The facts of the case fully appear in the opinion of the court, delivered
by Woodward, J.The only question on this record is, whether parol evidence was admissible -to prove that fourteen acres of land were included by mistake in the deeds under which the plaintiff claims; and, instead of attempting to educe from the multitudinous and jarring authorities on the subject of parol evidence, to vary written instruments, a rule that would be ap-, plicable to the question, I propose to treat it upon its elementary principles.
The plaintiff holds the legal title to the land in controversy, *56by virtue of two several deeds of the defendants, duly executed and delivered at different times, both of which describe the land conveyed'by metes and bounds which confessedly include the fourteen acres. Ejectment by him, therefore, is strictly an action at law. The defendants have no legal title. And at law Henry Hopple, the immediate grantor of the plaintiff is estopped by his deed from denying that he had title to the fourteen acres, or that he meant to convey it to the plaintiff. In like manner Clemens is estopped from denying the title he conveyed to Henry.
But in equity and good conscience those deeds ought to operate only so far as they express the intention and understanding of the parties ; and if, indeed, the fourteen acres were not bought and sold, but were included in the deeds by mistake, the defendants claim that the deeds should be reformed so as to conform to the intention of the parties. In other words, they are plaintiffs in chancery seeking a decree that the fourteen acres be reconveyed to them because included in their deeds by mistake. Notwithstanding the former action and the common law form of the present action, this is the proper light in which to regard the case — as upon bill and answer, in which the defendants on this record would be plaintiffs,, and the present plaintiff defendant — they alleging the mistake under oath, and he under oath denying it. If, then, the principle that a plaintiff cannot go into parol evidence for the purpose of obtaining a specific performance of a written agreement with a variation, though a defendant may to resist it, were applied here it would exclude the evidence on this ground at once, and dismiss the plaintiff’s bill. But this principle, though settled in many English cases, was successfully denied by Chancellor Kent in Gillespie v. Moore, 2 Johns. Ch. R. 598. He declared, what every one must feel to be true, that there would be a most deplorable failure of justice if mistakes could only be shown and corrected when set up by a defendant to rebut an equity. Ever since that case, which was decided in 1817, it has been a conceded jurisdiction of courts of equity in the United States to reform written instruments, at the instance of either plaintiffs or defendants, on the ground of fraud .or mistake, upon parol evidence, where no statutory provision intervened. It is obviously an appropriate branch of equity. A court of law may construe and enforce the instrument as it stands, or may set it aside altogether if there be adequate cause; but it cannot compel alterations to be made, and an avoidance of the entire instrument would be in most cases a nullification, and not an affirmance of what was really meant. Adams’ Equity, 406.
. How then would a chancellor regard this defence if presented *57in the form of a bill for the re-execution and correction of the deeds ?
In the first place, the statute of frauds and perjuries would not stand in his way, for, though the effect would be to pass an estate by parol, yet the statute must be so construed as to prevent frauds, and not to promote them. And this would apply where mistake and not fraud was the ground of the relief sought; for though a mistake does not necessarily include a fraud, yet to set up and use a written instrument for a different purpose from that for which it was made, would be as inequitable as to take advantage of an instrument fraudulently obtained.
But the chancellor would have to be satisfied that the mistake was on both sides — for if it be by one party only, the altered instrument will not express the intention of both. A mistake on one side may be a ground for rescinding a contract, or for refusing to enforce its specific execution, but it cannot be^a ground for altering its terms. Adams’ Eq. 411.
And what is the kind of proof a chancellor would require ? Chancellor Kent, after reviewing all the leading English cases, says, in Gillespie v. Moore, that the cases concur in the strictness and difficulty of the proof, but still they all admit it to be competent, and the only question is, does it satisfy the mind of the court ? He quotes Lord Hardwicke as saying that it must be proper proof, and the strongest possible proof, and Lord Thurlow’s remark that it must be strong, irrefragable proof, the difficulty of which was so great that there was no instance of its prevailing against a party insisting that there was no mistake.
We can get an adequate idea of the degree of certainty to which the parol proof must rise only by considering the value of the testimony afforded by deeds solemnly executed between the parties. The rule, in courts of law,.as already intimated, is that the written instrument contains the true agreement of the parties, and that the writing furnishes better evidence of the sense of the parties than any that can be supplied by parol. And let it be remembered that the only purposes for which deeds were invented, and by the statute of frauds a writing signed was rendered necessary in regard to land, were to secure evidence of contracts certain as to subject matter and interest. They become, when executed, the agreed evidence of the intent of the parties, as to what is conveyed, for what estate, and under what conditions or covenants. Domat, cited in Best, on Ev. 239.
All that precedes their execution is presumed to have been abandoned by the parties, except in the solitary instance of a deed which, by the very terms of a contract is intended as a partial execution, as in the case of Girard v. McCulloch, 4 W. C. C. R. 292. If a man, says Grlanville, acknowledges the seal *58attached to a deed to be his own seal, hete bound to warrant the terms of the deed, and in all respects to observe the compact expressed in the deed as contained in it without question. To overcome evidence of such dignity and worth, parol proof ought to come as near to absolute demonstration as any moral proposition can be brought, and hence chancellors everywhere, while asserting jurisdiction to reform deeds, have demanded a-clearness and fulness of proof to justify the exercise of so extraordinary a power, which in practice have amounted to an almost total abnegation of the power itself.
It is one of the peculiarities "of the present case that there are two deeds to be reformed. On the 7th April, 1858, Clemens Hopple and wife conveyed to Henry Hopple two pieces of land, containing together about one hundred and seventy-five acres, and including the fourteen acres in dispute. This is' the first deed in which the mistake is alleged to exist.
Henry Hopple had purchased thirty adjoining acres'of Henry Keogh, and obtained his deed therefor, on the 22d June, 1848, and at the same time the above deed, Clemens to Henry, was executed; there was indorsed on the deed from Keogh to Henry, the agreement between Clemens and Henry, which, without date or signature, is furnished to us in the defendant’s paper book. It is said that this is an agreement of exchange whereby the fourteen acres were secured to Clemens, though included in his deed to Henry. I should be sorry to doubt that the object of this agreement was just what is claimed for it, because a doubt would imply some comprehension of its meaning, and I confess its terms are unintelligible. Be it, however, what it is claimed to be, a reconveyance of the fourteen acres to Clemens, it proves no mistake in the deed, and if it did, how is Schettiger to be affected by it ? He was not privy to the agreement, it was not recorded, and he- had no notice of it. ■ I say it does not tend to prove a mistake in the deed, but the very reverse. On the same day that a grantor conveys one hundred and seventy-nine acres of land to another, the grantee executes a reconveyance of fourteen acres to the grantor. Does that prove a mistake in. the deed ? Quite the contrary. Both parties must have known that the fourteen acres had passed by the deed, else why the reconveyance? And if both knew that the fourteen acres passed there was no mistake.
But now as to the witnesses, Bender and Baker. Bender was present at the execution of the deed of Clemens to Henry, and witnessed it. “ I don’t recollect anything said when the deed was executed. Don’t mind anything said of the fourteen acres at that time. I wrote the agreement on back of the Keogh deed. The agreement and deed from Clemens and *59Henry were executed at the same time and place. Esquire Luther and I were present. I know the fourteen acres. It was agreed, as it states here, that the old man was to have the fourteen acres, more or less. I was not present when the deed was written. It was said there that day that the old man was to have that land: that was the agreement.” This evidence was objected to as irrelevant, and considering that the question at issue was mistake or no mistake in the deed, it was most truly irrelevant. Laying out of view the confusion and inconsistency of the statements of the witness, what more do they amount to than proof of a conveyance and re-conveyance? The witness proves the deeds executed between the father and son, and describes what he understood to be their effect. The question of mistake rests then just where it rested before — on the deeds — and this testimony goes for nothing. The deeds that day executed, interpreted as this witness interprets them, excludes the conclusion of a mistake. The testimony of Baker was, if possible, still more irrelevant. He tells us how Clemens acquired the land he conveyed to Henry, and how the latter got the thirty acres of Keogh, and then says, “ they both agreed that Henry should keep the fifty acres, and Clemens should have the thirty acres and the fourteen acres for himself, and he never sold it to Henry but kept it for himself.” When this agreement was made, whether before or after the deeds, we are' not informed, but 'the witness was not present at the execution of the deeds — does not speak of or allude to them, and of course, proves no mistake in them or either of them.
Such was the evidence of mistake on which the defendants relied. Let us notice in what circumstances of the parties it was offered and permitted to prevail.
Henry Hopple having obtained his father’s deed on the 7th of April, 1853, as before stated, entered into an article of agreement, on the 12th of September, 1853, for the sale of the same one hundred and seventy acres to Philip Schettiger, and agreed to make “ a good and indisputable title therefor, on the 1st April, 1854.” Accordingly, on the. 1st of April, 1854, he and his wife executed their deed to Schettiger for the land, in consideration of $2,500. Schettiger went into possession and farmed the fourteen acres — some witnesses think under Clemens Hopple, as they saw him haul grain to the old man’s barn. Browne says Schettiger told him he farmed the fourteen acres on the shares for the old man. Other witnesses say he refused to give the old man a share of the grain — inquired about the lines, and claimed to have bought all the land that was described in his deed. The only witness who was present at the execution of the deed says, Henry sold the one hundred and seventy acres and allowance — “ Nothing mentioned about the *60fourteen acres at the time the article was drawn. There was no reservation at all. Schettiger was to get what was in the original old deed — that was what was said, and nothing said of' any reservation.” Now where in all this was that clear and overwhelming evidence of mistake on which a chancellor would alone base a decree of re-execution ? The testimony of Bender and Baker related to the deed between the father and son, and failed to establish any mistake in that. If Schettiger farmed the fourteen acres under the old man, after the purchase from Henry, it may have been because of an unexpired lease — or of a misunderstanding as to line's and boundaries, or to avoid present dispute, or for other reasons. As evidence of mistake running through two deeds and an article of agreement — instruments executed with great deliberation and at intervals of several months — it is not worthy of a moment’s consideration. It has been said in Pennsylvania that what occurs at and immediately before the execution of papers may be proved by parol to establish a mistake; but all the evidence on the record that would come within that rule proves that there was no mistake. The defence does not rest on the muddy agreement that was indorsed on Keogh’s deed. If it did it would be necessary to interpret that instrument, and to bring home notice of it to Schettiger. But the case was put on the allbged mistake, and in respect to that there was a mistrial throughout.
The evidence relied on was entirely inadequate to establish the mistake, and the jurisdiction invoked being a chancery jurisdiction,'it was to be exercised by the court instead of the jury. The question, said Chancellor Kent, in all such cases, is whether the proof is satisfactory to the court. That question was not met at all in this case. If the evidence is conflicting it is for the jury to determine what is proved, but it is for the court to say whether the facts found establish such a mutual mistake as would make it a fraud to hold the parties to their • writing. And by means of special verdicts the court can always reach the real question in such a manner as to deal with it on its appropriate principles. In this case the discretion of the chancellor could have been safely exercised without a special verdict, by rejecting the inadequate evidence offered, and directing a verdict for the plaintiff.
The judgment is reversed and a venire de novo awarded.