This controversy involves the title to a tract of land in Pulaski County lying north of the present channel of the Arkansas River and said to be formed by accretion. Pfeifer has the title to the fractional part of section 12, township 1 north, range 11 west, abutting on the north bank of the Arkansas River, according to the Government’s plats of the official survey in 1818, and brought suit against Nix in the chancery court, claiming title to the land in controversy as accretion to the tract in section 12, and prayed that the boundaries thereof be declared, and that Nix be restrained from trespassing thereon. Nix has title to a part of section 13 abutting on the south bank of the river, according to the plat of said original survey, and, denying that the land in controversy is an accretion to Pfeifer’s land, now claims that it falls to him under the act of April 26, 1902, being within the lines of his orignal tract. From a decree in favor of plaintiff, the defendant appealed.
An attack is made on Nix’s title to the original tract, but we may dispose of that by a statement of the established doctrine that the plaintiff must recover, in an action of this kind, upon the strength of his own title, and not upon the weakness of the title of his adversary. Dawson v. Parham, 47 Ark. 215; Buse v. Russell, 86 Mo. 215; Apel v. Kelsey, 47 Ark. 413; Cox v. Arnold, 31 S. W. 593; Victoria v. Schott, 29 S. W. 681.
The lands of Gen. Churchill adjoin the Pfeifer land on the west, and in front of the Churchill property there is a tract called Owens Island. It is shown on the Government plats as an island, and there is such a distinct tradition concerning its *202formation that it still retains its designation as an island though it is now, and has been as far back as the memory of living witnesses extends, completely connected with the main shore on the north or Churchill side of the river, and there remains only a slight trace of the old water line between mainland and island. The east end of this so-called island is about at the line between Churchill and Pfeifer, and is a part of the Churchill farm. The tract in controversy is a body of land formed between the Pfeifer land, according to the old Government survey, and the north bank of the river, as it was located in 1877, when the river by sudden avulsion left its channel bed and turned southeasterly, forming what is known as the “Maumelle Cutoff”, and leaving a large body of land between the new and old channels known as “Jones’ Island”. Prior to 1877, perhaps 40 .or 50 years, the river began wearing to the south, by gradual erosion cutting away the south bank, thence deflecting northward to about the east line of the Pfeifer place, thence running east and again south, leaving what remained of section 13 as a peninsular extending north to a point opposite Pfeifer. In 1877 the river cut through the neck of this peninsular, thus forming Jones’ Island. Pfeifer’s claim is based upon the theory that the land has been made by gradual and imperceptible accretion, forming to and out from his original tra.ct, whilst Nix claims that the land first began forming out in the stream, and gradually made toward and joined to the north shore. Both concede that it is what is commonly called “made land”, but differ mainly as to the initial point of formation, whether from shore to stream or from mid-stream to shore.
The law governing the case is clearly established and entirely free from difficulty, and we need search no further than the decisions of this court to determine the rights of riparian landowners so far as the questions involved in this suit are concerned.
Land formed by gradual and imperceptible accretion, or by gradual recession of the water, belongs to the owner of the contiguous land to which the addition is made. The river line is a natural boundary, and its gradual advance or retreat carries the owner’s line with it, except in case of an avulsion, or sudden and perceptible change of the watercourse, in which latter case the line remains at the old water line, and becomes fixed by it, not *203subject to further change by the caprice of the river. St. Louis, I. M. & S. Ry. Co. v. Ramsey, 53 Ark. 314; Wallacc v. Driver, 61 Ark. 429; St. Louis v. Rutz, 138 U. S. 226; Nebraska v. Iowa, 143 U. S. 359.
In St. Louis, I. M. & S. Ry. Co. v. Ramsey, supra, the court held that the riparian owner takes only to high water mark, and the high water mark “is to be found by examining the bed and banks, and ascertaining where the presence and action of water are so common and usual and so long continued in ordinary years as to mark upon the soil of the bed' a character distinct from that of the bank in respect to vegetation and the nature of the soil itself.”
When the formation begins with a bar or an island detached and away from the shore, and by gradual filling in by deposit, or by gradual recession of the water, the space between bar or island and mainland is joined together, it is not an accretion to the mainland in a legal sense, and does not thereby become the property of the owner of the mainland. Holman v. Hodges, 112 Ia. 714, s. c. 84 N. W. 950; Perkins v. Adams, 33 S. W. 778; Victoria v. Schott, 29 S. W. 681; People v. Warner, 74 N. W. 705; Cooley v. Golden, 23 S. W. 100; Buse v. Russell, 86 Mo. 211.
So, guided by these settled principles, we have only to determine from the evidence how the formation began, and how it became joined to Pfeifer’s original tract. There is no doubt at all that it is now so joined, and that only a depression or disconnected slough marks the old bed -of the river. Upon the disputed point there is an irreconcilable conflict in the testimony of the witnesses. Nine witnesses introduced by the plaintiff testify positively that the formation began from the Pfeifer shore line, and graduall)’- made out therefrom; whilst eleven witnesses are introduced by defendant, most of whom testify that it began away from the shore, either as a sand bar or island originating from a “tow head”, and made toward the mainland.
.A careful analysis of the testimony of all these witnesses satisfies us that those introduced by the plaintiff in the main show better opportunities for accurate knowledge and more intelligent and definite recollection of the facts which they undertake to relate, and for that reason their testimony has greater weight. For instance, one who has owned and cultivated the *204adjoining farm since 1852, and his son 48 years of age who has known the place all his lifetime; another who has been on the Arkansas River for the past 55 years as pilot, engineer 01-master of a steamboat, and who says that he observed all the formations on the river and changes in the channel; another who has been a civil engineer in this locality since 1853, and has observed the lands all those years; another who, 43 years of age, was born and reared within two miles of the Pfeifer land; and other witnesses apparently of equal intelligence and opportunities for knowing the facts — all testify positively that the land in formation began at the shore line and made outward, and that there was no island or detached formation at all, and no channel next to the old shore line. Some point is made upon the fact that there is now and has1 always been a depression or slough along the old river line, and that the high bank of the old rivet-line is still visible, evidencing the fact that the' formation was originally an island ■ with a channel between it and the main shore which later filled in; but several of the witnesses explain that this is characteristic of such formations, that usually in case of alluvial deposits the height of the formation is greatest next to the water line and away from the shore, and always remains so. We think that the testimony clearly establishes the fact that the land in controversy was formed by alluvial deposit as an accretion to Pfeifer’s land, and that the finding of the chancellor is supported by the weight of the evidence.
Counsel for appellant urge further that, even though it be found that this formation was by deposit and accretion to the Pfeifer land, it was not by gradual or imperceptible process, but was by sudden and perceptible action, as much as 124.38 acres of Nix’s land having been washed away within less than three years. Mr. Justice Battue, in the opinion in Wallace v. Driver, supra, quoting in part from Rex v. Lord Yarbrough, 3 B. & C. 91, says: “In order to constitute accretion, it is not necessary that the formation be indiscernible by comparison at two distinct points of time. It is true that it is an addition to riparian land gradually and imperceptibly made by the water to which the land is contiguous; but the true test as to what is gradual and imperceptible in the sense of the rule is that, though the witnesses may see from time to time that progress has been made, they could not perceive it while the process was going on.”
*205Testing it by the rule thus laid down by this court as above, the land is accretion in a legal sense, though.it may have been rapidly formed, as no witness in the case claims to have “perceived it while the process was going on.”
The decree of the chancellor is affirmed.
Wood, J., not participating.