This appeal comes before their Lordships from the Calcutta High Court in a suit in which the respondent, the Secretary of State for India in Council, is the plaintiff. He claims a decree in his favour declaring his title to seven jalkars or fisheries which in his plaint he describes as "component parts" of his fishery known as Jalkar Gangapath Islampur (southern portion). His claim was originally to nine jalkars but has been restricted to seven.
 It is not disputed that the respondent as in right of the southern portion of Jalkar Gangapath Islampur is entitled to the fishings in some fifty miles of the river Ganges, which is a public navigable river. Nor is it disputed that he is entitled to all subsidiary fishings which the law recognises as "'adjuncts" of his fishery in the Ganges. But it is denied by the appellants, who are certain of the original defendants in the suit that the respondent has any right to the seven jalkars now claimed by him whether as "adjuncts" of his fishery in the Ganges or under any other title. The appellants who have contested the respondent's claim before their Lordships are the proprietors of an eight annas share of Pergunnah Ekbarabad, within whose limits the disputed fisheries lie, and they maintain that to the extent of their proportion the disputed fisheries belong to them as the owners of the underlying solum. The Secretary of State succeeded in his claim both before the Subordinate Judge and in the High Court.
 A description of the situation and geographical features of the jalkars in question is necessary in order to appreciate the problem. The Pergunnah Ekbarabad is an estate
lying between a river known as the Bhagirathi and a portion of the stretch of the Ganges in which the respondent admittedly owns the fishings. The Bhagirathi was originally not an independent river but was a diversion of part of the waters of the Ganges which it rejoined lower down. Subsequently, and at any rate before 1867, the Bhagirathi at its upper end ceased to be connected with the Ganges and became an independent river. Instead of flowing directly into the Ganges, as it originally did, it became united with another river called the Pagla and the united stream flows into the Mahananda river which in turn joins the Ganges. The Bhagirathi was at one time navigable but has ceased to be so. One of the seven jalkars claimed by the respondent is in a three-mile stretch of the Bhagirathi where it borders Pergunnah Ekbarabad some distance above its junction with the Pagla. Another of the disputed jalkars is in a channel between the Bhagirathi and the Pagla. The five other jalkars are bils, that is, ponds or swamps. Four of them lie at distances of from half a mile to two miles from the Bhagirathi with which in the rains they are connected by channels or ditches and the fifth is connected by a channel with the Pagla.
 The capricious habits of Indian rivers, due to the conformation of the land and the alternate extremes of dry and rainy seasons, result in frequent changes in their courses. Areas at one time covered with water dry up and areas of previously dry land become submerged, while connections with associated sheets of water open or close. Such alterations necessarily affect the fishing rights and have given rise to difficult legal questions which have been much litigated.
 It may now be taken as settled that in general the fishing rights in a river follow the river and that subsidiary fishing rights in waters associated with a river fishery, or "adjuncts" as they have come to be called, may vary with the changes in the river itself. In order that the fishings in any waters may be claimed as an "adjunct" of a river fishery these waters must be in continuous connection with the river throughout the year, in the dry season as well as in the wet season. Waters which cease to be continuously connected with the river cease to be "adjuncts" of it and the fishings in such disconnected waters cease to be part of the river fishery, and pass to the owners of the solum. On the other hand, the owners of a river fishery acquire right to the fishings in new areas which come to be covered with waters connected with the river and the owners of the submerged solum must recognise the creation of such fishing rights within their estates. Connections between existing sheets of water and the river may also open or close with consequent effects upon the fishing rights. There may thus be loss and gain as between the owners of the solum and the owners of the river fishery as the physical circumstances alter. Such their Lordships find to be the law as established after some fluctuations by the reported cases including the decision of this Board delivered by Lord Sumner in 41 IA 221.(1)
 Now it is manifest from a perusal of the respondent's plaint that he came into Court with a case based on the contention that he was entitled to the seven disputed jalkars as "adjuncts" or, as he put it, "component parts" of his fishery Jalkar Gangapath Islampur, that is of his fishery in the Ganges. Indeed in a petition which in the course of the proceedings he presented to the Court he expressly stated that "the jalkars in suit have been claimed in the suit as component parts of the Jalkar Gangapath Islampur" and asked leave to put forward an alternative argument should it be found that the jalkars in suit "can no longer be considered as connected with the Ganges." This shows the meaning which in his plaint the respondent attached to the words "component parts."
 When the respondent's averments are examined it appears that he nowhere claims that the water channel between the Bhagirathi and the Pagla or any one of the five bils is now connected throughout the year with the Ganges. On the contrary he avers only that they "remain connected with the Ganges for about three months every year during the rainy season." If these six jalkars, as he avers, are connected with the Ganges only for part of the year, the respondent cannot succeed in his claim to them as "component parts" or "adjuncts" of his fishery in the Ganges, in view of the law as their Lordships have above defined it. As regards the three mile stretch of the Bhagirathi river, the respondent puts his case in his plaint somewhat differently. He states that at its source the Bhagirathi is connected with Ganges only for about two-and-a-half months in the year in the rainy season. This would not make the fishings in the Bhagirathi an "adjunct" of his fishery in the Ganges. But he goes on to say that the Bhagirathi remains connected with the Ganges throughout the year because its waters flow into or ultimately find their way to the Ganges. This will plainly not do. The Bhagirathi is now a tributary, not a diversion of the Ganges and it is not even a direct tributary, as has been already explained. Once it is ascertained that the Bhagirathi is an independent tributary of the Ganges the fishings in it cannot in law be regarded as an "adjunct" or component part of the fishery in the Ganges. Counsel for the respondent very properly admitted that he could not contend that a right of fishing in a river included the right to fish in all its tributaries as "adjuncts". Such a contention if admitted would lead to manifest absurdity.
 The result is that the respondent's case as tabled on record necessarily fails and that the judgment of the Subordinate Judge in his favour so far as based on his finding that "the water of the jalkars in dispute is still the water of the Ganges", although none of them is connected throughout the year with the Ganges, is erroneous.
 The subordinate Judge proceeds, however to discuss the question of the respondent's title to the jalkars not as "adjuncts" of his fishery in the Ganges but as included in the grant of Jalkar Gangapath Islampur to his original predecessor. The respondent's plaint does not set out his title or deduce it through its transmissions to himself, but as the matter has been fully dealt with both by the sub-ordinate Judge and by the High Court their Lordships have thought it right to hear a full argument on the subject and to express their judgment upon it. Here, reference may again be made to the petition presented to Court on behalf of the respondent in which he asked leave to argue another ''aspect of the case", namely, that if it should be found that the jalkars in dispute were component parts of Jalkar Gangapath Islampur "at some point of time in the past and consequently the plaintiff had right to them then the plaintiff's right and claim to them should be upheld and he should get a decree in respect of them unless his claim is found to have been barred by limitation or his title extinguished by adverse possession."
 The defendants objected to the respondent "making a different case from that made in the plaint." What the Subordinate Judge did on the respondent's application does not appear, but he evidently allowed the question of title to be fully developed before him, although it scarcely seems to be the other "aspect of the case "referred to in the respondent's petition. Even on second thoughts it does not seem to have occurred to the respondent to make a case either that the jalkars in dispute were in express terms included in the title of the respondent's authors or that the respondent had acquired a prescriptive right to them by possession. The only allegation in the plaint as to possession is that the jalkars "were in peaceful possession of the plaintiff's predecessors, in-interest till about 40 years ago," since when the defendants have "been in wrongful possession of the same."
 The Subordinate Judge after hearing evidence and examining the documents came to the conclusion that the respondent had been able to prove "long possession from 1806 or 1807 to 1888" and that "apart from the question of title the plaintiff is entitled to succeed on the strength of long possession," but that "on a consideration of the documentary evidence ... there is no doubt that the disputed jalkars were included in the jalkar of Hoolas Chand [i. e., the Jalkar Gangapath Islampur] created by the permanent settlement, if not before that."
 When the case reached the High Court on appeal the learned Judges delivered a judgment discussing at considerable length the physical characteristics past and present of the jalkars in question and the law with regard to "adjuncts" of a river fishery. Bather surprisingly, they state that in his plaint "the plaintiff made the definite case that these waters are component parts of the Jalkar Gangapath Islampur" but that he did not thereby mean to claim that he had a right to the fishings in them as "adjuncts" of the river Ganges in the legal sense. They point out that the jalkars "cannot be considered as adjuncts because the connection directly or indirectly with the Ganges is not throughout the year"; and therefore they seem to infer that the plaintiff did not claim them as "adjuncts" but as parts of the same river system, whatever precisely that may mean. Passing from this topic apparently without pronouncing upon it, the learned Judges go on to say that the case in their "judgment depends upon two questions, namely "(1) the nature and extent of the right of the Crown to grant a several fishery to a private individual, and (2) the extent of the grant actually made by it in this case."
 Their Lordships see no reason to differ from the view of the High Court that there is nothing to prevent the Crown, when making settlements of land, from severing fishery rights in waters from the right to the subjacent soil or from granting the fishing rights to one private individual and the subjacent soil to another private individual. This being accepted, the question is next considered of the respondent's title to the seven jalkars in dispute, not as "adjuncts" of Jalkar Gangapath Islampur but as subjects of grant by themselves. Waiving the pleading point that the respondent has nowhere set out on record and deduced his title their Lordships have examined the documentary evidence on which the learned Judges of the High Court found that the seven jalkars must be held to have been included as substantive items in the respondent's title.
 The burden is upon the respondent to make out his title. He claims to be the successor in title of the grandmother of one Hoolas Chand. To this lady in pre-British times Nawab Aliverdi Khan in the middle of the eighteenth century made a grant of several fishery in the Ganges extending from Pointy to Sooty and designated Jalkar Gangapath Islampur. The fishery was then undivided and it was not until 1870 that it was divided into a northern and a southern portion. The grant was confirmed to the father of Hoolas Chand. Neither the original sanad nor the confirmatory sanad is forthcoming. The respondent is thus under the initial disadvantage of having to rely on secondary evidence to prove the contents of the original grant to his predecessor. The transmission to the respondent of the grant, whatever it included, is sufficiently traced.
 The first document is a letter of 2nd October 1789 from the Collector of Bhagalpur requesting instructions for the preparation of the Decennial Settlement. In his letter he purports to give a brief "account of Gungapath Islampore." He states that "the talook so called is a fishery on the Ganges extending from Pointy to near Sooty with one village in hand," was given in farm by the Nawab to one of his female domestics and is now held by her grandson Hoolas Chand. This was the original undivided fishery. The respondent can derive no aid from this document, and so far as it goes it is against his contention for the fishery is simply described as "on the Ganges" and while a village is mentioned as going with the fishery there is no mention of any other separate items or parcels. The Collector was directed to settle with Hoolas Chand who executed on 15th November 1790 a kabuliat of "Pargana Gangapath Islampur " without mention of anything more.
 On 25th October 1797 Hoolas Chand made an application to the Government in which he stated that the sanad granted to his father had been destroyed by a fire in his house. He described the subject of grant in the sanad as "Zemindary Mehal Perganah Gangapath Islampur ... within the boundaries specified below within the land of all Zemindars Rahati [still water] Bahati [flowing water] Shikasti [diluviated] Bahati [remaining] Dhobha [pond] and Chharan [deserted river bed]." In an attached schedule of particulars the boundaries of "Jalkar Khas Bahinch [direct fishing]", are stated to be "commencing from Painti [Pointy] to Shibgunge Turtipore [Sooty] with Phandi [branches]... the big river " "Ijaradars [lessee] of rahati and bahati," [still and flowing waters] are mentioned.
 Several other rivers with branches are included in the schedule but there is no mention of the Bhagirathi. After an inquiry the Collector ordered that Hoolas Chand should keep "in his possession the Mehals in the said Pargana as heretofore." From this also the respondent gains no assistance. The seven jalkars now claimed a shaving been included in the original grant are not named and the references to still and flowing waters, ponds and deserted river beds are in their Lordships' opinion no more than references to the then existing "adjuncts" of the fishery in the Ganges.
 The next document is one to which the High Court attached importance as "valuable evidence". It is a Dehabandi or return relating to the year 1804-5 made to the Collector of Purnea by a lady, describing herself as Zemindar of Mouzahs and Kismats in Pergana Gangapath Islampore "within the jurisdiction of District Purnea." It contains a long list of nearly two hundred jalkars among which are mentioned the seven disputed jalkars. Nothing appears to be known about this lady and counsel for the appellants raised a question as to the correctness of including these fisheries in the District of purnea, which however it is unnecessary to pursue.
 In the opinion of their Lordships, differing from that of the learned Judges of the High Court, this Dehabandi does not advance matters. It does not prove that this long list of jalkars was included expressly in the original sanad. If it is invoked to show that the seven jalkars in dispute were included expressly in the original grant, it is equally
evidence that all the other jalkars named in the long list were included expressly in the original grant which is most unlikely, and is not indeed suggested. All that it shows is that this lady in 1804-5 claimed a large number of jalkars as then included in Jalkar Gangapath Islampur but for aught that appears they may at that time all have been "adjuncts" of the river fishery and connected physically with it.
 A document of 1838 is said by the High Court to be of materiality. It is an extract from the Rubakari or proceeding in the Court of the Deputy Collector for the District of Malda regarding the Jalkar Pargana Gangapath Islampur. It examines the history to date of the fishery, and states that many of the jalkars included in the collection papers had been encroached upon and engulfed by the big river and cannot be identified and many others have been recently formed but that some 334 jalkars and dead rivers were correctly included and among others the jalkars in dispute. A temporary settlement for one year only was consequently made pending further inquiry. This is no doubt evidence of the jalkars in suit having been "component parte" of Gangapath Islampur in 1838 but it is no evidence that they were so at the date of the original grant or that they are so now and still less is it evidence that these jalkars were specified as items of the original grant. On the contrary it rather supports the view that the subsidiary jalkars were a fluctuating quantity varying with the physical conditions.
 In 1855 the Officiating Deputy Collector of Malda sought instructions from the Revenue Commissioner of the Rajshahi Division as to the fresh settlement of the Jalkar Gangapath Islampur. He alleged that the fishery included the fishings in all branches whose communication with the Ganges had become cut off and all beds which might be left by the river in future. "No sooner," he says, "does a branch of the Ganges become separated ... from the main stream than the right to fish it is immediately claimed by the zemindar whose lands surround the water; such claims have already been successfully asserted in too many instances." He accordingly suggested proceedings for the protection of the Government's rights, but his advice was not accepted and nothing came of the matter. Several subsequent suits to recover neighbouring jalkars with which this case is not concerned failed. When the Government advertised the fishery estate for sale in 1870 they described it as the southern portion of fishery Gangapath in the river Ganges, Pergunnah Gangapath, extending from south of Rajmehal Road to Soontee and Shibgarj Toortee. No separate items were mentioned.
 The learned Judges of the High Court find that "the documentary evidence on the record which we have noticed above leads to the conclusion that Pergunnah Gangapath Islampur comprises the jalkars now in suit. They have not changed their sites, at least there is no evidence to the contrary," Their Lordships find themselves unable to accept this view. Their examination of the documents, with the most important of which they have dealt above, satisfies them that the respondent has failed to discharge the burden upon him of proving that his title includes the jalkars in suit as specific items apart from his general right of fishing in the river Ganges. His title, in their opinion, gives him a right to all the "adjuncts" of his fishery as these may vary from time to time with the physical changes of the river, but does not entitle him to claim the fishing in waters which have ceased to be "adjuncts" of the river. It would be a singular right which entitled the respondent to claim as "adjuncts" the fishings in all newly formed water areas coming into existence in connection with the river, as he may undoubtedly do, and at the same time to retain a right to fishings in waters ceasing to be connected with the river. No precedent was cited to their Lordships of such an anomalous fishery right and it would require very precise evidence to establish such a right as a matter of title.
 As a last resort the respondent sought leave to make a case on prescriptive possession apart alike from his case on "adjunct" and his case on title. No such case is made on the record, no such issue was formulated and the High Court has not dealt with it. Their Lordships in these circumstances declined to allow this plea to be taken before them. Possession is a matter of fact and the appellants might well complain that if they were to meet a case of prescriptive possession they should have had notice of the precise period of possession and of the acts of possession on which the respondent proposed to rely. They could then have directed their evidence to this issue. The respondent has already been permitted to argue a case not made by him in his plaint. Their Lordships cannot permit him at this latest stage to raise yet another ease for which he has no record. It is true that a considerable body of evidence was adduced in the course of the case on the subject of possession but this was directed to support the respondent's other pleas and not to establish a separate right founded on prescriptive possession.
 Their Lordships will humbly advise His Majesty that the appeal be allowed, that the judgments of the Subordinate Judge and the High Court so far as affecting the present appellants be recalled and that the suit so far as directed against the present appellants be dismissed. The appellants will have their costs against the respondent both of this appeal and in the Courts in India.