Appellant seeks review of the chancellor’s location of the boundary between property of the parties. It was stipulated that appellant’s land was north and appellee’s south of the line dividing' townships 5 and 6. No question of adverse possession is involved. Appellant charges that appellees, through erroneous claim as to the location of the township. line, encroach upon her property and are illegally in possession of a strip running east and west along the south boundary in the form of a trapezoid 85 feet at the east end and 22 feet on the west.
Much testimony was introduced by each side in an attempt to determine the township line separating the respective properties. Seven surveys were made and *205numerous witnesses were called to describe physical indications of line and section corner locations. The result was a wide range of conflict.
The chancellor eventually directed Floyd Ragsdale, a surveyor first called as a witness by appellee, to ignore all other considerations, such as locating section lines and corners, and to trace township line east and west. In doing this the survey was conducted without reference to contiguous lines and corners. The line thus established adhered closely to physical evidence, mainly the course of an ancient fence row with traces of rusted wire overgrown by trees, and terminated at the north side of a lock on an abandoned levee. Resolving the conflict on the basis of this evidence, the chancellor adopted this line in his decree.
The line so established was not proved to coincide with physical features described in field notes of the original government survey, the explanation being that identifying markers were not found. Nor was any attempt made to support the location by ascertaining the corners of adjoining sections. The trial court found that the township line was straight, and this finding, together with objection to the lack of conformity with the original survey, forms the chief basis for appeal.
We are faced with a problem similar to the one confronting the trial court. The surveyors failed to agree as to the line and there was a notable lack of harmony respecting the procedure whereby accuracy might be achieved.
Our task of sifting the evidence and in endeavoring to reconcile opposing claims is rendered extremely difficult because in practically every instance where explanation of written diagrams or photographs was attempted there was insufficient identification of the details referred to, hence no clear understanding of the specific point in discussion is possible. Many witnesses, in testifying about maps, photographs or diagrams, used such expressions as “here” and “there” without making a corresponding notation on the writing. While the chan*206cellor was present and conld observe the particular portion indicated, we have no such advantage. Our review is thus restricted to a consideration of that part of the record and exhibits which can be identified as bearing directly on the testimony. In such cases, the finding of fact by the trial court must not be disturbed. East Texas Motor Freight Lines v. Dennis, 214 Ark. 87, 215 S. W. 2d 145.
At the area in issue the line between townships 5 and 6 divides sections 34, 35 and 36 on the north, in township 6, and sections 1, 2 and 3 on the south, in township 5. Controversy centers on the position of the northeast corner of section 2, which is “lost.” Its location is actually on the township line, between the northwest corner of section 2 and the Arkansas River. Appellant contends that establishing this corner is an indispensable element in locating the township line, and that only by following United States Land Office rules of procedure for restoration can such corners be located. These rules require measurement from known corners on the same line and known corners on a perpendicular line to a point of intersection. Appellee agrees that this method is proper where the known factors are present, but insists that here one of the factors is missing — that only one known section corner, the northwest corner of section 2, can be found.
Floyd Ragsdale and Elmer Smith each conducted an independent survey and then a joint one. The line finally adopted by the court was from this joint survey.
To project the township line, Ragsdale and Smith went west of a lake to “some land lines” and located a point of intersection by a north-south fence at the northeast corner of section 4. Here the east-west township line coincided with old fence-rows. The surveyors used this line as a starting point, checking their instruments by its course, and proceeded eastward. This led through a dense lake bottom and eventually to an iron pin at the northwest corner of section 2, and “hit reasonably close to it,” (the iron pin). They continued east, following very close to an old fence line, to a point on a levee, two *207feet south of a fence post, then along the same course to an old lock on a former levee. The terminus was a bolt on the north headwall of the lock. No markers were placed and no attempt was made to “tie in” the line with established locations or measurements to the north or south. This procedure is criticized by appellant as being inadequate to establish a line.
It is argued that no single line can be accurately established by relying primarily on compass readings. Compasses, it is pointed out, vary materially, not only as to particular instruments, but as to localities. Difference can be caused by ore fields, metal objects, or other factors. Furthermore, appellant insists, the survey finally accepted not only failed to coincide with the original government survey, made in 1825 and 1826, but did not agree with the line first found by Ragsdale himself in a prior survey where an 'attempt was made to check the location of the township line against other locations to the north and south.
Witnesses testified that lines did not conform to monuments mentioned in the original field notes — that none of the trees referred to could be found. While the county surveyor, who was called as a witness by appellant, reported that he began his work at a government corner, (the southeast corner of section 2} he admitted that he depended on information given by Norborn Jackson that a stake by a tree was in fact the marker for such corner.
The priority of original government surveys is not in issue, because no witness established, with sufficient certainty, that a particular survey was in conformity with the original government survey. We do not hold that a survey which establishes a line at variance with the original government survey can be accepted. We conclude only that such variation has not been shown here, because monuments on the government survey have, apparently been obliterated.
Ragsdale admitted that the line he established would cause the southern portions of sections 34-36 to have less
*208than regular acreage, whereas under normal circumstances shortages or irregularities would have pertained to the northern tier of sections 1-3, it having been the practice at the time of the original government survey to compensate for errors by adjustments in the northern tier of sections. This is not sufficient reason, however, to reject the Ragsdale line.
Failure to follow procedure prescribed by the U. S. Land Office for restoration of lost corners does not necessarily render a survey objectionable. The rules cited require presence of two known corners from which lines are run to create an intersection restoring the lost corner. Here, only one corner was known, the northwest corner of section 2. There appears to have been some dissension as to this fact, but we think it was resolved by the statement appearing in the transcript in the form of a question directed to appellant’s counsel:
“Q. Mr. Parker, do you accept the iron stake at the northwest corner of [Sec.] 2 as being a correct boundary line ? ”
“A. The northwest corner of Section 2 has been accepted by all parties as a correct starting point. ’ ’
It is noted that the final Ragsdale line missed this point by two feet. This variation was explained as having been caused by the fact that when proceeding east along the township line a dense growth was encountered in a lake bed before arriving at the northwest corner of section 2. It is recognized by all parties that absolute accuracy cannot be expected and the chancellor’s refusal to reject the line because of a two-foot differential at this point cannot be regarded as inconsistent with the preponderating evidence rule.
Appellant charges that the chancellor’s decretal finding that a township line is straight is not in accord with facts. To support this conclusion she refers to the testimony of Col. A. S. Turner, who teaches engineering at Arkansas Polytechnic College, and who stated that the east-west lines followed the curvature of the earth. As *209we understand this testimony, however, the line would be straight east and west, the curvature being subject to illustration as the inside of a series of parallel concentric circles.
Appellant urges that the Ragsdale line, if it coincides at certain locations with the old fence row, would have ended farther south on the eastern terminus.
Placing the line where appellant insists it should be would require acceptance of his location of the southeast corner of section 2, because this is the chief basis of surveys locating the line farther south. Since the findings of all surveyors varied considerably it cannot be said that the chancellor was compelled to accept any particular one. Directing the township line to be run was a method of securing further evidence as to the true location. Ample opportunity for examination of Ragsdale was given and utilized. The line found coincided with physical evidence and crossed the one recognized section corner location. We cannot say that the chancellor’s finding is contrary to, a preponderance of evidence.
Affirmed.