On June 11, 1927, Jean Cazaurang executed a will leaving nine-tenths of his estate to his second cousin Marie Lees, now Marie Lees Thomas. On January 2, 1929, immediately to the right of and below his signature on that will he wrote that date and provisions leaving $1 to the son of his niece and $1 to his wife, and beneath that he again signed his name. Between his signature to the original will and the new date he made a rough bracket. It was stipulated that both this original will and this addition thereto were entirely in the handwriting of the deceased. Both portions were written in French and the English translation of the entire document is as follows:

In the meantime, on April 25, 1928, the deceased executed a witnessed will in which he left his undivided one-half interest in a large ranch, which constituted the bulk of his estate, to his wife Marie Cazaurang, pointing out that the other half of this ranch already belonged to her. In that will he revoked all former wills by him made.

Both wills were offered for probate and the admission to probate of the first of these wills, with the addition thereto dated January 2, 1929, was contested by the estate of the decedent’s widow and by Jean B. Canchón. Among other things, the court found that the instrument dated June 11, 1927, with the addition dated January 2, 1929, constituted the last will of said decedent; that the purported will dated April 25, 1928, was revoked by the other will; that the holographic writing dated June 11, 1927, was revived, republished and reexeeuted as the will of the deceased by the holographic writing dated January 2, 1929; and that the two writings together constituted one will and the last will of said deceased. An order and judgment was accordingly entered admitting to probate the first of these wills, with the addition thereto, as the last will of the deceased, and this appeal followed.

The appellants contend that since the will dated June 11, 1927, had been revoked by the witnessed will of April 25, 1928, it could not have been republished or revived by the addition made on January 2, 1929, since that addition did not expressly refer to or adopt the language of the will of June 11, 1927. It is argued that the addition made on January 2, 1929, is a codicil but since it does not refer to the will of June 11, 1927, it necessarily must be taken as a codicil to the will executed on April 25, 1928.

Section 25 of the Probate Code provides that the execution of a codicil referring to a previous will has the effect of republishing the will, as modified by the codicil. In Estate of Plumel, 151 Cal. 77 [ 90 Pac. 192 , 121 Am. St. Rep. 100 ], an attempted holographic will, which was invalid because not properly executed, was held to have been incorporated into and made a part of a valid will by a codicil properly executed which was on the reverse side of the same paper. While the writing on the reverse side of the paper was labeled “codicil” that writing did not otherwise refer to- the invalid will on the other side of the paper. In that case, the court said:

In Estate of Finkler, 3 Cal. (2d) 584 [ 46 Pac. (2d) 149 ], it was held that interlineations in a holographic will were to be given effect though not accompanied by a redating, reexecution and republishing of the instrument.

While the addition with which we are here concerned is not, strictly speaking, an interlineation, with respect to the prior will to which it was attached, no reason appears why it should not be given an equal effect. Moreover, while the word “codicil” was not used in connection therewith the addition was not only written on the same sheet of paper as the prior will and on the same side, but it immediately follows the prior will with a rough bracket between the new date and the signature on the prior will. The addition is dated and signed and attached to the prior will not only physically, but by means of a rough bracket.

In addition to the instrument itself, Marie Lees Thomas testified that this original will of June 11, 1927, was handed to her by the deceased in June, 1927; that he took it back from her some time in December, 1928; that he again handed it to her on January 4, 1929, at which time the portion following the bracket had been added; and that she then retained it until after the testator’s death.

Whether we call this a codicil or a republication and reexecution, the deceased meant something by the words he added to this instrument on January 2, 1929. The • added words disclose a testamentary intent and he meant to make some change in the disposition of his estate. He could hardly have intended it merely as a change in his 1928 will because he made no reference to that instrument by attachment or otherwise, and the provision leaving his wife $1 was inconsistent with the 1928 will in which he left the bulk of his estate to her. It is a justifiable inference and we think the most reasonable inference, both from the instrument itself and the surrounding circumstances, that the testator, in making the addition on January 2, 1929, and in redating and resubscribing the same, intended thereby to adopt the provisions of the instrument to which it was attached, as modified by the addition, and to republish and reexecute the entire instrument as his last will. The manner in which the addition was made was a sufficient reference to, and adoption of, the language of the original will of June 11, 1927. The evidence sufficiently supports the inference and conclusion drawn by the trial court.

The appellants further contend that certain evidence offered by the respondent was impropérly received. Not only does the record disclose that this evidence was not considered by the trial court, but it clearly appears that it could not have affected the result.

The order appealed from is affirmed.

Marks, J., and Griffin, J., concurred.

Appellants’ petition for a hearing by the Supreme Court was denied April 11, 1941. Curtis, J., and Traynor, J., voted for a hearing.