CHIPMAN, C.

Jennie C. McCauley duly executed a will on February 12, 1900, in which she, among others, made several bequests to charitable institutions. On March 16, 1900, *434 she duly executed a codicil to this will. She died April 14, 1900, twenty-eight days after the execution of the codicil. The state, by the attorney-general, filed objections and contest to the petition for final distribution, so far as concerned the charitable bequests, and prayed that they be adjudged void, and that they be distributed to the state for the support of the common schools. The trial court adjudged the said bequests to be valid, and decreed distribution accordingly. The state appeals from the decree. The codicil did not attempt to change any of the charitable bequests or any of the general provisions of the will, but related solely to specific bequests and devises to certain individual legatees. It stated that “the foregoing codicil . . . was, at the date hereof, . . . signed, sealed, and published as, and declared to be, together with the will set forth on the preceding pages, to be her last will and testament,” etc.

Section 1313 of the Civil Code provides as follows: “No estate, real or personal, shall be bequeathed or devised to any charitable or benevolent society, or corporation, or to any person or persons in trust for charitable uses, except the same be done by will duly executed at least thirty days before the decease of the testator.” Section 1287 of the same code reads as follows: “The execution of a codicil, referring to a previous will, has the effect to republish the will, as modified by the codicil.” The testatrix left “no relatives or next of kin,” as she declared in her will, and, as seems to be conceded by respondents, the bequests in question will escheat, if as to such bequests the will is invalid.

Appellant’s contention is, “that the effect of the republication of the will by the codicil of March 16, 1900, and the testatrix dying in less than thirty days thereafter, is to invalidate all the bequests to charity contained in the will.”

Appellant cites numerous cases to the effect that the codicil brings the will to it, and makes it the will from the date of the codicil. Some of the cases speak of the codicil as a republication of the whole will at the date of the codicil. Still others hold that the codicil operates as a republication of the will, the effect of which is to bring down the will to the date of the codicil, so that both instruments are to be considered as speaking at the same date and taking effect at the same time. Payne v. Payne, 18 Cal. 292, (p. 302,) and In re Ladd, *435 94 Cal. 670, are cited as in line with the authorities elsewhere holding as above stated. In the Ladd case it was said that “the execution of the codicil had the effect ‘to republish the will as modified by the codicil’ (Civ. Code, sec. 1287), as of the date of the codicil. (Payne v. Payne, 18 Cal. 302.)” Again, “the effect of its execution was to republish the entire will, and not merely the clause so modified, ‘as if the testator had inserted in the codicil all the words of the will.’ ” (Doe v. Walker, 12 Mees. & W. 597.) In giving construction to the will it was said that “the whole of the original will and the codicil are to be construed as a single instrument executed at the date of the codicil, and of which all the parts are to be construed ‘so as, if possible, to form one consistent whole.’ ” (Civ. Code, sec. 1321.) But it was also said: “A codicil is never construed to disturb the dispositions of the will further than is absolutely necessary for the purpose of giving effect to the codicil. (1 Jarman on Wills, 176.) ‘The dispositions made by a will are not to be disturbed by a codicil further than is absolutely necessary in order to give it effect, and a clear disposition made by the will is not revoked by a doubtful expression or inconsistent disposition in a codicil.’ (Kane v. Astor, 5 Sand. 533.) ‘The different parts of a will, or of a will and codicil, shall be reconciled, if possible, and where a bequest has once been made, it shall not be revoked, unless no other construction can fairly be put upon the language used by the testator. (Colt v. Colt, 32 Conn. 446. See, also, Wetmore v. Parker, 52 N. Y. 462; Johns Hopkins University v. Pickney, 55 Md. 365.)’ ” No one for a moment can suppose that the codicil in the present will was intended to disturb the bequests made in the original will in aid of the charities named. These bequests were not only left untouched by the codicil, but the testatrix declared that “the foregoing codicil . . . was . . . published as, and declared to be, together with the will set forth on the preceding pages, to he her last will.” That the testatrix intended her bequests first made in the will to stand unaffected by the codicil can admit of no doubt, and yet we are asked to give such construction to section 1287 of the Civil Code as shall destroy a large number of her bequests and practically nullify the testatrix’s clearly expressed inten *436 tion with respect to them. We have seen that no such construction can be given to the codicil itself; can we, or, rather, are we compelled to, so construe the statute as to destroy these bequests, and thus thwart the design of the testator ? Section 1292 provides that, “Except in the eases in this chapter mentioned, no written will, nor any part thereof, can be revoked or altered otherwise than: 1. By a written will, or other writing of the testator, declaring such revocation or alteration . . .; or, 2. By being . . . destroyed, with the intent and for the purpose of revoking ...” And section 1317 provides that “A will is to be construed according to the intention of the testator. Where his intention cannot have effect to its full extent, it must have effect as far as possible.” In construing section 1287 we must keep in view the various sections relating to the subject of wills, and must so construe that section as to preserve the letter and spirit of all the provisions of the statute so far as possible. The section should have such construction, if it is possible in reason to do so, as will carry out the known intention of the testator. Section 1313 invalidates the charitable bequest, unless the will is “duly executed at least thirty days before the decease of the testator.” When the will is once “duly executed,” it remains the will of the testator until revoked. This may be done as prescribed in section 1292. A codicil does not disturb the will, except so far as it is inconsistent with it, or in terms, or by necessary intendment, revokes it. As was said in the Ladd case, “Where a bequest has once been made, it shall not be revoked, unless no other construction can fairly be put upon the language used by the testator.”

For some purposes, no doubt, the will speaks from the date of the codicil, but this is true only so far as the codicil requires that it should so speak. It is entirely consistent with the statute.and the codicil now before us, that the contested bequests should stand as made of the date of the will. The testatrix declared the will as first executed to be her will, except as to the changes made in the codicil, and the statute (sec. 1287) says that the effect of the codicil is “to republish the will, as modified by the codicil,” and not otherwise. To construe the statute as is contended for by appellant, is to leave a large part of the estate undisposed of, as well as to *437 defeat the object of the testatrix. We do not think it should be given any such construction.

A statute of Pennsylvania, referred to as the act of April 26, 1855, provided: “No estate, real or personal, shall hereafter be bequeathed, devised, or conveyed to any body politic or to any person in trust for religious or charitable uses, except the same be done by deed or will attested by two credible witnesses, at least one calendar month before the decease of the testator or alienor.” In 1879 an act was passed in that state by which it was provided “that every will shall be construed with reference to the real or personal estate comprised in it to speak and take effect as if it had been executed immediately before the death of the testator.”

In Carl’s Appeal, 106 Pa. St. 635, the will was executed September 6, 1877; a codicil was executed March 14, 1881, and the testator died April 4, 1881. It was contended that the charitable bequests of the will were void. In construing the act of 1879 with that of 1855, the orphans’ court, Gibson, J., speaking of the effect of the republication of a will by a codicil, said: ‘' There is a class of eases about which there can be no contention, such as Neff’s Appeal, 48 Pa. St. 501, in which the question was whether the codicil revoked a second will and republished an earlier will on which it was written. . . . But we find that in all other cases, since the modern statutes of wills, which may arise out of the execution of a codicil, questions regarding the intention of the testator are involved to such an extent as to qualify the rule invoked here materially,” citing instances, in one of which (Alsop’s Appeal, 9 Pa. St. 374) this.distinction is shown: “That though for some purposes a will and codicil are to be regarded as making but one testament, they will not be considered as a single instrument where a manifest intention requires otherwise.” Other cases are cited to show, for example, that the first will would have been revoked by the second will but for the codicil which gave it life, “and yet it did not affect the validity of the charitable bequests in the first will. This could only be on the ground of its being a separate instrument.” (Hamilton’s Estate, 74 Pa. St. 69; Brandish v. McClellan, 100 Pa. St. 607; Neff’s Appeal, 48 Pa. St. 501.)

The learned justice concludes: “The question raised here as to the period from which this will speaks, based upon *438 the doctrine o£ repnblieation by a codicil, would make void the charitable bequest given by the original will, contrary to the intention of the testatrix. The very act of republication, ipso facto, would make null and void that which republication by intendment of law reaffirms. I think the validity of the residuary bequest ... is not affected in any way by the codicil, arid that it does not bring the bequest within the prohibition of the act of 1855.” On appeal, the views of the orhpans’ court were fully indorsed. In further support of the decree it was said: “The fault of the opposing counsel lies in confounding a legal fiction with a physical fact. Of course, all wills must speak as of the time of the testator’s death. It is a pure legal fiction that they were executed at that time. The fact of actual execution remains, and is entirely unaffected by the fiction. Whenever the factum is material, of its own force, in determining results, it will be treated as of the date of its actual occurrence.” Our statute (Civ. Code, sec. 1287) certainly does not by its terms compel the construction urged by the attorney-general. To give to it such construction we must import into the statute a legal fiction. This we might do in some cases, but not where it would result in defeating the clearly expressed intention of the testatrix.

The judgment should be affirmed.

Gray, C., and Haynes, C., concurred.

For the reasons given in the foregoing opinion the judgment is affirmed.

McFarland, J., Henshaw, J., Lorigan, J.