The opinion of the court was delivered by

This is an appeal from a judgment of the district court approving a survey made by the county surveyor of Barber county upon a request to establish the proper line between the east and west halves of the northeast quarter of section 28, township 83, range 13 west. The east half is owned by O. F. Rowley who procured the survey to be made, and the west half was owned by Ellen E. McKee who appealed from the survey to the district court and was the plaintiff in this action.

The principal dispute between the parties concerns the original location of the northwest corner of section 28. This corner is called the Ball corner and for several years the only stone there . was one known as the Ball stone. It is claimed by plaintiffs that this stone marks the original government corner. McCleary, the county surveyor, did not so treat it, and the survey he, made resulted in establishing the corner in question at a point 1.55 chains west and 28 links south of the Ball stone.

A mile and a half north of the disputed corner is the quarter-section corner between sections 16 and 17, known as the walnut corner, and the parties agree that this is the true location of a government corner. A mile south of the disputed corner is a stone at the southwest corner of, section 28, which corner had been found and reestablished as the original government corner by a former county surveyor named Aber. To explain in a general way the method of McCIeary’s survey, it may be said that he ran a line between these two points, and he found the position of the temporary corner, in a north-and-south direction, upon this line by means of proportional measurements, To fix the position of this comer in an east-and-west direction, he surveyed a line from a stone at the northwest corner of section 26 to a stone on the northwest comer of section 30 on the range line, both of which points were familiar to- him and had *259been used by him in other surveys as having been established by Aber as government corners. On measurement of this four-mile line his survey showed a surplus of 30 links over the measurement mentioned in the government field notes, or 714 links to the section. On this basis of apportionment he then established the location of the northwest corner of section 28, west 15 links from the point of intersection of this line with the north- and-south line previously surveyed. The corners which he fixed were tested by running lines and making measurements from other known corners in the vicinity.

The plaintiffs insist that the Ball rock is a monument of the original survey, and that this fact is attested by government marks upon it, by its being in line with other comers, by the fact that its location is about the same distance from a certain creek as that mentioned in the government field notes, and by general recognition for many years. If the Ball rock is a government monument, it should have been taken as a guide in establishing the line between the lands of the contending parties ; but, while some of the evidence tended to support the claim of the plaintiffs that the Ball rock is a government stone, there was other testimony to the contrary. Upon the stone there were some government marks, as plaintiffs contend. According to the field notes, it was about the same distance from the creek that the government stone was said to be from that creek, and it had received recognition as a corner stone, and had in one instance been used as a starting point in surveying another tract of land. While something corresponding to government marks were found on this stone, the marks themselves tended to discredit the claim that it was a government monument. The marks did not fit the location, and indicated that if it was a government stone at all it belonged at a point miles away from its present location, and nothing was offered to show whether it had been removed or how it came to be misplaced. The field notes show that a sandstone was placed at this comer, while the Ball rock is a limestone. The stone set by the government was noted to be 12 x 12 x 5, while the Ball rock is 15 x 81/2 x 3, and, while the discrepancies in quality and size are not controlling, they tend strongly to support the finding of the trial court that it was not a government corner. (The State, ex rel., v. Manny, 99 Kan. 140, 160 Pac. 1014.) The correspondence of *260the distance of the Ball rock from the creek with that given in the government field notes between the government corner and the creek supports plaintiffs’ claim, but as the creek runs through a sandy bottom its course is uncertain and shifting, and a witness stated that such a stream could not be depended upon as a witness to a corner. Whether the creek had shifted since the original survey, and its value as a witness to the original comer, were questions of fact for the trial court. The fact relied on by plaintiffs, that no other stone was found at this corner, is also one for the court to consider in determining •the location of the original monument; but, since many of the government stones, are missing, it is not a controlling fact. As, is often the case in the settlement of disputed surveys of boundaries, the evidence was very conflicting. The plaintiff, sometime, before the trial, procured a surveyor named Gilmore to run the lines, and his result differed from that of the county surveyor and fixed the disputed corner nearer the Ball stone. It was made upon different variations from those used by the county surveyor, but there is much uncertainty as to the variations of the magnetic needle and no fixed law has yet been found governing variations. What allowance should have been made for variation was a question of fact to be determined by the trial court from the evidence. (Wilson et al. v. Inloes et al., 6 Gill [Md.], 121.)

While there is much complaint of the methods of the county surveyor in. making this survey, and considerable testimony tending to support the plaintiffs’ theory, there is sufficient evidence, we think, to uphold the finding of the trial court. The county surveyor took considerable pains to test and verify the locations made by him, from government and known comers, and appears to have followed the statutory rules of apportionment, and we' are unable to say that the survey and the judgment approving it should be set aside. We are not aided much by the authorities cited by counsel, since the case depends upon the preponderance of the evidence, and in instances of this kind each survey case is a rule for itself. As was saiel in The State, ex rel., v. Manny, supra, “the evidence presented a fair question of fact for the trial court, and its decision must be regarded as final.” (p. 142.)

The plaintiffs complain of the overruling of the motion for *261a new trial. Upon it additional testimony was presented as to the locations in question, and the affidavit of a surveyor was secured which tended to support the testimony of the surveyor, Gilmore, who testified in plaintiffs’ behalf. There was also testimony of a stone found near the dividing line between sections 27 and 22 upon which were no marks, and also testimony of another stone farther away. This testimony is largely cumulative in character, and, if introduced, would not, in our opinion, have changed the result. New evidence, to warrant the granting of a new trial, must, among other things, be such as would reasonably have compelled a different dicision. (Shores v. Surety Co., 84 Kan. 592, 114 Pac. 1062; Collins v. Bellford, 89 Kan. 92, 130 Pac. 662.) Some testimony was also offered relating to the rule governing the variation of the magnetic needle, but this was expert in character and . did not warrant the granting of a new trial. (Lillard v. Railway Co., 79 Kan. 25, 98 Pac. 213.)

The judgment is affirmed.