COLLIER, C. J.
1. The testimony of the witness Elisor, was properly excluded. He declared that he was unacquainted with the hand-writing of the plaintiff,-and had no other information upon the subject than was communicated to him by the defendant’s testator. This was altogether insufficient to establish the genuineness of a writing, which could only be admitted upon proof that it was subscribed by the party, who purported to have made it, or by his autho*462rity. The admission of its genuineness cannot be predicated of the advertisement by the witness, that he held a set-off against the note which he had previously made to the plaintiff. It does not appear, that the plaintiff ever read, or heard of that advertisement, and if he did, it would not have informed him, what was the character of the proposed set-off. Nor is the fact shown by the payment of the witness’s note to the plaintiff, by the testator. The testator was indebted to the witness, and transferred to him the note supposed to have been made by the plaintiff, that he might, by an exchange of paper, extinguish his indebtedness to the plaintiff; afterwards, the witness returned to the testator the note he had received from him, that he might obtain the note the plaintiff held against the witness, which the testator promised to do; and in a few days after the advertisement appeared, the testator brought to the witness, the note which the latter had given to the plaintiff. It may be, that the testator took up the witness’s note by the payment of money, or something else which the plaintiff esteemed an equivalent; but it would be a most unwarrantable presumption, in the absence of all legal proof that the testator held a note against the plaintiff, to infer that the payment was made by an exchange of paper.
2. Where the knowledge of the hand-writing of a party is acquired by having seen him write, the usual inquiry of the witness is, whether he has seen the party write, and after-wards, whether he believes the paper in dispute to be his hand-writing. The course of examination involves two questions; first, whether the supposed writer is the person of whom the witness .speaks; secondly, if he is the person, whether he wrote the paper in dispute. The first is a question of identity; the second, a question of judgment, or a comparison in the mind of the witness, between the general standard, and the writing produced. This kind of evidence, it is said, like probable evidence, admits of every possible degree, from the lowest presumption, to the highest moral evidence. It may be so weak, as to be unsafe to act upon ; •or so strong, as in the mind of every reasonable man, to produce conviction. But, whatever degree of weight his testimony may deserve, which is a question exclusively for the *463jury, it is an. established rule, that, if he has seen the person write, he will be competent to speak to his hand-writing; and this, although the impression on the mind of the witness may be faint and inaccurate. 1 Phil. Ev. 484. Mr. Green-leaf says, it is sufficient to let in the witness’s belief, that he has seen the party write but once, and then only his name. “ The proof in such case, may be very light; but the jury will be permitted to weigh it. 1 Greenl. Ev. § 577. In Carter v. Connell, 1 Whart. Rep. 392, the question proposed was, “are you able to say whether, to the best of your impression, this paper is in the hand-writing of'J C (the party) or not ? ” The court said, to inquire of a witness what is his impression, is descending to a test' too vague to form a judgment upon — it is like asking, what was his understanding of a conversation, instead of inquiring what the parties said., The learned annotators upon Phillips on Evidence, remarking upon this decision, say : “ Perhaps, under the circumstances, particularly as an objection was raised in due time, and the question not modified, the inquiry was to be understood as calling for the mere conjecture of the witness, without reference to any recollection of his,’in respect to the character of the party’s hand-writing. If the usual inquiry as to belief was purposely avoided with this view, the court were doubtless right in overruling the question put. But it is difficult to maintain, consistent with the cases, that a witness shall not testify, unless he will express a decided belief in respect to the hand-writing. If the witness’s knowledge appears to have been derived from proper sources, its degree respects the credence to be awarded to what he says, rather than its competency.” 3 Phil. Ev. C. & H’s Notes, 1321, et seq. See further, Garrells v. Alexander, 4 Esp. Rep. 37; Stranger v. Searle, 1 Esp. Rep. 14; Doe ex dem. Mudd v. Suckermore, 5 Adol. & Ellis’s Rep. 703; Utica Ins. Co. v. Badger, 3 Wend. Rep. 102; Redford’s adm’r v. Peggy, 6 Rand. Rep. 316; Jackson ex dem. Van Dusen v. Van Dusen, 5 Johns. Rep. 144; Freelove v. Fenner, 2 Gall. Rep. 170; 2 Stark. Ev. 512, 7th Am. ed.
In Riggs v. Tayloe, 9 Wheat. Rep. 486, in order to let in secondary evidence of an agreement alledged to be lost, a party made an affidavit, stating “ that his impression is, that *464he tore up the same:” Further, “he is not certain that he did tear it up, and does not recollect doing so, but such is his impression. If he did not tear it up, it has become lost or mislaid,” &c. It was insisted at the bar, that the affidavit was not sufficiently certain and positive — it stating the impression of the affiant, that he tore up the agreement, or the alternative, that it is lost or mislaid. The court said, “ we do not concur in this reasoning. An impression, is an image, fixed in the mind, it is belief; and believing the paper in question was destroyed, has been deemed sufficient to let in the secondary evidence ;” whether the writing was tom up, or lost or mislaid without being actually destroyed, the affidavit was sufficiently certain and positive. See also Doe ex dem. Mudd v. Suckermore, 5 Adol. & Ellis’s Rep. 703. In the last case, the judges who presided, delivered their opinions seriatim. Coleridge, J., said, “the rule as to the proof of hand-writing, where the witness has not seen the party write the document in question, may be stated generally thus : Either the witness has seen the party write on some former occasion, or he has corresponded with him, or transactions have taken place between them, upon the faith that letters purporting to have been written or signed by him, have been so written or signed. On either supposition, the witness is supposed to have received into his mind an impression, not so much of the manner in which the writer has formed the letters in the particular instances, as of the general character of his hand-writing ; and he is called on to speak as to the writing in question, by a reference to the standard so formed in his mind. It is obvious that the weight of this evidence may vary in every conceivable degree; but the principle appears to be sound, both in regard to the test of genuineness, and the acquisition of the means of applying it.” Patteson, J., remarked, that “all evidence of hand-writing, except where the witness sees the document written, is, in its nature, comparison. It is the belief which a witness entertains, upon comparing the writing in question with an exemplar in his mind, derived from some previous knowledge. That knowledge may have been acquired, either by seeing the party write, in which case it will be stronger, or weaker, according to the number of times, and the periods, and other *465circumstances under which the witness has seen the party write, but it will be sufficient knowledge to admit the evidence of the witness, (however little weight may be attached to it in such cases,) even if he has seen him write but once, and then merely signing his name; or the knowledge may have been acquired by the witness having letters, &c. of the party,” &c.
The witness, Brothers, who was introddced for the purpose of proving the genuineness of the receipt purporting to have been made by the plaintiff to the,-' testator, answered, that he had seen the plaintiff write once, but he did not know his hand-writing. The counsel and the court both informed the witness, that he was not required'to swear positively as to the writing, but if, from having seén the plaintiff write once, or oftener, he believed he was acquainted with his hand-writing, or would recognize it, then he was competent, and bound to give his opinion, or belief, whether the receipt in question was the writing of the plaintiff. The witness still insisted, that he was unacquainted with the plaintiff’s hand-writing, “ and did not believe that he could recognize or know it, if hie saw it.” Here the witness, after being distinctly informed of the character of the examination, and the latitude allowed to his answers, declared his inability to recognize the genuineness of the paper, even according to belief or opinion, and the court was not bound to permit the question to be propounded in a form merely varying in phraseology. If the witness seemed-to misapprehend the scope of the inquiry, then it would have been proper, and even the duty of the court to have permitted the examination to be prosecuted in a more intelligible form. But it was doubtless supposed, that no such necessity existed for protracting it, and that the witness had fairly shown himself incompetent. Assuming this hypothesis to be correct, as we must do upon the facts stated in the bill of exceptions, and the rejection of the questions proposed by the plaintiff’s counsel, was the exercise of a discretion, which this court cannot revsie. The general rules for the examination of witnesses, must necessarily be so often applied, or relaxed according to circumstances, apparent to no one except the *466presiding judge, that a strict uniformity is not at all times to be expected, and in some instances, would prove injurious to the interests of justice. Much, therefore, is confided to the discretion of the judge of the primary court, which, though it should not be exercised by an arbitrary strictness on the one hand, or arbitrary indulgence or relaxation on the other, should always be applied with a due regard to the rights of parties. 1 Phil. Ev. 268.
In respect to the frame of the questions, we cannot perceive that they were objectionable, the witness having first stated, that he had seen the plaintiff write; they seem to have been drawn with studied precision, with the view of informing the witness of the nature of the impression, belief or opinion, he was asked to express, and considering how difficult it is sometimes, to make a very conscientious witness understand the line between knowledge and belief, we think they are not only unexceptionable in point of form, but are in terms most appropriate. It is a mistake to suppose they call for mere impressions, too indistinct to rest a conclusion on; true, they ask the witness to state impressions, yet these impressions must, in the nature of the subject, amount to belief or opinion. This proposition seems to us too plain to require illustration. But the defendant had previously asked equivalent questions — both his counsel and the court informing the witness of the character of the examination, and manner in which the answer might be made, and we cannot say that the refusal to permit a repetition of the inquiry, is an available error. Its allowance, or rejection, was a matter within the discretion of the circuit court, depending upon considerations of which it might be difficult to inform .us by the record, and which, if informed, we perhaps could not control.
3. The jury are generally the appropriate arbiters of the quantity and measure of proof, or rather of the effect of evidence. It is for the court to intercept such testimony as would tend to prejudice or mislead, beyond this, the law confides in the integrity and good sense of the jury. “ In the ordinary affairs of life,” says Mr. Greenleaf, “ we do not require demonstrative evidence, because it is not consistent with the nature of the subject, and to insist upon it, would be unreasonable and absurd. The most that can be affirmed *467of such things, is, that there is no reasonable doubt concerning them. The true question, therefore, in trials of fact, is not, whether it is possible the testimony may be false, but whether there is a sufficient probability of its truth; that is, whether the facts are shown by competent and satisfactory evidence. Things established by competent and satisfactory evidence, are said to be proved.” Again: “None but mathematical truth is susceptible of that high degree of evidence, called demonstration, which excludes all possibility of error, and which, therefore, may reasonably be required in support of every mathematical deduction. Matters of fact are proved by moral evidence alone; by which is meant, not only that kind of evidence which is employed on subjects connected with moral conduct, but all the evidence which is not obtained from intuition, or from demonstration.” Greenl. Ev. § 1, 2d ed. “ By satisfactory evidence, which is sometimes called sufficient evidence, is ihtended that amount of proof, which ordinarily satisfies an unprejudiced mind, beyond reasonable doubt. The circumstances which will amount to this degree of proof, can never be previously defined ; the only legal test of which they are susceptible, is their sufficiency to satisfy the mind and conscience of a common man; and so to convince him that he would venture to act upon the conviction, in matters of the highest concern and importance to his own interest.” Id. § 2. Further: “ Circumstantial evidence is of two kinds, namely, certain, or that from which the conclusion in question necessarily follows; and uncertain, or that from which the conclusion does not necessarily follow, but is probable only, and is obtained by a process of reasoning.” “ In civil cases, where the mischief of an erroneous conclusion is not deemed remediless, it is not necessary that the mind's of the jurors be freed from all doubt; it is their duty to decide in favor of the party, on whose side the weight of evidence preponderates, and according to the reasonable probability of truth. But in criminal cases, because of the serious and irreparable nature of the consequences of a wrong decision, the jurors are required to be satisfied, beyond any reasonable doubt, of the guilt of the accused, or it is their duty to acquit him; the charge not being proved by that higher degree of evidence *468which the law demands. In civil cases, it is sufficient, if the evidence, on the whole, agrees with and supports the hypothesis, which it is adduced to prove; but in criminal cases, it must exclude every other hypothesis, but that of the guilt of the party.” Id. § 13, (a).
In Ross v. Gould, 5 Greenl. Rep. 204, upon the trial of a writ of entry, the genuineness of a deed for the premises demanded, adduced by the tenant, was contested by the demandant, and evidence was offered by both parties, for and against it. Whereupon, the judge instructed the jury, that the tenant having introduced the evidence purporting to be a deed, and claiming under it, it was not the duty of the demandant to prove it to be a forgery; but it was the duty of the tenant to give them reasonable satisfaction that the deed was genuine. And if he had failed thus to satisfy them upon this point, they would return their verdict for the demand-ant. In the argument in the supreme court, it was insisted, that as the tenant proved the deed by one witness, and it was read to the jury as the deed of the demandant, the onus was cast on the demandant to show it was a forgery; and the evidence offered afterwards to this point being contradictory, the jury should have been instructed that it was the duty of the tenant to prove it a genuine deed. The court said, t! in the examination of the contested fact, the onus probandi may, in the course of a trial, be thrown from one party upon the other several times, according as the complexion of the proof may change. But when it is said, as was stated by the judge at the trial, that the onus probandi is on the party who offers a paper as a genuine deed, under which he claims, the plain, sound, common sense and legal meaning is, that it has reference to all the evidence in the cause respecting the alledged genuineness of the contested paper; or, in other words, it means that the party affirming the paper or instrument to be genuine, must furnish to the jury so much evidence, as to leave a balance of proof in favor of the genuineness of the instrument, after making all due allowance for the influence of the proof adduced on the other side to produce a different conviction in the minds of the jury.”
In Brooks et al. v. Barrett et al. 7 Pick. Rep. 94, the court said, the “shifting of the onus probandi is quite familiar in *469the course of trials. Iu a suit upon a written contract, on the general issue, the plaintiff produces his note, proves the defendant’s signature, and stops; the defendant then alledges payment, want of consideration, or other matter in defence, and the burden of proof is upon him.”
In the Inhabitants of Attleborough v. the Inhabitants of Middleborough, 10 Pick. Rep. 378, the jury were instructed, that as the amount of the highway tax was not carried to the succeeding year, the presumption was, that it was paid ; but that this presumption might be repelled by evidence ,• that they would consider the evidence in relation to this point, and if the defendants had failed,- upon the whole, to satisfy them, beyond a reasonable doubt, that the tax was paid, they must find for the plaintiffs. It was contended, in the supreme court, that the instruction ought to have been, that the presumption was conclusive ; but the court said, “there was nothing in the statute, and certainly nothing in the circumstances of the case, to render this presumption conclusive. It was evidence, to be weighed and considered by the jury, and they might well consider it as fully counterbalanced by the express negative testimony of, the surveyor. The jury were also rightly instructed, that the burden of proof was on the defendants, and that if they doubted of the truth of the fact relied on, to wit, the payment of the tax, the verdict must be for the plaintiffs.” In Crail v. Crail, 6 Penn. State Rep. 480, it was adjudged not to be error to instruct the jury, that they should require clear and satisfactory evidence, to find a tenancy from parol declarations, when the alledged tenant had a legal title from the commonwealth, and it appeared an ancient agreement, for a sale by him to the landlord had been abandoned.
We may pass over the first charge, and direct our attention to the second; for if the latter can be supported, notwithstanding the stringency of the terms in which it is expressed, it must be conceded, that no objection will lie to the former. In the second charge, the courHinstructed the jury, “ that if the evidence, although it preponderated in favor of the receipt being genuine, still, if upon a fair and full examination of it, their minds were left in a state of doubt and uncertainty as to its being genuine, they should find in favor of the *470plaintiff upon that point.” This instruction is in very broad and general terms, and does not particularize the nature of the “doubt and uncertainty” which should occupy the minds of the jury, so as to induce them to withhold their sanction from the preponderating evidence in favor of the genuineness of the receipt. The fair inference is, that any degree of “ doubt and uncertainty,” was sufficient to produce that effect, although it was not the result of the application of the ordinary tests by which truth, or strong and overbalancing probabilities, are evolved from controverted facts. If there was sufficient probability, upon an examination of all the testimony to the point, to convince the judgment, and induce an unprejudiced mind to affirm the genuineness of the receipt, this was enough to incline the scale in favor of the defendant; yet the jury were not left free to act upon evidence thus potent, unless they felt certain beyond doubt, that the conclusion to which they were led, was founded in truth. They were authorized to infer, that demonstrative evidence was required, and that moral evidence was insufficient. A rule thus stringent would impair, to a great extent, the value of circumstantial evidence — frequently the best that can be obtained, and often the most satisfactory and convincing. It is directly opposed to what we have seen is said by Mr. Greenleaf, whom we all agree is a text writer of great exactness and accuracy. He expressly affirms, that in civil cases, it is not necessary that the minds of the jurors should be freed from all doubt; but they should decide according to the reasonable probability of truth, and in favor of the party on whose side the weight of evidence preponderates.
The citation from 7th Pickering lays down no proposition in respect to the quantum of evidence, and furnishes no aid in the explication of the question before us. In the case in 10th Pickering, no objection appears to have been taken to the terms in which the charge to the jury was expressed, and the revising court did not consider that point — the only inquiry was, whether a presumption from a given fact was conclusive, or could be countervailed by evidence. The decisión, then, is worth but little as an authority, for any other purpose than to show what was the effect of the evidence be*471fore the court. In the case from 6th Penn. State Reports, the court certainly were correct in saying it would require “ clear and satisfactory evidence ” to establish a tenancy in favor of the plaintiff in the absence of a writing, where the alledged tenant had a legal title from the State, which he had conveyed many years previously to the defendant — and when, too, it was shown that an ancient contract for a sale to the ancestor of the plaintiff (the assumed landlord), had been abandoned. The citation from 5th Greenleaf, intimates in no equivocal terms, that a balance of evidence in favor of one party, after giving all due credits to the other, is usually sufficient to authorize the jury- to find a verdict in a civil case. The view we have taken, instead of being opposed, is really supported by Starkie. This learned author says, that in criminal cases, a mere preponderance of proof will not suffice —all reasonable doubt must be excluded; but in cases of a civil nature, where no presumption, of law or prima facie right operates in favor of either party, a mere preponderance of evidence on either side, may be sufficient to turn the scale. True, a mere preponderance, where it falls short of fully disproving a legal right once admitted or established, or of rebutting a presumption of law, is insufficient. “ One who seeks to charge another with a debt, must do so by full and satisfactory proof; and on the other hand, where a debt has once been established by competent proof, the debtor cannot discharge himself but by full proof of satisfaction. Again, where the law raises a presumption in favor of the fact, the contrary must be fully proved, or at least such facts must be proved, as are sufficient to raise a contrary and stronger presumption.” 1 Stark. Ev. 543, 7th Am. ed. These are legal propositions, well founded in reason. It must be admitted, that when the plaintiff showed a grima facie right to recover, it devolved upon the defendant to establish a payment or discharge, “ by full and satisfactory evidencebut the second charge required more than this — it supposed that the defendant should discharge himself by evidence so conclusive, as to exclude “ doubt and uncertainty ” from the minds of the jury. This requisition, we have seen, was too exacting, and cannot be supported. The judment of the circuit court is consequently reversed, and the cause remanded.