Atkinson, J.
1. Two of the grounds of the motion for new trial were as follows: "9th. The court, having correctly charged the jury as to the propounders’ contention, that the instrument purporting to be the last, will and testament of Mary L. Spencer is as a matter of fact her last will, and having correctly charged the contention of caveator that said instrument is not as a matter of fact her last will and testament, but on the contrary a forged paper and a forged will, proceeded to charge as follows: 'The propounders hold the affirmative, and the burden of proof is on them to prove that Mary L. Spencer herself signed the instrument which, is alleged to be her will, and that.it has been proven as the law requires! When this is done, the burden is shifted to the caveator to make good his contentions as set up in his caveat. Movant alleges that the language in the charge above quoted, 'when this is done, the burden is shifted to the caveator to make good his contentions as set up in his caveat,’ is error, for the reason that the parties had arrived at an issue and proposition, as to the paper being the will of Mary L. Spencer, — whether she signed it or not; this proposition was affirmed by propounders and denied by caveator, and on the issue thus made both sides had introduced evidence; and the jury should not have been instructed that when propounders prove that Mary L. Spencer signed the instrument which is alleged to be her will and that it has been proven as the law requires, then the burden is shifted to caveator and it devolved on *127him to make good his contentions that the paper was a forgery. Movant says that the burden in a ease like this never shifts, and it is not obligatory on the jury to divide the evidence into two parts, and determine their belief provisionally by one of the parts only, but they may consider the evidence for both sides as a whole in all the stages of their deliberations and are not bound to distinguish the effect of propounders’ evidence separately considered. 10th. The court charged the jury as follows: ‘Before you can do this’ .(set up the instrument presented as the last will and testament of Mary L. Spencer), ‘and before the case is made out on the part of the propounders, it must be shown to your satisfaction that Mary L. Spencer herself signed the instrument which is alleged to be her will, and signed it in the presence of all three of the witnesses whose names appear thereon, and that they signed it in her presence/ Movant says the foregoing charge is correct, but the court immediately proceeded to charge as follows: ‘When this is done, the burden of proving that said instrument is a forgery or that the said instrument is not the last will and testament of Mary L. Spencer, and that it has not been proven as the law requires, is shifted to caveator. It is then upon the caveator to carry this burden and to satisfy the jury of the truth of his contentions, before you would be authorized to find in his favor. Whether caveator has carried this burden and established his contentions to the satisfaction of the jury, that said instrument is a forgery and that it is not the last will and testament of Mary L. Spencer, is for the jury under 'the evidence to determine, which if done to your satisfaction would authorize at your hands a finding in favor of the caveator/ Movant alleges that the charge last quoted above is error. For that, as he contends, the parties had arrived at an issue and proposition, af.firmed by propounders on one side and denied by caveator on the other, and the burden, under the law, of establishing the affirmative of that proposition by a preponderance of evidence was on propounders, and never shifted during the course of the trial, but remained with them to the end. That the proof on both sides applies to the affirmative or negative of the issue of facts, and the propounders, whose case requires the proof of that fact to the satisfaction of the jury, had all along the burden of the proof. The burden was not upon caveator to establish to the satisfaction of the jury that the alleged will was a forgery; all the law required of'him was, *128that his evidence, taken with that for propounders and all the circumstances surrounding the case, should be sufficient to render the affirmation that Mary L. Spencer signed the will unsatisfactory to the jury, they, the jury, being the judges of whether or not all the evidence and circumstances satisfactorily settled the proposition that she did sign it, or whether it was insufficient to establish that proposition to their satisfaction. It was not incumbent on caveator to establish to the satisfaction of the jury that Mary L. Spencer did not sign the alleged will. If the evidence for propounders and caveator and all the surrounding circumstances were sufficient to prevent the jury from being sufficiently satisfied that Mary L. Spencer did in fact sign the will, then that was all the law required, for the caveator to prevail. Caveator did not have to prove to the satisfaction of the jury that it was not the will of Mary L. Spencer, as the charge complained of required him to do. Propounders had to prove to the satisfaction of the jury that it was the will of Mary L. Spencer, and all that was required of caveator was to prevent the propounders from doing this to the satisfaction of the jury. The jury should have been told that they must be satisfied from the whole case, the evidence for the propounders and for caveator and all the surrounding circumstances, that Mary L. Spencer did actually sign the will, and if they were not so satisfied from the whole case, then propounders had failed to establish their contention and caveator was entitled to a verdict.”
Much confusion has arisen out of the fact that the expression “burden of proof” has been used in two senses, viz: 1. The necessity which rests upon a party at any particular time during a trial to create a prima facie case in his own favor, or overthrow one when created against him. 2. The necessity of establishing the existence of a fact, or state of facts, by evidence which preponderates to a legally required extent. 16 Cyc. 926. In the charge complained of in the ninth ground of the motion, the judge instructed the jury that the propounders held the affirmative and the burden of proof was upon them to prove that the alleged testatrix herself signed the instrument propounded, and that “it has been proven as the law requires.” In another part of the charge this expression was explained so as practically to mean proved by the attesting witnesses. He then informed them that when this has been done the burden is shifted to the caveator “to make good” his contention as *129set up -in his caveat. In the 'charge complained of in the 'tenth ground, the judge first instructed the jury that'before the case was made out on the part of the propounders it must be shown to the satisfaction of the jury that the alleged' testatrix herself signed the will propounded, in'the presence of all three attesting witnessesj and that they 'signed in her presence. He then added that when this was done' the burden of proving that the instrument was a‘ forgery or was not the last will and testament of, the alleged testatrix shifted to the caveator. He described the burden which the caveator carried on the subject of whether the will was signed by the alleged testatrix or not, by saying,'“whether the caveator has carried this burden 'and established his contentions to the satisfaction of the jury” was for the jury to determine under the evidence, “which if done to your satisfaction would authorize at your hands a finding in favor of the caveator.” This clearly informed the jury that certain things would shift the burden of proof to the caveator, and in order to successfully carry the burden thu's placed, upon him he must satisfy the jury of the truth of his contentions. It is also true that in describing what was necessary on the part of the propounders to make óut a .prima facie case the judge used the expression “to your satisfaction.”- But this did hot relieve the fact that to meet a prima facie case the caveator must satisfy them of his contention. This placed upon him a greater birrden than that required by law. The propo’unders alleged thát the will was executed by the purported testatrix. This was an inherent part of their case. The burden' was on them to establish that fact by a preponderance of evidence. The caveator could, of course, undertake to rebut the evidence on the subject of the'factum or‘execution of the instrument introduced by the propounders, but he did not undertake an affirmative burden of satisfying the jury that the will was not signed by the alleged testatrix. In the final determination of the' issue, the jury should take into consideration all of the evidence introduced both by the propounders and the caveator, and from a consideration of it arrive at a conclusion in regard to that issue. In several cases which have been before this court, where the caveat was based on the ground of insanity, undue influence, fraud, or the like, there have been discussions as to what evidence was necessary to be introduced' by the propoundef, in addition to proof of the factum of the will, in order to make out'a *130prima facie case, and wlien the burden of proof to establish grounds of the character just indicated devolves upon the caveator. In none of them which have come to our attention was there involved the question of the duty of establishing by a preponderance of the evidence the signing of the will. Freeman v. Hamilton, 74 Ga. 317 ; Slaughter v. Heath, 127 Ga. 747 (9), (57 S. B. 69), and cases cited. It is unnecessary to enter into a-discussion of what will make oüt a prima facie case, or in regard to the duty to rebut such a case, or as to the quantum of evidence necessary to be introduced by a propounder in regard to sanity and voluntary action on the part of a testator. The issues here raised were merely in regard to whether the will was executed by the alleged testatrix as by law required, or was a forgery, and, if her signature was actually placed on the instrument by her, whether it was.in ignorance of its contents. If in a trial the question is whether a certain person signed a certain instrument or not, the party who asserts that it was so signed and to whose case this is essential carries the burden of proving it. If he makes out a prima facie case, that would authorize the jury to find in his favor, if nothing further appeared. But if the defendant introduces evidence tending to disprove the prima facie case thus made, all the evidence should be considered by the jury in finally determining where the preponderance lies, the burden being upon the person who alleged the fact of the signature, in the sense that such allegation must be supported by the preponderance of the evidence in order for the party making it to have a verdict in his favor. This being so, it was misleading to charge, on an issue of whether the alleged testatrix signed the instrument propounded as her will, that the caveator carried the burden of establishing his contention "to the satisfaction” of the jury, although previously the court had charged that at the outset the propounders must show that the alleged testatrix signed the -instrument in the presence of the three witnesses whose names appear thereon and that they signed in her presence. There was a sharp conflict in the evidence, and we can not say that this error was harmless.
2, 3. The rulings announced in the second and third headnotes do not require discussion.
4. The defendants in error filed a cross-bill of exceptions, assigning error upon the admission of certain evidence. One of the *131witnesses whose name appeared to- be signed to the will was J. A. Stover. The evidence showed that he was dead, and the propounders introduced testimony to prove his handwriting. The caveator introduced counter-testimony for the purpose of proving that the purported signature of Stover was not genuine. On this question there was much conflict in the evidence. The caveator offered evidence of the son of Stover that he had several conversations with his father, occurring after the death of the testatrix, in which his father said that it seemed very strange that Mrs. Spencer did not make a will, and asked the son if he had heard Mrs Spencer say anything about a will. He also offered a letter written by J. A. Stover to Spencer, the husband and sole heir of testatrix, whose administrator was the caveator, in which Stover stated that if Spencer wished to sell his farm and would make a price and terms with regard to it, and the writer could afford to pay for the home for himself and family, he would consider the proposition. The will was dated in 1891, and Mrs. Spencer died in 1892. The letter was written in 1905. The proceedings to probate the will were not begun until 1901. Evidence was introduced tending to show that the farm referred to in the letter was land covered by the will. Under the terms of the will a life-estate only was left to Spencer, with remainder over to the propounders. This evidence was admitted over objection upon the part of the propounders.
The objections may be reduced- to two substantial points: (1) was the evidence inadmissible as hearsay; (2) was it inadmissible on the ground that the attesting witness being dead he could not be asked touching conflicting statements, and therefore could not be impeached by their introduction? We do not think either reason sufficient ground for excluding the evidence. In several earlier English cases evidence of this character was admitted, but in Stobart v. Dryden, 1 M. & W. 116, which was decided in 1836, declarations of a deceased attesting witness to a mortgage deed whose handwriting had been proved were offered, as amounting to an acknowledgment of forgery, and were rejected by the Court of Exchequer. Parke, B., declared that the evidence of the handwriting in the attestation was not used as a declaration of the witness, but to show the fact that he put his name in that place and manner in which in the ordinary course of business he would have done if he had seen the deed executed. This case stood for some *132time as a leading case, and has been referred to approvingly by some'eminent text-writers. 1 Greenleaf on Evidence (15th ed.), § 226 ; 1 Reclfield on Wills (4th ed.), *270. In the United States a doctrine diifereht from 'that' asserted in the Strobart’ case was early anhounced. In' 1819 it was said by 'Kirkpatrick, C. J., in the case of Newbold v. Lamb, 2 Southard (5 N. J. L.), 499-501: "The only reason why proof of the handwriting of a witness is taken as sufficient proof of the execution of a deed is founded upon the presumption that what' an honest man has attested under his hand, is true.” See also Krauss v. Miller, 10 S. & R. 158; Clark v. Boyd, 2 Ohio, 280; Losee v. Losee, 2 Hill (N. Y.), 609; Kirk v. Carr, 54 Pa. 285" court="Pa." date_filed="1867-03-04" href="https://app.midpage.ai/document/kirk-v-carr-6232914?utm_source=webapp" opinion_id="6232914">54 Pa. 285-90; Boyen’s Will, 23 Iowa, 254-57; Egbert v. Egbert, 78 Pa. 326" court="Pa." date_filed="1875-02-08" href="https://app.midpage.ai/document/egbert-v-egbert-6235013?utm_source=webapp" opinion_id="6235013">78 Pa. 326; Neely v. Neely, 17 Pa. 227" court="Pa." date_filed="1851-01-19" href="https://app.midpage.ai/document/neely-v-neely-6228995?utm_source=webapp" opinion_id="6228995">17 Pa. 227; Reformed Dutch Church v. Ten Eyck, 25 N. J. L. (1 Dutch.) 40; M'Elwee v. Sutton, 2 Bailey (S. C.), 128; Colvin v. Warford, 20 Md. 357" court="Md." date_filed="1863-12-11" href="https://app.midpage.ai/document/colvin-v-warford-7891736?utm_source=webapp" opinion_id="7891736">20 Md. 357. In Runyan v. Price, 15 Ohio St. 1 (86 Am. D. 459), and Baxter v. Abbott, 7 Gray, 71, some of-the reasoning apparently conflicts with the authorities above recited; but in the Ohio case, on the trial of an appeal 'from the probate of a will, the testimony of a subscribing witness taken on the original probate was read in evidence under a statute of that State, the witness being dead. Evidence was offered to show declarations of the witness respecting the capacity of the testator, for the purpose of impeaching his testimony. It was-held that before such statements could be given' in'evidence the witness must'be interrogated as to them, and the fact that he died after giving his original testimony, and that thus the opportunity for such examination had been, cut off, did not form an exception to the general rule. In the Massachusetts case it was said that no inference as to the opinion of an attesting witness to a will could be drawn from the mere fact of his signing it, and that therefore evidence of a contradictory opinion expressed by him was inadmissible. The subject is considered’ at some length in 2 Wigmoreon Evidence, §§ 1505, 1514.
In Deupree v. Deupree, 45 Ga. 415, it was held that if the testamentary paper be proved by the witnesses to have been subscribed by the testator in their presence, and they further state that they signed as witnesses ■ immediately thereafter, but they are unable from want of recollection to state affirmatively whether the testator remained in the room or not whilst they were signing, and seven *133or eight years have elapsed, there is a presumption of law arising from the attestation that it was duly attested; and that it was error in the court to charge the jury that this presumption did not arise unless the attestation clause recited the presence of the testator during the subscription of the witnesses. In the opinion it was said:- “In 2 Redfield on Wills, 35, it is said that even where there is no attestation clause, or it is defective, there still remains the presumption that all which appears upon the paper occurred in the order stated, and as the law requires it shall be done." The proof of attestation, therefore, carries with it something more than the mere fact that the witness signed the paper. If the witnesses are placed on the stand, they can be cross-examined and can be asked if they have not made statements conflicting with other testimony as given. If one or more of them is dead, and evidence is introduced to prove his signature, the purpose and effect' is not solely to prove that such witness or witnesses signed the paper, but from that fact to derive inferences largely dependent upon the presumption that when they purported to sign properly they did so. When this added effect is to be used, it can not be counteracted, it not being possible to cross-examine such attesting witness or to lay the foundation for impeachment. If proof of the handwriting of the attesting witness is to carry with it the force of an assertion by him that the instrument was executed because he witnessed it, this implied assertion should be impeachable by showing that he had made statements conflicting with it. The rule as to making a preliminary examination and calling the attention of the witness to conflicting statements before introducing them can not be applied in such a case. Nor is it like the Ohio case, above cited, where a witness testified and then died, and it was desired to show subsequent statements conflicting with his former testimony. A somewhat similar relaxation as to the requirement of preliminary examination before introducing impeaching conflicting statements was recognized in Battle v. State, 74 Ga. 101. It was there held that where dying declarations were proved, it was admissible to show that after the deceased was wounded he made statements concerning the transaction conflicting with such dying declarations. We recognize the fact that there may be some danger arising from the admission of such impeaching testimony, but there is also danger in admitting dying declarations, statements claimed to be *134part of the res gestae, opinions of witnesses, and proof of the signature of an attesting witness itself as having probative value- in lieu of the introduction of the witness. But the danger of abuse arising from the admissibility of the evidence can not destroy such admissibility or outweigh the counter-danger arising from admitting mere proof of the handwriting of a witness to have evidential value of the execution of a will without the introduction of the witness, and at the same time absolutely shutting oil any practical mode of impeaching or destroying such evidential value, thus in effect relaxing the rule requiring the production of the witness in favor of one person without relaxing the rule that the witness, when produced, must be asked about the conflicting statements before proving them.
Judgment reversed on main-bill of exceptions; and affirmed on cross-bill.
All the Justices concur.