Rankin, C. J.
1. These two appeals arise out of two suits for ejectment brought by the same plaintiffs as landlords against the tenant, defendant, Nando Lal Duli.
2. The land in question is homestead or bastu land and the area is in each ease about two cottahs. The annual rent is in one case 18 annas and 9 pies and in the other case one rupee six annas and three pies.
3. The sole question for decision in each case is whether or not the tenancy of the defendant is a permanent tenancy. Both tenancies are very old and have been traced back to the time of the defendant's grandfather and grandfather's brother. Kabuliats put forward by the plaintiff dated in 1876 and 1846, have been rejected as spurious by both the Courts below. The origin of the tenancies is unknown and the finding is that they have been in existence for about a hundred years. It is proved that they have been held throughout for residential purposes; it is proved that they have been held at the same rate of rent for forty years and according to the plaintiff's case, the rent has not varied during the last 60 years. There is nothing to show that the rentals have ever been changed. The holdings are in a village and near to a road. The defendant is a labourer. The land is entirely occupied by mud-walled huts, some of which are very old and six or seven of which have been raised by the defendant according to his evidence since his father's death, some 32 years ago. The two plots of land, according to the defendant, are in one place and a way runs between them. Four or five of the huts are let out to sub-tenants so as to produce income to the defendant from which he maintains himself.
4. In these circumstances it is contended for the defendant that the tenancies are very old, that they are of unknown origin, that rent has been paid at a uniform rate for a very long time, that the tenancy is one for residential purposes and that it has passed from one generation to another in the same family by succession twice, if not three times. In view of the fact that a labourer in the position of the defendant would not be likely to have pucca or masonry buildings, it is said that the absence of such permanent structures upon the land is a circumstance which in no way goes to show that the tenancy, was not a permanent one.
5. The landlords contend on the other hand that before an inference can legitimately be drawn to the effect that the defendant has a permanent right, facts must be proved which are unequivocally referable to a right of this character and that the mere fact that a tenant has been allowed to hold homestead land for a long time without any enhancement of his rent is not a sufficient basis for an inference that he has a permanent right even when it is coupled with the circumstance that the holding has remained in the same family for three generations.
6. The principles applicable to cases of this class may be stated as follows: (1). When a person claims to hold land as a tenant under a landlord it is for him to prove the existence, the nature and the extent of the interest which the owner of the full rights has granted to him: (2) The terms of a holding as between; landlord and tenant must in these cases be a matter of contract, either expressed or implied (3). The legislature, as regards this province, has regulated the terms of agricultural holdings. The letting of land for residential, purposes is regulated by the Transfer of Property Act of 1882, but from the operation of this statute old tenancies, such as those now in question, are excluded by Section 2. (4). Ordinarily the person who sets up a contract will be required to give reasonable particulars and direct proof of the contract relied upon, but in the case of tenancies proved to be of long standing this principle is inapplicable, and from the history of the tenancy and the circumstances of the case it is open to the tenant to show that the origin of the tenancy being unknown the correct inference is to the effect that the right granted to the tenant and enjoyed by him is a permanent right.
7. In addition to these principles of law we have to bear in mind certain general considerations of fact applicable to Indian conditions prior to the Transfer of Property Act that is to a period which ended. about fifty years ago: (a) It was not unusual and is in no way incredible either that an owner of land should mean to give a permanent right for residential purposes to a tenant in an Indian village or that the tenant should be content to take such a right by mere oral agreement coupled with possession. (b). The fact that a tenancy was for residential purposes in no way involved of itself that the tenant's right to the land was to be permanent as the land could be used for kutcha structures to be ereeted by the tenant, (c) In ordinary circumstances documents are difficult to preserve in India and in a case like the present are little likely to survive for a hundred years (d) Lastly there is the consideration remarked upon by Chakravarti, J., in Abdul Hakim v. Elahi Baksh : AIR1925Cal309
The fact that a tenant is allowed to continue in possession of land for generations without alteration of the rent is a common occurrence in this country and is usually attributable to the reluctance of a landlord to eject a tenant from his home so long as he does not make himself objectionable and regularly pays his rent.
8. Quite recently it was held by the Judicial Committee of the Privy Council in Dhannamal v. Motisagar . that the question whether a tenancy is permanent is not a question of fact upon which the opinion of the lower appellate Court is conclusive in second appeal.
It is clear that the proper effect of a proved fact is a question of law and the question whether a tenancy is permanent or precarious seems to them, in a case like the present to be a legal inference from facts and not itself a question of fact. The High Court has described the question here as a mixed question of law and fact a phrase not unhappy if it carries with it the warning that in so far as it depends on fact the finding of the Court of first appeal must be accepted.
9. This ruling must be accepted and applied: it is in consonance with much Indian authority: but it creates a difficulty in cases which are near the boundary line. In those, at all events, it is not easy to see that the ultimate inference is not itself a question of fact. As a matter of law it is clear that if the origin of the tenancy is known if, for example, it was created by an agreement which is evidenced by writing inferences from the history of the tenancy and the circumstances obtaining during the last fifty years cannot be drawn upon the same principle as are applicable to a tenancy if unknown origin. It is clear also that for the purpose of ascertaining the character of the tenancy an inference based merely upon facts which are in themselves ambiguous or equivocal, in the sense that they are not necessarily referable to the existence of a permanent right in the tenant, is an inference which is unwarranted in law. In numerous and various respects such an inference may be erroneously drawn and the error be error in law. But on the other hand, we are in this class of case dealing not with the effect of a fact upon the rights of the parties but with the inference to be drawn from subsequent facts as to the nature and extent of the right previously granted by the landlord to the tenant. For this purpose it would rather seem that the proper effect of a proved fact is a question of evidence. So long as it was thought that cases of this kind fell witi in a few readily recognizable classes it was possible to regard these cases as though the inference to be drawn could be ascertained as a matter of law. If the tenant could bring his case within one or other of these classes he was held to be entitled to a presumption of law. This is really the standpoint of the earlier Calcutta decisions e.g., Prosunno Coomaree Debea v. Button Bepari  3 Cal. 696 and of Rampani, J., in Nabu Mondul v. Cholim Mullick  25 Cal. 896 (F.B.). But it is clear now, if only from the decided cases, that the present question is constantly arising in widely different circumstances and has to be answered in each case upon the circumstances of that case. Lord Eobertson in Nilratan v. Ismail Khan  32 Cal. 51 at 61 said:
The question here as in other similar cases is whether the true inference from the facts is that the tenure is permanent or precarious, the burden of proof being on the tenant.
10. If this be so the ultimate inference that the right of a tenant was a permanent right must be open to the Court in point of law, but it is difficult to see how in its nature it can be other than an inference of fact.
11. I make these observations because I recognize that in the present case on the facts found by the lower appellate Court, it may be thought to be difficult to say that the conclusion at which it arrived was either inadmissible in law or compulsory in law : to say that there is no evidence upon which the Court could find that the tenancy was permanent or to say as a matter of law that the facts upon which it did so find Were evidence with which it was obliged to be satisfied.
12. In view of Dhanna Mal's case, however, I think that we must decide the case for Ourselves on the facts found by the Subordinate Judge. We cannot Confine ourselves to saying as was said by Garth, C.J., in Ganqadhur v. Ayimuddin  8 Cal. 960. (at p. 962) that
upon these circumstances we think that the Courts below were at liberty to presume, if they thought fit....that the grant itself was of a permanent character.
13. In the judgment of Chakravarti, J., in Abdul Hakim's case A.I.R. 1925 Cal. 43 (supra) the principle relied on by the tenant isstated to be 'the principle of a lost grant.' This seems to me to introduce from the English law of easements an element which is both unnecessary and confusing. For the present purpose we are troubled by no rule that a permanent tenancy must have been created by a written instrument nor are we concerned to find a lawful origin for acts which would otherwise have been unlawful. The continuance of an old tenancy is equally lawful whether the tenancy be permanent or precarious. The doctrine of lost grant when examined in the case of Angus v. Dalton 4 Q.B.D. 162, was found to bristle with difficulties and in any case it is necessarily based upon an arbitrary period which the law of England puts at twenty years.
14. The references to the principle of lost grant are however only an incident in the judgment in Abdul Hakim's case A.I.R. 1925 Cal. 43. The cases which illustrate the facts from which and the principles on which an inference as to a permanent tenancy can properly be made are very carefully analyzed by Chakravarty, J., in a clear and highly useful statement of the law.
15. On second appeal and in the lower appellate Court the absence of permanent masonry structures upon the land of these suits has been explained by the circumstance that the defendant is a labourer and that people of his position would not generally expect to reside in buildings of such character. This consideration is sufficient to show that the absence of permanent structures is not to be regarded as evidence negativing the permanency of the tenancy. If, tharefore, in the other circumstancess there is sufficient evidence from which an inference of permanency can be drawn no difficulty would arise. But the mere circumstance that the tenant and his family have for a very long time been allowed to continue residing in the same place without any variation in the rate of rent is a circumstance which by itself is an insufficient foundation for holding that the tenant's right was permanent in its origin.
16. In Abdul Hakim's case A.I.R. 1925 Cal. 43 Chakravarti, J., as a result of his analysis of previous decisions considered that the absence of permanent pucca buildings on the land would ordinarily be fatal to a claim for permanency. What I think he meant by this statement was that unless permanent pucca buildings existed on the land the tenant would not as a rule be able to point to anything more than matters which can be explained by the reluctance of a landlord to eject a reasonable tenant i.e., to point to any other element showing that the tenant's long occupation at a uniform rate of rent is unequivocally referable to a permanent right.
17. In my opinion it cannot be laid down that the existence of permanent structures is the only unequivocal or unambiguous fact for the purpose of an inference in favour of the tenant. In this respect I think that the case law has broadened somewhat since Prosonna Coomaree's case  3 Cal. 696 was decided. In Secretary of State v. Luohmeswar Singh  16 Cal. 223 at p. 23, Lord Hobhouse said:
All they offer is some conjecture of such an agreement founded simply on their long possession at a uniform rate of payment. If we could not find out the origin of those things, there would be strength in that argument.
18. The rent in that case had come to be far below the value of the land. In the case of Naba Kumari v. Behari Lal  34 Cal. 902, Sir Arthur Wilson delivering the judgment of the Judicial Committee referred to the circumstances that
the rent was almost a nominal one and had never been enhanced though the value of the holding as measured by the sale price had greatly increased.
19. Again in Bireswar v. Troilakya Dasi  30 C.W.N.709. the land was situated in the Howrah Municipality and the rent remained unchanged for 65 years though the value of the land had increased abnormally. An inference of permanency was drawn mainly upon this consideration.
20. The effect of successions cannot in my opinion be rated very high. No reasonable landlord in a case like the present would be minded to eject the family from its homestead merely because the father or the grandfather had died. Assuming that prior to the Transfer of Property Act a tenancy which was not permanent would not be heritable, still these successions are in themselves poor evidence that the tenancy is being treated, as a tenancy heritable as matter of right. The effect of such evidence is in no way comparable to the effect of a recognition by a landlord of a tenant's claim to transfer the original tenancy to an assignee.
21. The rent in the present eases has remained unchanged, according to the plaintiffs' admission, for 60 and 80 years respectively. It is in evidence that the land of these suits has been occupied not merely by the defendant's family for their own residence, but of late years at least by sub-tenants who have occupied mudwalled huts paying rent to the defendant therefor. But it is difficult to be sure whether for two cottas of land in this village a rent of 13 annas 9 pies or 1 rupee 6 annas 3 pies is a reasonable or economic rent, or should be regarded as a very low rent. None of the Courts below appear to have proceeded upon this consideration. I notice that the defendant in his evidence gave as a reason for which his father had been wont to say that he had a permanent right that the defendant talked to him of the low rent of the land. This seems to be the only evidence that the rent was in fact low and we cannot say whether the economic or rack-rent would be likely to have tempted the landlord to raise the rent on threat of ejectment. It is no doubt possible to call a mud-walled hut a substantial structure but the existence of such huts adds nothing to the fact that the tenancy was for residential purposes. It is not such a structure as the tenant could not without imprudence raise upon the land in the absence of a permanent right nor such as gives notice to the landlord that the tenant claims permanent right. It will not suffice to say that these tenants were labourers and that mudwalled huts were something of a luxury for them.
22. I am sorry that the defendant's possession should be disturbed but I should be still more sorry to administer the law in such a way as to cause landlords to raise tenants' rents or to eject tenants merely to preserve their rights as landlords. This consequence will certainly ensue unless the Courts hold very firmly to the principle that it is not the business of the plaintiff to explain the possession, it is the business of the defendant to show that it leads to the inference of a permanent tenancy, Secretary of State v. Maharajah Luchmeswar Singh  16 Cal. 223.
23. On the whole it appears to me that these appeals should be allowed and the suits decreed. No costs in any of the Courts.
24. I am entirely of the same opinion. At one time during the arguments I was strongly inclined to the view that the defendant's possession of the homestead extending over a hundred years or more with payment of the same rent for sixty years, if not more, should not be disturbed, and that the concurrent opinion of three Courts such as there has been in this case, that the tenancy was a permanent one should not be dissented from. On considering the matter further and in the light of the authorities, bearing upon it, I can see no escape from the conclusion that the appeal should be allowed and the plaintiff's suit should be decreed.
25. Direct proof of the contract which was the origin of the tenancy not being available, the question is what is the correct inference to be drawn from the circumstances proved in the case. The onus of proving permanency being unquestionably on the defendants, the question, properly formulated, is whether the defendants have succeeded in proving circumstances which go to raise in their favour a presumption' or perhaps more correctly an inference of permanency. Whether the defendant has succeeded in discharging this burden is to be judged by the well recognized rule of circumstantial evidence that such evidence should not merely point to the inference that is to be drawn, but that the evidence must be of such a nature that it can possibly lead to no other inference. Were this inference an inference of fact only an error in drawing the inference in view of the definition of 'proof' as given in Section 3, Evidence Act, would in a case where it cannot be said that there is no material in support of it be at best a question of sufficiency of proof, and consequently a question which may not be gone into in second appeal. In any event, were this the position, I should be extremely reluctant to interfere with the affirmance which my learned brother Mitter, J., has lent to an inference of this nature which the first two Courts have concurrently drawn in defendant's favour.
26. The decision of the Judicial Committee in the case of Dhannmal v. Moti SagarA.I.R. 1927 P.C., however, makes it plain that the question is a legal inference from proved facts and not itself a question of fact, and and as my Lord the Chief Justice has said in his judgment that decision must be accepted and applied. We have, therefore, in this appeal to draw our own inference from the proved circumstances. Applying to the case the rule of circumstantial evidence to which I have already referred I find that the only facts proved are that the tenancy was for residential purposes, that possession has extended for over a hundred years with constant instances of inheritance and succession and that there has been payment of uniform rent for sixty years or more, or in other words that there is no proof that the rent varied at any time. Presence of permanent masonry structures would have been a circumstance in the chain in defendant's favour, but their absence, though not necessarily against him, is obviously of no use to him. Instances of transfer would have assisted the defendant, but they are entirely absent. All that has been proved, therefore, is long possession on the part of the tenant together with abstention on the part of the landlord to enhance the rent. Abstention to enhance the rent, to lead to an inference, should be accompanied by circumstances making it unlikely to be due to mere inaction on the part of the landlord: but no such circumstances have been proved. What has been proved, in my judgment is not unequivocally referable to the permanency of the holding and permanency, therefore, is not the correct inference to be drawn.