delivered the opinion of the Court.
This suit Avas brought by the State of Tennessee against W. A. Cissna and the Muncie Pulp Company, in the chancery court of Tipton county, Tennessee, to recover about one thousand acres of land, charged in the bill to be situated in that county and then in the possession of W. A. Cissna, who claimed to own the same in fee, and the Muncie Pulp' Company, his lessee. An injunction was also asked to stay waste in cutting and removing timber, being committed by the Mun-cie Pulp Company. These defendants made defense by plea in abatement to the jurisdiction of the court, in *57that the lands sued for Avere not situated in the State of Tennessee, but in the State of Arkansas. The defendant Cissna in his plea says that these lands were formerly, about 1823, on the Tennessee side of the middle of the Mississippi river, Avhich, as the river then ran, AAras and is the boundary line between Tennessee and Arkansas, but by gradual and imperceptible erosion upon the Tennessee bank, and accretion upon that of Arkansas, they became in the course of time and are now within and a part of the territory of the State of Arkansas. The defendant Muncie Pulp Company simply says the lands are not within the State of Tennessee, but within the boundaries and a part of the State of Arkansas. The defendants also filed answers to complainant’s bill under an agreement of record that in so doing they would not waive their pleas to the jurisdiction of the court. The case, after issue, was by consent of parties transferred from the chancery court of Tipton county to that of Shelby county, and there heard by the chancellor upon the pleas in abatement and the proof offered by the parties upon the issues thus made. The cháncellor avús of the opinion that the case was with the defendants, and sustained the pleas and dismissed the bill. Complainant has ap-peáled from this decree and assigned error.
The case is before us alone upon the question of jurisdiction presented by the pleas in abatement, but the decision of this question necessarily involves the title of the complainant to the lands sued for, since she *58claims them as a sovereign State, under the same grants, treaties, and legislation by which its western boundary is defined, declared, and established. The location of the boundary line between Tennessee and Arkansas, and the right of the former to recover the lands in question, are practically the same question, and will therefore be considered together.
The lands described in the bill and sought to be recovered confessedly were at one time, about 1823, under the waters of the Mississippi river. This is admitted in the plea of W. A. Cissna, and is so clearly and conclusively established by the proof, that it is not now controverted by any one. The Mississippi- at this point a.t that time and for many years thereafter made a great bend, forming a .tongue or. peninsula extending northwestward from a direct north and south line, the distance around which was more than twenty miles, but across the neck connecting it with Tennessee less than two miles. This peninsula was separated by McKenzie’s Chute, an arm of the river, and the northern part was known as “Island 37.” ■ The whole, called “Devil’s Elbow,”'was part of Tipton county, Tennessee. The river began this bend at the southern point or apex of Dean’s Island, which was between the main channel of the Mississippi river and Barnay’s Chute, and is a part of the territory of Arkansas, and property of the defendant W. A. Cissna, and ran first westward, then northward betwéen Dean’s Island and the main land of the península and Island 37, then westward and *59southward around Island 37, then in a northeastern direction until it came within about two miles of the' place where it started northward, and then resumed its general course southward. The main channel of the river was at this time southwest and west of Dean’s Island, about one mile, or g little less, in width. The location of the islands here mentioned and the course of the river are difficult to describe, and can best be seen and understood from an inspection of a map made by Maj. J. H. Humphreys, a civil engineer, a copy of which is exhibited with complainant’s bill and here reproduced. See the following page.
The river continued to run between Dean’s Island and the peninsula and Island 37 opposite it until March 7, 1876. Considerable changes, however, had taken place in its bed at this point in the meantime. The width of the channel, by erosion and caving in of the Tennessee bank south, southwest, and wrest of Dean’s Island along the main land and Island 37, had increased from its former width to that of one and one-quarter miles or one and one-half miles, and a towhead, which seems to be a formation upon the bottom of the river, appearing at times, but not abvays above its surface, and neither a bar, nor yet land, had appeared off the apex of Dean’s Island, a navigable chute running between it and the island, and a sand bar and mud flats, only seen in very low wrater, had also formed in the river near the bank of that island, perhaps beloAV the towhead. A steamboat reconnaissance of the river, *61under the direction of the War Department of the United States, was .made by Col. Suter in 1874, and a. map of the place which we are now describing was prepared by him or his assistants and is in evidence. There is no proof of any material changes in the river between 1874 and 1876, and this map, while it is not shown to be altogether correct and accurate, may be-said to present the general situation as it existed in the latter year. It is also here reproduced. See the following page.
*60
*61Upon the date referred to, March 7, 1876, the river suddenly and with great violence, within about thirty hours, made for itself a new channel directly across the neck opposite the apex of Dean’s Island, then reduced in width to about a mile, which new and shorter channel, thus made, it continues to occupy to this time. The new channel was called the “Centennial Cut-off,”, and the island made by it “Centennial- Island.” The change of the channel, as stated, was sudden and violent. About two thousand acres of valuable cultivated lands were swept aAvav, with the farmhouses, ginhouses, and other improvements upon them, in a few hours, and the inhabitants with difficulty saved their lives and personal property. The old chiannel around the bend of the elbow was abandoned by the current of the river, but remained, for a few years, covered with dead water, becoming a lake or lagoon. It was no longer navigable,, except in time of high water for small boats, and this continued only for a short time. The fall in the river: around the elboAV, from six to eight feet, was all con*63densed in tbe one mile of tbe cut-off, and made a strong current there. This, of course, drew tbe water from tbe old channel rapidly, and greatly reduced its depth. The old bed immediately began to fill with sand, sedi: ment, and alluvial deposits, and bars formed in it. It became dry land, cottonwood and willow trees began to grow upon it, and it is now for the most part covered with valuable timber and susceptible of cultivation. It is very valuable, both on account of tbe timber growing upon it and the fertility of its soil. This suit is brought to recover a portion of tbe main channel lying between tbe middle of it, as it existed when tbe cut-off took place, and tbe Tennessee bank. Tbe claim of the State is that its sovereignty and territory extended to a line drawn along tbe middle of tbe Mississippi river, and that, it being a navigable stream, it bad title to the lands within its boundaries covered by tbe waters of tbe river, and, when tbe waters abandoned the bed, they remained its property, and it is entitled to recover them in this suit.
*62
*63We will now proceed to consider this boundary of Tennessee and tbe title which she acquired and has to tbe lands claimed. Tbe territory constituting tbe State of Tennessee, with perhaps small areas upon her northern and southern boundaries, acquired by conventions with adjoining States, was originally the western part of tbe colony and State of North Carolina, and her boundaries are tbe same as they were before ceded by that State. Charles II of England, in tbe sec*64ond and effective rojal charter of North Carolina, granted June 30, 1667, to Edward, Earl of Clarendon, and his associates, described the territory granted as “all that portion, territory or tract of land situated and being within our dominion of America aforesaid, extending north and eastward as far as the north end of Currituck river or inlet upon a straight westwardly line to Wyonoke creek, which lies within or about the degrees of thirty-six and thirty minutes northern latitude; and so west in a direct line as far as the south seas, and south and westward as far as the degrees of twenty-nine inclusive of northern latitude, and so west in a direct line as far as the south seas, together with all and singular the ports, harbors, bays, rivers and inlets belonging in the province and territory aforesaid, and also all the soils, lands, fields, woods, mountains, farms, lakes, rivers, bays, islets situate or being within the bounds limits last before mentioned, etc.”' 2 Cobb & Haywood’s Compilation, p. 1.
The western boundary of the territory granted was then unknown, but extended to the western boundary of the possessions of Great Britain in North America at that period. This boundary was, by the treaty between Great Britain, Prance, and Spain, made in February, 1763, fixed irrevocably upon a line drawn along the “middle of the Mississippi river.” 3 Jenkinson’s Treaties, 177; Iowa v. Illinois, 147 U. S., 2, 13 Sup. Ct., 239, 37 L. Ed., 55; Louisiana v. Mississippi, 202 U. S., 41, 26 Sup. Ct., 408, 50 L. Ed., 913. This line *65was afterwards recognized as the western boundary of the original thirteen States, or those whose territory extended to the Mississippi river, in the treaty made by them with England September 3, 1783. 8 Stat., 81, 82.
Virginia and North Carolina then owned all the territory bordering upon the east bank of the Mississippi river from near its source to the southern boundary of'Tennessee, now composing the States of Illinois, Kentucky, and Tennessee, and afterwards ceded it to the United States for the purpose of forming new States to be admitted to the Union. North Carolina ceded her part of the territory in December, 1789, and authorized her senators in the congress of the United States to convey it, which they did February 25, 1790, and the conveyance was accepted by an act of congress passed for that purpose April 2, 1790. 1 Stat., 106, c. 6. The territory ceded and conveyed is described in the cession act and conveyance a*s follows:
“All right, title and claim which this State [North Carolina] has to the sovereignty and territory of the lands situated within the chartered limits of this State, west of a line beginning on the extreme height of the Stone Mountain, at the place where the Virginia line intersects it, running thence along the extreme height of said mountain to the place where the Wautauga river breaks through it; thence a direct course to the top of the Yellow Mountain, where Bright’s Road crosses the same; thence along the ridge of said moun*66tain, between the. Avaters of Doe river and the waters of Rock creek, to tlie place where the road crosses the Iron Mountain; from thence along the extreme height of said mountain to where the Nolichucky river runs through the same; thence to the top of the Bald Mountain; thence along the extreme height of said mountain to Painted Rock, on the French Broad river; thence along the highest ridge of said mountain to the place where it is called the Great Iron or Smoky Mountain; thence along the extreme height of said mountain to the place where it is called Unicoi or Unaka Mountain, between the Indian towns of Cowee and Old Chota; thence along the main ridge of said mountain to the southern boundary of this State.” Cobb & Haywood’s Compilation, 7, 8, 9, 10.
The inhabitants of this Territory, through their representatives, organized as a State and adopted a constitution February 6, 1796, which described the territorial boundaries of the new State of Tennessee in the language of the cession act; and this State, Avith this constitution, was by congress admitted into the union as a sovereign State June 1, 1796. 1 Stat., 491, c. 47. The act of congress" does not define the limits of the State, further than to declare that it shall have and be composed of all the territory ceded by North Carolina and they are therefore controlled by tiie cession act and the constitution of the State. They are repeated in substantially the same language as in *67those instruments in the constitutions adopted in 1834 and 1870. The description given in the latter is in these words':
“That the limits and boundaries of this State being-ascertained, it is declared that they are as hereinafter mentioned, that is to say, beginning on the extreme height of Stone Mountain at the place where the line of Virginia intersects it, in latitude thirty-six degrees thirty minutes north, running then with the extreme height of said mountain [and then with other mountains therein stated and named] to the southern boundary of this State as described in the act of cession of North Carolina to the United States of America; and that all territory, land and waters, lying west of said line as before mentioned 'and contained within the chartered limits of North Carolina are within the limits and boundaries of this State, over which the people have the right of exercising sovereignty and the right of sale, so far as it is consistent with the constitution of the United States, the bill of rights, constitution of North Carolina, the cession act of said State and the ordinances of congress for the government of the territory northwest of the Ohio.” Const. Tenn., art. 1, sec, 31.
The general description of the boundaries of the State, preceding a specific description contained in the Code adopted in 1858 is in the language of the constitution. Code, sec. 60. In the same chapter (section *6869) the boundary between this State and the State of Arkansas is described as follows:
I “The western boundary of the State of Tennessee is in the middle of the stream of the Mississippi river including within the State of Tennessee all such islands as are held under grants from the States of Tennessee and North Carolina.”
This section must be construed to mean the same as the cession act and provision of the constitution; otherwise it is invalid. Congress first authorized the State of Tennessee as its agent to dispose of all unappropriated and ungranted lands within its territory for certain purposes, and afterwards in 1846 released and surrendered to it all right and title of the United States to the lands within the State acquired by them from North Carolina then ungranted and unappropriated.
The State of Arkansas as well as those of Missouri and Iowa were part of the territory of Louisiana owned at various times by Prance and Spain, and finally acquired by the United States from the former by purchase in 1808. Arkansas was admitted into the union as a sovereign State by an act of congress approved 'June 15, 1836 (5 Stat., 50, c. 100), and its eastern boundary was designated and defined as the “middle of the main channel” of the Mississippi river. This boundary is also embodied in the several constitutions of that State subsequently adopted. There is no difference between the “middle of the Mississippi river,” as the western boundary line of Tennessee is described, *69and the “middle of tbe main channel” of that river, as the eastern boundary line of Arkansas is defined. They mean the same thing, and the words “main channel” were evidently intended to make the common boundary more definite by designating the larger channel, where there existed two or more channels on account of the numerous islands to be found in the river. The use of the words “middle of the main channel” could not have been intended to designate a different boundary line than that of Tennessee as it then existed, because congress had no power to change the boundaries of Tennessee as fixed by it when that State was admitted to the union in 1796. Const. U. S., art. 4, sec. 8; Louisiana v. Mississippi, 202 U. S., 40, 26 Sup. Ct., 408, 50 L. Ed., 913.
While complainant and the defendants agree that the western boundary line of Tennessee is as declared and fixed by the treaties and legislative enactments which we have briefly stated — that is, that the middle of the Mississippi river as it ran in 1763 is the line that separates the jurisdiction of Tennessee from that of Arkansas — yet they disagree as to what was meant by the expression “middle of the river” and how it is now to be interpreted. Complainant insists that the contracting parties and legislative bodies, establishing this boundary by these words, “middle of the river,” meant the middle of the main channel of the river, or a line along the river bed equidistant from the visible, defined, and substantially established banks confining *70tbe waters on either1 side — that is, the line between it and Arkansas; while the defendants contend that they meant the center of that part of the waters or stream of the river which is deepest and is usually used by steamboats and other craft plying the river, or, in other words, the center of the channel of commerce. This is an important question, affecting the States of Tennessee and Arkansas in their sovereign capacity, and their jurisdiction along this entire joint boundary line, and the decision of this case, since at the point where the water flowed over the lands in controversy the deep-water channel used by boats in ascending and descending the river previous to 1876 ran much nearer to one bank than the other, the Tennessee bank, and it is entitled to the most careful consideration by the court. The defendants insist that under the laws of nations and the weight of decisions of the courts of this country, where a navigable river or the middle of such river is made the boundary line separating coterminous States or nations, the center of the channel of commerce is the true and correct line between them, and that this rule should apply in this case. The - State controverts this, and maintains that the weight of authority in such cases is that the separating line is midway the channel or bed of the river and equidistant from the visible and established banks within which the waters are confined and flow. The State further insists that not only the general rule is in favor of this contention, but that, if it were otherwise, it would not *71be applicable to this case, because the line has been fixed as claimed by it by treaties, legislation, and long usage, acquiescence, and possession by the two sovereign States interested, complainant and Arkansas. Before proceeding to the direct question involved, we think it will throw some light on the authorities we will discuss to notice what are the constituent parts of rivers or other streams as defined and used by the courts. In Lux v. Haggin, 69 Cal., 417, 10 Pac., 770, it is said:
“A watercourse is defined to consist of bed, hanks, and water. It .must be made to appear that the water usually flows through a regular channel with banks or sides. The bed and banks, or the channel, is in all cases a natural object to be sought after, not simply by application of any abtract rule, but, like other natural objects, to be sought for and found by the distinctive appearance it represents. Whether, however, worn deep by action of the water or following the exact depression without any marked erosion of soil or rock; whether distinguished by difference of vegetation or otherwise rendered perceptible — a channel is necessary to the constitution of a watercourse.”
In Benjamin v. River Improvement Company, 42 Mich., 628, 4 N. W., 483, it is said that the channel of a river is the passageway between the banks through which its waters flow; and in Larrabee v. Cloverdale, 131 Cal., 96, 63 Pac., 143, a channel is said to include, not only all the channels through which under existing conditions of the country the water naturally flows, *72but new channels through which it may afterwards flow.
We, think;, from examination of a number of cases bearing more or less upon this subject, that the channel of the river and the bed of the river ordinarily mean the same thing, and are understood to describe that depression on the earth’s surface in which the waters of the stream are confined and flow in its ordinary stages, unaffected by freshets or droughts. Houghton v. Railroad Co., 47 Iowa, 370; Cessill v. State, 40 Ark., 504; Railroad v. Ramsey, 53 Ark., 314, 13 S. W., 931, 8 L. R. A., 559, 22 Am. St. Rep., 195; Stover v. Jack, 60 Pa., 339, 100 Am. Dec., 566; Howard v. Ingersoll, 13 How. (U. S.), 381, 14 L. Ed., 189; Alabama v. Georgia, 23 How. (U. S.), 505, 16 L. Ed., 556; Branham v. Turnpike Co., 1 Lea (Tenn.), 704, 27 Am. Rep., 789; Dunleith & Dubuque Bridge Co. v. Dubuque, 55 Iowa, 558, 8 N. W., 443.
The precise question we are now considering was before the supreme court of Arkansas in 1883, and that court construed the treaties we have referred to, and the act of congress admitting Arkansas into the union, as contended for by Tennessee, and held the line between that State and Tennessee to be the middle of the main channel or bed of the Mississippi river, equidistant from the visible banks confining its waters, and not one along the so-called center of the channel of commerce. The opinion is an able and interesting one, and since it is a decision of the direct question here involved by *73the highest court of one of the two sovereign States interested, in a case to which that State was a party, we quote from it at length. Cessill, the plaintiff in error, was indicted and convicted of illegally selling liquor from a boat anchored in the Mississippi river, and appealed. Eakin, J., speaking for the court, said:
“It will be observed that the principle upon which the court proceeded is that the line of deepest water in the river bed is the boundary of the State, and continues such as it fluctuates.
“The' act of congress admitting the State into the union, approved June 15, 1836 (5 Stat., 50, c. 100), designated for the eastern boundary ‘the middle of the main channel’ of the Mississippi river, between latitude thirty-six degrees north and the northeast corner of the State of Louisiana, at a point to be determined by extending the north line of the latter State to the middle of the said channel. This description was embodied in the constitution of 1836, and repeated in that of 1864. It was also adopted in the constitution of 1868, with the explanation that the said boundary should include a certain island known as ‘Belle Point Island.’ In addition to this, the present constitution provides generally that the State shall embrace ‘all other land originally surveyed and included as a part of the territory of the State of Arkansas.’ No question arises in this case upon either of the two qualifications, and the sole matter left for us to decide is this: What is meant by the ‘main channel,’ and what is the middle of it?
*74“The channel of a river, bay, or sound is, in boatmen’s parlance, the course over its bed over which the water is deepest and the navigation safest. This may be irrespective of the current or distance from the shore. In questions of geography or boundaries, however, it is more generally used to designate the depression of a bed below the permanent banks, forming a conduit along which waters flow, and which may be at some times full and at others nearly, if not quite, dry. In this sense it is of common use in law. It is the more obvious signification in connection with boundaries, inasmuch as it presents something of a permanent nature, or at least at all times visible, and, when changed, leaving traces of the old landmarks. In this sense we speak of bayous — Bartholomew and Atchafalaya — as old channels ¿of the Arkansas and Red rivers. They have permanent features independent of water; whereas, channels in the sense of the river pilot are ever shifting, invisible, discoverable only by patient soundings, and then imperfectly. We cannot suppose that such channels would be adopted as State boundaries, or as references to determine them.
“The Mississippi river is full of islands, having water beds on each side. The object of the description of the boundary was to afford the means of determining whether or not any given island was within the State, by taking the largest of these water conduits as the true river. The middle of the main channel, then, must mean the point or line long the river bed equidistant *75from tbe permanent and defined banks of tbe ascertained channel on either side. Even this line is a fluctuating one, but in a far less and no very inconvenient degree. Gradual attrition on one side, with accretion on tbe other, making a change in tbe permanent banks, might perhaps change tbe boundary with regard to absolute space. But it is not necessary, for practical purposes, that a boundary should be a fixed mathematical line, and this could only apply to changes in the banks of a channel which remains substantially the same. For, if the main body of the water were to find a new channel and abandon the old one, leaving intervening lands in a natural state, the old boundary would be still ascertainable, and would govern. This has been decided in the case between Kentucky and Missouri (infra), and results, with regard to surveyed lands, from the additional clause, above noted, in the constitution of 1874. It seems that the largest channel determines which is the river, and the central line of that makes the State boundary.
“The boundary line in question is a very old one, and does not concern this State alone. It originated with the treaty between England, France, find Spain, in February, 1768, which made the middle of the Mississippi river the boundary between British and French territories. This line has been ever since observed in subsequent treaties, in federal legislation, in State constitutions, and in judicial decisions, and there are not lacking unmistakable indications of the meaning *76of tbe middle of the riyer. For instance, in the treaty between the United States and Spain, in October, 1795, before onr purchase of Louisiana, the fourth article provides That the western boundary of the United States, which separates them from the Spanish colony of Louisiana, is in the middle of the channel or bed of the river Mississippi, from the- northern boundary of said States to the completion of the thirty-first degree of latitude north of the equator.’
“In the case of Myers v. Perry et al., 1 La. Ann., 372, which resulted from a steamboat collision on the Mississippi, it became necessary to ascertain the locus in quo as affecting jurisdiction between the States of Louisiana and Mississippi. The middle of the river was taken as the boundary line, without any reference to depth of the water. See, also, on the same subject, a case very replete with historical learning, that of Morgan & Harrison v. Reading, reported in 3 Smedes & M. (Miss.), 366, in which this great empire boundary is described, with reference to the treaty Of 1763, as ‘a line drawn along the middle of the Mississippi.’ This would not be a good description of a steamboat track, zigzagging from hank to bank amongst sand bars in low water. . . .
“In the case of Missouri v. Kentucky, 11 Wall. (U. S.), 395, 20 L. Ed., 116, which was a contest between States for jurisdiction over Wolf Island, in the Mississippi, Mr. Justice Davis said that by virtue of the treaties above named, together with the treaty of- peace *77with England in 1783, the ancient right of Virginia, to which Kentucky had succeeded, extended to the middle of the bed of the Mississippi river.
“It seems that, where there are several channels, the principal one is considered the river, and in this the medium filum makes the boundary.
“There was only one channel in this case, which was the river bed between the Arkansas and Tennessee shores at Osceola. The court and attorneys treated the case throughout as if channel meant the line of the deepest water sought by boatmen, and the instructions were given on one side and refused on the other with reference to this idea. The river bed being the same as in 1784, no question could arise as to change of channel. The instructions asked by the defense were erroneous, but those given for the State were equally so, being based on a false theory as to the meaning of channel. It should have been left to the jury to determine whether the position of the boat was nearer to the Arkansas or the Tennessee main bank, and to have found the defendants guilty or innocent accordingly.” Cessill v. State, 40 Ark. 501.
We concur fully with the supreme court of Arkansas in the construction given the treaties of 1763 and 1783 in that opinion, and hold, as held by that court, that the boundary line between the British possessions in America, which then included all the territory now composing the States bordering upon and having for their western boundary the Mississippi river, and the territory of *78Louisiana then belonging to Spain, was fixed and defined as a line along tbe middle of tbe main channel of tbe river, equidistant from tbe visible and permanent bants confining its waters, and that tbe several acts of congress admitting into tbe union the States lying upon both sides of tbe river at various times, in calling for tbe middle of tbe river and tbe middle of the main channel. or stream of tbe river, bad reference to these treaties and must be construed to mean tbe same thing. This question has not before been before this court; but in a case involving property rights upon an unnavigable stream, called for as a boundary line of private estates, it was held that “tbe thread of tbe stream is tbe middle line between tbe shores, irrespective of tbe. depth of the channel, taking them in tbe natural and ordinary stage of tbe water, at medium height, neither swollen by freshets nor shrunk by droughts.” Branham v. Turnpike Co., 1 Lea (Tenn.), 706, 27 Am. Rep., 789. Tbe general understanding of tbe people and tbe constituted authorities of Tennessee has been and is that tbe line separating tbe State from Arkansas is as defined in tbe case of Cessill v. State, supra. This appears from an act of the general assembly of tbe State approved April 15, 1903 (chapter 420, p. 1215, Acts 1903), in which the lands in controversy and all others lying upon the Tennessee side of tbe middle of tbe old bed of the river are declared to be tbe property of tbe State, and tbe governor authorized to appoint commissioners, to act with other commissioners to be appointed by tbe State of Arkan*79sas, to run and'mark tbe line, and also to report to tbe governor tbe extent and value of sucb lands. Tbe general assembly of Arkansas passed a similar act, but it was vetoed by tbe governor of that State, and therefore no commissioners were appointed under tbe act passed by tbe legislature of Tennessee. This suit was brought by direction of tbe governor of this State, and is not only an acquiescence in tbe boundary line as defined by tbe authorities of Arkansas, but an assertion of jurisdiction up to that line and title to property within it. We think, whatever may be tbe construction of tbe treaties defining this great boundary line, or tbe acts of congress admitting other States bordering upon it, that tbe concurrence of- Tennessee and Arkansas in tbe interpretation of tbe treaties and legislation affecting their boundary line is effective between them, and controlling-in this and other cases involving tbe question.
These same treaties, we have seen, which define tbe common boundary line of all tbe States bordering upon both sides of tbe Mississippi river, in connection with tbe acts of congress admitting those States into tbe union, have been frequently construed by other courts, and in every case that has been called to our attention, with two exceptions, all these courts have concurred with tbe conclusions reached in tbe case of Cessill v. State, supra.
Tbe boundary line separating tbe States of Louisiana and Mississippi, Missouri and Kentucky, Missouri and Illinois, and Iowa and Illinois, where tbe Mississippi flows between them, is defined in tbe several acts of con*80:gress admitting these States into the union in words similar to those defining the line between Tennessee and .Arkansas,* that is, “the middle of the Mississippi river,” or “the middle of the main channel of the river.” We have seen, from cases cited in the opinion of the court in Cessill v. State, supra, of Myers v. Perry, 1 La. Ann., 372, and of Morgan & Harris v. Reading, 3 Smedes & M., 366, that the supreme courts of Louisiana and Mississippi have both construed the treaty of 1763, and the acts of congress in relation to it, to define and fix the line equidistant from the banks of the river. The supreme court of Iowa has so held in a case involving the line between it and the State of Illinois. In relation to what is meant by the middle of the channel it is there said:
“The course of navigation, which follows what boatmen call the channel, is extremely sinuous and often •changing, and is unknown except to experienced navigators. On the other hand, the bed of the main river, designated by the word ‘channel5 used in its primary sense, is the great body of water flowing down the stream. It is broad and well defined by islands or the main shore. It cannot be possible that congress and the people of the State in describing its boundary used the word ‘channel5 to describe the sinuous, obscure, and changing line •of navigation rather than the broad and distinctly defined bed of the main river. The center of this river bed -channel may be readily determined, while the center of the navigable channel often could not be known with *81certainty. Tbe first is a fit boundary line of a State. The second cannot be.” Dunleith & Dubuque Bridge Co. v. Dubuque County, 55 Iowa, 558, 8 N. W., 443.
The case of Missouri v. Kentucky, 11 Wall. (U. S.), 395, 20 L. Ed., 116, involved a question of the jurisdiction oyer Wolf Island in the Mississippi river. Construing the treaty of 1763 between England, Prance, and Spain, Mr. Justice Davis, speaking for the court, said:
“It is unnecessary for the purposes of thisi suit to consider whether on general principles the middle of the channel of the navigable river which divides coterminous States is not the true boundary between them, in the absence of express agreement to the contrary, because the treaty between Prance, Spain, and England in February, 1763, stipulated that the middle of the Mississippi river should be the boundary between the British and French territories on the continent of North America. And this line, established by the only sovereign powers at that time interested in the subject, has remained ever since as they settled it. It was recognized by the treaty of peace with Great Britain in 1783, and by different treaties since then, the last of which resulted in the acquisition of the Territory of Louisiana (embracing the country west of the Mississippi) by the United States in 1803. The boundaries of Missouri, when she was admitted into the Union as a State in 1820, were fixed on this basis, as were those of Arkansas in 1836. And Kentucky succeeded in 1792 to the ancient *82right and possession of Virginia, which extended by virtues of these treaties to the middle of the bed of the Mississippi river.”
The act of congress passed April 18,1818 (3 Stat.,429, c. 67), enabling the people of the Territory of Illinois to adopt a constitution and organize a State, defined the western boundary of the State as follows: “Starting-in the middle of Lake Michigan, at north latitude forty-two degrees, thirty minutes, thence west to the middle of the Mississippi river, and thence down the Mississippi river- to its confluence with the Ohio river.”
The case of St. Louis v. Rutz, 138 U. S., 226, 11 Sup. Ct., 337, 34 L. Ed., 941, was brought to recover an island in the Mississippi river, and involved the location of the line separating the States of Missouri and Illinois. The island was found to be upon the eastern side of the center of the main channel and bed of the river, and the decree was for the defendant. It is there said:
“As the law of Illinois confers upon the owner of land in that State which is bounded by or fronts on the Mississippi river the title in fee to the bed of the river, to the middie thereof, or so far as the boundary of the State extends, such riparian owner is entitled to all the lands in the river which are formed on the bed of the river or of the middle of its width. That being so, it is impossible for the owner of an island which is situated on the west side of the middle of the river, in the State-of Missouri, to extend his ownership by mere accretion to land situated in the State of Illinois, the title in fee *83to which is vested by the law of Illinois in the riparian owner of the land, in that State.”
In the cases of Jones v. Soulard. 24 How. (U. S.), 41, 16 L. Ed., 604, and St. Louis Public School v. Risley. 10 Wall. (U. S.), 91, 19 L. Ed., 850, it is held that, under the act of Congress admitting Missouri to the union and defining its eastern boundary as the middle of the main, channel of the Mississippi river, the line was the middle of the river, without any reference whatever to where the channel of commerce ran, and presumably to a line midway between the established banks of the river.
In the cases of Nebraska v. Iowa, 143 U. S., 359, 367, 12 Sup. Ct., 396, 36 L. Ed., 186, and Missouri v. Nebraska, 196 U. S., 23, 25 Sup. Ct., 155, 49 L. Ed., 372, both of which involved controversies of jurisdiction growing out of sudden and violent changes made by the Missouri river in its channel similar to the one made by the Mississippi in this case, it was held that the boundary line between the commonwealths, which were parties to those cases, respectively, remained fixed in the center of the old river bed, thus in effect holding that previous to the avulsions by which the channel of the river was changed the State line was the middle of the channel; that is,, a line midway between the banks of the river. In the case of Nebraska v. Iowa, 143 U. S., 361, 12 Sup. Ct., 396, 36 L. Ed., 186, the following quotation is made, with approval, from the opinion of Attorney-General Gushing-in a matter of dispute between the United States and Mexico as to the international boundary at the place-*84where the Eio Grande had raade a change in its channel:
“But, on the other hand, if, deserting its original bed, the river forces for itself a new channel in another direction, then the nation through whose territories the river thus breaks its' way suffers injury by loss of territory greater than the benefit of retaining the natural river boundary, and that boundary remains in the middle of the deserted river bed. For, in truth, just as a stone pillar constitutes a boundary, not because it is a stone, but because of the place in which it stands, so a river is made the limit of nations, not because it is running water bearing a certain geographical name, but because it is water flowing in a given channel and within given banks, which are the real international boundary.”
The only cases that have been called to our attention supporting the contention of the defendants are those .of Buttenuth v. St. Louis Bridge Co., 123 Ill., 535, 17 N. E., 439, 5 Am. St. Rep., 545, and Iowa v. Illinois, 147 U. S., 1, 13 Sup. Ct., 239, 37 L. Ed., 55, both involving the line in the Mississippi river separating the States of Iowa and Illinois. The former was decided first, and is cited and approved in the latter. Mr. Justice Field, delivering the opinion of the court, says:
“When a navigable river constitutes the boundary between two independent States, the line defining the point at which the jurisdiction of the two separates is well established to be the middle of the main channel of the stream. The interest of each State in the navigation of *85the riyer admits of no other line. The preservation by each of its equal right in the navigation of the stream is the subject of paramount interest. It is therefore laid doAvn in all well-recognized treatises on international law of modern times that the middle of the channel of the stream marks the true boundary between the adjoining States up to which each State on its side will exercise jurisdiction. In international law, therefore, and by the usage of European nations, the term ‘middle of the stream/ as applied to a navigable river, is the same as the middle of the channel of such stream, and in that sense the terms are used in the treaty of peace between Great Britain, Prance, and Spain, concluded at Paris in 1763. By the language, ‘a line drawn along the middle of the River Mississippi from its source to the River Iberville/ as there used, is meant along the middle of the channel of the River Mississippi. Thus Wheaton, in his Elements of International Law (8th Ed.), section 192, says: ‘Where a navigable river forms the boundary of coterminous States, the middle of the channel, or “thalweg,” is generally taken as the line of separation between the two States, the presumption of law being that the right of navigation is common to both; but this presumption may be destroyed by actual proof of prior occupancy and long, undisturbed possession, giving to one of the riparian proprietors the exclusive title to the entire river.’
“And in section 202, while thus stating the rule as to the boundary line of the Mississippi river being the mid-*86dlé of the channel, he states that the channel is remarkably winding, ‘crossing and recrossing perpetually from one side to the other of the general bed of the river.’
“Mr. Creasy, in his First Platform on International Law, p. 222, section 231, expresses the same doctrine. He says:
“ ‘It has been stated- that, where a navigable river separates neighboring States, the “thalweg,” or middle ■of the navigable channel, forms the line of separation. Formerly a line drawn along the middle of the river, the “medium filum aquae” was regarded as the boundary line, and still will be regarded prima facie as the boundary line, except as to those parts of the river as to which it can be proved that the vessels which navigate those parts keep their course habitually along some channel different from the medium filum. Where this is the case, the middle of the channel of traffic is now considered to be the line of demarcation.’ ”
And after citing several other works on International Law, Justice Field proceeds:
“The reason and necessity of the rule of international law as to the midchannel being the true boundary line of a navigable river separating independent States may not be cogent in this country, where neighboring States are under the same general government, as in Europe, jet the same rule will be held to obtain, unless changed by statute or usage of so great length of time as to have acquired the force of law.
“As we have stated, in international law and by the *87usage of European nations, tbe terms ‘middle of tbe stream’ and ‘midchannel’ of a navigable river are synonymous and interchangeably used. Tbe enabling act of April 18, 1818 (3 Stat., 428, c. 67), under which Illinois adopted a constitution and became a State and was admitted into tbe union, made tbe middle of tbe Mississippi river tbe western boundary of tbe State. Tbe enabling act of March 6,1820 (3 Stat., 545, c. 22, section 2), under which Missouri became a State and was admitted into tbe union, made tbe middle of tbe main channel of tbe Mississippi river tbe eastern boundary, so far as its boundary was coterminous with tbe western boundary of Illinois. Tbe enabling act of August 6, 1846 (9 Stat., 56, c. 89), under which Wisconsin adopted a constitution and became a State and was admitted into tbe union, gives tbe western boundary of that State, after reaching tbe River St. Croix, as follows: ‘Thence down the main channel of said river to tbe Mississippi, thence down tbe center of tbe main channel of that [Mississippi] river to tbe northwest corner of tbe State of Illinois.’ Tbe northwest corner of tbe State of Illinois must therefore be in tbe middle of tbe main channel of tbe river which forms a portion of its western boundary. It is very evident that these terms, ‘middle of tbe Mississippi river,’ and ‘middle of tbe main channel of tbe Mississippi river,’ and ‘center of tbe main channel of that river,’ as thus used, are synonymous. It is not at all likely that tbe congress of the United States intended that those terms, as applied to tbe Mississippi *88river separating Illinois from Iowa, should have a different meaning when applied to the Mississippi river separating- Illinois from Missouri, or a different meaning when used as descriptive of a portion of the western boundary of Wisconsin. They were evidently used as signifying the same thing.”
He then quotes extensively from the case of Dunleith & Dubuque Bridge Co. v. County of Dubuque, supra, and Buttenuth v. St. Louis Bridge Co., supra, and concludes :
“The opinions in both these cases are able, and present, in the strongest terms, the different views as to the line of jurisdiction between neighboring States, separated by a navigable stream; but we are of the opinion that the controlling consideration in this matter is that which preserves to each State equality in the right of navigation in the river. We therefore hold, in accordance with this view, that the true line in navigable rivers between States of the union which separates the jurisdiction of one from the other is the middle of the main Channel of the river. Thus the jurisdiction of each State extends to the thread of the stream; that is, to the ‘midchannel,’ and, if there be several channels, to the middle of the principal one, or rather, the one usually followed.”
This case is in direct conflict with the previous cases of Missouri v. Kentucky, St. Louis v. Rutz, Jones v. Soulard, St. Louis Public School v. Risley, and Nebraska v. Iowa, above cited, which involved practically the same *89question, and the first four construed the same treaties and acts of congress. They are not differentiated, overruled, or even referred to in the case of Iowa v. Illinois. The decision in these cases is based upon the proper construction of the treaties and acts of congress admitting the States into the union, and not upon the laws of nations. We are better satisfied with the reasoning of these cases than Ave are Avith that of the last one, and prefer to folloAV them. We think they correctly construe and interpret the treaties and legislation controlling and defining the boundary lines of the coterminous States upon the Mississippi river according to the intention of the powers making the treaties and of congress in admitting those States into the union.
The case of Iowa v. Illinois was 'decided avowedly upon the rules of international law, and was not a construction of the treaties defining the boundaries under consideration, with a view of ascertaining the intention of the parties making them, and the controlling consideration with the court in the application of the rules of international law to the case was the preservation of equality in the right of navigation to the river to the co-terminous States.
There is much conflict in the opinions of test-writers upon the law of nations upon this question; but the weight of authority is that at the time the treaties in question were made, in the absence of a convention establishing it otherwise, the true boundary between nations bordering upon navigable waters was a line midway be=*90tween the visible and fixed banks of the stream. Mr. Creasy, in his work on International Law, the chief authority cited by the court in Iowa v. Illinois, says that “formerly the line drawn along the middle of the water, the medium filum aquae, was regarded as the boundary line, and still will be regarded prima facie as the boundary line except as to those parts of the river as to which it can be proved that the vessels which navigate those parts keep their course habitually along some channel different from the medium filum ”
It will not be amiss here to call attention to what Mr. Angelí, in his work on Water Courses, says upon this question:
“By the middle of the channel is meant the thread of the stream, the filum aquae; that is, the middle line between the shores upon each side, without regard to the channel, or lowest parts or deepest parts of the water. In ascertaining the shores, the water 'line on each side to measure, it will be proper to find where these lines are when the water is in its natural and ordinary stage, at medium height, neither swollen by freshets nor shrunk by droughts.”
Another author says:
“A river that separates two jurisdictions is not to be considered barely as water, but as water confined in such and such banks and running in such and such channel ; hence there is water having a bank and a bed over which the waters flow in its channel, meaning by the word ‘channel’ the place where the river flows, includ*91ing the whole breadth of the river.” Grotiiis, p. 18, c. 2.
We think some confusion has arisen, both in the textbooks and in the decisions, in relation to this matter, by failure to properly differentiate those cas.es where there are several channels caused by the existence of islands in the stream, where it is held that the line is the center of the main channel, meaning the largest division of the river at that place, from those where there is only one channel'to be followed.
We do not deem it necessary, however, to enter into a discussion of the laws of nations upon the subject.
Whatever may be the general rule, we do not think it applies or is controlling in this case. General rules of international law cannot be invoked when the matter in question has been settled by the parties in interest otherwise, either by agreement, convention, acquies- ■ cence, or long and undisturbed occupancy and possession. Twiss, International Law, 127; 1 Halleck, International Law, 50.
We do not think the high contracting parties to the treaty between Great Britain, France, and Spain, made in 1763, in which the line separating the British possessions in North America and the Territory of Louisiana, defined to be the “middle of the river,” meant a channel of commerce as it varied and shifted from side to side of the stream, but a line midway between the banks. We understand from Mr. Greasy, as above quoted, that at that day the call for a navigable stream *92as tbe boundary line between two nations was construed to be tbe middle of tbe bed of tbe river, a line equidistant from tbe respective banks, and will be regarded prima facie so at tbis day. Wben these treaties were made tbe country along tbe banks of tbe Mississippi river above tbe city of New Orleans was practically unknown and uninhabited, save by tbe Indians. Tbe river was not navigated. There were no boats upon tbe river, and its various windings bad not been surveyed or mapped, nor its depth sounded. There was no known fixed channel of commerce in tbe river." Had a question of jurisdiction arisen under tbe construction of tbe treaty given in Toma v. Illinois, it would have been impossible to determine whether tbe occurrence took place within tbe territories of Great Britain or those of Prance, or Spain. We think, considering tbe then existing conditions, every presumption is that- tbe parties intended tbe middle of tbe channel between tbe banks which control the waters of tbe stream, tbe bed of tbe river, should be the line separating their respective Territories. Tbe banks were visible, and a line midway between them could be ascertained wben occasion required it. No channel of commerce existed, and a line in tbe center of it could not possibly be located. The first constructions of these ancient treaties were in accordance with tbe contention made by Tennessee in tbis case. In tbe treaty made by tbe United States with Spain in October, 1795, before tbe purchase, of Louisiana, tbe fourth article provided “that tbe western boundary of tbe United’ States, *93which separates them from the Spanish colony of Louisiana, is in the middle of the channel or bed of the River Mississippi, from the northern boundary of said States to the completion of the thirty-first degree of latitude north of the equator.” Cessill v. State, 40 Ark., 505. This was not only an interpretation of the former treaties, but it superseded them. The decisions of all the courts of last resort of the several States, as well as those of the United States, involving this boundary line, with the exception of those of Buttenuth v. St. Louis Bridge Co., supra, and Iowa v. Illinois, supra, have been favorable to the contention that the line runs midway between the banks of the river, and it is only at a late day by those cases that a doubt was suggested or arose as to the true and correct line which formerly separated the British possessions in America from those of Prance and Spain, and subsequently a number of the largest and most influential States of the union. The former construction has become a rule of property and should not be disturbed. We are not impressed with the argument that it is necessary to hold the line to he along the so-called channel of commerce in order to preserve to the several States interested in the question equality in the right of navigation of the river. We do not think such necessity exists. Where a navigable river constitutes the boundary between two States, the middle of the- channel separating their respective jurisdictions, both are presumed to have free use of the whole of it for the purposes of commerce. The whole river is of *94right common to both nations as a public highway. The Apollon, 9 Wheat. (U. S.), 362, 371, 6 L. Ed., 111; Handly v. Anthony, 5 Wheat. (U. S.), 374, 5 L. Ed., 113; Wharton, Int. Law Digest, section 30; Gould on Waters, 202; Wheaton, Int. Law, sections 192, 193. The free navigation of the Mississippi river by citizens of the United States was expressly provided for and preserved in the treaty made by the United States with Great Britain in 1783, and again in that made by the United States with Spain, October 27, 1795. The right of the citizens of the several States of the union to navigate the waters of this river has been frequently declared by acts of congress. It is asserted in the constitutions of all the states bordering upon its waters, and is so well established that no possible apprehension can be entertained that it will be interfered with. Act Gong. May 18,1796,1 Stat., 464, c. 29; Act. Cong. June 1,1796, 1 Stat., 490, c. 46; Act Cong. March 3,1803, 2 Stat., 229, c. 27; Act Cong. March 26, 1804, 2 Stat., 277, c. 35; Act Cong. Eeb. 20, 1811, 2 Stat., 641, c. 21; Act Cong. March 3, 1811, 2 Stat., 662, c. 46; Act. Cong. April 8, 1812, 2 Stat., 701, c. 50; Act Cong. June 4, 1812, 2 Stat., 743, c. 95; Act. Cong. March 1, 1817, 3 Stat., 348, c. 23; Act Cong. May 8, 1817; Gould on Waters, section 68; Const. Tenn., 1870, art. 1, section 29.
The commerce clause of the constitution of the United States, all other things aside, affords ample protection to the right of every citizen to the free navigation of the river, whether the current be in one State or another, *95without fear of hindrance or burdens' imposed by such States. There can be no doubt of this.
The reasons for haying a fixed, certain, and visible line, such as the middle of the channel as measured from the respective banks of the river, we think, greatly outweigh those advanced in support of the decision of the case of Iowa v. Illinois. It is of the highest importance to the adjoining States that the location of the boundary line between them be certain and susceptible of easy proof; otherwise, they will be greatly embarrassed in the enforcement of their criminal laws, the assessment and collection of taxes, and many other things in the ordinary and common exercise of sovereignty. It is easy to conceive cases Avhere so much doubt could be thrown upon the location of the channel of commerce.that the jurisdiction of either State to punish crime committed upon the river would be entirely defeated.
But the question has been settled by the duly constituted authorities of Tennessee and Arkansas by judicial decisions, legislation, and other authorized official actions, long acquiescence, the exercise of jurisdiction unchallenged, and other acts amounting to an agreement or convention. The highest court of Arkansas, in a case to which the State was a party, and at its instance, in the assertion of its sovereignty and jurisdiction, has defined the limit between the two States to be the line midway between the visible banks of the river, and enforced the criminal laws of the State up1 to that line. The general assembly of Tennessee has claimed title to the lands *96formerly covered by tbe waters of tbe river up to tbe same line, and tbe governor of tbe State lias directed and authorized that tbe proper officers institute tbis suit to recover such lands. Both States agree upon it as tbe true and correct line separating' tbeir territories, and others cannot be beard to complain. Such a course of conduct has been held to be conclusive upon States in controversies concerning tbeir boundaries. Indiana v. Kentucky, 136 U. S., 479, 10 Sup. Ct., 1051, 34 L. Ed., 329. Tennessee is now before tbis court as a suitor, asserting that her western boundary line lies midway between the visible banks of tbe Mississippi river. Arkansas, as a suitor in tbe case of Gessill v. State, supra, asserted that her eastern boundary line ran at tbe same place, and exercised jurisdiction to such line. Tbis is binding upon both States, and neither can now recede from such solemn admissions of tbe location of tbe line separating them.
Tennessee acquired title to all tbe soil under tbe waters of tbe Mississippi river to tbe limits of her jurisdiction. It is well-settled law that soil under the waters of navigable rivers, as well as tbe waters, are held by the States for tbe use and in trust of tbe public, so long as tbe river continues navigable.
Tbe United States has always recognized tbis rule in its disposition of tbe public domain. Tbe grants made by it lying upon navigable streams to private parties are limited by high-water mark, and the soil between that and tbe rivers and under tbeir waters vested *97in the States in which it lies. The States, however, may dispose of this property as they may allow, subject to the general control of congress over all navigable waters; this being a matter within their discretion and governed by the State authorities.
In Hardin v. Jordan, 140 U. S. 371, 372, 11 Sup. Ct., 808, 838, 35 L. Ed., 428, the title of the State to the soil under navigable waters within their boundaries, and their right to control and dispose of the same, before and after abandoned by the waters, is held and stated in these words:
“With regard to grants of the government for lands bordering on tide water, it has been distinctly settled that they only extend to high-water mark, and that the title to the shore and lands under water in front of the lands so granted inures to the State within which they are situated, if a State has been organized and established there. . Such title to the shore and lands under water isrregarded as incidental to the sovereignty of the State — a portion of the royalties belonging thereto and held in trust for public purposes of navigation and fishery — and cannot be retained or granted out to individuals by the United States. Pollard v. Hagan, 3 How. (U. S.), 212, 11 L. Ed., 565; Goodtitle v. Kibbe, 9 How. (U. S.), 471, 13 L. Ed., 220; Weber v. Harbor Commissioners, 18 Wall. (U. S.), 57, 21 L. Ed., 798. Such title being in the State, the lands are subject to State regulation and control, under the conditions, *98however, of not interfering with the regulations which may be made by congress with regard to public navigation and commerce. The State may even dispose of the usufruct of such lands, as is frequently done by leasing oyster beds in them, and granting fisheries in particular localities; also by the reclamation of submerged flats, and the erection of wharves and piers and other adventitious aids of commerce. Sometimes large areas so reclaimed are occupied by cities, and are put to other public or private uses, State control and ownership therein being supreme, subject only to paramount authority of congress in making regulations of commerce and in subjecting the lands to the necessities and uses of commerce. See Manchester v. Massachusetts, 139 U. S., 240, 11 Sup. Ct, 559, 85 L. Ed., 159; Smith v. Maryland, 18 How. (U. S.), 71, 15 L. Ed., 269; McCready v. Virginia, 94 U. S., 391, 24 L. Ed., 248; Martin v. Waddell, 16 Pet. (U. S.), 367, 10 L. Ed., 997; Den v. Jersey Co., 15 How. (U. S.), 426, 14 L. Ed., 757.
“The right of the States to regulate and control the shores of tide waters, and the land under them, is the same as that which is exercised by the crown in-England. In this country the same rule has been extended to our great navigable lakes, which are treated as inland seas, and also, in some of the States, to navigable rivers, as the Mississippi, the Missouri, the Ohio, and, in Pennsylvania, to all the permanent rivers of the State; but it depends on the law of each State to what waters *99and to what extent this prerogative of tbe State over the lands under water' shall be exercised.”
In Tennessee it has uniformly been held that grants to lands lying upon navigable streams extend to ordinary low-water mark only, and that the title to the bed of the stream remains in the State. Martin v. Nance, 3 Head, 649; Posey v. James, 7 Lea, 98; Goodwin v. Thompson, 15 Lea, 209, 54 Am. Rep., 410; Holbert v. Edens, 5 Lea, 204, 40 Am. Rep., 26; Stockley v. Cissna, 119 Fed., 829, 56 C. C. A., 324; Taylor v. Commonwealth, 102 Va., 759, 47 S. E., 875, 102 Am. St. Rep., 865; Holman v. Hodges, 112 Iowa, 714, 84 N. W., 950, 58 L. R. A., 673, 84 Am. St. Rep., 367.
In the case of Holbert v. Edens, supra, it is said:
“If a watercourse be navigable in a legal sense, the soil covered by the water, as well as the use of the stream, belongs to the public.”
In the case of Goodwin v. Thompson, 15 Lea, 215, 54 Am. Rep., 410, this court held, not only that the soil in navigable streams belonged to the State, but that it was not subject to entry or grant as other lands ; the statute providing for disposition of public lands not authorizing such grants. In that case it is said:
“We think the public use of our navigable rivers imperatively requires that the soil under the water should be in the State in trust for the public, and that title to the soil under such terms was not intended to be secured by individuals under our general land laws, and *100that any person setting up claim thereto must be able to show an express legislative grant.”
It is also well-established law that when the waters recede, or land is formed upon the bed of navigable rivers, as in case of islands forming in navigable waters, the property in such dry land is in the State, to be disposed of by it as its authorities may determine and direct. Morris v. Brooke (Del.), cited in Mulry v. Norton, 53 Am. Rep., 215, note; Hardin v. Jordan, 140 U. S., 371, 372, 11 Sup. Ct., 808, 838, 35 L. Ed., 428; Packer v. Bird, 137 U. S., 666-672, 11 Sup. Ct., 210, 34 L. Ed., 819; 2 Black. Com., 261; 11 Am. & Eng. Enc. Law (1st Ed.), 865.
The case of Morris v. Brooke is an instructive one, and the conclusions of the court are well supported by authority. We quote from it:
“New islands arising in the sea or in a navigable river prima facie belong, according to the common law, to the king in England, and in this country to the State. But this rule is not universal.
“The right to the new islands, and also to lands gained by alluvion or dereliction (in cases where they are not gained by insensible degrees), all of which are' governed by the same principles, follows the right to the soil which is covered with water. As the king is the proprietor in general of the soil covered with the sea or a navigable river, it is reasonable that he should have the soil where the water leaves it dry; and this stands on .the ground of the prerogative.
*101“But, where the right to the soil when covered with water belongs to a subject, he is entitled to all these increments. 2 Bl. Com., 262; Hale, cle Jure Maris, chs. 4 .and 6.
“This is illustrated by the law relative to islands aidsing in private rivers. If an island arises in the middle of such a river, it belongs in common to those-who have lands on each side thereof; but, if it be nearer to one bank than to the other, it belongs only to him who is proprietor of the nearest shore. Yet this, says Sir William Blackstone (2 Com., 261), seems only to be reasonable when the soil of the river is equally divided between the owners of the opposite shores; for, if the whole soil is in the freehold of any one man, as it usually is, wherever a several fishery is claimed, there it seems just (and so is the constant practice) that the lyotts, or little islands arising in any part of the river, shall be the property of him who owneth the piscary and the soil. The rules relative to the sea and navigable rivers are formed on the same principles.
“This subject is very satisfactorily explained by Lord Hale in his Treatise de Jure Maris, chs. 4 and 6, to the whole of which I generally refer for the proof of the rule I have stated, that the right to a new island follows the right to the soil on which it was formed. This will be found from those chapters to be the rule with regard to all maritime increments. I Avill state here a few passages from them: ‘If a subject hath had by prescription the property of a cer*102tain tract, or creek, or navigable river, or arm. of the sea, even while it is covered' with- water, by certain known metes and extends, though, it should be relicted, the subject will have the propriety in the soil relicted.’ Harg. Law Tracts, 15. ‘If a subject hath land adjoining the sea, and the violence of the sea swallows it up, but so that yet there be reasonable marks to continue the notice of it, or though the marks be defaced, yet if by situation and extent of quality and bounding upon the firm land the same can be known, though the sea leave the land again,' or it be regained by art or industry, the subject doth not lose his property; and accordingly it was held by Cooke & Poster, M., 7 Jac. C. B., though the inundation continue for forty years. If the marks remain or continue, or. extent can reasonably be certain, the case is clear.’ Id., 15. '
“The case of the town of Shinbridge, in 18 Hen., Ill, is stated in page 15: ‘The river of Severn had gained upon the town of Shinbridge so much that its channel ran over part of the Shinbridge lands, and lost part thereof unto the other side (Aure), and then threw it back to Shinbridge. It shall not belong to Aure, neither ■was it at all claimed by the king, though Severn be in that place an arm of the sea; but it was restored to Shinbridge as before. The propriety of the soil was not lost to the owners who had it before.’
“ ‘The soil under the water must needs be of the same propriety as it is when it is covered with the water. If the soil of the sea while it is covered with *103•water be the king’s, it cannot become the subject’s because the water has left it. But when the land, as it stood covered with water, did by particular usage or prescription belong to a subject, then recessus maris, so far as the subject’s particular interest went while it was covered with water, so far the recessus maris, vel hr a chid ejusdem, belongs to the same subject.’ Id., 31.”
We think it may be considered as settled that the soil under the Mississippi river, to the western boundary of the State, belongs to complainant, and that whenever it is abandoned by the water flowing over it, and no longer suitable or required for the purposes of commerce and navigation, when not done imperceptibly and in process of accretion, may be taken in possession and disposed of by the State as her authorities may see fit.
The change made by the river March 7, 1876, in its channel, did not alter the boundary line separating complainant and Arkansas, or affect the respective rights of those States, or those of the owners vof lands abutting upon the river in the abandoned channel or bed. The channels of the rivers and other streams and bodies of water may and do become changed and their physical location altered by the forces of nature operating upon their shores or banks. When the change is made insensibly, by gradual and imperceptible washing away of one shore and the formation in like manner upon the other shore, it is said to be “by erosion and accretion.” When it is made suddenly and violently, *104and is visible and the effect certain, it is called “avul-sión.” Where the boundary lines between individuals, as well as States and nations, are marked by streams, and the location of the stream is altered by erosion and accretion, it continues to be the boundary line; but, where the alteration occurs as the result of an avulsion, no change is made, but the limits of the private estates or national territory and jurisdiction remain as before. These principles are well settled at common law, and have been frequently applied by the courts of the various States and those of the United States. All the authorities in relation to this doctrine were reviewed by Mr. Justice Brewer in the great case of Nebraska v. Iowa, 143 U. S., 360, 12 Sup. Ct, 396, 36 L. Ed., 186, a case involving a dispute between those States concerning their joint boundary line, where the Missouri river, which marked the limits between them, had, in time of a great freshet, suddenly made a change in its channel similar to that in this case, across the 'neck of a bend therein, in the form of an ox bow, which was held to be an avulsion, and to leave the line between the States in the center of the old channel or bed of the river, where it had previously existed. This case is. so exhaustive of the subject and ably considered that we make no apology for quoting from it at length. It is there said:
“It is settled law that when grants of land border on running water, and the banks are changed by that gradual process known as ‘accretion/ the riparian *105owner’s boundary line still remains the stream, although, during the years, by this accretion, the actual area of his possession may vary. In New Orleans v. United States, 10 Pet. (U. S.), 662, 717, 9 L. Ed., 573, this court said: ‘ The question is well settled at common law that the person whose land is bounded by a stream of water which changes its course gradually by alluvial formations shall still hold by the same boundary, including the accumulated soil. No other rule can be applied on just principles. Every proprietor whose land is thus bounded is subject to loss by the same means which may add to his territory; and, as he is without remedy for his loss in this way, he cannot be held accountable for his gain.’ See, also, Jones v. Soulard, 24 How. (U. S.), 41, 16 L. Ed., 604; Banks v. Ogden, 2 Wall (U. S.), 57, 17 L. Ed., 818; Saulet v. Shepherd, 4 Wall. (U. S.), 502, 18 L. Ed., 442; St. Clair County v. Lovingston, 23 Wall. (U. S.), 46, 23 L. Ed., 59; Jefferis v. East Omaha Land Co., 134 U. S., 178, 10 Sup. Ct., 518, 33 L. Ed., 872.
“It is equally well settled that where a stream, which is. a boundary, from any cause suddenly abandons its old and seeks a new bed, such change of channel works no change of boundary, and that the boundary remains as it was, in the center of the old channel, although no water may be flowing therein. This sudden and rapid change of channel is termed, in the law, ‘avulsion.’ In Gould on Waters, section 159, it is said: ‘But if the change is violent and visible, and arises from a known *106cause, such as a freshet, or a cut through which a new channel is formed, the original thread of the stream continues to mark the limit of the two estates.’ 2 Bl. Com., 262; Angell on Water Courses, sec. 60; Trustees of Hopkins Academy v. Dickinson, 9 Cush. (Mass.), 544; Buttenuth v. St. Louis Bridge Co., 123 Ill., 535, 17 N. E., 439, 5 Am. St. Rep., 545; Hagan v. Campbell, 8 Port. (Ala.), 9, 33 Am. Dec., 267; Murry v. Sermon, 1 Hawks (N. C.), 56.
“These propositions, which are universally recognized as correct where the boundaries of private property touch on streams, are in like manner recognized where the boundaries between States or nations are by prescription or treaty found in running water. Accretion, no matter to which side it adds ground, leaves the boundary still the center of the channel. Avulsion has no effect on the boundary, but leaves it in the center of the old channel. In 8 Op. Attys. Gen., 175, 177, this matter received exhaustive consideration. A dispute arose between our government and Mexico, in consideration of changes in the Rio Brava. The matter having been referred to Attorney-General Cushing, he replied at length. We quote largely from that opinion. After stating the case he proceeds:
“ ‘With such conditions, whatever changes happen to either bank of the river by accretion on the one or degradation on the other — that is, by gradual, and, as it were, insensible, accession or abstraction of mere particles — the river as it runs continues to be the boundary. *107One country may, in process of time, lose a little of its territory, and the other gain a little; but the territorial relations cannot be reversed by such imperceptible mutations in the course of the river. The general aspect of things remains unchanged; and the convenience of allowing the river to retain its previous function, not-Avithstanding such insensible changes in its course, or in either of its banks, outweighs the inconveniences, even to the injured party, involved in a detriment, which, happening gradually, is inappreciable in the successive moments of its progression.
“ ‘But, on the other hand, if, deserting its original bed, the river forces for itself a new channel in another direction, then the nation through whose territory the river thus breaks its way suffers injury by the loss of territory greater than the benefit of retaining the natural river boundary, and that boundary remains in the middle of the deserted river bed; for, in truth, just as a stone pillar constitutes a boundary, not because it is a stone, but because of the place in which it stands, so a river is made the limit of nations, not because it is running water bearing a certain geographical name, but because it is water flowing in a given channel and within given banks, which are the real international boundary.
“ ‘Such is the received rule of law of nations on this point, as laid down by the writers of authority. See, e. g., Puffend, Jus Nat. lib. lv, c. 7, sec. 2; Gundling, Jus. Nat., p. 248; Wolff, Jus Gentium, secs. 106-109; *108Vattel, Droit des Gens, lib. 1, c. 22, secs. 268, 270; Stympanni, Jns Marit, c. 5, notes 476-552; Rayneval, 'Droit de la Nature, tom. 1, p. 307; M'erlin, Repertoire, ss. voc. alluv.’ ”
Further reference is made in the opinion to many authorities, among them Vattel, who states the rule thus (hook 1, c. 22, secs. 268, 269, 270) :
“If a territory terminating on a river has no other boundary than that river, it is one of those territories that have natural or indeterminate hounds (territorio, arcifinia), and it enjoys the right of alluvion; that is to say, every gradual increase of soil, every addition, which the current of the river may make to its hank on that side, is an addition to that territory, stands in the same predicament with it, and belongs to the same owner. For, if I take possession of a piece of land, declaring that I will have for its boundary the river which washes its side, or if it he given to me on that footing, I thus acquired beforehand the right of alluvion; and, consequently, I alone may appropriate to myself whatever additions the current of the river may insensibly make to my land. I say ‘insensibly,’ because in the very uncommon case called ‘avulsion,’ when the violence of, the stream separates a considerable part from one piece of land and joins it to another, but in such manner that it can still be identified, the property of the soil so removed naturally continues vested in its former owner. The civil laws have thus provided against and decided this case, *109when it happens between individuáis and individuals. They ought to unite equity with the welfare of the State, and the care of preventing litigations.”
This full and able presentation covers all the law upon the subject of accretion and avulsion, and it seems useless to further discuss it; but we will cite some other cases in -which the same doctrine is announced and applied: Moss v. Gibbs, 10 Heisk., 283; Posey v. James, 7 Lea, 98; Stockley v. Cissna, 119 Fed., 812, 56 C. C. A., 324; Missouri v. Kentucky, 78 U. S., 410, 20 L. Ed., 116; Missouri v. Nebraska, 196 U. S., 23, 25 Sup. Ct., 155, 49 L. Ed., 372; Indiana v. Kentucky, 136 U. S., 508, 10 Sup. Ct., 1051, 34 L. Ed., 329; Rees v. McDaniel, 115 Mo., 145, 21 S. W., 913; Holbrook v. Moore, 4 Neb., 437; Collins v. State, 3 Tex. App., 823, 30 Am. Rep., 142; Buttenuth v. St. L. Bridge Co., 123 Ill., 546, 17 N. E., 439, 5 Am. St. Rep., 545.
We have, in the light of these authorities, no hesitancy in holding that, the change made in its channel by the Mississippi river in 1876 at Centennial Cut-Off was an avulsion, and the limits of Tennessee and Arkansas, their respective rights in the abandoned channel, and those of individuals who owned lands lying and abutting upon it, all remained as they were before the formation of the new channel. The cut-off or formation of the new channel worked a great and important change in the course of this great river, shortening its length nearly twenty miles, driving the owners of nearly two thousand acres of valuable cultivated *110lands from their property, washing away the surface, and occupying it as a bed for its waters, and so affecting the old channel that it necessarily filled up in the usual way of beds of rivers abandoned by the stream, and became dry land. At the same time it separated from the other portions of Tennessee a large part of a civil district of one of its counties. The change was visible, and the ultimate effects certain and inevitable. It was accompanied with great and uncontrollable force and violence, and occurred in less than two days^ a remarkably short time when the importance of its effects and results are considered. The change was complete. It began Friday morning, and before the next Sunday morning a channél of the usual width of the river, about a mile wide, had been washed out, and a steamboat passed through it that day in the usual course of navigation of the river. What had before been the channel of a great river, and a highway of the nations, became a lagoon and slough. It is difficult to conceive of a stronger or more conclusive case of avulsion, both in respect to the new channel thus made and the old one abandoned.
We are now to determine where the line between Tennessee and Arkansas should be located at the place where the lands sued for lie and are bounded by it. We are of the opinion that the true and correct line is midway between the banks of the river as they existed in 1823, as shown by the map of Maj. J. H. Humphreys. *111We are led to this conclusion by the following considerations :
We have seen that the line between the States was midway between the bants of the river as they existed in 1763. There is no direct evidence where they then were, and none can now be obtained. The earliest record of the location of the banks of the river is as they were in 1823, or between that date and 1830. The territory now composing the State of Arkansas was then a territory, and the lands belonged to the United States. Those bounded by the Mississippi river at the point in question, including Dean’s Island, were surveyed and' laid off into townships and sections, and these surveys and maps then made are now of record in the General Land Office of the United States. The lands upon the Tennessee side of the river, including what is now known as “Centennial Island” and “Island 37,” which are directly opposite Dean’s Island, were granted by the State of Tennessee, under the authority vested in it by congress, to various individuals, between 1822 and 1830. These grants and the entries and surveys upon which they were made are found in the proper offices of Tennessee. These surveys, covering both sides of the river, included all the lands there lying. They called for and adjoined each other, and other grants lying back of and behind them upon the main land. The original corners, landmarks, and lines are known and can be pointed out by those residing in the neighborhood. Maj. J. H. Humphreys, a competent civil *112engineer, surveyed, tlie townships and sections upon Dean’s Island, and the grants made by Tennessee upon Centennial Island and Island 87, and has constructed a map showing how all of the several tracts lie, and their location in respect to the bants of the river upon both sides, as they were originally surveyed. The width of the river between these banks was not shown in the original grants and surveys; but, when all the lines of the townships and grants were run and located upon the premises, it was an easy, matter to measure the distance between those lines and thus ascertain it.
The correctness of the survey of. Maj. Humphreys is not seriously controverted in this record, and we do not think it could be. It was evidently made in a careful manner, and is accurate and correct. The defendant Cissna concedes in his plea that this was the situation in T823. The presumption is in favor of the permanency of boundary lines, and the burden of proof is upon the party averring that the location of a line has been changed by the action of the forces of nature. The defendant has undertaken to prove that a change took place in this case by accretion to Dean’s Island, and erosion upon the opposite Tennessee bank. The exact contentions are that by erosion upon the banks of what are now Centennial Island and Island 37, and accretion to the banks of Dean’s Island, since 1823, both before and after the cut-off in 1876, the middle of the river and the line separating the two States had advanced gradually westward towardsi the Tennessee *113bank, and that at the time of the cut-off the middle of the river was where the eastern boundary line of the Huddleston and Trigg lands had been before they were washed away and became part of the bed of the river, and, that being the boundary between the two States, complainant can recover nothing east of it, and, having previously granted that portion of the channel covering the Huddleston and Trigg lands, it cannot recover that, because those who hold under the original grants are entitled to such lands since restoration or reappearance, caused by the abandonment of the channel by the waters, and therefore the bill of complainant must be dismissed. We will dispose of these contentions in the order they are stated.
The great volume of the testimony introduced in this case by both parties was for the purpose of proving that the channel of the river at that place where the lands sued for now lie increased in Avidth since 1823 and prior to 1876, and the extent of such increase, and by the complainant to prove that no accretions had formed upon Dean’s Island after 1823, and by the defendants that the area of the island had in this way, since that date, been greatly increased and extended westward. Witnesses were examined who had lived and owned lands in the vicinity of the premises in dispute for many years before and after 1876, others who had navigated boats upon the river as captains and pilots during that period, and whose duty it was to be familiar with the *114river, its banks and channel, and still others who had never seen or known the premises until alter the abandoned bed had filled np and had overgrown with timber. The chart made of the survey of Ool. Suter in 1874, and others made by authority of the war department between 1878 and 1884, of the river, were introduced; and a number' of civil engineers, including Ool. Suter, who testified in the case, and undertook to read and interpret them. These witnesses differed considerably in their • reading of the charts and what they show the condition of the premises to have been when the surveys upon which the charts, are predicated were made. TVe will not undertake to analyze all this evidence. It would serve no good purpose, and only unreasonably extend the length of this opinion. We will only state the ultimate facts which we find to be established.
When the avulsion took place, by erosion from the Tennessee side, the width of the river south and west of Dean’s Island had greatly increased, much more immediately south of that island than west of it, where the premises sued for are situated. While there is some conflict in the evidence, we find that at this place it had increased from perhaps a little less than one mile in 1823, to between one mile and a quarter and one mile and a half, and that the most, if not all, of this was the result of erosions from the Tennessee bank. This, we think, is clearly established by the testimony of the witnesses who had resided upon the lands in the neighborhood, and especially upon Centennial Island *115and Island 37, and of captains and pilots of steamboats navigating the river for many years previous to 1876, when the change in the channel took place. The lands lying upon the river, on Centennial Island and Island 37, were originally granted to Simon Hnddleston,-Chalmers, and John Trigg. John Trigg had two grants, one for thirty-seven acres and one for 152 acres, lying immediately north of the first tract. -A considerable portion of the eastern parts of the Hud-dleston grants, all of the Trigg thirty-seven-acre tract, and nearly all of the tract of 152 acres, and a part of the Chalmers tract, had been washed away, and the channel of the river at that time flowed, to that extent, over them. The location of these lands and their relative position to the river as it flowed in 1823 will be seen by inspection of the map of Map Humphreys. The eastern bank of the river, lying on the Tennessee side at this point, was rather a high bank, and when the cut-off took place, and the current of the river was changed and no longer flowed against this bank, erosion upon it ceased, and no change' was subsequently made in it. It can now be seen, and its identity and location are conceded. The lands between this bank and the bank of Dean’s Island as it was in 1823, which is also located with reasonable certainty, is all now dry land covered with timber, and, as before stated, is now the subject of this controversy.
We do not think that there were any accretions to Dean’s Island previous to 1876. This is also clearly *116established by the evidence of witnesses who were living in the neighborhood and navigated the river immediately preceding the cut-off, and were thoroughly familiar with the situation as it then existed. They testify from their own observation and knowledge of the facts. They all state that while the towhead had appeared south and southwest of Dean’s Island, and near and below it a sandbar and mud fiats had formed, which were beneath the surface except in times of very low water, and that no land had formed along the banks of the island; that in medium stages of the river boats ran over this bar and these flats, and along the west and south banks of Dean’s Island, and through the channel between it and the towhead; and that this was done by the largest boats then navigating the river. It is also clearly proven that the width of the channel of the river had increased fully, and perhaps more than, the erosions upon the Tennessee bank, and therefore there was no room for any accretions to the Arkansas bank. These are facts clearly established in this record, and to our minds they demonstrate that in 1876 there had been no appreciable change in the banks of Dean’s Island since 1823.
Much stress is laid upon the chart made in 1874, under the- direction of Col. Suter, and his interpretation of the topographical signs and tracings appearing upon it, tending to establish that at that time there was timber growing upon what is shown on the chart to be bars and banks in the river. Complainant also *117examined civil engineers, who undertook to interpret these maps and state what they showed in relation to accretions upon Dean’s Island and the width of the river at the time they were made. This evidence is not entitled to very great weight. The chart is not- the result of a careful survey of the river and its banks, but, in the main, from an inspection of it made from the deck of a steamboat. It was a mere steamboat recon-noissance. Col. Suter describes it as follows:
“I was assigned to what was called the Transportation routes of the seaboard,’ and the part assigned to me was the Mississippi river, from Cairo to the Gulf. In the summer of 1874 — the summer and fall of that year — a certain sum of money was given to me to make an examination, and a party was put on a steamboat belonging to the government and instructed where to make a reconnaissance. The funds did not allow of an actual survey, and that was the best we could do. The idea was to get some idea of the condition of the river and the portions of it needing improvements. That party was organized in the latter part of the summer of 1874, I have not any data at hand that Avould give the exact date, but, near as I can recollect, they started in August. They went down the river from Cairo to Vicksburg, and then returned, and subsequently went over the samé ground again, extending the examination as far as New Orleans. This particular part of the river which you allude to was passed over four times, twice downstream and twice upstream. The methods *118followed, were somewhat crude, but were tbe best we could do. Tbe party being, as I said, in a steamboat, tbe course of tbe boat was taken by a compass. Tbe distance was determined by tbe speed of tbe boat, wbicb bad been accurately gauged before tbe party started. Tbe widths of tbe river were, of course, estimated; but, where it was possible to stop for any length of time to get instructions ashore, tbe triangulations, and get tbe widths, it was done. That was used as a check, and there were other points that enabled some kind of a check on tbe width. Tbe greatest difficulty was in tbe length, which, of course, tbe speed of tbe boat, varying with tbe current and all that, rendered it somewhat uncertain. Tbe best that could be done was to take points, say thirty or forty or fifty miles, where anything could be recognized as a town that was shown on tbe State maps. It enabled tbe distance in tbe longitude and latitude to be determined approximately, and tbe lengths were determined by this reconnoissance. If tbe distance varied, they were shortened up all . along tbe lines, according to tbe judgment of him who made tbe actual observations. There is one thing, of course, to be borne in mind in a case of this kind; that tbe examination was not made with any idea of determining actually by metes and bounds. That was not the idea, at all. It was to get a sketch, at any rate; something that tbe river looked like, and a general idea of its shape, direction, and location of tbe channel, and show points like that.”
*119Col. Suter does not testify from his personal recollection of the river and its banks. There was nothing about Dean’s Island to attract his special attention to it, and its hanks were a very small part of the recon-noissance made by him. He had not been there for nearly thirty years, and could only testify what he understood the topographical signs upon the chart to mean. The civil engineers examined for complainant read these signs differently, and under their interpretation the chart tended to support ‘the insistence of complainant that no accretions had formed on Dean’s Island at that time. These witnesses were never upon the premises until after this litigation began, and knew nothing of the real facts of the case. The testimony of all these witnesses is largely conjectural and speculative, and of that character that can only be relied upon in the absence of better testimony and from the necessity of the case. The chart which we have reproduced at a former page of this opinion, to our minds, so far as it shows anything, corroborates the statements of the witnesses for the complainant that, at the time of the survey, there were no accretions to Dean’s Island. It shows the towhead and the sand bar or mud flats as part of the river, and not part of the land. This reconnoissance was made when the waters were at a very low stage, almost unprecedented in the history of the river, and the showing is therefore as favorable as it possibly could be to the theory that accretions had then formed.
*120The defendants have introduced much testimony to show that cottonwood trees of an age which antedates the cut-off are found growing upon what was the old bed of the river, as evidence that there were accretions to the island previous to the cut-off. Sections cut from such trees have been produced in court, and the number of rings or circles in them, which, it is said, show the annual growth, pointed out as conclusive evidence of their age. We do not, under the facts of this case, attach much importance to this evidence. There is testimony in the record that the cottonwood in the alluvial bottoms of the Mississippi grows faster than any but one other known tree, and that trees of the size of these found on this land have been known to grow within the time elapsing since 1876. It also appears that dry land first appeared in the old channel along the mud flats and sand bars where the water was shallowest, upon the Arkansas side of the river, and that the formation there was black alluvion and very fertile, while at other places, near the Tennessee bank, appearing above the surface later, there was more sand, and, of course, less fertility. It is therefore, reasonable to suppose that the cottonwoods first began to grow upon the side next to Dean’s Island, and have attained greater size there than they' have on the Tennessee side. This, we think, clearly accounts for the difference in the size of the timber at the different places in the old channel, although there is conflict in the evidence as to whether this is in fact true. What are called annual *121rings or circles in these trees, according to the evidence,, is not reliable testimony of their age. It appears that the trees are; much affected by wet and dry seasons, and' by cold weather, and that in some years more than one ring is formed. The sections .produced in court, were also evidently cut very near the surface, where the tree is abnormally enlarged, and are not fair specimens of the growth upon the land.
There is also testimony of several witnesses tending to show that there is an elevation along the old river channel, considerably west of the original Dean’s Island bank, which they took to be and called the bank of 1876. This is mere speculation upon the part of' these witnesses. They did not reside in the neighborhood previous to 1876, and they know nothing of the condition of things as they, then existed. The witnesses examined in the case, old men who have lived in the neighborhood all their lives, and are familiar with the country and with the effects of freshets in the Mississippi river, say that there is no such bank; that what the defendants’ witnesses took for banks are mere ridges- or banks thrown up by the action of the water of the' river during freshets, when the old bed was flooded,, and the depressions near those banks mere channels that were washed out on such occasions.
While.all the matters which it is insisted this character of testimony tends to establish are circumstances to be considered in ascertaining the ultimate fact of whether there were or were not accretions to Dean’s-*122Island previous to the avulsion, yet they are of a conjectural and speculative character, and cannot be held to outweigh or even equal the positive and uncontra-dicted testimony of witnesses who testify from personal knowledge and observation of the events and facts which they had the opportunity and which it was their interest and duty to observe and know. The statements of these living witnesses of the condition and location of the river and its banks at the place in question are reasonable and consistent with the admitted facts and the history of such occurrences, and we have no doubt but that they are true. ■
The question involved is the location of a boundary line. Its location in 1823 may be said to be a conceded fact. Every presumption is in-favor of the permanency of the location of such lines. It is of the highest importance that their location should be certain and fixed. When a claim is made that a line of this character has been changed by the forces of nature, it must be supported by the clearest and most satisfactory evidence. This has not been done by the defendants in this case.
We are clear, also, that there were no accretions to the Arkansas bank after 1876. The doctrine of accretions has no application to the filling up of the old channel, abandoned by the river for a new one, as the result of an avulsion. The rights of the parties, in every respect, remain • as they existed prior to the change. The proof conclusively shows that this change *123Avas sudden, violent, and complete; that it resulted in drawing off much of the waters of the old channel, and leaving those remaining still and stagnant, a mere lagoon; that it was no longer used by boats navigating the river, except occasionally by small craft when the waters were up; that the channel immediately began to fill with sand and other alluvial deposits; and snags, bars, and banks appeared in a short time above the surface, and willows and cottonwoods began to grow upon them. This process of filling up went on over the entire channel, but it is probable that land first appeared and vegetation first began to grow next to the Arkansas side, upon the sand bars and mud flats there previous to the cut-off, and the waters shallowest. This steadily proceeded until the old bed of the river ■ became dry land, covered with valuable timber. This final result was evident from March 7, 1876, when the cut-off was made. The new channel was upon the direct general course of the river, and only about one-twentieth the length of the old one, and the fall in it was so great that it was a physical impossibility for the waters of the river to return to their old bed. In the nature of things and the history of such occurrences upon the Mississippi river, no one could doubt but that it would be only a few years at the furthermost until the old bed should be entirely filled'up and almost every vestige of its formerly being the channel of a great river gone. This effect of the avulsion was necessary, fixed, and certain, and inevitable. It is immaterial whether *124the land first appeared on one or the other side, because it was known and certaifi that the entire bed would be filled up and become dry land. The underlying reason for holding that boundary lines upon running streams may be changed by erosion and accretion, and that the States or individual proprietors separated by the stream may lose or gain territory and land, is that the loss and gain are so gradual and imperceptible that it is impossible to identify and follow the soil lost, or to prove where that gained came from. This is illustrated in the case of Nebraska v. Iowa, supra. In that case it was insisted that the doctrine of accretion had no application to the Missouri river, because of the rapid and great changes constantly going on in respect to its banks'. The court, in disposing of this insistence, while admitting the facts, said:
“Notwithstanding this, two things must be borfie in mind, familiar to all dwellers on the banks of the Missouri river, and disclosed by the testimony: That, while there may be an instantaneous and obvious dropping into the river of quite a portion of its banks, such portion is not carried down the stream as a solid and compact mass, but disintegrates and separates into particles of earth borne onward by the flowing water, and giving to the stream that color which, in the history of the country, has made it known as the ‘muddy* Missouri; and also that while the disappearance, by reason of this process, of a mass of a bank may be sudden and obvious, there is no transfer of such a solid *125body of earth to the opposite shore, or anything like a visible and instantaneous creation of a bank on that shore. The accretion, whatever may be the fact in respect to the diminution, is always gradual and by imperceptible deposit of floating particles of earth. There is, except in such cases of avulsion as may be noticed hereafter, in all matter of increase of bank, always a mere gradual and imperceptible process. There is no heaping up at an instant, and while the eye rests upon the stream, of acres or rods on the forming side of the river. No engineering skill is sufficient to say where the earth in the bank washed away and disintegrating into the river finds its rest and abiding place. The falling bank has passed into the floating mass of earth and water, and the particles of earth may rest one or fifty miles below, and upon either shore. There is, no matter hoAV rapid the process of subtraction or addition, no detachment of earth from the one side and deposit ,of the same upon the other.”
Thus in effect it was held that the loss must be suffered, because it was impossible for the losing party to follow and identify his property. It would hardly be contended, if the avulsion in this case had immediately resulted, by great deposits of alluvion and drawing off of the waters of the abandoned channel, in drying them up, that because land first appeared upon the Arkansas bank, where the waters were shallowest, it was an accretion to that bank. The principle is not changed because it took a period of several years *126to accomplish the same fixed, known, and inevitable result. The filling up of the old channel in this case was independent of the riparian rights and worked no change in them. In the case of Willey v. Lewis, 28 Wkly. Law Bul., 104, decided by the court of common pleas of Ohio, where a stream changed its channel and the old bed gradually dried up, the question was whether the'doctrine of accretion applied. The court said:
“In the case at bar, until the new channel was cut • through, the water ran in the old channel as above located. When the new channel was cut through, the river ran through that, and the old channel became an abandoned channel. The change was sudden and rapid; was avulsion, as distinguished from an imperceptible change, or accretion. A change of channel could not, in the nature of things, be instantaneous. It must require a certain time. But if it is rapid, sudden, and distinguishable from an imperceptible change, I think . under this late case it must be controlled by the law of avulsion. I see no middle course. It must be either accretion or avulsion. I do not think it an answer to say that some water still ran in the old channel until it eventually dried up. That must necessarily be the case in every change of channel; and, if it were an answer, then the proposition that the boundary remains as it- was would be a myth.”
It is not every gradual change of the channel of streams caused by filling- up with deposits cast by the *127waters that will, as an accretion, change the boundary lines of either States or riparian proprietors. This was held in the case of Missouri v. Kentucky, 11 Wall. (U. S.), 395, 20 L. Ed., 116. “ The main channel of the Mississippi riyer originally ran west of Wolf Island, and the island Ayas part of the territory of the State of Kentucky. Through a period of several years the western channel gradually filled up, and the main channel shifted to the east of the island. The State of Missouri claimed jurisdiction oyer the island on this account, and brought suit against the State of Kentucky to establish such jurisdiction. It was held, notwithstanding the change was slow and gradual, yet, since there AAras no doubt where the State line ran, it was not changed by' the change in the flow of the waters, and the suit dismissed.
The case of Indiana v. Kentucky, 136 U. S., 479, 10 Sup. Ct., 1051, 34 L. Ed., 329, was a similar case and involved jurisdiction over Green River Island in the Ohio river. When the State of Kentucky was originally admitted into the union, the main channel of the Ohio river, upon the northern bank of Avhich, at Ioav-water mark, the line separating Kentucky and the Territories north of it ran, was north of this island. Afterwards by slow process of accretion and filling up the channel changed to the south of the island, and In-diana claimed jurisdiction over it, and brought suit to enforce such jurisdiction. It Avas held that, it appearing that the island was originally within the limits of *128.Kentucky, the jurisdiction over it was not lost by the change in the channel of the river.
The case of Hughes and Others v. Heirs of Birney and Others, 107 La., 664, 32 South., 30, is also analogous to this. Previous to 1876 the Mississippi river, opposite the city of Vicksburg, Mississippi, reversed its course and ran northwards for some distance, then eastward, then southward, pursuing its general course, forming a long, narrow tongue of land, called “De Soto Point,” and owned by some eight or more different proprietors. In that year, 1876, the river made a new •channel for itself, cutting across the tongue of land near its northern extremity. This cut-off by erosion .gradually swept away all the land south of it, until it reached the southern extremity of the tongue, where it made for itself a new and permanent channel, through which the curreut ran, and all the old channel, including that made in the erosion, became a lake, called “Lake Centennial.” About eight months elapsed from the time the first cut-off was made until the permanent channel was reached and formed. ' The lake then, where there was formerly land upon the tongue, began to fill up, bars appeared above low water in about three years, and gradually became dry land fit for cultivation and habitation. The owner of the northern ■extremity of the tongue, which was not washed away, •claimed it as accretion to her land. The court held that this claim could not be maintained, but that the land, being subject to survey and identification, was *129the property of those who owned it at the time the surface was washed way; that the ownership' of the soil carries with it all that is directly above and under it; that ground upon which the river rested temporarily, in going over that portion of De Soto Point, never ceased to belong to the defendants, the heirs of Birney, and deposits placed upon it by the river in retiring from it, having been put upon land belonging to them, became likewise their property; and that the doctrine of reappearance of land after submergence controlled the case.
The formation of dry land in the old channel of the river opposite Dean’s Island, was not an accretion to either bank of that channel, but a filling up by deposit from the bottom of the old bed of the river, until it emerged from the water and became habitable and susceptible of cultivation. It was not in any way built upon the banks or aided by them. The new soil did not accrete to the banks, but built up on that of the owners of the old bed. It was not an accretion to anything, but an emergence of land, that, had been theretofore covered by the waters, caused by an avulsion, and was and is the property of those who held it in its submerged condition. The channel of the river as it flowed in 1876, when the cut-off took place, covered the channel occupied by it in 1823, and part of the grants of Simon Huddleston,-Chalmers, and John Trigg, formerly on the eastern shores of Centen*130nial Island and Island 37. When the waters of the river abandoned these lands, and they emerged and became dry land, the owners, in this case the State and the grantees, Huddleston and others, and not the then abutting riparian proprietors, were entitled to them. This is well-settled law. In the case of Mulry v. Norton, 100 N. Y., 426, 3 N. E., 585, 53 Am. Rep., 206, it is said:
“It is not, however, every disappearance of land by erosion or submergence that destroys the title of the true owner or enables another to acquire it; for the erosion must be accompanied by transportation of the land beyond the owner’s boundary to effect that result, or the submergence followed by such lapse of time as will preclude the identity of the property from being established upon its reliction. Land lost by submergence may be regained by reliction, and its disappearance by erosion may be returned by accretion, upon which the ownership temporarily lost will be regained. When portions of the main land have been gradually encroached upon by the ocean, so that navigable channels have been extended therefrom, the people by virtue of their sovereignty over public highways, undoubtedly succeed to the control of such channels — to ownership of the land under , them in cases of its permanent acquisition by the sea. It is clearly true, however, that when the waters 'disappear from the land, either by its gradual retirement therefrom or the elevation of the land by avulsion or accretion, or even the exclusion of *131the water by artificial means, its proprietorship- returns to the original riparian owners.”
To the same effect are the cases, above cited, of Morris v. Brooke (Del.), cited in Mulry v. Norton, 53 Am. Rep., 215, note; Hughes et al. v. Heirs of Birney et al., 107 La., 664, 32 South., 30; Hardin v. Jordan, 140 U. S., 382, 11 Snp. Ct., 808, 838, 35 L. Ed., 428; St. Louis v. Rutz, 138 U. S., 226-246, 11 Snp. Ct., 337, 34 L. Ed., 941; Stoekley v. Cissna, 119 Fed., 831, 56 C. C. A., 324.
This was the rule of the common law, and it applies, as is fully shown in the authorities we have cited, in favor of the State and of individuals, and as well to cases of emergence of lands which have in all known times been covered by the sea or navigable rivers, as well as those which have been submerged and reappeared again. If the soil under the waters belonged to an individual, the dry land appearing is his property; and if the submerged soil belonged to the State, when it is abandoned by the waters and becomes habitable and susceptible of cultivation, it remains her property. Clearly, the position of the defendants, that the State is not entitled to recover the portions of the channel covered by the grants to Huddleston, Trigg, and Chal-mers, is sound. These parties, or their assigns, are entitled to. them. Morris v. Brooks, supra; 2 Bl. Com., 262; Hale, de Jure Maris, chs. 4 and 6.
What, then, are the rights of the States of Tennessee and Arkansas in the premises in controversy, and what *132is the true location of the line between them? We think, unquestionably, that the bed of the abandoned river should be divided between them, for we apprehend that the Arkansas side belongs' to that State, since the title of ripaiúan owners under its laws is limited to high-water mark. Railroad v. Ramsey, 53 Ark., 314, 13 S. W., 931, 8 L. R. A., 559, 22 Am. St. Rep., 195. The line separating their respective jurisdictions is to be run along the channel midway between the banks as they existed and were surveyed in 1823, as shown in the map made by Maj. J. H. Humphreys, and exhibited with the bill of complainant. This was the line between Tennessee and the Territory of Louisiana when the former became a State and was admitted into the union. It was the line between Tennessee and that Territory after it was purchased by the United States, as is shown by the surveys and grants upon both banks of" the river made between 1822 and 1830; and the only occurrence tending to show a change in it since that date is the widening of the river between then and 1876 by erosion on the Tennessee bank to the extent of submerging the lands there, constituting a part of the grants made to Simon Hud-dleston, -Chalmers, and John Trigg. The same rule that entitles those parties to their lands when abandoned by the river also entitles Tennessee to its original one-half of the river bed. This is the natural and necessary result of the avulsion. The effect of it was to press back the line of the State, as it ran at low-water mark, to the eastern boundary line along the river bank to the *133grants it had made, so as to restore the grantees and their assigns to their property, and at the same time to press back to the center of the old channel, as it ran previ ons to the submergence of those grants, the line between the two States, so- as to restore to Tennessee what it held before the erosions upon its banks. The right of restoration to their lands was one of the vested rights of those grantees, and the right of Tennessee to be restored to her share of the original channel was one of her vested rights. These were the rights of the parties that existed at the time of the avulsion, and were fixed and settled by it, and which they had the right to have worked out and adjusted.
It restores all parties to their original status, and does justice to them all. If the result of the avulsion had only affected the waters of the river, so far as to cause them to recede from the lands of the riparian proprietors on the Tennessee bank and occupy the channel as it existed in 1823, it.would not be denied that the line would now be the center of the bed as it was in 1823. That the entire old bed was abandoned cannot” change the rights of the parties. The others interested cannot be restored to their OAvn by the forces of nature, and Tennessee entirely eliminated and denied any benefit of the reliction of the waters. She cannot in this way be deprived of the property, when the same, can' without doubt be identified and located.
■ It is said that complainant only sued for the land lying west of the center of the channel as it was in 1876, *134and therefore cannot recover to the center of the channel of 1823. This is true; hut this case must he remanded, for a hearing upon the answers of the defendants, and, if it is desired, the- hill may then be amended, so as to make the proper averments to entitle her to recover under the principles here settled.
Eeversed and remanded.