Although I may yield to the opinion of others, I never shall be reconciled to the doctrine that an absolute deed can, at law, be turned into a mortgage by parol evidence; nor, that it can be done in a court of equity, except on the ground of fraud or mistake. It is contrary to a first principle in the law of evidence to allow a deed or other written instrument to be contradicted by parol proofs. And in the case of a deed of lands, it is also contrary to the spirit, if not the letter, of the statute of fiauds to allow snch evidence to be given. No estate or interest in lands, other than leases for a year, nor any trust or power over or concerning lands, can be created, granted, assigned, surrendered or declared, unless by act or operation of law, or by a . deed of conveyance in writing. (2 R. S. 134, § 6.) Now what was the effect of the parol evidence given in this case but to divest the plaintiff of his formal legal title, and give the lands to another 1 The estate of the plaintiff is made to pass to the defendant Aldrich without “a deed or conveyance in writing.” A trust concerning the lands, to wit, that Moore should have them again in a certain event, is created or declared by parol proof.
I have already expressed my dissent from this doctrine in the case of Swart v. Service, (21 Wendell, 36,) where some of the evils of admitting such evidence are very briefly suggested. On looking again at that case, I find, contrary to what was supposed at the time, that it was not then necessary to pass upon this question; and I therefore feel myself at liberty to renew that dissent. I am the more encouraged to do so on finding that my brethren agree with me in principle, whatever they may think on the score of authority. In Patchin v. Pierce, (12 Wendell, 61,) it was held, that parol evidence was inadmissible to show that the sum intended to be secured was less than that mentioned in the condition of a mortgage of per*609sonal property, because it contradicted the written instrument. The principle of that case is not only applicable to this, but the remarks of the present chief justice in delivering the opinion of the court, are directly to the present purpose. “ An absolute deed,” ■ he says, “ may in equity be turned into a mortgage by parol proof; but that is on the assumption of fraud in the grantee, upon which ground the action of the court is sustained. If there is a mistake in the mortgage as to the amount of indebtedness of the mortgagor, the remedy, as in all cases of this kind, is to be sought in a court of equity. Here, the party must be bound by his agreement.” In Swart v. Service, already referred to, Cowen, J, after mentioning two cases in this court where such evidence was held admissible, says, “For one, I was always at a loss to see on what principle the doctrine could be rested, either at law or in equity, unless fraud or mistake were shown in obtaining an absolute deed where it should have been a mortgage. In either case the deed might be rectified in equity; and perhaps even at law, in this state, where mortgages stand much on the same footing in both courts. Short of that, [fraud or mistake] the evidence is a direct contradiction of the deed” A court of law could not, I think, undertake to reform the deed on the ground that it deals with a mortgage upon equitable principles, because the instrument is not a mortgage, but a deed, and the very point in controversy is, whether it can be changed from the one to the other by parol proof. My brother Cowen has also plainly manifested his opinion on this question in the Notes to Phil, Ev. p. 1429 to 1434, where most of the cases are collected. . See also Meads v. Lansing, (1 Hopk. 124,) which shows that this deed could not be reformed, even in a court of equity, without some evidence in addition to' that which was given on the trial.
I shall not examine the cases, nor go at large into this . question. The principle which forbids a court of law to listen to such evidence as that on which the defendants rely,- is a familiar one. It has been adjudged over and over *610again, and cannot be departed from without letting in a long train of bad consequences. Let us look for a moment at the facts of this case. Moore not only gave an absolute deed, but he intended to give such a deed. There was neither fraud nor mistake. The parties did just what they intended to do. But now, Moore, (or his grantee,) says, it is not a deed, but a mortgage, because the agreement was, if I repaid the consideration money, or paid' the note which you endorsed for me—it is impossible from the evidence to say what the pretended agreement was.—the deed was to be given up, or I was to have the land back aghin. And this parol defeasance is made out, not by proving what was done and said at the time, but by the mpst suspicious and least reliable of all kinds of evidence—what 'some one heard, or supposed he heard the plaintiff say about the transaction. And thus the plaintiff’s deed is turned into a mortgage, and his title to lands overthrown, by a conversation which may have been misapprehended at the time, or imperfectly recollected on the trial. No man can feel secure of the enjoyment of his house or lands, when his title may be overturned by such evidence.
Suppose the plaintiff should yield, and agree that his deed is a mortgage. He could not then bring ejectment. (2 R. iS. 312, § 57.) His only remedy would be to foreclose in equity; and how, when he files his bill for that purpose, is he to prove that his deed is a mortgage 1 What he has said on that subject will not be evidence in his favor. His conveyance will then be what the parties made it, an absolute deed, and he will be sent back again to a court of law to try his le-' gal title there. But suppose he makes out a parol defeasance to the deed, he will then, I presume, be met by the objection that the conveyance was not recorded as a mortgage, but as a deed, (1 R. S. 756,) and that it is consequently void as against .the subsequently recorded deed to the defendant Aldrich. Notwithstanding the case of James v. Morey, (2 Cowen, 246,) there will be two sides to that question.
If it is now too late to say, that parol evidence to turn *611an absolute deed into a mortgage, is inadmissible in a court of law, it certainly is' not too late to say, that the evidence given in this case was improper and insufficient. In Roach v. Cosine, (9 Wendell, 227,) "and Walton v. Conly, (14 id. 63,) where parol evidence was admitted, it was evidence of what the parties did and said at "the time—it was direct proof, and that too, not only of an agreement that the conveyance should be held as a mortgage security, but of such facts as plainly went to show that such was the nature of the transaction. But what kind of evidence have we" here ? The two witnesses, Knowles and Parker,- were allowed to testify, not to what took place at the time, nor to facts as they then existed, but to what they had subsequently heard the plaintiff say. They knew nothing of - a pre-existing debt/a loan of money, or the endorsement of a note, for which the conveyance should be held as a security. And then, what kind of a defeasance do they make out? Knowles says, the plaintiff told him, that Moore had deeded the lot to him, and if Moore paid a lank note, he would deed the lot lack to him. What note, for how much, and when payable? All these matters are left in the dark; and besides, the conversation goes to prove a present intention, rather than an original agreement to reconvey. The evidence of Parker is much of the same character; but whether the plaintiff spoke of" the payment of the consideration money, or a certain note, the witness is unable to state. The account which Moore himself gives of the matter is no better, except that he may be supposed to speak of what took place at the time the deed was made. What endorsement, for how much, and when payable, he has not told us. But what is worse than all the rest, his evidence goes to prove an express trust, rather than a mortgage, contradicting the language and nature of the conveyance. He says it was expressly understood that the deed was not to be put upon record, and whenever he called for the deed, the plaintiff was to deliver it up.
Upon this case, the judge decided, that if the jury lelieved the testimony of Moore, ■ or that any defeasance to *612the deed was established, although the terms of that defeasance were not proved, the defendants were entitled to a verdict. There is, I think, no case in any court, and certainly not at law, which goes so far as this. It is bad enough to allow the parties to set up a parol defeasance to an absolute deed, when the terms of the defeasance are clearly established. But here the deed has been contradicted and overthrown on proof that there was either a trust, or a condition of some kind, without telling us what in particular that condition was.
I think there should be a new trial.
New trial denied.