N.G. Chandavarkar J.
1. The suit, which has led to this appeal, was brought by the appellant, Bai Kashi, for arrears of maintenance against Mansukh, alleging that she was his natra wife, by whom he had several children, but that he had expelled her from his house in 1901 and had refused to maintain her.
2. The Subordinate Judge, First Class, at Ahmedabad, who tried the suit, has found upon the evidence that the appellant Bai Kashi, was a Brahmin widow, when she married the respondent Mansukh, a Shudra by caste, and that some time after marriage, both of them became converts to Christianity. The principal issue raised at the trial was whether the marri age was legal and valid. The Subordinate Judge found in the negative upon that issue, and, finding on some of the subsidiary issues also against the appellant, he has dismissed the suit.
3. The Subordinate Judge's finding upon the evidence that the parties as a fact married according to the rites of Hindus, that at the time of marriage the appellant was a Brahmin widow, aged sixteen, and the respondent a Shudra, that thereafter they lived as wife and husband, having several children, both before and after their conversion to Christianity, has not been assailed before us by the respondent. These facts found by the Subordinate Judge are amply borne out by the evidence. Three questions of law have been argued in support of the appeal. They are, first, is the marriage of a Brahmin woman with a man of the Shudra caste, such as the marriage in dispute was, legal and valid, according to Hindu Law? Secondly, if it is not legal and valid, is the appellant entitled to maintenance as at all events the kept mistress of the respondent under that law? Thirdly, is the respondent estopped from maintaining that the appellant is not his lawfully married wife and as such is not entitled to maintenance
4. At the outset Mr. Shah, the learned pleader for the appellant, urged that, whether the marriage was valid or not according to Hindu law, the parties having lived as husband and wife for more than twenty-five years and having had children by the marriage contracted in fact, the Court should raise every presumption in favour of the legality of the marriage after the lapse of so long a period. It is true that once we get to a marriage in fact, ' there would be a presumption in favour of there being a marriage in law;' Inderun Valungypooly Taver v. Ramasawmy Pandia Talaver (1869) 13 M.I.A. 141; and that is so, especially where a great length of time has elapsed since its celebration. But the legal presumption can be rebutted by ' strong, satisfactory and conclusive ' evidence : Morris v. Davis (1837) 5 C & F. 163. In the present case it is so rebutted. The evidence is conclusive that the woman was a Brahmin, and the man a Shudra, by caste; and upon that the question arises, whether such a marriage is legal according to Hindu Law.
5. It would appear from some of the Smritis that at one period in the growth of the Hindu community in ancient times intermarriage between a man of a higher and a woman of a lower caste was allowed, under certain restrictions. Thus in Manu, Chapter III, Verse 13, it is stated:- ' It is declared that a Shudra woman alone (can be) the wife of a Shudra, she and one of his own caste (the wives) of a Vaishya, these two and one of his own caste (the wives) of a Kshatriya, those three and one of his own caste (the wives) of a Brahmin' (Sacred Books of the East: Vol. 25, p. 77). According to that, a man of any of the three twice-born castes could legally marry a Shudra woman. But Manu condemns in verse 14 of the same Chapter the marriage of a Shudra woman with a Brahmin or a Kshatriya or a Vaishya. Medhatithi, one of Manu's commentators, construes the condemnation as being not an absolute prohibition of the marriage but in the nature of a recommendation to a Brahmin or a Kshatriya or a Vaishya not to marry a Shudra woman. He argues that as verse No. 13 allows and verse No. 14 condemns the marriage, there is an option given. Medhatithi's view has the support of the maxim of the Mimamsa (the Hindu Law of Interpretation), viz., Virodhe the vikalpaha, which means that where there is a conflict of texts we are free to adopt any text we choose (See the Mitakshara, Moghe's Edition No. 3, p. 3).
6. Kulluka Bhatta, another of Manu's commentators, seems to take the same view as Medhatithi. He says that as in Verse 13 of Manu, Ch. HI, marriage according to the descending order of castes is permitted, the prohibition in verse 14, 'should be understood to refer to a marriage in the reverse order of castes.' Vasistha in his Smriti says that intermarriage between a man of a higher caste and a woman of a lower caste is mentioned y by one Acharya but he himself pronounces against it. Narada allows such marriages, and is not against a man of one of the regenerate castes matrying a Shudra woman (XII, 5, Dr. Jolly's Institutes of Narada, p. 81). Yajnyavalkya in his Smriti, on which the Mitakshara is a commentary, recognises by implication the conflict of authority on the question of marriage between a man of one of the regenerate castes and a Shudra woman and says that his own opinion is against such marriages, (Mitakshara, Moghe's Edition No. 3, p. 15). Commenting upon that text of Yajnyavalkya, Vijnaneshvara explains in the Mitakshara that the text is a prohibition (nishedha) of the marriage of a woman of the Shudra caste with a man of any of the twice-born castes.
7. From this gloss of the Mitakshara, from which Nilakantha, the author of the Vyavahara Mayukha, expresses nowhere any dissent, it is reasonable to infer that, according to the leading authorities on Hindu Law as recognised in this Presidency, a Shudra wife is not permitted to a Brahmin, a Kshatriya, or a Vaishya. If that is so, it is a necessary corollary from it that the marriage of a Brahmin woman with a Shudra is also prohibited. This inference is supported by a text cited by Madhavacharya in his commentary on the Parashara Smritis which is regarded by Hindus as an authority of special force in these times called Kaliyug. In that commentary, after quoting the conflicting texts on the question of marriage between persons of different castes, Madhavacharya concludes as follows:-
Distinguishing the different ages, the law is clearly established by a Smriti, which, referring to the marriage of persons of any of the twice-born castes with persons of other castes, declares. The learned say that these practices must not be followed in the age of Kali.
8. It is not a case, however, of mere inference, so far as the marriage of a Brahmin woman with a Shudra is concerned. Section 4 of the Chapter on Rituals in the Mitakshara deals with the question how the different castes of mixed offspring are formed and must be distinguished. Vijnaneshwara begins his glosses on the smritis of Yajnyavalkya dealing with that question with the following introductory remarks :-
Having stated (in the previous section) that a Brahmin may have four, a Kshatriya three, a Vaishya two, wives, and a Shudra one wife ,(the author) pointed out that sons were to be begotten thereby. Now he (the author) proceeds to explain what kind of son is begotten and by whom and from whom.
9. In this introductory remark Vijnaneshwara would seem to have contemplated the possible validity of the marriage of a man of one of the regenerate castes with a Shudra woman, though he had in an earlier chapter construed Yajnyavalkya's pronounced opinion against it as a nishedha or prohibition. Assuming that this conflict in Vijnaneshwara's views makes it uncertain whether, in his opinion, such marriages were absolutely illegal, the introductory remark does not touch the question as to the legality of the marriage of a Brahman woman with a Shudra. That question is dealt with in Verse No. 93 of the Yajnyavalkya Smriti, where it is stated that ' one begotten on a Brahmin woman by a Shudra becomes aChandala, outcaste to all religion' (chandalaha sarvadharma bahishkrutaha).
10. To understand the meaning and significance of this text of Yajnyavalkya and its bearing on the question now under discussion, we must compare its language with that of the two immediately preceding texts (Nos. 91 and 92). No. 91 says (see the translation of the Yajnyavalkya Smriti in Mandlik's Hindu Law, page 173):-'The son begotten by a Brahmin upon his Kshatriya woman is called a Murdhavasikhta, one upon a Vaishya woman is called an Ambasta, one upon a Shudra woman is a Nishada or Parasava.' Then No. 92 states:- 'The sons borne by a Vaishya and a Shudra woman to a Kshatriya are called Mahishya and Ugra respectively; the son born of a Vaishya by a Shudra woman is a Karana. This rule refers to regularly married wives.' The last sentence, viz., 'this rule refers to regularly married wives,' relates to both the texts, 91 and 92, and shows that the marriages there referred to are legal. Then we come to No. 93. It is as follows in translation:-'One begotten on a 'Brahman 'woman by a Kshatriya is a Sita; by a Vaishya, a Vaidehaka; and by a Shudra, a Chandala, outcaste to all religion.'
11. The original word for ' religion' is dharma, Though Mr. Mandlik, from whose Hindu Law the translation of the above-cited text of Yajnyavalkya is taken, translates dharma as meaning ' religion,' the term has a wider construction and means also law or any customary observance. Hence 'outcaste to all religion, law and customary observance ' would be a more appropriate translation of the last portion of the text. Apararka, in his commentary on Yajnyavalkya, explains it to mean that the offspring of a man of the Shudra caste by a Brahmin woman is excluded from the rights and observances of the four castes called varnadharma but not from the rights and obligations common to all human beings, which are designated sadharana dharma, or from the duties prescribed for Chandalas. Apararka mentions those duties from a text of Shankha. It is unnecessary to mention them here, because the exclusion abovementioned from 'all religion' necessarily implies that the connection of the parents, being sinful, is a prohibited connection and that the parents become on that account degraded from their respective castes.
12. It will be noticed that Yajnyavalkya does not say in text No. 93, as he did in the preceding texts Nos. 91 and 92, that the rule prescribed in it refers to ' regularly married wives '. The omission must be regarded as deliberate and can be accounted for on no other ground than that, in Yajnyavalkya's opinion, such marriages are not legal and the offspring of them must be regarded as the result of adultery.
13. Why then, it may be asked, does he refer to the sons born of such connections as are dealt with in Verse No. 93, in the Chapter on Rituals while settling the status of sons born of marriages valid in law? The answer is that the section, in which Verse No. 93 occurs, deals with the various kinds of sons, including what are called mixed off-spring, for the purpose of assigning to each kind its appropriate caste, place, and occupation, in Hindu society. The caste, the place, and the occupation, had to be determined whether the marriage of the parents was legal or not, having regard to the existence of such offspring in the community.
14. This interpretation of text No. 93 of Yajnyavalkya Smriti is borne out by a reference to other authorities. That text and the text which immediately follows (No. 94) deal with the place in society of the sons of parents, of whom the father is of a lower and the mother of a higher caste. In neither of the texts does Yajnyavalkya say that the two texts in question refer to 'regularly married wives' as he had said with reference to the two immediately preceding texts, which relate to the sons of parents of whom the father is of a higher and the mother of a lower caste. Now, in text No. 94 Yajnyavalkya states that the son of a Shudra by a Kshatriya woman is called Kshattara. According to Vyasa, he is called Nishada, and he is to be regarded as born in adultery. In a work called Jati Viveka (Examination of Castes) it is said:-' A Kshatriya woman who gives birth to a son by a Shudra produces a son who is called Kshattara or Nishada. He is outcaste to all religion.'
15. These authorities support the conclusion that Yajnyavalkya's texts Nos. 93 and 94, dealing with marriages in the reverse order of castes, refer not to marriages valid in law but to adulterous connections of the most degraded character.
16. Further, Yajnyavalkya says that a son begotten by a Shudra on a Brahman woman becomes a Cha?idala, the most degraded of human beings, and therefore ' outcaste to all religion.' That is to say, he regards the progeny as something worse than illegitimate; and that can only be because they are the offspring of a connection which is sinful. Where a particular act is condemned as sinful, that is, as rendering the doer liable to absolute and inexpiable forfeiture of all religious and customary rights of the four castes, the necessary implication is that it is sinful and is therefore prohibited. That is the converse of the rule that where a particular act is prescribed and the omission to do it is condemned as sinful, the implication is that the act must be done. For an illustration of this rule, see the Mitakshara, Ch. I, Section VII, placitum 10 (Stokes, p. 400), where Vijnaneshvara points out that the rule of Hindu Law, which requires brothers to allot a share to their sisters when they effect a partition inter se of their deceased father's property, is mandatory, because reufsal to make the allotment ' is pronounced to be a sir; by the text 'they who refuse to give it shall be degraded.' ' So also, when it is said that a son born of a Brahmin woman by a man of the Shudra caste becomes a Chandala (most degraded) and is 'outcaste to all . religion,' the meaning is that he becomes degraded and outcaste to all religion because his parents are such; and the parents are such because the connection is contrary to the Shastras and therefore unlawful.
17. It may indeed be argued that, as in texts Nos. 93 and 94, where Yajnavalkya mentions the kinds of sons born of marriages in the reverse order of castes, he declares that the son born of the marriage of a Brahmin woman with a Shudra becomes ' a Chandala,' ' outcaste to all religion,' but as he does not say that of the other kinds of sons mentioned in the same texts, we must infer either that he was in both the texts dealing with marriages valid in 'law but condemned-as inferior or that he intended to declare all the marriages specified in the texts except the marriage of a Brahmin woman with a Shudra as being not objectionable and prohibited absolutely. It may be maintained that if Yajnyavalkya had intended the latter inference to be the proper* one to draw from his texts, he would not have mentioned the other marriages with the marriage of a Brahmin woman with a Shudra in the same texts; that, therefore, the former inference is the proper one to draw. This argument finds a cogent answer from Vijnaneshvara's explanation in another section of the Mitakshara, that all the kinds of sons mentioned in texts Nos. 93 and 94, including the Chandala, stand upon the same footing, because they are all, he says on the authority of Angira Rishi, antyas, i. e., men of the lowest and most degraded castes, inferior even to Shudras. (See the Mitakshara, section on 'Penance ') In Verse No. 95 of his Smriti Yajnyavalkya describes the progeny of marriages in the reverse .order of castes as asantas, i. e., nonentities (Moghe's Edition No. 3, p. 381).
18. So far then the conclusion from the Mitakshara is against the legality of the marriage in dispute. As to Nilakantha, the author of the Vyavahara Mayukha, which prevails in Guzerat, whence this case comes, he cites verse No. 93 of Yajnyavalkya in his Samskara Mayukha and says that he does not dilate upon the subject of the distinction of castes of the various kinds of mixed offspring, because ' it is useless ' and for fear of diffuseness. That is to say, the subject has no practical value and interest and therefore it is unnecessary to write at length upon it. It could have no practical value, because, in the author's opinion, such marriages being prohibited had become obsolete. Nilakantha must, therefore, be regarded as holding the same opinion as the author of the Mitakshara.
19. This conclusion from the treatment of the question under discussion, by Vijnaneshvara in the Mitakshara and by Nilakantha in the Samskara Mayukha, derives support from other authorities on Hindu Law. For instance, Medhatithi, one of the commentators of Manu, in his gloss on Manu's text, Chapter III, 13, declaring that a Brahmin may marry a woman of the Shudra caste, and on text No. 14 condemning that kind of marriage, says. 'But it does not follow that woman belonging to a higher caste can be married by a Shudra.' Kulluka Bhatta, another of Manu's commentators, says: 'This prohibition should be understood to refer to a marriage in the reverse order of castes.' Narada does not mention in his Smriti marriages in the reverse order, though he allows a man of one of the three regenerate castes to marry a Shudra woman. Anga Rishi declares as follows:-' A son born of a Sanyasi (ascetic), a son born of a Brahmin woman by a Shudra, a son born of the marriage of a couple belonging to the same gotra-all these three are Chandalas.' Now, the collocation of these three shows that just as the last kind of marriage-i. e., of sagotras, a couple belonging to the same gotra or gensis undoubtedly void and illegal, according to Hindu Law, so the first two are also void and illegal. To the same effect is Yama: ' The son of a degraded ascetic, the son of a Brahmin woman by a Shudra and the son born of a marriage of parents of the some gotra, are all three Chandalas.'
20. For these reasons, the marriage of the appellant with the respondent must be held to have been illegal and void, according to Hindu Law.
21. The next question is whether the appellant can claim maintenance as the dasi or kept mistress of the respondent. That question must also be answered in the negative. The appellant was discarded by the respondent before his death and her connection with him cannot be said to have been of a continuous character to entitle her to maintenance as his dasi or kept mistress, as that term is understood and explained as law for this Presidency in the leading case of Rahi v. Govinda ILR (1875) 1 Bom. 97. The appellant's claim in the alternative as the kept mistress or continuous concubine of the respondent must be disallowed also on the authority of Ningareddi v. Lakshmawa (1901) 26 Bom. 163.
22. Lastly, it was contended for the appellant that, in any case, the respondent must be held estopped from denying to her the status of his wife and the right of maintenance, after he had married her in fact and lived with and borne children by her as his lawfully wedded spouse for a large number of years. To this argument there are two serious objections of law.
23. If the argument means that the respondent by the marriage in fact and his subsequent conduct impliedly contracted to maintain her as his wife, the contract was unlawful, and contrary to public policy according to Hindu Law.
24. If, on the other hand, the argument is one of bare estoppel, it must come under one of two heads. Then it must be either representation made by the respondent to the appellant at the time of marriage that he would treat her as his wife, or representation by subsequent conduct on his part that she was his wife. As to the first, ' the doctrine of estoppel by representation ', as said by Lord Selborne in Maddison v. Alderson (1883) 8 App. Cas. 467 : 3 Bom. L.R. 647 ' is applicable only to representations as to some facts alleged to be at the time actually in existence, and not to promises de flUuro, which, if binding at all, must be binding as contracts'. The representation by the respondent to the appellant at the time of marriage to treat and maintain her as his wife was a promise infuturo and as such it was a contract, which was illegal.
25. As to the second head of the argument, viz., estoppel by subsequent conduct, Section 115 of the Indian Evidence Act requires to create such estoppel intentional conduct on the part of the respondent which caused the appellant to believe and act on the belief that a certain thing was true, i. e., true in fact, not merely true in law. Here the representation by conduct related not to a fact but to the belief of both the parties as to the law that the marriage was lawful. Ex hypothesis the respondent believed and represented that the marriage was valid and that the appellant was his wife, according to Hindu Law. The appellant also believed likewise. But neither party could plead ignorance of law.
26. Had it been the case of the appellant that the respondent had misled her at the time of marriage by a representation that he was a Brahmin and could, therefore, legally marry her there would have been estoppel against him. But that has not been her case either in the Court below or here.
27. In Sarat Chunder Dey v. Gopal Chunder Laha , the Privy Council held that a person making an adoption, which was not valid and legal according to Hindu Law, was estopped from maintaining its invalidity and illegality because of a series of acts on his part which had induced the belief on the part of the boy adopted that the adoption was valid and in consequence of which the adopted son had abstained from claiming a share in the inheritance of his natural family. The principle laid down in that case by their Lordships, that a series o acts by which an adoption is professedly made and sub sequently recognised constitutes a representation not in law only but of fact, is inapplicable here because no question of estoppel having been raised at the hearing in the Court below, no issue was framed to try it and no evidence was led to show what representation of fact was made or what act was done by the respondent, to lead the appellant to believe that she was his lawful wife and how in consequence of suet representation she had acted and had her position altered the her prejudice. On the facts now before us it is as probable that the appellant had induced the respondent to marry he: and treat her as his wife as that he had persuaded her to marry him and treat him as her husband. From the mere fact that the parties went through the Hindu ceremony of marriage and thereafter lived together for several years as husband and wife it is impossible to draw an inference of estoppel. An estoppel is a substantial plea, which must be set up and proved by unambiguous evidence by the party relying upon it.
28. For these reasons the decree must be confirmed, but as the case is one of great hardship to the appellant, we direct that each party bear her or his own costs throughout.