delivered the opinion of the court.
This is the sequel to the caveat case of McClung v. Folkes, decided by this court in November, 1917, reported in 122 Va. 48, 94 S. E. 156. The nature of the controversy and its history up to the date of that decision are tersely stated in the opinion delivered by Judge Burks, and we need not here repeat that statement in full, but the following extract therefrom will afford a helpful introduction to the present phase of the litigation. After referring to the mass of conflicting evidence in the case, Judge Burks said: “While there was this great volume of evidence relating to the issue submitted to the jury, the parties finally narrowed the issue between them to a single point and each of them staked the fate of the casey on whether the corner of one of the tracts was at a point which we may call A, or at a point which we may call B. At the. request of the plaintiff, and without objection from the defendant, they were instructed that if they believed the corner was at B, they should find for the defendant. Other instructions were given at the instance of each of the parties without objection from the other. Under these *262circumstances, the jury found for the defendant, and we cannot disturb their verdict.”
Shortly after the decision in this court was announced, McClung brought a suit in equity to enjoin the enforcement of the judgment against him in the caveat proceedings, and to obtain a new trial on the grounds of fraud and after-discovered evidence. The bill was filed at the first February rules, 1918, and at the next succeeding term of the court the defendant, Folks, appeared and demurred to the bill. In the meantime the plaintiff had taken and filed certain depositions. The circuit court having heard the cause on the bill and exhibits and the demurrer thereto, and, “the examination of witnesses,” sustained the demurrer and dismissed the bill. From that action this appeal was allowed.-
The alleged newly discovered evidence, which related solely to the location of the pivotal corner, consisted of a certain survey made for one Thomas Dixon in the year 1793, and recorded in “Bath County Journal No.. 1, p. 22.” The materiality of this survey as evidence for the plaintiff is manifest. At the second trial of the law cause, in July, 1914, and again at the third trial in July, 1915, the defendant Folks sought to fix the corner in question at a point known in the record as Terry’s gate, referred to in the former opinion in this case as point B. The evidence at both of these trials was such as that a finding by the jury either for or against this contention on the part of the defendant would not have been interfered with by the court. The plaintiff’s paper title, which was superior to the defendant’s, ran back to a certain Bradshaw survey, and the location of the disputed corner determined the location of that survey and of the plaintiff’s land. If the corner was at Terry’s gate, the defendant prevailed; if it was not at Terry’s gate, his theory failed and the plaintiff would probably have recovered a verdict. *263To maintain his contention, the defendant undertook to show that a certain marked white oak tree or stump at Terry’s gate indicated the corner in question. There was no positive evidence to identify this stump as being the corner of the Bradshaw survéy. Persuasive but not conclusive evidence was found in the fact that the Bradshaw survey called for white oak timber at that corner; that this tree was marked as a corner; that a block therefrom was exhibited to the jury and the annulations appeared *to carry the date of the marks (after accounting for the period subsequent to 1899, when it was said the tree was killed by lightning) back to the date of the Bradshaw survey, which was in 1791; and that according to the county surveyor, Beveridge, a principal witness for defendant, there was no other known survey of that period covering lands in that section of Highland county which might account for the marks on the white oak stump at that point.
There was material and strongly persuasive testimony in behalf of the plaintiff, that the white oak at Terry’s gate was not a corner of the Bradshaw survey, but there was no satisfactory explanation of the marks on the tree in the absence of evidence of. any other survey of that period.
The plaintiff, always insisting that the true corner was not .at Terry’s gate, but at a point which would so locate the Bradshaw survey, under which he claimed, as to cover the land in controversy, apparently relied upon the theory that the marks on the white oak stump were probably marks of a “Tomahawk survey” or of some surveyor’s work which had never been made the' basis of a grant from the Commonwealth. At any rate, the importance of showing that the marks were not made in the course of the Bradshaw survey was manifest and was fully understood.
[1, 2] The bill alleges that the Dixon survey was not *264only material, but conclusive evidence in the plaintiff’s behalf, and descends into particulars which present a most convincing argument in support of this allegation. The defendant, relying on the record of the former case, which was expressly made a part of the bill, earnestly contends that the new evidence was merely cumulative, and, further, that it could not produce a different result on another trial. We have no difficulty in holding that the evidence is not merely cumulative. Barsa v. Kator, 121 Va. 290, 298, 93 S. E. 613; Johnson v. Commonwealth, post, p. —, 101, S. E. 341, decided at this term. Whether we would be warranted in saying that, conceding its very manifest materiality, it “ought to produce” a different result, may be a debatable question, but it is one which, for reasons now to be stated, we need not decide.
The grounds upon which the demurrer was sustained are not desclosed by the record, but in the petition upon which this appeal was granted it is said that the judge of the circuit court delivered an oral opinion in which he took this position:
“That the date of the Bradshaw survey was known, and the annulations in the snag at Terry’s gate showed that these remarks were of a survey approximately the same date; that the very crux of plaintiff’s contention was that these were the marks of some other survey; that the court must take judicial cognizance of the historical fact that Bath county was cut off of Augusta; that in fact the “Dixon” survey was engrossed in the Surveyor’s Book of Bath county; and, under these circumstances, the court must say, as a matter of law, that plaintiff’s counsel ought to have discovered it, and failure to do so was such an omission as to debar plaintiff from any relief on the ground of after-discovered evidence.”
We concur in this view of the case. It is true that the *265plaintiff had made a very diligent search of the records for all title papers of every description affecting the title of either the plaintiff or defendant, and had been very thorough in this investigation, carrying it back to the Commonwealth. This investigation resulted in the discovery and production of a large number of old surveys. If the Dixon survey had been a title paper constituting a link in the chain of either the plaintiff’s or defendant’s title, then there would have been no obligation on the part of the plaintiff to make any unusual search, and an examination of the records in the clerk’s office by means of reference to the indexes would have been a sufficient examination. This is the usual method of examining titles by careful and capable lawyers, and thus meets the test of reasonable care and diligence. But a stray survey like the Dixon survey could never have been found by the ordinary methods of search, and the very theory and contention of the plaintiff was such as to call his attention sharply to this fact. A diligent search of all the title papers involved in the chain of either plaintiff or defendant had failed to disclose any other than the Bradshaw survey which could plausibly account for the corner at Terry’s gate. The annulations on the tree at that point showed with reasonable certainty that this corner was marked at about the date of the Bradshaw survey, which was in 1791. Bath county was carved out of Augusta county by an act of the legislature of December, 1790, effective in May, 1791. If there was a recorded survey, other than the Bradshaw survey, which could have been represented by the marks at Terry’s gate, and which' was never carried into grant, it would in all probability have been recorded among the records of Augusta county just before, or among the records of Bath county just after, the formation of the latter county. The usual method of examining title, as we have pointed out, would never have disclosed any “hid*266den” survey, as this one is called in the bill. • The only way to find such a survey would be to turn the pages of the books in which the same would probably have been recorded. This not only.would appear to have been a comparatively easy and brief task as an independent proposition, but it now actually appears as a matter of fact that the record which was found of the Dixon survey was on page 22 of Journal No. 1 in Bath county.
[3, 4] It is argued on behalf of the plaintiff that all this looks plain enough now, but that until the survey was actually discovered neither the plaintiff nor his counsel could have been expected to think of looking for the survey in this way. We cannot eccept this view. Of course, such a survey as this might never have been recorded anywhere, but it is certainly not improbable that it would have been, as this one was, recorded and then abandoned. With knowledge of the importance of finding the evidence-.of some survey which had never figured in the title of either party—a knowledge which the plaintiff certainly had as early as the trial in July, 1914—it seems to us that reasonablé diligence called for, and would have suggested, an examination of the surveyor’s books, page by page, in an endeavor to find the record thereof. ' Reasonable diligence always depends upon the facts and circumstances of the case.
[5] This controversy has been in court since 1912, and the caveat case was tried three times. The interests of society require that’ there be an end of litigation. Ño maxim of the law is more firmly established. For this reason the rule has always béen strictly adhered to, that however material may be the newly discovered evidence, a new trial will not be granted on that account unless such evidence could not have been discovered bv the exercise of reasonable diligence in time for use at the former trials. The instant case, in our opinion, is peculiarly one for the *267application of this rule. The evidence in question was a matter of record, its importance was known to the plaintiff for at least a year before the final trial, and it could have been discovéred by the exercise of diligence reasonably commensurate with the. manifest exigency of the situation.
“Where a new trial is sought on the ground of newly discovered evidence, it must appear not only that the evidence has been discovered since the former trial, but also that it is such, that by the exercise of reasonable diligence on the part of the applicant it could not have been procured for the trial. The authorities are unanimous to the effect that a new trial will not be granted on the ground of newly discovered evidence which could have been discovered before the trial by the exercise of reasonable diligence.
“(2.) Discovery of Evidence on Record.—Thus a new trial will not be granted if the newly discovered evidence is a matter of record, as it could have been discovered by reasonable diligence. Records are always accessible to both parties, and diligence requires that they be examined and procured in time for the trial. Especially is this rule applicable when it does not appear that the movant was ignorant of the existence of the record in question.” 14 Ency. Pl. & Pr. 798-802.
[6] “The discovery after the trial of a judgment, or judicial reeord, or of an ordinance or resolution of a municipal corporation, or of a deed or other instrument of record, or of any other matter of public record, is not ground for a new trial, unless, on diligent research in the proper office such record was not discovered before the trial.” 29 Cyc. 894.
“It is said to be a general rule that courts of equity will not grant new trials on the ground that a receipt or other material document was lost or missing, and has been dis*268covered since the verdict; and this is declared to be a rule of policy intended to secure care and vigilance, and prevent parties from coming forward subsequently with evidence which close investigation would have disclosed at the time; for it is truly said that a failure of justice in a particular instance is not so great an evil as that there should be no certain end to litigation.” 1 Bart. Chy. Pr. 46-7. See also Burks’ Pl. & Pr. 565; 10 Michie’s Dig. 447.
We are of opinion, therefore, that the plaintiff is not entitled to a new trial upon the ground of after-discovered evidence.
The bill contains a further allegation, which, although controlled by considerations already expressed herein, ought not to-be left unnoticed. That allegation is as follows : “Your orator is advised, believes and charges that the defendant Folks, in using said blocks and marks as evidence of the location of a comer in the Bradshaw survey, and as being surveyor’s marks, made in the course of that survey, did so with full knowledge that such were not the true facts, and that said marks in fact belonged to the Dixon survey, and that in this regard he acted knowingly and with! the intention of deceiving and defrauding your orator in the premises.”
This raises the question as to the circumstances under which a court of equity will grant relief against a judgment obtained upon false testimony.
In 1 Black on Judgments, (2d ed.), sec. 372, it is said:
“Whether relief will be granted in equity on the ground that the judgment was procured by the perjury of the plaintiff or a witness, is disputed. There are some English and American cases holding that such action is proper, where adequate redress cannot be had at law, and where the proof to convict the perjured witness could not be obtained in time to be used on the trial. And in at least *269one State -(Minnesota) the statutes authorize a suit to be brought to set aside a judgment obtained by the fraud or perjury of the prevailing party. In some other jurisdictions, it is thought that if a party to a suit intentionally procures and produces false testimony, suborning his witnesses to perjury and conspiring'with them to secure a judgment, this amounts to such fraud as will enable the adverse party, if defeated in the suit, to secure an injunction against the judgment. But this doctrine is denied in other States, and indeed the general current of authority is now in favor of the rule that perjury committed by the successful party or his witnesses at the trial is no sufficient ground for vacating the judgment or enjoining its enforcement.”
In Pico v. Cohn, 91 Cal. 129, 25 Pac. 970, 27 Pac. 537, 13 L. R. A. 336, 25 Am. St. Rep. 159, a proceeding in equity to annul a former decree on account of false swearing, the court expressed the following views, which appear to be in accord with the current of authority." “That a former judgment or decree may be set aside and annulled for some frauds, there can be no question; but it must be a fraud extrinsic or collateral to the questions examined and determined in the action. And we think it is settled beyond controversy that a decree will not be vacated merely because it was' obtained by forged documents or perjured testimony. The reason of this rule is, that there must be an end of litigation and when parties have once submitted a matter, or have had the opportunity of submitting it, for investigation and determination, and when they have exhausted every means for reviewing such determination in the same proceeding, it must be regarded as final and conclusive,' unless it can be shown that the jurisdiction of the court has been imposed upon, or that the prevailing party, by some extrinsic or collateral fraud, has prevented a fair submission of the controversy. What *270then is an extrinsic or collateral fraud, within the meaning of this rule? Among the instances given in the books are such as these: Keeping the unsuccessful party away from the court by a false promise of a compromise, or purposely keeping him in ignorance of the suit; or where an attorney fraudulently pretends to represent a party, and connives at his defeat; or being regularly employed, corruptly sells out his client’s interest. United States v. Throckmorton, 98 U. S. 65, 66, 25 L. Ed. 93, and authorities cited.
“In all such instances the unsuccessful party is really prevented, by the fraudulent contrivance of his adversary, from having a trial; but when he has a trial, he must be prepared to meet and expose perjury then and there. He knows that a false claim or defense can be supported in no other way; that the very object of the trial is, if possible, to ascertain the truth from the conflict of the evidence, and that, necessarily, the truth or falsity of the-testimony must be determined in deciding the issue. The trial is his opportunity for making the truth appear. If, unfortunately, he fails, being overborne by perjured testimony, and if he likewise fails to show the injustice that has been done him on motion for a new trial, and the judgment is affirmed on appeal, he is without remedy. The wrong, in such case, is, of course, a most grievous one, and no doubt the legislature and the courts would be glad to redress it if a rule could be devised that would remedy the evil without producing mischiefs far worse that the evil to be remedied. Endless litigation, in which nothing was ever finally determined, would be worse than occasional miscarriages of justice; and so the rule is,- that a final judgment cannot be annulled merely because it can be shown to have been based on prejured testimony; for if this could be done’ once, it could be done again and again ad infinitvm.”
*271In U. S. v. Throckmorton, 98 U. S. 61, 25 L. Ed. 93, 96, Mr. Justice Miller, after reviewing some of the cases on the subject, said: “We think these decisions establish the doctrine on which we decide the present ease; namely, that the acts for which a court of equity will on account of fraud set aside or annul a judgment or decree, between the same parties, rendered by a court of competent jurisdiction, have relation to frauds, extrinsic or collateral, to the matter tried by the first court, and not to be a fraud in the matter on which the decree was rendered.
“That the mischief of retrying every case in which the judgment or decree rendered on false testimony, given by prejured witnesses, or on contracts or documents whose genuineness or validity was an issue, and which are afterward ascertained to be forged or fraudulent, would be greater, by reason of the endless nature of the strife, than any compensation arising from doing justice in individual cases.”
In 15 Ruling Case Law, sec. 223, p. 771, the text says: “Since every losing litigant may consider himself the victim of false testimony, and believes it to be willfully false, and that his adversary knew it to be so, if relief should be freely granted in equity against a decision in another judicial tribunal on the ground that it had been procured through perjury and subornation of perjury, courts of equity would be engaged a greater portion of their time in retrying questions of fact, on the suggestion that their trial in the original action had been affectecl by perjury. • To prevent this result the doctrine of intrinsic and extrinsic fraud has been resorted to by some courts, and while conceding the fact that extrinsic fraud furnishes a foundation for equitable relief, the position is taken that false testimony amounts to intrinsic fraud, proof of which is inadmissible. It must be admitted that there is considerable confusion among the cases as to just how far *272the courts are justified in going in the granting of the relief for false testimony, and to what extent equitable interference amounts to an improper relitigation of questions determined in the original trial at law resulting in the judgment. The confusion appears to be due to the fact that no hard and fast rule can be .safely enunciated that will fit all cases; each case must, to a marked decree, depend upon its own facts and circumstances, and the granting or refusal to grant relief in such cases is largely a matter of discretion, which will seldom be interfered with on appeal.”
The authorities upon this question are multitudinous. See the citations in those above referred to, and in 10 L. R. A. (N. S.) 216, et seq.; 8 Ann. Cas. 83-86.
i [7] - We need not go further in the present case than to say that the authorities are practically harmonious in holding that before a court of equity will interfere with ,a judgment at law on account of perjured testimony at the trial in which the judgment was rendered, it must appear (1) that without such testimony the result ought to have been different, and (2) that the fact of its falsity could not have been discovered by the exercise of proper diligence on the part of the party complaining. This latter requirement can be made to account for some of the .apparent conflict in the decisions. For example, in the case of Marshall v. Holmes, 141 U. S. 589, 12 Sup. Ct. 62, 35 L. Ed. 871, the enforcement of a judgment obtained by means of a false and forged letter and false testimony was enjoined, but it affirmatively appeared that the evidence of the falsity was discovered after the judgment was rendered and could not have been anticipated or discovered before its rendition. In such cases the relief is granted on the just theory that, as expressed by Mj. Justice Miller in U. S. v. Throckmorton, supra, “there has *273never been a real contest in the trial or hearing of the case.”
[8] In the case at bar, we have already pointed out that the plaintiff knew for at least a year before the trial just what the defendant would contend and offer proof to show about the marks on the tree at Terry’s gate, and! that the exercise of reasonable diligence would have enabled! him to find the Dixon survey, which, as he claims, would have conclusively defeated the defendant’s theory. This being true, it appears not, so far as the legal result is concerned, whether the defendant’s evidence on the point was willfully false or not. Equity relieves only the diligent, however strongly the action assailed may be condemned by the conscience of the court.
The complainant claims that the depositions taken by him before the demurrer was passed upon or even filed, are sufficient to establish this charge in the bill. In view of this contention if we would say from the record before us that the case on this point had been fully developed, and that the complainant would not be able to add anything to his proof as already introduced, we might have disposed of this branch of the case by saying that the evidence was insufficient. It does not measure up to the clear and convincing character required in such cases. It does show that at some time the defendant and his, witness, Beveridge, learned of the Dixon survey, but it does not go far enough to establish the grave accusation that they knew its contents and effect in time for use before the trial was ended. We cannot say, however, that the plaintiff meant to commit himself to the proposition that he had completed his depositions when the demurrer was interposed and passed on. The argument of his counsel indicates that they think the alleged fraud is fully established, but the petition more than once complains that the court acted arbitrarily and by its action “on the de*274murrer made impossible the development of the true facts in the case.” The defendant’s counsel in their oral argument, while insisting that the proof of fraud was insufficient, took the position that the case ought to be heard on the bill and demurrer independent of the depositions. In their brief they very earnestly deny that the defendant had any knowledge of the Dixon survey, but no answer containing this denial has been filed. In this state of the record we have deemed it proper to take the case simply as it is made by the bill and exhibits and the demurrer thereto.
There is no error in the decree complained of, and it will be affirmed.
Affirmed.