concurred in the opinion of Moncure, J. except as' to the proof of the set off of 27 dollars 50 cents.

Allen, J. concurred with Moncure, J.

Judgment affirmed.

In my view of this case it is unnecessary to decide whether the Court below erred in permitting the plea of the act of limitations and list of offsets to be filed; being of opinion that it did not err in deciding the other questions which arose in the case.

1 think the Court did not err in permitting the replication to the plea of the act of limitations to be amended. The demurrer to the original replication was sustained, solely because it contained no proferí in curiam of the covenant therein mentioned. After the demurrer was sustained the replication was amended by the insertion of the proferí therein. This amendment cured a defect which was merely technical, produced no delay nor inconvenience, and was necessary to the ends of justice. Without stopping to enquire how far it would have been authorized by the English practice, it is sufficient to say that it was fully authorized by the practice and decisions of Virginia. Cooke v. Beale's ex'ors, 1 Wash. 313; Graham v. Graham, 4 Munf. 205.

Nor do I think the Court erred in overruling the demurrer to the amended replication. Two proposi*390tions are presented by the demurrer. First, that the promissory note on which the action was brought was merged in the covenant; and if not, secondly, that the action on the note accrued at its date, and the act of limitati°ns began then to run, and was not afterwards arrested or suspended by the covenant.

First. Was the promissory note merged in the covenant ? The doctrine of merger is a technical doctrine founded upon the presumed intention of the parties. “ A simple contract debt is merged in a bond or covenant taken for or to secure the claim, because in legal contemplation the specialty is an instrument of a higher nature, and affords a higher security and a better remedy than the original demand presented. But this does not hold, even in favour of a surety by simple contract, if it appear on the face of the subsequent deed that it was intended only as an additional or collateral security, and there is nothing in the deed itself expressly inconsistent with such intention.” Chitty on Con. 783. It plainly appears on the face of the covenant in this case that the note was not intended to be merged therein. It was deposited in the hands of Bowles to indemnify him against his responsibility as the bail of Elmore ; and was to be returned to the latter as soon as the former should be released from such responsibility. Its existence as an independent security was thus preserved; and though its animation was suspended during the continuance of such responsibility, it was to become again an active security when the responsibility should be determined. The covenant was not given for the payment of the debt, but for the return of the note. It was collateral to the note. The delivery of the note would be a discharge of the covenant, and a suit upon the covenant might produce merely nominal damages. The note therefore was not merged in the covenant.

*391Secondly. Was the action on the note barred by the act of limitations ? The note bears date on the 13th of June 1817, and is payable on demand. The covenant bears date on the 2d of October 1818. The plain object of the covenant was to suspend the right of action on the note until Bowles should be released from responsibility as the bail of Elmore, and then to restore such right of action. Was not that object legal ? Was it not competent for the parties to agree, on sufficient consideration, that the time for payment of the note should be postponed ; or that it should be payable on a future contingency ? Could not the debtor, for any consideration however valuable, by any contract however solemn, arrest or suspend the running of the act of limitations against a subsisting debt ? Where a new promise is made to pay a debt barred by the act of limitations, it has long been a vexed question whether the action should be brought on the old or new contract; or perhaps it would be more proper to say, whether or not it might be brought on the old ; for I suppose there never was a question but that it might be brought on the new contract. Two of the Judges of this Court contended in Butcher v. Hixton, and The Farmers Bank v. Clarke, 4 Leigh 519 and 603, that the action should be brought on the new contract; but the question was not decided in those cases, and has never been decided in Yirginia. That the old cause of action is revived, and may be sued upon, was recently decided by the Supreme court of Massachusetts in the case of Ilsley v. Jewet, 3 Metc. R. 439, which was ably argued, and in which all the authorities on the subject were reviewed. Our statute of 1838 regards the original cause of action as revived and brought down by the new promise, and our new Code expressly authorizes the creditor at his option to sue on the original cause of action or on the new promise. But whether, prior to the operation of the *392statute of 1838 and of the new Code, (which do not apply to this case,) the action in such cases should have been brought on the old or new contract, is a question which does not affect this case. Generally, ^ere is nothing in the new promise which shews an intention of the parties to set up the old contract; and an action on the new promise answers every purpose of an action on the old contract. Were such an intention plainly expressed in the new promise, there is no case which shews that the action could not accordingly be brought on the old contract; and I imagine there can be no doubt but that it could. In this case it is plainly expressed in the covenant that the note should be preserved, and set up as the cause of action when the responsibility of Bowles as bail should be ’determined. This too is a stronger case than that of a mere promise to pay a debt barred by the act of limitations. It is the case of a covenant made on valuable consideration by the maker of a note a few months after its date, to return it to the payee on a future contingency, in order that he might then have a right of action thereon. That the action in this case was rightly brought on the promissory note, and was not barred by the act of limitations, is, I think, clearly shewn by the case of Irving, &c. v. Veitch, 3 Meeson & Welsby, p. 90. This is not like the case of a creditor covenanting not to sue for a limited period on a note of the debtor remaining in the hands of the creditor; in which case it might be said that the institution of a suit upon the note during such period would merely be a breach of the covenant. In this case the note was surrendered by the creditor to the debtor, for the purpose of effectually preventing it from being used as a subsisting security until the debtor should be released from his responsibility as bail of the creditor. If the debtor should be so released, the note was to be returned to the creditor, and to become again an active *393security. If he should not be so released, but be subjected to loss as bail, the note would be at home, and the amount of it in the hands of the debtor for his indemnity. The effect then of the covenant, in my opinion, was to restore the right of action on the note on the 25th of February 1832, when Bowles was released from his responsibility as the bail of Elmore by the latter’s death; but there being no personal representative of Elmore until the 3d of August 1836, the act of limitations did not begin to run until that day. Hansford v. Elliot, 9 Leigh 79. The action having been instituted in September 1837, was of course not barred by the act.

I am also of opinion that the Court did not err in excluding the evidence mentioned in the two bills of exceptions taken on the trial. The defendant being unable, as he averred, to produce direct proof of the handwriting of Elmore, offered the evidence mentioned in the first of the said two bills of exceptions, to prove that the order therein mentioned was drawn by Elmore and paid by Bowles; but it was excluded by the Court. Was the evidence admissible for the purpose for which it was offered ? The question is not whether independent parol evidence of the payment of the money by Bowles at the request of Elmore would have been admissible. There is a class of cases, it is true, in which parol is primary evidence, notwithstanding the existence of written evidence of the same fact. When a receipt is given for a payment, the general opinion is that the payment may be proved as well by parol evidence of the fact as by the production and proof of the receipt, though the case of Hamlin's adm'r v. Atkinson, 6 Rand. 574, has thrown some doubt on that question in this State. But I doubt whether the case under consideration falls in that class. If Bowles claims to have paid this money to Burton upon Elmore’s written order, the order itself would *394seem to be the highest evidence of the fact of its having been given; and must be produced and proved as primary evidence, or its absence accounted for before secondary evidence would be admissible. There are reasons for the admissibility of parol, as primary evidence in the case of a receipt which do not apply to the case of an order; and I do not see why parol evidence is admissible in the case of an order, any more than in the case of any other written contract. However this may be, certainly the written order having been produced and relied on to prove the request, and po attempt having been made to prove a parol request, the order should have been proved precisely in the same way as if it had been the only admissible evidence of the request. It should have been proved that Elmore executed or acknowledged the order, or that the signature thereto was his handwriting. It may have been impossible for the defendant to have proved either of these facts. If so it was his misfortune or the consequence of his neglect. He should have supplied himself with legal evidence. As to the questions mentioned in the second bill of exceptions, they were, I think, improper, and with the answers thereto were rightly excluded by the Court. The first question was clearly inadmissible. The witness after stating in effect that he had no knowledge of the account or order, was asked “ if he had any doubt that the said account was due according to its face.” I think the other question, “ Whether he would have drawn the order without the direction of Elmore,” was also inadmissible. The answers to these questions would have been mere opinions of the witness ; and a very unsafe foundation for the verdict of a jury. Every man who brings a suit for a claim must prove it by legal evidence ; and no opinions, however numerous or well founded, that he would not assert an unjust claim, will be sufficient to warrant a verdict in his favour.

I think the judgment ought to be affirmed.

Upon the effect of the covenant between the parties in regard to the promissory note sued upon to prevent the action from being barred by the statute of limitations, and the other questions arising upon the pleadings, I concur in the opinion of Judge Moncure. .

But I think the Circuit court erred in excluding from the consideration of the jury the evidence offered by the defendant in the action to prove the following item in his account of set offs: “1819, July 1st. Cash paid Dr. Burton, 27 dollars 50 cents.” The defendant’s two bills of exception on this subject, one of which refers to the other, may be properly treated together, and as presenting substantially the question, whether the proposed evidence was admissible for the purpose of proving the payment of so much money by the defendant’s testator, for the plaintiff’s intestate at his instance and request. The defendant produced a paper consisting of several parts, unattested by any subscribing witness, and purporting to be, 1st. An account of Dr. Burton against Elmore, the plaintiff’s intestate, for medical services. 2dly. The order of Elmore endorsed thereon addressed to Bowles, the defendant’s testator, requesting him to pay the amount of the account to Dr. Burton. 3dly. The acceptance of Bowles at the foot of the order. And to authenticate the paper as genuine, the defendant offered the parol evidence mentioned in the bills of exception. But the Court excluded from the jury the whole evidence written and parol, on the ground it seems that there was no proof of the handwriting of Elmore’s signature to the order, nor of his admission that it was his signature.

In this decision the Court overlooked, I think, a distinction between writings attested by subscribing witnesses and those not so attested. Where an instrument is attested by a subscribing witness, the law requires him to be called to prove its execution; the parties *396having agreed that he shall be their witness of the fact J and the presumption being that he has a better knowledge than any other of the attending circumstances : and the omission to call him cannot' be supplied Pro°f anY acknowledgment or admission of the party against whom the instrument is adduced, or that the signature of the party is in his handwriting. If, however, the production of the subscribing witness has from his death or other circumstances become impracticable, the next best evidence is proof of the handwriting of his attesting signature, which covers the entire execution of the instrument, from the presumption that he would not have put his name there in authentication of what in truth did not occur.

The case of an instrument unattested by any subscribing witness stands upon a quite different footing. There the parties have not made any one the special depository of the fact of execution. The fact may be proved by any witnesses. And the evidence may be direct and positive or indirect and circumstantial. It may relate to what occurred at the time of the transaction, or to subsequent acknowledgments or admissions of the party, either by words or by circumstances of recognition or acquiescence. Or it may be of the handwriting of the party; but that rests upon no higher ground in regard to competency than any other evidence of the fact of execution, whether direct or circumstantial; if it did ■ then, as being of a higher class, it would exclude all evidence not only of subsequent acknowledgments or ■ admissions, but even of what occurred at the time of the transaction. And besides to require proof of the handwriting of the party, presupposes that the signature of his name was written with his own hand, whereas, in simple contracts, at least, if written by some other person at his request or by his authority, it would have been equally effectual.

*397In the present case it cannot, I think, be doubted that it would have been competent for a witness to prove that he presented the account upon which the order is endorsed to Elmore, and at his request, or by his permission, wrote the order, which was signed by Elmore with his own hand, or by some other person at his request; and that such proof of the execution of the order would be direct, positive and complete, without any evidence whatever of the handwriting of Elmore. Here the evidence of these facts was not direct and positive, but indirect and circumstantial; and no one supposes that the law makes any distinction in point of competency between these two kinds of evidence, or that the sufficiency of either is not matter for the consideration of the jury. And it is obvious that the evidence of a witness which would otherwise be positive, may be rendered circumstantial by his want of recollection of parts of the transaction, though he still remembers circumstances from which the inference is safe and satisfactory.

The circumstances appearing from the bills of exception are these: 1. The authentication of Dr. Burton's account upon which the order i? endorsed. His brother was examined as a witness, and deposed that the account was copied by him from Dr. Burton’s books by his direction. 2. The authentication of the body of the order from its being in the handwriting of the same witness. 3. That the witness would not have written the order without the request of Elmore, (which we are warranted in assuming from his being prevented by the Court from speaking to that point.) See 1 Stark. Ev. pt. 2, p. 332-3, n. and the cases there cited. 4. That the account with the order endorsed, was in the possession of but not retained by Elmore, but came to the possession of Burton or his agent, and was presented to Bowles for his acceptance, and was accepted by him at the foot of the order. 5. That the order was after-*398wards paid by Bowles. The two last circumstances are inferences from, but warranted by the evidence, as will be seen by inspecting it. The evidence moreover "warrants the inference, that the medical services charged in the account, were rendered by Burton, and that the account itself was presented to Elmore for payment by the agent of Burton.

I cannot doubt that if the evidence offered had been submitted to instead of being excluded from the jury, they might very properly have deemed it sufficient to sustain the set off claimed by the defendant. But whether they would or not, it seems to me that it was proper for their consideration, and I can perceive no principle, nor am I aware of any authority, upon which it ought to have been rejected. Evidence of the handwriting of the party is doubtless proper in such cases, (though in itself circumstantial,) and is perhaps most usual, because most convenient; and the failure to produce it when obtainable, may in a doubtful case be proper for the consideration of the jury,"though entitled to but little weight, for it is equally in the power of either party, and is in its nature of equal force when used negatively as when used affirmatively. But after the lapse, as in this case, of twenty-six years, the fair presumption is that such evidence has perished, if it ever existed; and if so, then the exclusion of other evidence whether direct or circumstantial, to prove the execution of the instrument, has to my mind no colour of reason.