The proceeding is one seeking the *56probate of a will purporting to be executed by one H. L. Tillman, tlie validity of which was contested by the appellant. The main issue is as to the genuiness of the signature. of the testator.
1. Two of the three subscribing witnesses, who attested the execution of the paper, were shown to be dead. When this is the case as also where such witnesses have become incompetent to testify, or are insane, or absent from the State, it is allowed to prove the paper by evidence of a secondary nature. This may be either by proving the attestation to be in the handwriting of the deceased witness; or by proving the testator’s own signature by some person who is acquainted with it. Foote v. Cobb, 18 Ala. 585; Cox v. Davis, 17 Ala. 714; Guice v. Thornton, 76 Ala. 466. The court properly permitted the signatures of the attesting witnesses, Glenn and Wicker, both of whom were deceased, to be proved by persons who knew their hand-writing.
2. The court, also, properly refused to allow the contestant to exhibit' to a witness the signatures of other papers, purporting to have been made by the subscribing witness Wicker, not in evidence in the case, for the purpose of comparison with the alleged signature of Wicker to the will, and for the purpose of determining the genuiness of the latter. Kirksey v. Kirksey, 41 Ala. 626; State v. Given, 5 Ala. 747 ; 1 Greenl Ev. § 581; 1 Whart. Ev; § 712.
3. The question is raised in this case as to whether, in a proceeding of this kind, the proponent of a will, who is a party and interested as a legatee, is a competent witness under the statute, to prove the execution of the paper. Code, 1886, § 2765; Code, 1876, § 3058. The case of Kumpe v. Coons, 63 Ala. 448, decided by this court in 1879, determines this question in the affirmative, holding that a legatee or devisee under a will, who was an attesting witness to the paper, was competent to testify to its execution in any suit or proceeding in which the validity of the paper as a will may be involved. It was said that the statute put under the ban of exclusion evidence only of a particular character; viz: transactions with, or statements by, any deceased person, whose estate is interested in the result of the suit. The controversy, which was one to determine the status of the estate of the decedent, and the condition in which he died — whether testate or intestate — was one between living parties, affecting only their interests and it was observed that in no proper sense was the estate of the testator interested in the result.
*57We are deeply sensible o£ the bad results that may flow from such a rule of evidence, but this is for legislative rather than judicial correction. We are disposed to adhere to the case of Kumpe v. Coons for the following reasons: (1) The section of the Oode of 1876 (§ 3058), which was thus construed to-authorize the admission of such witnesses as legally competent, was re-enacted as section 27 65 of the Oode of 1886, without change, and this was a legislative sanction of the judicial construction previously placed upon it. (2) A sufficient time has elapsed, being some nine years, since the decision was promulgated, to justify us in believing that it has been acted on by the legal profession, and others, and that wills have been drawn and attested under its authority, and titles by devise to some extent acquired under it, which ought not to be disturbed by its reversal, even if we doubted the soundness of the decision. (3) It is often more important that a rule of law should be fixed, even though with less of reason in it, than subject to the uncertainty of fluctuating judicial decisions. “Certainty,” said Lord Hardwick, “is the mother of repose, and, therefore, the law aims at certainty.” — Sheddon v. Goodrich, 8 Ves. 497; Morton v. N. O. & Selma Railway Co., 79 Ala. 616.
The Probate Court committed no error in allowing Mrs. Burks, the proponent of the will, to testify as to the fact of its execution, although she was the sole devisee under its provisions.
1. The Probate Court erred in many of its rulings on the evidence in this case, for which we are compelled to reverse the judgment. It had no relevant bearing upon the issue in this case — which was the simple inquiry whether the paper in controversy was executed by H. L. Tillman, the decedent —that he was a person of reticent habits in his business; or that he owed debts to a larger amount than that stated in the alleged will. The latter fact, if true, rather tended to show that he neglected to pay or did not remember such other debts, or was not disposed to recognize them. The note and mortgage of January 15th, 1880, held by the witness W. L. Tillman against the testator should have been, excluded; and also the statement as to how much money belonging to the decedent’s estate had come into the witness’ hands, and was claimed by the proponent; and so should the letter of May 31, 1883, written by this witness to the proponent on 'this subject, with all other evidence of a like kind bearing on these collateral points, which was too remote to shed any light on the issue on trial.
*585. The court correctly charged the jury that no mere par- . tiality, caprice, or unequal distribution of property by the testator, or his failure to provide for the payment of his debts, could be regarded by the jury as a reason for setting-aside or vitiating a will, in other respects valid. The second charge given at the instance of the proponent only asserted this proposition in effect.
6. The third charge was correct if we construe it to assert that the measure of proof as to the proper execution of a will, in the particulars required by the statute, need only be such as is reasonably sufficient to satisfy tlio minds of the jury as to such fact. The first charge Avas defective in omitting, as one of these requirements, that the attesting witnesses should subscribe their names to the instrument in the presence of the testator. — Moore v. Spier, 80 Ala. 130.
7. The general rule is that where the will of a testator is proved to have been in the testator’s possession, and can not afterwards be found, the prima, facie presumption is that it was destroyed by the testator before his death animo revocandi. But this presumption may be rebutted by proper legal evidence. — 2 Greenl. Ev. § 688a. But if the instrument is shown to have been out of his possession, the party asserting the fact of revocation “must show that it came again into his custody, or was actually destroyed by his direction.” — 2 Greenl. Ev., § 681.
8. Where the will is executed in duplicate, as in this case, a somewhat different rule obtains. If, in such case, the testator destroys one of the duplicates, and this is the only one in his possession, an intent to revoke is to.be presumed —Mr. Greenleaf says, “is very strongly to be presumed;” but he adds, “if he was possessed of both copies and destroys one, it is weaker; and if he alters one, and then destroys it, retaining the other entire, the presumption has been said still to hold, though more faintly; but the. contrary,” he adds, “has also been asserted.” — 2 Greenl. Ev., § 682. We take this to be a correct enunciation of these general principles thus stated. The proponent testified that the testator executed his will in duplicate, and that he kept one copy himself, and gave the other to her. While it does not appear what she did with the one given to her, it certainly is not shown that it ever went back into the possession of the testator. There can, therefore, be no presumption of its revocation by him, under this state of facts. The fair inference from the evidence is that the will here sought to be *59probated, was the one retained by the testator. As to this point the evidence may, on another trial, possibly be made more clear, or at least less equivocal.
The third charge requested by the contestant was, under this view of the case, misleading, and was properly refused.
9. The first and second charges were purely argumentative, and there was no error in the refusal of the court to give them. Charges of this character — asserting that the jury “may look to” this fact, or “may consider” that fact, or “are authorized to” infer certain formulated conclusions from the evidence, and especially from specified parts of it — have often been condemned by us as objectionable, and should never be given, although either the giving, or the refusal of such instructions may not be a reversal error. They are legitimate arguments to the jury, not announcements of legal principles proper to be in the forms of instructions by the court.
The judgment is reversed and the cause remanded.
STONE, C. J.I participated in the decision of the case of Kumpe v. Coons, 63 Ala. 448, and think it right.