Upon the question, whether there was legal evidence of the delivery of the deed, the cases of Vines, v. Brownrigg, 4 Dev. 265, and Andrew v. Shaw, 4 Dev. 70, are in point. They lay it down, that the act of 1806 does not create any new rule, as to the proof of the execution and delivery of a deed of gift of slaves ; and that, if the subscribing witness, from want of integrity, will not, ©r, from want of memory or knowledge, cannot, prove the signing, sealing and delivery of the deed, the. deficiency in his *213evidence may be supplied by that of the other witnesses.' Those adjudications and the reasons for them are attacked on the ground, that the statute requires a deed of gift to be attested by at least one credible witness, and that he shall prove the due and fair execution of it on the trial. It is argued, that delivery is an essential part of the execution of a deed, and, indeed, that it is no deecl until delivery : and thence, that the subscribing witness must attest the delivery as well as the signing and sealing. But that seems to be rather a play on words, and an • adherence to the letter, without regard to the sense and purpose, of the Statute, which would render it absurd and inoperative. It is true, that, technically, delivery forms part of the execution of a Seed ; that is, it. is not a deed without delivery. But, in common speech, execution means generally signing and sealing a paper, as contradistinguished from its delivery.' It seems plain, that it is to be understood in that sense in statutes, which require subscribing witnesses ; for no one ever thought of delivering a deed before its attestation.— This verbal criticism, overlooking the context and nature of the thing, would destroy the attestation of Heeds delivered as escrows, unless the same person happened to be the witness to the signing and sealing and to both the first .delivery and the final one ; for, until the latter, the instrument is not a deed, and, so, the attestation could not be that of a •witness to the deed. Thus, also, the statute of devises uses the language, that no last will shall be good, unless such last will be written in the testator’s life and signed by him, and be subscribed in his presence by two witnesses at least;, and, then, that the same shall be proved by at least one of the subscribing witnesses, but, if contested, it shall be proved by all. What is to be subscribed by the witnesses ? The will, answers the statute. But, by the same statute, litera-tim, it is not a will, until it be subscribed by the two witnesses ; and then, according to the argument, the attestation must be null, since it was-nol a will — that is, a perfect will *214—upon !he subscription of the first witness, nor indeed until the death of the testator. That cannot be the meaning of the statute. On the contrary, it is manifest, that “ such last will shall be subscribed by two witnesses” means, that the paper writing, purporting to be the will, shall be thus subscribed. Accordingly, it has been supposed to be perfectly settled, that the two witnesses need not even subscribe together, but may do so at different times and not in the presence of each other. The ground, on which Vines v. Browning and Andrews v. Shaw are impeached, thus seems to the majority of the Court altogether unsatisfactory in itself, and to leave those cases with all the authority, to which, as judicial precedents, they are entitled. The point decided 'distinctly arose in each case; and, upon mature consideration, the judges held, that it was not the purpose of the act of 1806, more than that of 1792, to require more to be proved by the witnesses to the writings mentioned in them, than by the witnesses to other instruments, but that the intention was merely to restore the rule of the common law, that upon trials such instruments were to be read upon proof of them then made by the witness, and not upon the proof, on which they were registered. That was thought to be the whole scope of the act of 1806, and, the more especially, as it was but applying to it the construction known to have been invariably put on that of 1792 couched in like language. Those decisions were made in December 1833, and have been fully acquiesced in, we believe, ever since, by the profession, and not questioned until the present time. They have, besides, received the sanction of the legislature. In revising the body of the statute laws in 1836, not only is the first section of the act of 1806 re-enacted without alteration, but, with those decisions before the legislature, the third section of the act of 1792, and the second of that of 1806 are incorporated into one section, saying that on trials the due and fair execution of written conveyances of slaves, by way of gift or sale,. *215shall be proved by the subscribing witness; thus expressly putting the two on the same footing, as the Court had by inference before held it had been intended to do. Rev. St. ch. 37, sec. 21. It would be a public mischief, in this state of the matter, to over-rule those cases ; for, if the point now agitated is not to be considered as1 thus put to rest, it would seem that nothing is to be deemed settled in our law.
It was, however, further contended in the argument, that although that may be generally true, here the deficiency is not supplied because the wife was not competent to accept the delivery, nor to prove it. "As to the competency of Mrs. Gaskill to give evidence : It is true, she could not have done so in her husband’s life time, in a controversy to which he was a party, both in respect of his interest and person. But when he died, her exclusion, as far as it arose Irom the interest of the husband or the policy of the law, ceased ; and she became competent in any suit by or against the husband’s administrator, to give evidence against the administrator, though not for him. The first she can do, because she swears against her interest, which is always allowable ; the latter she cannot, beeause the effect of her evidence would be to increase the fund, out of which she is to have a distributive share, or repel a charge on it. But the question of her capacity to give evidence in this case is much like that started in Harrison v Burgess, 1 Hawks, 384: whether, upon a caveat of a husband’s will, the widow could prove, that he depositedit with her for safe-keeping, so as to bring it within the act of 1784: on which the opinion of the Court was, undoubtingly, in the affirmative. In truth, this communication must, from its nature, have been made to the wife for the express purpose that she should make it known so as to effectuate the deed; and therefore, the case falls directly within that of Hester v Hester, 4 Dev. 228. As to the other part of the proposition, namely, that the delivery to the wife was not a delivery of the deed in law: This is said on the ground, that husband and wife *216iirc one, and thence is derived the idea, that the delivery to her was retaining the custody in his own hands; and, so, lie did not part from the control of the instrument, and it never became his deed. The Court cannot adopt the reasoning. In the case just cited, this notion was repudiated. For, although, to many purposes husband and wile are one, yet they are naturally two persons, and to many other purposes they ate distinct in a legal sense, both criminaliter and cioilüer; and among those instances it was there held, that a wile might be a depository of her husband's will as any other friend might be. That seems to be directly applicable to the case in hand. It is common doctrine, that a wife may be her husband’s attorney, and, with the husband’s assent, the attorney of another person. There is nothing in their relation to forbid that. If a third person, then, had made this deed, and delivered it to the wife for the son, there can be no question oí the delivery. , For, not only is it delivery to the donee’s attorney a delivery to the donee, but it has been long laid down, that a delivery to a stranger for the donee is suilicient and makes it a deed presently, and until he disagree thereto. The old cases were looked into in Tate v Tate, 1 Dev. and Bat., Eq. 22, and the conclusion there drawn from them is, that when the maker of a deed parts from the possession of it to any body, there is a presumption, that it was delivered for the benefit of the gaantee, and it is for the maker, to show the contrary. If, indeed, the husband had professed to deliver the deed to'his wife, or to any one else, to keep for him, or subject to his disposition/the presumption just mentioned would not arise. But that was not the fact in this case. The deed was expressly delivered to the mother for the son. The wife did not take it, as wife, or the servant of the husband, but exactly the contrary. The intention was, that it should operate as a deed to the son. It is said, however, that he had a legal dominion over her, and might have compelled her to give him up the paper, Whether he could or not, de*217pends upon the question, whether she held the instrument as the husband’s or the son’s property. He might, it is true, by superior strength and his authority over the person of the wife, have forcibly compelled her to part from the paper. But he could not have done so rightfully, if he parted from the instrument as his deed for one instant; for, he would have no more authority, legal or moral, to take from his wife a deed made by him to his son and in her custody for the son, than he would have to take a deed made by a third person and left with her by the donor or donee to keep for the donee. The dominion of a husband over the wife is given for his security and her protection, and to those ends •will be upheld. But the law will not allow it to be abused and perverted to the prejudice of other persons, and such of their rights as are founded in justice and law. In this case, by the act of delivering the deed to the mother for her son, the husband expressed in the strongest manner he could, that she might act on behalf and for the benefit of their child in taking and keeping the deed as the son’s; and it became at once as operative as if it had been put into the hands of the infant himself, and could not be recalled.