I am of opinion that the will of William Bunnavant deceased was well proved, there being no evidence of fraud or unfairness in the transaction.
It is a will attested in 1828, and offered for probate in 1837. After such a lapse of time, the forgetfulness of the witnesses as to some of the circumstances attending its execution, affords no fair presumption tha* they did not exist. New persons witnessing a paper would, after eight or nine years, be able to recall every fact that might be necessary to give it legal validity; and if their defect of memory is, without other impeachment, to prejudice the rights of parties claiming under it, the mischief would be greater than any that can result from this decision. The law regulating devises requires reasonable proof that every statutory provision has been complied with, but it does not prescribe the mode of proof, nor that the will shall be proved, as well as attested, by two or more credible witnesses; nor that frail memory shall change its nature, and perform impossibilities. A will over thirty years old is said to *23prove itself, although the testator may have died recently, because of the unreasonableness of requiring strict proof of so old a transaction. (Said by counsel, arguendo, in Calthorpe v. Gough, 4 T. R. 707. 709. n. to have been decided by lord Kenyon in Mackay v. Newbolt. And see Ld. Rancliffe v. Lady Parkins, 6 Dow’s P. C. 202.) What then if it be 29 years old, or has been attested so long before as to afford a fair presumption, derived from our experience, that many things actually existing may have faded from the memory of the witnesses ? Will a jury, or a court of probate, which acts upon the same principles of evidence as a jury, reject the will because those circumstances are not all recollected ? If the witnesses are dead, we infer all the necessary requisites from proof of their handwriting, although the memorandum of attestation is silent as to material ones. Hand v. James, Com. Rep. 531. Croft v. Pawlett, 2 Stra. 1109. Brice v. Smith, Willes’s Rep. 1. We also infer the signing of the devisor, from the fact of his acknowledging the instrument to be his will, (Ellis v. Smith, 1 Ves. jun. 11.) and the signature of the witnesses in his presence, from the fact that they were all in one room. Why then may we not presume the same thing, from the proof of any other facts or circumstances by which those to be inferred are usually accompanied ? This is a familiar rule of evidence, illustrated by all the writers on the subject, and applicable to every case where the law has not positively prescribed the degree of proof. Why, in this case, should we make an exception ? Why exclude circumstantial evidence, or deduction of fact from fact, and confine ourselves to positive proof, so difficult to be obtained, and more difficult from conscientious than from unscrupulous witnesses ? It is admitted there is no authority for this innovation, and I think it rests on no sound principle. The question is not what facts are to be proved (for about that we all agree), but through what *24media those facts may be impressed upon the minds of . . _ _ _ . , those appointed to try and determine them.
In the case at bar, the will is signed by the testator (for his mark is a signing within the statute; Harrison Harrison, 8 Ves. 186. Addy v. Grix, 8 Ves. 604.) and there are three subscribing witnesses to it, whose names are in juxtaposition ; and from the absence of proof to the contrary, we may presume that the ink has the same appearance in all the signatures. All the witnesses prove their handwriting, and declare that they would not have attested but at the request of the devisor. The first witness, Lipscomb, declares that he signed the testator’s name recognizing the mark. He proves that he witnessed the will when other persons not recollected were present, and he specifies minutely the particular occasion when that took place. He also remembers that the will was read to the testator at his desire; that he subscribed it within a few paces of him; that he is positive it must have been acknowledged at the time, or he would not have attested it; and that the whole transaction took place outside the house in the open air, where they were weighing Sudbury’s fodder, on the occasion of his removal from the county. The two other subscribing witnesses recollect that they were at the testator’s house on the very occasion specified by Lipscomb ; and none of the witnesses ever attested any other will or paper for Dunnavant, as far as we learn. Here is a chain of evidence very satisfactory to my mind, in proof of the only doubtful fact, that the will was signed by the witnesses, or at least two of them, in the presence of the testator; for Lipscomb signed in his presence, and the others were at the same place, at the same time, attesting after him. But suppose Lipscomb only proves the due execution of the will ás to himself, and the other witnesses establish no more than the genuineness of their signatures, being oblivious of all other circumstances; and that one of them would not *25have attested, “ unless requested bv the testator, and ,, m t> he had thought all things were regular.” (See Roberts s evidence.) Is not this sufficient proof to authorize a court of probate to admit a will to record, whemthe rejection of the motion will conclude one party, whilst the granting of it will not preclude the other from asking an issue of devisavit vel non ? I clearly think so. The forgetfulness of the witnesses ought not to be permitted to defeat the will; and for this position the case of Dagwell v. Glascock, Skinner 413. is a very strong authority. But there is a case in 19 Johns. Rep. 386. where the reasoning of chief justice Spencer in delivering the opinion of the court is very apposite to this, and covers indeed the whole ground. I cite it because I think the opinion he expresses is strictly in accordance with well established principles of evidence. It is the case of Jackson d. Legrange v. Degrange. The will in question was witnessed by Quackenbush, Lansing and Wendell. Lansing was dead; Wendell was living, and a resident of the state; but (Quackenbusk was the only witness called to prove the will. He established the genuineness of his own signature, and proved the handwriting of Darning; but he had forgotten every circumstance of the transaction, and did not even recollect that he had ever seen the testator. Yet from the circumstance of his name being subscribed as a witness, he supposed he must have seen it executed; and he knew the requisites to a good execution of a will, being at that time about to be admitted an attorney in the courts of New York. It was held that this was not sufficient proof of the will, because Wendell was not produced. The chief justice said, “ Wendell ought to have been called, inasmuch as Quackenbusk did not prove the facts essentially necessary to the valid execution of the will. If Wendell had been called, he might have either proved or disproved these facts. If his recollection should also have failed him, still, if he could have proved his signature, then, on *26proving the signature of the testator, I should be of the opinion that the will had been sufficiently proved toentitle it to be read. The law does not require impossibilities ; and therefore, when the will has been executed a l°n§ Peri°d before the trial, it is not ordinarily to be expected that the witnesses will be able to remember all the material facts. In this respect a will may be compared to a deed the signature of which is denied.” The will in the case in Johnson was older than in this, but it was not thirty years old, and the circumstance of more or less antiquity is no otherwise material than as it may be supposed to affect more or less the memory of the witnesses. The principle is, that positive, direct proof of material facts may be dispensed with in every case, where “it is not ordinarily to be expected that the witnesses will be able to remember” them. The period between the attestation and the probate might be so recent as to afford very strong, nay irresistible evidence that facts not remembered did not exist; and then all ground for presuming a due execution would be taken away.
A circumstance noticed in the case cited may be supposed to have had some weight; to wit, that Quaclcenbush knew the requisites to a good execution of a will. Here, Roberts did not know them; and Robertson the other witness says, if the testator had requested him to witness the will, he would have done so, whether he were present or not. It must be admitted that such declarations impair our confidence in the certainty of the inferences deducible from the other facts and circumstances;, but still'enough remains to authorize us to pronounce in favour of the will. The fact remains, that wills are usually witnessed in the presence of testators ; and the other circumstances shew that the ordinary concomitants of attestation and subscription existed in this case; it being satisfactorily proved that in respect to the first witness every thing was regular. In *27this particular, the case contrasts favourably with Jackson v. Legrange, where the first witness only proved his signature; and if he knew the requisites to a good execution, it did not follow that Wendell was equally well informed. Yet the chief justice held that if, in addition to Qiiaclcenbush’s testimony, Wendell should prove h,is signature merely, his recollection failing him. as to every other circumstance, it would be sufficient.
For these reasons, I am of opinion that the order of the court below should be affirmed.
Brooke, J.I do not dissent from the opinion of a majority of the court, because I know the inutility, as regards the public, of a difference of opinion as to the fact on any state of the evidence submitted to a jury or a court of probate. As to the principle that all the requirements of the statute must be proved in order to establish a will, there can be no difference of opinion. As to the kind and degree of proof that is necessary to satisfy the statute, I differ with the majority of the court; nor do I think there is anything exorbitant in the requirement of proof to satisfy the statute, either in the case of Burwell and others v. Corbin and others, 1 Rand. 131. or in my opinion in the case of Dudley v. Dudley, 3 Leigh 436. In the first case, judge Roane did not differ with the rest of the court on the construction of the statute, but on the degree of proof of its requirements, as we differ in the case before us. None of the court in the case of Burwell and others v. Corbin and others denied the rule, that, upon evidence, inference may be drawn from the proof of one fact or several facts to prove another fact, in a proper case. In the case before us I do not object to that rule, but I am opposed to making such a use of it as to repeal the statute. I do not think that proof of a will in all particulars by one witness is enough to establish by inference that another, who is examined, and recollects *28nothing but his attestation of a recent will, was a witness t0 that proved by the first. To prevent perjury, the statute requires two witnesses. Now, in the case before us, is it more probable that Robertson'and R°berts forgot the presence of the testator and that his mark was made by his direction, if in truth he was present and directed his mark to be made, than other facts not. so impressive, which they relate ? I will not notice the english cases : there is no difference between them and our own cases, except in the degree of proof which is required to satisfy the statute. Inference from one fact to prove another has been carried too far by the courts of both countries : the statute has been left out of sight, and the evidence treated as if all that was necessary was to prove that the testator executed the will. Putting it on that ground, one credible witness is as good as the two required by the statute; and if inference from his testimony is to supply the proof of another witness, required by the statute, the will is proved by one witness instead of two. The cases which, from necessity, admit of inferior proof to that which by the letter of the statute seems to be required, have no application to the case before us: as where all or some of the witnesses are dead, or out of the jurisdiction of the court; or as in the case in Skinner, where they were suborned to deny their signatures. In the first case, proof is admitted of their signatures; and in the second, to contradict the attesting witnesses, and to prove their attestation of the will. But that is not our case. All the witnesses are alive, and examined to prove the will; no secondary evidence is called for. Only one of them proves the presence of the testator when he attested, and that his mark was made by the witness. The other two only prove their handwriting, and that they would not have attested, they think, unless the testator had requested them, and all things had been regular. ' They do not prove that he made his mark as his signa*29lure, or that it was made in his presence and by his direction, nor that he was present when they subscribed their names; two material requisites of the statute. These particulars they could not recollect, though they remembered other matter not so impressive. In the case of Burwell and others v. Corbin and others, if the evidence of one witness was enough to satisfy the statute, the evidence of Scrimger the first witness proved every thing that was required; but if two witnesses were necessary, Barrióle the second witness proved nothing but the acknowledgment of the testator that the paper writing was his will. He knew nothing of the signature, the factum, of signing required by the statute: , though Scrimger the first witness proved it, he did not. For the grounds of my opinion in that case, I refer to the cases there cited. My opinion in the case of Dudley v. Dudley, 3 Leigh 436. was founded on the same course of reasoning; that is, that however full the testimony of one witness may be to prove a will, our statute requires two witnesses to the facts which are necessary to be proved. In England, three witnesses are required: and in Westbeech v. Kennedy, 1 Ves. & Beam. 362. two witnesses proved the will in every particular; and though the third witness did not prove the factum of signing, he proved that the testator sealed the will in his presence, which was regarded as proof of the signature. Now it does appear to me that the proof of the two witnesses in that case would not have sufficed, if the testator had not sealed the will in the presence of the third witness. In the case before us, one witness proves what is required by the statute. Suppose his testimony out of the case, and the testimony of jRobertson and Roberts the only testimony; it could not be said that theirs was sufficient. Neither of them proves the presence of the testator when they attested, nor that his mark, which stands for his name, was made in his presence and at his request. All they prove is their *30handwriting as attesting witnesses. But that is held ..... „ . . .... by a majority ox the court to be sufficient, m addition t0 the testimony of Lipscomb the first witness; and I will not dissent from them, for the reason before stated.
Tucker, P.The probate of this will depends mainly, I think, upon the question how far the defects of the memory of the witnesses can be supplied by mere inference from their attestation; for although there are some of the requisites of the statute proved, there are others unquestionably at which we can only arrive by implication.
There is no question, I think, that if the subscribing witness to a deed or will be dead, or absent from the country, or otherwise so situated that his testimony cannot be had, the proof of his handwriting is all that is necessary, and his attestation is a sufficient ground for presuming that the instrument has been executed with all the solemnities and ceremonies required by the law. This is confessedly the case with deeds and obligations, as appears from the numerous cases cited in Phillips on Evidence, p. 362. It is not less the case with wills; for if the witnesses are dead, or insane, or out of the jurisdiction, so as not to be amenable to process, their handwriting may be proved, and it is then a question for the jury, whether, -under the circumstances of the case, it is probable that all the formalities of the statute were regularly observed. This was expressly decided in Hand v. James, Com. Rep. 531. Croft v. Pawlett, 2 Strange 1109. Brice v. Smith, Willes 1. and as we shall presently see, our own court has acted upon similar principles. It is true that in these cases all the witnesses must be dead, or not to be had, for otherwise the due execution of the will cannot be established by proof of the handwriting of one or more. In a trial at law one of the subscribing witnesses, and in a question of probate two of them, may suffice to prove a will, provided they can *31prove all the solemnities required by the statute. But ; . . , , if they can only prove their own signatures, the other witnesses, if living, must be produced, or if dead, their handwriting must be proved ; and it is then, as I have said, a question for the jury or the court of probate, w’hether, under the circumstances, all the requisites of the statute have been complied with. See Jackson d. Legrange v. Legrange, 19 Johns. Rep. 386.
In the case at bar, then, if all the witnesses were dead, the proof of their handwriting wmuld have justified the court of probate in inferring that the w'ill had been executed with all the solemnities required by the statute. But the witnesses have all been examined. They have all proved their own attestations. They all swear, too, that they would not have attested the will, unless it had been acknowledged by the testator, and they had supposed all things regular. This has been held not only sufficient proof of a deed (1 Binney 436. 2 Hay 338. 1 Cox 10. cited Phillips on Evid. 358. in note. 2 Dali. 96.) but it has also been considered adequate in the case of a will. In Jackson v. Legrange, 19 Johns. Rep. 389. Spencer, chief justice, said, “ If the subscribing witnesses prove their signatures, though they may not be able to recollect the delivery, yet if they declare that they never subscribed as witnesses without a due execution of a deed by a grantor or obligor, such proof would be sufficient. In this respect, a will may be compared to a deed, the execution of which is denied.” I think, then, from the evidence in this case, we must take it as sufficiently proved that the testator acknowledged the instrument in question as his last will.
The question still recurs whether, as the witnesses have been actually examined, and have failed to prove a compliance with all the requisitions of the statute, that compliance can be inferred from their attestation.
It is very clear that if the witnesses denied their attestation, or would not swear that they saw the testator *32execute and publish his will, or alleged the incapacity of the testator, their handwriting might be proved and their testimony controverted. Skin. 413. Lowe v. Jollife, 1 W. Black. Rep. 365. But in these cases there mus,; he evidence produced to contradict their testimony. How is it where, as in this case, they do not negative a compliance with the statute, but merely have forgotten the circumstances ? It seems to me, that upon fair analogy, the question should be decided as it would be if the witness were dead and his handwriting proved. And so accordingly, I think, are the authorities to be understood. Thus, in the case of Dagwell or Dagnee v. Glascock, Skinner 413. Cases Temp. Holt. 741. there were three subscribing witnesses. One of them would not swear that he saw the testator seal and publish his will. Now, three witnesses were absolutely necessary to the validity of the will, so that here was a failure of proof as in our case ; but the court, upon proof of his handwriting, held the will to be sufficiently proved. This case seems to me to be in point. So in the case of Jackson v. Legrange, before cited, chief justice Spencer, speaking of the necessity of calling the other witness to the will, observes, “ If his recollection should also have failed him, still if he could have proved his signature, then, on proving the signature of the testator, I should be of opinion that the will had been sufficiently proved to entitle it to be read.” That is to say, where the memory of the witness fails, the proof of his signature will justify an inference that all the requisites of the law have been complied with. Then comes the following remark: “ The law does not require impossi-' bilities; and therefore, where the will has been executed for a long period before the trial, it is not ordinarily to be expected that the witnesses will be able to remember all the material facts;” and in this regard he likens a will to a deed, where, if the memory of the witness fails as to actual proof of delivery, it may still be in*33ferred from his attestation, and his swearing that he would not have subscribed it as a witness without a due execution.
Then comes the case of Smith v. Jones, 6 Rand. 33. In that case the witness, mrs. Jones, fell short of the full proof of the requisites of the statute. She neither proved that the name of the testator was written by the other witness at the testator’s request, nor did she prove the attestation to have been in his presence. Yet judge Carr, delivering the opinion of the court, says, if Pendergast (the other witness) had proved the same facts, “ we think the evidence of these subscribing witnesses would have proved circumstances from which a court of probate ought to have concluded that the will was signed by Pendergast for the testator, in his presence and by his direction, and moreover was attested by two witnesses in his presence.” Here, then, these important substantive requisitions of the statute were inferred, although the witnesses were present at the trial and examined, and yet failed to prove them. And this seems truly reasonable; since it never could have been the design of the statute to vacate a wall which was duly executed, whenever the witnesses to it have forgotten any material circumstances attending the attestation. They are indeed always called upon to prove the will, not because the statute requires that they shall prove a compliance with all the requisites of the law, but because they would be most likely to prove or disprove them, and because, upon principles of common law, the attesting witness to every instrument must be produced, if living and within the power of the court.
The case of Bond & wife v. Seawell, 3 Burr. 1773. cited by judge Carr, is deservedly considered by him as a strong case in support of this doctrine of inference. In that case the will was on two sheets. It was deemed essential that both sheets should have been in the room at the time of the acknowledgment. Yet two of the *34witnesses “ never saw the first sheet, nor was it produced to them, nor was it upon the table.” Nevertheless lord Mansfield was of opinion that the jury might presume, and ought to presume from the circumstances, that essential fact, though the witnesses did not and could not prove it. '
Next comes the case of Boyd v. Cooke, 3 Leigh 32. In that case the position taken in Smith v. Jones, that a court of probate occupies the place of a jury as to questions of fact, and that its province is to draw all just inferences from the evidence, was distinctly maintained. Lastly, in the case of Dudley v. Dudley, 3 Leigh 436. the judges reiterated this doctrine, and moreover declared that in the absence of all proof to the contrary, the acknowledgment and attestation gave rise to an irresistible inference that the instrument in that case had been previously signed; and they sustained the will upon this proof, although the testator’s name was not signed by himself, and although only one of the witnesses proved that it was signed by another person for him, in his presence and” by his direction.
From the whole of these cases, then, I deduce that on a question of probate, the defect of memory of the witnesses will not be permitted to defeat the will, but that the court may, from circumstances, presume that the requisitions of the statute have been observed; and that they ought so to presume from the fact of attestation, unless the inferences from that fact are rebutted by satisfactory evidence. Upon these principles, there can be no difficulty in the present- case. Lipscomb's testimony establishes his own attestation, and his conviction that he never would have attested if the testator had not acknowledged the will. The other witnesses prove the same. The acknowledgment, then, is 'established; and from this acknowledgment of the will, we infer an acknowledgment and recognition of the signature. To use the language of judge Cabell, “It *35is not usual for men to acknowledge papers, either as deeds or wills, and to call on others to attest them, before they are signed. Such a thing may happen; and when it is proved to have happened, the acknowledgment and attestation will be disregarded. But in the absence of all proof to the contrary, the acknowledgment and attestation give rise to an irresistible inference that the instrument had been previously signed.” It is moreover a recognition of the signature: for when, taking the paper already signed by his mark and with his name written by Lipscomb, he acknowledges it as his will, he of course acknowledges the whole, and not a part only; and still less can we presume that he designed to exclude the signature from that acknowledgment, without which it would have been no will. But if the signature is acknowledged as his, we are inevitably led to conclude from that acknowledgment and all the other circumstances, that Lipscomb affixed it at his request; and thus the proof would be full, even if his mark was not made with his own hand. But that it was so made I cannot doubt, both because of its being most usual for the party to make his mark, and because the paper calls it his mark, besides the recognition of the signing and sealing by the last clause in the will, which he must have heard when the rest of the will was read to him. See 7 Taunt. 251.
The execution by the testator being thus established, we come next to the attestation, and as there is no proof rebutting the inference that the attestation was according to law, we must take it to have been so; and thus the execution of. the will must be considered as fully proved.
Upon a very careful consideration, therefore, of the whole case, I am of opinion to affirm the sentence. Nor do I apprehend any evil from this decision. It may, perhaps, sometimes lead to the establishment of wills not duly executed, as doubtless may be the case *36also where the witnesses are all dead or absent, and every thing is presumed from their attestation. But far greater mischiefs would arise from a contrary decision, which should make the rights of every devisee and leSatee depend not only upon the honesty, but also upon the slippery memory of witnesses. Under such a decision, no man could be sure of dying testate, since the forgetfulness of a witness would frustrate all his precaution ; and a question of title by will, which, in the spirit of the statute of frauds, the legislature designed to rest upon written evidence alone, would after all depend upon the integrity and the memory of those who were called on to attest the instrument. The consequence would be, that I may have a good will by me to-day, but if I live five years if may become a void instrument, because one of the witnesses to it cannot recall some ceremony of its execution. Such a consequence I would not aid in bringing about. It would tend, I have no doubt, to multiply the attempts, already too common, to set aside wills; since the chances of success must be very much increased, if the frailty of human memory is to be called in to the aid of the discontented heir.
Sentence affirmed.