This case involves a boundary dispute between adjoining landowners. Appellants claim that two boundaries between their property and appellees’ were established through either adverse possession, boundary by agreement, or the doctrine of boundary by acquiescence. The chancellor found that appellants had failed to establish boundary line by acquiescence or adverse possession and that a 1996 survey com*210missioned by appellees accurately established the boundary lines. We find no error and affirm.
Appellants and appellees are next-door neighbors in the town of Garfield. Appellants acquired their property from the Lawson family on January 14, 1994. Mrs. Reba Lawson had lived on the property since 1985. Appellees purchased their land from the estate of Ruth Carter on July 28, 1995. The Carters had resided on the property since at least 1951. Appellants’ property bordered appellees’ property on both the east and the north.
The controversy that gave rise to this case began when appellees procured a survey of their property in October 1996. Upon receipt of the survey, appellee Gary Eckel began construction of a fence along his eastern and northern borders. Appellants immediately filed a petition in Benton County Chancery Court to stop the construction. They alleged that the predecessors in title to both pieces of property — Mrs. Carter and Mrs. Lawson — had agreed upon boundary lines that differed from those shown on the survey. Specifically, appellants claimed that a small ridge running between the properties had been “acquiesced” in as the east/ west boundary and that a stone wall had been acknowledged as the north/south boundary for more than seven years. They stated further that they had adversely possessed the disputed area along the north/south border. Appellees generally denied appellants’ allegations. The case went to trial, and the chancellor made the following findings:
2. That the plaintiffs have failed to prove by a preponderance of the evidence that a boundary line dividing the parties [sic] property has been established by acquiescence or that the plaintiffs have adversely possessed the defendants property.
3. This court further finds that the survey which was stipulated as exhibit “4” and which has been filed for record in November of 1996 is the accurate boundary which divides the parties [sic] property.
It is from this order that appellants bring their appeal.
Chancery cases are reviewed de novo on appeal. Summers v. Dietsch, 41 Ark. App. 52, 849 S.W.2d 3 (1993). We will not reverse a chancellor’s finding of fact in a boundary dispute case *211unless the finding is clearly erroneous. Id. A finding is clearly erroneous when, although there is evidence to support it, the reviewing court is left with the definite and firm conviction that a mistake has been committed. Id. Location of a boundary fine is a question of fact. Kittler v. Phillips, 246 Ark. 233, 437 S.W.2d 455 (1969); Jennings v. Burford, 60 Ark. App. 27, 958 S.W.2d 12 (1997).
Appellants’-first argument on appeal is styled as follows: THE TRIAL COURT WAS IN ERROR IN HOLDING THAT THE APPELLANTS FAILED TO MEET THEIR BURDEN OF PROOF IN ESTABLISHING THE NORTHERN AND EASTERN BOUNDARIES BY ACQUIESCENCE. Our review of this case is complicated by appellants’ use of the terms “boundary by agreement” and “boundary by acquiescence” as though they were one and the same. They are not. See Seidenstricker v. Holtzendoff, 214 Ark. 644, 217 S.W.2d 836 (1949). In the case of a boundary by agreement, the landowners have made a parol agreement as to the location of the boundary. See, e.g., Nunley v. Orsburn, 312 Ark. 147, 847 S.W.2d 702 (1993); Moeller v. Graves, 236 Ark. 583, 367 S.W.2d 426 (1963). For a valid oral boundary line agreement to exist, four factors must be present: (1) there must be an uncertainty or dispute about the boundary fine; (2) the agreement must be between the adjoining landowners; (3) the line fixed by the agreement must be definite and certain; and (4) there must be possession following the agreement. Nunley v. Orsburn, supra; Fields v. Griffen, 60 Ark. App. 186, 959 S.W.2d 759 (1998). The agreement is binding even if the parties entering into possession pursuant to it do not occupy the land for the full statutory period of seven years. Rabjohn v. Ashcraft, 252 Ark. 565, 480 S.W.2d 138 (1972).
By contrast, a boundary by acquiescence arises not by a parol agreement but from the actions of the parties. See Jennings v. Buford, supra. It is more in the nature of an implied agreement presumed to exist by the long acquiescence of adjoining landowners who apparently consent to a dividing fine between their properties. Raborn v. Buffalo, 260 Ark. 531, 542 S.W.2d 507 (1976). The concept is based upon the landowners’ tacit acceptance of a fence fine or other monument as the visible evidence of *212their dividing line. Walker v. Walker, 8 Ark. App. 297, 651 S.W.2d 116 (1983). The acquiescence need not occur over a specific length of time, although it must be for “many years” or a “long period of time.” Jennings v. Burford, supra. But see Rabjohn v. Ashcraft, supra, holding that the acquiescence must exist for a period of seven years. Most boundary by acquiescence cases involve time periods of at least twenty years. See Kitler v. Phillips, supra (over fifty years); Clay v. Dodd, 238 Ark. 604, 383 S.W.2d 504 (1964) (fifty years); Gregory v. Jones, 212 Ark. 443, 206 S.W.2d 18 (1947) (thirty-four years); Jennings v. Burford, supra (twenty years); Summers v. Dietsch, supra (twenty years).
The caption of appellants’ argument refers to boundary by acquiescence, as does the chancellor’s decree. However, the majority of appellants’ proof below and their argument on appeal are directed to the theory of boundary by agreement. Because appellants and the chancellor treated the theories of boundary by agreement and boundary by acquiescence interchangeably, we consider the merits of the appeal as to both theories.
Appellants’ first argument focuses on two matters: (1) whether a boundary agreement actually existed between appellants’ and appellees’ predecessors in title, and (2) if an agreement did exist, whether it fixed a definite and certain boundary fine. The chancellor made no specific finding regarding the existence of an agreement. But at the close of the evidence at trial, he found that the proof was not sufficient to allow him to establish a definite, agreed-upon boundary line. We cannot say that the chancellor’s finding was clearly erroneous.
Appellant Allen Lammey testified that the east/west boundary between the properties was agreed upon by Mrs. Lawson and Mrs. Carter as follows: The southern point of the boundary line began at a maple tree. It extended northward along a ridge that ran part of the length of the property. The boundary was “aimed down” by a lilac bush that was in line with an iris garden, other flower gardens (some of which were no longer in existence), and three elm trees. The northernmost corner was represented by the corner of a rock wall that was no longer there. Lammey had marked the spot with a rock. Rick Lawson, the son of Reba Law*213son, testified that the maple tree at the southern point of the fine was “close to the boundary.” On cross-examination, he marked the southernmost point on a photograph at a location other than the maple tree. He testified that the line went through the lilac bush, through the flower beds, and to a group of elm trees. He further testified that the northernmost point of the east/west boundary was marked by a pin in the ground about a foot from a walnut stump. Appellee Gary Eckel testified that there had been a lilac bush on the property other than the one referred to by Lam-mey and Lawson and that he had torn it down.
Regarding the north/south border between the properties, Lammey testified that a rock wall, since torn down, represented the border fine. He said that the wall extended from the easternmost point westward to some remnants of an old fence. The remnants were located about four feet behind a barn that was no longer there. On cross-examination, Lammey admitted that the rock wall he referred to was actually a large rock pile. Lawson testified that the easternmost point of the border began at the pin near the walnut stump and ran westward to a point about three feet behind where the barn had been. He acknowledged that the “rock wall” was treated as a boundary, although he could not say if Mrs. Carter had considered it a boundary.
It is essential to the validity of a binding boundary line agreement that the boundary line fixed by agreement be definite, certain, and clearly marked. Walters v. Meador, 211 Ark. 505, 201 S.W.2d 24 (1947). Appellants cite Disney v. Kendrick, 249 Ark. 248, 458 S.W.2d 731 (1970), in which the chancellor’s finding of boundary by agreement was affirmed, in support of their argument that their proof was sufficient to establish a boundary line. The Disney case involved a boundary dispute between the Disneys and the Kendricks, who shared one common boundary between them. The chancellor found that, even though at one point no specific marker or monument plainly marked the boundary fine, a boundary by agreement was established when the Disneys and Mr. Fink, the Kendricks’ predecessor in title, installed concrete stobs at the southern and northern border points of their properties.
*214The proof in this case regarding the location of the boundary, or at least the end points of the boundaries, is not as conclusive as that in Disney. Approximate points are not sufficient. See DeClerk v. Johnson, 268 Ark. 868, 596 S.W.2d 359 (Ark. App. 1980). Appellants’ evidence may have provided a general idea regarding the location of what they contended was the agreed boundary. But we cannot say that, given the conflicts in Lammey’s and Lawson’s testimony and the destruction of certain landmarks relied upon by appellants, that, the chancellor’s decision should be reversed.
Appellants’ next argument on appeal is that the chancellor erred in holding that they were required to establish open and notorious possession for a period of seven years to prove “boundary by acquiescence.” Appellants are referring to the following remarks made by the court at the close of the evidence:
I mean there has to be an exact definitive location. It can’t just be out there somewhere. It has got to be exact. There are no monuments out there. I mean this rock wall was the best, but that turned out to be — I believe the general consensus now is it was probably more of a berm of rocks thrown out of a garden plot. Somebody mowed up to it on one side, and of course you couldn’t mow the rocks, so then somebody trimmed the rocks on the top, but that is not a definite marker. Of course, the other thing you have too is that case law has consistently said it requires a long period of time, and the case law has consistently used the term seven years which is for adverse possession. Also, as we are well aware, adverse possession requires more then [sic] just occupancy. That’s part of it, but then it has to be adverse, hostile, continuous possession for over a period of seven years. Here your people can only state for certainty their occupancy on the property, which was less than four years.
The court’s remarks again reflect the confusion both appellants and the chancellor were under in this case regarding the difference between boundary by acquiescence and boundary by agreement. Appellants are correct that possession for a period of seven years is not necessary to prove a boundary by agreement. Rabjohn v. Ashcraft, supra. However, a boundary may be established by acquiescence in a “clearly established line as the boundary over a period in excess of seven years” (emphasis added), and without *215the necessity of a prior dispute or adverse usage up to the line. Id. When the court’s remarks are viewed as a whole, it was the lack of certainty regarding the boundary line that made up the court’s mind to rule against appellants, and the remarks regarding adverse possession were offered as merely an additional reason to deny appellants’ petition. On de novo review, we cannot say that the trial court’s finding of uncertainty about the location of the line, which precludes appellants from prevailing on either a theory of boundary by agreement or boundary by acquiescence, is clearly erroneous.
Affirmed.
Neal and Meads, JJ., agree.