delivered his opinion.

The sole question put in issue by the pleadings, is the freedom of the plaintiffs. If they caii establish, by legal evidence, their title to freedom, they must succeed. If they . be slaves (no matter to whom,) they must fail. To prove their right, they rely on a deed of emancipation, executed by Pennington, their former master, dated the 25th of June, 1796. To defeat this evidence, the defendant produces an office copy of a deed executed by the same Pennington, and dated the 20th of November, 1774, purporting to sell to Martha Pennington, on certain terms and conditions, the slave Betty, &c. from whom it was agreed that the plaintiffs are descendants. To the admissibility of this copy, the plaintiffs objected on two grounds, onlv *541one of which is worthy of consideration; that is, because the copy could not he read, unless the loss of the original was established. To remove this objection, the defendant made oath that he had searched the Clerk’s office, and all other places where he supposed the original deed might probably he found, and had not been able to find it. The Court over-ruled the objection, and permitted the copy to go to the jury. Was this correct? The general rule is, that the best evidence must be given, of which the nature of the thing is capable. The deed here is the best evidence. Hut, it is said to be lost. If so, the law is not so unreasonable as to require its production, but is satisfied with the next best evidence, a copy. To lay a ground, however, for the introduction of this secondary evidence, it must he shewn with reasonable certainty, that the original is lost. Were the Court correct in receiving the defendant in this case, to furnish by his own oath, evidence of the loss of tiie original; and if so, was the evidence sufficient to lot iu the copy?

As to the first, I was strong-ly inclined to think, when I commenced the examination of this subject, that the de fondant ought not to have been suffered to give evidence. It seemed to me to innovate seriously upon that fundamental rule, that no body shall testify in his own cause; and the innovation, I feared, would endanger the purity of evidence. Further examination has shewn me, that the practice is against my first impressions; and though I do not cite the following cases as binding authority, I am disposed to yield to their weight. I had not adverted to the well settled distinction between evidence offered to the Court, upon a collateral point, not for the consideration of the jury, and evidence in chief. In the latter case, no interested witness can be heard. In the former, the parties themselves are often examined by the Court. Forbes v. Wate, 1 Black. Rep. 532. The issue was non est factum in a suit on a bond. It became necessary to prove, that the subscribing witnesses were dead; and the plaintiff him*542self was examined as a witness to that point, and aspreparatory to the proof of the hand-writing. In Jordan v. Cooper, Sergeant and Rawle’s Rep. 504, the question was, whether the Court below had done right in admitting a party to prove by his own oath, notice to the othei’ party to produce a certain deed. Chief Justice Tin gum aw, considered it the settled practice, in collateral matters of this kind, to admit the evidence of the party. He cites several cases from Yates, to the same point. See also Douglas’s lessee v. Saunderson, 2 Dall. 116, citing 1 Black. Rep. 532; Godb. 193, 326. In Butler v. Warren, 11 Johns. Rep. 57, the Court considered the admission of an interested witness, to prove service of a notice on the defendant to produce a paper on the trial, preparatory to giving evidence of the contents of the paper, as an infraction of the rule of law, which precludes the admission of an interested witness to give evidence on the trial. But in Jackson, &c. v. Frier, 16 Johnson’s Rep. 193, the subject is again considered. Chief Justice Spencek, delivers the opinion of the Court, and in a clear and sensible manner, states the reasoning on the subject, cites all the authorities, overrules the decision in Butler v. Warren; and decides that a party may be examined by the Court, to lay a ground for the introduction to the jury of secondary evidence. The evidence of the loss of a deed (he says) is addressed to the Coui’t alone, and it is not a subject on which the jury are to pass. See also Givens, &c. v. Manns, 6 Munf. 201, where the Court say that a party, by affidavit, may prove the loss of a bill of sale. I conclude that the Court) in admitting the defendant to give evidence as to this collateral matter, did not err.

Was the evidence of loss sufficient to let in the copy? The defendant swore, that he had searched the office, and every other place where the deed would probably be, and had not been able to find it. He also proved by the Clerk of the Court, where it had been recorded, that he had examined his office and could not find it. I think this was *543sufficient to let in the secondary evidence; for, the “ law exacts nothing unreasonable in such a case. If the parol proof of loss establishes the fact with reasonable certainty, it is sufficient.” 8 East, 289; 10 Johns. Rep. 374; 16 Johns. Rep. 193.

We must next enquire, whether the secondary evidence offered in this case, was sufficient? It is an office-copy, without other proof of the execution of the original deed, than its being recorded upon the acknowledgment of the grantor. If this were a deed, required by law to be recorded, there would be no difficulty in the question; copies of such deeds being every day admitted, without other evidence, than their having been recorded. But, from the best examination I have been able to give the subject, I cannot find any act of Assembly, directing that a bill of sale of slaves, whether taking effect in preesenti or in futuro, shall be recorded. The act of 1758 does not touch the subject, as it relates to deeds of gift of slaves; and this is a deed of bargain and sale. The statute of frauds and perjurios does not apply, because it speaks of deeds on consideration not deemed valuable in law; this is a deed for valuable consideration. The act for regulating conveyances, in the 4th section, speaks of settlements; and this may be called a settlement; but, certainly it is not a marriage settlement, of which alone that section speaks. The question is, therefore, free from the influence of our recording laws. In Lee v. Tapscott, 2 Wash. 281, and Rowlettsv. Daniel, 4 Munf. 473, it is said, that the copy of an ancient deed, where possession has followed the deed, and, (in the latter ease,) where the deed was recorded on the acknowledgment of the grantor, shall be received, without any proof, that the original is lost or destroyed. I do not consider these cases as in point; for, though Ours is an ancient deed, it wants that corroboration arising from possession. In Stip v. Turner, 1 Wash. 322, a deed had been admitted to record, on the certificate of two persons, styling themselves Justices of the Peace *544for South' Carolina; but, the Governor’s testimonial was wanting. On a trial in ejectment, the'plaintiif relied on . this deed, and produced a witness to prove the execution The Court below rejected the witness and the deed. This Court said, the deed was neither legally proved, nor legally recorded; and, as a recorded deed, would have been properly rejected; but, that such deeds were valid between the parties; that the actual execution of the deed, was a fact which the plaintiff was at liberty to prove, as in other cases, by evidence satisfactory to the jury.

Applying this law, it would seem, that if the party be at liberty to prove the actual execution of the deed, as he does other facts, the acknowledgment of the grantor in a Court of record, might be considered pretty strong evidence of such execution. The case of Maxwell v. Light, 1 Call, 117, may also be considered as having some bearing on the present; the Court deciding there, that if the deed of lease was admitted to record at the instance of the appellant, a copy might, under the circumstances, be received as evidence. The English books give us much more light upon this subject. Gilbert, in his Law of Evidence, 86, says, that where a deed needs enrolment (as deeds of bargain and sale by 27th H. 8, ch. 16,) there the enrolment is the sign of the lawful execution of such deed, and a copy shall be a sufficient attestation; but, where a deed needs no enrolment, the inspeximus of such enrolment is not evidence. Buller, in his Nisi Prius, commenting on this passage, doubts whether deeds of bargain and sale enrolled, ought to be given in evidence without being proved; though, he admits that such was'the practice at Nisi Prius. He adds “the case of Smartle v. Williams, is much relied on in support of this practice; but, that case is wrong reported; for, it appears from the report in 3 Lev. 347, that the acknowledgment was by the bargainor, and so is stated in iSalleeld’s manuscript. Besides, it appears from both the books, that it was only a term that passed, and, consequently, it *545was not an enrolment within the statute.” After some furiher remarks to shew that the bare enrolment ought not, in all eases, to make the deed evidence, Justice Ltnvuci! adds, “ on the other hand it seems as absurd to say, that a release which has been enrolled, upon the acknowledgment of the re-Ioasor, shall not be admitted in evidence against hiro, without being proved to he executed, because such release does not need enrolment: and, in fact, such deeds have often been admitted, and that was the case of Smartle v. Williams. The deed there, did not need enrolment, yet being enrolled, on the acknowledgment of the bargainor, it was read against him without being proved.” In 1 Salk. 280, is the case of Smartle v. Williams, referred to by Fuller. A deed of bargain and sale, acknowledged by the bargainor and enrolled, (by which a term for years was assigned) was given in evidence, without any proof of the bargainor’s sealing and delivery thereof; and, after debate, it was allowed by C. J. Holt, Eyre and tot. cur.; for, the acknowledgment of a party in a Court of record, or before a master extraordinary in the country (as this case was) is good evidence of its being sealed and delivered; and, such an acknowledgment, estops a man from pleading non est factum.” In Lady Halcroft v. Smith, Freem. 259, a distinction was made between deeds of bargain and sale (enrolled under stat. H„ 8,) and other deeds enrolled; and, it was held that a copy of a deed enrolled for safe custody, would not be evidence, otherwise than against the party who sealed it, and all claiming under him. Phillips, 410, after citing the cases, -concludes thus: (i The rule concerning copies of enrolments appears then to he, that a copy of the enrolment of a bargain and sale of freehold in lauds, &«. is as good evidence as the original itself; but, that a copy of the enrolment is not evidence of a, bargain and sale of a chattel interest, or of the contents of any other deed enrolled for safe custody, except as against the party acknowledging the deed; and, that against such party, and against all claiming under him, *546a copy of the enrolment of any deed is admissible in evidence. ” From these authorities, I think it may be safely concluded, as a general proposition, that the office copy of a deed recorded on the acknowledgment of the bargainor, (though there be no law requiring such record) is admissible evidence against the barg'ainoi’, and those claiming under him. This, I say, may be admitted as a general proposition; but, it still remains to enquire, whether the particular circumstances of the case before us, do not take it out of the rule.

The deed of bargain and sale was executed in 1774, In whose possession it remained afterwards, we are not informed;' nor does the record give us any information, whether the defendant who produced it is in any way connected with it, or claims under it at all; or whether any other persons set up claims under it. It is produced merely to shew, that at the time the deed of emancipation was executed, the grantor had no title. That deed, (under which the plaintiffs claim their freedom) was executed in June, 1796, and duly recorded in the August following. The deed of 1774, is recorded, on the acknowledgment of the grantor, in 1812, thirty-eight years after its execution, and sixteen years after the execution of the deed of emancipation. And the question is, must this acknowledgment be taken as proof of the execution of the deed, against the plaintiffs ? I strongly incline to think not. It is clear, that after the execution of the deed of emancipation, the grantor could not, by any act or deed of his, revoke that deed, or divest the title to freedom, which it had vested in the plaintiffs; and yet, nothing will be easier than to effect this indirectly, if his subsequent acknowledgments are admissible against the plaintiffs. Suppose the grantor, after emancipating his slaves, to change his mind, and repent of the act. He has only to execute to a purchaser a hill of sale for them, take no witnesses to the transaction, give the bill of sale a date prior to the deed of emancipation, and have it recorded on his acknowledgment. In the case *547before us, there is not a tittle of evidence of the existence or execution of the deed, but the acknowledgment of the grantor, made 16 years after the execution of the deed of emancipation; and, although the rule of law ho, that that acknowledgment is to be taken as evidence against him, and those claiming under him, I believe we must limit the meaning to those claiming under him, by title derived subsequently to the acknowledgment. This may be illustrated by analogy. It is laid down in the books, that the recital of a deed in another deed is evidence against the party who executed the reciting deed, or against any person claiming under him. 1 Phill. Evid. 411; yet, when this rule comes to be applied, it is restricted to those claiming under the grantor, by title subsequent to the deed. In Penrose v. Griffith, 4 Binney, 230, this point came directly before, the Supreme Court of Pennsylvania. C. J. Tn.oirvj said: “The rule of law is, that a deed containing a recital of another deed, is evidence of the recited deed, against the grantor, and all persons claiming by title derived from him subsequently. What is the reason of this rule ? It is this: the recital amounts to a confession of the party; and that confession is evidence against himself, and those who stand in his place. But, such confession can ho no evidence against a stranger; it can be no evidence against one, who claims by title derived from the person making the confession, before the confession made; because ho docs not stand in the place of the person making the confession; he claims paramount the confession. One who has conveyed his right, can, by no subsequent confession, affect; the right which he has conveyed.” The rule thus laid down, is admitted by the rest of the Court. In the same book, Garwood v. Dennis, p. 327, the rule is again stated with the same restriction. In 2 Bay’s Rep. 93, it is decided, that t"v ••1,1 of a bond is an incompetent witness to prove payment, so as to destroy the right of the assignee to recover. In Frear v. Evertson, 80 Johns. Rep. 142, the question was, whether *548the admissions of a party after he had assigned his interest to another, could be given in evidence. The Court say, “that, having assigned his interest in the chose in action, he could not impair that interest by any confessions made by him, to the prejudice of his assignee.'” The same principle may be found in several of our own Reports. In Vaughan’s adm’r. v. Winkle’s ex’r., 4 Munf. 136, in trespass for goods taken away, proof by witnesses that the person of whom the plaintiff bought the goods, was heard to say, before the institution of the suit, that when he sold them they belonged to the defendant, was decided to be inadmissible. In Givens, &c. v. Mann, 6 Munf. 191, a suit for freedom, the plaintiffs claimed under a deed of emancipation from Thomas Reynolds-, the defendants were purchasers of the paupers, as slaves, from John Reynolds, the son of Thomas. The Court decided, that the admissions of Thomas, with respect to John’s title and possession, so far as the}’ were made prior to the deed of emancipation, might be received; but that no subsequent acknowledgments were admissible. These authorities well warrant the conclusion, I think, that in the case before us, the acknowledgment of Pennington, of the execution of the deed, is no evidence of that fact, as against the plaintiffs; and, consequently, that the copy of the deed, resting solely on that acknowledgment, ought not to have been admitted by the Court below.

I am of opinion, therefore, that the judgment be reversed, and the cause sent back, with directions to the Court, not to receive the copy, unless the execution of the deed be proved by sufficient evidence, exclusive of the acknowledgment of the grantor.

The' other Judges concurred; and the judgment was reversed, and the cause remanded. *