Mookerjee and Cuming, JJ.:— The subject matter of each of the three litigations, which have culminated in these appeals, formed part of the estate of one Taraprasanna Roy, brother of the appellant Sarada Prasanna Roy. Tara Prasanna made a testamentary disposition of his properties on the 3rd July, 1914, and died on the 3rd October, 1914. He left a widow, Trilokmohini Debi, who executed three conveyances on the 10th September, 1915, 25th April, 1916 and 12th July, 1916, in favour of Manmohan Pande, and dealt with the estate left by her husband on the footing that she was under the will absolutely entitled thereto. The three conveyances constitute the foundation of the title set up by the plaintiffs. They alleged that the first defendant Sarada Prasanna had unlawfully kept them out of possession of the properties vested in them by virtue of the conveyances. The defendant resisted the claim on a variety of grounds which need not be enumerated in detail for our present purpose. It is sufficient to mention two of the defences which were unsuccessfully urged in the Court below and have been reiterated in this Court, namely, first, that Tara Prasanna was governed by the Mitakshara law and was not competent to make a testamentary disposition of his estate which was held jointly by him with his brother and upon his death passed to him by survivorship; and, secondly, that in respect of a property called Rautara, the brothers obtained it as a testamentary gift from their paternal grandfather and held it as joint tenants with right of survivorship. The Subordinate Judge has overruled these contentions and decreed the suits. Upon the first point he has found, first, that Tara Prasanna was governed by the Dayabhaga law and was competent to make a testamentary disposition in favour of his wife; and secondly, that even if he were held to be governed by the Mitakshara law, he could make such disposition, as he was not joint with his brother. Upon the second point, the Subordinate Judge has found that, in respect of Rautara, the brothers were not joint tenants with right of survivorship. On the present appeal, the questions mentioned have been argued with great elaboration, and our attention has been drawn to all the relevant materials on the record. We have minutely considered the evidence, in spite of the obvious imperfections of the paper book, and we shall now state our conclusions with the reasons therefor.

The first question we have to take up for consideration, is, whether Tara Prasanna was governed by the Dayabhaga or the Mitakshara law. He was resident in Bengal and consequently we start with the initial presumption that he was governed by the Dayabhaga law; for it is well settled that a Hindu family, residing in a particular province of India, is presumed to be governed by the law of the place where it resides: Ramdas v. Chandra . This presumption is rebutted where the family is shown to have migrated from one province to another; the presumption then arises that the family carried with it the laws and customs as to succession and family relation prevailing in the province from which it came: Soorendro Nath v. Heeramonee , Parbati Kumari v. Jagadis Chunder . In the case before us, the existence of tradition has been established by evidence, that in the time of Pratapaditya, that is towards the end of the sixteenth century, this family, together with several others, migrated to Bengal from a place in the North-West Provinces where the Mitakshara law, as modified in the Mithila School, prevailed. Consequently, the initial presumption is displaced, and the presumption arises that Tara Prasanna was governed by the Mitakshara law, using that expression, for the sake of brevity, as indicating for the purposes of this case the Mitakshara law as modified in the Mithila School. The burden thus shifts, and now lies upon the plaintiffs to prove that the family, after migration, has adopted the law and usages of the place to which it has migrated: Govind Chandra v. Radha Kristo , Jagannath v. Narayan , Mailathi v. Subba Raya , Kuladaprasad v. Haripada .

The genealogical table marked “A” and annexed herewith, shows the ancestry of Tara Prasanna Roy and his brother Sarada Prasanna Roy. They are descended from Raghunath Roy, who had two sons Kartik Chandra and Radhamohan. Chandrasekhar, one of the sons of Radhamohan, was the paternal grandfather of the appellant and his brother. On the 13th October, 1863, Chandrasekhar instituted a suit for establishment of title to the estate of his maternal uncle, Totaram, who had died in 1825, leaving a widow, Subhadra, who succeeded to her husband's estate and died in 1851. The genealogical table of the family of the mother of Chandrasekhar is marked “B” (1) and is annexed herewith. In the litigation commenced by Chandrasekhar the question arose whether the deceased and his nephew were governed by the Dayabhaga law or by the Mitakshara law. Chandrasekhar asserted that the Dayabhaga law had been adopted by the families, and his contention was upheld. The judgment delivered by Bay ley and Sambhunath Pandit, JJ. in that suit on the 20th February, 1865, has found its way into the reports and is regarded as a leading authority on the subject of change of personal law by migration: Chundrosekhar v. Nobin Soondur . Reference is made in that decision to an earlier judgment in a litigation commenced by Abhoy Charan, the son of the daughter of Baidyanath, brother of the very Totaram who was the maternal uncle of Chandrasekhar. The decision in the suit by Abhoy Charan was pronounced on the 30th December, 1862, by Raikes, Seton-Karr and Jackson, JJ., and has also found its way into the reports: Ootum v. Abhoy Charan . In the suit by Abhoy Charan, an opinion had been expressed that the evidence did not establish that the family of Totaram had abandoned the Mithila law in favour of the Dayabhaga law: Bayley and Pandit, JJ., considered the grounds for this view, and held, on the fuller materials, produced before them, that the Dayabhaga law governed the parties. There is one passage in the judgment which may usefully be recalled here:

“The plaintiff has clearly shewn by the evidence of friends and relatives that in the family of the parties before the Court, in several cases where succession, according to the Bengal law, was to go to one, and according to the school of Mitakshara to another, the succession was according to the former or Bengal School. It is proved that the ancestors of these parties migrated into this country about twelve or thirteen generations ago, with a few of their own priests. It is quite natural that some among them may still have for their priests descendants of the original family priests; as the descendants of these priests cannot find it convenient to attend upon and accompany all the numerous descendants of the original emigrants scattered by degrees in different places from time to time, to many of the later generations, it is only left to accept the Bengali priests available. Thus too, the descendants of the family priests by degrees, after a few generations, acquiesced in performing the external ceremonies (or kria) according to the Bengal School. In such cases the Nagri character is abandoned for Sanskrit writing after 5 or 6 generations; old habits and customs fall one by one; and the priests and their jujmans both gradually become Bengalis in every sense. In this state of change for a long time, even after a complete conversion, the priests and their jujmans remain in theory a separate class, though for all other practical purposes, either of ceremonies, or of succession, they have long adopted the Bengali School of the Hindu law.

“Such emigrants may become Bengalis in their ceremonies, but cannot and would not be allowed to intermarry with Bengalis of their own class. The fact, therefore, of want of proof of intermarriage with Bengalis, or the fact of some families of the descendants of these original emigrants having still for their priests some of the descendants of the priests that had accompanied their ancestors, does not afford proof in support of the fact of the original laws and customs being still observed by these people as a class and tribe. The witnesses for the plaintiff distinctly state that the necessary rites in the families of both the parties are now performed according to the Bengali School; and, what is more important, they distinctly swear to the fact that within their knowledge succession is guided by the same school. The evidence given by the defendant to counteract this case of the plaintiff is the deposition of two witnesses who are not of the family of the tribe to which both the parties belong, and who distinctly admit that they are ignorant of the school of the Hindu law under which succession is guided in the family. It is to be borne in mind that the fact of some members of the family having Bengali priests is by itself a strong proof of all having adopted Bengali ceremonies for all essential ceremonies, though it is admitted that these persons of different branches of the original family intermarry only among themselves. It is clear that generally there can be no intermarriage between two classes of people who perform their essential rites, such as marriages, funerals, tonsure, etc., under two different schools of Hindu law. When these people are found to have lived here so long as to have by degrees adopted the language, the dress (particularly for women), and the ordinary food for the people of Bengal—the two last of which are materially different in the two countries among such people, when all intermarriage with the parent country has long ceased—it is nothing surprising that they should be proved, as deposed to by competent witnesses, to have by degrees been compelled to abandon their old forms of ceremonies now impracticable, and what is of more importance, of having, whenever such a contingency arose that one more nearly allied was not the heir by their old law,—allowed him or her to succeed according to the laws of Bengal.

“The cases of succession quoted by the witnesses of plaintiff do not, as to the rights of the daughter's son, show sufficiently clear proof of the statement of the plaintiff with regard to the application of the law of Bengal, as these witnesses were not made to state that, when such succession took place to the deceased, he was a member of a joint family. There cannot, however, be any mistake regarding the succession of a sister's son. He has no right to succeed except under the law of the Bengal School. This was understood to be the law before the decision of the Judicial Committee in Girdhari v. Bengal Government and of the Full Bench in Amrita v. Lakhi which, explained away the contrary decision of the Judicial Committee in Thakoorain v. Mohun . It is important that more than two instances of such succession were pointed out by the witnesses. The defendant has not contradicted the fact, or proved that, in any other similar cases of succession, any other rule of succession was adopted. The instances given by the witnesses took place in the class to which both the parties belong, and the statement of the plaintiff is corroborated by the fact of Subhadra's undisputed succession in that very branch of the family of which Tota and his brother were the members. Some attempt was made to raise doubts regarding the correctness of the claim of the plaintiff, on the ground that he asserted wrong facts in the case, and did not venture to bring his claim for more than 10 year after the death of Subhadra, and that he had pleaded otherwise in the other case brought against Nobin. But in the face of such clear proof of the law of succession applicable to the family as now found by evidence, we cannot attach any importance to this objection. It is not at all surprising to find that plaintiff at first entertained doubts regarding his ultimate success, and so delayed to sue for that or other reasons.”

These observations may be taken along with the apt comment of Babu Kishenkishore Ghose, who argued on behalf of the respondent in the case of Ootum v. Abhoy , that families situated like that of the defendant commonly use the customs of Bengal for everyday purposes and produce the law of Mithila for their law suits. The weight of the decision in Chundro Seekhur v. Nobin was fully realised by the legal advisers of the defendant, and an endeavour was accordingly made to minimise its effect by the contention that it is at best a decision merely that the family of the maternal uncle of Chandrasekhar was governed by the Dayabhaga law. In support of this position, reference has been made to a passage in the judgment of O'Kinealy and Hill, JJ., in the case of Parbati Kumari v. Jagadis which was ultimately affirmed by the Judicial Committee and, which, it was suggested, indicated the possibility of marriages between the members of two families, one governed by the Dayabhaga, the other governed by the Mitakshara. We need not consider, whether, even if permissible, this is probable; the fact remains that in the suit of Chandrasekhar, no such distinction was made between the families of his paternal and maternal ancestors. The judgment of the High Court makes it abundantly clear that evidence was adduced to prove that both the families had abandoned the Mitakshara in favour of the Dayabhaga, and this evidence was accepted by this Court as reliable. It may be observed parenthetically, that the contrary view might lead to unforeseen complications. The principle of mutuality applies in the determination of the question of heirship between a maternal uncle and his nephew, and if one of them were governed by the Mitakshara and the other by the Dayabhaga, the result would be very different from what would be the case if both were governed by the Dayabhaga. The appellant has contended that the recitals and findings in the judgment of this Court in the suit of Chandrasekhar do not conclude him; the answer is that the respondents do not rely upon the judgment for such purpose. Assume for the moment that the principle recognised in Kashinath v. Jagat Kishore and Tripurana v. Rokkam applies, and the recitals in the judgment cannot be used as evidence, still, the judgment is evidence as a relevant fact in issue or as a transaction. The judgment so used shows that the question was raised and decided, whether the families of the paternal and maternal ancestors of Chandrasekhar were or were not governed by the Dayabhaga law. The evidence then accessible to the Court is no longer available, partly by reason of lapse of time, partly from destruction of portions of the record in accordance with statutory rules. There can, in our opinion, be no room for doubt as to the propriety of the course adopted by the Subordinate Judge, when he attached great weight to the circumstance that, in 1865, this Court found that the Dayabhaga law had been adopted in the families of the paternal and maternal ancestors of Chandrasekhar.

It is not a matter for surprise that Chandrasekhar, who had successfully asserted that the families of his paternal and maternal ancestors had adopted the Dayabhaga law, should have consistently adhered to that position, which had secured for him the estate of his maternal uncle after the death of his maternal aunt. On the 29th January, 1864, while his suit against Nobin Sundar was still pending in the Court of the Principal Sadar Amin of Jessore, he made a testamentary disposition of his properties. The terms of the will, which has been produced before us, leave no room for doubt that the testator made the disposition on the footing that he was subject to the Dayabhaga law and had absolute dominion over his properties, ancestral as well as self-acquired. Chandrasekhar died in 1868, and his estate was disposed of as directed by him. Many years after his death, on the 23rd February, 1876, an ekrarnama was executed between his sons Hari Prasanna and Mahendra Nath on the one hand, and the descendants of his brother Gobinda Chandra and cousin Kali Kishore on the other. The document was in essence a deed of partition amongst members of the family, and proceeded on the assumption that the testamentary disposition by Chandrasekhar was valid and operative. Along with this, we must remember that the brick built house of Chandrasekhar at Krishnagar was sold by Hari Prasanna and Mahendra Nath, long after Sarada Prasanna had grown into manhood; this is consistent only with the theory that Hari Prasanna and Mahendra Nath considered that they were competent to deal in this manner with ancestral property, as in reality they would be under the Dayabhaga law. We find, again, that on the 8th July, 1886, Hari Prasanna and Mahendra Nath executed a deed of gift in favour of Panchanan, the son of their sister Joytara; the subject-matter of the gift was the property obtained by their father Chandrasekhar as the result of his suit against Nobin Sundar already mentioned. This cannot be reconciled with the theory that the family was governed by the Mitakshara law. We find next that, on the 7th September, 1895. Hari Prasanna made a testamentary disposition of his estate. The will recognises the validity of the disposition by Chandrasekhar and proceeds to give directions consistent with the Dayabhaga and inconsistent with the Mitakshara. There is really no room for doubt that Hari Prasanna like his father Chandrasekhar, acted throughout on the assumption that the family was governed by the Dayabhaga law. Hari Prasanna died on the 1st December, 1895, and then, for the first time, Sarada Prasanna made an attempt to get up the theory that the family was still governed by the Mitakshara law. Hari Prasanna had left, by his second wife, two sons, namely, Sarada Prasanna and Tara Prasanna and, by his fourth wife six sons, namely, Guruprasanna, Jyotiprasanna, Jagatprasanna, Lalgopal Joygopal and Bejoygopal, the last four of whom were minors. There can be little doubt that Sarada Prasanna took full advantage of the situation, created disputes, and, on the 22nd December, 1895, just three weeks after his father's death, secured a reference to arbitration. Here, for the first time, we find an assertion that the members of the family were “up-country Brahmins and governed by the Mitakshara law”, coupled with an allegation that Hari Prasanna, the father of the executant, “out of misapprehension used to consider the paternal properties and the other properties acquired by the proceeds thereof as his self acquired properties”. We need not discuss what weight should be attached to a statement of this character by a person who had passed the prime of life and Lad never ventured to set up such a case so long as his father was alive. But we note that the arbitration was carried out with uncommon rapidity; in the course of three days, on the 25th December, 1895 the arbitrators made their award, which was made a rule of Court on the 10th February, 1896. The Subordinate Judge has held that this was engineered by the appellant, with a view to nullify the testamentary disposition made by his father. That would be an obvious and immediate advantage; but foundation would also be laid for a remote benefit; as Tara Prasanna was childless, Sarada Prasanna would be entitled to claim his interest by survivorship, if the family were held to be governed be the Mitakshara. In this connection, it cannot be ignored that he was an experienced member of the legal profession, while his brother and his step-brothers, four of whom were minors, were not familiar with the devious paths of law and lawyers. It is desirable to add that we are not now concerned with the validity of the settlement; the transaction is relevant, only upon the question whether the family was governed by the Dayabhaga or the Mitakshara. Upon that point, as the Subordinate Judge has held, it is relatively of little consequence, when contrasted with the lifelong and self-consistent conduct of Chandrasekhar and Hari Prasanna. We pass on now to the conduct of Tara Prasanna himself. On the 3rd July, 1914, he made a testamentary disposition of his properties, and gave directions consistent with the Dayabhaga and inconsistent with the Mitakshara. We need not, however, attach much weight to this fact, as this is the very disposition which has led up to the present litigations. There is no room for dispute that the conduct of the immediate ancestors of the appellant cannot be reconciled with the theory that the family was governed by the Mitakshara law. But our attention has been drawn to three instances of succession in the family, which, it is alleged, do not support the position that the Dayabhaga law is applicable. On examination, however, each of these turns out to be of an inconclusive character. Kalikanand survived his father, Gobinda Chandra, for less than a month (A). After his death, his widow, Katyani, contented herself with an allowance and resided, as is not uncommon, in her paternal residence. The income of her husband's share was very small and the amount paid as maintenance might have been regarded as an adequate substitute. Instances of exclusion are of value, only if there is proof of demand and refusal. The widow of Girijanand, another son of Gobinda Chandra, was excluded from inheritance by the will of her husband, which is said to have contained a provision for her maintenance, and she committed suicide within two or three years after she had become a widow (A). Nanibala, the widow of Saileswar, was a minor at the date of the present suit, and she had time yet to enforce her right of succession (A). On the other hand, there are instances of the grant of succession certificate and guardianship certificate, such as is inconsistent with the theory of the existence of a joint family governed by the Mitakshara law [Beejraj v. Bhyropersaud , Bissen v. Chatrapat , Gharibullah v. Khalak Singh ]; this happened in fact, with regard to the estates of Mahendranath and Hariprasanna after their demise (A).

The conclusion thus drawn by the Subordinate Judge from instances of succession in the family, has been fortified by him by reference to evidence of the observance of rites and ceremonies at marriages, births and deaths, which indicate the relinquishment of Mitakshara law and the adoption of the Dayabhaga law in the family. The value of such evidence was emphasised by the Sudder Court in Raj Chunder v. Gocul and by the Judicial Committee in Parbati Kumari v. Jagadis Chunder , but it was perhaps not fully appreciated in Huro Pershad v. Shibo Shunkuree . The evidence on this part of the case is conflicting, but the Subordinate Judge has come to the conclusion that the balance of testimony is in favour of the plaintiffs. There is some indication that Sarada Prasanna, since the death of his father in 1895, his endeavoured to conform to the Mitakshara rather than to the Dayabhaga; the evidence shows, his scholarly attainments in Sanskrit and his acquaintance with authoritative works of both the schools. The evidence at the same time shows, however, that the officiating priests have not much of learning, and one of them, in the stress of cross-examination, made the astonishing statement that his knowledge of Smriti of Mithila was derived from Raghunandan (the authoritative exponent of the Bengal School of Hindus law). The evidence further shows that the Maithili priests employed perform the puja in the same manner as Bengal priests. The inference is legitimate that the priests, like the members of the family, have adopted the Bengal system of rites and ceremonies, even though they were themselves Maithili in origin. The Sradh ceremony is performed according to the system prevalent in Bengal; while the attempt to prove that, the Sapindakaran ceremony, in an exceptional case was performed according to the Mithila school on the twelfth day and not according to the Bengal school on the thirteenth day, completely broke down when the fact was brought out in evidence that it was in fact performed on the anniversary of the day of death. Much stress was laid on the performance of the Tilak ceremony prior to marriage; but it cannot be overlooked that it has a singular resemblance to what is known as the Asirbad in Bengal. In our opinion, the Subordinate Judge has correctly held that as regards language, dress, food and rites and ceremonies, the members of the family have adopted the style prevalent in Bengal. This, indeed, was fully established more than half a century ago in the suit brought by Chandrasekhar for recovery of the estate of his maternal uncle; this is manifested in the passage from the judgment of Bay ley and Pandit, JJ. set out above.

We cannot leave this part of the case without some reference to the evidence which has been adduced to elucidate the history and status of what has been called connected families. Such evidence, where a family is one of a group, may be admitted under section 11 of the Indian Evidence Act, which lays down that facts not otherwise relevant are relevant if, by themselves or in connection with other facts, they make the existence or non-existence of any fact in issue or relevant fact highly probable or improbable; see Mohesh Chunder v. Satrughan where Lord Macnaghten pointed out that the rule of the custom of lineal primogeniture obtained in all the Dhalbhoom families, though some of them were governed by the Mitakshara, others by the Dayabhaga, notwithstanding that there were intermarriages between them: see also Shimbhunath v. Gayanchand , Harnabh v. Mandil . In the case before us, it is not necessary for the plaintiffs to establish that every one of the families which belong to the group that migrated towards the end of the sixteenth century had abandoned the Mitakshara in favour of the Dayabhaga; the fact that some of them have done so may lend additional support to the allegations of the plaintiff, and can be used only to supplement direct evidence, for we cannot prove the unknown by doubtful or hypothetical analogies. One of the families mentioned is that of the Missirs of Samta (B.2); a daughter of Mahendra married into that family, while two daughters of that family married Durgananda and Girijananda, the nephews of Chandrasekhar. In a litigation relating to that family, the question arose, whether the Dayabhaga or Mitakshara was the governing law; Honatun v. Huttun . Pandit, J. in remanding the case, referred to the decision in Ootum v. Abhoy and after remand, it was held by the District Judge, Mr. Rivers Thomson, on the 28th July, 1864, that the family was governed by the Dayabhaga law. Again, in the case of the family of the Chaudhuris of Panchpota (B. 7) it was held, on the 27th June, 1901, that the members were subject to the Dayabhaga and not the Mitakshara. The Subordinate Judge, Mr. Srigopal Chatterjee, pronounced an elaborate judgment and reviewed the result of litigations in connected families, including the suit of Chandrasekhar previously mentioned. This judgment was affirmed by this Court on the 8th July, 1904, by Rampini and Mitra, JJ. Patitpaban Chaudhuri, the plaintiff in that case, who unsuccessfully urged that the family was governed by the Mitakshara, is a son-in-law of the appellant. In two subsequent suits, however, in 1905 and 1907, Patitpaban succeeded in securing a contrary verdict; but it is noticeable that the judgment of the High Court in the earlier suit was not produced in the first of these cases, while it was erroneously excluded from evidence in the second case. In the family of the Rays of Gangadharpore, it was found in 1915 that the members had adopted the Dayabhaga system; two of the daughters of the appellant have married into this-family. Similar remarks apply to the Pradhan family of Samta (B.3). In that family, a succession certificate was taken out on the death of Pramatha, by his mother Nistarini, such as is permissible only if the Dayabhaga applies. These Pradhans were the agnates of the maternal uncle of Chandrasekhar; and, a daughter of the appellant has married into that family, while a daughter of that family has married a son of Mahendra, the paternal uncle of the appellant. We do not overlook, however, that the Pradhans of Piplee (B.5), the Rays of Beria (B. 6) and the Rays of Jahanabad appear to have adhered to the Mitakshara. It also appears that the Misras of Chatra belong to the same category (B.4); the mother of the appellant was a daughter of this family. We need not deal in detail with other families, which are very remotely connected, such as the family of the Rays of Lalgola. In a litigation concerning the members of that family the Mitakshara was assumed to be applicable in 1870; Mooktakeshee v. Oomabutty ; but the real question in issue there was whether the family was joint or separate. On a review of the entire evidence on this part of the case, we agree with the Subordinate Judge that the preponderance is distinctly in favour of the view maintained by the respondents. It is not necessary to decide with regard to each of the individual families, whose cases are not really before us for adjudication, whether they are governed by the Dayabhaga or the Mitakshara; it is sufficient to state that the impression left on our minds is that in the allied, or connected families, the Dayabhaga has been very widely adopted, and the Bengal system of rites and ceremonies has been generally recognised. This renders it highly probable, within the meaning of section 11 of the Indian Evidence Act, that Tara Prasanna Roy, the grandson of Chandrasekhar Roy, was governed by the Dayabhaga like his two immediate ancestors.

We have indicated above that the appellant, since the death of his father, has made an attempt to revert to the Mitakshara, which had been abrogated in the family for generations. This may raise the interesting question, whether an individual member of a family, which has migrated and adopted the law of what may be called its new domicile, may, of his choice, revert any day to the law which governed the family in its original home. Neither authority nor principle has been invoked in favour of the possibility of such reverter, except by the growth of a new family usage. Where a Hindu family migrates from one province to another, the presumption is that it carries with it the laws and customs as to succession and family relations prevailing in the province from which it came; but this presumption may be rebutted by proof that the family has adopted the law and usages of the place to which it has migrated. The reason is that the Hindu law is not a merely local law, but is essentially personal law, an integral factor of the status of every family which is governed by it. In its new domicile the family may, by the reflex action of manners and customs prevalent in resident families, consciously or unconsciously modify its own governing rules; there may thus be an acceptance of a new law, not due to sudden change by choice or agreement, but by the gradual evolution of a family usage: see Tarachand v. Reebram , Abraham v. Abraham . When a new family usage has thus grown up in the course of generations, possibly with the concurrence or acquiescence of families of the same group, it famishes the governing law of the family: see Rutcheputty v. Rajunder , Pudmavati v. Doolarsing , Soorendro Nath v. Heeramonee , Mohesh v. Satrughan , Parbati v. Jagadis , Abdur Rahim v. Halimabai , Balwant v. Baji Rao (7), Rana Mahatab v. Badan Singh (8), Kulada v. Haripada (9), Pitamoar v. Nisikanta (10). When a family usage has thus developed into a binding law, let it be conceded that it may be superseded, by the growth of a contrary usage or by legislative enactment: Hammerten v. Honey (11). But it would obviously lead to much confusion and abundant litigation, if the law permitted arbitrary attempts to revive and give effect to the original usages, after they had been clearly abandoned, and the abandonment had been acted upon so long as to result in the adoption of new usages. The view cannot consequently be maintained that it is open to a member at any time to disclaim his personal law and adopt another at his choice; if this were permissible, chaos would be the only result; it is not necessary to discuss here whether a family custom can be put an end to by the consensus of opinion of all the members or of the majority of the family: Mahomed v. Haji Ahmed (12), Rojkishen v. Ramjoy , Sarabjit v. Indarjit , Ayyakutti v. Krishna . In the case before us, there is no evidence of discontinuity of the personal law adopted by the family in the course of generations, and the growth of a new family usage, constituting its suppression and a reverter to the personal law of the family in its original domicile.

The second question we have to take up for consideration is, whether the appellant and his brother were joint or separate in estate. The Subordinate Judge has answered the question against the appellant. We are of opinion that this conclusion cannot be successfully assailed. The award of the 25th December, 1895 clearly effected a severance of interest among the eight sons left by Hari Prasanna Roy. The award was made pursuant to the ekrarnama of the 22nd December, 1895 and was transformed into a decree of Court on the 10th February, 1896. The contention that the arbitrators acted in excess of their authority is futile. The division they effected was within the scope of their authority; but if they really exceeded their powers, the appellant should have objected when his step-brothers enforced the award under sections 525 and 526 of the Code of Civil Procedure, 1882. There was then not only no opposition on the part of the appellant, the decree was in fact made with his consent; and it cannot now be impeached as irregularly or improperly obtained. The Subordinate Judge has further held on the oral and documentary evidence, that the decree based on the award has been carried into effect. The appellant did not thereafter live with his brother or step-brothers as members of a joint family. The income has been separately enjoyed, in defined shares, though the properties may not all have been partitioned by metes and bounds. The evidence has been placed before us, and has been minutely commented upon. We see no reason to doubt that the Subordinate Judge has correctly estimated the effect of the evidence on this part of the case. There is no room for controversy that according to the principle enunciated by the Judicial Committee in a long series of decision, a severance was effected in this case; and there is neither allegation nor proof of a subsequent reunion. As stated by Lord Westbury in Appovier v. Rama Sabba the true test of partition of property is the intention of the members of the family to become separate owners. Consequently, there may be partition, even though there be no actual division by metes and bounds. This has been repeatedly reaffirmed: see Balabux v. Rukhmabai , Balkishen v. Ramnarain , distinguished in Durga v. Balmakund , Parbati v. Naunihal , Surajnarain v. Ikbalnarain , Girja Bai v. Sadashiu Kawal v. Prabhu , Nageshar v. Ganesha , Surajbunsi v. Sheo Prashad , Lakshman v. Ram Chandra , Mudit v. Ranglal , Ajodhya v. Mahadeo (13), Anandkisore v. Daiji , Rajaratna v. Anji , Palaniammal v. Muthu Venkata .

The third question, which requires examination, is, whether the appellant and his brother were, in respect of the property known as Rautara, joint tenants with right of survivorship. The Subordinate Judge has answered this question against the appellant. He is clearly right as to the true construction of the will of Chandrasekhar. The property was given by Chandrasekhar to his grandsons, Sarada Prasanna and Tara Prasanna, as Yaluk or affectionate gift. This could not make the donees joint tenants with right of survivorship. We need refer only to the decision of the Judicial Committee in Jogeswar v. Ramchund , where Lord Watson expressed his disapproval of the importation of an extremely technical rule of English conveyancing, because, first, the principle of joint tenancy is unknown to Hindu law apart from the Mitakshara doctrine of survivorship, and, secondly, even according to the law of England, a conveyance, or an agreement to convey his or her personal interest by one of the joint-tenants operates as a severance. Reference may also be made to the text and also to the decisions in Rewun Persad v. Radha Beeby , Navroji v. Perozbai , Rai Diwali v. Patel , Kishori v. Mundra , Harprasad v. Sukhdevi , Karuppai v. Sankara , Chakkra v. Kunhi , Bhoba v. Peary . We hold accordingly that the interest of Tara Prasanna in Rautara did not pass by survivorship to the appellant, independently of the question whether the parties are governed by the Dayabhaga or the Mitakshara.

After the most careful consideration of all the materials on the record we see no escape from the conclusion that the Subordinate Judge has rightly decreed the suits and that the appeals must be dismissed with costs.

B.M.S

Appeals dismissed.