Multiple appeals are taken from a judgment admitting to probate a 1956 holographic will of Carolina Cuneo, deceased.

'When Carolina died in 1958 she left surviving her six children: Julia Cuneo, Clara Mondani, Sarah Carboni, Louise Yarni, Suzanne Dobbert and Andrew Cuneo. Her estate, appraised at $67,698.98, consisted primarily of farm properties, livestock, and cash deposits. Beginning in 1943 she made four successive wills, three of which have been offered for probate. With one exception the testamentary writings are holographic and written in Italian. They provide generally as follows:

A second translation of the foregoing document, submitted in evidence, varies the wording in minor respects only.

In 1957 De Paoli, who had possession of the 1954 will, surrendered it to Henry Mondani, Clara’s husband, at decedent’s request. The will was delivered to decedent who stated that “this is not what I want,” and that she intended to burn it. The original will was not found. However a copy thereof was identified and introduced into evidence.

The 1956 instrument, when found with the 1951 will among decedent’s valuable papers, contained penciled strike outs of the phrases “my will is in the possession of De Paoli, ’ ’ and ‘ ‘ take my will from De Paoli. ’ ’

Following decedent’s death her daughter Sarah first petitioned for letters of administration, alleging that her mother died intestate. The litigation thereafter developed in the following order: Clara filed the 1951 holographic will and petitioned for probate thereof; Andrew filed a contest of the 1951 will; Clara filed the 1956 instrument, alleging it to be a codicil to the 1951 will and petitioned for probate of the two instruments together; Andrew filed a contest of the two instruments ; De Paoli petitioned for probate of the 1954 will, alleging the 1956 instrument to be a codicil thereof, and opposing the other petitions; Andrew filed a contest of the 1954 will and alleged codicil.

After trial of the consolidated causes the court found that the 1951 will had been revoked by the general revocation clause of the 1954 will; that the 1954 will was destroyed by decedent with intention to revoke, and was so revoked; that it was the decedent who made modifications to the 1956 instrument, and as modified it was intended by decedent to remain in effect; that it was testamentary in character and was legally sufficient to stand by itself. The court concluded and ordered that the 1956 instrument alone be admitted to probate and that Clara be appointed administratrix with will annexed.

Turning first to Andrew’s appeal it is his contention that since the 1956 instrument was intended only as a codicil, and since “ [t]he revocation of a will revokes all its codicils” (Prob. Code, § 79), it was therefore revoked by reason of the destruction of the 1954 will, and hence his mother died intestate.

De Paoli, who also appeals, contends that the 1956 instrument was originally intended as a codicil to the 1954 will; that the 1954 will was later destroyed and revoked only under the mistaken belief that the 1956 instrument could be given effect as a testamentary instrument standing alone. He argues further that the 1956 instrument could not, by itself, be given the effect intended by the testatrix because the pretermitted heir statute (Prob. Code, § 90) requires that the estate be distributed in part to those children not mentioned in the 1956 instrument. It is his conclusion that the doctrine of dependent relative revocation requires the revival of the 1954 will, and it should be probated as modified by the 1956 codicil.

Finally there is the appeal by Clara Mondani who also relies on the doctrine of dependent relative revocation. It is her contention that the 1951 will, revoked by the 1954 will, was revived by operation of the doctrine when the 1954 will was revoked and the decedent made modifications to the 1956 codicil.

It is apparent that all contentions raised herein relate, in one manner or another, to the propriety of the trial court’s conclusion that the 1956 instrument became a self-contained will upon modifications thereto and the revocation of the 1954 will.

We cannot disregard the trial court’s finding that the decedent intended the republished codicil to be her sole testamentary instrument. The modifications of the original instrument evidence a desire to divorce it from the 1954 will, and her revocation of the latter must be deemed in aid of her intention that the 1956 instrument stand alone. Although there is no direct evidence that the decedent destroyed her 1954 will, other substantial evidence supports such a conclusion. Similarly, there is no direct evidence as to who made the changes in the 1956 instrument, but where changes and alterations are made in a holographic will in the possession of the testatrix, the inference is that they were made by her. (Estate of Stickney, 101 Cal.App.2d 572, 575 [ 225 P.2d 649 ].) Cancellations, modifications and interlineations made in a testator’s hand in a holographic will do not render it invalid. Eather, they become a part thereof. (Estate of Finkler, 3 Cal.2d 584, 600 [ 46 P.2d 149 ].)

Manifestly, the document is testamentary in character, and while it may have lost something in its translation and perhaps is inexpertly drawn, nevertheless it appears to be sufficiently clear on its face as to decedent’s intentions to render them ascertainable and thus avoid intestacy. . . . [T]he rule is well settled that where the construction given to an instrument by a trial court is reasonable and appears ■to be consistent with the intent of the party making it, courts of appellate jurisdiction will not substitute another interpretation, even though it may seem equally tenable with that accorded by the trial court.” (Estate of Northcutt, 16 Cal.2d 683, 690 [ 107 P.2d 607 ].)

The contention herein urged by Andrew that the 1956 instrument must fall under section 79 of the Probate Code because, as a codicil, the will to which it related was revoked, is not tenable. As a codicil the statute would have required its revocation with that of the will heretofore determined to have been revoked on substantial evidence. But merely because it was intended as a codicil when first published does not compel its continuance as such. A document, originally a codicil, which is a complete testamentary instrument in all respects is not dependent upon the continued existence of its basic will to retain its testamentary character. (Estate of Benson, 62 Cal.App.2d 866, 870 [ 145 P.2d 668 ].) Where the document satisfies the legal requirements of a testamentary instrument the discernible intent of the testator as to which instrument or instruments constitute the final testamentary expression must govern. (Estate of Danford, 196 Cal. 339, 342 [ 238 P. 76 ] ; Prob. Code, § 101.) For reasons heretofore set out we are persuaded that the trial court properly concluded that the decedent intended her 1956 instrument to continue in effect as her final testamentary expression after revocation of the 1954 will.

We are aided to the foregoing conclusion by the established rule that a will must be construed according to the intention of the testator and that effect must be given to the expressed intention so far as possible to avoid intestacy. (Estate of Heard, 25 Cal.2d 322, 326 [ 153 P.2d 553 ]; Prob. Code, § 102.)

As previously noted De Paoli and Clara Mondani urge divergent views as to the effect of the application of the doctrine of dependent relative revocation when applied to the facts of the present case. “Under the doctrine ... an earlier will, revoked only to give effect to a later one on the supposition that the later one will become effective, remains in effect to the extent that the latter proves ineffective. [Citations.] The doctrine is designed to carry out the probable intention of the testator when there is no reason to suppose that he intended to revoke his earlier will if the later will became inoperative. [Citations.] ” (Estate of Kaufman, 25 Cal.2d 854, 858, 859 [ 155 P.2d 831 ].)

The doctrine thus requires that, in revoking a prior and executing a subsequent mil, it be specifically intended that certain provisions in the former testament have a continuing effect, either through similar provisions in the new will or because it is intended to make only conditional changes which subsequently do not become effective for the reason that the conditions on which predicated fail to come into being. It is De Paoli’s contention herein that the 1954 will was thus only conditionally revoked by the testatrix, the condition being that the 1956 instrument be legally sufficient to accomplish the testatrix’s obvious purpose to provide for Clara and Julia to the full extent of the decedent’s estate. This cannot be accomplished, it is claimed, because the pretermitted heir statute requires that each heir not mentioned in the 1956 instrument be distributed the share to which he would be entitled under the laws of succession.

In Estate of Kaufman, supra, 25 Cal.2d 854 , a testator wished to make minor changes not going to the dispositive provisions of the will, and executed a second will for this reason. His death occurred within 30 days thereafter, invalidating a substantial gift to a church under the applicable mortmain statute. The court concluded at page 860: “The record makes clear the intention of the testator. He did not change his testamentary purpose but only minor details in its execution. . . . Since the second will was virtually identical with the first in the disposition of the testator’s estate, it is clear that the first will was revoked only because the second duplicated its purpose and that the testator would have preferred the first will to intestacy as to a substantial part of his estate. Where a testator repeats the same dispositive plan in a new will, revocation of the old one by the new is deemed inseparably related to and dependent upon the legal effectiveness of the new.” In that ease the first will was ordered admitted to probate together with the second.

It is argued that in the instant case the testatrix’s intention that the first will be admitted to probate in the event of the insufficiency of the second is not as clear. The 1956 instrument became effective as a self-contained will only at the time the testatrix republished it as a will, at which time or prior thereto the 1954 will must be deemed to have been revoked. The changes in dispositive provisions between the 1954 will and the republication of the 1956 instrument were to eliminate a $500 gift to a former daughter-in-law; provide at least life estates, if not gifts in fee, for Clara and Julia instead of an estate for Julia alone for her life or until she married; and eliminate provisions for remaindermen interests, whereas originally the six children and a grandchild were specifically named to share therein. A change in the person designated to administer the estate was also made and a clause disinheriting anyone who contested the will was eliminated.

The foregoing are, manifestly, substantial changes and standing alone cannot be said to evidence an intention on the part of the testatrix that the 1954 will was to continue in effect in the event the 1956 instrument could not, as a self-contained will, be effective in accomplishing her purpose. It is significant, too, that upon the delivery of the 1954 will to her from De Paoli she stated that it was not what she wanted, and that she intended to burn it.

Consistent with the foregoing there are, nevertheless, good and substantial grounds for the application of the doctrine of dependent relative revocation, if we recognize that it is not the 1954 will which is sought to be revived, but instead the 1954 will as modified by the 1956 codicil. Prior to her revocation of the 1954 will, and republication of the 1956 codicil as a will, the testatrix had effectively modified the 1954 will with the 1956 instrument, conceded by all parties to then be a codicil. Although such codicil did not formally modify specific provisions of the will, those provisions of the will were nevertheless effectively modified which were inconsistent with the codicil. (Estate of Salmonski, 38 Cal.2d 199, 209 [ 238 P.2d 966 ].) This included the whole of the dispositive plan of the testatrix and the substitution of Clara to “manage” or administer the estate in the place of De Paoli.

In 1957 then, when the testatrix revoked the will, she substituted in its place a testamentary document which made identical provisions for disposition of the estate. It is not significant that the testatrix evidently misunderstood the effect of the 1956 codicil on her will and was thus under an impression that the dispositive provisions of the will continued to be effective. Only such a misunderstanding can account for her statement that the 1954 will was not what she wanted prior to her substitution in its place of the republished 1956 instrument, by which she sought to make the identical disposition as provided in the 1954 will as modified. Thus the will, as modified by the codicil, was what she wanted regardless of whether she so understood it to have that effect. Any conclusions as to her intentions when she revoked the 1954 will and republished the 1956 codicil as a will must be considered in the light of her knowledge at that time. (Estate of Kaufman, supra, 25 Cal.2d 854, 861 .)

The situation herein is not distinguishable in principle from that in the Kaufman case. The testatrix, in 1957, did not change her dispositive plan, but only minor details in its execution. As the 1956 instrument, republished as a will, was virtually identical with the 1954 will as modified, it is clear that the latter was revoked only because the testatrix thought that, by the republished instrument, she could accomplish what the 1954 will as modified did in fact accomplish. It must be deemed incontrovertible that the testatrix would have preferred the 1954 will, as modified, to intestacy as to a substantial portion of her estate. “When a testator repeats the same dispositive plan in a new will, revocation of the old one by the new is deemed inseparably related to and dependent upon the legal effectiveness of the new. [Citations.] ” (Estate of Kaufman, supra, 25 Cal.2d 854, 860 .)

It affirmatively appears that the 1956 instrument as republished by the testatrix, standing alone, will fail to accomplish her intended purposes. Those purposes are discernible over a period of years commencing in 1943. Since that time the testatrix has evidenced a continuous, consistent testamentary plan wherein she provided the bulk of her estate for Julia, her incompetent daughter, and Clara who was to care for Julia. This plan was reasserted when the testatrix published the 1956 codicil to the will, and again when she republished the codicil as a will in 1957. However, the 1956 instrument, standing alone, fails to mention four of the decedent’s six issue. Section 90 of the Probate Code imposes intestacy upon a testatrix as to any child who is not provided for by the testator “in his will” unless it “appears from the will” that the omission was intentional. Extrinsic evidence is not admissible to establish the intent to omit. (Estate of Trickett, 197 Cal. 20 [ 239 P. 406 ].) On the other hand, if the 1954 will as modified is revived and admitted to probate together with the 1956 instrument as republished, the testatrix’s intentions with respect to all her issue expressly appear within the testamentary documents, and such children do not achieve the status of pretermitted heirs. (Payne v. Payne, 18 Cal. 291 .)

It is also urged herein that the 1951 will was revived by application of the doctrine of dependent relative revocation or otherwise, and that it should be admitted to probate together with the 1956 “codicil” thereto. The 1951 will was revoked by the 1954 will, and there is no substantial evidence which requires a conclusion that at such time the revocation was conditionally dependent upon the sufficiency of the 1954 will. The further facts that decedent kept the 1951 will with her other valuable papers and that after the strike outs in the 1956 instrument there still remained an implication that the document was a codicil, do not require that the trial court should have found, as a matter of law, an intent to republish the 1951 will through the alleged codicil thereto. In any event the admission of the 1951 will to probate would serve no purpose, in view of the conclusion herein that the later documents are to be admitted.

There remains unresolved at this time the question whether the dispositive provisions of the 1956 instrument conferred upon Julia and Clara life estates or full fee interests. The trial court made no findings thereon, and properly so. “ When there is presented for probate one or more instruments claimed to constitute a will, the only question legitimately before the court is whether or not the propounded instrument or instruments constitute a will; and in determining that question the court ordinarily has nothing to do with the construction of the will, resolving inconsistencies in the disposition of property or interpreting the testamentary provisions.” (Estate of Salmonski, supra, 38 Cal.2d 199, 207 .)

The judgment is reversed with directions to admit the 1954 will to probate with the 1956 codicil as republished in 1957.

Gibson, O. J., Traynor, J., Schauer, J., McComb. J., Peters, J., and Tobriner, J., concurred.