Decedent made a will in 1906 and kept it in his desk until his death. As some of the legatees died he drew pencil lines through their legacies, This was effective as a *576“cancellation” of those legacies within section 20(a) of the Wills Aet of June 7, 1917, P. L. 403, and they dropped out of the will. The residuary clause (or rather an alternative residuary clause) in favor of the proponents was left intact, and the signature was left intact. There was enough left of the will of 1906 to entitle the remnant to probate, if no more had appeared ; and the register did admit it to probate, omitting the canceled, clauses.
In 1939 decedent made another will duly executed. This also he kept in his desk. But the first part is missing. As thus produced it contains no express clause of revocation of prior wills, but it does make a disposition of his estate which is totally inconsistent with the will of 1906. An unsigned paper was produced by the contestant which was in the handwriting of decedent, and which contained a clause of revocation. The hearing judge, for convincing reasons, found that this paper was not the missing part of the 1939 will.
The will of 1939 was “repealed” (to use the language of the Wills Act) by drawing pencil lines through the signature of testator and the signatures of the witnesses; and this, there can be no doubt, was a cancellation of that will. Evidence was given of the oral and written declarations of decedent from which the hearing judge found that the cancellation of this will was done with the intention of making still another will; and he concluded that this brought about the revocation of the first will. He, therefore, reversed the decision of the register and set aside the probate, and as no other executed and uncanceled testamentary paper was produced, there is an intestacy.
There has been a great variety of opinion as to the effect of the revocation of .a second will upon a first will which has been preserved. A full account of these opinions can be found in the textbooks and the annotated cases; and particularly down to the year 1900 *577in an article in 48 Am. Law Reg. 505 by Mr. Justice Roberts, then an instructor in the Law School at the University of Pennsylvania. One of the conclusions of the author was adopted by the Supreme Court in Ford’s Estate, 301 Pa. 183, 30 years later. This was that when there is a second will containing an express clause of revocation of prior wills the first will is immediately annulled, and the revocation of the second will does not “revive” the first will. It is considered that the clause of revocation is not testamentary and does not depend for its effectiveness on the taking effect of the second will as the last will of .testator. It is a “repeal” by other writing “executed and proved in the manner hereinbefore provided” for the execution of a will: Wills Act, sec. 20(a), which is quoted in full below. This conclusion is abundantly supported by the authorities cited in the opinion of the court.
Contrasting with this is the case in which the second will contains no express clause of revocation of the first will, but disposes of the estate in a manner inconsistent with the first will. That is this case; and it is also the case of Flintham v. Bradford, 10 Pa. 82. It is often said that the inconsistent will “revokes” the prior will. But Coulter, J., in that case presented the proper view thus (p. 90) :1
“All wills are in their nature inchoate and ambulatory until testator’s death, at which time, and not before, the testament becomes operative and complete. The will of 1824 was an inchoate intention, mutable and inconstant, and, by the wilful and deliberate act of cancellation on the part of the testator, it became as if it never had been. The prior will of 1821, being preserved by testator entire, and without intentional or apparent blemish, became the will for the time being, which would be consummated at testator’s death, unless before that time he manifested a change of intention, according to the rules of law.
*578“That change of intention would be manifested by-making a new will, or a revocation of the prior will might be presumed even from subsequent acts of the testator; but only from those acts which of themselves have been adjudged to afford sufficient evidence of an entire change of intention, and, therefore, constituted of themselves independent revocations.
“These acts are such as burning, destroying, cancelling, or obliterating the instrument. Or by marriage of testator and birth of a child, conveying away the whole of the estate, &c. But none of these acts were present here, as regarded the will of 1821. So that if it was restored by the act of cancellation of the posterior will of 1824, it must remain the will of deceased.
“The language of Lord Mansfield in Goodright v. Glazier, 4 Burr. 2514, in which Justices Yates and Willes concurred, is strong and clear to the point. £A will (he says) is ambulatory till the death of the testator. If testator lets it stand till he dies, it is his will. If he does not suffer it to do so, it is not his will. Here he had two. He has cancelled the second. It has no effect, no operation. It is as no will at all, being can-celled before his death. But the former, which was never cancelled, stands as his will.’ ”
The argument of the opinion in Ford’s Estate also supports this conclusion. In that case there was a will of 1924, and wills of 1926 and 1927, both containing express clauses of revocation. The will of 1926 dropped out because it was executed in duplicate and one copy was not forthcoming: Ford’s Estate, supra, at page 187; Bates’ Estate, 286 Pa. 583. Testator tore up the 1927 will on his deathbed, making statements which tended to show that he wished to die intestate. This will was inconsistent with the will of 1924. The court said (p. 198) : “Without the existence of the 1926 and 1927 wills, his declaration would have amounted to *579nothing so far as the 1924 will is concerned . . That is to say, without the express clause of revocation in the 1926 and 1927 wills, the later will merely superseded the prior will; and the destruction of the later will would have left the prior will in force; and it could not be revoked by oral declarations.
In 1 Page on Wills (3rd ed.) 868, §475, there is the following comment on the distinction between the two cases:
“As far as this distinction can be justified from consideration of strict logic, it rests on the theory that revocation by an express revocation clause takes effect at the execution of the revoking will; while revocation by a later inconsistent will does not take effect unless testator leaves such revoking will in existence and unrevoked at his death. It would follow, therefore, that the first will was never revoked where the second will is merely inconsistent therewith. Some of the courts, however, speak of the destruction of the later will as a revival of the prior will.”
A leading case is Scott v. Fink, 45 Mich. 241, 246, which contains the following much-quoted passage:
“There seems to have been a material distinction, and on good ground, between the state of a former will after a second one merely inconsistent with it, and its state after a second one with a declaration expressly revoking it.
“In the first case the only chance for the second to operate in revocation of the first, according to the prevalent theories of the courts, was by its coming to a head as an active will, which it could do only by surviving its author. Being the last expression of the decedent and at the same time practically inconsistent with the prior one, the intent to repeal the first by it was to be implied. In case, however, of its being recalled by the testator in his lifetime, it could not, on the theory referred to, be taken to have had the effect *580to do away with its predecessor. Being cut off before having its dispositions of property awakened into life, it could haye no affirmative operation through its dispositions upon the estate.”
We know of no dissent from the general statement as to the effect of a later inconsistent will. However, a large .part of the cases actually deal with wills containing an express clause of revocation, and what is said about an inconsistent will is dictum. The writer has not often dealt with a subject so plagued with dicta, though the actual decisions may be right. Of the cases examined, the following concern a later inconsistent will which contained no clause of revocation, and support the conclusion in Flintham v. Bradford, supra; Peck’s Appeal, 50 Conn. 562; Dawson v. Smith, 3 Houst. (Del.) 92; Dingman v. Dingman, 199 Mich. 384; Cheever v. North, 106 Mich. 390; Johnston’s Will, 69 Hun. (N. Y.) 157. See also Mr. Justice Roberts’ article in 48 Am. Law Reg. 505.
It will conduce to clear thinking on this subject if the case of an inconsistent will is spoken of as such, and not as a revocation. The cancellation of the superseding will does not then “revive” the prior will, it merely removes the obstacle to the effectiveness of the prior will. The hearing judge concluded from Ford’s Estate and Flintham v. Bradford and other cases “that while a cancellation of a later will may revive an earlier will, this is not a conclusion, but a mere presumption. In other words, cancellation of a later and inconsistent will is a revival, unless an intention to the contrary is shown.” From the oral and written (non-testamentary) declarations of testator in evidence he concluded that the cancellation of the prior will “was made with the intention of making a new will”; and that this overcame the presumption, and made the revocation of the second will a revocation of the prior will also.
*581This, it seems to us, Is contrary to the theory of the effect of a later inconsistent will without a clause of revocation, and to the letter and spirit of section 20 (a) of the Wilis Act, supra, which is as follows:
“Section 20(a). No will in writing, concerning any real estate, shall be repealed, nor shall any devise or directions therein be altered, otherwise than by some other will or codicil in writing, or other writing declaring the same, executed and proved in the manner hereinbefore' provided; or by burning, cancelling, obliterating, or destroying the same by the testator himself, or by someone in his presence and by his express direction.”
Section 20(6) declares the same with respect to personal estate excepting nuncupative wills. And it is not authorized by any of the decisions referred to, as will appear in the discussion below.
In 1 Page on Wills (3rd ed.) 869, §475, is contained a forceful passage on this subject as follows:
“The courts which distinguish between the -cases in which the later will has an express revocation clause, and those in which it is merely inconsistent with the prior will, probably come nearer to the intention of the greater number of testators. The courts which have adopted the ecclesiastical law rule, or some modification thereof which treats the actual intention of the testator as of paramount importance, attempt to do justice by giving effect to the actual intention of the testator; but they do it by throwing open the gates for oral evidence of testator’s declarations, upon which the validity of the first will is to depend. As the whole tendency of legislation has been to require form in the execution of a will, and as the courts are careful to prevent testator from incorporating a document to be made in the future, no matter how clear his intent may be to incorporate it, it would seem that a situation in *582which testator might put one or the other of his wills in effect by oral declarations was a violation of the spirit of modern legislation on the subject of the execution of wills, and probably a violation of the letter as well.
“No rule can be worked out which will avoid the dangers of oral evidence on the one hand and which will give effect to the actual intention of the testator in the particular case, on the other.”
Mr. Justice Roberts in his article, after approving the doctrine that an express revocation is not ambulatory, says (p. 520) :
“It is believed that either one of two positions is inevitable; that there is a revocation and no revival; or no revocation, and the former will remains unimpaired. Nevertheless, some courts have, as we have seen, taken cognizance of the intention of the testator, for or against the revival. This may be well enough where no statutory method of revocation is prescribed, and where no formality is necessary to the making of a will, but if a will is revoked by a clause inserted in a later will, then it is gone. However much the testator may intend to set it up again by destroying the revoking will, his intention can have no effect, and a court should not hear evidence of intention, for the statute provides that no testamentary intent is sufficient unless evidenced in certain ways. . . .
“On the other hand, if the second will is no revocation, its destruction as matter of law must leave the first in force, and the fact that testator said he did not intend it to be in force, but intended to die intestate, cannot alter the status of the first will. The statute provides against the revocation of wills by word of mouth, and it nowhere provides that the cancellation of a revoking will, with intent to die intestate, shall revoke a former will. Therefore, evidence of intent is *583immaterial and irrelevant here. On this ground, Flint-ham v. Bradford seems to be erroneous.
“A very different question is raised where the second will contains no clause of revocation, but simply devises the property in a different manner from the first. In such case, if the second will takes effect upon the death of the testator, the first cannot, and, of course, the second does, take effect as being the expression of the testator’s latest intent. Many of the cases recognize this distinction, which is very well brought out in the Michigan cases. In fact, the Michigan Supreme Court takes the most logical and consistent view of the problem, and one which, it is submitted, ought to prevail in jurisdictions where the Statute of Frauds or similar enactments are in force, and where there is no such statute as I Victoria, C. 26, Sec. 22.”
The animadversion on Flintham v% Bradford, supra, .in this passage is not justified. The court in that case did have before it an offer to show certain circumstances and declarations which were so remote and inconclusive as to be ineffectual even if admissible. Of this proposed evidence Coulter, J., says (p. 91) :
“The other rejected evidence is all of the same class and character, and would, in effect, if allowed to prevail, defeat the rule in regard to the effect which the cancellation of a posterior will has on a prior one preserved by testator; and would also amount to a revocation by word of mouth. I will admit that if it was clearly proved that the testator, at the time he can-celled the posterior will, intended to die intestate, but had not the prior will then in his possession or power,. so as to annul or destroy it, that such intent existing at the time of cancellation, and connected with it, might be given in evidence as part of the transaction or res gestse; just in the same manner and for the same reason that a cancellation by accident or misadventure, *584such as throwing a wrong paper in the fire, obliterating the name by mistake, or accidentally throwing ink instead of sand on the paper, could be proved as explanatory of the act. But the intent and the acts evidencing it ought to be contemporaneous with the act of destruction. Because the very preservation of the prior will beyond that period, when in his possession and power, at the time he annulled the posterior one, was evidence of an intent to die testate, and keep a will in readiness to be consummated by his death. It was in fact and in law set up and restored, as if the posterior will had never existed. And it cannot be repealed or revoked; by word of mouth, or subsequent circumstances and declarations, unless those circumstances amount to a revocation in law, such as marriage and birth of a child, alienation of the estate, &c.: Moritz v. Brough, 16 S. & R. 403.”
We are not able to follow the learned justice in his statement that a first will which is not at hand to be destroyed might be revoked by the revocation of a second will plus “intent”. We need not go further into this because there is no evidence in this case that the first will was not at hand, and the first will was preserved for some time by testator in his desk, where he had opportunity to destroy it — which Justice Coulter says is evidence of an intent to die testate.
This passage is quoted because it discloses a necessary qualification of the declarations to show intent, if they be receivable at all, viz, that they must be contemporaneous with' the act of revocation of the second will — part of the res geste thereof. In the Ford case the declarations attended the destruction of the second will and they supported the supposed presumption. To the same effect is Blackett v. Ziegler, 153 Iowa 344, 133 N. W. 901. In Kerchner’s Estate, 41 Pa. Superior Ct. 112, cited by the hearing judge, the declarations *585were contemporaneous and they supported the supposed presumption.
We have been unable to find any case in which the so-called presumption that the prior will is to stand has been overcome by any declarations or' any circumstances, contemporary or otherwise. Such a result does not accord with the theory as to the effect of an inconsistent later will which does not contain an express revocation clause. Revocation by such means is not effected by a writing or by an act of cancellation. It is effected against a presumption, and by parol. This is entirely contrary to the letter and spirit of section 20(a) of the Wills Act.
The possible exception to the foregoing statement may be found in Manning’s Estate, 46 Pa. Superior Ct. 607. That opinion proceeds on the theory that the first will had been “revoked” (whether by an express clause of revocation or not does not appear) and that the destruction of the revoking instrument did not “revive” it, especially in the light of declarations denying an intention to revive. Since Ford’s Estate has cleared the air, we believe that the problem should be approached in a different way, as has been done above.
In the present case there is no evidence as to the act of revocation — no finding and no evidence as to when it was done. There is oral evidence that at the time he signed the will of 19S9 “He said he was going to re-write the will because I had witnessed it in the wrong place, and because he had given us these rings which he had stated he was going to in some clause in the will. He said, ‘Since I have settled that, that need not appear in the will when it is probated.’ ” Another witness testified to substantially the same. This will was undated, but the testimony was that it was executed at Ocean City in the middle of August 1989. On August 22, 1939, testator wrote to the ladies who witnessed the 1939 will, saying, inter alia, “have begun *586new Pages to save time much as possible.” The'hearing judge says, “It is a reasonable conclusion, however, that he cancelled it [the 1939 will] on or about the time the above letter was written.” We think this inference is only a surmise, and that finding of fact on the subject cannot be supported. He began scratching up the 1939 will before he signed it, and no one knows when he finished doing so.
The conclusion of the hearing judge in substance is that the cancellation of the 1939 will was done with the intention of making a new will. If the theory of inconsistent wills is correct, that fact is immaterial. It was canceled. The obstacle to the efficacy of the prior will — which he preserved in his desk — is removed, and it stands as the only executed will he left behind him, until he makes another will or revokes it under the statute. If the result is to admit to probate the remnant of an old will, the fault is that of a very eccentric testator.
The only difficulty in this case results from loose expressions in many opinions that revocation of a second will — sometimes an inconsistent will, and sometimes a will with a revocation clause — plus parol evidence of “intent” — sometimes contemporary, and sometimes not — can revoke or not revoke a prior will. So far as these expressions sanction a revocation of a will by actions of which an essential part is proved by parol, they are all dicta, and we believe that they are unsound. It is a lesson in the difference between decision and dictum that no court, so far as we are aware, has ever taken the responsibility of adjudging that a written will, under the Wills Act, has been revoked by parol.
The exceptions are sustained, and the appeal from the register is dismissed.
Sinkler and Hunter, JJ., dissent.*587 Decree
And now, December 24, 1943, the appeal from the decree of the register of wills admitting to probate the writing dated November 26, 1906, as the last will and testament of Thomas A. Burtt, deceased, is dismissed, and the record is remitted to the register.