The two questions in this ease are, first, whether the cancellation of a will, which was duly executed, and which contained a clause expressly revoking former wills, has the effect, as matter of law, to revive a former will which has not been destroyed, or whether in each instance it is to be regarded as a question of intention, to be collected from all the circumstances of the case; and secondly, if it is to be regarded as a question of intention, whether subsequent oral declarations of the testator are admissible in evidence for the purpose of showing what his intention was. These are open questions in this Commonwealth. In Reid v. Borland, 14 Mass. 208, the second will was invalid, for want of due attestation. In Laughton v. Atkins, 1 Pick. 535, the second will was adjudged to be null and void, as having been procured through undue influence and fraud; and the whole decision went upon the ground that it was never valid, and could not be.
The first of these questions has been much discussed, both in England and America; and it has been often said that the courts of common law and the ecclesiastical courts in England are at variance upon it. See 1 Wms. on Executors (5th Am. ed.) 154-156, where the authorities are cited. The doctrine of the ecclesiastical courts was thus stated in 1824, in Usticke v. Bawden, 2 Add. Ecc. 116, 125: “ The legal presumption is neither adverse to, nor in favor of, the revival of a former uncancelled, upon the cancellation of a latter, revocatory, will. Having furnished this principle, the law withdraws altogether; and leaves the question, as one of intention purely, and open to a decision, either way, solely according to facts and circumstances.” See also Moore v. Moore, 1 Phillim. 406; Wilson v. Wilson, 3 Phillim. 543, 554; Hooton v. Head, 3 Phillim. 26; Kirkcudbright v. Kirkcudbright, 1 Hagg. Ecc. 325; Welch v. Phillips, 1 Moore P. C. 299. In Powell on Dev. (ed. of 1827) 527, 528, a distinction is taken between the effect of the cancellation of a second will which contains no express clause revoking former wills, and of a will which contains such a clause; and in respect to the latter it is *254said that, “ if a prior will be made, and then a subsequent one expressly revoking the former, in such case, although the first will be left entire, and the second will afterwards cancelled, yet the better opinion seems to be, that the former is not thereby set up again.” Jarman’s note questions the soundness of the above doctrine (p. 529, n.). While this apparent discrepancy in the respective courts remained not fully reconciled, in 1887, the English Statute of Wills was passed, St. 7 Will. IV. & 1 Vict. c. 26, § 22 of which provided, that “ no will or codicil, or any part thereof, which shall be in any manner revoked, shall be revived otherwise than by the reexecution thereof, or by a codicil executed in manner hereinbefore required, and showing an intention to revive the same.” Since the enactment of this statute, the decisions in all the courts have been uniform, that after the execution of a subsequent will which contained an express revocation, or which by reason of inconsistent provisions amounted to an implied revocation, of a former will, such former will would not be revived by the cancellation or destruction of the later one. Major v. Williams, 3 Curt. Ecc. 432. James v. Cohen, 3 Curt. Ecc. 770, 782. Brown v. Brown, 8 El. & Bl. 876. Dickinson v. Swatman, 30 L. J. (N. S.) P. & M. 84. Wood v. Wood, L. R. 1 P. & D. 309. In order to have the effect of revocation, it must of course be made to appear that the later will contained a revocatory clause, or provisions which were inconsistent with the former will; and the mere fact of the execution of a subsequent will, without evidence of its contents, has been considered insufficient to amount to a revocation. Cutto v. Gilbert, 9 Moore P. C. 131. See also Nelson v. McGiffert, 3 Barb. Ch. 158.
In the United States, there is a like discrepancy in the decisions in different States, though the clear preponderance appears to be in favor of a doctrine substantially like that established in the ecclesiastical courts. This rule was established in Connecticut, in 1821, in James v. Marvin, 3 Conn. 576, where it was held that the revocatory clause in the second will, proprio vigore, operated instantaneously to effect a revocation, and that the destruction of the second will did not set up. the former one; and the like rule was declared to exist in New York, by the Supreme Court of that State, in 1857, in Simmons v. Simmons, 26 *255Barb. 68. The question was greatly considered in Maryland, in 1863, in Colvin v. Warford, 20 Md. 357, 391, and the court declared that “ a clause in a subsequent will, which in terms revokes a previous will, is not only an expression of the purpose to revoke the previous will, but an actual consummation of it, and the revocation is complete and conclusive, without regard to the testamentary provisions of the will containing it.” The court further held that the cancellation of a revoking will, prima facie, is evidence of an intention to revive the previous will, but the presumption may be rebutted by evidence of the attending circumstances and probable motives of the testator. In Harwell v. Lively, 30 Ga. 315, in 1860, a similar rule was laid down, and maintained with great force of reasoning. The opinion of the court concludes with the following pertinent suggestion: “It must be conceded there is much law adverse to the doctrine.....Calculated as it is to subserve and enforce the tenor and spirit of our own legislation, and to give to our people the full benefit of the two hundred years’ experience of the mother country, as embodied in the late act, is it not the dictate of wisdom to begin in this State where they have ended in England ? We think so.” See also Barksdale v. Hopkins, 23 Ga. 332. The courts of Mississippi, in 1836, and of Michigan, in 1881, adopted the same rule. Bohanon v. Walcot, 1 How. (Miss.) 336. Scott v. Fink, 45 Mich. 241. It is to be observed, that some of the foregoing decisions are put expressly on the ground that the later will contained an express clause of revocation. 45 Mich. 246. 20 Md. 392. An examination of the cases decided in Pennsylvania leads us to infer that a similar rule would probably have been adopted in that State, if the question had been directly presented. Lawson v. Morrison, 2 Dall. 286, 290. Boudinot v. Bradford, 2 Yeates, 170; S. C. 2 Dall. 266. Flintham v. Bradford, 10 Penn. St. 82, 85, 92.
On the other hand, in Taylor v. Taylor, 2 Nott & McC. 482, in 1820, it was held in South Carolina that the earlier will revives upon the cancellation of the later one; and the same rule prevails in New Jersey, as is shown by Randall v. Beatty, 4 Stew. (N. J.) 643, and cases there cited.
In various States of the Union, statutes have been enacted substantially to the same effect as the English statute above *256cited, showing that wherever, so far as our observation has extended, the subject has been dealt with by legislation, it has been thought wiser and better to provide that an earlier will shall not be revived by the cancellation of a later one. There are, or have been, such statutes in New York, Ohio, Indiana, Missouri, Kentucky, California, Arkansas and Virginia, and probably in other States. Concerning these statutes of New York, it is said in 4 Kent Com. 532, that they “have essentially changed the law on the subject of these constructive revocations, and rescued it from the hard operation of those technical rules of which we have complained, and placed it on juster and more rational grounds.”
On the whole, the question being an open one in this State, a majority of the court has come to the conclusion that the destruction of the second will in the present case would not have the effect to revive the first, in the absence of evidence to show that such was the intention of the testator. The clause of revocation is not necessarily testamentary in its character. It might as well be executed as a separate instrument. The fact that it is inserted in a will does not necessarily show that the testator intended that it should be dependent on the continuance in force of all the other provisions by which his property is disposed of. It is more reasonable and natural to assume that such revocatory clause shows emphatically and conclusively that he has abandoned his former intentions, and substituted therefor a new disposition of his property, which for the present, and unless again modified, shall stand as representing his wishes upon the subject. But when the new plan is in its turn abandoned, and such abandonment is shown by a cancellation of the later will, it by no means follows that his mind reverts to the original scheme. In point of fact, we believe that this would comparatively seldom be found to be true. It is only by an artificial presumption, created originally for the purpose of preventing intestacy, that such a rule of law has ever been held. It does not correctly represent the actual operation of the minds of testators, in the majority of instances. The wisdom which has come from experience, in England and in this country, seems to point the other way. In the absence of any statutory provision to the contrary, we are inclined to the opinion that *257such intention, if proved to have existed at the time of cancel-ling the second will, would give to the act of such cancellation the effect of reviving the former will; and that it would be open to prove such intention by paroi evidence. Under the statute of England, and of Virginia, and perhaps of other States, such revival cannot be proved in this manner. Major v. Williams, and Dickinson v. Swatman, above cited. Rudisill v. Rodes, 29 Grat. 147. But this results from the express provision of the statute.
In the present case, there was no evidence tending to show that the testatrix intended to revive the first will; unless the bare fact that the first will had not been destroyed amounted to such evidence. Under the circumstances stated in the report, little weight should be given to that fact. The will was not in the custody of the testatrix, and the evidence tended strongly to. show that she supposed it to have been destroyed.
The question, therefore, is not very important, in this case, whether the subsequent declarations of the testatrix were admissible in evidence for the purpose of showing that she did not intend, by her cancellation of the second will, to revive the first; because, in the absence of any affirmative evidence to prove the existence of such indention, the first will could not be admitted to probate. Nevertheless we have considered the question, and are of opinion that such declarations were admissible for the purpose of showing the intent with which the act was done. The act itself was consistent with an intention to- revive, or not to revive, the earlier will. Whether it had the one effect, or the other, depended upon what was in the mind of the testatrix. It would in many instances be more satisfactory to have some decisive declaration made at the very time, and showing clearly the character of the act. Evidence of declarations made at other times is to be received with caution. They may have been made for the very purpose of misleading the hearer as to the disposition which the speaker meant to make of his property. On the other hand, they may have been made under such circumstances as to furnish an entirely satisfactory proof of his real purpose. It is true, that it may not be proper to prove the direct act of cancellation, destruction or revocation in this manner. But when there is other evidence of an act of *258revocation, and when the question of the revival of an earlier will depends upon the intention of the testator, which is to be gathered from facts and circumstances, his declarations, showing such intention, whether prior, contemporaneous or subsequent, may be proved in evidence.
In the great case of Sugden v. St. Leonards, 1 P. D. 154, the question underwent full discussion, in 1876, whether written and oral declarations made by a testator, both before and after the execution of his will, are, in the event of its loss, admissible as secondary evidence of its contents; and it was decided in the affirmative. It was admitted in the argument, at one stage of the discussion, that such subsequent declarations would be admissible to rebut a presumption of revocation of the will; but, this being afterwards questioned, it was declared and held, on the greatest consideration, not only that these, but also that declarations as to the contents of the will, were admissible. See pages 174, 198, 200, 214, 215, 219, 220, 225, 227, 228, 240, 241. The case of Keen v. Keen, L. R. 3 P. & D. 105, is to the same effect. See also Gould v. Lakes, 6 P. D. 1; Doe v. Allen, 12 A. & E. 451; Usticke v. Bawden, 2 Add. Ecc. 123; Welch v. Phillips, 1 Moore P. C. 299; Whiteley v. King, 10 Jur. (N. S.) 1079; Re Johnson’s will, 40 Conn. 587; Lawyer v. Smith, 8 Mich. 411; Patterson v. Hickey, 32 Ga. 156; 1 Jarm. Wills, (5th Am. ed. by Bigelow,) 130, 133, 134, 142, and notes. The question was also discussed, and many cases were cited, in Collagan v. Burns, 57 Maine, 440, but the court was equally divided in opinion. Many, though not all, of the cases, which at first sight may appear to hold the contrary, will be found on examination to hold merely that the direct fact of revocation cannot be proved by such declarations.
The result is, that, in the opinion of a majority of the court, the will should be disallowed, and the decree of the Probate Court Reversed.