delivered the opinion of the Court:
i. The first assignment of error is founded on the exceptions taken to the offer in evidence by the defendants of the declarations of Hollingshead and Luby, the subscribing witnesses to the bond. As shown in the preceding statement, these witnesses were both dead, and the bond was admitted in evidence on proof of their signatures, as well as that of the principal, who was also deceased.
The question is, can the declarations of a deceased attesting witness to an instrument be admitted to show, in dis- • paragcment of the evidence of its execution afforded by his signature, that the signature so attested was forged?
The admissibility of such evidence was long a matter of doubt in England, but was finally ruled against in a well-considered case that has not since been questioned in that country. Stobart v. Dryden, 1 M. & W. 615. The reasoning of Baron Parke, who delivered the opinion in that case, has met with the unqualified approval of Mr. Greenleaf, and has been admirably summarized in a section of his great treatise on the Law of Evidence, which we quote as follows :
“An exception to this rule has been contended for in the admission of the declaration of a deceased attesting witness to a deed or will in disparagement of the evidence afforded by his signature. This exception has been asserted on two grounds : First, that as the party offering the deed used the declaration of the witness, evidenced by his signature, to prove the execution, the other party might well be permitted to use any other declaration of the same witness to disprove it; and, secondly, that such declaration was in the nature of a substitute for the loss of the benefit of a cross-examination of the attesting witness, by which either the fact confessed would have been proved, or the witness might have been contradicted and his credit impeached. Both these grounds were fully considered in a case in the Exchequer, and were overruled by the court; the first, because the evidence of the handwriting in the attestation is not used *447as a declaration by the witness, but is offered merely to show the fact that he put his name there in the manner in which attestations are usually placed to genuine signatures ; and the second, chiefly because of the mischiefs which would ensue if the general rule excluding hearsay was thus broken in upon. For the security of solemn instruments would thereby become much impaired, and the rights of parties under them would be liable to be affected at remote periods by loose declarations of the attesting witnesses which could neither be explained nor contradicted by the testimony of the witnesses themselves. In admitting such declarations, too, there would be no reciprocity, for, though the party impeaching the instrument would thereby have an equivalent for the loss of his power of cross-examination of the living witness, the other party would have none for the loss of his power of re-examination.” 1 Greenleaf Ev., Sec. 126; see also 1 Wharton Ev., Sec. 731.
Counsel for appellees have made a forcible argument on behalf of the admissibility of this evidence, founded largely on the hardships that must result, in some cases, from the rigid enforcement of the rule of exclusion ; and they have the support of decisions by some courts of high authority in this country.
After careful examination of those cases, and after much consideration, our conclusion is that the rule of exclusion is supported by the strongest reason. The hardships that might result from the exclusion of the declarations are more than set off by the greater mischiefs that would follow the establishment of this exception to the general rule against the admissibility of hearsay evidence. Courts that have held to the contrary have been forced to admit the weakness and unreliability of such evidence, and some of them have said that without strong support it would not justify the rejection of an instrument.
2. Instead of instructing the jury, as was done without qualification, that “the Government is bound, in order to make out its case, to satisfy you by a preponderance of the *448evidence that this is Robert Boyd’s genuine signature,” the court should have told them that evidence of the death of the subscribing witnesses and proof of their signatures would make a prima facie case of the due execution of the •bond by the principal and all of the sureties. Clark v. Courtney, 5 Pet. 319, 344; Sicard v. Davis, 6 Pet. 124, 137; Stebbins v. Duncan, 108 U. S. 32, 44; Hanrick v. Patrick, 119 U. S. 156, 172.
3. The last point for consideration arises on the exception taken to thg^charge to the jury, to the effect that they must find for the defendant Slater also, in the event that they should find that the defendant Boyd did not execute the bond.
We think this, too, was error. By the execution of the bond a?id its return to the principal or his agent for delivery to the obligee,'Slater became estopped to set up any condition not known to that obligee, upon which his signature had been obtained. Dare v. United States, 16 Wall. 1. Nor can the forgery of Boyd’s signature, if it be established, have the effect to discharge Slater from liability on the bond. He should have inquired into that before signing his name. He trusted to the principal in the bond, for •whose accommodation he signed, and must suffer the consequence. of his fraud. He cannot shift the loss upon the' obligee who received the bond in reliance upon the genuineness pf all the signatures, and to whom the signature of Slater, following that of Boyd, might well be regarded as a representation of its genuineness. York Co. M F. Ins. Co. v. Brooks, 51 Me. 506; State v. Peffer, 31 Ind. 76; Helmes v. Wayne Agr. Soc., 73 Ind. 325; Selzer v. Brock, 3 Ohio St. 302, 308; State v. Baker, 64 Mo. 167; Stern v. The People, 102 Ill. 540, 553; Lombard v. Mayberry, 24 Neb. 674, 685.
For the reasons given, the judgment must be reversed, with costs to the appellants, and the cause remanded for new trial. It is so ordered.