This is the third appearance in this court of a chain of litigation that has been in progress between the appellant Bass and the appellee Willey and others for some eighteen years. The dispute is over the ownership of a tract of land, originally formed by accretion, that now lies at some distance north of the Arkansas River, in Arkansas County. This appeal is from a decree finding that the appellant Bass has no claim to the land in controversy (and thus, by implication, recognizing the ownership of the appellees, who derive their title from C. F. Willey, now deceased).
*783The earlier phases of this litigation bear upon the present appeal and must be briefly reviewed. In 1946 C. F. Willey, in a suit to enjoin Bass from cutting timber, obtained a default decree finding that Willey was the owner of Section 1, Township 8 South, Range 4 West, and “all accretions adjoining or contiguous thereto.”
In 1947 Bass was cited for contempt, for an asserted violation of the court’s injunction. Bass defended the contempt citation by attempting to prove that the 1946 decree was a nullity. It was Bass’s theory then, and it is his theory today, that long ago—perhaps as far back as the 1880’s—the tract that had been originally surveyed by the Government as Section 1 was completely eroded away by a gradual northward movement of the Arkansas River, which eventually crossed all of Section 1 and ate away not only that section but also most of Section 36, Township 7 South, Range 4 West, lying just north of Section 1 and being later owned by Bass.
Bass has attempted throughout the litigation to prove that the river, after having reached its line of maximum progress to the north, then gradually retreated southward and re-created land that re-emerged not as Section 1, which had become nonexistent, but as an accretion to Section 36 and to other lands that Bass now owns along the line of the river’s farthest advance to the north. Upon this hypothesis Bass contended that the 1946 default decree, in referring to Section 1 and its accretions, actually described no land at all, so that the decree was void.
The chancellor rejected Bass’s attack upon the 1946 decree. Upon the first appeal we affirmed the chancellor’s action but confined our decision to a single point, holding that Bass was estopped to deny the existence of Section 1 for the reason that Bass himself had recognized the existence of the section by having purportedly conveyed it to Willey in 1930. Bass v. Willey, 216 Ark. 553, 226 S. W. 2d 980.
Soon after our decision upon the first appeal Willey again instituted contempt proceedings against Bass. By filing a counterclaim Bass became the real plaintiff in *784the case; that is still his position. He contends that even though he is estopped to question the existence of Section 1 the estoppel does not extend to the accretions thereto, because those accretions were not mentioned in the 1930 deed that gave rise to the estoppel. Hence Bass insists that he be permitted to prove that the tract now in dispute accreted not to Section 1 but to the more northerly land owned by Bass.
This contention was at first rejected by the chancellor, upon a plea of res judicata. On the second appeal we reversed that decree, holding in essence that whether this tract accreted to Section 1 or to Bass’s lands was an issue of fact upon which Bass had never had his day in court. Bass v. Willey, 227 Ark. 1025, 304 S. W. 2d 943.
On remand the parties built up an extensive record, with many exhibits, in their efforts to trace the wanderings of the Arkansas River since this area was surveyed by the Government in 1819. It is still Bass’s contention that Section 1 was eaten away long ago by the northward movement of the river, that when the river retreated southward even past its 1819 channel the land in dispute was formed as an accretion to lands now owned by Bass, and that the estoppel stemming from Bass’s 1930 conveyance of Section 1 does not preclude him from asserting that the tract in controversy was never an accretion to Section 1. The chancellor, by the decree now on review, dismissed Bass’s claim to the land.
The facts are not simple. That they may be more easily understood we are inserting in this opinion, as Figure 1, a greatly simplified reproduction of one of the principal exhibits in the record.
Both Section 1 and Section 36, as surveyed by the United States in 1819, are shown in heavy lines. The irregular shape of these small fractional sections was due to the fact that in 1819 they were bounded on the west by the Arkansas River and on all other sides by Spanish land grants that had been made before the Louisiana Purchase. The tract in dispute, marked by the corners
*786ABC®, is also shown in heavy lines. It should he added that although the common boundary between Section 36 and Section 1 appears to extend all the way to the western border of the land in dispute, the testimony indicates that the northwest corner of Section 1 is actually about 180 feet east of the western edge of the land in dispute, so that the land in dispute is contiguous to Section 36 for a distance of about 180 feet.
Deferring for the moment the appellant’s arguments with respect to the 180 feet of contiguity just mentioned, we are of the opinion that Bass has not succeeded in escaping the estoppel that came into being in 1930 when he undertook to convey Section 1 to Willey. Even though that deed made no express reference to accretions, we held in Towell v. Etter, 69 Ark. 34, 59 S. W. 1096, 63 S. W. 53, in the opinion on rehearing, that a conveyance of a tract of land by its legal description carries the accretions thereto unless they are specifically excepted. Hence it is not possible to limit Bass’s estoppel to the original boundaries of Section 1, for the reference in his deed to Section 1 must be taken to be a reference to the accretions as well.
Moreover, the 1946 default decree confirmed Willey’s title to Section 1 and “all accretions adjoining or contiguous thereto.” As a result of Bass’s estoppel Section 1 must be deemed to have existed, as far as the litigants and the court were concerned. We must attach some meaning to the explicit decretal reference to accretions. Unless that reference was meant to encompass the greater part of the tract in dispute (all except the west 180 feet), we are unable to see that this language in the decree had any significance at all. ■ •-
As we have said, the testimony indicates that the northwest corner of Section 1 is really about 180 feet east of point A on the western border of the land in dispute. This leaves a corridor through which the tract in controversy may be connected with Bass’s land in Section 36. The appellant forcefully argues that, under the rule governing the apportionment of accretions, he should be awarded a strip 180 feet wide along the west side of the *787tract in dispute, upon the theory that this strip was an accretion to Bass’s Spanish Grant No. 2334.
The basic rule for apportioning accretions is well understood. It is first necessary to determine what proportionate part of the old bank was owned by each riparian proprietor. The new bank is then divided in the same way by assigning to each proprietor his proportionate share. The division is completed by drawing-lines to connect the points so fixed upon the new bank with the corresponding points of ownership upon the old bank. Hamilton v. Horan, 193 Ark. 85, 97 S. W. 2d 637.
Bass relies upon the apportionment shown in Figure 1, which is based upon one of Bass’s exhibits. This particular apportionment was made in 1918 by a surveyor named Keaton. What Keaton did was to treat the Line of Maximum Recession (the line of the river’s greatest northward progress) as the old bank. The 1918 bank of the river, in between its two points of intersection with the Line of Maximum Recession, was treated as the new bank. Keaton then followed the basic rule of apportionment in dividing the new bank and in drawing his lines. It will be seen from the lines of apportionment on Figure 1 that most of the land in dispute was considered to be an accretion to Spanish Grant No. 2334 and to the small triangle of Section 36 that lay north of the Line of Maximum Recession.
There are two fatal defects in the appellant’s argument. In the first place, there is no sound reason why Keaton’s 1918 apportionment should be accepted as controlling today. The record does not show why Keaton undertook the apportionment. Perhaps his purpose was to fix the boundary between the land now in dispute and the land lying immediately to the west (referred to as the Anderson-Tully Company property)- The fact that Keaton’s line is still the boundary between these two tracts suggests that his purpose was to fix that boundary.
There is no showing, however, that Keaton’s apportionment was accepted by or even known to the own*788ers of the land now in question. If the apportionment had become binding upon those landowners, either by agreement or by court action, then of course title would have vested and the division would have become a new starting point for the apportionment of future accretions. There is no such proof. The river seems to have been changing its course continually ever since 1819. An apportionment made at any particular time would not divide the accretions in the same way as another apportionment made several decades later. There is no reason today to go back to the particular point in history when Keaton made his survey and to declare that the conclusions he reached somehow became binding for all time upon the owners of the tract now in dispute.
The second flaw in Bass’s argument is equally serious. The Keaton apportionment, in order to be an equitable division of the accretions, must rest upon the assumption that this whole segment of the river constituted a fairly straight line as it gradually moved northward, until it finally flowed in a direct course along the Line of Maximum Recession, etching out that escarpment before beginning its retreat to the south.
We are convinced by the record that this is not what actually happened. Instead, the Line of Maximum Recession was eroded over a period of many years by the northernmost tip of a loop in the river that traveled from west to east. This loop was part of the main channel until it was cut off by an avulsion in 1926 and became an oxbow lake, shown upon Figure 1 as Moody Old River.
On this point the most convincing proof in the record is a number of aerial photographs, the earliest one having been taken in 1930. In these photographs there are plainly discernible lines of erosion, on the upstream side only, that exactly parallel the curving sides of the loop that is now Moody Old River. It is hardly possible for one to study these pictures without becoming convinced that this loop in the stream, which may have had its origin in the bend that appears in Figure 1 as the 1819 channel, did in fact travel downstream, from west to east.
*789This conclusion is entirely reasonable. It is a matter of common knowledge that a bank is eroded by the pressure of the current against the shoreline. This pressure is naturally apt to be greater on the downstream sides of the loop, causing the oxbow to travel in that direction. Apparently this phenomenon is perfectly well understood, for textbooks upon the subject merely observe as a matter of accepted fact that individual meander loops tend to shift downstream. See Schultz & Cleaves, Geology in Engineering (1955), p. 151; P. R. Van Frank, Random Notes on Improvement of Rivers (1933), p. 124.
The general rule of apportionment, involving the drawing of lines from the old bank to the new, is not to be followed inflexibly, even to the point of injustice. It will not be applied, for example, if there are such irregularities in the shoreline as to make the resulting division inequitable. Malone v. Mobbs, 102 Ark. 542, 146 S. W. 143, Ann. Cas. 1914A, 479.
Keaton’s apportionment is untenable, because it involves the mistaken assumption that these accretions were formed as the river moved from the north to the south. But if, as we think to have been the case, the loop really moved from west to east, then the accretions formed in that direction, and for the purpose of an apportionment the old bank would be the trailing edge of the loop at some selected point in its progress. In that event the lines of apportionment would necessarily run in an easterly or southeasterly direction, so that the land in dispute could not be considered as an accretion to Bass’s lands to the northeast.
As the real plaintiff, Bass had the burden of recovering upon the strength of his own title. The chancellor was right in concluding that that burden was not met.
Affirmed.
Harris, C. J., not participating.