Mookerjee and Chotzner, JJ.:— This is an appeal by the plaintiffs in a suit instituted to test the legality of an assessment of revenue made under the Bengal Alluvion and Diluvion Act (Act IX of 1847). The case for the plaintiffs is that the disputed lands, which have been constituted into new estates and assessed with revenue, belonged to them by virtue of their rights in Zamindari No. 11492 of Dacca Collectorate and in Taluq No. 4058 of Faridpur Collectorater the plaintiffs maintain that the assessment could not have been validly made under the provisions of Act IX of 1847 and has not in fact been made in accordance with those provisions. The plaintiffs further seek consequential relief by way of refund of revenue paid under compulsion together with damages thereon. The claim is resisted on behalf of the Secretary of State substantially on the grounds that the disputed lands were accretions to the zamindari and taluq mentioned by the plaintiffs, that they were liable to assessment as they had never before been assessed with revenue, and that the assessment had been made in due exercise of statutory powers. The Sub-ordinate Judge has dismissed the suit except in respect of an area which it was admitted on behalf of the defendant, had been included in the assessment by mistake. On the present appeal the decision of the Sub-ordinate Judge has been assailed on the ground that the assessment made by the revenue authorities was ultra vires, first, because the conditions essential for the assumption of jurisdiction by the revenue authorities as prescribed by Act IX of 1847 were not in existence; and secondly because the lands were not liable to assessment. The first of these objections manifestly goes to the root of the decision of the Board of Revenue which was pronounced on the 29th December, 1913 and summarises, though not in full detail, the proceedings adopted by the revenue authorities for the assessment of the disputed lands.

It appears that on the 13th April, 1861, a notification was issued in the Calcutta Gazette for sale of the zamindari rights of Government in more than two hundred tracts, of land in khas mahals. One of these No. 180, was described as “Touzi No. 9234—two pieces of land of kismat Khagatia in Chur Madansankar”. As appears from the robakari of Mr. J.C Dodgson, Collector of Dacca, dated 13th August, 1861, the sale was held on the 20th May, 1861 and was confirmed on the 13th August, 1861. The area was not mentioned in the robakari, but reference was made to Touzi No. 9234 recorded as bearing a sadar jama of Rs. 1,242-11-7. It was expressly directed that the mahal be recorded as permanently settled. There is, we think, no room for controversy that the permanently settled estates came into existence as such on the 20th May, 1861, and we are not at this stage concerned with the exact area intended to be included in this newly created permanently settled estate.

Some years later, on the basis of a robakari of the Deputy Collector of Dacca, dated 27th April, 1870, and an order of the Board of Revenue, dated 17th November, 1870, a notification was issued in the Calcutta Gazette for the sale of another tract of khas mahal land described as “Touzi No. 9691 Pergana Gunanandi, Taluk Basiram Sarma, alluviated Chur formed out of diluviated lands”. The sale was held on the 15th March, 1871 and a sale-certificate was issued to the purchasers on the 3rd April, 1871. We need not at this stage investigate what lands were intended to be covered by the sale-certificate. It is sufficient to state that the taluk as such came into existence on the 15th March, 1871.

The plaintiffs contend that in respect of the zamindari and the taluk thus created in 1861 and 1871 respectively, the proceedings initiated by the revenue authorities were not taken in conformity with the requirements prescribed by Act IX of 1847. This, renders necessary an examination of the statutory provisions.

The provisions of Act IX of 1847 were analysed in the judgment of the Judicial Committee delivered by Lord Herschell in the case of Secretary of State v. Fahamidannissa , the Judicial Committee affirmed the decision of the majority of the Full Bench in Fahamidannissa v. Secretary of State , which had overruled in part the decision of Wilson, J. in Sarat Sundari v. Secretary of State (3). Act IX of 1847 was framed for the assessment of lands gained from the sea or from rivers by alluvion or dereliction. Section 3 empowers the Government to direct new surveys of riparian lauds, and provides as follows:

“Within the said Provinces, it shall be lawful for the Government of Bengal, in all districts or parts of districts of which a revenue-survey may have been or may hereafter be completed and approved by Government, to direct from time to time, whenever ten years from the approval of any such survey shall have expired, new survey of lands on the banks of rivers, and on the shores of the sea, in order to ascertain the changes that may have taken place since the date of the last previous survey, and to cause new maps to be made according to such new survey”.

Section 4 lays down that the approval of the revenue surveys of districts or parts of districts which may be hereafter surveyed, shall be deemed to have taken place on such day as may be specified as the clay of such approval in the Calcutta Gazette. Sections 5 and 6 deal, respectively, with the question of deduction from jama of estates from which lands have been washed away, and the question of the assessment of increments to revenue paying estates. Section 6 which is relevant, in the case before us, provides as follows:

“Whenever on inspection of any such new map, it shall appear to the local revenue authorities that land has been added to any estate paying revenue directly to Government, they shall without delay assess the same with a revenue payable to Government according to the rules in force for assessing alluvial increments and shall report their proceedings forthwith to the Board of Revenue whose orders thereupon shall be final”.

The expression “any such new map” plainly refers to the “new map” made according to the “new survey” as contemplated in section 3. That section provides for periodical surveys at intervals of not less than ten years, after a revenue survey has been completed and approved. The object of the “new survey “is to ascertain the “changes” that may have taken place since the date of the last previous survey, that is changes by alluvion or dereliction (not changes by possession): Wakilan v. Deonandan . Section 6, then imposes upon the revenue authorities the duty to assess what may be called added land, whenever, on inspection of the new map, it appears that land has been added to an estate paying revenue directly to Government. There must consequently be a comparison between two maps, made at an interval of not less than ten years and each showing the revenue paying estate concerned. That estate must, accordingly, be in existence as a revenue paying estate, if not before, at least on the date of the first of the two maps taken as the basis for comparison. We may usefully recall here the following passage from the judgment of Wilson, J., in Sarat Sundari v. Secretary of State , which except in one particular, remains unaffected by the decision of the Full Bench and of the Judicial Committee in [Fahamidannissa v. Secretary of State and Secretary of State v. Fahamidannissa . “The object of the Act is to provide for the assessment of riparian estates from time to time, in accordance with the changes which periodical surveys may show to have taken place in their area and boundaries. Section 3 of the Act refers to a revenue survey which is to be approved by Government as fixing the boundaries of estates, and provides that at intervals of not less than ten years, fresh surveys of such estates may be made. Section 5 then provides for a reduction in the Sudder jama when, on a comparison of two successive surveys, it appears that the area of an estate has been diminished, and section 6 provides for an addition to the jama when, on inspection and comparison of the new map, land appears to have been added to the estate since the last survey. In every case the starting point is to be the revenue survey which, it would appear, is to be taken as representing the boundaries of the estate as they existed at the time of the permanent settlement, and it is apparently not open to the revenue authorities to go behind that survey and enquire whether in fact the boundaries at the time of settlement were not other than therein represented”.

Wilson, J., in this passage, had apparently in view the case of an estate which was in existence at the time of the Permanent Settlement of 1793. Section 3 of Act IX of 1847 is, however, all comprehensive in scope, and sections 5 and 6 both refer to all estates paying revenue directly to Government, no matter whether they were or were not in existence in 1793. “What is essential to attract the application of Act IX of 1847 is that there should have been, in the case of the estate concerned, a revenue survey. This primâ facie furnishes the boundaries, as presumably, though not conclusively, accurate; see the judgment delivered by Wilson, J., on behalf of the majority of the Full Bench in Fahamidannissa v. Secretary of State , which, to this extent, overruled his previous decision in Sarat Sundari v. Secretary of State . The true position is that the revenue survey map is taken as the basis of comparison; but the comparison of the maps is not conclusive. The comparison sets the revenue authorities in motion, and they may, then, on the best materials they can procure, proceed to assess what land they deem to be assessable. This view is confirmed by the observations of Lord Herschell in Secretary of State v. Fahamidannissa , section 3, according to him, empowers the Government of Bengal, in any district in which a revenue survey has been completed and approved by the Government, to direct decennially a new survey of lands on the banks of rivers, and on the shores of the sea, in order to ascertain the changes that may have taken place since the last previous survey and to cause new maps to be made according to such new survey. Section 6 then provides that whenever, on inspection of any such new map, it shall appear to the local revenue authorities that land has been added to any estate paying revenue directly to Government, they shall, without delay, duly assess the same according to the rules in force for assessing alluvial increments. Such added land cannot obviously be land which was already comprised in a permanent settlement, but had become derelict of the sea or a river; for it would be a contradiction in terms to maintain that such land had been “added” to the estate to which they already belonged. Lord Herschell then refers to section 5 which deals with the question of deduction from jama of estates from which lands have been washed away, and points out that the Act provides no machinery for making such abatement where the land was covered with water at the time of the original survey; it is only “when on inspection of the new map” it appears that the land has been washed away that there is any legislative authority for making an abatement. These remarks apply equally to as case under section 6, and it is only when on inspection of the new map it appears that land has been added, that there is legislative authority for assessment of additional revenue. Lord Herschell finally adds that it would be an erroneous interpretation of Act IX of 1847 to hold that it rendered the Board of Revenue supreme and enabled them to make valid and effectual a proceeding on their part which the law had declared to be wholly illegal and invalid.

In the case before us, Mr. Bose has forcibly argued on behalf of the appellants that the record indicates that the proceedings for assessment, the legality whereof is in question in this suit, were not taken in conformity with the provisions of Act IX of 1847. He has argued that neither in the case of the zamindari created in 1861 nor in the case of the taluk created in 1871, is it shown that there was a revenue survey contemporaneous or subsequent in date; nor is it shown; that there was a new survey after the lapse of not less than ten years from the date of such revenue survey. There could not consequently have been a comparison of two requisite maps in terms of section 6 before the revenue authorities set in motion the machinery at their disposal. The difficulty of the situation has been fully appreciated by the advisers of the Crown, and the suggestion has been thrown out that a survey made by Parbati Charan Ray, which is mentioned in the course of the proceedings, may fulfil the requirements of the initial survey contemplated by section 3. We have been further informed that there are resolutions of Government some of them published in the Calcutta Gazette which lend support to this theory. The appellants have, however, strenuously opposed the reception of additional evidence at this stage, specially as they may have no opportunity to adduce rebutting evidence; and in this connection, they have invited our attention to O. XLI, r. 27 of the Civil Procedure Code, 1908, and to the observations thereon by the Judicial Committee in Kessowji Issur v. G.I.P Ry. Co. , see also Sreeman v. Gopaul where the Judicial Committee emphasised the importance of exercising cautiously and sparingly the discretionary power vested in an Appellate Court to allow the production of additional evidence. The appellants have, as might have been expected, dwelt further on the fact that the objection as to non-compliance with the essential provisions of Act IX of 1847 was urged, not only in the lower Court, but also before the revenue authorities, and yet had not been heeded. We have anxiously considered what course should be adopted in such circumstances; we have arrived at the conclusion that the matter should be further investigated and an opportunity should be afforded to the Crown, on terms presently to be stated,—to establish in full detail tlie’ steps taken by the revenue authorities for the assessment which purports to have been made in accordance with Act IX of 1847 and is challenged as illegally made. The matter is involved in obscurity chiefly from two standpoints; first, we have no information on the record as to the survey by Parbati Charan Ray; secondly, we are not apprised whether the survey by Parbati Charan Ray was a “revenue survey” or a “new survey” within the meaning of section 3, and whether the map prepared by him was in fact taken as the basis for comparison, under section 6, with a subsequent “new map”. It is, we think, necessary that the proceedings of the authorities in relation to the survey by Parbati Charan Ray, as also the proceedings by the revenue authorities which culminated in the order for assessment made by the Board of Revenue on the 29th December, 1913, should be placed, each in its entirety, before the lower Court. It does not appear to have been realised that as the jurisdiction of the revenue authorities was challenged, it was incumbent on the advisers of the Crown to place before the Court all the materials available, with a view to establish that jurisdiction had been assumed in strict compliance with the statutory requirements. The papers relating to the survey of Parbati Charan Ray as also those relating to the assessment under Act IX of 1847 are all in the custody of the officers of the Crown, and the plaintiffs could not be expected to produce them. The attempt at discovery of documents made by the plaintiffs in relation to another part of the case was infructuous, and its fate was by no means calculated to encourage them to renew their effort. We trust that to ensure the production of the necessary papers, we need only draw attention to this point and to refer to the emphatic disapproval, expressed by Lord Shaw in Murugesam v. Manickavasaka , of the practice which has grown up, in Indian Procedure, of those in possession of important documents or information lying by, trusting to the abstract doctrine of the onus of proof and failing accordingly to furnish to a Court of justice the best material and the best assistance for its decision. We are consequently of opinion that there must be a remand for investigation of the question, whether the proceedings of the revenue authorities, which terminated in the controverted assessment, fulfilled the requirements of Act IX of 1847. In the view we take of this question, it is not desirable that we should express a final opinion upon the various points which require consideration on the merits. If, on behalf of the Crown, it is not established after remand that the proceedings were initiated in conformity with the requirements of Act IX of 1847 (the burden must lie upon the defence, as a primâ facie case has been made out by the plaintiffs on this point), the suit will stand decreed. In that event, the question of liability of the disputed lands to assessment will be left open; and, as upon a fresh assessment in a proceeding lawfully taken, that very question may require examination, the parties should obviously be left unembarrassed by superfluous findings. On the other hand, if the defendant should be able to satisfy the Court, after remand, that the proceedings under Act IX of 1847 conformed strictly with the statutory requirements, the questions on the merits will have to be reconsidered by that Court; and it is from this point of view that we shall now refer to some points which, amongst others, require, in our opinion, further elucidation.

As regards Khagatia Char, the question of what, was included in the permanently settled estate requires re-examination, was it a settlement of lands within defined boundaries as set out in the robakari of Mr. O.C Mallik, Deputy Collector, dated 21st May, 1858 or was it a settlement of a specific area of land. Besides this, the question must be faced, was it a settlement in accordance with the condition of things as they existed in 1857 when the lands were surveyed, or according to the state of things as they existed when the sale took place in 1861. Three possible views have been suggested; (i) what was sold was land as in 1858 on revenue as assessed in 1858; (ii) what was sold was land as in 1861 on revenue as assessed in 1858; (iii) what was sold was land as in 1861 on revenue as assessed in 1858 with an implied liability to submit to additional assessment for extra area. The solution of the question in controversy is by no means free from difficulty, and the authorities applicable will be found reviewed in the case of Secretary of State for India v. Narendra Nath Mitter . It is manifestly of vital importance to the parties whether 1857 or 1861 is taken as the appropriate date, and whether the settlement is deemed to have been made in respect of a specific area or of land comprised within specific boundaries. The burden of proof, as pointed out in Secretary of State v. Jatindra , lies upon the Crown to establish that the land now attempted to be assessed is added land, that is, land not included in the original assessment; see the decisions of the Judicial Committee in Secretary of State v. Fahamidannissa , Haradas Acharjya v. Secretary of State (4), Chandrika v. Indar (5), Secretary of State v. Maharaja of Burdwan (6). Here, again, the relevant papers are all in the hands of the officers of the Crown, and it is incumbent on them to place before the Court all the materials available, so as to facilitate the determination of the exact situation of the lands comprised in the settlement when the estate was created. In this connection, the question of title to what is called the “cipher touzi” must also be reconsidered, and the point determined, whether the Permanent Settlement of 1861 was made in respect of land within defined boundaries to the exclusion of a sheet of water, as has been purged on behalf of the defendant.

The observations we have made with regard to Char Khagatia apply, more or less, to the Basiram Char, which was created a taluk in 1871. Here, again, the question arises with reference to the robakari dated 27th April, 1870, whether a specific area or bind within specific boundaries was settled. There is further the question, whether the settlement was made in view of the conditions as they prevailed at the time when the char was surveyed or at the time when the settlement took effect. But whatever view may ultimately be adopted, as the relevant papers are in the possession of the officers of the defendant, they should be produced to enable the Court to fix the exact situation of the lands coveted by the sale-certificate of 3rd April, 1871. The burden, it must be remembered, lies upon the defendant to establish that the lands sought to be assessed are added lauds within the meaning of section 6.

Finally, the history of Kartikpur requires much fuller investigation than has been found possible. We shall not, for reasons already assigned, express a definite opinion upon the question, whether the site of the lands now in dispute was dry land at the time of the Permanent Settlement of 1793 and was presumably included in a permanently settled estate. But this much may be stated that the maps placed before us from Rennell's Bengal Atlas, and Rennell's Supplementary Atlas, primâ facie, strengthen the case for the appellants. How far this may justify a presumption that the state of things in 1764-1773 continued substantially unaltered till the time of the Decennial Settlement in 1789 (which was made permanent in 1793), is a matter for consideration in the light of the principles enunciated by the Judicial Committee in Jagadindra Nath v. Secretary of State , Haradas v. Secretary of State and Secretary of State v. Maharaja of Burdwan , see also Prafulla v. Secretary of State . In the consideration of such a question, the subsequent history of Kartikpur, the submergence and re-appearance, may well be borne in mind; but it has been urged on behalf of the appellants, that this must be taken along with the fact that the river Brahmaputra did not change its course till the beginning of the nineteenth century, and, that consequently, the river Ganges was not so erratic in 1789 as it has been since that period. The papers, which might have thrown light upon the question of the condition of Kartikpur at about the time of the Permanent Settlement have, however, been kept back: on behalf of the defendant. The robakari of 7th September, 1842 refers to the batwara papers of 1792, prepared by Thomson, and to the quinquennial papers of 1792; these have not been produced, and the Chauhaddibandi papers also have been kept back. The Sub-ordinate Judge should take steps to secure the production of those and other papers relating to the Permanent Settlement of Kartikpur. If they are still withheld, the Court must not hesitate to draw an adverse inference, such as legitimately arises from the omission to produce relevant papers: Murugesam v. Manicka . The history of Kartikpur must, in our opinion, be reinvestigated in the light of further papers, which we trust will be produced on behalf of the Secretary of State.

We may add that as the decision of the Board of Revenue was pronounced on the 29th December, 1913 and the present suit was instituted on the 2nd January, 1915, the objection of limitation was abandoned in the Court below, and we are consequently not called upon to examine the decisions in Secretary of State v. Prafulla Nath and Prafulla Nath v. Secretary of State .

The result is that this appeal is allowed, the decree of the Sub-ordinate Judge set aside in so far as it dismisses the suit, and the case remanded to him for retrial, with liberty to both sides to adduce fresh evidence and to apply for fresh local investigation. The decree of the Sub-ordinate Judge will stand confirmed in so far as it is in favour of the plaintiffs appellants in respect of 72.55 acres of land. The appellants will have their costs in this Court in full, and also the costs in the lower Court, excluding the court-fees; the remainder of the costs already incurred by the plaintiffs in the lower Court, and the costs to be incurred by both sides after remand will be in the discretion of that Court.

B.M.S

Appeal allowed.