Judgement Appeal (No. 32 of 1919) from a judgment and decree of the High Court (January 12, 1915) varying a decree of the Subordinate Judge of Madura (December 23, 1911). The parties to the litigation were members of a Hindu joint family belonging to the caste of Athangudy Chettis, and were resident at a village called Vallalapatti in the Madura district. The suit was brought by the appellant against the respondents, his father and half-brothers by a second wife, for partition. By paras. 6 and 13 of his plaint the appellant alleged a custom of his caste, resident at Vallalapatti and certain other villages, in the terms set out in the judgment of the Judicial Committee. By para. 7 he relied on an agreement which " following the custom " was executed by his father in 1885 before marrying his second wife, stipulating to give for the jeshtabhagam of his male children by his first wife certain [@ page LRIA 540] properties and Rs. 500, and to divide and give a moiety of the remaining family properties. He prayed to recover the properties named as jeshtabhagam, also Rs. 500 for moopu, and to have it declared that he was entitled to a moiety of the remainder of the property he also prayed that, if the Court should be of opinion that the father was entitled to maintenance, a direction for payment of a one-fifth share or otherwise. The respondents by their written statements denied the appellants right to a larger share than that allowed on partition by the ordinary Hindu law. Both Courts in India held that the agreement, not being registered, was not enforceable; also that the appellant was entitled to recover moopu. The present appeal was confined to the validity and effect of the custom with regard to which oral and documentary evidence was adduced at the trial. The Subordinate Judge held that there was a custom of the caste whereby upon a second marriage the family property became divisible equally between the male issue of the two wives; he however was of opinion that the father could not be excluded from a share. He declared that the plaintiff was entitled to a one-third share. The Subordinate Judge held that there was a custom of the caste whereby upon a second marriage the family property became divisible equally between the male issue of the two wives; he however was of opinion that the father could not be excluded from a share. He declared that the plaintiff was entitled to a one-third share. The half-brothers (the present respondents 2 to 4) appealed to the High Court, and the plaintiff (the present appellant) filed cross-objections under Order xli., r. 22, maintaining his right under the custom to a moiety of the property. The appeal was allowed, it being held that the plaintiff was entitled to a one-sixth share only ; in other respects the decree of the trial judge was affirmed. The learned judges (Spencer and Sankara Nair JJ.) came to the conclusion " that division according to the number of wives having sons has not been established as a custom, and cannot be enforced in this case." 1921. April 28, 29 ; May 2. Macquisten K.C. and Ingram for the appellant. The evidence established that there was a caste custom in the villages referred to whereby on a second [@ page LRIA 541] marriage the property was divisible according to wives. In the various instances proved effect had been given to that custom; no instance to the contrary was proved. A custom of this nature is not unknown in Hindu law, and is prevalent in Southern India Sumrun v. Khedun ((1814) 2 S. D. A. 116, 147.) ; Temmakal v. Subhammal (( 1864) 2 Madr. H. C. 47.) ; Stranges Hindu Law, vol. ii., p. 237; Maynes Hindu Law, 8th ed., para. 473w. Both Courts upheld the custom as to the moopu, and the custom with regard to the division of the rest of the property rested on the same evidence. It is true that there is no record of a division in the lifetime of the father, but the appellant was entitled to partition and the custom has to be applied. The plaint conceded the fathers right to maintenance. In any case the appellant was entitled to a one-fifth, not a one-sixth, share. It is true that there is no record of a division in the lifetime of the father, but the appellant was entitled to partition and the custom has to be applied. The plaint conceded the fathers right to maintenance. In any case the appellant was entitled to a one-fifth, not a one-sixth, share. Another sharer could not be introduced after suit ; the institution of the suit operated as a partition Suraj Narain v. Ikbal Narain (( 1912) L. R. 40 I. A. 40.); Girja Bai v. Sadashiv Dundhiraj (( 1916) L. R. 43 I. A. 151.) ; Kawal Nain y. Prabhu Lai. (( 1917) L. R. 44 I. A. 159.) De Gruyther K.C. and Dube for the respondents. The custom contended for is not in accordance with that pleaded, which ought to have been strictly proved. In any case the evidence did not establish any custom affecting the shares to which the parties were entitled on partition. A custom varying the ordinary law must be proved by clear and unambiguous evidence Ramalakshmi Ammal v. Sivantha Perumal Sethurayar (( 1872) 14 Moo. I. A. 570, 585.) ; Abdul Hussein Khan v. Bibi Sona Dero. (( 1917) L. R. 45 I. A. 10, 15.) The custom relied on here differs so widely from ordinary Hindu law that particularly clear proof is requisite. Only four instances of divisions in accordance with the custom were attempted to be proved, of which two only were sup ported by documents. The agreements relied on did not show that the property referred to was actually divided in accordance with them, nor was it clear that the property referred to was joint property. The agreements really indicate that there [@ page LRIA 542] was no binding custom Rama Nand v. Surgiani. (( 1894) I. L. R. 16 A. 221.) The code of rules drawn up by the Nagarathars indicates at most that some such division of family property when a second wife was taken was regarded as a matter of social propriety. There is no reported case in which a custom of patnibhaga has been proved. In Mootoo Vengada Chellasamy v. Toombayasamy (( 1849) 1 S. D. A. (Madr.) 27, 30.), decided by the Madras Sadr Diwani Adalat in 1849, and coming from the Tinnevelly district, which adjoins Madura, the custom was held not to be deserving of recognition in that part of India. In Mootoo Vengada Chellasamy v. Toombayasamy (( 1849) 1 S. D. A. (Madr.) 27, 30.), decided by the Madras Sadr Diwani Adalat in 1849, and coming from the Tinnevelly district, which adjoins Madura, the custom was held not to be deserving of recognition in that part of India. So far as the agreements show a provision for the maintenance of the first wife by the managing member they are consistent with Hindu law. A " moopu " in the case of a Hindu marrying a second wife is an ancient part of Hindu law see Bannerjees Hindu Law of Marriage and Stridhan, p. 136 ; Stranges Hindu Law, vol. i., p. 52 ; Mitakshara 2.11.2. It is for that reason, and because of the small sum involved, that the allowance of the moopu is not contested on this appeal. Macquisten K.C. in reply. In the case before the Sadr Diwani in 1849 there appears to have been no evidence given of a custom in the family or caste. [As to proof of custom reference was made to the Indian Evidence Act (I. of 1872), s. 13, and Kokla v. Piari Lai. (( 1913) I. L. R. 35 A. 502.)] June 9. The judgment of their Lordships was delivered by LORD PHILLIMORE. The plaintiff in this case, the present appellant, instituted a suit for partition of family property against his father and three half-brothers. During the pendency of this suit, a fourth half-brother was born, and all these are defendants and respondents. The right of the plaintiff to have partition was not seriously questioned; the dispute concerns the extent of his share. The Subordinate Judge decreed to him a third share, but the High Court only gave him one-sixth, on the theory that the father would have one share, the plaintiff one, and each of his four step-brothers, one. There is a possible third view that the plaintiff should [@ page LRIA 543] have one-fifth, his youngest born step-brother not being counted for a share, as having been born since the unequivocal statement by the plaintiff of his desire to have partition. But in the circumstances, their Lordships do not find it necessary to pronounce upon this contention. There is a possible third view that the plaintiff should [@ page LRIA 543] have one-fifth, his youngest born step-brother not being counted for a share, as having been born since the unequivocal statement by the plaintiff of his desire to have partition. But in the circumstances, their Lordships do not find it necessary to pronounce upon this contention. It is not disputed that, according to the ordinary Hindu law of the Mitakshara, upon the death of the father and a subsequent partition, the five children of the two marriages would each take an equal share, and that if there were a partition during the fathers lifetime, he would count as one with the five, so that the shares would be in sixes. The appellant has contended before their Lordships that by the usage and custom of the sub-caste to which he belongs, the children of each wife take as a unit and sub-divide their share among themselves, so that he as the only son of the first marriage would, if is father were dead, be entitled to half, and each of the four half-brothers to an eighth. Applying this principle to partition during the fathers lifetime, he now claims to divide the property into three shares, and take one for himself, leaving one-third for his four half-brothers, and one-third for his father. In addition, before the property was divided he claimed the sum of Rs. 600 as a first charge in his favour on the whole property as representing the moopu provided according to the caste custom for a first wife when her husband married a second time, and descending from her to her son or sons. He did not in his plaint state the custom as to application on partition in the precise form in which he insisted upon it in his evidence, and in which it was found in his favour by the Subordinate Judge, and it has been contended by the respondents that this variation is fatal to his case. The paragraphs of his plaint which relate to the partition are the following " 6. The paragraphs of his plaint which relate to the partition are the following " 6. It is the long standing custom obtaining among the people of the said caste, that any one of the said caste desiring to marry a second wife during the lifetime of the first wife gets the assent of the first wife for it and sets apart certain properties suitably to his position for [@ page LRIA 544] the jeshtabhagam of the male children by first wife born or to be born, and divides the properties, giving one share to all the male children together by the first wife on one side and a share equal thereto to all the male children together by second wife on the other, irrespective of one wife begetting a greater, and the other a less, number of male children. " 13. According to the custom of caste, 1st defendant should live with his children by his second wife. Plaintiffs have no objection to make proper arrangements for him out of the common properties till his death if he were to claim separate maintenance. Plaintiffs believe that Rs. 20 per mensem will be a sufficient allowance for him." He also relied in para. 7 upon an agreement executed by his father on the occasion of his second marriage, which he asserted was made in pursuance of the custom by which the father stipulated that he would give for the jeshtabhagam, which seems practically the same thing as the moopu, the Rs. 600, or their equivalent, and further stipulated to divide the remaining family property between the families of his two wives, giving each a moiety. When it came to giving evidence it appeared that there were few known instances of partition during the fathers lifetime, and that there was a lack of authority for the contention that the father was in such a case reduced to maintenance, as contended by the plaintiff in para. 13 of his plaint. The evidence offered mainly related to division after the fathers death and to the caste custom applying in this event. 13 of his plaint. The evidence offered mainly related to division after the fathers death and to the caste custom applying in this event. It appears to their Lordships that the case in the Court of first instance was conducted upon this footing, and that what the Court was asked to decide was whether the caste custom as to division after the fathers death was or was not a proved and binding custom ; and that the mode of partition during the lifetime of the father was treated as consequential upon the custom, so that the plaintiff would be not precluded by the language of his plaint from proving the custom which would operate upon the death of the father by reason of his having carried his claim too far and endeavoured to reduce [@ page LRIA 545] the fathers interest upon a partition in his lifetime to one of maintenance only. The Subordinate Judge who had control of the matter and could have allowed an amendment in the pleading, if it were necessary, and if no injustice was thereby done, stated among others the following issues " (1) Whether the custom set up by the plaintiff is true and valid ; (8) To what share is the plaintiff entitled ? " He found the first issue in the affirmative, and as to the eighth issue he said " 8th Issue — The division is as per number of wives having sons. This will naturally be after the husbands death. In some cases to which plaintiffs exhibits relate, the husband has made a division retaining nothing to him. It is open to him not to claim a share, but where the son enforces a partition, it can never be said that the father is not entitled to a share. The number of shares will be taken to be the number of wives having sons plus the father, if he is alive. So that in this case the property will have to be divided into three equal shares, of which the first plaintiff will be entitled to one-third, first defendant one-third, and defendants 2 to 4 and 24th, one-third, first plaintiff is therefore entitled to one-third share." No doubt in their memorandum of appeal to the High Court the defendants took the objection that " the custom set up in paras. 6 and 13 of the plaint could not be split up piecemeal, and that the whole must stand or fall together." But there is no trace in the judgment of the High Court of this view having been taken by the learned judges in that Court. Their Lordships therefore think that it was open for the appellant to contend at their bar, as he did, that the effect of the custom would entitle him on a partition effected during his fathers lifetime to one-third. This being so, it is necessary to look somewhat closely into the custom and to the evidence given in support of it, evidence which was accepted by the Subordinate Judge, but considered insufficient by the High Court. That the two modes of division between sons are both [@ page LRIA 546] known in Hindu law is unquestionable. There are appropriate words for them. When the division is by number of sons, it is called putrabhaga, when the division is according to wives it is known as patnibhaga. That putrabhaga is now the recognized rule of Hindu law is not to be questioned ; but there are traces of the other view. In the appendix to Stranges Hindu Law, vol. ii., p. 351, the answer of the pundit to whom the case concerning the Zillah of Chittore was referred, states that there is much dispute in books as to which is the true view of division ; and the pundit in that case proceeded to decide in favour of the rule of division by wives as being the law in the superior castes and the custom in the vaisya and sudra castes. The parties in the particular case were, as it happened, sudras. This decision, as a statement of the general Hindu law, was incorrect, as is pointed out by Mr. Colebrooke and Mr. Ellis in their notes on the case ; but both these writers agree that there might be such a custom, and that it would support the pundits opinion. Strange himself, in the first volume of his book, the first edition of which was in 1825 and the second in 1830, speaks of the two modes—patnibhaga, or division by wives, and putrabhaga, or division by sons. Strange himself, in the first volume of his book, the first edition of which was in 1825 and the second in 1830, speaks of the two modes—patnibhaga, or division by wives, and putrabhaga, or division by sons. He speaks of the first as an " unnatural division " and says it "is therefore allowed only among sudras, nor among them but where there is a custom for it, which must, of course, be strictly proved, though it is said to prevail in the Southern territories of India as much did formerly the custom of gavelkind in Kent, thus to a certain extent but still in the sudra class only superseding the law of the sastras ; and to this opinion the frequency with which references of the kind appear to have been made in the courts of the company in the Peninsula seems to give countenance." There being this divergence of thought, it is not wonderful, that in a land where there are so many customs, appropriate to certain areas of territory, families, or castes, though the prevailing law is that of putrabhaga there should be in certain cases a customary law of patnibhaga. As was observed by [@ page LRIA 547] their Lordships in the case of the Collector of Madura v. Mootoo Ramalinga Sathupathy (( 1868) 12 Moo. I. A. 397, 436.) "Under the Hindu system of law clear proof of usage will outweigh the written test of the law " Mayne in his Hindu Law, para. 473, states the general law as to the right to shares which pass by survivorship as follows "Each class will take per stirpes as regards every other class, but the members of the class take per capita as regards each other. This rule applies equally whether the sons are all by the same wife or by different wives." But in a note he adds "In some families, however, a custom called patnibhaga prevails of dividing according to mothers, so that if A had two sons by his wife B and three sons by C, the property would be divided into moieties, one going to the sons by B, and the other to the sons by C Sumrum v. Khedun. (2 S. D. A. 116, 147.) This practice prevails locally in Oudh, as evidenced by numerous wajib-ul-arz, which I have seen in cases under appeal to the Privy Council." The case before the Sadr Diwani Adalat to which Mayne refers carries him but a little way. It was decided in 1814, and held that two instances of the alleged peculiar usage of the family in which distribution had been regulated by the number of wives were not enough to prove the custom and that the general Hindu law must prevail. However, it shows the existence of the idea. Their Lordships have not been referred to any other authority for the existence of the custom in Northern India, but as regards Southern India, it seems fairly prevalent. Mr. Ellis, on p. 357 of vol. ii. of Strange, is quoted as speaking strongly of the prevalence of the custom in many parts of Southern India; and at p. 167, a paper of his, written in 1812, is set out, in which he gives certain deviations by the Dravidian people from the ordinary Hindu law, the third of which is as follows "The division of estates in case of one person having several families by different women among the families in equal shares without reference to the number of persons ?n each.” He explains all these deviations as showing that [@ page LRIA 548] though the Brahmans were " successful in extirpating the aboriginal religion of the South .... they succeeded but partially in introducing the laws of the smritis, and were obliged to permit many inveterate practices to continue....." It is possibl that the matriarchal theories of the earlier inhabitants of Southern India may have led to the prevalence of this custom and caused the difficulty in the way of its being extirpated by the Brahmans. If this theory were sound it would naturally lead us to expect that the extirpation of the custom would be less effective in the lower castes. In the case of Temmakal v. Subhammal (2 Madr. H. C. 47.), decided in 1864, it was held to be within the power of a guardian to refer to the panchayat the question which of the two principles of division should apply. Incidentally it may be mentioned that the panchayat held that the division should be by mothers. In the case of Temmakal v. Subhammal (2 Madr. H. C. 47.), decided in 1864, it was held to be within the power of a guardian to refer to the panchayat the question which of the two principles of division should apply. Incidentally it may be mentioned that the panchayat held that the division should be by mothers. Their Lordships therefore have to approach the evidence in this case with a knowledge that such a custom does exist, and was not an improbable one in the particular case, the parties coming from Southern India and belonging to the sub-caste of Chettis. The Chettis are generally deemed to be sudras. The judgment in the High Court in this case describes the parties as vaisyas, but apparently without any foundation for this in the evidence. The explanation for this may be, as was suggested at the bar, that the Chettis now forming a prosperous class of the community are gradually claiming to be considered as vaisyas ; but whether they are vaisyas or sudras does not make much difference for the purpose of considering the probability of the custom. The evidence offered for the plaintiff was to the effect that in seven villages, inhabited by this particular class of Chettis, there were several peculiar customs, two of them relating to the case where a man marries a second wife in the life of the first. One custom is that the first wife is entitled to have some property set aside for her maintenance which would descend to her son if she had one, and is then called moopu. The other, that the property is upon the second marriage, [@ page LRIA 549] notionally divided, one moiety going to the sons of the first wife, and the other to the sons of the second. The custom further appears to have put some restriction on the liberty of a second marriage, and these marriages seem infrequent where the first wife has already a living son. It appears to be usual to execute an agreement or settlement, with the consent of the relatives and nagarathars or villagers, represented by the heads of houses, providing for the moopu, and at the same time stating that the property will be or is divided in moieties between the two families. It appears to be usual to execute an agreement or settlement, with the consent of the relatives and nagarathars or villagers, represented by the heads of houses, providing for the moopu, and at the same time stating that the property will be or is divided in moieties between the two families. In this particular case, such an agreement or deed of settlement was produced and proved, and the plaintiff did alternatively base his case upon this agreement. It was said, however, and for the present purpose it may be taken, that the document. being unregistered, could not be enforced as a conveyance. This document, however (exhibit C), is only one of a series of documents lettered to M, where provisions of a like character were made. The earliest is exhibit M, and goes back to 1864. In that case the first wife had a son, but she was considered to be incapable of carrying on household affairs. The document proceeds to show that her relations had been consulted and that with their consent and with the consent of the wife the husband was building a separate house for her and her son, arranging to mess with each wife in turn, to allow the first wife and her son to take certain lands and goods as stridhanam, and that the rest of the mans property was to be divided one moiety to the first wife and the other moiety to the second. The last document is exhibit B, in 1900. Here the first wife had no children, and it is provided that there should be certain lands for her moopu, that should she have a male child, the land shall be appropriated for moopu, which apparently means that her son would succeed to it, and that the issues of the two marriages should take the properties remaining after excluding the moopu lands, in equal shares. Various comments were made upon these documents by counsel for the respondents. Various comments were made upon these documents by counsel for the respondents. It was suggested that several [@ page LRIA 550] of them were mere agreements, and that it did not follow that they had ever been carried out that where there was no son by the first marriage, the father, if at the moment separate from his family, was dominus of his property, and could arrange to placate his first wife by providing for her possible issue it did not however appear that the father was separate from the other members of his family and in some cases he certainly was not. Then it was suggested that the very existence of these agreements proved that there was no legal custom, that it was a mere social usage, and matter of propriety among the inhabitants of the seven villages that a partition of this kind should be made, and therefore it was occasionally made by people who wished to stand well with their neighbours. But on the other hand, the witnesses who speak to these documents also give oral evidence to the effect that it was and had always been the custom in their villages that property should be divided in this way. Two special settlements showing a different arrangement were produced by the defendants, but the cases were special, being cases where a senior wile adopted the child of a junior wife, and provision might have to be made for preventing the adopted son from getting a double portion. One settlement was put in where there was only a conditional moopu and no provision as to the sons. In this case there was no son by the first wife, and it was suggested that perhaps she had passed the age of probable child-bearing. But with these exceptions there were no documents pointing to a different custom. As to the oral evidence the case in favour of the custom largely outweighed the evidence against it. Except the defendant, the father, no witnesses spoke without qualification to the negative. The father had executed this particular deed, and tried to get out of it by saying that it was not meant to be acted upon, but as a blind to induce his wife to consent to the second marriage. Except the defendant, the father, no witnesses spoke without qualification to the negative. The father had executed this particular deed, and tried to get out of it by saying that it was not meant to be acted upon, but as a blind to induce his wife to consent to the second marriage. One of the defendants witnesses said, " I do not know of any instance of partition according to the number of sons in all my seventy-five years experience. [@ page LRIA 551] I do not even remember to have heard of division by the number of sons." As regards the argument that the existence of the several agreements shows that without them the law would have been otherwise, their Lordships on consideration are not inclined to attach much importance to it. It would be necessary to fix the moopu ; and this being so, it would be convenient at the same time that the settler should state his recognition of the custom. At any rate, the inference from the existence of settlements, that settlements are required, is not enough to outweigh the very positive evidence of the custom. In the High Court it was said that there was no evidence that among the Chettis where settlements were not made division of property among male children per stirpes was observed ; but the learned judges must have omitted to notice some of the positive evidence given on both sides. Weight was also attached by them to a document drawn up in 1898 called an agreement between the nagarathars of the seven villages, stating various matters of conduct and of business on which they desired to come to an agreement among themselves, such as that they would refer all their disputes to the nagarathars and not have recourse to the magistrates or even to the village panchayat. One provision is thus expressed " Should any wish to take a second wife during the lifetime of his first wife, he shall do so after informing the nagarathars of his village and after making sufficient provision for maintenance of his first wife. One provision is thus expressed " Should any wish to take a second wife during the lifetime of his first wife, he shall do so after informing the nagarathars of his village and after making sufficient provision for maintenance of his first wife. Any violation of this rule shall be communicated by the nagarathars of that locality to the nagarathars of all seven villages who shall all assemble at Navinapatti, and the party shall abide by their decision." This no doubt does not state the custom now sought to be proved ; but the document is not a record of laws, but a provision de futuro as to social conduct ; and one of the witnesses says that the rules as to moopu and division were so well known that it was not necessary to express them. The High Court have treated the case as if it were an [@ page LRIA 552] attempt to set up a local custom, and say it would be unreasonable to impose it upon all persons dwelling in the area. But their Lordships conceive that the custom is one of the particular class of Chettis, who happen, it is true, to dwell in and probably are at the moment the only dwellers in the area of the seven villages. It was pointed out for the appellant that when the High Court came to consider the second matter, said to be proved by custom—namely, the giving of moopu—they thought that it was proved by the evidence oral and documentary. For the respondents, it was said that an arrangement similar to moopu was known to ordinary Hindu law. This may or may not be so; but the High Court considered moopu as established by custom, and the evidence for the one matter is substantially as strong as for the other. There is a curious possible effect of the custom upon general Hindu law, which may have some day to be considered. Upon a partition during the fathers lifetime, the general Hindu rule is that he gets a share equal to that of one son, which if the partition were by sons would be one-sixth. The effect, in the view taken by the Subordinate Judge, of a division patnibhaga is that the father counts for one share, and the children of each wife for one share, and so he gets one-third. The effect, in the view taken by the Subordinate Judge, of a division patnibhaga is that the father counts for one share, and the children of each wife for one share, and so he gets one-third. Their Lordships, however, have not to determine this point. If the division is patnibhaga, the plaintiff as the only son of the first wife is certainly entitled to one-third. Their Lordships think that he proved the custom of patnibhaga, and that he was entitled first to the moopu (which the respondents do not now question), and secondly to a third share of the residue. Their Lordships will therefore humbly advise His Majesty that the decree of the High Court be reversed and the decree of the Subordinate Judge restored, and that the appellant have the costs in the High Court and in the Privy Council.