid marriage may take place between S (6) and D (6) because sapinda relationship of both is in this case traced through their mothers and they are both removed from the common ancestor by more than five degreeṣ

No. 3 A S (2) S (2) S (3) S (3) S (4) S (4) D (5) D (5) D (6) D (6) D (7) S (7) Here a marriage may take place between S (6) and D (6) because their sapinda relationship is to be traced through their mothers and both are removed from the common ancestors by more than five degreeṣ But a marriage cannot take place between S (7) and D (7) as the sapinda relationship of D (7) is to be traced through her father and she is not more than 7 degrees from the common ancestor.

No. 4 A S (2) S (2) S (3) S (3) S (4) S (4) D (5) S (5) D (6) S (6) Here a marriage cannot take place between D (6) and S (6) as the sapinda relationship in the latter's case is to be traced through his father and he is not more than seven degrees removed from the common ancestor, though D (6) whose sapinda relationship is to be traced through her mother is more than five degrees from the common ancestor. According to Bālabhaṭṭa and some others marriage will take place as D(6) is beyond five degrees (tracing through her mother), though S(6) is within 7 degrees (tracing descent through his father) and so is not outside sapinda limitṣ

All these four examples are taken from the Dharmasindhu (III Pūrvārdha p. 226–227). No. 3 illustrates what is called sāpiṇḍa relationship by ‘frog’s leap’.⁶⁸ Just as a frog leaps from one spot to another leaving intervening objects untouched, so in this example No. 3, there is sāpiṇḍya between D(5) and D(6), but S(6) and D(6) are left unaffected by sāpiṇḍya (as relationship is traced through the mothers of both), while sāpiṇḍya reverts to affect D(7) and S(7). The maxim of ‘frog’s leap’ is a very ancient one, being exemplified by the Mahābhāṣya of Patañjali.⁶⁹

The rules about prohibition of marriage on the ground of sāpiṇḍa relationship, particularly where cognates (persons connected through females) were concerned, have not been observed in numerous instances over wide areas from very ancient timeṣ One striking illustration is the question whether a man can marry his maternal uncle's or his paternal aunt's daughter, particularly the former. On this point opinion has been sharply divided from ancient timeṣ Āp. Dh. S. I. 7. 21. 8 includes among patanīya actions (mortal sins) sexual intercourse with the uterine relations (mothers and sisters) of one's parents and their childreṇ This would prohibit marriage with one's maternal uncle's or paternal aunt's daughter. Bauḍ Dh. S. I.⁷⁰ 1. 19-26 notes that there were five practices peculiar to the south viz. eating in the company of one whose upanayana has not been performed, to eat in the company of one's wife, to eat stale food, to marry the daughter of the maternal uncle or of a paternal aunt and there were five practices peculiar to the north. Then he proceeds to state the argument that those who follow the practices mentioned by him in countries other than those where they prevail incur sin, that Gautama (XI. 20) holds this last argument to be wrong and then Bauḍ states his own view that one should pay no heed to either set of practices, as both are opposed to smṛtis and the views of śiṣṭaṣ From this it is clear that a marriage with one's maternal uncle's daughter or paternal aunt's daughter was in vogue in the south (below the Narmadā probably) long before the Bauḍ Dh. S. (i. e. several centuries before the Christian era) and that North India did not go in for such marriages and that orthodox sūtra writers like Gautama and Baudhāyana reprobated such practiceṣ Manu (XI. 172-173) forbids such marriages "on approaching (for carnal intercourse) the daughter of one's father's sister or of one's mother's sister or of mother's full brother, a man must undergo the penance called 'cāndrāyaṇa'. A wise man should not take as his wife (any one of ) these; they are not fit to be wedded because they are (sapiṇḍa) relatives, for by wedding them one sinks low (i. e. falls into hell or loses caste)". Haradatta on Āp. Dh. S. II. 5. 11. 16 quotes a verse from Šātātapa ⁷¹ prescribing the penance of cāndrāyaṇa for marrying one's maternal uncle's daughter or a girl having the same gotra as one's mother's gotra (i. e. maternal grandfather's gotra) or a sapravara girl and a text of Sumantu forbidding marriage with the daughter of the brother of one's mother or step-mother. It would be seen that all these relatives are third from the common ancestor and so are very much within the prohibition against marrying a sapiṇḍa, whatever be the number of prohibited degrees for marriage that are accepteḍ Kumārila in his Tantravārtika while discussing the scope and validity of sadācāra (on Jaimini I. 3. 7) enumerates many lapses from good conduct attributed to great men and heroes of the past and tries to explain away those violations of good conduct. The charge brought against Vāsudeva (Kṛṣṇa) and Arjuna, the central heroes of the Mahābhārata, is that they respectively married Rukmiṇī and Subhadrā who were their maternal uncle's daughters (this being forbidden). Kumārila makes very scathing remarks against this practice ⁷² prevalent among southerners and then refutes this charge by saying that although in the Mahābhārata (Ādiparva 219. 17-18) Subhadrā is described as the daughter of Vasudeva and sister ⁷³ of Kṛṣṇa, she was really not so, but that she was Vāsudeva's mother's sister's daughter or was the daughter's daughter of the sister of the father of Vāsudeva's mother (or step-mother Rohiṇī).⁷⁴ Kumārila argues that such female cousins one degree or several degrees removed are often spoken of as sisterṣ Vide my notes to the Vyavahāramayūkha pp. 200-202 where this matter is fully explaineḍ This reply of Kumārila well illustrates the method of dealing with inconvenient texts adopted by mīmāṁsakas like Kumārila when the texts run counter to their cherished viewṣ Viśvarūpa on Yāj. III. 254 quotes Manu XI. 172 and verses of Saṁvarta prescribing the penance of Parāka for carnal intercourse with one's maternal uncle's daughter.⁷⁵ Medhātithi on Manu II. 18 refers to this practice as prevalent in some countries, and says it is opposed to the rule of Gautama (IV. 3) about prohibition on the ground of sāpiṇḍya and explains how such practices might have ariseṇ Persons making love to the handsome daughters of their maternal uncles married them for fear that they might otherwise be punished by the king; it may be that others literally following the words of Manu IV. 178 held by the practice which their forefathers resorted to for fear of punishment.⁷⁶ Among medieval writers, some condemn this practice, others justify it. Aparārka (pp. 82-84) after a lengthy discussion condemns marriage with maternal uncle's daughter. The Nirṇayasindhu also does the same (p. 286).

Among those who support marriage with a maternal uncle's daughter the most prominent are the Sṃ C. (I. pp. 70-74) and the Par. M. (I. 2. pp. 63-68). They both admit that there are Passages of Manu, Śātātapa, Sumantu (all quoted above) and others condemning marriage with a maternal uncle’s daughter, yet they say that there are passages of the Veda and some smṛtis in support of it and śiṣṭas (learned and respectable persons) observe this practice and so such marriages are sadācāra and not forbiddeṇ They rely upon two Vedic passages; one is from the Śat. Br. I. 8. 3. 6 (S. B. E. vol. XII. p. 238) “thus the separation (of the eater and the eaten) is effected in one and the same act; hence from one and the same man spring both the enjoyer (the husband) and the enjoyed (the wife); for now kinsfolk live sporting and rejoicing together saying ‘in the fourth or third man (generation) we shall unite’. And this is so in accordance with that (separation of the spoons).” ⁷⁷ The mention of the union in the third generation is understood by the Sṃ C. as referring to the marriage of a person with his maternal uncle’s or paternal aunt’s daughter. Viśvarūpa also refers to this passage (on Yaj. I. 53), but he does not draw the inference that such marriages are authorized by the Veda. Another Vedic verse is relied upon by the Sṃ C. and Par. M and other workṣ It is a verse from a Khila sūkta ‘Come, O Indra, by commended paths to this our sacrifice and partake of your portioṇ They (the priests) have offered the fat seasoned with ghee that is thy portion, as the maternal uncle’s daughter or the paternal aunt’s daughter (is one’s lot in marriage)’. ⁷⁸ This verse is referred to by Viśvarūpa on Yaj. (I. 53) but his explanation of it is different viz. that it really indicates one out of four different alternatives as regards prohibited degrees for marriage.⁷⁹ Aparārka (p. 83) on Yāj. I. 53 reads the latter half somewhat differently and so interprets it as to spell out of it a prohibition against marrying a maternal uncle's daughter. This text is quoted in I. L. R. 7 Maḍ 548 at p. 550. The Smṛticandrikā relies upon a passage of the Caturviṁśatimaṁta that allows a marriage of persons who are third or fourth in descent from a common ancestor on both sides (on the mother's or father's side). There is also a text of Bṛhaspati which prescribes that the practices of the countries, castes and families should be guarded (or enforced) by the king as they have been in vogue from past times, otherwise the subjects become inflamed and among such practices he instances 'Brāhmaṇas in the south marry the maternal uncle's daughter'.⁸⁰ So there is smṛti authority for such marriageṣ Besides southern śiṣṭas deeply read in the Vedas and acting according to the meaning of the Veda practise marriage with a maternal uncle's daughter. Here both Sṃ C. and Par. M. are on very firm grounḍ The Smṛtimuktāphala of Vaidyanātha says 'among the Andhras śiṣṭas deeply read in the Vedas follow the practice of mātulasutā-pariṇayana and among the Draviḍas respectable people allow marriage of a person with a girl who is fourth in descent from the common ancestor'.⁸¹ Among several castes in the Deccan and the Madras Presidency, not only is marriage with a maternal uncle's daughter allowed, but it is highly commendeḍ Even certain brāhmaṇa castes like the Deśastha brāhmaṇas of the Karṇāṭaka and Karhāḍā brāhmaṇas observe this practice in modern timeṣ The Saṁskārakauṣṭubha (pp. 616-620) and the Dharmasindhu support mātulasutā-pariṇayana. Both the Sṃ C. and Par. M. say that though the Vedic passage from the Śatapatha is a mere arthavāda (a laudatory statement) in praise of a Vedic act, yet on the analogy of the words 'he holds (the samidh) above (the sruc) for gods', are construed by Jaimini (III. 4. 15) as ā vidhi ( as a positive rule) on account of their apūrvalva ( not being known from any other source), these words of the Śatapatha are also to be construed as a vidhi ( of the marriage with maternal uncles daughter ).⁸² The Sṃ C. also presses into service the Vedic verse ( Rg. X. 16. 5 ) addressed by Yamī to Yama. But that mantra does not clearly refer to cross-cousin marriageṣ

The smṛti passages that condemn cross-cousin marriages are explained in a peculiar manner by the Sṃ C. and the Par. M. When a woman is married in one of the four forms, brāhma &c., she passes into the gotra of her husband, becomes a sapiṇḍa in the husband’s family and so she is severed from her father’s family (as to gotra and sapiṇḍa relationship); but when a woman is married in the āsura, gāndharva and other forms, she does not pass over into the gotra of her husband, but remains in the gotra of the father and her sapiṇḍa relationship with her father and brother continueṣ Therefore the son of such a woman, if he marries the daughter of his mother’s brother, would be marrying a girl who is a sagotra and sapiṇḍa of his mother. The Sṃ C. and the Par. M. and other works say that the smṛti texts forbidding marriage with maternal uncle’s daughter refer to a person whose mother was married in the gāndharva, āsura and the other two forms, but not to a person whose mother was married in the brāhma and the three other approved formṣ This is the reasoning of the Sṃ C. and the Par. M.⁸³ These rules about prohibition on the ground of sapinḍa relationship cause great difficulties in modern times, when owing to the rapid spread of co-education and the rise in the marriageable age of girls, love marriages have become somewhat frequent. Lovers do not stop to consider such ( to them ) trivial matters as sapinḍa relationship, but the law is often inexorable. The narrowing of the limits of sapinḍa relationship is permitted by the authors of commentaries and digests only on the ground of immemorial family, caste or local usage. For example, the same authors, who hold that marriage with the maternal uncle's daughter is sanctioned by Vedic texts and custom, condemn and forbid marriage with one's paternal aunt's daughter or with one's maternal aunt's daughter ( vide below ). There are a few verses cited from the smṛtis which favour the narrowing down of sapinḍa relationship. For example, the Caturviṁśatimata⁸³ states 'According to Śākaṭāyana there is no blame in marrying a girl who is 7th or 6th or 5th (from the common ancestor); similarly Manu, Parāśarya, Yama and Aṅgiras say that one may marry a girl who is third or fourth (from the common ancestor) on both sides (i. e. from her father's side or mother's side ). Whoever enters into such marriage alliances by relying on the usage of his country or of his family is always entitled to associate with people and this is seen from the Veda '. Parāśara (as quoted in the Nirṇayasindhu and other works) says 'a man who is himself fourth or fifth (from the common ancestor) may according to Parāśara's view marry a girl who is 4th or 6th (from the common ancestor), but one who is himself fifth cannot marry a girl who is also fifth (from the common ancestor ).' The Saṁskārakaustubha, the Sāpiṇḍyadīpikā and a few digests hold these verses authoritative and allow such marriages provided they are in accordance with local or family usage. As against this the following facts have to be remembereḍ One of the maxims of interpretation is that where there is conflict between smṛti texts, the preferable rule is to follow the opinion of the majority.⁸³ Gaut., Manu, Yaj., Marīci and numerous other smṛtikāras are opposed to these dicta. The second rule is that Manu has the highest authority among smṛtis and that smṛti which is opposed to the dicta of Manu is not commendeḍ⁸³ The above passages are opposed to the words of Manu on sapinḍa relationship and about marry-ing a maternal uncle's daughter (Manu V. 60, III. 5, XI. 171-172). A third point about the above texts is that several eminent works like the Nīrṇayasindhu and the Dharmasindhu⁸³ hold that some of these verses are baseless and not genuine and that they really refer to persons adopted or persons descended from wives of different castes of the same common ancestor; and lastly even those works that support such narrow-ing of sapinḍa relationship for marriage do not advocate it for all and sundry, but only where there is a local or family usage to that effect. Therefore, marriages among persons who are 3rd or 4th or 5th from the common ancestor cannot be regarded as valid in general, but only on the ground of usage. The following diagram will furnish some examples of the narrowing down of sapinḍa relationship.

A (common male ancestor) | --------------------------------- | | D (2) D (2) | | S (3) D (3) | | D (4) S (4) | S (5)

Here ordinarily there can be no valid marriage between D (4) and S (4) or S (5) because they are both not beyond even the fifth degree from the common ancestor; but if the verses of the Caturviṁśatīmata and Parāśara be followed the marriages are valiḍ It should be noted that even under the Special Marriage Act (III of 1872) marriage between D (4) and S (4) or S (5) will not be valid according to the 2nd proviso to sec. 2. The Nirnayasindhu⁸³ was not prepared to concede that such marriages, though not the best, were at least allowable as inferior (anukalpa). The Dharmasindhu⁸³ stated that only in calamities when one is unable to secure a girl one may enter into a marriage where sapinda relationship is narrowed down, but that those who can secure another girl should not go in for it, as the sin of incest would be incurreḍ Such marriages did not take place except during the last few years and so the courts have not yet pronounced on their validity. The preceding remarks are made by way of caution only and it is quite possible that courts may find out some way to uphold the validity of marriages when they come before them many years after their celebratioṇ But it is well to remember that the validity of such marriages in castes that have no usage to that effect is, to say the least, doubtful.

There is a great⁸⁶ deal of discussion in the smṛtis and nibandhas on the gotra of a womaṇ Āśv. gr. S. I. 8. 12 is interpreted by some as laying down that husband and wife become of one gotra one year after marriage. Laghu Hārīta appears to refer to this and also proposes an option that she takes up the husband's gotra immediately after marriage. Yama 86 and Likhita 25 say that after marriage on the 4th night a bride becomes one with her husband as to gotra, pinda and āśauca; while Yama 78 and Likhita 26 state that she loses her father's gotra on taking the seventh step. The Mit. on Yāj. I. 254 has a long note on this subject, states that there are two views and finally reaches the conclusion that a woman retains her father's gotra even after marriage for pindadāna, if she was a putrikā or was married in the āsura and the following forms; but if she was married in the brāhma and other approved forms, there was an option viz. piṇḍa may be offered to one’s mother by one’s father’s gotra or by her father according to family usage. Vide also Aparārka pp. 432, 542, Sṃ C. I. p. 69.

Both of them allow a person’s marriage with his paternal aunt’s daughter. Certain inscriptions at Nāgārjunikoṇḍa (of about the 3rd century A. D.) establish that the king Siri Chāntamūla who had performed the Vājapeya, Aśvamedha and other sacrifices, married a daughter of his paternal aunt (E. I. vol. 20 at p. 4). There are other authors, however, who, while allowing marriage with a maternal uncle’s daughter, disallow it with a paternal aunt’s daughter. For example, the Nirṇayasindhu says that, though it follows from the same Vedic passages that one’s paternal aunt’s daughter may be married, yet such a marriage should not be contracted as it is hateful to the people, as there are no other passages laying down the advissability of such marriageṣ⁸⁷ Similarly the Sṃ C. (I. p. 71) and Par. M. (I. 2. p. 65) say that though marriage with one’s mother’s sister or mother’s sister’s daughter also should be allowed on the same grounds as those urged in favour of the marriage with the maternal uncle’s daughter, yet it is forbidden since it is condemned by śiṣṭas and is hateful to the people and both again rely on Yāj. I. 156.⁸⁸

Another and a very striking instance of the limits of sapiṇḍa relationship not being observed is the practice among even brāhmaṇas (such as some Deśasthas in Karṇāṭaka and Mysore) marrying their own sister’s daughter (i. e. their own niece). Vide Mandlik’s Hindu Law p. 425. In Ramangavda v. Shivaji, the Bombay High Court held that marriage with one’s sister’s daughter was invalid among brāhmaṇas; but in that case no custom as to the validity of such marriages was alleged⁸⁹. In Vellanki Ramakrishna v. Kotagiri Subbamma 43 Maḍ 830 at p. 834 it is stated that in the Velama caste a man may marry his sister's daughter.

On account of these divergences about the limits of prohibited degrees for marriage it appears that the remarks of the Saṁskārakāustubha (p. 620) and of the Dharmasindhu (p. 228) are most sensible and practical. They say that even in the Kali age⁹⁰ those, in whose families or countries the limits of sapinda relationship are narrowed down and marriage with the maternal uncle's daughter has been in vogue for ages, may do (among whom there is no such custom) may without blame invite such persons for śrāddhas and may take girls from their families in marriage and that the passages quoted by Hemādri forbidding their being invited at śrāddhas only apply to those who marry a maternal uncle's daughter even though they have no such family or local usage.⁹¹

Another question is how far sapinda relationship holds good in the case of girls belonging to the family from which one's step-mother comeṣ The Udvahtattva (p. 118), the Nirnayasindhu (p. 289), the Saṁ. Pr. (pp. 695-699), Saṁskāra- Kaustubha (pp. 621-630) and the Dharmasindhu (p. 230) deal with this matter. They all rely on a text of Sumantu⁹² 'all the wives of one's father are mothers, the brothers of these are one's maternal uncles, their sisters are one's mother's sisters, the daughters of these are one's sisters and the children of these latter are one's sister's children; otherwise there would be saṁkara (confusion)'. Two interpretations are placed on this, one view (and that is held by most) is that this lays down sāpiṇḍya only with the persons specifically mentioned,⁹³ while another view (this is held by the Sāṃ Pr.) is that there is sāpiṇḍya for four generations from the father of one's step-mother. According to the first view a man cannot marry the daughter or grand-daughter of his step-mother's brother or sister, but his own daughter can marry the son of his step-mother's brother; while according to the second view the latter marriage also would contravene the rule about this extension (atideśa) of sāpiṇḍya.

Marriages with certain girls were forbidden by certain writers on the ground of what is called ‘viruddhasambandha', although there is no question of sapinda relationship in such caseṣ The Gṛhya-pariśiṣṭa⁹⁴ (quoted in the Nirnayasindhu p. 239) prescribes 'one should marry a girl with whom there is no viruddhasambandha, e. g. the daughter of one's sister or the sister of one's paternal uncle's wife (these should not be married). Viruddhasambandha ( incompatible relation ) occurs where the standing of the proposed bridegroom and the bride would resemble that of father or mother to the other ( or of brother and sister). In modern times such marriages take place as a matter of course and no court would declare them to be invaliḍ Vide I. L. R. 20 Maḍ 283 and I. L. R. 43 Maḍ 830 where marriage with one's wife's sister's daughter was held to be valid, among brāhmaṇas as well as śūdras in the Telugu and Tamil districts, relying on Mandlik's Hindu Law pp. 484-485 and other authoritieṣ In Ramchandra v. Gopal 32 Boṃ 619 at pp. 630-631 the court approves of the decision in 20 Maḍ 283 and holds that the dicta as to viruddha-sambandha are only recommendatory. About the sāpinḍa relationship of the adopted son as regards marriage, āśauca and śrāddha a good deal is said in many works such as Saṃ K. (pp. 182-186), Nirṇayasindhu (pp. 290-291), the Vyavahāramayūkha, the Saṁ. Pr. (pp. 688-694), Saṃ R. M. (pp. 453-456). Sāpinḍya for āśauca and śrāddha will be dealt with later oṇ As to sāpinḍya for marriage in the case of adoption there is great divergence of viewṣ The Saṁ. (p. 690) says that the son given in adoption has sāpinḍya for seven generations with his genitive father and for three generations with the adoptive father (pp. 693-694).¹¹⁰³ The Saṁ. K. appears to hold, after quoting several views, that if the upanayana of the adopted boy was performed in the family of birth, then he has sāpinḍya with the genitive family for seven generationṣ But if the ceremonies from jātakarama to upanayana are performed in the adoptive family then he has sāpinḍya in the adoptive family for seven generations, but only for five generations if only upanayana is performed in the family of adoptioṇ The Nirṇayasindhu gives its own view that there is sāpinḍya in both families for seven generationṣ The Vyavahāramayūkha holds (following Gautama IV. 3) that in the case of kevala dattaka (the ordinary adopted son) sāpiṇḍa relationship extends to seven degrees in the adoptive father’s family and five degrees in the adoptive mother's family.¹¹⁰⁴ The Dharma-sindhu follows the Saṁ. K. generally but it makes one significant remark. It says that several writers hold that in the case of dattaka, sāpinḍya is to be observed for a lesser number of degrees (than seven or five) in both familieṣ¹¹⁰⁵ The Saṁ. R. M. (p. 454) says that the limits of sāpinḍya spoken of as applying to an adopted person extend to his children also.¹¹⁰⁶ As the limits of sāpinḍya for marriage are very much narrowed down when marriage with one's maternal uncle's daughter or one’s niece is allowed, on the other hand certain communities extend such limits too much. The Deśastha brāhmaṇas of the Mādhyandina śākhā in the Deccan do not marry a girl whose father's gotra is the same as the gotra of the bridegroom's maternal grandfather. Manu III. 5 lays down 'that girl who is not a sapiṇḍa of the mother (of the bridegroom) and who is not a sagotra of the father (of the bridegroom ) is commended in the case of twice-born classes'. Many com- mentators and digests like Kullūka, the Madana-pārijāta (p. 132), the Dīpakalikā, the Udvāhatattva (p. 107 ) understand the word asagotra ( in Manu III. 5 ) after ‘mātuḥ’ also, and thus forbid marriage with a girl whose gotra is the same as that of one’s mother ( i. e. of one’s maternal grandfather ). Medhātithi⁹⁹ on Manu III. 5 quotes a text of Vasiṣṭha which prescribes the penance of cāndrāyaṇa for marrying a girl who has the same gotra as the bridegroom’s mother ( i. e. his mother’s father ) and abandonment of her. Haradatta on Āp. Dh. S. II. 5. 11. 16 quotes a similar verse from Śātātapa. Kullūka, the Sṃ C. (I. p. 69), Haradatta on Āp. Dh. S. II. 5. 11. 16, the Gr. R. (p. 10), Udvāhatattva p. 107 and other digests quote a verse of Vyāsa ‘some do not desire for marriage a girl who has the same gotra as one’s mother; but one may marry without question (or fear) a girl when her birth and name (as connected with one’s mother’s original gotra ) are unknown’. A woman on marriage loses her original gotra and passes over into the gotra of her husband; so the words ‘sagotra’ of the mother simply mean ‘having the same gotra as one’s mother once had in her maiden state’. The Sṃ C. (I. p. 69) takes this verse to refer only to a woman who was made an appointed daughter (putrikā) by her father. Halā- yudha also held the same view. But other writers do not approve of this ( vide the Gr. R., p. 10 ). The Nirnayasindhu¹⁰⁰ (III. pūrvārdha p. 302 ) forbids marriage with a girl who is sagotra of one’s mother, but the Saṃ K. (p. 693) and the Dharmasindhu both say, after quoting a text of Satyāṣāḍha, that this restriction applies only to those who study the Mādhyan- dina śākhā. The Bengal school represented by the Dāyabhāga and Raghunandana differ from the Mitākṣarā in the interpretation of the word 'sapiṇḍa'. In this system the word 'piṇḍa' is taken to mean 'the ball of rice' that is offered in śrāddhas to deceased ancestors &c. (while under the Mit. interpretation 'piṇḍa' means body or particles of body). Sapiṇḍa means 'one who is connected with another through oblations of food'. The author of the Dāyabhāga propounds his theory with reference to inheritance and he himself says that with reference to āśauca¹⁰¹ sapiṇḍa relationship is to be differently understooḍ Further Jīmūtavāhana does not set forth his theory of sapiṇḍa relationship with reference to marriage. His theory is that in matters of inheritance the guiding principle is the spiritual benefit ( upakārakatva ) conferred on the deceased through oblations of food, and he relies on Manu (IX. 106) for this propositioṇ For his sapiṇḍa theory he principally relies on two passages, Bauḍ¹⁰² Dh. S. I. 5. 113-115 and Manu IX. 186-187. Baudhāyana's words are 'the paternal great-grand-father, the paternal grand-father, the father, the man himself, his full brothers, his son, grandson and son's son from a woman of the same varṇa: all these participating in undivided dāya (heritage) are called sapiṇḍaṣ Those who participate in divided dāya are called sakulyaṣ Thus issue of the body existing, wealth goes to them; on failure of sapiṇḍas, the sakulyas'. The text of Manu (IX. 186-187) is 'To three, libations of water must be given; towards three piṇḍa proceeds; the fourth is the giver of these ( of water and piṇḍas ), there is no fifth properly (concerned in this). Whoever is the nearest among sapiṇḍas, his becomes the wealth of him (who dies). After him the sakulya becomes (the heir) &c'. Jīmūtavāhana's explanation of this text is as follows: 'A man when living offers piṇḍas to his three male ancestors; but when he himself dies, his son performs The sapindikarana¹⁰³ śrāddha for him, he is made one with the pitṛs and he then begins to participate along with his father and grandfather in the three piṇḍas offered by his son to the latter's three ancestorṣ Thus those to whom he offers piṇḍas and those who offer piṇḍas to him are called, since they share in undivided oblations, avibhaktadayāda sapiṇḍaṣ Several objections can be raised against the theory of Jīmūtavāhana. In the first place he assigns the meaning of piṇḍa to the word dāya in Baudhāyana's passage for which there is no warrant. Baudhāyana really means that those enumerated by him are called sapiṇḍas who take undivided wealth i. e. who constitute an undivided family. In the next place on his interpretation the word sapiṇḍa has to be interpreted in entirely different ways for āśauca and marriage. Besides he himself is not sure of his ground, since he says that although learned men may not approve of his theory that spiritual benefit is the guiding principle in taking property by inheritance they must admit that the order of heirs as stated by him relying on Manu IX. 186-187 is the proper one.¹⁰⁴

The following table illustrates the sapiṇḍas expressly so stated by the Dāyabhāga where P stands for the propositus, S for son and F for father : GGF12 — S13 — S14 — S15 | GF8 — S9 — S10 — S11 | F — S5 — S6 — S7 (4) | P — S1 — S2 — S3 Śrīkrsna, a commentator of the Dāyabhāga and author of Dāyakramasaṁgraha, and Raghunandana, author of Smṛtitattva, and others elaborate these ruleṣ The Full Bench of the Calcutta High Court in Guru Govind Shaha Mandal¹¹¹³ v. Ananad Lal Ghose Mazumdar gives an elaborate statement of the several sapindaṣ But as those rules have reference only to succession, they are passed over here. Raghunandana in his work on marriage called Udvāhatattva¹¹¹⁴ quotes the famous verse of the Matsyapurāṇa, 'the fourth and (two beyond him) among ascendants are partakers of lepa (the leavings of the boiled rice that become attached to the hand of him who offers pindas), the father and the rest (i. e. two more beyond him) are partakers of the pinda; he who offers the pinda to them is the seventh; sāpindya extends to seven generations' and refers the readers to his Śuddhitattva for elucidatioṇ Raghunandana does not give any definition of sapindas for marriage, but discusses the conflict about the several texts such as ‘7th on the father’s side and 5th on the mother’s side’. He expressly says (p. 110 of Udvāhatattva) that the words ‘who is not a sapinda of one’s mother’ are still applicable even if several females intervene between the common ancestor and the girl proposed to be marrieḍ He then adds a special rule that descendants of what are called pitṛbandhus and mātṛbandhus also are forbidden if they are within 7 degrees and 5 degrees respectively from these bandhuṣ The pitṛbandhus¹¹¹⁵ of a person are his paternal grand-father’s sister’s sons, his paternal grandmother’s sister’s sons and his father’s maternal uncle’s sons; while mātṛbandhus are a person’s mother’s father’s sister’s sons, mother’s mother’s sister’s sons, mother’s maternal uncle’s sonṣ These two latter may be illustrated by two diagramṣ No. I. Paternal line 7 6 5 4 3 Father's paternal aunt S (1) (1) father A (bridegroom)

Father's maternal line 7 6 5 4 3 father's maternal-grand-father Grandfather = Grandmother father's grand maternal mother's uncle S (3) mother's sister S (2)

N. B.—Here S1, S2 and S3 are the three pitrbandhus of A, the bridegroom, and they are the starting points for calculating prohibited degrees among paternal cognateṣ In the ascending line only the descendants of the common ancestors are excludeḍ For example, S1 is a pitrbandhu and his descendants up to 7 degrees are excluded; but S1's father is not a bandhu of the bridegroom; therefore S1's father's sister may be married by the bridegrooṃ Under this rule the 6th descendant (a girl) of S1 will be ineligible for marriage with A; but she will be 9th from A's grandfather's father who is the common ancestor. So it will be seen that this goes far beyond the limits of sāpiṇḍya generally prescribed and there is no valid reason assigned for thiṣ

No. II. Mother's paternal line 5 4 3 2 maternal great grand-father mother's paternal aunt S1 (1) maternal grand-father = maternal grand-mother mother A (bridegroom)

Mother's maternal line 5 4 3 mother's maternal grandfather mother's mother's sister S2 mother's mother's brother S3 N. B.—Here S1, S2 and S3 are mātṛbandhus of the bride-grooṃ The maternal great-grandfather is the starting point in calculating prohibited degrees in the mother's paternal line. In the mother's maternal line the starting points are S2 and S3. In the ascending lines of the mātṛbandhus the descendants of only the common ancestor are excludeḍ For example, the girl descended from S3's maternal ancestors may be married by the bridegroom and so also a girl descended from the paternal ancestor of S2 or S1.

Another rule propounded by Raghunandana is that even within prohibited degrees a valid marriage may be contracted if three gotras intervene. In the case of girls descended from pitṛbandhus and mātṛbandhus the computation of gotra must be made from theṃ For want of space it is not thought advisable to illustrate this by citing several exampleṣ But one example is given to illustrate this rule.

A (common ancestor) ┌───────────────────┐ │ │ D (1) S (1) │ │ D (2) S (2) │ │ S S (3) │ │ D (3) S (4) │ D (4)

Here according to the Bengal school S (4) can marry D (4) because three gotras intervene between her and the common ancestor, although S4 is only 5th in descent from the common ancestor; for according to the Bengal school it is not necessary that both the bridegroom and the bride be beyond the limits of sapindaśip, but only the bride need be so; while according to many writers of the Mitākṣarā school both must be beyond the limits of sapindaśip from the common ancestor.¹⁰⁸ The Dāyabhāga does not rely upon any Vedic passages for its theory of sapiṇḍa relationship. The Mit. (on Yaj. I. 52) relies upon three Vedic passages only in propounding the theory of sapiṇḍa relationship viz. ‘indeed the man himself is born from himself (as son, Ait. Br. 33. 1)’ ; ‘thou art born again ( or reproduced) in the offspring’ (Tai. Br. I. 5. 5 ) and ‘this body is made of six kośas (sheaths) ; three are derived from the father, three from the mother; bones, muscles and marrow from the mother’ (Garbhopanisad). These passages at the most say that particles of the bodies of the parents continue in their offspring; but they do not say anything about the meaning of sapiṇḍa or the limits of the sapiṇḍa relationship for marriage or inheritance. Even in the Ṛg. the words jñāti and bandhu which occur in the dharmasūtras ( e. g. Āp. Dh. S. I. 3. 10. 3, I. 5. 11 17. and Gaut. II. 44, IV. 3 and 5, VI. 3 ) frequently occur ( vide Ṛg. VII. 55. 5 and X. 85. 28 for jñāti and Ṛg. I. 113. 2, V. 73. 4, VII. 72. 2, VII. 67. 9 for bandhu ). All that we can say is that both meanings of sapiṇḍa were implicit in the word piṇḍa from the earliest times and that the sūtra writers were conscious of both meaningṣ

As to the grounds on which marriages between near sapiṇḍas were prohibited various theories have been advanced by anthropologistṣ Vide Westermarck in his ‘History of Human marriage’ (eḍ of 1921, vol. II. pp. 71–81 ) and Rivers on ‘Marriage of cousins in India’ in J. R. A. S. for 1907 pp. 611–640. Some think that the prohibition was due to the abhorrence which men in the primitive times felt for incest. To me it appears more probable that in India at least the prohibition was due to two causes; firstly, the observed fact that, if near relatives marry, their defects are transmitted with aggravation to their offspring and secondly the fear that, if marriages between near relatives by blood were allowed, there may be clandestine love affairs and consequent loss of morals and it would be difficult to secure husbands for girls who would be living under the same roof with several near or distant cousinṣ

The Par. M. I, part 2, p. 59 expressly says that only such a girl who is not a sapinḍa of the bridegroom within the prohibited degrees on any of the two theories of sāpiṇḍya (viz. by connection with particles of the body or with the balls of rice) is eligible for marriage.

The question what support Vedic Literature lends to the two interpretations of the word sapinḍa may now be briefly discussed here.

The word piṇḍa occurs ¹¹⁰ in the Ṛgveda (I. 162. 19) and the Tai. S. IV. 6. 9. 3 where it seems to mean ‘a part of the body of the sacrificial animal thrown into fire as an offering’. Here it is clear that the word piṇḍa is not used in the sense of ‘ball of rice’. But in the Tai. S. II. 3. 8. 2 and in the Śat. Br. II. 4. 2. 24 the word piṇḍa means ‘ball of rice’ offered to the Maneṣ The Nirukta III. 4 and 5 twice employs the words ‘piṇḍadānāya ’ (for offering balls of rice). But the word sapinḍa hardly ever occurs in the Vedic literature and we have no means of judging in what sense it was used in the Vedic literature. In the dharmasūtras the word sapinḍa occurs frequently and the dharmasūtras show a close connection between offering piṇḍa and the taking of inheritance (vide Gaut. 14. 13, 28. 21, Āp. Dh. S. II. 6. 14. 2, Vasiṣṭha IV. 16–18, Viṣṇu Dh. S. 15. 40).

It was shown above (pp. 436–37) that some sages prohibit marriage with a sagotra girl while others prohibit it with a sapravara girl. A number of sages and works like Viṣṇu Dh. S. 24. 9, Yāj. I. 53, Nārada (stripumsa, verse 7) require that the girl to be married must not be a sagotra nor a sapravara. Therefore it is necessary to understand the meaning of gotra and pravara. A detailed treatment of the topics of gotra and pravara would extend to a large treatise. Only a few salient points can be gone into here. The subject of gotra and pravara in the Vedic Literature has been treated at length by me in the J. B. B. R. A. S. ( New series vol. II. for 1935 ). Only the conclusions there drawn will be set out here. The word gotra in the Ṛg. means 'cow-stable' or 'herd of cows' in a few passages (I. 51. 3, II. 17. 1, III. 39. 4, III. 43. 7, IX. 86. 23, X. 48. 2, X. 120. 8). By a natural metaphor 'gotra' was applied to a cloud (in which waters are pent up as cows in a cow-stable) or to a cloud demon and also to a mountain range or peak which conceals water-yielding cloudṣ Vide Ṛg. II. 23. 3 (where Bṛhaspati's car is styled 'gotrabhid'), Ṛg. X. 103. 7 (=Tai. S. IV. 6. 4. 2, Atharva-veda V. 2. 8. and Vāj. S. 17. 39), Ṛg. VI. 17. 2, X. 103. 6. In some of these verses it is possible to take 'gotra' in the sense of 'fort'. In some cases 'gotra' probably means only "assemblage' (समूह) e. g. Ṛg. II. 23. 18, VI. 65. 5. From this last sense of 'assemblage' the transition to the meaning of 'a group of persons' is both easy and quick. There is no positive instance of the word 'gotra' being unchallengeably used in the sense 'descendants of a common patriarchal ancestor' in the Ṛgveda; but the concep- tion underlying the idea of gotra was, it is plain, quite familiar even in the age of the Ṛgveda. In the Atharvaveda V. 21. 3¹¹¹ the word 'viśvagotryaḥ' (belonging to all families) occurṣ Here the word 'gotra' clearly means 'a group of men connected together' (by blood). The Kauśika sūtra IV. 2 quotes a mantra in which gotra undoubtedly means 'a group of persons'.

Several passages of the Tai. S. show that descendants of great sages were called after those sageṣ In Tai. S. I. 8. 18. 1 it is said 'the Hotṛ is a Bhārgava' (descendant of Bhṛgu). The commentator explains that this is so only in the Rājasūya. It is quite possible that in those days descent was traced through teacher and pupil as well as through father and soṇ But there being a very few occupations only it is most probable that the son generally learnt from his father the lore of the latter. In Tai. S. VII. 1. 9. 1 we read 'therefore one does not find (or know) two Jāmadagṇiyas (in succession) who are poor (or grey- haired )'. From this it is clear that in the times of the Tai. S. Jamadagni was regarded as a very ancient sage, that several generations of Jamadagni's descendants had passed away by that time, that they were all known as Jāmadagṇiyas (or-gṇiyas) and that no two descendants were found to be poor (or grey- haired ) in successioṇ In numerous mantras of the Ṛgveda the descendants of well-known sages are denoted by the plural form of the names of those sageṣ In Ṛg. X. 66. 14 we read 'the Vasiṣṭhas have raised their voices like their father'. In Ṛg. VI. 35. 5 the Bharad-vājas are referred to as Āṅgiraseṣ According to Āśv. Śrauta-sūtra Bharadvāja is a gotra falling under the Āṅgirogaṇa. In the Brāhmaṇa Literature there are ample indications that priestly families had come to be formed into several groups named after their (real or supposed) founders and that such families differed in details of worship according to the group they belonged to. The Tai. Br. (I. 1. 4) prescribes that the consecration (ādḥāna) of the sacred Vedic fires is to be performed for Bhṛgus or Āṅgirases with the mantra ‘ bhṛgūṇām (or aṅgi-rasām) tvādevānāṁ vratapate vratenā-dadhāmi,’ that for other brāhmaṇas with the words ‘ādityānāṁ tvā devānāṁ vratapate’ &c. The Tai. Br. II. 2. 3 speaks of the ‘Āṅgirasi prajā’ (people of the Āṅgiras group). The Tāṇḍya Brāhmaṇa ¹¹² prescribes that the camasa (cup) made of udumbara was to be given as dakṣiṇā to a sagotra brāhmaṇa. The Kauṣītaki Br. ¹¹³ (25. 15) says that one who has performed the Viśvajit sacrifice ( in which everything that the sacrificer owned was gifted away) should stay for a year with a brāhmaṇa of the same gotra. The Ait. Br. (30. 7) contains the story of Aitāśa and his son Abhyagni and it is stated that the Aitaśāyanas Abhyagnis are the most sinful of the Aurvaṣ In the Kauṣītaki Br. where the same story occurs it is said that the Aitaśāyanas became lowest among Bhṛgus, as they were cursed by their father. According to Baudhāyanaśrauta-sūtra the Aitaśāyanas are a sub-section of Bhṛgugaṇa. Śunahśepa, when he was accepted as a son by Viśvāmitra, came to be called Devarāta and the Ait. Br. (33. 5) says that the Kāpileyaś and Bābhravas were affiliated to Devarāta. According to Bauḍ śrauta-sūtra Devarāta and Babhru are sub-sections of Viśvāmitra gotra. Śunahśepa is said to have been an Āṅgirasa by birth (Ait. Br. 33. 5 ) ¹¹⁴. So this makes it clear that gotra relationship was by birth in the times of the Ait. Br. ( and not from teacher to pupil ). In the Upaniṣads the sages when expounding the knowledge of brahma addressed their pupils by the gotra names e. g. by the Bhāra- dvāja, Gārgya, Āśvalāyana, Bhārgava and Kātyāyana gotras in Praśna I. 1, Vaiyāghrapadya and Gautama in Chāndogya V. 14. 1 and V. 16. 1; Gautama and Bharadvāja, Viśvāmitra and Jamadagni, Vasiṣṭha and Kaśyapa in Bṛ. Up. II. 2. 4. All this shows that the system of gotras with several sub-sections was well established in the times of the Brāhmaṇas and the ancient Upaniṣadṣ But the gotras are in these works referred to in connection with sacrifices or educatioṇ There is hardly any distinct reference in these to gotra or sagotra in relation to marriage. In the Lātyāyana-śrauta-sūtra ¹²² VIII. 2. 8 and 10 it is prescribed that one who has given away everything in the Viśvajit sacrifice should stay three nights with niṣādas and may partake of their jungle diet and then again three nights with 'jana' and then the sūtra gives several views about the meaning of ‘jana’ one of which, that of Dhanañjapya, is that 'jana' means a person with whom (i. e. with whose daughter) marriage is possible while one who is sagotra is called ‘samāna- jana’. This clearly proves that long before the Lāṭyāyana- śrautasūtra marriage with a sagotra had been forbiddeṇ Besides several of the gṛhya and dharma sūtras prohibit marriage with a sagotra girl. It is impossible to hold that this was a new conception that arose only about the time of these sūtraṣ Therefore, it must be supposed that prohibition as to sameneness of gotra in marriage had its origin long before the period of the sūtras in the times of the Brāhmaṇa works (if not earlier).

Gotra was of supreme importance in several fundamental matters and it largely entered into several practices of the ancient Aryanṣ A few examples may be given here. (1) In marriage sagotra girls were forbidden, vide above pp. 436-37. In the Lājahoma at marriage two offerings were to be made by all except Jamadagnyas, who had to make three (Āśv. gṛ. I. 7. 8-9).

(2) In matters of inheritance the wealth of one dying without issue went to his near sagotras ( Gaut. 28. 19 ).

(3) In śrāddha the brāhmaṇas to be invited should not belong as far as possible to the same gotra as the person invi- ting (Āp. Dh. S. II. 7. 17. 4, Gaut. 15. 20 ). (4) In pūrvana sthalipāka and other pākayajñas, all were to cut off oblations from the middle and fore-half of the havis, but for Jāmadagnyas ( who are pañcāvattins ) they were to be cut off from the middle, the fore-part and the hind part ( vide Āśv. gr. I. 10. 18-19 ).¹¹⁷ (5) In offering water to a preta ( a person recently dead ) his gotra and name were to be repeated ( Āśv. gr. IV. 4. 10 ). (6) In the caula ceremony tufts of hair were to be left in accordance with the gotra and practice of the family ( Khādira gr. II. 3. 30 ). (7) At the time of performing one's daily sandhyā prayer, one has to repeat even in modern times one's gotra and pravara, the Vedaśākhā and sūtra which one studieṣ As regards śrauta sacrifices a few interesting examples may be giveṇ Jaimini establishes that sattras ( sacrificial sessions extending over 12 days and more ) could be performed only by brāhmaṇas and that among brāhmaṇas the Bhrgus, Śunakās and Vasiṣṭhas are not entitled to perform them ( VI. 6. 24-26 ). Those of the Atri, Vadhryaśva, Vasiṣṭha, Vaiśya ( Vainya? ), Śaunaka, Kaṇva, Kaśyapa and Saṁkṛti gotras took Nārāśamsa as the second prayāja, while others took Tanūnāpāt as the second ( vide Śabara on Jaimini VI. 6. 1 ). The conception of pravara is closely interwoven with that of gotra from very ancient timeṣ The two have to be studied together. 'Pravara' literally means 'choosing' or 'invoking' (prārthanā).¹¹⁸ As Agni was invoked to carry the offerings of ṛṣis ( his remote ancestors ) who in former times had invoked Agni, the word pravara came to denote one or more illustrious ṛṣis, ancestors of a sacrificer. A synonym of pravara is ārṣeya or ārṣa ( as in Yāj. I. 52 ). Pravara entered into several domestic ceremonies and practices according to the gṛhya and dharma sūtras. For example: (1) a bride was to be chosen whose father's pravara was not the same as that of the bridegroom's father. Vide above p. 437. (2) In upanayana the girdle (mekhalā) was to have one, three or five knots according to the number of ṛṣis constituting the boy’s pravara (vide Śāṅ. gr. II. 2). (3) In Caula the tufts of hair to be left on the head depended on the number of sages constituting the pravara of the boy’s family (Āp. gr. 16. 6). The mass of material on gotra and pravara in the sūtras, the purāṇas and digests is so vast and so full of contradictions that it is almost an impossible task to reduce it to order and coherence. The learned author of the Pravara-mañjarī (which is the leading work on the subject) wrote in despair ‘Here, in the parts of sūtras that have been quoted there is a great diver-gence in the order (of the names of pravaras) of the texts of the several sūtrakāras, this being specially so in the text of Āśvalāyana (śrauta)-sūtra. Thus, though divergence is clearly established, yet following the order of the texts of the majority of writers such as Baudhāyana, Āpastamba and Kātyāyana we shall declare (the rules) about marriage or no marriage’. ¹¹⁹ We have first to understand what gotra in the sūtras and digests means and how it is inter-related to pravara. Among the sūtras that treat of gotra and pravara the śrauta sūtras of Āśvalāyana (Uttaraṣaṭka VI, khaṇḍas 10-15), Āpastamba (24th praśna) and Baudhāyana (B. I. eḍ vol. III pravarādhyāya at end) are the most important. The Pravaramañjarī (p. 5) has a verse to the effect that Baudhāyana’s pravarādhyāya is the best on the subject. The Śrautasūtra of Satyāṣāḍha Hiraṇyakeśi (21st praśna) has a section on this subject, which is the same as Āpastam-baśrauta with a few omissions and variationṣ The Baudhā-yanaśrauta-sūtra says ¹²⁰ ‘Viśvāmitra, Jamadagni, Bharadvāja, Gautama, Atri, Vasiṣṭha and Kāśyapa are the seven sages and Agastya is the eighth; the progeny of these eight sages is declared to be gotra’. These seven sages are probably derived from what is stated in the Br. Up. II. 2. 3-4 (=Śat. Br. XIV. 5. 2. 6 ) where these very seven sages are enumerated as those intended in the śloka ' arvāg-bilāścamasa¹²¹ &c. ' quoted there. The same work ( Bauḍ śrauta ) states that there are thousands, tens of thousands and arbudas ( millions of millions ) of gotras, but the pravaras are only 49. Besides the sūtra works, some purāṇas like the Matsya ( chap. 195–202 ), the Vāyu ( chap. 88 and 99, Ānaṇ eḍ ), Skanda III. 2 ( Dharmāraṇya kāṇḍa ) contain elaborate enumerations of gotras and pravaraṣ The Mahābhārata sets out at length the subdivisions (such as Madhuc-chandas, Devarāta ) of Viśvāmitragotra in Anuśāsanaparva 4. 49–59. Digests like the Smṛtyarthasāra ( pp. 14–17 ), the Smṛ. Pr. ( pp. 591–680 ), the Saṁ. K. ( pp. 637–692 ), the Nirṇaya- sindhu, the Dharmasindhu, the Bālaṁbhaṭṭī contain a vast material on this topic. There are also special works like the Pravaramañjarī on this topic. The general conception about gotra is that it denotes all persons who trace descent in an unbroken male line from a common male ancestor. When a person says ' I am Jamadagni-gotra ' he means that he traces his descent from the ancient sage Jamadagni by unbroken male descent. As stated by Bauḍ cited above, from very ancient times these male founders were supposed to be eight. This enumeration of eight primary gotras seems to have been known to Pāṇini. Patañjali says ' there were eighty thousand sages who observed celibacy. The accepted opinion is that the spread of progeny was due to eight sages includ- ing Agastya. The offspring ( apatya ) of these eight are gotras and others than these are called gotrāvayava'.¹²² Pāṇini defines gotra for grammatical purposes as 'apatyam pautra- prabhṛti gotram' (IV. 1. 162), which means 'the word gotra denotes (in my work on grammar) the progeny (of a sage) beginning with the son's son'. For example, the son of Garga would be called Gārgi, but the grandson would be called Gārgyaḥ and the plural Gargāḥ would denote all descendants of Garga (downwards from Garga's grand-son). But this definition is a technical one in grammar and is meant to indicate how derived (taddhita) words are to be formed by means of termina- tionṣ Even Pāṇini makes use of this technical sense only in the apatyādhikāra, but elsewhere in his work he uses the word gotra in the popular sense as comprehending all descendants of a common male ancestor. Vide the Kāśikā on Pāṇini II. 4. 63, IV. 2. 39 and IV. 3. 80. The Sām Pr. (pp. 591-592) lucidly explains ¹²³ thiṣ That a man belongs to a particular gotra is known by him only from tradition, from his father and other elders or from people about him, just as he knows that he is a brāhmaṇa from the same source ¹²⁴. Medhātithi on Manu III. 5. 194 has a very lucid and interesting discussion on this topic. His argument is: just as, though all persons are men, some are called brāhmaṇas, so among brāhmaṇas certain persons are known by immemorial usage (or convention) as belonging to certain gotras like Vasiṣṭha and the sūtra-kāras lay down that a certain gotra has certain pravaras; so the word gotra is applied to Vasiṣṭha and other sages by rūdhi (by convention or long-standing usage). It cannot be supposed that a person called Parāśara was born at a certain time and then his descendants came to be called Parāśaraṣ In that case the Veda would not be anādi (beginningless), as it is sup- posed to be, since it mentions Parāśara, Vasiṣṭha &c. So gotra is anādi like the brāhmaṇa caste and the Veda. The word is also secondarily used to denote a person, who is very illustrious on account of his learning, wealth, valour or generosity, who thereby gives a name to his descendants and then becomes the founder of the family. This is laukika gotra. But this is not the meaning of gotras which brāhmaṇas have. The secondary meaning may apply to the word gotra when used in the case of kṣatriyas.'¹²⁵ The Mit. on Yaj. I. 53 says 'gotra is that which is known from tradition handed down in the family.'¹²⁶ Each gotra is associated with one, two, three or five sages ( but never four or more than five ) that constitute the pravara of that gotra.¹²⁷ The gotras are arranged in groups, e. g. there are according to the Āśvalāyana-śrauta-sūtra four sub-divisions of the Vasiṣṭha gaṇa, viz. Upamanyu, Parāśara, Kuṇḍina and Vasiṣṭha ( other than the first three ). Each of these four again has numerous sub-sections, each being called gotra. So the arrangement is first into gaṇas, then into pakṣas, then into individual gotraṣ The first has survived in the Bhṛgu and Aṅgirasa gaṇa. According to Bauḍ the principal eight gotras were divided into pakṣas. The pravara of Upamanyu is Vasiṣṭha, Bharadvāsu, Indrapramada; the pravara of the Parāśara gotra is Vasiṣṭha, Śāktya, Pārāśarya; the pravara of the Kuṇḍina gotra is Vasiṣṭha, Maitrāvaruṇa, Kauṇḍinya and the pravara of Vasiṣṭhas other than these three is simply Vasiṣṭha. It is therefore that some define pravara as 'the group of sages that distinguishes the founder ( lit. the starter ) of one gotra from another'.¹²⁸

Though the word pravara does not occur in the Ṛgveda, the word ' ārṣeya ' occurs therein and the system of pravara goes back almost to the Ṛgveda. Ṛg. IX. 97. 51 has ' thereby may we acquire wealth and ārṣeya resembling Jamadagni'ṣ' Sometimes the idea of invoking Agni is conveyed without using the word pravara or ārṣeya. In Rg. VIII. 102. 4¹²⁹ it is said 'I invoke Agni just as Aurva, Bhṛgu and Apnavāna diḍ' It is remarkable that these are three of the five pravara sages of the Vatsa-Bhṛgus according to Bauḍ (3). Rg. I. 45. 3 has ‘O Jātavedas (Agni), give heed to the summons of Praskanva, as in the case of Priyamedha, Atri, Virūpa and Angiraṣ'¹³⁰ In Rg. VII. 18.21 it is said 'they who from house to house gladdened thee, being desirous of thee, viz. Parāśara, Śatayātu and Vasiṣṭha, will not forget the friendship of a liberal patron (like thee)'.¹³¹ It deserves to be noted that this mantra mentions Parāśara (who in later mythology is the grandson of Vasiṣṭha and son of Śakti), Śatayātu (who is Śakti according to Sāyana), and Vasiṣṭha. Parāśara, Śakti and Vasiṣṭha constitute the pravara of Parāśara gotra (according to Aśv. and Baud). In the Atharvaveda (XI. 1.16, XI. 1.25, 26, 32, 33, 35, XII. 4. 2 and 12, XVI. 8.12–13) ārṣeya means 'descendants of sages or those who are related to sages'. In the Tai. S. both ārṣeya and pravara occur in the sense of the sūtraṣ In the Tai. S. II. 5. 8. 7 (which refers to the recita- tion of the Sāmidheni verses) we read "he says 'choose (or invoke) ye the fire called havyavāhuna'; he chooses him (the fire) of the gods ; he chooses the ārṣeya ; in so doing he does not depart from the relationship (by blood) and doing so serves for conti- nuity. He chooses the later ones beginning from the remoter ones". In this passage 'ārṣeyam' appears to be used in the sense of 'one or more illustrious ancestors of the sacrificer' and reference is made to one of the two modes of mentioning the sages constituting the pravara. 'Ārṣeya' may also be taken here as an adjective (qualifying Agni), the meaning being 'he invokes Agni by the names of the illustrious ancestor sages of the yajamāna'. In one mode the remotest ancestor is named first in a taddhita (derivative) formation, then his descendant and so on, the sage nearest the sacrificer being mentioned last. For example, the pravara of Bhṛgu Vatsa is ' Bhārgava-cyavana- apnavānaurva-jāmadagnyeti'. This method is employed by the Hotṛ priest when he invokes fire as the divine Hotṛ with the pravara-mantra 'Agne mahān-asi brāhmaṇa bhārata deveddha manviddha rṣiṣṭuta &c.' (vide Tai. S. II. 5. 9. and Śat. Br. I. 4. 2, Aśv. Śr. I. 2. 27-I. 3. 6). In the other mode the affix 'vat' is used after the name of each pravara sage and the remotest one is mentioned last (e. g. Jamadagnivat, Ūrvavat, Apnavānavat, Cyavanavat, Bhṛguvat). This mode is employed by the adhvaryu when he chooses the Hotṛ priest.¹¹³⁸ The Tai. S. II. 5. 11. 9 appears to refer to one (i. e. 2nd) mode. The Kauṣītaki brāhmaṇa explains the purpose of taking the names of ancestors 'as the gods do not partake of the offering of him who has no list of (illustrious) ancestors, therefore he pronounces the ārṣeya of the sacrificer.' The Ait. Br. (34. 7) has an interesting passage on pravara. When a brāhmaṇa is initiated for a sacrifice, that fact is announced in these words 'a brāhmaṇa has been initiated for a sacrifice.' How is the initiation of the kṣatriya to be announced ? The reply of the Ait. Br. is 'even in the case of the kṣatriya the announcement is to be in the same form (viz. a brāhmaṇa has been initiated); but with the pravara of the family priest. Therefore they should proclaim the kṣatriya's initiation as a sacrificer with the ārṣeya of his family priest and should invoke Agni with his family priest.'¹¹³⁹ The Aśv. Śr. (Uttara śatka VI. 15. 4-5) and Bauḍ Śr. (pravarapraśna 54) say that in the case of kṣatriyas and vaiśyas the pravara of their purohita was to be employed or the pravara 'Māṇava-Aila- Paurūravasa' or simply 'Manuvat.' The origin of that rule is to be found in the above passage of the Ait. Br. There is another similar passage in the Ait. Br. (35. 5). The Śat. Br. I. 4. 2. 3–4¹³⁴ shows that the illustrious ancestors to be invoked were supposed to be related as father and son and not by apostolic successioṇ The Mahābhārata says that the original gotras were only four viz. Aṅgirās, Kāśyapa, Vasiṣṭha, Bhṛgu,¹³⁵ The verses are rather abruptly introduced in the epic and there is nothing to show on what this statement was based and it appears that it is due to the imagination of the writer. Bauḍ as shown above states that the original gotras are eight. But it is remarkable that Bhṛgu and Aṅgirās (whose divisions and sub-divisions are many) are not included by him in these eight. Therefore, it appears that even Baudhāyana is not correctly stating the number of original gotraṣ Gautama and Bharadvāja are stated to be two out of the original eight, but both of them instead of being separately dealt with are grouped under the comprehensive Aṅgirasagaṇa. So even Bauḍ is not to be implicitly followeḍ The Bālambhaṭṭī mentions eighteen principal gotras (eight as in Bauḍ plus ten more some of which are names of mythical kings¹³⁶). Bauḍ himself says that there are millions of gotras and in the Bauḍ Pravarādhyāya there are over 500 names of gotra and pravara sages; while the Pravaramañjarī quotes a verse that 'there are three crores of them¹³⁷ and so the gotra system is difficult to comprehend', and it mentions about 5000 gotraṣ Therefore, as the Smṛtyarthasāra says the nibandhas endeavour to place the innumerable gotras under groups and distribute them among 49 pravaras (mentioned by Bauḍ).¹³⁸ Some idea of these gotras and their distribution among the pravaras is given below. The appendix under note No. 1144 collects together the 49 pravara groupṣ

The Bhrgugaṇa and the Aṅgirogaṇa are very extensive. The Bhrgus are of two sorts, Jāmadagnya and non-Jāmadagnya. The Jāmadagnya Bhrgus are again towfold, Vatsas and Bidas (or Vidas); the non-Jāmadagnya Bhrgus are fivefold viz. Ārṣṭiṣeṇas, Yāskas, Mitrayus, Vainyas and Śunakās (these latter five are called kevula Bhrgus). Under each of these sub-divisions there are many gotras, on the names and number of which the sūtra- kāras are not agreeḍ The pravara of Jāmadagnya Vatsas is constituted by five sages according to Bauḍ and by three according to Kātyāyana. The pravara of the Bidas and the Ārṣṭiṣeṇas also has five sageṣ¹³⁹ These three (viz. Vatsas, Bidas and Ārṣṭiṣeṇas) are styled pañcāvattin (Bauḍ5) and they cannot inter-marry (the reason will be explained below). The five non-Jāmadagnya Bhrgus have each of them numerous sub-divisionṣ These divisions of Bhrgus are given here according to Bauḍ Āp. has only six of them (and not seven as he excludes Bidas from this group). According to Kātyāyana, Bhrgus have twelve sub- divisions (vide Sāṃ Pr. p. 625 ).

The Aṅgirogaṇa has three divisions, Gautamas, Bhara- dvājas and kevalāṅgirases; out of whom Gautamas have seven sub-divisions, Bharadvājas have four (Raukṣāyaṇas, Gargas, Kapis and Kevala-Bharadvājas ), and Kevala-Aṅgirases have six sub-divisions and each of these again is sub-divided into numerous gotras. This is according to Bauḍ Other sūtrakāras differ as to the subdivisionṣ Atri (one of the eight primary gotras) is subdivided into four (Atris proper, Vādbhūtakas, Gaviṣṭhiras, Mudgalas ). Viśvāmitra is sub-divided into ten, which are further sub-divided into 72 gotras. Kāśyapas are sub- divided into Kāśyapas, Nidhrūvas, Rebhas, and Śāṇḍilaṣ Vasiṣṭha has four sub-divisions (Vasiṣṭhas with one pravara only, Kuṇḍinas, Upamanyus and Parāśaras ) which are further sub- divided into 105 gotras. Agastya has three sub-divisions (Agastyas, Somavāhas and Yajñavāhas ), the first of which is further sub-divided into twenty gotras. When it is said that marriage with a sagotra or a sapravara girl is forbidden, each of these is separately an obstacle to marriage. Therefore, a girl, though not sapravara, may be yet sagotra and so ineligible for marriage or though not sagotra may yet be sapravara and not eligible. For example, the gotras of Yaska, Vādhūla, Mauna, Mauka are different, yet a marriage between persons belonging to these gotras is not possible, because the pravara of all these is the same, viz. ‘Bhārgava-Vaitahavya-Sāve-tasa-iti’.¹⁴⁰ So also though the gotras Samkṛti, Pūtimāṣa, Taṇḍi, Sambu and Śaṅgava are different, there can be no marriage between them as the pravara is the same, viz. ‘Āṅgirasa, Gaurivita, Sāmkṛtya’ (acc. to Āśv. Śrauta). When it is said that samāna-pravaras cannot marry, sameness may be due to only one sage being the same in the pravaras of two gotras or there may be two or three or more sages that are commoṇ The general rule is that if even one sage is the same¹⁴¹ in the pravaras of two different gotras, then they are sapravara, except in the case of the Bhṛgu group and the group of Aṅgiraseṣ In these two latter unless there are at least three sages common (when the pravara is constituted by five sages) or at least two sages are common (when pravara consists of only three sages), there is no sapravaratva and no bar to marriage. It will be noticed that among the five pravara sages of the Vatsas, Bidas and Ārṣți-ṣeṇas quoted above there are three sages that are common and so they cannot inter-marry.

Though the vast majority of gotras have three pravara sages, a few have only one pravara sage or two sages or five. Those who have only one sage are Mitrayus (pravara Vādhrya-śva according to Āśv.), the Vasisthas (other than Kuṇḍina, Parāśara and Upamanyu) have only one pravara Vasiṣṭha, the Śūnakas have one pravara Gṛtsamada (according to Āśv.) or Saunaka or Gārtsamada (according to Bauḍ 9); the Agastis have one pravara Agastya (according to Āp. Śr. 24. 10. 9); Āśv. optionally allows three pravaras to Mitrayus (Bhārgava-Daivodāsa-Vādhryaśva) and to Śūnakas (viz. Bhārgava- Śaunahotra-Gārtsamada and Āp. also allows to Agastis three pravaras (Viz. Āgastya-Dārdhacyuta-Aidmavāha-iti). According to Āp. and the purāṇas, Dhāpayantas have two pravaras (Vaiśvāmitra-Paurana-iti ), Aṣṭakas Lohitas have two (Vaiśvā-mitra-Aṣṭaka iti), and Śaṇḍilas also have two (Daivala-Asita iti). But he says that according to some the latter have three pravaras also (Kāśypa-Daivala-Asita-iti), while Baudhāyana gives four optional groups of three sages each for Śāṇḍilas (Bauḍ 43). For Vāri-Dhāpayantas Āśv. prescribes three only (Vaiśvāmitra-Devarāta-Paurana) and also for the Aṣṭakaṣ According to Bauḍ the Vatsas, Bidas and Ārṣṭiṣeṇas (among the Bhrgus) have five pravara sages; so also among Aṅgirases, the Kaumaṇḍas, Dīrghatamasas, Rauksāyaṇas and Gargas have five pravara sages, though the last have optionally three also.

The Āp. śr. ¹⁴² (24. 5 6) says ‘the sages to be invoked were to be three and they were to be seers of mantras’. It is therefore that the number of pravara sages is limited; there was no such requirement for a gotra and so gotras multiplied to an unlimited extent.

There are certain families that are dvigotras (i. e. have two gotras). Āśv. uses the word ‘dvipravācanāh’ for them ¹⁴³. They are principally three viz. Śauṇga-Śaiśiris, Sāmkr̥tis and Laugākṣiṣ From a Śuṇga, a sub-division of Bharadvāja gotra, was born a son to the wife of a Śaiśiri, a sub-division of Viśvā-mitra (by niyoga); the son, therefore, came to be called Śauṇga-Śaiśiri. Therefore Śauṇga-Śaiśiris cannot marry in both Bharadvāja and Viśvāmitra gotraṣ Their pravaras are given below ¹⁴⁴. They have to select one pravara sage from one gotra and two from the other gotra or two from the first and three from the other. A pravara of four sages is not allowed nor of one beyond five. For the other dvigotras, the Saṁ. K. (pp. 682-686), Nirnayasindhu p. 300 ff. may be consulteḍ In the case of the adopted son also, on the analogy of the Śauṇga-Śaiśiris, both gotras and the pravaras of both gotras have to be considered and the dictum of Manu (IX. 142) that ‘the son given does not share the gotra and inheritance of the genitive father and the svadhā (śrāddhas etc.) of the giver ceases’, is restricted only to matters of inheritance, śrāddha and the like and does not apply to marriage. ¹⁴⁵

A few words must be said about the gotra and pravara of kings and other kṣatriyas. It appears from the Ait.Br. quoted above (p. 488) that in the case of kṣatriyas the pravara of their purohita was employed in religious acts where pravara had to be reciteḍ This leads to the inference that most kṣatriyas had forgotten their gotras and pravaras by that time. The Śrauta sūtras allow ¹⁴⁶ an option to kṣatriyas (to kings according to Āśv.). ‘They may employ the pravara of their purohitas or all kṣatriyas may employ the same pravara viz. Mānav-Aila-Paurūravasa-iti’. Medhātithi on Manu III. 5 states that the distinctions of gotras and pravaras concern primarily brāhmaṇas alone and not kṣatriyas and vaiśyas and quotes Āśv. Śr. (I. 3) in support. The Mit. and other nibandhakāras rely on the first alternative mentioned in the sūtras and say that in marriages of kṣatriyas and vaiśyas the gotras and pravaras of their purohitas should be considered, as they have no specific gotras of their own ¹⁴⁷. This is carrying the doctrine of atideśa (extension) too far or with a vengeance. The Saṃ K. (pp. 689–690) assigns reasons for this attitude of the Mit. but they are not satisfactory. Ancient literature and epigraphic records show that kings had gotras of their owṇ In the Mahābhārata it is said that Yudhishthira. went to the court of Virāṭa in the guise of a brāhmaṇa and was asked by the king to state his gotra he said that he was of the Vaiyāghrapadya gotra ( Virāṭa-parva 7. 8-12 ). That this was the gotra of the Pāṇḍavas also follows from the fact that in the Bhīṣmatarpaṇa¹⁴⁸ that is performed on the eighth day of the bright half of Māgha, Bhīṣma's gotra is given as Vaiyāghrapāda or-padya and pravara as Sāṁkṛti. Jaimini ( VI. 6. 12-15 ) establishes that the Kulāyayajña was to be performed jointly by a king and his purohita ( and so they could use different pravaras it appears ). The Pallavas of Kāñcī had Bhāradvāja as their gotra ( vide E. I. vol. I p. 5 ). The Cālukyas are often described as 'mānavya-gotra' ( vide E. I. vol. VI. p. 337 ). In a copper-plate grant of Jayacandra ( dated samvat 1233 i. e. 1176 A. D. ) the donee was a kṣatriya described as Rāuta-śrī-Rājyadharavarman of the Vatsa gotra and of five pravaras viz. Bhārgava-Cyavana-Apnavāna-Aurya-Jāmadagnya.¹⁴⁹ In the Garra (Bundelkhand) plate of the Candella king Trailokyavarma of Kālañjara a village is granted to Rāuta Sāmanta of the Bhāradvāja gotra in recognition of his father's death in battle with the Turuṣkas ( E. I. vol. 16, p. 274 ).

In many inscriptions and copper-plate grants hundreds of donees with their gotras and pravaras occur and it would be an interesting study to compare the latter with the material derived from the sūtras and nibandhaṣ For example, vide E. I. vol. 19, pp. 115-117 and 248-250 for about 205 donees and their gotras in the plates of Bhāskaravarman, E. I. vol. 14, p. 202ff ( 500 donees with gotras in the time of Candradeva Gāhaḍavāla, samvat 1150 ), E. I. vol. 13 p. 237, E. I. vol. 8 pp. 316-317 ( śake 1346 ), E. I. vol. 9 p. 103 ( 32 donees with gotras, pravaras, śākhās &c. ), E. I. vol 12 pp. 163-167 ( 120 donees with gotras, śake 1508 ), Gupta Inṣ No. 55 Cammak plate of Pravarasena &c. Even the Buddhists kept the system of gotras ( vide E. I. vol. 10, Lüder’s list, No. 158 ).

According to Āp. Śr. the vaiśyas had¹⁵⁰ a single pravara 'Vatsapra', while according to Bauḍ they had three, viz. Bhālandana-Vatsapra-Māńktila’. They could also employ the pravvara of their purohitaṣ The Sāṃ Pr. (p. 659) says that Bhālandana is the gotra of vaiśyaṣ

If a person does not know his own gotra and pravara he should take those of his ācārya (teacher of Veda), according to Āp.¹⁵¹ Though he takes his teacher's gotra, it is only the teacher's daughter that is forbidden to him in marriage and not other girls of the same gotra as the ācārya'ṣ Both Sāṃ K. and Sāṃ Pr. (p. 650) quote a verse that when one does not know one's gotra one should call oneself of Kāśyapa gotra. This is so when he does not know even his teacher's gotra. The Sṃ C. (śrāddha section p. 481) says that this is so even when one does not know the gotra of one's maternal grand-father (i.e. he should offer piṇḍa with Kāśyapa gotra to his maternal grand-father).

The word gotra also came to mean any family name (surname). In the inscriptions we find this usage very ofteṇ For example, in the Bannahalli plate of the Kadamba king Kṛṣṇavarma II (E. I. vol. VI, p. 18) a śreṣṭhin (a merchant) is said to have been of the Tuṭhiyalla gotra and pravara. The Reḍḍi king (a śūdra) Allaya Vema of Rajahmundry was said to have been of the Polvola gotra (E. I. vol, XIII. p. 237, of śake 1356).

A few interesting matters about pravaras are noted below. It is noteworthy that even as to the same gotra, there is great divergence among the sūtrakāras about the sages constituting the pravara e. g. as to Śāṇḍilyagot-ra. Āśv. gives two groups of the sages 'Śāṇḍila-Asita-Daivala-iti' or 'Kāśyapa-Asita-Daivala-iti', but Āp. states his own view that there are only two sages in the pravara viz. ' Daivala-Asita iti' and that some say they are three, Kāśyapa-Daivala-Asita-iti; while Bauḍ states four groups, 'Kāśyapa-Avatsāra-Daivala-iti, Kāśyapa-Avatsāra-Asita-iti; Śāṇḍila-Asita-Daivala-iti; Kāśyapa-Avatsāra-Śāṇḍila-iti.' No adequate reasons can be given why even at so early an age as the sūtras, not only the order of the names in the pravara, but the very names in the pravara and their number should have varied so much. Bauḍ¹⁵³ (pravarādhyāya sec. 44) notes that the Laugākṣis (or Laukākṣis) are Vasiṣṭhas by day and Kāśyapas by night and their pravaras also show this double relation-ship. The Smṛtyarthasāra says that this picturesque description of them has reference to the prayājas and the like i. e. by day they follow the procedure of prayājas peculiar to Vasiṣṭhas and by night that peculiar to the Kāśyapaṣ

Among the gotras there are names of mythical kṣatriyas and kings like Vitahavya and Vainya and among the names of pravara sages many legendary kings like Māndhātṛ, Ambarīṣa, Yuvanāśva, Divodāsa appear. Vitahavya figures even in the Ṛgveda as closely connected with the Bhṛgus (Ṛg. VI. 15. 2-3). In the Mahābhārata it is narrated that Vitahavya, being a king, attacked Divodāsa, whose son Pratardana pursued Vita-havya, whereupon Vitahavya took shelter with Bhṛgu and that when Pratardana asked Bhṛgu whether there were any kṣatriya in the hermitage, Bhṛgu replied that there were all brāhmaṇas and that by this Vitahavya became a brāhmaṇa (Anuśāsana chap. 30).

Similarly the pravara of the Hārītas is either Āṅgirasa-Ambarīṣa-Yauvanāśva-iti or Māndhātṛ-Ambarīṣa-Yauvanāśva-iti. These are mythical royal sageṣ Among ¹⁵⁴ the Bhṛgus is a sub-division called Vainya which is further subdivided into Pārthas and Bāskalaṣ The story of Pṛthu who milked the earth is well-known (Droṇa-parva 69); he is called ‘ādirāja’ in Anuśāsana 166. 55. The Vāyupurāṇa in several places narrates that some kṣatriyas became the pravaras of brāhmaṇaṣ Vide chap. 88. 72-79, ( about Viṣṇuvṛddha, ¹⁵⁵ who was descended from Purukutsa, whose son was Trasadasyu ), chap. 88. 6-7, 92. 6, 99. 158-161, 99. 169-170 ( Ānaṇ eḍ) for other examples ¹⁵⁶. How and why kṣatriya names were adopted as brahmanical pravaras is obscure and difficult to understanḍ If one may hazard a guess, it is probably due to the fact that the purāṇas retain very ancient traditions of times when there were no water-tight varṇas and that ancient kings were learned in the Vedic lore and maintained śrauta fires also, became famous as sages in whose name fire was to be invoked to carry offerings to Gods even by brāhmaṇas who came ages after theṃ

The connection of gotra and pravara may be stated thus: Gotra is the latest ancestor or one of the latest ancestors of a person by whose name his family has been known for genera- tions; while pravara is constituted by the sage or sages who lived in the remotest past, who were most illustrious and who are generally the ancestors of the gotra sages or in some cases the remotest ancestor alone.

It has been seen (pp.437–38) that marriage between parties that are sagotra or sapravara is no marriage and the woman does not become the man's wife. What were the consequences of such a void union? Bauḍ¹⁵⁸ (pravarādhyāya 54) says that if a man has intercourse with a sagotra girl he should undergo the penance of cāndrāyaṇa, after that he should not abandon the woman, but should only maintain her as if she were a mother or a sister; if a child is born it does not incur sin and it should take the gotra of Kaśyapa. Aparārka quotes (p. 80) Sumantu and another smṛti to the effect that if a person inadvertently marries a sagotra or samāna-pravara woman he should give up intercourse with her, should maintain her and undergo cāndrāyaṇa.¹⁵⁸ But if he knowingly marries a sagotra or sapravara girl the penance was heavier (viz. that for incest) and if he has intercourse with her or begets a child from her he loses his caste and the child will be a caṇḍāla.¹⁵⁹ The rule of Bauḍ that there would be no blemish and the child will be of Kaśyapa gotra is restricted to inadvertently marrying such a girl.¹⁶⁰ The Saṃ Pr. quotes a verse of Kātyāyana to the effect that if a marriage is gone through with a sagotra, the girl may be again given in marriage to another. But the Sāṃ Pr. gets rid of that inconvenient text by saying that it does not apply to the present age. So the poor girl for no fault of hers had to pass her whole life in enforced celibacy, being neither an unmarried woman nor a widow.¹⁶¹ Questions about the validity of sagotra or sapravara marriages have not yet come before the courts; but it is likely that in the near future courts will have to deal with such caseṣ If a marriage takes place under the Special Marriage Act of 1872 (as amended in 1923) no difficulty will arise; but recently several marriages have been celebrated under the old Śāstric procedure between persons who are sagotra or sapravara. Upon the strict letter of the dharmaśāstra texts such marriages are absolutely voiḍ But it seems that the legislature should intervene and declare sagotra and sapravara marriages valiḍ Whatever may have been the case thousands of years ago when there were no means of communication and when there were small communities, the prohibition of sagotra relationship had some plausibility and real feeling of close kinship about it; but now the prohibition has become meaninglesṣ A man from Kashmir may marry a girl from Madras and the parents of both may have the same gotra. Granting for argument that the gotra sage was a common ancestor, one does not know how many generations have intervened between that remote ancestor and the intending spouses and particles of the ancestor's body, if they have survived at all in the intending spouses of the same gotra, must be in the present generation in the most attenuated state. The prohibitions based upon gotra and pravara are said by orthodox people to be prescribed upon unseen (adṛṣṭa) grounds and so they argue that they must be held to be absolute and not admitting of any evasion by any one calling himself a Vedic Hindu. One may be permitted to reply to these people that when almost every-body has given up the ancient cult of the śrauta and gṛhya fires and when hundreds of other innovations in ancient practices have been accepted without demur, there is hardly any justification for sticking to this one remnant of ancient practiceṣ As a matter of fact many brāhmaṇas now do not remember their pravara themselves but have to be told by the priests what their pravara iṣ Certain prohibitions against marriage between near sapinda relations may be respected on account of their universal acceptance. Even the Special Marriage Act of 1872 prescribes that there can be no marriage between parties when they trace descent to a common ancestor who is the great-grandfather or great-grandmother ( or who is nearer even than these ) of any one of theṃ

Sagotra relationship is in one direction wider than sapinda relationship and narrower than it in another so far as marriage is concerneḍ A man cannot marry the daughter of any sagotra, however distant the sagotra may be. Similarly even an adopted man cannot marry the daughter of a sagotra of his genitive father for two reasons, firstly because, though on adoption he becomes severed from his natural family for inheritance and offering of pinda ( vide Manu IX. 142 ), his other rela tionships with the natural family remain intact ; and secondly because Manu ( III. 5 ) says that the girl must not be a sagotra of the bridegroom's father ( and so even if by adoption a man goes into another gotra, it is the father's gotra that is to be considered ). Sapinda relationship prohibiting marriage extends only to seven or five generations, but prohibitions on the ground of sagotra relationship extend to any number of generationṣ On the other hand sapindas may be either of the same gotra ( i. e. sagotra ) or of a different gotra ( i.e. bhinnagotra ). Thus up to a certain limit sapindas include both sagotras and bhinnagotraṣ The latter are called bandhus by the Mit.; they are all cognate relations and have an important place in inheritance. We have seen ( p. 436–37 ) that sagotra marriages were totally forbidden in the śrauta sūtras ( in the śrauta sūtras like Lātyāyana śrauta and in gṛhya and dharma sūtras²³⁶). A passage of the Śat. Br. ( quoted above at p. 461 ) is relied upon by Professors Macdonell and Keith ( Vedic Index vol. I. p. 236 ) for holding that marriages within the third and fourth degrees on both maternal and paternal sides were allowed in the days of the Śatapatha and that therefore a man could marry in those days his paternal uncle's daughter. This latter is a startling propositioṇ The passage in the Śat. Br. is no doubt expressed in general words ( 'one may be united in the third or fourth generation' ). But that passage is gene rally applicable even if only a marriage with a maternal uncle's daughter or paternal aunt's daughter is meant. The passage does not expressly allow paternal cousins' marriageṣ The dharma-sūtras prohibit sagotra marriageṣ Both the learned Professors, as most western scholars do, probably hold that The Śatapatha is not separated from the sūtras by more than a few centurieṣ If a paternal uncle’s daughter had been eligible for marriage in the times of the Śatapatha, but became forbidden in the times of the sūtras we shall have to suppose that an usage died out from one end of the country to another and an opposite usage became prevalent throughout within a few centurieṣ The smṛtis are not afraid of stating ancient practices which they themselves do not accept (i. e. niyoga). The smṛtis do not say that marriages with paternal uncle’s daughters were ever allowed anywhere. So it appears that the Śatapatha is not referring to marriage with paternal uncle’s daughter, but to marriage with maternal uncle’s or paternal aunt’s daughter. It must however be pointed out that Aparārka (pp. 15. 63), the Sṃ C. (I. p. 12), Par. M. I. part 1 p. 133 and other digests quote a passage from the Brahmapurāṇa ¹⁶² that sagotra and sapiṇḍa marriages are forbidden in the Kali age. It may be argued with some force that this implies that sagotra marriages once took place. But there are various ways of explaining thiṣ It is possible that in the purāṇa the word gotra is not used in the technical sense, but only in the sense of family or surname. Among śūdras there is no gotra in the strict sense, but they also do not marry a girl who is believed to be of the same family, though the exact relationships or generations are not knowṇ Besides it has been shown above that if a woman was a putrikā or married in the Gāndharva or Āsura form she retained the gotra of her father and the son of the putrikā would have the gotra of his maternal grand-father and yet being of a different family his marriage with the daugther of a sagotra of the maternal grand-father might have taken place in ancient times and was forbidden in the Kali age by the purāṇa. Similarly the implied reference to marriages of sapiṇḍas as taking place in former ages has probably marriage with maternal uncle’s daughter in view. When in the kalivarjya texts it is said ‘ these dharmas are declared to be prohibited by the sages in the Kali age ’ (vide Vyavahāra-mayūkha p. 242), it is not proper to argue that everyone of the practices forbidden in the kali age was valid in former ageṣ All that is meant seems to be that most of them were allowed in former ages and these along with others enumerated are not to be practised in the Kali age, just as when in a crowd many have umbrellas and a few have not, a speaker still says with reference to the whole crowd ‘chatrino gacehanti’ (here go persons with umbrellas).

There are certain other prohibitions about marriage. The Smṛtimuktāphala¹⁶³ quotes a verse of Hārīta that one should not give one’s daughter in exchange to another’s son and receive that other’s daughter in marriage for one’s son, one should not give two daughters to the same man (at the same time), nor should one give one’s two daughters to two persons who are brotherṣ But these dicta will now certainly be held as merely recommendatory. Besides there is no objection in India in modern times to marrying the deceased wife’s sister, though even in England the deceased wife’s sister could not be married until 1907 (when the Deceased Wife’s Sister’s Marriage Act, 1907, Edw. 7 chap. 47 legalised such marriages).

The next question is as to who have power to arrange for the marriage of a girl and to give her away. The Viṣṇu Dh. S. (24. 38-39) specifies the order of persons who are entitled to exercise this right of guardianship in marriage ‘the father, the paternal grand-father, a brother, a kinsman, a maternal grand-father and the mother are the persons, by whom the girl may be given away in marriage. In the absence of the preceding one (the right) devolves upon the next in order, in case he is able’. Yāj. (I. 63-64) gives a slightly different order viz. he omits the maternal grand-father and adds that the right can be exercised only when the guardian is not affected by lunacy and similar defects and that in the absence of these the girl should perform svayaṁvara (i. e. choose a husband herself). Nārada¹⁶⁴ (stripumsa verses 20–22) gives the order as father, brother (with father's consent), paternal grand-father, maternal uncle, agnates, cognates, mother (if sound in mind and body), then distant relations, then the maiden may perform svayaṁvara with the king's permissioṇ To give away a girl in marriage was not only a right but was rather a heavy responsibility, as (Yāj. I. 64) and others declare that if a girl is not got married by the guardian at the proper time the latter incurs the sin of the murder of an embryo. The practice of svayaṁvara is well known from the Rāmāyaṇa and the Mahābhārata, but it was confined mostly to the princely familieṣ Manu (IX. 90–91) required a girl to wait for three years after she became marriageable before she could choose her husband herself. But Viṣṇu Dh. S. 24. 40 says that a girl should wait only for three monthly periods after she attains puberty and that after that period she has full power to dispose of herself in marriage as she thinks best.

There are no rules in the smṛtis as to who is to arrange the marriage of a male, since in ancient times they did not contem-plate the marriages of minor maleṣ

The mother has been put low in the order of guardians probably because of the dependent status assigned to women and because of the fact that in the ceremony of kanyādāna she cannot personally engage but has to get it done through some male relative. The courts in modern India, however, have held that the mother is entitled to select a husband for her daughter even when the girl's paternal grand-father is living, though the actual gift may be made by a male. Vide Bai Ramkore v. Jamnadas, I. L. R. 37 Boṃ 18, where Yāj. I. 63 is interpreted as only laying down who are to make a gift of the girl and as not intended to take away altogether the mother's right of selecting a bridegroom in favour of even distant relationṣ Vide also Indi v. Ghania (I. L. R. 1 Lahore 146), Jawani v. Mula Ram (I. L. R. 3 Lahore 29, where it was held that, after the father, the mother of a girl has the preferential right to select a bride-groom for her and that she is under no obligation to consult the girl's paternal kindred such as a paternal uncle), Ranga-naiki v. Ramanuja I. L. R. 35 Maḍ 728 (where all the authorities are examined). The Dharmasindhu states the important proposition that when the girl performs svayaṁvara or when the mother is to give away the girl in marriage, the girl or mother should perform the Nandiśrāddha and the princi-pal saṁkalpa is to be pronounced by her and the rest of the rite is to be performed through a brāhmaṇa.¹⁶⁵ Nārada states the general rule that if anything is done by one who is afflicted with lunacy or similar defects, what he does is as if not done. So a marriage settled even by the father if he is a lunatic need not be performeḍ¹⁶⁶ If an unauthorized person (like a maternal uncle) were to give away a girl in marriage though her father is alive and fit, what is the result? The digests state¹⁶⁶ that if the marriage has been completed by the performance of saptapadī, it cannot be set aside merely on the ground of the want of authority in the giver, since marriage rites are the principal matter and the authority to give is a very subsidiary matter, the absence of which cannot affect the principal matter. But before the marriage takes place a person who wants to give away a girl, though persons better qualified exist, can be prevented from doing so. Courts in modern India have followed these rules, relying on the doctrine of 'factum valet quod fieri non debuit' (what ought not to be done when done is valid) and holding that when once a marriage is duly solemnized and is otherwise valid, it is not rendered invalid because it was brought about without the consent of the proper guardian for marriage or in contravention of an express order of the court. Vide Khushalchand v. Bai Mani (I. L. R. 11 Bombay 247) and Bai Diwali v. Moti (I. L. R. 22 Boṃ 509).

A few words must be said about the sale of girls in marriage. We read in the Maitrāyaṇīya S. I. 10. 11 'she indeed commits falsehood (or sin) who being purchased by her husband roams about with other maleṣ'¹⁶⁷ There is another passage of the Veda relied upon along with the above by the pūrvapakṣa (the plausible view-point) in Jaimini (VI. 1. 10-11) which denies that women have a right to take part in Vedic sacrifices, viz. 'one should give to the daughter's father a hundred (cows) plus a chariot.' Jaimini replies (VI. 1. 15) that the giving of a hundred is not for purchasing a bride, but it is only a duty and a hundred must be offered as a present ( whether the girl is beautiful or not ).¹⁶⁸ This shows that, even if some girls were purchased for marriage in the times of the Maitrāyanīya S., there was a popular revulsion of sentiment about this practice and the sale of girls was severely condemned by the time of the sūtrakāraṣ The Āp. Dh. S. (II. 6. 13. 10–11) also makes interesting remarks on this point ¹⁶⁹ "there is no gift and the incidents of purchase about one's children; in marriage the gift ordained by Veda to be made to the daughter's father in the words 'therefore one should give a hundred (cows) besides a chariot to the girl's father and that (gift) should be made to belong to the (married) couple' is due to the desire (of the father to give a status to the daughter and her sons) and is meant as a fulfilment of duty (and not as a sale transaction). The word 'purchase' applied to such a transaction is merely figurative, since the relationship (as husband and wife) arises (not from the so called purchase but) from dharma." Vaṣ Dh. S. (I. 36-37) quotes the two passages of the Veda (from Mait. S. and the other about the gift of one hundred cows) in support of the Mānuṣa (i. e. Āsura) form of marriage. The Nirukta VI. 9 ¹⁷⁰ while explaining Ṛg. I. 109. 2 ('O Indra and Agni, I have heard you to be greater donors than a partially fit son-in-law or a brother-in-law') remarks that "the word 'vijāmātā' means among the southerners the husband of a woman who is purchased; what is meant is that he is a bridegroom who is deficient and not endowed with all good qualitieṣ" So Yaska implies that in the south girls were sold for substantial sums of money to persons who (either because they were old or wanting in some qualities desirable in a good bridegroom) were therefore really deficient as bridegroomṣ In the Nirukta (III. 4) while discussing the several views about the obscure verse of the Ṛgveda III. 31. 1 (sāsad-vahnir &c.) one of the reasons assigned for the view that women do not inherit is that gift, sale and abandonment in the case of women exist, but not in the case of men, to which some reply that these (gift, sale &c.) can be made of males also as is seen in the story of Sunahṣepa (in the Ait. Br. 33, Tai S. V. 2. 1. 3, Tai. Br. I. 7. 10.).

These passages lead to the inference that in ancient times girls were sometimes purchased for marriage, as was the case in many other countrieṣ But gradually public feeling entirely changed and not only was the sale of daughters by the father or brother severely condemned, but even taking of presents by them was looked down upoṇ Āp. Dh. S. has already been quoted above (note 1174). The Bauḍ Dh. S. (I. 11. 20-21)¹⁷¹ quotes two verses 'that woman who is purchased with wealth is not declared to be a legally wedded wife (a patnī); she is not (to be associated with the husband) in rites for the gods or manes and Kaśyapa declares that she is a dāsī (slave girl). Those, who, blinded by greed, give their daughters in marriage for a fee (śulka), are sinners, sellers of their own selves and perpetrators of great sin and they fall into hell &c.' In another place Bauḍ says 'he who gives his daughter (in marriage) by sale (as a chattel) sells his merit (punya)'. Manu (III. 51, 54-55) strikes¹⁷² a tender note about daughters when he says 'a father should not take even the smallest gratuity for his daughter; if he takes a gratuity through greed he becomes the seller of his child; when relations do not take for themselves wealth given by the bridegroom as gratuity (but hand it over to the girl) there is no sale (of the girl); the wealth so taken is for honouring the maidens and is only taken from the bridegroom out of loving concern for theṃ Fathers, brothers, husbands and brothers-in-law desiring their own welfare should honour women and should give them ornaments'. Manu (IX. 98) further recom- mends that 'even a śūdra should not take a gratuity when giving his daughter (in marriage), since in taking a gratuity he clandestinely sells his daughter'. Yāj. III. 236, Manu XI. 61 include the sale of children among upapātakaṣ The Mahā- bhārata (Anuśāsana 93. 133 and 94. 3) condemns the taking of a gratuity for giving a girl and Anuśāsana (45. 18-19) speaks of gāthās of Yama contained in dharma-śāstras 'who- ever sells his son for a price, or gives a daughter for the sake of his own liveli-hood in return for a gratuity, would fall into a most horrible hell called Kālasūtra'. Verse 23 of the same chapter says 'even a stranger cannot be sold, what of one's own children'. Verse 20 (= Manu III. 53) condemns even the ārṣa form of marriage as a sale because a pair of cattle is therein taken by the girl's father. In Kēraḷa or Malabar it is believed that the great teacher Śamkara laid down 64 ācāras, among which are prohibition of the sale of girls, prohibition of satī &c. Vide I. A. vol. IV, pp. 255-256 and also Atri v. 389 and Āp. (in verse) IX. 25 (eḍ by Jīvānanda). This practice, however, persisted till modern timeṣ For example, in an inscription dated about 1425 ¹⁷³ A. D. from Paḍaivīḍu (North Arcot District) we find an agreement signed by the representatives of Karṇāṭa, Tamil, Telugu and Lāṭa (South Gujarat) brāhmaṇas that they would give up taking gold for their daughters and get them married by the simple kanyādāna (the Brāhma form) and that the father who accepted gold and the bridegroom who paid gold were to be punished by the king and were to be excommunicated from the brāhmaṇa caste. The Peshwa ¹⁷⁴ issued orders (about 1800 A. D.) addressed to the brāhmaṇas of Wai ( in the Satara District ) forbidding them to take money for giving their daughters in marriage and prescribing fines for the father who took money, for the giver and the intermediary who brought about the marriage. Among certain castes and among the śūdras even now money or money's worth is often taken when a daughter is married, but generally the money is meant as a provision for the girl and for defraying the expenses of the girl's father.

The question of the sale of girls in marriage has been from the remotest ages bound up with the question of the father's power over his children Rg. I. 116. 16¹⁷⁵ and I. 117. 17 refer to the story of Rjrāśva who was deprived of his eye-sight by his father because the former gave a hundred rams to a she-wolf. The verses refer to some natural phenomena under a metaphorical garb and cannot be used for drawing the inference that a father could in law deprive his son of eye-sight at his will. The story of Sunahśepa (Ait. Br. 33) shows that in rare cases the father did sell his soṇ The passage of the Nirukta about the power to sell, to gift away or to abandon daughters has been already cited (p. 505). The Vaṣ Dh. S.¹⁷⁶ (XVII. 30-31) says 'Sunahśepa is an example of the son bought' (one of the twelve kinds of sons). The same sūtra (XVII. 36-37) defines the apaviddha kind of son as one, who being cast off by his parents, is accepted (as a son) by another. Manu IX. 171 also defines the 'apaviddha' in the same way. Vaṣ Dh. S. (XV. 1-3) propounds the absolute power of the parents over their children in the words 'man produced from seed and uterine blood springs from the father and the mother; (therefore) the parents have power to give, to sell or to abandon him; but one should not give nor accept an only son'. Here Vaṣ states the right of patria potestas several centuries before Justinian, who makes the vain boast (Institutes, Lib. I. Tit. IX. 2) that 'no other people have a power over their children such as we have over ours'. Manu (VIII. 416) and the Mahābhārata ¹⁷⁷ (Udyoga 33. 64) both state that the wife, the son and the slave are without wealth and that what-ever they acquire belongs to him whose they are. Manu (in V. 152) says that 'gift (by the father of the bride) is the source of (the husband's) ownership (over her)'. But gradually the rigour of the father's power was lessened by other competing considerations such as the ideas that the son was the father himself born again, and that the son conferred great spiritual benefit on the souls of the father and his ancestors by the balls of rice offered in śrāddha. So gradually the father's power over the son became restricteḍ Kauțilya (III. 13) gives the interesting information that mlecchas incur no blame by selling or pledging their children, but an ārya cannot be reduced to the state of slavery. Yāj. II. 175 and Nārada (dattāpradānika 4) both forbid the gift of one's son or wife. Kātyāyana ¹⁷⁸ says that though the father has powers of control over the wife and the son, he has not the power to sell or make a gift of his soṇ Yāj. (II. 118-119) modified the rule about the son's acquisitions also. Manu VIII. 389 prescribes a fine of 600 paņas for abandoning one's mother, father, wife or son when they are not sinnerṣ Vide Yāj. II. 237, Viṣṇu Dh. S. V. 113-114, Kauțilya III. 20 (p. 199) for a similar provisioṇ Manu (VIII. 299-300) restricted a man's power to award corporal punishment for misconduct to his wife, son or slave to striking with a rope or a thin piece of bamboo.

One question discussed by dharmasūtra writers is whether one has ownership over one's wife and childreṇ In Jaimini ¹⁷⁹ (VI. 7. 1–2) it is decided that in the Viśvajit sacrifice where one is to give away all that one has, one cannot give away one's parents and other relatives as one can make a gift only of what one is master of. The Mit. on Yāj. II. 175 says ‘though one cannot make a gift of one's wife or child to another, one is still owner of theṃ’ The Vīramitrodaya ¹⁸⁰ (vyavahāra p. 567) is of the same opinioṇ On the other hand the Tantraratna (of Pārthasārathimiśra) says that the word gift with reference to the son and the like is used only in a secondary sense viz. that of passing to another the power of control over the son Ch. IX

or the daughter. The Vyavahāramayūkha is also of the same opinioṇ

A few words may be said about infanticide. Westermarck in his ‘Origin and Development of Moral Ideas’ vol. I. (1906) pp. 393-413 furnishes an exhaustive account of this practice in ancient and modern times among barbarous and civilised communities from various countries, e. g. in Sparta (for the purpose of securing strong and healthy fighters), among the Rajputs (from family pride and fear of the crushing burden of expenses at marriage).¹⁸¹ He is wrong in saying that in the Vedic times infanticide or exposure of children was practiseḍ Ṛg. II. 29. 1 ¹⁸² is of no use on this point; it says ‘cast off from me sin as a woman who secretly gives birth to a child (casts it off).’ This is not a reference to infanticide of children born in wedlock, but refers to the exposure of a child by an unmarried woman which is clandestinely practised everywhere and as regards which even in England a very lenient attitude is shown as manifested by the passing of the Infanticide Act (12 and 13 Geo. V. Chap. 18). The most important passage on which some European scholars like Zimmer and Delbrück rely upon for this proposition is Tai. S. VI. 5. 10. 3 ‘They go to the arabhṛtha¹⁸³ (the final sacrificial bath); they keep aside the sthalis (pots) and take up the vessels for vāyu : therefore they (the people) keep aside the girl when she is born and lift up ( i. e. greet with pride and joy ) the son’. This simply refers to the fact that a daughter was not greeted as much as the soṇ It has nothing to do with exposure or infanticide. That passage only expresses the sentiment contained in the Ait. Br.¹⁸⁴ (33. 1) 'the wife is indeed a friend, the daughter is distress (or humilia- tion), the son is light in the highest heaven'. The Mahābhārata (Ādi. 159. 11) in a similar vein says 'the son is one's self, the wife is one's friend, but the daughter is indeed a difficulty'. But all the same the Gṛhyasūtras like Āp. (15. 13) call upon the father to greet his daughter also with a mantra when he returned from a journey, the difference being that in the case of the son there is kissing of the son's head and muttering of certain mantras in his right ear. Manu (IX. 232) advises the king to award death sentence to him who kills a woman, a child or a brāhmaṇa. Manu IX. 130 (=Anuśāsana 45. 11) expressly says that 'just as the son is one's own self, so is the daughter like the son; how can another person take the wealth (of the deceased) when his daughter who is his own self is there to take it'? Nārada¹⁸⁵ (dāyabhāga v. 50) reasons that both the son and the daughter continue the line of the father and therefore in the absence of the son the daughter suc- deeds as heir. Bṛhaspati exclaims 'the daughter is born from the limbs of the parents as much as the son; when she is alive how can another take her father's estate'? Bāṇa, who as a great poet had unparalleled sympathy with the emotions and feelings of the average man, makes even the king Prabhākara- vardhana say about his daughter what every Indian father has felt for thousands of years 'this rule of law laid down by some one viz. that one's own children (daughters) sprung from one's body, fondled on one's knees and whom one would never forsake, are taken away all of a sudden by persons (husbands) who till then were quite unfamiliar. It is on account of this sorrow that although both (son and daughter) are one's own children the good feel sorrow when a daughter is born and who offer water in the form of tears to their daughters at the very time of their birth'. A daughter was not greeted at birth, not because the father had no love for her, but because he felt that a daughter meant a source of anxiety all her life to hiṃ It was the great concern felt for a daughter's well-being in life and her character that made the parents anxious that no daughter be born to theṃ Society expected a very high moral rectitude from women and treated lapses of men with leniency. This is finely expressed in the Rāmāyaṇa (Uttara-kāṇḍa 9. 10–11). Ancient literature did not everywhere treat women with scorn and contempt. It has already been shown how highly the wife was regarded even in the most ancient days as a man's half. Ṛg. III. 53. 4 speaks of the wife as a haven of rest (jāyed-astam). The Chāṇ Up.¹¹⁹¹ looks upon the sight of a woman in a dream as very auspicious and as prognosticating success in religious rites already undertakeṇ Manu (III. 56 = Anuśāsana 46. 5), though he has said, as will be shown later on, some very hard things about women, was not unmindful of the honour due to them and says in a chivalrous spirit 'where women are honoured there the gods love to reside; where they are not honoured, there all religious acts come to nought.' Maidens were regarded as pure (vide p. 296 above) and auspiciouṣ When the king passed through his capital it was customary to greet him with fried grain showered by maidens (Raghuvamśa II. 10). The Śaunaka-kārikā¹¹⁹² (mṣ in Bombay University Library folio 22 b) includes a maiden among the eight objects which were auspiciouṣ The Droṇaparva (82. 20–22 ) mentions numerous objects which Arjuna looked at and touched as auspicious when starting for battle, among which well-decked maidens are mentioneḍ Gobhila-smṛti II. 163 says that one that sees on rising from bed in the morning among others a woman whose husband is living is free from all difficultieṣ The Vāmana-purāṇa (14. 35–36) mentions several objects which are auspicious when one is about to leave home, among which figure brāhmaṇa maidenṣ Vide Sṃ C. I. p. 168.

Some remarks must be made about the times auspicious for marriage. In the marriage hymn (Ṛgveda ¹¹⁹³ X. 85. 13) the words occur 'the cows are killed on the Aghās and (the bride) is carried away (from her father's house) on the Phalgunis'. The cow was killed in Madhuparka which was offered to the bridegroom on the day of marriage. Or this may be a reference to the giving of cows by the bridegroom to the bride's father (as in the form later called ārṣa). So it appears probable that this is a reference to marriage being performed on the day when the moon was in conjunction with the constellation of Aghās (i. e. Maghā). The two Phalgunis follow immediately after the Maghā nakṣatra. There is an echo of this in the Āp. gr. III, 1-2 which says 'cows are accepted and (the bride) is carried (to the bridegroom's house) on the Phalgunis'. This means that the marriage (probably in the ārṣa form) is celebrated on the Maghās and the bride goes from her father's house on the next day after marriage or after one day more. The Āśv. gr. (I. 4. 1) says ¹⁸⁹ that 'in the northward passage of the sun, in the bright half of a month and on an auspicious lunar mansion, caula, upanayana, godāna and marriage are to be performed and that according to some teachers marriage may be celebrated at all times' (not necessarily in northward passage &c). The Āp. gr. (2. 12-13) prescribes that all seasons except the two months of Śiśira (i. e. Magha and Phālguna) and the last of the two months of summer (viz. Āṣādha) are fit for marriage and all nakṣatras which are declared to be auspicious (or holy). Āp. gr. further ¹⁹⁰ adds (3. 3) 'a father who desires that his daughter should be dear (to her husband) should give her in marriage on the Niṣtyā (i. e. Svāti constellation); thus she becomes dear (to her husband), she does not return (to her parent's house); this is a procedure based upon the Brāhmaṇa passage.' The Brāhmaṇa passage is Tai. Br. I. 5, 2. The Bauḍ gr. (I. 1. 18-19) is the same ¹⁹¹ as Āp. gr. about months and adds that the nakṣatras for marriage are Rohini, Mṛgaśīrsa, Uttarā Phalguni, Svāti; while Punarvasu, Tiṣya (Puṣya), Hasta, Śravaṇa and Revati are suitable for other ceremonies of an auspicious character. The Mānava gr. (I. 7. 5) says that the nakṣatras Rohiṇī, Mṛgaśiras, Śravaṇa, Śraviṣṭhā (Dhāniṣṭhā) and the Uttarās (i. e. Uttarāṣāḍhā, Uttarā Phalguni and Uttarā Bhādrapadā) are fit for marriage and for taking the bride from her parent's house and whatever other (nakṣatra) is declared to be auspiciouṣ Kāthaka gr. 14. 9-10 and Vārāha gr. 10 are similar. The Rāmāyaṇa¹⁹² (Bālakāṇḍa 72. 13 and 71. 24) speaks of marriage being per- formed on Uttara Phalguni, of which Bhaga is the deity. The Mahābhārata also speaks of marriage on a nakṣatra presided over by Bhaga (Ādiparva 8. 16). The Kauśika sūtra (75. 2-4) makes an approach to modern practice when it prescribes that marriage should be celebrated after the full moon of Kārtika and up to the full moon of Vaiśākha or one may do as one likes but should avoid the month or half month of Caitra.¹⁹³

Medieval digests introduce many detailed rules derived from astrology which it is not possible to set out here. A few only will be indicateḍ The Udváhatattva (p. 124) quotes Rājamārtaṇḍa¹⁹⁴ and Bhuja-balabhīma to the effect that all months are auspicious for marriage except Caitra and Pauṣa and that when a girl is very much grown-up one should not wait for auspicious seasons, but she should be given away on any day when the moon and the zodiacal sign rising at the moment of marriage are favourable, and that one should go into questions of auspicious ayana, month, day &c. up to only the tenth year of girlṣ The Saṁskāraratnamālā (p. 460) says that as there is conflict among the dicta of sūtra and smṛti writers about months one should follow the usage of one's country. One should not celebrate the marriage of the eldest son with the eldest daughter of a man in the month of Jyeṣṭha (or on Jyeṣṭhā nakṣatra) nor should one celebrate a marriage in the month of birth or on the day of birth or nakṣatra of birth (of the eldest son or daughter). Wednesday, Monday, Friday and Thursday are the best days of the week, but the Madanapārijāta says that any day is good if the marriage is celebrated at night.¹⁹⁵ In marriage the moon must be in a strong position for girlṣ One should avoid Jupiter being the 4th, 8th or 12th from one's rāśi (zodiacal sign in which the moon was at the time of birth). The Nirṇayasindhu quotes a verse that when the girl has reached puberty, then one should not wait till Jupiter is favourable, but marriage should be performed even when Jupiter is 8th from the zodiacal sign of birth, propitiatory rites however being performed for Jupiter.¹⁹⁶ Upanayana and marriage cannot be performed when Jupiter is in the zodiacal sign called Leo (Siṁha), but this applies only to the tract between the river Godāvarī and the Gangeṣ¹⁹⁷ From the nakṣatra and zodiacal sign at the time of birth (in the case of both the girl and the bridegroom) certain astrological calculations were made in eight ways¹⁹⁸ called kūṭaṣ They were Varṇa, Vaśya, Nakṣatra, Yoni, Graha (planets governing the twelve rāśis), Gaṇa, Rāśi, and Nāḍī, each later one of which was more powerful than each preceding one and marks (guṇa) from 1 to 8 were respectively assigned to each of these if the conditions laid down were satisfieḍ Two of these viz. gaṇa and nāḍī are attached great importance even now among brāhmaṇas and other classes also and so they may be illustrated here, though very briefly. The 27 nakṣatras are arranged in three groups of nine each, each group being assigned to Deva-gaṇa, Manusyagaṇa and Rākṣasagaṇa as follows: | Devagaṇa | Manusyagaṇa | Rākṣasagaṇa | | :--------------- | :------------------ | :---------------- | | Aśvinī | Bharaṇī | Kṛttikā | | Mṛgaśiras | Rohiṇī | Āśleṣā | | Punarvasu | Ārdrā | Maghā | | Puṣya | Pūrvā | Citrā | | Hasta | Uttarā | Viśākhā | | Svāti | Pūrvāṣāḍhā | Jyeṣṭhā | | Anurādhā | Uttarāṣāḍhā | Mūla | | Śravaṇa | Pūrvabhādrāpadā | Dhaniṣṭhā | | Revatī | Uttarabhādrāpadā | Śatatārakā |

If the bride and bridegroom are born on nakṣatras that belong to the same class out of these three groups it is the best thing but if their nakṣatras of birth belong to different groups, then the rules are: it is middling if their nakṣatras belong to the deva group or the human group respectively, or if the bridegroom's nakṣatra being of the devagaṇa or rākṣasagaṇa, the bride's is of the manuṣyagaṇa, while if the nakṣatra of the bride is of the rākṣasa group and the bridegroom's of the manuṣya type, then death would result. Similarly if the nakṣatras of the pair respectively belong to the deva and rākṣasa groups, there would be quarrels and enmity between the two.

For the purpose of nāḍī the nakṣatras are divided into three groups of nine each, ādyanāḍī, madhyanāḍī and antyanāḍī, as follows:

| Ādyanāḍī | Madhyanāḍī | Antyanāḍī | | :---------------- | :------------------ | :----------------- | | Aśvinī | Bharaṇī | Kṛttikā | | Ārdrā | Mṛgaśiras | Rohiṇī | | Punarvasu | Puṣya | Āśleṣā | | Uttarā | Pūrvā | Maghā | | Hasta | Citrā | Svāti | | Jyeṣṭhā | Anurādhā | Viśākhā | | Mūla | Pūrvāṣāḍhā | Uttarāṣāḍhā | | Śatatārakā | Dhaniṣṭhā | Śravaṇa | | Pūrvabhādrāpadā | Uttarābhādrāpadā | Revatī |

If the nakṣatras of the intended pair belong to the same nāḍī, then that portends death and so there is to be no marriage in such a case. The respective nakṣatras of birth in the case of the two should belong to different nāḍīṣ The anxieties of the girls' father did not end here. If after the marriage was settled but before it actually took place any relative in any of the two families died, then the marriage agreement was to be broken off according to some writers, but Saunaka¹⁹⁹ mercifully states a more sensible and reasonable rule viz. if the father or mother or paternal grand-father or paternal grand-mother or paternal uncle, brother or unmarried sister of the intended bride or bridegroom dies, or the bride groom's first wife or his son from another wife dies, then only it is pratikūla and the marriage should not be performed but the death of any one else presents no obstacle.

If before the rites of marriage begin (i. e. before the per formance of nāndiśrāddha ), the mother of the bride or of the bride-groom has her monthly illness then the marriage has to be postponed till she becomes pure (till the fifth day after the illness).

Forms of marriage: From the times of the gṛhya sūtras dharmaśūtras and smṛtis the forms of marriage are said to be eight, viz. Brāhma, Prājāpatya, Ārṣa, Daiva, Gāndharva, Āsura, Rākṣasa and Paiśāca (vide Āśv. gr. I. 6, Gaut. IV. 6-13, Bauḍ Dh. S. I. 11, Manu III. 21 = Ādi-parva 73. 8-9), Viṣṇu Dh. S. 24.18-19, Yāj. I. 58, Nārada (stripuiṁsa, verses 38-39), Kauṭilya III. 1, 59th prakaraṇa, Ādi-parva 102. 12-15 (they are described but not named) ; some of these arrange the first four differently e. g. Āśv. gr. arranges them as Brāhma, Daiva, Prājāpatya and Ārṣa, while Viṣṇu arranges them as Brāhma, Daiva, Ārṣa and Prājāpatya; Āśv. gr. I. 6 places Paiśāca before Rākṣasa. The Mānava gr. speaks of only two Brāhma and Śaulka (i. e. Āsura) probably because these two were the forms most current. Āp. Dh. S. (II. 5. 11. 17-20—II. 5. 12. 1-2) speaks of only six omitting Prājāpatya and Paiśāca; while Vāṣ Dh. S. I. 28-29 expressly says that there are only six forms of marriage viz. Brāhma, Daiva, Ārṣa, Gāndharva, Kṣātra and Mānusa (the last two being the same as Rākṣasa and Āsura). It is impossible for want of space to set out the various definitions of the several forms given by the several authorṣ There is general agreement on the special characteristics of each and it is sufficient to point out these as given in Manu III. 27-34. The gift of a daughter, after decking her (with valuable garments) and honouring her (with jewels &c.), to a man learned in the Vedas and of good conduct, whom the father of the girl himself invites, is called Brāhma. When the father gives away his daughter after decking her (with ornaments &c.) to a priest, who duly officiates at a sacrifice, during the course of its performance, it is the Daiva forṃ²⁰⁰ When there is a gift of one’s daughter, after taking one pair of cattle (a cow and a bull) or two pairs only as a matter of fulfilling the law (and not as a sale of the girl ), that is named the ārṣa forṃ The gift of a daughter, after the father has addressed (the couple with the words ‘may both of you perform your religious duties together’) and after he has honoured the bride-groom (with Madhuparka &c.), is declared to constitute the Prājāpatya forṃ Yāj. I. 60 calls this ‘kāya’, because in the Brāhmaṇa works ‘ka’ means ‘Prajāpati’. When the girl is given away at the father’s will after the bride-groom gives as much wealth as he can afford to pay to the relatives of the girl and to the girl herself, that is called the Asura forṃ The union of a girl and the bride-groom by their mutual consent is known as Gāndharva, which springs from the passion of love and has intercourse as its purpose. The forcible abduction of a maiden from her house, while she weeps and cries aloud, after her kinsmen have been slain (or beaten ), wounded and (their houses or fortresses) are broken open, is called the Rākṣasa forṃ When a man has intercourse with a girl stealthily while she is asleep or intoxi- cated or disordered in mind (or unconscious), that is the Paiśāca form, which is the basest and the most sinful of all formṣ

In the first four forms there is the gift of the girl (kanyādāna) by the father or other guardian to the bride-grooṃ The word ‘dāna’ here is used in a secondary sense (as stated above on p. 504) viz. in the sense of transfer of the father’s right of guardianship and control of the maiden to the husbanḍ All gifts are to be made with water in the case of brāhmaṇas as stated by Manu ( III. 35 ) and Gaut. V. 16-17.²⁰¹ Similarly in all the four forms where there is kanyādāna the girl is to be well dressed and decked with ornamentṣ The essence of the brāhma form is that the girl is given without receiving anything from the bride-groom, who is invited and honoured by the girl's father. It is called brāhma either because brahma means the holy Veda and this mode being sanctioned by the most ancient texts it is the holiest and best form, or brahma means dharma and being the best of all forms it is called brāhma ( vide Smṛtimuk- tāphala part I. p. 140 ). In the ārṣa form a pair of cattle is received from the bride-groom and it is somewhat inferior to the brāhma forṃ But the pair is given not as a price ( i. e. there is no purchase ), but because that is one way of effecting a marriage laid down by the śāstras ( vide the passage quoted above at p. 504 'therefore one should give a hundred cows &c' and Āp. Dh. S. II. 6. 13. 11, which is opposed to Vaṣ Dh. S. I. 36 ) and the gift is made to show one's regard for the girl ( vide Manu III. 53-54 ). The daiva form is only appropriate to brāhmaṇas as only²⁰² a brāhmaṇa could officiate at a sacrifice. It is so called because the bridegroom is engaged at the time in rites for the gods and it is inferior to the brāhma inasmuch as in it there is some trace of benefit to the father ( that the priest may do his best in the rite for the gods ). Govindasvāmin on Bauḍ Dh. S. says 'the bride is in this form part of the sacrificial fee'. In all forms of marriage the husband and wife have to perform all religious acts together, as forcibly put by Āp. Dh. S.²⁰³ ( there is no separation between husband and wife, since from the time of taking the hand of the wife there is joint perform- ance of all religious actions ). But in the Prājāpatya the words used indicate according to the commentators ( vide Sāṃ Pr. p. 852 and Haradatta on Gaut. IV. 5 ) one or more of several things, viz. that the husband will remain a house-holder all his life and will not become a recluse while the wife is living or that he will not marry another wife i. e. it will be a strictly monogamous marriage which is defined in Hyde v. Hyde (1866), P. and M. p. 130 as ‘ the voluntary union for life of one man and one woman to the exclusion of all others ’ ( p. 133 ), or that he will associate the wife with himself not only in sacrifices but also in works of charity ( like building tanks, wells &c. ). This form is inferior to brāhma in that the father, as it were, makes a special stipulation with the bride-groom, while in the brāhma there is no such special stipulation, but the bride-groom promises that he will not break faith with his wife in the matter of the three puruṣārthas, dharma, artha, kāma.²⁰⁴ In the āsura form there is practically a sale of the girl for money or money’s worth and so it is not approveḍ The ārṣa is distinguished from āsura in this that in the latter there is no limit to what is taken from the bride-groom, while in the former a pair of cattle is offered as a matter of forṃ In the gāndharva there is no gift by the father; on the contrary the father’s authority is set at naught by the girl for the time being. Marriage was a sacrament according to ancient sages, and its principal purposes as stated above (pp. 428–29) were the acquisition of merit by the performance of religious duty and of progeny. In the gāndharva form the principal object was gratification of carnal desires and so it is held inferior to the first four forms and is disapproveḍ This kind of marriage is so called because it is prompted by mutual love and the Gandharvas were known to be libidinous, as the Tai. S. VI. 1. 6. 5 (strīkāmā vai Gandharvāḥ) and Ait. Br. V. 1 state. In this form at least the girl’s feelings are consulteḍ In the rākṣasa and the paiśāca, there is no gift by the father and both are or may be against the wish of the girl. The forcible carrying of the girl is the essence of the rākṣasa (even if there is no fight because the girl’s father takes no retaliatory steps in fear of the abductor’s strength). It is called rākṣasa because rākṣasas (demons) are known from legends to have been addicted to cruel and forceful methodṣ The paiśāca is so called because in it there is action like that of piśācas (goblins) that are supposed to act stealthily by night. It is not to be supposed that when ancient sages held that rākṣasa and paiśāca were forms of marriage, they legalized marriage by capture or stealth. What they meant was that these were the means of securing ²⁰⁵ wives and that there are not really eight kinds of vivāhas, but rather there are eight ways in which wives may be secureḍ It is for this reason that Vatsa says that if a fine girl cannot be secured by any means she may be approached even in private by stealth and marrieḍ The sages condemned in no measured terms the paiśāca. From the fact that Āpastamba and Vasiṣṭha both ignore the Paiśāca and Prājāpatya it may be inferred that these two had ceased to be recognised by their time and that the other sages enumerated them only because they appeared in ancient works and for the sake of completeness of treatment. Vasiṣṭha (17. 73) expressly repudiates the idea that legal wifehood can arise by forcible seizure of a girl 'if a damsel ²⁰⁶ has been abducted by force and not wedded with sacred texts she may lawfully be given to another man (in marriage); she is just as good as a maiden'. The smṛtis out of regard for the future welfare of the girl preferred to blink at the wrong done, but insisted upon the abductor or stealthy seducer performing the rites of homa and saplapadī in order to confer on the girl wronged the status of a legally married wife. But if the girl may be given to another in marriage and pronounced very heavy punishments for the abductor or seducer of a girl (vide Manu VIII. 366 and Yāj. II. 287-288). Sir Gooroodas Banerjee ²⁰⁷ rightly expresses surprise at Macnaughten's saying that fraud was legalised by the Hindu Law in the case of the paiśāca forṃ Manu VIII. 366 says that if a man has inter-course with a girl of his own caste with her consent, he will have to give a fee to the father if the latter so desires and Medhātithi²⁰⁸ adds that if the father does not desire money the lover will have to pay a fine to the king, that the girl may be given to him or if she has lost her love for him, she may be given to another and if the lover himself wants to discard her he should be forced to accept her (as a wife). Nārada (strīpuṃsa, verse 72) similarly says that if a man has sexual intercourse with a maiden who is a consenting party, then there is no offence, but he must marry her after decking her (with ornaments) and treating her with honour.

The Sṃ C. and other digests state that in the gāndharva, āsura, rākṣasa and paiśāca homa and saptapadī are necessary and they quote Devala and the Gṛhyapariśiṣṭa in support. The Mahābhārata (Ādi. 195. 7)²⁰⁹ expressly says that even after svayaṃvara religious rites had to be performeḍ Kālidāsa in Raghuvamśa VII describes how after the svayaṃvara of Indumatī the principal religious rites of madhuparka, homa, going round the fire, pāṇigrahaṇa took place. Since Āśvalāyana first speaks of eight forms and then prescribes the performance of homa and saptapadī, he implies that these are necessary in all formṣ

The smṛtis contain several views about the suitability of these eight to various varṇaṣ All are agreed that the first four, brāhma, daiva, ārṣa and prājāpatya, are the approved forms (praśasta or dharmya). Vide Gaut. IV. 12, Āp. Dh. S II.5.12. 3, Manu III. 24, Nārada (stripuṃsa, verse 44) &c. Most say that each preceding one out of the first four is superior to each succeeding one and that thus brāhma is the best (vide Āp. Dh. S. II. 5. 12. 4, Bauḍ Dh. S. I. 11. 11). Almost all are agreed that paiśāca is the worst. Manu III. 23-26 refers to several viewṣ One view is that the first four (brāhma &c.) are the proper forms for brāhmaṇas (Bauḍ Dh. S. I. 11. 10, Manu III. 24). Another view was that the first six (out of the eight i. e. all except rākṣasa and paiśāca) are allowed to brāhmaṇas and the last four to kṣatriyas, and the gāndharva, āsura and paiśāca to vaiśyas and śūdras (Manu III. 23). A third view was that prājāpatya, gāndharva and āsura may be resorted to by all varṇas and paiśāca and āsura should not be resorted to by any one of any varṇa (Manu III. 25 = Anušāsana 44. 9–10), but in another verse (III. 24) Manu allows āsura to vaiśyas and śūdras²⁰⁹. Manu mentions the view that the gāndharva and rākṣasa are proper (dharmya) for a kṣatriya or a mixture of these two viz. where the girl loves the bride-groom, but her parents or guardians disapprove or cause obstacles and the lover takes away the girl after a fight with her relations (Manu III. 26 and Bauḍ Dh. S. I. 11. 13). Bauḍ Dh.S.¹²¹⁵ (I. 11. 14–16) recommends āsura and paiśāca to vaiśyas and śūdras and assigns the interesting reason ‘ for the vaiśyas and śūdras do not keep their wives under restraints, they having to do the work of ploughing and waiting upon (the other varṇas) ’. Nārada (strīpuṁsa, verse 40) says that gāndharva is common to all varṇaṣ The Kāmasūtra (III. 5. 28) first speaks of the brāhma as the best (following the view of the dharmaśāstra writers) and then, true to its own particular subject, gives its own opinion that gāndharva is the best (Kāmasūtra III. 5. 29–30). The gāndharva form was very much in vogue among royal familieṣ In the Śākuntala,¹²¹⁶ Kālidāsa gives expression to this practice. In the Mahābhārata¹²¹⁷ Kṛṣṇa says to Arjuna who was in love with Subhadrā that carrying away by force one’s lady-love is commended in the case of valiant kṣatriyaṣ In the Sanjan plates of Amoghavarṣa (dated śake 793) it is stated that Indrarāja married the daughter of the Cālukya king at Kaira by the rākṣasa form of marriage (E. I. vol. 18, p. 235 at p. 243). Another and a very famous historic example of the rākṣasa form is the forcible abduction after a most dashing and valiant fight by Pṛthvīrāja Cohan of the daughter of Jayacandra, the king of Kanoj.²¹⁴ It is said that the daughter of Jayacandra was a consenting party; in that case this would be a mixture of the two forms of gāndharva and rākṣasa (compare Manu III. 26 ). The svayainvara very often spoken of in the dharmaśāstras was practically the gāndharva as stated by the commentary Viramitrodaya.²¹⁵ It had several varietieṣ The simplest form of svayāmvara occurs when, as described by Vaṣ Dh. S. 17. 67-68, Manu IX. 90, Bauḍ Dh. S. IV. 1. 13, a girl, who attains puberty and whose father does not find a proper husband for three years, herself seeks her husband ( or after three months from puberty according to Gaut. 18. 20, Viṣṇu Dh. S. 25. 40-41 ). Yāj. I. 64 also recommends svayamvara to every girl when there is no parent or other guardian who can find out a worthy husband for her. When a girl chose her own husband as above she had to return all ornaments given to her by her parents or brother and the husband who married her had to pay no śulka ( dowry ) to the father as the latter lost his power over her by not giving her away in time ( vide Gaut. 18. 20 and Manu IX. 92 ). This simple svayamvara was applicable to girls of all casteṣ Sāvitrī indulged in this kind of svayamvara, when she went about in a chariot to find out a suitable husband for herself. But the svayamvaras described in the two great epics are often most elaborate and spectacular affairs and were confined to royal fainilieṣ The Ādiparva says that kṣatriyas commend svayamvara and resort to it, but they prefer a girl who is carried off after subduing her relativeṣ Bhīṣma carried off the three daughters of the king of Kāśī and got two of them ( Ambikā and Ambālikā ) married to his ward Vicitravīrya.²¹⁶ The svayamvara of Sītā or Draupadī did not depend upon the will of the bride, but the bride was to be given in marriage to whomsoever showed a certain skill as a warrior. In the case of Damayantī it was a real choice of her husband by her ( Vanaparva 54 ff. ) though she chose Nala in a vast and splendid assembly of royal suitorṣ Kālidāsa also gives us a fine description of the svayamvara of Indumatī in the Raghuvamsa. Bilhaṇa in his Vikramāńkadevacarita (canto IX) gives a description of the historic svayamvara of Candralekhā (or Caṇḍaladevi) daughter of the Śilāhāra prince of Karahāṭa (modern Karad), when she chose Āhavamalla or Vikramāńka, the Cālukya king of Kalyāṇa (latter half of 11th century). Such a svayamvara, it appears, was thought to be unsuitable to brāhmaṇas according to the Ādiparva.²¹⁷ In the Kādambarī (Pūrvabhāga, penultimate para) Patralekhā says that svayamvara is ordained in the Dharmaśāstraṣ²¹⁸

The Āp. Dh. S. II. 5. 12. 4 makes a general statement that the progeny of a couple partakes of the character of the form in which they were married ²¹⁹ (i. e. if the marriage is in the best or in an approved form the son is good; if the marriage is in a condemned form, the son bears a low character). Manu (III. 39–42) expands the same idea by stating that sons born of marriages in the brāhma and other three forms are full of spiritual eminence and are endowed with beauty, virtues, wealth, fame and very long life, while sons of marriages of the other four forms are cruel, are liars, haters of the Veda and of dharma. Some sūtras and smṛtis state how many generations are rendered holy by a son born in one of the first four formṣ For example, Āśv. gṛ. (I. 6) says that a son born of parents married in the brāhma, daiva, prājāpatya or ārṣa forms respectively brings purification to twelve descendants and twelve ancestors on both sides (i. e. his father’s and mother’s), to ten descendants and ten ancestors on both sides, to eight descendants and ascendantṣ Manu (III. 37–38) and Yāj. (I. 58–60) put the matter somewhat differently. According to them the son of a brāhma marriage brings purification to ten paternal ancestors, to ten male descendants and to himself (in all 21), of the daiva marriage to seven paternal ancestors and seven male descendants, of the prājāpatya marriage to six male ancestors, to six male descendants and to himself (in all 13), of the ārṣa marriage to three male ancestors and three male descendantṣ Gaut. (IV. 24-27) has similar provisionṣ Commentators like Viśvarūpa and Medhātithi explain that these verses are not to be taken literally. They merely praise the extreme desirability of the brāhma forṃ Viśvarūpa, however, alternatively proposes that one may accept the texts of the holy sages as literally true, following the dictum of Śabara that there is nothing too heavy (or impossible) for a holy text.²²⁰ One may laugh at these texts about the virtues of the several forms of marriage, but they are really intended to emphasize the high importance to the future of the race and to society of noble ideals of marriage, of morals and of a decent and peaceful mode of life. The forms of marriage have their roots deep down in the Vedic Literature. Rg. X. 85 gives expression to a marriage in the brāhma form (there is kanyādāna and so forth). The āsura form (by payment of money) is referred to in Rg. I. 109. 2 and Nirukta VI.9 (quoted above in note 1175). The gāndharva form or svayaṁvara is indicated by Rg. X. 27. 12 (quoted above p.439) and Rg. I. 119. 5. The story of Śyāvaśva narrated in the Bṛhad-devatā (V. 50) in connection with Rg. V. 61 makes an approach to the daiva forṃ It is related in the Bṛhad-devatā that Ātreya Arcanānas who officiated as a priest at a sacrifice for king Rathāviti asked for his son Śyāvaśva the hand of the king's daughter. In modern times two forms are in vogue, the brāhma and āsura. In the brāhma form it is a gift of the girl pure and simple; in the āsura form it is like the sale of the bride for pecuniary consideration paid to the father or other guardian for his benefit. If this element of pecuniary consideration paid to the guardian exists, its effects cannot be undone by the form of a gift being gone through.²²¹ The gāndharva form is said to be obsolete now, yet in some cases before the courts it has been held that it is still in vogue.²²² There can be no gāndharva if the girl is a minor. Further if a widow remarries, that marriage will be ordinarily regarded as gāndharva, because there will be no gift of a kanyā ( as she is a widow ) and because she herself will generally arrange such a marriage.

Before proceeding to set out in detail the ceremonies of marriage, it would be best to analyse the contents of Ṛgveda X. 85, which is a marriage hymn redolent of the highest ideal of marriage and conjugal felicity and several verses of which are recited even now in the marriage riteṣ The hymn refers to a mythical marriage of Sūryā, daughter of Savitṛ, with Soma and the important features of the marriage, though not arranged in a regular sequence in the hymn, are : the two Āśvins went to ask for Sūryā as a bride for Soma ²²³ ( verses 8-9) ; Savitṛ agreed to give her ( v. 9), the bride-groom was treated with honour, presents were made to him and cows were killed for (or presented to) him ; Soma took hold of her hand with the verse (36) ‘ I ²²⁴ take thy hand for prosperity (or love) so that you may grow to old age with me thy husband; the gods, Bhaga, Aryaman, Savitṛ, the wise Pūṣan have given thee to me for performing the duties of a house-holder ’; the bride is a gift by her father in the presence of gods and the fire (v. 40–41) ; the girl passes from the dominion and control of her father and becomes united with her husband (v. 24) ; the bride is best as follows: ‘ may you stay here together, may you not be separated, may you compass all life (long life), happy in your own house and playing with your sons and grandsons; O Indra ! make her endowed with worthy sons and prosperity; bestow on her ten sons and make her husband the eleventh ( male ) ; may you be queen over your father-in-law, mother-in-law, over the husband’s brother and sister (vv. 42, 45-46)’. ²²⁵ Among subordinate items it is noteworthy that Raibhyā was sent along with Sūryā as her friend ( anudeyī ) to make her time not hang heavy on her handṣ (v. 6) when she first went to her husband's family (just as in modern times in Western India some woman accompanies the bride for a few days as pātharākhiṇ ‘one who guards’), that even a female servant accompanied Sūryā.²²⁶

In connection with the rites of marriage it is necessary to observe that the greatest divergence prevailed from very ancient timeṣ Āśv. gr. (I. 7. 1–2) says ‘various indeed are the observances of ( different ) countries and villages; one should follow those in marriages ; what, however, is common ( to all or most ) that we shall declare’. Similarly Āp. gr. (2. 15 ) declares ‘people should understand from women ( and others ) what procedure is ( to be observed according to custom )’ and the commentator Sudarśanācārya notes that certain rites like the worship of planets, aṅkurāropana and the tying of pratisara (the marriage string or ribbon round the wrist ) are usual and are performed with Vedic mantras, while others like Nāgabalī, Yakṣabalī and the worship of Indrāṇī are performed without Vedic mantraṣ The Kāthakagṛhya 25. 7 allows usages of countries and families to be observed in marriage and the commentators mention several such usageṣ As the gṛhya-sūtra of Āśvalāyana contains perhaps the shortest account of marriage rites and as that sūtra is probably the most ancient among the gṛhya sūtras I shall set out below the entire ceremony of marriage from that sūtra. Then a few important details from other gṛhya sūtras will be added and it will be pointed out how in modern times a marriage is celebrated, particularly among higher classeṣ It must be remembered that there is not only great divergence as to the number of separate ceremonies that constitute the saṁskāra of marriage, but the sequence of even the most important ceremonies is different in the several sūtras and the mantras also are different (though some like Rg. X. 185. 36 are common to all ). It is remarkable that out of the 47 verses of Rg. X. 85, the Āp. Mantrapāṭha employs as many as 29 (most of which are quoted in connection with marriage ). The main outlines of the marriage saṁskāra show a remarkable continuity for several thousand years from the times of the Ṛgveda down to modern timeṣ

The Āśv. gr. (I. 7. 3–I. 8. ) deals with all the marriage rites as follows^† : Having placed to the west of the fire ( that is kindled on the altar as described already) a mill-stone and to the north-east (of the fire) a water jar, he should offer sacrifice (with the sruva), while the bride takes hold of him (touches his right hand). Standing with his face turned to the west, while the bride is sitting and has her face turned to the east, he should seize only her thumb with the mantra ‘I take hold of thy hand for happiness’ (Rg. X. 85.36 quoted in note 1228), if he desires that only male children be born to him; he may seize her other fingers if he is desirous of female children; he may seize her hand on the hair side together with the thumb if he be desirous of both (male and female children). Leading her thrice round the fire and the water jar so that their right sides are turned towards (the fire &c.) he murmurs ²²⁷ ‘I am ama (this), thou art sā (she), thou art sā, I am ama; I am heaven, thou art the earth; I am the sāman, thou art the ṛk. Let us both marry here. Let us beget offspring. Dear to each other, bright, having well disposed minds, may we live for a hundred years!’ Each time he leads her round (the fire) he makes her tread on the stone with (the words) ‘tread on this stone; be firm like a stone; overcome the enemies; trample down the foes’. ²²⁸ Having first poured clarified butter over her joined hands, the bride’s brother or some one who is in the place of brother pours fried grain twice over the bride’s joined hands, three times in the case of those whose gotra is Jamadagni (i. e. if the bridegroom is of that gotra). Then he pours clarified butter over what has been left of the havis (sacrificial material or offering) and over what has been cut off (separated from the aggregate). This is the rule about the portions to be cut off (in every case where there is avadāna). ²²⁹ With the following verses (recited by the bride-groom) viz. 'to god Aryaman the girls offered sacrifice; may he the god Aryaman loosen her from this (i. e. her father's) and not from that place (the bride-groom's), svāhā! to god Varuna the girls have offered sacrifice; may he, god Varuna &c. To god Pūṣan the girls have offered sacrifice, to Agni; may he, god Pūṣan &c. ; with these (the bride) should sacrifice (the fried grain) without opening her joined hands, as if they were the spoon called sruc. Without going round the fire the bride sacrifices fried grain a fourth time silently with the neb of the śūrpa (winnowing basket) towards herself. Some²³⁰ lead the bride round each time after fried grain has been poured out, so that the two last oblations do not follow immediately after each other. Then he loosens her two locks of hair if they are made up (that is if her hair has been bound round on two sides with two tufts of wool), (the right one) with the mantra 'I release thee from the fetter of Varuna' (Rg. X. 85. 24) and the left one with the following mantra (Rg. X. 85. 25). Then he causes her to step forward in a north-eastern direction seven steps with the words 'may you take one step for sap, second step for juice (or vigour), third step for the thriving of wealth, fourth step for comfort, fifth step for offspring, sixth step for seasons, may you be my friend with your seventh step! May you be devoted to me; let us have many sons, may they reach old age'.²³¹ Bringing the heads of the two (bride and bride-groom) together, he (the ācārya) sprinkles their heads with water from the water jar. And the bride should dwell that night in the house of an old brāhmaṇa woman whose husband and children are alive.²³² When she (the bride) has seen the Pole star, the star Arundhatī and the seven sages (the constellation of Ursa Major), let her break her silence and say ‘may my husband live and may I secure offspring’.²³³ If (the newly married couple) have to make a journey (to their home in another village), let him cause the bride to mount a chariot with the verse ‘may Pūṣan lead thee from here holding thy hand’ (Rg. X. 85. 26); he should make her ascend into a boat with the hemistitch ‘carrying stones (or the river called Aśmanvatī) flows; get ready’ (Rg. X. 53. 8) and let him make her descend from the boat with the following hemistitch; if she weeps let him pronounce the verse ‘they weep for the living’ (Rg. X. 40. 10). They constantly carry the nuptial fire in front. At pleasing places, trees and cross roads, let him mutter ‘may no way-layers meet us’ (Rg. X. 85. 32). At every dwelling place (on the way) let him look at the onlookers with the mantra ‘this newly married bride brings good luck’ (Rg. X. 85. 33). He should make her enter his house with the verse ‘here may happiness increase unto you through offspring’ (Rg. X. 85. 27). Having kindled with fuel sticks the nuptial fire and having spread to the west of it a bull’s hide with the neck turned towards the east and the hair outside he makes oblations, while she (the bride) is sitting (on that hide) and takes hold of him, with the four verses (one oblation with each verse) ‘may Prajāpati create offspring to us’ (Rg. X. 85. 43-46); and then he partakes of curds with the verse ‘may all the gods unite our hearts’ (Rg. X. 85. 47) and gives (the remaining curds) to her or he besmears the heart (of both) with the rest of the āJya (with part of which he had already sacrificed). From that time they should not eat kṣāra²³⁴ and lavaṇa, should observe celibacy, wear ornaments and sleep on the ground (not on cots) for three nights or twelve nights or for a year according to some (teachers); thus (those teachers say) a ṛṣi will be born as a son (to them). When he has fulfilled these observances he should give the bride's shift to a brāhmaṇa who knows the Sūrya hymn (Rg. X. 85) and food to brāhmaṇaṣ Then he should cause the brāhmaṇas to pronounce auspicious wordṣ

In this description of the saṁskāra of marriage there are three partṣ There are certain rites that are preliminary, there are then a few rites that are of the essence of the saṁskāra viz. pāṇigrahaṇa, homa, going round the fire and the saptapadī, and there are certain rites like the seeing of the Pole star &c. that are subsequent to the central riteṣ The essential rites are mentioned by all sūtrakāras, but as to the preceding and sub- sequent rites there is a great divergence in the detailṣ Even as regards the essential rites the sequence in which they take place differṣ For example, the Āśv. gr. (I. 7. 7) describes going round the fire before saptapadī, while the Āp. gr. describes saptapadī (IV. 16) before the act of going round the fire (V. 1). The Gobhila gṛhya (II. 2. 16), the Khādira gr. (I. 3. 31), and Bauḍ gr. (I. 4. 10) describe pāṇigrahaṇa after saptapadī while many other sūtrakāras describe it before saptapadī. Then again there are many acts reference to which is altogether omitted in the Āśv. gr. e. g. there is no reference to madhuparka (which is mentioned in Āp. gr. III. 8, Bauḍ gr. I. 2. 1, Mānava gr. I. 9) nor to kanyādāna (which is referred to in Pār. gr. I. 4 and described in detail in Mānava gr. I. 8. 6-9). Āśv. probably omitted express mention of kanyādāna, because in defining the first four forms he uses the word 'dadyāt' while in the last four there is no kanyādāna and Āśv. wanted to describe ceremonies that were common to all formṣ

Taking as many gṛhya sūtras as I could read, the following is a fairly exhaustive list of the different matters described in the saṁskāra of marriage. A few notes are added against those that are deemed important.²³⁵

Vadhuvara-guṇaparīkṣā: (examining the suitability of a girl or a bride-groom). This has been dealt with already (vide pp. 429-436 above).

Varapreṣaṇa: (sending persons to negotiate for the hand of the girl). The ancient custom seems to have been to send some person or persons (Rg. X. 85. 8-9) to the father of the girl for asking her in marriage. The same was the practice in the sūtraṣ Vide Śāṁ. gr. I. 6. 1-4 (where Rg. X. 85. 23 is the mantra recited when sending them), Bauḍ gr. I. 1. 14-15, Āp. gr. II. 16 and IV. 1-2 and 7. Even in medieval times, particularly among kṣatriyas, the bride-groom was the first to seek for the hand of a girl. In the Harṣacarita, prince Grahāvarma of the Maukhari race sent messengers for the hand of princess Rājyaśrī, sister of Harṣa¹²⁴⁰. In modern times among the brāhmaṇas and many other castes, the girl's father has to seek a bride-groom, though among the śūdras and several other castes the old practice is retaineḍ

Vāgdāna or Vān-niścaya: (settling the marriage). This is described by Śāṁ. gr. I. 6. 5-6. Medieval works like the S. R. M. pp. 529-533 describe this ceremony at great length.

Maṇḍapa-karaṇa: Erecting a paṇḍal where the ceremonies are to be performeḍ Pār. gr. I. 4 says that marriage, caula upanayana, keśānta and sīmanta are to be performed outside the house in a paṇḍal. Vide Sām Pr. pp. 817-818.

Nāndi-śrāddha and Puṇyāhavācana: These are referred to by Bauḍ gr. I. 1. 24; most of the grhyasūtras are silent about these. Vide for these pp. 216, 218 above.

Vadhūgrhagamana: (bride-groom's going to the bride's house). Vide Śāṁ. gr. I. 12. 1.

Madhuparka: (reception of the bride-groom at the bride's house). Āp. gr. III. 8, Bauḍ gr. I. 2. 1, Mānava gr. I. 9 and Kāṭhaka gr. 24. 1-3 prescribe thiṣ Vide below chap. X for madhuparka. Śāṁ. gr. I. 12. 10 appears to refer to two madhuparkas, one before marriage and one after marriage (when the bride-groom returned to his own house). The commentator Ādityadarśana on Kāṭhaka gr. 24. 1 refers to the opinion of some that madhuparka should be offered at the close of the marriage, but states his own view that in all countries it is offered before marriage.

Snāpana, Paridhāpana and Saṁnahana: (making the bride bathe, put on new clothes and girding her with a string or rope of darbha.) Vide Āp. gr. IV. 8 and Kāṭhaka gr. 25. 4. Pār. gr I. 4 refers only to the putting on of two garments, Gobhila gr. (II. 1. 17–18) refers to bathing and putting on a garment, Mānava gr. (I. 11. 4–6) refers to paridhāpana and samnahana. Strangely enough Gobhila gr. II. 1. 10 speaks of the sprinkling of the girl’s head with the best of surā (wine), which the commentator explains as water.

Samāñjana : ( anointing the bride and bride-groom ). Vide Śāṇ gr. I. 12. 5, Gobhila gr. II. 2. 15, Pār. gr. I. 4, in all of which Rg. X. 85. 47 is cited as the mantra with which anointing or sprinkling is done.

Pratisarabandhā : ( tying an amulet string on the bride's hand ). Vide Śāṇ gr. I. 12, 6–8; Kauśika sūtra 76. 8.

Vadhūvara-niṣkramaṇa : ( the coming out into the pandal of the bride and bride-groom from the inner part of the house). Pār. gr. I. 4.

Paraparasamīkṣaṇa : ( looking at each other). Vide Pār. gr. I. 4, Āp. gr. IV. 4, Bauḍ gr. I. 1. 24–25. Pār. gr. I. 4 says that the bride-groom recites Rg. X. 85. 44, 40, 41 and 37 at this time. Āp. gr. IV. 4 and Bauḍ gr. say he recites Rg. X. 85. 44. The Āśv. gr. pariśiṣṭa I. 23 says that first of all a piece of cloth is held between the bride-groom and bride and that at the proper astrological moment it is removed and then the two see each other. Laghu-Āśvalāyana-smṛti (15. 20) also says the same. This practice is observed even now. When the interposed cloth is held between the bride and bride-groom verses called maṅga-lāṣṭakas are repeated by brāhmaṇas, the last of which verses is 'tad-eva lagnam sudinam tad-eva' &c.

Kanyādāna : ( the gift of the bride). Vide Pār. gr. I. 4, Mānava gr. I. 8. 6–9, Vārāha gr. 13. The Āśv. gr. pariśiṣṭa sets out the procedure about the kanyādāna which is the same even now. The Śaṁ. K. p. 779 notes about half a dozen different methods of uttering the formula in kanyādāna. It is in this rite that the father of the girl says that the bride-groom should not prove false to the bride in dharma, artha and kāma and he responds with the words 'I shall not do so' ( nāticārāmi).²³⁷ This is done even now. Agnisthāpana and homa: (establishing the fire and offering of ājya oblations into fire). Here there is great divergence about the number of āhutis and the mantras to be reciteḍ Vide Āśv. gr. I. 7. 3 and I. 4. 3–7, Āp. gr. V. 1 ( 16 āhutis with 16 mantras), Gobhila gr. II. 1. 24–26, Mānava gr. I. 8, Bhāradvāja I. 13 &c.

Pāṇigrahaṇa: (Taking hold of the bride's hand).

Lājahoma: (Offering of fried grain into fire by the bride). Vide Āśv. gr. I. 7. 7–13, Pār. gr. I. 6, Āp. gr. V. 3–5, Śāṅ. gr. I. 13. 15–17, Gobhila gr. II. 2. 5, Mānava gr. I. 11. 11, Bauḍ gr. I. 4. 25. &c. Āśv. gr. says that the bride makes three offerings of fried grain when mantras are repeated by the bridegroom and a fourth is made of the remaining lājās by the bride silently. Some others speak of only three offerings by the bride.

Agniparīṇayana: (the bride-groom going in front takes the bride round the fire and water jar ). It is while doing this that he utters the words ‘amohasmi’ &c. (vide Śāṅ. gr. I. 13. 4, Hir. gr. I. 20. 2 &c ).

Aśmārohaṇa: (making the bride tread on a mill-stone ). These three are done thrice viz. lājahoma, then agniparī- ṇayana and aśmārohaṇa, one after another.

Saptapadī: (taking seven steps together ). This is done to the north of the fire; there are seven small heaps of rice and the bride-groom makes the bride step on each of these seven with her right foot beginning from the west.

Mūrdhābhiṣeka: (sprinkling holy water on the head of the bride and of the bride-groom according to some and on the head of the bride only according to others ). Āśv. gr. I. 7. 20, Pār. gr. I. 8, Gobhila gr. II. 2. 15–16 &c.

Sūryodīkṣaṇa: (making the bride look towards the sun ). Pār. gr. I. 8 speaks of this and employs the mantra ‘tac-cakṣur’ ( Ṛg. VII. 65. 16 = Vāj. S. 36. 24 ).

Hṛdayasparśa: (touching the bride' heart with a mantra ). Pār. gr. I. 8, Bhār. gr. I. 17, Bauḍ gr. I. 4. 1.

Prekṣakānumantrana: (addressing the spectators with reference to the newly married bride ). Mānava gr. I. 12. 1 (which employs Ṛg. X. 85. 33 ). Pār. I. 8 employs that verse for reciting over the bride.

Dakṣiṇādāna: (gifts to the ācārya ). Pār. gr. I. 8, Śāṅ. gr. I. 14. 13–17 (both prescribe a cow as the fee in the case of brāhmaṇas, a village in marriages of kings and nobles, a horse in marriages of vaiśyas &c. ). Gobhila gr. II. 3. 23, Bauḍ gr. I. 4. 33 speak of only a cow.

Grhapraveśa : (entering the bride-groom's house). Grhapraveśanīya homa : (sacrifice on entering the bride- groom's house ). Sāṇ gr. I. 16. 1–12, Gobhila gr. II. 3. 8–12, Āp. gr. VI. 6–10.

Dhruvārundhati-darśana : (pointing out the Pole star and Arundhati to the bride at night on the day of marriage ). Aśv. gr. I. 7. 22 speaks of the seven sages in addition, Mānava gr. I. 14. 9 speaks of the same three and adds Jīvantī, Bhār. gr. I. 19 speaks of Dhruva, Arundhati and other nakṣatras, Āp. gr. VI. 12 (only Dhruva and Arundhati) Pār. I. 8 (only Dhruva). According to Sāṇ gr. I. 17. 2, Hir. gr. I. 22. 10 both the bride and the bride-groom remain silent till night. According to Aśv. gr. only the bride does so. Gobhila gr. II. 3. 8–12 describes Dhruvārundhati-darśana before grhapraveśa.

Āgneya Sthālīpāka : (mess of cooked food offered to Agni). Vide Āp. gr. VII. 1–5, Gobhila II. 3. 19–21, Bhār. gr. I. 18, Hir. gr. I. 23. 1–6.

Trirātrvrata : (keeping for three nights after marriage certain observances). Vide Aśv. above p. 530 for the observances which are enumerated by almost all sūtrakāraṣ Āp. gr. VIII. 8–10, Bauḍ gr. I. 5. 16–17 contain the interesting injunction that the newly married pair should sleep on the ground on the same bed for three nights, but should interpose between them a staff of udumbara wood anointed with perfumes and wrapped round with a garment or a thread and that on the fourth night it should be removed with the verses Rg. X. 85. 21–22 and thrown into water.

Caturthīkarma : (rite on the fourth night after marriage). This has been described above (pp. 202–204 ).

In the medieval digests certain other ceremonies are mentioned and they are observed in modern times also. A few of them will be noted below. Here again the order is not the same in all workṣ The Dharmasindhu p. 265 refers to this divergence.

Simāntapūjana : (honouring the bride-groom and his party on their arrival at the bride's village ). This is done before vāgniścaya in modern timeṣ Vide Sāṃ K. p. 768; Dharma- sindhu III. p. 261. . _Gauri-Hara-pūjā:_ (worship of Śiva and his consort Gaurī). Sāṃ K. p. 766, S. R. M. p. 534 and 544, Dharmasindhu p. 261 ( notes that there are several views as to when kanyādāna takes place ) describe thiṣ Images of Gaurī and Hara are to be made of gold or silver or pictures of them on a wall &c. or on a piece of cloth or stone are to be drawn and worshipped by the intending bride after puṇyāhavācana and before kanyādāna. Vide Laghu-Āśvalāyana 15. 35.

_Indrāṇī-pūjā:_ (worship of Indrāṇī, the consort of god Indra ). Vide Sāṃ K. p. 756, S. R. M. p. 545. This seems to have been comparatively an ancient practice as Kālidāsa in Raghuvamśa VII. 3 seems to refer to it ( ‘there was absence of disturbers of svayaṁvara on account of the presence of Śacī’ ). Probably Śacī was worshipped before the svayaṁvara begaṇ

_Taila-haridrārōpaṇa:_ (Applying turmeric powder to the bride-groom’s body from what is left after the bride’s body has been so treated). Vide Sāṃ K. p. 757, Dharmasindhu III. p. 257.

_Ardrākṣatāropaṇa:_ (mutual showering of wet unbroken rice grains by the bride and bride-groom). In a vessel of some metal like silver a little milk is poured and clarified butter is sprinkled over it and unbroken wet rice grains are poured thereiṇ The bridegroom applies milk and ghee to the joined hands of the bride twice and thrice places rice grains in the joined hands of the bride so that her anjali becomes filled up and twice sprinkles ghee over her joined handṣ Some other person does the same to the joined hands of the bride-groom and the bride’s father places a golden piece on the joined hands of both. Then the bride-groom places his joined hands on those of the bride whose father then repeats a mantra and raises her up; she then pours the rice over the head of the bridegroom who follows her in the same way. This is done thrice by each and then lastly by the bride ( i. e. seven times in all ). Then the priest sprinkles on their heads water with an udumbara twig together with dūrvā grass after reciting verses ‘ Āpo hi ṣṭhā &c. ( Rg. X. 9. 1-3 )’. Then the couple make a tilaka mark on each other’s forehead, garland each other and tie a thread with a turmeric piece on each other’s hand ( which is variously called ‘kaṅkaṇa-bandhana’ or ‘kautuka-bandhana’). Vide Sāṃ Pr. pp. 828-829, S. R. M. p. 556. Kālidāsa in Raghuvamśa VII describes ārdrakṣatā-ropaṇa as the last of the rites of marriage and in VIII. 1 speaks of the kautuka. Maṅgalasūtra-bandhana: (tying a string having golden and other beads on it round the neck of the bride). This is now regarded as the most important ornament which no woman will part with as long as her husband is alive. But the sūtras are entirely silent about it. Among the earliest references is one from Saunaka smṛti ²³⁸ (mṣ in Bombay University library, folio 39 b). The Laghu-Āśvalāyana smṛti 15. 33 also prescribes it and the mantra to be employed when doing so. Gadādhara on Pār. gr. I. 8 says that maṅgala-sūtra should be worn and garlands be placed round their necks by the bride and bride-groom, though the sūtra of Parāskara is silent on the point. The Bauḍ gr. śeṣa sūtra V. 5 in describing 'arkavivāha' speaks of maṅgalya sūtra to be tied to the plant. It is not clear whether it means the same as the maṅgalasūtra now tied by married women round their neckṣ About the nose-ring or nose ornament to which all women whose husbands are living attach the greatest importance in modern times, the sūtras, smṛtis and even the early digests are entirely silent. Dr. Altekar in his recent work 'Position of Hindu women in Hindu Civilization' pp. 362-64 holds from the evidence of the sculptures throughout India-that the ‘nose-ring was unknown throughout the whole of India during the entire Hindu period' (i. e. till about 1000 A. D.). Mr. P. K. Gode in Annals of B. O. R. I. vol. XIX (for 1939) pp. 313-34 adduces evidence to show that a nose-ornament was known to literary works from at least about 1000 A. D.

Uttarīyaprānta-bandhana: (Tying turmeric pieces and betelnut on to the end of the upper garments of both and making a knot of their two garments together). Vide Saṃ K. p. 799, S. Pr. p. 829.

Airinīdāna²³⁹ : (Presenting the mother of the bride-groom with several gifts in a large wicker-work basket with lamps lighted and requesting her and the relatives of the bride-groom to treat the bride affectionately). Vide Saṃ K. p. 811 Dharmasindhu p. 267. A basket of bamboo (vaṁśa) is used probably to symbolize the continuity of the family (vaṁśa) of the bride-grooṃ This is done when the bride is about to leave the father's place to go to the bride-groom's place after marriage.

Devakottāpana and Maṇḍapodvāsana : (taking leave of the deities that had been invoked before the ceremonies began and taking down the pandal). Saṃ K. pp. 532-533, S. R. M. pp. 555-556.

Two interesting questions arise viz. when can a marriage be said to be final and irrevocable and what would happen if a marriage is brought about by force or frauḍ

Manu says (VIII. 168) 'what is given by force, what is enjoyed by force, also what has been caused to be written by force and all other transactions done by force, Manu has declared void;' and in VIII. 165 he places fraudulent transactions on the same footing as those brought about by force. There is great difficulty in applying these dicta to marriageṣ We saw above that Vaṣ Dh. S. (17. 73) and Bauḍ Dh. S. declare that if a girl has been carried off by force and has not been wedded with the repetition of sacred texts she may be given to another man in marriage. Viśvarūpa (p. 74) and Aparārka (p. 79) add a gloss that this can be done only after she has undergone a prāyaścitta. From this it appears that if the marriage rites (like saptapadi) have been performed, the ancient law-givers would not have declared the marriage null and void even if the girl had been carried away by force or married by frauḍ In modern works on Hindu Law the proposition is stated²⁴⁰ that 'a marriage, though performed with the necessary ceremonies, may be set aside by the court, if it was brought about by force or frauḍ' This opinion is based on what was said in some decided cases e. g. Aunjona Dasi v. Prahlada Chandra (6 Bengal L. R. 243 at p. 254), Venkatacharyulu v. Rangacharyulu I. L. R. 14 Maḍ 316 at p. 320, Mulchand v. Bhudhia I. L. R. 22 Boṃ 812. But in none of these cases was a marriage duly solemnized by the performance of the rites of pāṇigrahaṇa, going round the fire and saptapadī, set aside. There are mere obiter dicta in these cases to the effect that a marriage may be set aside by the court for force or frauḍ

The Vaṣ Dh. S. (17. 72) goes so far as to observe ' when a girl has been promised in marriage (and the promise has been confirmed) with water, if the intended bride-groom dies, but the Vedic mantras have not been recited, that girl still belongs to her father (and may be given to another)'. Kātyāyana ²⁴¹ has a similar verse 'if after choosing a girl as his bride, a man dies (or is unheard of) the girl after the lapse of three monthly periods may marry another'. And another verse of the same author says that if a person after giving a gratuity for a girl and strīdhana to her goes abroad, the girl may be kept unmarried for a year and then may be given to another. Manu (VIII. 227) says 'the Vedic mantras recited in the marriage rite are a sure indication of wife-hood; but their completion should be under- stood by the wise as occurring on taking the seventh step'. Aparārka p. 94 (on Yaj. I. 65) quotes a similar verse from Nārada (strīpuṁsa v. 3). The Udvāhatattva p. 129 quotes Yama to the same effect. So it follows that the marriage be- comes complete and irrevocable the moment the saptapadī rite is performed, but before that rite is gone through there is a locus pænitentiæ or a power to resile from what has been under- takeṇ Before the saptapadī is performed, if the bride-groom dies, the bride is still a maiden and not a widow and can be married again²⁴². The most essential ceremonies of marriage are the homa and saptapadī. The Droṇaparva²⁴³ says that promise of a daughter and giving a daughter with water are not certain means of knowing wife-hood but saptapadī is known to be the completion of marriage. If any of the other ceremonies are wanting that would not vitiate the marriage. In I. L. R. 12 Cal. 140, it has been held that the vṛddhi-śrāddha is not an essential ceremony and its absence would not vitiate marriage. Even the Kāmasūtra²⁴⁴ quotes the unanimous opinion of the ācāryas that marriages celebrated before fire as a witness cannot be revokeḍ In the case of śūdras there are no Vedic mantras and so in their case the completion of marriage will be determined according to custoṃ The digests like Gr. R. say (p. 57) that in the case of śūdras the marriage will be complete when the śūdra girl holds the fringe of the garment of the bride-grooṃ

Manu (IX. 47) declares 'once is the partition of inheri- tance made, once is a maiden given in marriage'. This rule really means that once the ceremony of marriage is completed by saptapadi, the marriage is irrevocable and the girl cannot thereafter be given to any one else. But if a girl is only promised in marriage and if a more worthy suitor subsequently presents himself, then the father may commit breach of promise and give her in marriage to another. Vide also Manu (IX. 71 and VIII. 98). Yāj. I. 65 states the rule and the exception 'A girl is given only once; a person who after having promised to give, deprives that man of her, is liable to the punishment of a thief; but if a more worthy suitor approaches, the father may deprive the former (suitor) of her (hand) though promised'. The Mahābhārata (Anuśāsana 44. 35) says that up to pāṇigrahaṇa any one may ask for the hand of a maiden (even though promised before to another). Nārada (stripumsa vv. 30 and 32) contains similar provisionṣ²⁴⁵ Conversely, if a man agrees to marry a girl and subsequently discovers defects in her or if she is diseased or already deflowered or given by practising fraud he may refuse to marry her. Vide Manu IX. 72. If a guardian gives away a girl without declaring the defects of a girl (and they are subsequently discovered) the guardian should be fined in the highest ammercemment (according to Yāj. I. 66 and the lowest, according to Nārada, stripumsa v. 33). Aparārka²⁴⁶ (p. 95) and others add that the defects referred to must be latent and not apparent or patent to view. If a suitor repudiates a girl who is free from defects he should be fined in the highest ammercemment and if he falsely accuses her of defects he should be fined one hundred panas (Yāj I. 66 and Nārada, strīpumsa v. 34). Nārada adds that he who abandons a faultless girl should be punished and should be made to marry the same girl.²⁴⁷ Some of the smṛtis and digests are very much exercised over the question of the bride having menstruation while the marriage ceremonies are in progress and homa is about to be performeḍ Atri²⁴⁸ ( Jīvānanda part 1, p. 11, chap. V ) prescribes that having made the girl bathe with the Havismatī ( with a verse in which the word havis occurs, probably Ṛg. X. 88. 1 or VIII. 72. 1 ) and to put on other garments and having offered an āhuti of clarified butter with the verse 'Yuñjate manaḥ' ( Ṛg. V. 81. 1 ) the ceremonies should be proceeded with. The Smṛtyarthasāra ( p. 17 ) first quotes two verses ( the same as Atri’s ) and proposes an alternative method viz. that the bride and the bride-groom should stay separate from each other for three days and on the fourth day after ceremonially bathing themselves should perform the homa in the same fire.