This action of ejectment was tried to a jury and a verdict rendered in favor of defendants. The writ of error to this court is based mainly upon exceptions to certain instructions given to the jury by the trial judge and to his refusal to give certain instructions requested by the plaintiffs. The nature of the questions thus raised will appear from the following statement:

The plaintiffs claim title to the lands in dispute under a survey of 480,000 acres made for Wilson Cary Nicholas by one J. A. Taylor in the year 1794. The report of this survey is dated October 19, 1794, and recites that it was completed on the 10th of September of that year. Upon this survey a patent, dated March 23, 1795, was issued to Robert Morris, as assignee of Nicholas. The tract of 36,750 acres described in plaintiff’s declaration, which is included for the most part in this 480,000 acres, was conveyed to Jonathan Patterson, Richard C. Ridgeway, and William G. Boulton, trustees, by Michael Bouvier and wife, by deed dated January 7, 1865. It is known and referred to herein as the “Dasher Tract.”

The defendants claim title under a survey of 500,000 acres, also made by Taylor for Nicholas, which purports to have been completed on the 9th of September, 1794, and upon which a patent was issued to him under date of June 25, 1795. This tract subsequently became forfeited to the state of West Virginia for nonpayment of taxes, and was proceeded against by the commissioner of school lands of Wyoming-*367county, who divided the tract into lots or sections, known as school sections, and sold the same under decrees of the circuit court of that county. The individttal defendants herein, or some of them, claim the surface of the particular school sections involved in this case, and the Pocahontas Coal & Coke Company claims the underlying coal and minerals.

There is no dispute as to the regularity of the paper title of the plaintiffs or the paper title of the Pocahontas Coal & Coke Company, and the latter appears to cover the lands in controversy. The real dispute is one of location. The plaintiffs contend that the school sections here referred to, or portions of them, overlap and conflict with the eastern boundary of their said tract of 36,750 acres, the Lasher tract, whilst the defendants contend that this tract does not extend far enough east to include these school sections, or any part of them, and therefore there is no conflict or overlapping.

The. common starting point of the two surveys above mentioned, the first running around to the west and the second around to the east, is designated A on the map used at the trial, and the location of this point is known and conceded. In the 480,000-acre survey its northwest corner is described as “three sugar trees and a buckeye by a small branch of Guyandotte river,” and it closes as follows:

“Thence N. 89 degrees E. 10,050 poles, crossing Guyandotte and several branches thereof, to a line of survey of 500,000 aeres made for said Nicholas, and with the same S. 9,700 poles, crossing Guyandotte, Branson’s fork, and Indian creek to the beginning.”

In the 500,000-acre survey its northeast corner is described as “two poplars and a buckeye tree on a ridge that divides the branches of Guyandotte,” and the last two calls are:

“N. 79° W. 8,040 poles, crossing several branches, and thence with a line of survey of 480,000 acres, made for said Nicholas, south 11,260 poles, crossing Guayandotte, Branson’s Fork of Indian creek, to the beginning.”

It thus appears that these two surveys, which were completed about the same time, have a common boundary line, described as running north and south, which extends from A, the common point of beginning, to the northerly line, of the 480,000-acre tract. There was also another survey by Taylor of a tract of 320,000 acres on which a pat•etii was issued to Robert Morris, assignee of Nicholas, under date of March 4, 1795, the beginning point of which was “5 chestnut trees” on the top of a described ridge. This point is designated D on the map referred to, and its actual location on the ground is known and conceded.

By various transfers of title, which need not be recited, the two tracts of 480,000 acres and 320,000 acres, which were patented to Morris, came into the possession of Charles Feinour, Jr., in November, 1846. He conveyed 50,000 acres out of the 320,000-acre survey to one Beck, leaving in the twTo tracts, as was supposed at the time, an aggregate •of 750,000 acres. Subsequently, and after other changes of ownership, these two tracts were mortgaged in 1847 to John W. Tilford, who later assigned his mortgage to Michael Bouvier. On the 7th of December, 1850, Bouvier entered into an agreement with John Herman, Eus*368tache Bouvier, Oakes Terrill, Jr., Edwin C. Searles, and William A, Bull, who then owned the two tracts as tenants in common and in the .following proportions, namely: Herman rV1B0, Eustache Bouvier 3B/i50, Terrill 1B/iso, Searles 1B/ioo and Bull. 8/ioo. This agreement was to the effect that each of these owners would give his promissory note to Bouvier' for such proportion of the amount due on the mortgage as equaled his undivided interest in the mortgaged lands; that Bouvier would foreclose his mortgage and take title under the foreclosure sale; that he would then convey to each of the owners who had paid his note his proportionate share of the lands as follows, namely: to Herman 85,000 acres, to be surveyed in a square lot as nearly as migh,t be and located next to lands of H. Stiles; to Terrill, 75,000 acres next to Herman’s lot; to Searles, 75,000 acres next to Terrill’s lot; to Bull, 40,000 acres next to Searles’ lot; to Eustache Bouvier, 175,000 acres next to Bull’s lot; and also to Herman the balance-of the lands amounting to 300,000 acres; that in case a survey should show that there was not enough land to give each of the parties named the number of acres mentioned, then there should be a proportionate abatement of their respective shares, and if the quantity exceeded the estimate of 750,000 acres the share of each should be correspondingly increased. It was also provided that, if either of the parties failed to pay his note at maturity, the agreement to convey to him would become void, and Bouvier would be at liberty to sell to any one he chose, or to keep for himself, the share of the one so in default.

In pursuance of this agreement Bouvier foreclosed his mortgage, bid in the property at the sale, and got a deed therefor under decree of the court in October, 1852. In the meantime, or about that time, one Henry B. Harman was employed to make a survey of the lands and divide them into lots according to the respective interests of the parties to whom they were to be conveyed. He found, however, or reported, that the two tracts in question, after deducting the 50,000 acres which Feinour had sold to Beck, contained only 157,500 acres, instead of the supposed quantity of 750,000 acres. Under the agreement, therefore, the former owners, provided they paid their notes, were entitled to receive as follows, namely: Herman (7T/ioo) 80,850 acres to be divided into two lots, Terrill and Searles each (1B/i5o) 15,-750 acres, Bull (8/ibo) 8,400 acres, and Eustache Bouvier (ffB/iso) 36,-'750 acres. It appears that the notes of Herman, Terrill, and Searles were paid, and accordingly, in March, 1853, Michael Bouvier conveyed two lo.ts aggregating 80,850 acres to Herman, one lot of 15,750 acres to Terrill, and one lot of 15,750 acres to Richard Warren, who had succeeded to the interest of Searles.. The notes of Bull and Eustache Bouvier were not paid,' and consequently Michael Bouvier was left free to do as.he pleased with the lots which had been surveyed for their respective shares. Nearly 12 years later, in January, 1865, he. conveyed to Patterson, Ridgeway, and Boulton, by the deed above mentioned, under which the plaintiffs claim title, the lot of 36,750 acres whid} had been surveyed for Eustache Bouvier.

The plaintiffs’ case rests upon two propositions: First, that the common boundary of the 480,000-acre tract, under which plaintiffs *369claim, and the 500,000-acre tract, under which defendants claim, is a line which in 1794 ran due south to the common starting point of the two surveys, and that the proper way to locate this common boundary at the present time is to allow such variation of the compass as is necessary to retrace on the ground a line running exactly south in 1794; second, that the lands described in the conveyance of the Lasher tract to Patterson, Ridgeway, and Boulton, under which plaintiffs also claim and must claim, extend eastwardly to the common boundary so located of the two tracts mentioned. It is also contended in effect that the location of this common boundary and of the eastern line of the Lasher tract, as claimed by the plaintiffs, appears from an inspection of the surveys and deed in question and the language therein contained. The refused instructions of which plaintiffs complain virtually involve a construction of the records in accordance with this contention, since, if granted, they would logically lead to a verdict for the plaintiffs. For example, take plaintiffs’ instruction No. 7:

“The court instructs the jury that the common line between the 500,000-acre patent, under which defendants claim, and the 480,000 Robert Morris patent, under which the plaintiffs claim, is a lino that runs due south to the admitted beginning corner at the mouth of Adkins branch, and that in extending the location of said line on the ground at the present time the line should be run with such variation of the compass as is shown by the evidence to be necessary in retracing at the present time a line run in 1795.”

Again, plaintiffs’ instruction No. 11 is as follows:

“The court instructs the jury that according to the evidence in this case the eastern line of the Lasher tract, which is claimed by the plaintiffs, coincides and runs with the easlern line of the 480,000-acre patent offered in evidence, and if they believe from the evidence that according to the true location of the eastern line of said 480,000-acre tract said tract embraces the land in controversy, then they should find that said land in controversy is embraced within the lines of the Lasher tract.”

In short, the plaintiffs insist that the construction of this documentary evidence was a matter of law for the court, that its proper construction involved affirmance of the propositions embodied in the quoted requests, and that it was therefore error to hold that these questions of location were questions of fact for the determination of the jury. Stated in another way, the assignments of error here considered are based upon the theory that there was practically no issue of fact for the jury, because the instructions requested would be consistent only with a verdict for the plaintiffs.

As the verdict was in favor of the defendants, it is not necessary to review the facts and arguments which support the plaintiffs’ contention. The real dispute here is whether the opposing facts and circumstances relied upon by defendants rendered the true location of the school sections in controversy so doubtful and uncertain as to make their location a question of fact for the jury. We proceed, therefore, to take up some of the considerations which are urged to sustain the submission of the case and the judgment entered upon the verdict.

First, as to the common division line of the two large surveys made in 1794. Should this line now be located as matter of law by running north from the common starting point with suitable allowance for *370variation of the magnetic needle, which is said to be 5° east? It will be observed that the closing line of the 480,000-acre tract is 9,700 poles long, while the closing line of the 500,000-tract is 11,260 poles long, so that the latter tract extended much further north than the former. Neither of these surveys calls for any natural object at the point from which the last course is stated to run south to the place of beginning. The northwest corner of the 480,000-acre survey is located by reference to described natural objects from which the distance is given in an.easterly direction. In like manner the northeasterly corner of the 500,000-acre survey is located by reference to described natural objects from which the distance is stated in a westerly direction. Now, while each tract is said to be bounded by the other, from the northerly line of the former to the place of beginning, it does not follow, if the distances given for their respective northerly lines were now measured on the ground, one running easterly 10,050 poles from its northwest corner and the other westerly 8,640 poles from its northeast corner, that lines run from those points to the place of beginning would either be due south or coincide with each other. Indeed, it is quite conceivable that, if thé stated distance be measured from the northeast corner of the 500,000-acre survey, á line run from the end of that distance to the starting point would pass west of the lands in dispute, although a line run to the same place of beginning from a point at the stated distance from the northwest corner of the 480,000-acre tract might run approximately south and pass east of the lands.in dispute.

Although the Morris patent was earlier in date than the Nicholas patent, the- survey of the Nicholas tract bears an earlier date than the other survey, and it might well be claimed that the older survey would control, or have to be first satisfied, if they were found to be conflicting. In the nature of the case these surveys of vast tracts of land, which were then a wilderness, must have been more or less inaccurate, and it would not be in the least surprising to find "that the descriptions therein given, particularly those here referred to, could not now be made to conform to a common boundary which was stated more than 100 years ago to run due south. Nor does it seem certain that a line run at the present time from the common starting point north 5° east would coincide with a line that was due south in 1794. Doubtless the 5° east represents the recognized variation in the magnetic needle and would be theoretically correct. But Wagner, the official surveyor, testified that other lines of these surveys, when run according to their actual location upon the ground, were found to vary from 5° to 25° from the true magnetic bearing. In order to retrace a line and locate it accurately by this method—that is, by a given allowance for magnetic variation—it would be necessary to know that the original survey was correct, as well as the assumed variation. Obviously, in lines of such considerable length as those here in question, a slight change in the allowed variation would materially change the location of the reproduced line. Moreover, it appears to be the settled rule .of law that the location of a line by retracing its described course is not allowable, unless there is a conflict in the described calls which can thus *371be harmonized. Without referring to other matters in this connection, we think enough has been said to show that the proper location of the division line between these old surveys was sufficiently uncertain u> justify the trial court in holding that it was a question of fact for the jury, and that it was not error to refuse instructing the jury that it should be located in the manner contended for by plaintiffs.

Second, as to the location of the Rasher tract. Does this tract extend easterly to the division line between the two original surveys, and so take in the school sections in controversy ? This was the tract con-vet ed by Michael Bouvier to Patterson, Ridgeway, and Boulton by deed of January 7, 1865, which describes a tract of 36,750 acres by metes and bounds according to the survey of Harman above mentioned, and was one of the subdivisions of the two large tracts referred to, lying south of the Herman 63,000 acres. As above stated, deeds were given to Herman, Terrill, and the successor of Searles in 1853 in pursuance of the agreement entered into in 1850. In all these deeds there is full reference to the 1850 agreement, and in each of them the intention is expressed to convey the proper fractional part of the 750,000 acres, found by Harman to be only 157,500 acres, be the number of acres more or less, whichever survey proved to be correct. In other words, these deeds all show on their face that they were executed in pursuance of this agreement, and were intended to give to the respective grantees their proper share of all the lands acquired by Michael Bouvier, whether the two tracts contained the supposed quantity of 750,000 acres as stated in the original surveys of Taylor in 1794, or only the i 57,500 acres reported by Harman in 1852; that is to say, if the tract conveyed to John Herma.ii, for example, by metes and bounds, proved for any reason to he. less than his rightful share of the entire lands, be would nevertheless be entitled to his distributive portion of whatever remained unconveyed. But the failure of Eustache Bouvier to pay liis note left Michael Bouvier free to dispose of that share as and when In; saw fit; and it is noticeable that the deed to Patterson and others in 1865 omits all reference to the agreement of 1850 and the other matters relating thereto which were recited in the deeds given in 1853. In other words, the grantees of the Rasher tract were not parties to the contract with Michael Bouvier, and could therefore claim only so much as was deeded to them by metes and bounds in the conveyance of 1865.

Moreover, the boundaries of this tract are those which were located and marked on the ground by Harman in 1852. The deed to Herman for the 63,000 acres recited a line running along the north side of Porks Ridge to a stake “on the old marked line of said 480,000 -acre survey,” and the deed to Patterson, Ridgeway, and Boulton calls likewise for “an old marked line” as the eastern boundary of the Rasher tract. In both these deeds the northerly line of the Lasher tract and the southerly line of the Herman tract are described by distances which are very much less than the northerly line of the 480,000-acre tract as described in the original Taylor survey. In short, there are such differences in the descriptions and distances, including the repeated reference to the old marked line, as to justify doubt at least wheth*372er Harman in his survey of 1852 did not stop considerably short of the eastern boundary of the 480,000-acre tract; or, if this old marked line was in fact on its eastern boundary as originally located, then that boundary was considerably west of a line running north from the starting point. It is difficult on any other theory to account for the great discrepancy between the 750,000 acres given by Taylor and the 157,500 acres found by Harman. There is certainly ah inconsistency between the descriptions in the original survey and the metes and bounds fixed by Harman some 58 years later, and that inconsistency seems of such significance, and permits such varying inferences, as to make the true location of the eastern line a question of fact, and to justify the trial court in refusing to instruct the jury that the Basher tract extended to the old division line between the surveys of 1794. In other words, the evidence as a whole showed such a degree of uncertainty respecting the actual location by Harman of the eastern line of the particular tract in question that it became a question of fact which was properly submitted to the jury.

It is not to be denied that the facts and arguments advanced by the plaintiffs are of such persuasive character that the jury would have been clearly justified in returning a verdict in their favor. But we are nevertheless, convinced, after painstaking study of the record, that the plaintiffs’ proofs were not sufficiently convincing to entitle them to the refused instructions, and that the trial court did not err in holding that the location of the school sections in controversy was a question of fact for the jury to determine.

In view of the verdict for defendants and the conclusions reached by us respecting the assignments of error, it is not necessary to consider the questions raised by the defense of adverse possession.

We are of opinion that no reversible error has been made to appear, and the judgment below will therefore be affirmed.