Deemer, J.
Elizabeth W. Lewis, now deceased, made a will in the year 1895. Thereafter, and in 1904, she made another, which disposed of her estate, and contained a clause revoking the will of 1895. She retained both wills until the year 1905, when she destroyed the one of 1904 by burning it, and about the same time executed the following paper: “To the Judge of the District Court of Eayette County, Iowa: I hereby request you in case of my death to appoint S. B. Ziegler, of West Union, Iowa, as administrator of my estate without giving bonds.” This she signed, and her signature was witnessed by two witnesses. The will of 1904 was left in the possession of a Mr. Preston, who drew it, until December of that year, when testatrix requested Mr. S. B. Ziegler, who had drawn the first will and had it in his possession, to get the second one from Preston, which he did, and Ziegler took both wills to Mrs. Lewis’ home, where each one was read to her, clause by clause, in the presence of Mr. and Mrs. Caldwell, the persons who witnessed the paper which we have heretofore set out, and after some discussion Mrs. Lewis said that she wanted the first will to stand, and she directed Ziegler to destroy the second one, which he did in her presence, and in the presence of the Caldwells. After the destruction of the second will, Mrs. Lewis said to Mr. Ziegler: “There is no executor in this will that stands. ... I want you to act as my executor,- you have always done my business for a great many years, and I want you to act.” Mr. Ziegler made the objection that he did not care to act as executor of the will, because of the necessity of giving a bond, whereupon Mrs.. Lewis said that he did not need to give a bond. The writing was then prepared by Mr. Ziegler, and was executed and witnessed as shown. This so-called codicil does *347not seem to have been attached to the will, but it was kept by Mr. Ziegler, and was presented for probate with the will. The first will was admitted to probate, and. this action is brought to set aside the order. The trial court denied the relief asked, and plaintiff appeals.
The questions presented are new to this court and some of them are the subject of many conflicting and irreconcilable decisions. In the absence of statute governing some of the matters arising upon the appeal, it may be said there is no general rule, and that each court for itself has found it necessary to fix the rule for its jurisdiction. The relevant statutes of this state are as follows:
All other wills, to be valid, must be in writing, signed by the testator, or by some person in his presence and by. his express direction writing his name thereto, and witnessed by two competent persons; but if a codicil is duly executed to a will defectively executed and clearly identified in such codicil, the will and codicil shall be considered one instrument and the execution of both sufficient. Code, section 3274.
Wills can only be revoked in whole or .in part by being canceled or destroyed by the act or direction of the testator; with the intention .of so revoking them, or by the execution of subsequent wills. When done by cancellation, the revocation must be witnessed in the same manner as the making of a new will. Code, section 3276.
Section 3274 is quoted because of its bearing upon the claim that the paper heretofore set out is a codicil to the first will, and, having been executed after the destruction of the will of 1904, it amounted to a republication of the first will. As to that, more hereafter.
It is admitted that the first will was never destroyed by the maker, and it is also conceded that the second will contained an express revocatory clause of the first will, and that this second will was absolutely destroyed by burning. These being the undisputed facts., the questions involved are: (1) Was the first will republished by the paper hith*348erto quoted, which, it is claimed, is a codicil to that will? (2) Did the destruction of the second will under the circumstances disclosed amount to a revivor of the first will?
I. Wills: revivah repubHcation. Every one concedes that a will expressly revoked by a subsequent will or other instrument of revocation may be republished or revived by the re-execution thereof, or by a codicil executed in accordance with » /» i • j? statutory requirements for the execution ox x t t 4 wills, showing an intention to revive the same. When done by a codicil, an intent to republish or revive the former will must be shown, and this may be inferred from any reference which makes such intent obvious, as, by reference to “my will,” or to the will by date. Crosbie v. McDonal, 4 Ves. 610; Payne v. Payne, 18 Cal. 291. And the codicil need not be attached to the will. Van Cortlandt v. Kip, 1 Hill (N. Y.) 590; Pope v. Pope, 95 Ga. 87 (22 S. E. 245) ; Appeal of Wikoff, 15 Pa. 281 (53 Am. Dec. 597). But, if not attached, there must be such reference to the will intended to be republished as to identify it, or to furnish the means for identification, without resort to any other testimony, save to show that the document sought to be incorporated is identical with that referred to in the will. The codicil itself must refer to the paper sought to be incorporated, if it be then in existence. Newton v. Seamen’s Society, 130 Mass. 91 (39 Am. Rep. 433) ; Damon v. Bibber, 135 Mass. 458; Parrott v. Avery, 159 Mass. 594 (35 N. E. 94, 22 L. R. A. 153, 38 Am. St. Rep. 465) ; In re Sohor, 78 Cal. 477 (21 Pac. 8); Crosby v. Mason, 32 Conn. 482; Booth v. Church, 126 N. Y. 215 (28 N. E. 238); Allen v. Maddock, 11 Moo. P. C. 427 (4 Gray’s Cases, 198) ; Goods of Sunderland, 4 Gray’s Cases on Property, 217; In re Young’s Estate, 123 Cal. 337 (55 Pac. 1011); In re Andrews’ Will, 162 N. Y. 1 (56 N. E. 529, 48 L. R. A. 662, 76 Am. St. Rep. 294).
This is doubtless the rule intended to be announced *349by Code, section 3274, hitherto quoted. In other words, all wills'must, as a general rule, be in writing, duly signed and attested; and if a codicil is relied upon for a publication it must clearly identify the will, and parol testimony is not admissible, in the absence of any attempt to identify the will in the codicil. These being the rules announced by all of the authorities, it is apparent that the written instrument executed by Mrs. Lewis, either contemporaneously with or after the destruction of the second will, cannot be treated as a codicil to the first one, because it was not attached to nor did it refer in any manner to the prior will.
Moreover, the paper itself does not indicate any in-, tent on the part of the maker to revive a former will. Construed without reference to the other testimony, it indicates a thought on the part of the maker that Ziegler should act as administrator of her estate without bond. If it means anything, this would seem to indicate that the maker wished her estate to be administered upon according to law, and not under any will. Manifestly this so-called codicil can not be considered a republication of the first will.
2. sumption6as to revival. The second question presented is much more difficult of solution. Shortly stated, it'is this: Does the destruction or cancellation of a second will, containing an express revocation of a former one, in itself revive the fii’st or former one? Upon no subject _ , x 0 relating to the law of wills are the author!ties in such hopeless and irreconcilable conflict. A learned text-writer has thus stated the law of England as it existed prior to Jjie adoption of the statute known as 1 Viet. ch. 26, section 22:
The English law prior to statutes upon this difficult question was in great confusion. The ecclesiastical courts seemed disposed to hold, in cases of testaments, that no presumption arose, either for or against the validity of the first will, upon such a state of facts, and that the ques*350tion was to be settled by the intention of the testator as disclosed by the evidence. The common-law tribunals in dealing with wills were inclined to adopt the theory that the revocation of the second will raised a presumption that testator thereby intended the first will to be in full force and effect. This was a prima facia presumption only, and might be rebutted by evidence of a contrary intention. The two sets of tribunals thus seemed to agree that the testator might revive his first will by the revocation of his second, if he intended to do so. Further doubt, however, arises upon attempting an analysis of the earlier English cases for two different reasons: First, it is not always clear whether the English courts are discussing a case where the second will expressly revoked the first, or where it was merely inconsistent with it. Second, in many of the cases, especially in the ecclesiastical courts, the declarations of the testator might have been sufficient to republish his first will, as no set form was required for the execution of wills of personal property. It is therefore at times hard to determine whether the first will is valid, because it has been republished after the revocation of the second will, or whether the mere revocation of the second will, with intent to revive the first, revived it without republication.
This condition of uncertainty upon an important and often occurring question was -ended in England by the statute 1 Viet, chapter 26, section 22, which provides in substance that a will once revoked can be revived, but by a new codicil, or re-execution. This statute has always been held to apply with equal force to a will revoked either by a later will containing a clause of revocation, or by a later inconsistent will. Where such a statute is in force, the revocation of a later will by a testator who intends thereby to revive his earlier will, and who so declares his intention, has no effect to revive his earlier will, unless there is a reexecution or republication, as contemplated by the statute. Page on Wills, sections 211, 212.
The same writer, in speaking of the law in this country, said:
In the United States, in the absence of a statute on this subject, the decisions are by no means uniform. The *351better line of authority made a distinction between the cases where the later will contained an express revocation clause, and where it was merely inconsistent with the earlier will. ‘There, seems to have been 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 the effect to do away with its predecessor. Being cut off before having its disposition of property awakened into life, it could have no affirmative operation through its dispositions upon the estate.’ Where such distinction is recognized, the destruction of a later will inconsistent with an earlier will, but containing no clause of express revocation, revives the first will. Where the second will contains a clause of revocation, it is held in many jurisdictions in the United States, in accordance with the distinction already given, that the destruction of the second will doe's not revive the fi*st will. Page on Wills, section 273.
Authorities are cited in support of these views which need not be reproduced here, for we regard the statements of the text a correct exposition of most of the cases cited. See, also, Gardner on Wills, 271-274. Any reference to the authorities upon the subject would be incomplete without incorporating therein a citation to two cases which may well be regarded as leading ones upon this perplexing problem. These are Pickens v. Davis, 134 Mass. 252 (45 Am. Rep. 322), and Williams v. Miles, 68 Neb. 463 (94 N. W. 705, 96 N. W. 151, 62 L. R. A. 383, 110 Am. St. Rep. 431, 4 Am. & Eng. Ann. Cas. 306). Opinion by *352Pound, C. This latter opinion carefully reviews all the decisions, both English and American, and finally adopts the rule of the ecclesiastical- courts, making revivor of the former will a question of testator’s intent, to be deduced from all the circumstances of the case. There is a valuable note to the case in 4 Am. & Eng. Ann. Cas., commencing at page 313. We may remark, parenthetically, that the English statute referred to, which was passed in the year 1837, is no part of our written law, and can not be considered as a part of the common law which we-inherited at the time of the Revolution.
3' íation ofaw¡Hs: It should also be noted that we have no local statute upon the subject of the effect to be given the destruction of a second will upon a former one, which is yet preserved. We have already copied the section of our Code with reference to the revocation of wills, and it is deemed proper, in view of some claims made by counsel, to indicate our views with reference to the proper interpretation of that statute. In the first place, it indicates that wills may be revoked, either in whole or in part. Next it indicates that revocation may be made by cancellation or destruction, or by the execution of subsequent wills; and the cancellation referred to is a written one, which must be executed in the same manner as the making of a new will. This- may bo then simply an instrument of cancellation or revocation, or it may be a part of another and subsequent will, which contains an express clause of revocation. The word “cancellation” may have had a different meaning at common law, ■but in our statute it manifestly means revocation by a written instrument. Again, physical destruction of a will amounts to a revocation, when so intended by the testator. Again, a will may be revoked under this statue, either in whole or in part, by the execution of subsequent wills. This follows when there is an inconsistency between the two wills, and there is no express revocatory clause. See, *353as sustaining this view, Fry v. Fry, 125 Iowa, 424; Schillinger v. Bawek, 135 Iowa, 131; In re Will of Dunahugh, 130 Iowa, 692; McCarn v. Rundall, 111 Iowa, 406; In re Browns Will, 143 Iowa, 649. These matters are pointed out in order that we may distinguish, harmonize, explain, and, perhaps, criticise some of the cases relied upon by appellee. If a will be revoked by destruction, as by burning, and a second one is executed in its place, there is nothing to revive upon the destruction or revocation of the second one. A will once executed may be revoked by the execution of an instrument of revocation or cancellation, and this instrument may be a new will, containing an express clause of revocation, or by an instrument of revocation alone.
4. Same. Upon the execution of such an instrument, the prior will is revoked, no matter whether the instrument of revocation be probated or not. It is the execution of the instrument in proper form which effectuates - . , , . . the revocation. I his view is sustained by reason and the weight of authority, although disapproved by a minority of the courts. See, as sustaining the rule, some of our own cases already cited and the following from other jurisdictions: In re Barnes’ Will, 70 App. Div. 523 (75 N. Y. Supp. 373); Brown v. Brown, 8 El. Bl. 876 (92 Eng. C. L. 876); Matter of Cunningham, 38 Minn. 169 (36 N. W. 269, 8 Am. St. Rep. 650); Stevens v. Hope, 52 Mich. 65 (17 N. W. 698) ; Cheever v. North, 106 Mich. 390 (64 N. W. 455, 37 L. R. A. 561, 58 Am. St. Rep. 499); Marsh v. Marsh, 48 N. C. 77 (64 Am. Dec. 598). Clearly this is the doctrine heretofore announced by us in McCarn v. Rundall, and In re Dunahugh’s Wills, supra.
5. SAME’ If revocation, either in whole or in part, is to be implied from the execution of a second will, this revocation does not become effective if the second will is destroyed or revoked before probate, for the reason that every will, as such, is ambulatory in *354character, and if not in existence at the time of testator’s death, and there being nothing to probate except the original will, there is no inconsistency and no revocation by implication. This is the doctrine of the Connecticut and other courts, although, as we think, they unduly extend it by making it apply to instruments expressly revoking prior wills. Doubtless the reason for this was to find some ground upon which to base the doctrine of revivor, although not calling it by that name. See Peck’s Appeal, 50 Conn. 562 (47 Am. Rep. 685), as explained in Blakeman v. Sears, 74 Conn. 516 (51 Atl. 517) ; Stetson v. Stetson, 200 Ill. 607 (66 N. E. 262, 61 L. R. A. 258) ; Flintham v. Bradford, 10 Pa. 90; Randall v. Beatty, 31 N. J. Eq. 643; Sewall v. Robbins, 139 Mass. 164 (29 N. E. 650).
6. Same: revivor: evidence. In the instant case, there was an express revocation of the former will in the one executed in the year 1904. But this second will, containing the revocatory clause, was destroyed by burning. The first will had not destroyed, but had been kept by the testatrix, and at the time of the burning of the second there is no doubt, under the testimony, that she intended to revive the first. To be logical and consistent, we must hold that the second will, when executed with its clause of express revocation, revoked or canceled the first will, and we are required now to formulate a rule with reference to the revivor of the first will by reason of the destruction of the instrument of revocation. The safest doctrine, we think, is that announced by the ecclesiastical courts of England, to the effect that it is a question of testator’s intent, to be gathered from admissible parol testimony. It would not do to hold that the former will was absolutely revived by the destruction of the second, for that may have been entirely contrary to testator’s intent. Having made the second will, and laid aside and, perhaps, forgotten the first, it would be dangerous to hold that the destruction of the second ipso facto revived the *355first, no matter if testator did not so intend. On the other'hand, it would in many cases4 frustrate testator’s intent, were we to hold that the former will could only be revived in such case by a re-execution or a republication thereof after the destruction of the second will, which had not been admitted to probate, and which could not be, because of its destruction. There is no danger, then, it seems to us, in holding it' to be a question of testator’s intent, to be arrived at from all the circumstances in the case. Testimony to establish such intent could only come from disinterested witnesses, and we can perceive of no harm which would result in submitting such question as one of fact.
In the construction of wills, testator’s intent has always been regarded as controlling, and so with reference to what should be regarded as his will. There can be no valid objection to a rule leaving the question of revivor in such cases as this to be found as a matter of intent upon permissible parol testimony.
Declarations of testator at' the time of revoking a will have generally been admitted, when testified to by disinterested parties. Boyle v. Boyle, 158 Ill. 228 (42 N. E. 140); Behrens v. Behrens, 47 Ohio St. 323 (25 N. E. 209, 21 Am. St. Rep. 820) ; Steinke’s Will, 95 Wis. 121 (70 N. W. 61). Even where a contrary rule prevails, admissions are admissible when part of the res gestae. Caeman v. Van Hacke, 33 Kan. 333 (6 Pac. 620) ; Hayes v. West, 37 Ind. 21; Waterman v. Whitney, 11 N. Y. 157 (62 Am. Dec. 71) ; Townshend v. Howard, 86 Me. 285 (29 Atl. 1077) ; Pickens v. Davis, supra; Williams v. Williams, 142 Mass. 515 (8 N. E. 424). 'These are' the conclusions announced in Barksdale v. Hopkins, 23 Ga. 332; McClure v. McClure, 86 Tenn. 174 (6 S. W. 44) ; Carpenter v. Miller, 3 W. Va. 174 (100 Am. Dec. 744) ;. In re Gould, 72 Vt. 316 (47 Atl. 1082) ; Rise v. Scott, 88 Minn. 386 (93 N. W. 109); Colvin v. Warford, *35620 Md. 357; Lane v. Bill, 68 N. H. 275 (44 Atl. 393, 73 Am. St. Rep. 591).
We do not think there is any presumption one way or the other from the destruction of the instrument of revocation. The whole matter is one of fact, dependent upon the testimony which may be offered to show testator’s intent. This is the rule announced by the later and better authorities, as shown in Williams v. Miles, supra, and the one best calculated to effectuate justice. It is the rule by statute in New York and Indiana. See In re Forbes’ Will (Sur.) 24 N. Y. Supp. 841; Kern v. Kern, 154 Ind. 29 (55 N. E. 1004). The result of its application to the case at bar is to affirm the judgment of the court below, and it is so ordered. — Affii'med.
Sherwin, C. J., concurs in the result reached herein.