delivered the opinion oí the Court. The de~ fence in this case arose under a notice given by the defendant, that in support of his plea of payment, he should offer in evidence a receipt dated in If 85. signed “ N. Beake” On the Trial the counsel for the defendant produced this receipt, and offered it to the jury, as the receipt and signature of N. Beake the elder, deceased; examined a witness to the hand writing, who proved all the requisites to establish it as the hand writing of Beake the elder. It was read to the jury as such, and they rested their evidence without mentioning Beake junior as the subscriber of the receipt; indeed the witness on his cross-examination, expressed his belief that it was not young Beake's hand writing.

After plaintiff had gone through his testimony and rested his cause, the counsel for the defendant stated, that in the opening they had endeavoured to prove the receipt to be the hand writing of the old man; but if they had failed to establish this fact, they conceived it competent on this issue, to prove it to be the signature of Beake the younger, the plaintiff in the present cause, as the agent or attorney for his father.

This attempt was opposed by the opposite counsel, and the court overruled the testimony; the question argued on the present motion is, whether we were not wrong in that opinion.

I have considered the case at different times since, have searched the books for further light upon the subject, and have consulted gentlemen of the bar of a neighbouring State: the result of my own experience at the bar for 30 years, my researches and inquiries furnish no case or precedent which bears the least upon the opinion delivered at the Trial. I still think the attempt was monstrous, — not in the counsel, who no doubt proposed it as in their opinion a legal defence,— but as a proposition of law, in the opinion of the court, leading to consequences immoral and injurious in the highest degree.

The only doubt I had at the Trial was, whether it might sot have been proper to admit the evidence, and to have *14charged the jury to disregard all the evidence in support of the receipt as contradictory, and calculated to mislead thenv j,y t|ie proof 0f two propositions, one of which must have been absolutely false within the knowledge of the party.

It is an acknowledged doctrine, that a party shall not be permitted to discredit a witness whom he has himself called. Adams v. Arnold, 12 Mod. 375. 1 Morgan’s Essays 441. Rapp v. Le Blanc, 1 Dall. 63. (a) By producing a witness in court, he must always be taken, as against the party producing him, to be credible; and a man ought not to be allowed after introducing testimony to prove his case in one shape, to turn about and endeavour to attain the same end by the production of evidence which proves the falsity ox the •first witness.

It is a rule in pleading, that on the trial a party cannot desert his issue: if a man plead non detinet, he shall not be allowed to give in evidence that he holds the goods in pawn; on non est factum, he cannot give in evidence a release: on a plea of nullum vas turn defendaixt would not be permitted to prove a release. These rules are established, not only to prevent surprise to the opposite party, but for the sake of certainty and consistency, to keep the party to his point in pleading. When on a trial, a party offers to prove he paid the money to the testator himself, and that he has a receipt signed by him, and produces witnesses to prove the signature to be his; there is as much, if not more reason in such a case, to disallow a departure from the allegations and proofs, than to go out of the plea: It is true, as has been argued, that this was not a desertion of the issue, — but it was deserting the defence set up under this issue, which went to the whole action, and attempting to prove that the same signature was the act of two distinct persons; — an utter impossibility, — and which necessarily involved in it the perjury of one or the other of the witnesses adduced by the party, and left the fact altogether questionable and unsettled.

*15There could perhaps be cases imagined, in which it might •e permitted the defendant to pursue such a course. An executor finding a receipt among the papers of his testator, ■md ignorant whether it bore the signature of the original obligee or his son, or agent, and so shaping his defence, might be permitted to establish it as the writing of one or the other, according to the preponderance of the testimony: but this case is widely different, and to be judged by different principles.

The case cited from Strange (a) is not applicable! it is besides short and confused. On the probate of a Will, it is certainly true, that the Executor or devisee may controvert the testimony of the subscribing witnesses, although they have been produced by him: but this is permitted them, because such witnesses derive no authority from being produced by the party on the ground of his trusting them. He is compelled to have recourse to those whose names he finds subscribed to the instrument, and if after having assisted at the solemn execution of the paper, they attempt to destroy its validity, there is no principle of law or reason, or policy, which is impugned by leaving him at perfect liberty to repel their testimony and to establish the writing. He did not invite their attestation, and they arc brought before the court, not as his witnesses, but from the mere necessity of the case. It is evident this case will not bear the plaintiff out, in the unprecedented latitude which he contends should be given to the principle.

This is the only case in the books in which a party is allowed to produce contradictory evidence, and it is one altogether anomalous in its nature, and confirms the idea adopted by the court on this occasion.

Upon the whole, the court adheres to the opinion first dc« livered on the trial, that the evidence offered and overruled^ ®ught not to have been admitted, and therefore for that cause, no new Trial should be granted. Yet we are of opinion *16for another reason this rule should be allowed. At the first (a) trial of this cause, the defendant was permitted without opposition to try his receipt in both ways. The objection there-f°re at the second trial, was a surprise on the defendant: if it had been made at first, he would have had his election and ceme prepared accordingly on the second trial. For this reason we think a new trial should be granted, which is no more than having the cause reconsidered by another jury. But this must be on terms. — Isi defendant must pay costs — %d he ■ ust confess judgment with release of errors, to avoid the •. ■ i rquences of his own death in the interim. 3d A trial to ■; • .-'•'d whether the parties die or not, so that the death of either shall not be made a ground of error. On these terms let a new trial be granted. (b)

Rule absolute on Conditions.