delivered the opinion of the Court.
This appeal raises two issues of first impression in the Commonwealth: (1) whether the applicable statute permits partial revocation of a formally attested will by marks placed upon it by the testator with intent to partially revoke the will, and (2) whether the presumption that a testator marked on his will with intent to revoke portions of it can arise where two duplicate originals were in the possession of the testator from their execution until found after his death and one is marked while the other is not. The trial court ruled in the affirmative on each issue deciding, in effect, that the testator successfully “cut” Carolyn Etgen out of his will. Etgen appeals.
The facts were stipulated. Frank I. Whitten, Jr., the testator, died on March 2, 1978. In oral argument, all counsel represented that he died by his own hand. On September 6, 1977, the date his will was executed, Whitten was cohabiting, in his home, with Etgen. However, a short time prior to Whitten’s death, Etgen moved out and the relationship between the two was severed.
The will propounded by the administrator c.t.a. of the marked will, J. Grant Corboy, was found among the personal effects of Whitten at his home. According to the stipulation, the marked will “was in the custody of the testator, Frank I. Whitten, Jr., *415after its execution until it was found among his personal effects after his death.”
When the marked will was found, two paragraphs concerning Carolyn Etgen were marked through in pencil though the paragraphs remained legible. All of the sixth paragraph and much of the thirteenth paragraph were so marked. The sixth paragraph originally provided as follows:
Sixth. I give the entire balance of my estate, both real and personal, to my friend Carolyn Elizabeth Etgen, who shares my household here on rural route 1 out of Narrows, Virginia. If she predeceases me, or is engaged or married to another man as of the date of my death, I assign my property as follows in sections seventh through twelfth.
This paragraph was marked through in its entirety with diagonal marks. The thirteenth paragraph originally provided as follows:
Thirteenth. I hereby appoint my friend Carolyn Elizabeth Etgen executrix of this my last Will and Testament. If she predeceases me, or is engaged or married to another man as of the date of my death, I appoint her brother, Michael W. Etgen, of Shawsville, Virginia, as executor of my Last Will and Testament.
(Emphasis added.) Everything was marked through with horizontal lines except the language emphasized. Further, “executrix” was changed to “executor.” In addition to the changes in paragraphs 6 and 13, all paragraphs after the fifth paragraph were renumbered. However, there were omissions in the numbering sequence. For example, the renumbered paragraphs skip from ten to thirteen. It was stipulated that at the time the marked will was discovered, no other will was found.
On August 30, 1978, the marked will was admitted to probate by the Clerk. In the probate order, the court noted that the subscribing witness called to prove the will testified that “no pencil interlineations were present on the paper writing when he and the other subscribing witness, witnessed the testator’s signature.”
Because of the marks on the will, the administrator c.t.a. sought instructions from the court regarding the distribution of the proceeds of the estate. As indicated above, the case was submitted to the court on stipulations. All counsel agreed that the issue was the *416legal effect of the interlineations and deletions found on the marked will, “with regard to whether the testator effected a partial revocation or total revocation of his Will or whether the interlineations were of no effect.” In a letter opinion dated November 4, 1981, the trial court ruled that Whitten had effected a partial revocation which, in essence, eliminated Etgen as a beneficiary.
Thereafter, in January 1982, a safe deposit box belonging to Whitten was discovered. Since no key was found for the deposit box, it was drilled open. Inside was a duplicate original of the marked will. This version of the will was identical to the marked will except it bore no markings.
Upon discovery of the unmarked will, Etgen’s counsel requested and received a hearing on the effect of the newly discovered writing upon the trial court’s earlier ruling of partial revocation. The hearing on the unmarked will was held on July 7, 1982. No evidence was presented except the unmarked will.
In an order dated September 7, 1982, the trial court confirmed its earlier ruling. The court said that the discovery of the unmarked copy of the will had no effect on the previous ruling. The court ruled that Code § 64.1-58 permits partial revocation of an attested will by cancellation and that Whitten “intended to and did effect a partial revocation of his will which eliminated Ms. Etgen as a beneficiary.”
On appeal, Etgen attacks the trial court’s order on two grounds. First, she argues that the statute and the cases do not permit partial revocation of an attested will without further formalities. Second, she contends that the evidence was insufficient to establish a partial revocation.
To resolve the first issue, we must analyze the applicable statute, for it is a question of statutory law whether partial revocation of an attested will can be accomplished in the Commonwealth without further formalities. At the time the marked will was probated, Code § 64.1-58 provided, in pertinent part, as follows:
No will or codicil, or any part thereof, shall be revoked, unless by a subsequent will or codicil, or by some writing declaring an intention to revoke the same, executed in the manner in which a will is required to be executed, or by the testator, or some person in his presence and by his direction, cutting, tearing, burning, obliterating, canceling, or destroy*417ing the same, or the signature thereto, with the intent to revoke.
(Emphasis added.) The words emphasized above clearly indicate that the legislature intended to permit partial revocation of an attested will. In Annot., 62 A.L.R. 1367, 1392-93 (1929), the authors discuss partial revocation of attested wills and write as follows:
[T]he matter of the revocation of a will in toto or pro tanto is statutory, depending largely upon the language of the statute. ... If the prohibitory provision only designates “wills,” ... a pro tanto revocation is not authorized. On the other hand, if this provision designates, in effect, “wills or any portion or clause thereof,” a revocation pro tanto is authorized, providing the act of the testator is one of the methods designated by the statute to revoke a will, or part thereof.
Our statute, as it then existed, was of the kind that permits partial revocation.1
Etgen cites two Virginia cases which, according to her, make clear that this Court has interpreted Code § 64.1-58 and its predecessors as not allowing partial revocation of attested wills without further formalities. Yet, those cases are distinguishable.
Etgen relies on Triplett’s Ex’or v. Triplett, 161 Va. 906, 172 S.E. 162 (1934), for the proposition that changes made on an attested will prior to its execution can be probated while changes made after execution can not be probated. Triplett, however, involved a holographic will which was signed in the presence of witnesses. Though it was not admitted to probate as an attested will, it was admitted as a holographic will. Consequently, the discussion of the will as an attested will was dictum. Nevertheless, that dictum does not support Etgen’s argument. In Triplett, the testator sought to insert new language. Here, the changes are essentially cancellations, an effort to delete one beneficiary. As a result, the rule referred to in Triplett has no application here.
Moreover, other language in Triplett establishes that this Court has long recognized that partial revocation of an attested will could, in certain circumstances, be achieved without further for*418malities. In the opinion, the Court said the following about changes made to attested wills:
“ ‘When a will is presented for probate and discloses interlineations and erasures, it is obvious that, if they were made before the will was executed, they form a part of it, and the document, as so changed and altered, is the one which should be received; while, if they were made after execution, they form no part of the will and are of no effect whatever, unless they sufficiently support an inference of cancellation.’’ ”
161 Va. at 917, 172 S.E. at 166 (quoting City National Bank v. Slocum, 272 F. 11, 16 (6th Cir.), cert. denied, 257 U.S. 637 (1921)) (emphasis added). Here, Etgen has failed to focus upon the italicized language. That language makes clear that though post-execution additions to an attested will are meaningless, the same is not true of post-execution cancellations. In short, Triplett does not support Etgen’s argument that partial revocation of a formally attested will is impermissible in the Commonwealth.
Nor does Harris v. Wyatt, 113 Va. 254, 74 S.E. 189 (1912), support Etgen’s position. That case involved a formally attested will. A will and codicil were offered for probate. The codicil specifically revoked clause five of the will. But the will bore erasures of certain language in the residuary clause. In the probate order, the trial court noted that certain words had been erased from the will. The clerk thought the order meant that the erased words were to be omitted from the will book. A suit followed. The trial court ruled that the probate order did not exclude the erased words and said further that “it would have been erroneous to have excluded them.” 113 Va. at 258, 74 S.E. at 190. This statement might appear to suggest a rule against partial revocation of attested wills but that suggestion dissipates once the reasons for the statement are explored.
The reason the erased words had to be probated was the lack of evidence that the testator made the erasures. We wrote as follows on that point:
The paper offered for probate, including the italicized words, and proved to have been properly executed as a will, was to be treated by the probate court as the last will and testament of the deceased, in the absence of some evidence that the cancellation was done by the deceased, or by some *419person in his presence and by his direction, which would be sufficient to show that fact, or that the instrument was found after the testator’s death among his repositories in the mutilated condition it was in when offered for probate, and under such circumstances that the fact of revocation might, to the extent of the cancellation, be presumed.
Id. at 259, 74 S.E. at 190 (emphasis added). The last phrase, “to the extent of the cancellation,” is further recognition by this Court that there can, in proper circumstances, exist a partial revocation, by cancellation, of a formally attested will.
We hold, with regard to the first issue, that Code § 64.1-58, as it existed at the time the marked will was probated, permitted partial revocation of a formally attested will through cancellation by the testator, with intent to partially revoke. Therefore, the trial court’s ruling on the first issue will be affirmed.
Etgen’s second argument is that even if partial revocation was permitted by the statute, the evidence in this case was insufficient to establish that Whitten made the cancellations with the intent to revoke. We agree.
The trial court’s ruling that the testator cancelled portions of his will with intent to partially revoke the will is based on a presumption which the court determined arose from certain facts stipulated by all counsel prior to the discovery of the unmarked will. The stipulation was that the marked will was in the testator’s custody “after its execution until it was found among his personal effects after his death.” In the usual case, a stipulation such as this would give rise to the following presumption: Where a will is shown to have been in the custody of the testator after its execution and is found among his effects after his death mutilated, obliterated, cancelled or otherwise altered as provided by statute then, in the absence of evidence to the contrary, it will be presumed that such act or acts were performed by the testator with the intent to revoke. See, Jessup v. Jessup, 221 Va. 61, 267 S.E.2d 115 (1980); Timberlake v. State-Planters Bank, 201 Va. 950, 115 S.E.2d 39 (1960); Ruth v. Jester, Adm’r., 198 Va. 887, 96 S.E.2d 741 (1957); Franklin v. McLean, 192 Va. 684, 66 S.E.2d 504 (1951).
However, this case is not usual. This case involves duplicate originals of the same will, one marked, the other unmarked, and both in the possession of the testator. The presumption described *420above does not address a situation where two wills have been in the custody of the testator after execution and both are found among his effects after his death with one altered and the other in its original form.
The facts here do not differ materially from a situation where a testator dies clutching duplicate originals, one in each hand, one marked, the other unmarked; which will should be presumed to be the true will? In our opinion, the answer is neither. In a situation of the kind existing here, it is illogical to conclude that either will is entitled to greater deference than the other. Nothing about the circumstances of the discovery of the two duplicate originals permits a court confidently to conclude that one will is more meritorious than the other. The only reasonable conclusion is that neither will is entitled to the evidentiary advantage provided by the presumption the trial judge here applied.
The trial court concluded that the discovery of the unmarked will had no bearing upon the court’s previous ruling that the marked will was the true will and that the marked will had been partially revoked. The trial court reached this conclusion because of the stipulation concerning the possession and discovery of the marked will. That conclusion was erroneous. The discovery of the unmarked will does not contradict the stipulation concerning the marked will; it merely provides new evidence that not only the marked will but also the unmarked will was in the testator’s possession from execution until discovery after his death.
We hold, with regard to the second issue, that where duplicate originals of a formally attested will are in the possession of the testator from the time of execution until discovery among his effects after death and one version is altered while the other is in its original condition, then neither will is entitled to a presumption that it is the true will of the testator. In such a situation, the proponents of the different versions of the will must prove that their version is the true will.
With regard to the evidentiary holding, we are unpersuaded by Corboy’s reliance upon Blalock v. Riddick, 186 Va. 284, 42 S.E.2d 292 (1947). That case involved duplicate originals. There, one copy of the will was in the testator’s possession while the other was in the possession of the testator’s lawyer. In that situation, we wrote as follows:
*421Where a will has been executed in duplicate, and a copy has been retained by the testator and a copy left in the custody of another person, if the copy retained by the testator cannot be found after his death, there is a rebuttable presumption that he has destroyed his copy for the purpose of revoking the will. Since the revocation of a will necessarily revokes a copy thereof, the copy left in the possession of. the other person may not be admitted to probate, unless or until the proponent proves that the copy retained by the testator was not destroyed by him animo revocandi.
Id. at 292, 42 S.E.2d at 295 (emphasis in original). Blalock, however, is readily distinguishable. There, the facts established that the testator was in possession of only one copy of his will. Here, the testator was in possession of both copies. As our prior discussion makes clear, this factual distinction prevents application of the normal presumption. It also means that there is no copy in the possession of another person which can be deemed to be revoked or cancelled to the same extent as the copy in the testator’s possession. Blalock has no application here.
In light of the foregoing rulings, we find it necessary to remand this case. At the time of trial, Corboy relied on a presumption to which he was held entitled by the trial court. Neither party develop any evidence concerning the placement of the marks on the marked will or the testator’s intent with respect thereto. Given our disposition of this appeal, we think the respective parties are entitled to offer such relevant evidence on these points as may be admissible. Therefore, the trial court’s judgment will be affirmed in part, reversed in part, and remanded for further proceedings consistent with this opinion.
Affirmed in part, reversed in part, and remanded.
Code § 64.1-58 was repealed in 1985. The new provision, Code § 64.1-58.1, is materially different.