FRANK HOLT, Justice.

This is a boundary line dispute which involves three contiguous tracts of land. The appellants constructed a fence along the south line of their property which adjoins appellee Whiting's land. The appellants constructed another fence along the west line of their property which is adjacent to the lands of the appellee Hudson estate. The appellees filed separate suits alleging that the appellants' fences were encroaching upon their property. The appellants counterclaimed in each case and asserted that the fences were constructed upon the true boundary lines between the three parties. The cases were consolidated for trial. The chancellor found for the appellees "to the extent of adopting the lines as shown by the Quertermous Survey." This survey was relied upon by both the appellants and appellees. However, the court agreed with the appellees' theory or version of this survey and found that appellants' fence encroached 140 feet on the lands of the appellee Hudson estate and, further, that the appellants' fence encroached upon appellee Whiting's lands a distance that varied in width from 81 feet at one corner to 107 feet at the other corner of their adjoining lands. The total encroachment upon both tracts involves about five acres of unenclosed timber and swamp land.

On appeal the appellants contend for reversal that the decree is not supported by the evidence and is, in fact, contrary to the evidence. We do not agree.

It was stipulated, and the court found that: Except for 20 acres, appellee Whiting is the owner of the lands described as all of Spanish Grant No. 2346 (82.08 acres, more or less); appellee Hudson estate is the owner of the lands described as Spanish Grant No. 2353 (no acreage given); and that the appellants are the owners of their property which is described as fractional quarters (approximately 78 acres).

Appellee Whiting acquired his land from the State in 1946 and the appellee Hudson estate lands were acquired in 1901. The appellants acquired ownership of their lands in 1962. In 1967 appellants retained a surveyor, Mr. Keefe, to survey their property in order to accurately locate and fix their true boundary lines. Following his survey appellants, several months later, constructed the fences that caused this controversy. Mr. Keefe testified that he fully and completely surveyed the appellants' property and established all four corners of their sectionized tract and pin-pointed the location of their boundary lines. It was his testimony, corroborated by another surveyor who reviewed his work, that his survey "closed" within acceptable limits. The starting point of Mr. Keefe's survey was the so-called "Crafton-Conners" corner, a section corner, and was "tied" to this point and also "tied" to other well established corners. The plat of his survey was made an exhibit to his testimony. According to Keefe, his survey conforms to the official U.S. Government Land Office Survey, as well as to a 1934 survey made by Mr. Quertermous, then the County Surveyor of Arkansas County. It appears that Keefe did not make a survey of either of the Spanish Grants. His survey was based upon appellants' sectionized lands. The only field notes he used consisted of the Government Land Office plat. According to him, this plat reflects the same information as field notes with the exception of bearing or witness trees. He testified that he didn't expect to find any 150-year-old bearing trees and that he found no monuments upon the lands in question.

It appears that the crucial starting point of the conflicting surveys made in this boundary line dispute revolves around where the southwest corner of appellants' lands adjoin both of appellees' lands. It is the appellees' theory and version that an accurate survey depends upon the correct starting point or location of the southeast corner of Spanish Grant No. 2353 and that this corner is common to the southwest corner of appellants' sectionized tract. Appellees' surveyor, Mr. Strode, the local county surveyor since 1945, testified that he was familiar with the property having made surveys in this vicinity in 1945, 1954, 1960, 1966, 1967, and the present survey in 1968. He testified that he was personally familiar with and found a witness tree and an iron pipe as representing the established corner from which he made his measurements. The 1934 Quertermous survey was introduced as an exhibit and relied upon by both the appellants and the appellees. According to Mr. Strode, the southeast corner of Spanish Grant No. 2353 is shown by the Quertermous survey as being a monument corner. This corner was marked by an iron pin as indicated on this survey. There were witnesses who testified to the effect that this corner was once marked by witness trees and that at the time of the trial the trees had been cut. Appellant Claude Rice testified that he cut certain witness trees upon the advice of his surveyor, Keefe, that they were not representative of the true boundary line. Strode testified that the southeast corner of Spanish Grant No. 2353 and the southwest corner of appellants' tract are one and the same corner. Using this corner as a starting point, Strode, as he interpreted the Quertermous survey, went from monument to monument using official field notes in establishing appellants' encroachment upon both appellees' lands to the extent as found by the court. Strode did not survey the appellants' property and Keefe, appellants' surveyor, appears to have given no information on his plat or to the court about the correctness of the appellees' Spanish Grants, Nos. 2346 and 2353, which adjoin appellants' lands.

Numerous lay witnesses testified that the blazed line running north (between appellee Hudson estate and appellants' lands) from the common corner established by Strode had been painted and was clearly marked and monumented and regarded by the appellees, timber cutters, and others since 1929 as being in conformity with the Strode survey. According to appellee Whiting and other lay witnesses, Whiting's north line of Spanish Grant No. 2346 (appellants' south line) had also been a well marked line for any years with actual monuments along this line and it also conforms to the Strode survey. It was stipulated that the testimony of certain other witnesses would be cumulative with respect to the boundary lines as contended by the appellees. There was evidence that the appellant, Claude Rice, had stopped clearing his lands, digging ditches, and cutting timber at the blazed lines between his lands and both of appellees' lands until he had his survey made in 1967 by Keefe.

The Court found, inter alia, in a memorandum opinion that:

"The defense contended that there was greater merit in the survey of a tract than there was of a survey of a line. The defendant made no attack upon the correctness of the Spanish Grant lines other than the conflict which resulted from the establishment of new lines of his own tract. The Spanish Grant lines were not surveyed by the defendants' surveyor.

* * *

Monuments of the prior official surveys, either government or county surveys, were marked by witness trees and the location and direction of the witness trees from the exact corner was recorded in field notes by the surveyor and placed in the records of the county. The defendants' surveyor did not resort to nor use the field notes but simply worked from a certain corner and surveyed a tract according to the deeds of acquisition of the defendant.

* * *

The court finds that Honorable T. J. Strode, County Surveyor for Arkansas County, used the field notes and located the witness trees as well as blazed lines and monuments and tied his survey into those witness trees and monuments and the Quertermous Survey as shown by the plat introduced in evidence. [From the formal decree]

* * * The failure of Mr. Keefe to survey all the tracts involved tends to limit the effectiveness of his survey. Mr. Mehlburger says that because the survey closes it is much more apt to be accurate. That is true provided they start from the right corner and use the correct angles and go the right distances. But the mere closure of a survey without all of the other elements being completely accurate would not necessarily be proof that the survey was correct.

* * *

In the instant case T. A. Strode, County Surveyor, admitted to what he says is a small error. Likewise, Surveyor Keefe admitted to a mistake * * *

* * *

In the instant case Rice did not present the surveyor from Monticello who had preceded Keefe.

Spanish Grants were established in the area in which these lands are situated before it was transferred to France and ultimately to the United States through the Louisiana Purchase. The establishment of sectionized method of measuring and identifying land was accomplished in the United States a considerable period of time after the acquisition of the Louisiana Purchase. The fact that our government did not change Spanish Grants but allowed them to remain as they had been laid out originally, is significant.

* * *

This court will therefore find for the plaintiffs to the extent of adopting the lines as shown by the Quertermous Survey. This will have the effect of depriving Plaintiff Whiting of about twenty feet along the north side of his Grant No. 2346 if one of his witnesses was correct in saying that Whiting had gone over the line. The ancient land markings are to be preserved no matter which party is affected. The southwest corner of the Rice land as marked by the witness tree should be reestablished and a permanent monument put in place, and the Quertermous Survey shall control."

We cannot say that the chancellor's findings are contrary to the preponderance of the evidence inasmuch as it appears that the location of the boundary lines was at most a fact question based upon two conflicting theories. Burkhart v. Watson, ( 247 Ark. 1030) 448 S.W.2d 954. Certainly the chancellor had the right to consider the theory advanced by the appellees because in the location of land boundaries, where there are inconsistent theories, resort may be had first to natural objects or landmarks, next to artificial monument and then to adjacent boundaries and thereafter to courses and distances. 12 Am.Jur.2d Boundaries 65 and 67; also, 55. See Clark on Surveying and Boundaries, 3d ed., 293, where it is said:

"* * * An accurately established line-tree is a permanent monument of the first order and conclusively shows the true location of the line. Natural objects called for in a description may be located by parol testimony. After they have been so identified they become evidence of the exact location of the corner or line. That is what the surveyor desires. * * * Courses and distances yield to fixed monuments or natural and located objects. Courses and distances yield to a call for a natural object * * *. It is for that reason that the surveyor is required to search for natural objects, such as a bearing tree, a monument well planted in the soil for future reference."

See, also, 3 A Ark. D-522, Boundaries, Key No. 3(3).

The appellants next assert that the trial court's decree did not decide the case or settle the dispute between the parties. The decree orders that the boundary lines between the three parties "shall be fixed in accordance with the blazed line and monuments set forth in the Quertermous Survey, which survey and field notes were followed by the present County Surveyor, Honorable T. J. Strode, in making his survey, that the Southwest Corner of the Rice property in Frl. NW1/4 of Section 35, Township 7 South, Range 3 West, Arkansas County, Arkansas, should be reset and replaced by the use of an iron axle or iron or concrete marker for a permanent marker and said Southwest Corner of said property shall be set and fixed in accordance with the Quertermous Survey." We agree with the appellants that the court's decree did not sufficiently "fix the boundary line with such certainty that it can be identified by reference to the decree." McEntire v. Robinson, 243 Ark. 701, 421 S.W.2d 877 (1967).

Therefore, the cause is remanded to the trial court with directions to fix and define the boundary lines, which establish appellants' encroachment, with such certainty that the boundary lines can be identified by reference to the decree.

Affirmed and remanded.

FOGLEMAN, J., concurs.