Kinne, J.
{dissenting). — I. I can not agree to the result reached by the majority of the court in this case. With all due respect, it seems to me that the opinion to a certain extent ignores the issue as presented in the case, and, while conceding that there could be no reformation of the will on the ground of mistake, accident, or surprise, the conclusion reached is attempted to be justified on the theory that, after rejecting the false description, sufficient remains to lead to the identification of the land in controversy. It is said to be a case of latent ambiguity. I do not so regard it. To properly understand the case as presented it becomes *60necessary to fully state the facts. This is an action of right at law to recover possession of the “southwest quarter of section fourteen (14), township ninety-eight (98), range seventeen (17), in Mitchell county, Iowa.’7 It is prosecuted by plaintiffs as residuary legatees, under the will of Janet Eckford, executed June 14, 1889. She died May 17, 1890, seized of several tracts of land, including that in controversy. She also left a valuable personal estate. It is conceded, and also appears from the will itself, that testatrix intended by it to dispose of all her estate. She did in fact make an effectual disposition of all her estate, including the land in controversy, either to defendant James Eckford for life, etc., or to plaintiffs, unless there was such a lapse or failure at law as to both said parties that it descended to her heirs generally as intestate property. The land in controversy (“S. W. 1-4,77 14 — 98—17) was owned by testatrix when she executed her will as well as when she died. In her will she devised the “S. E. 1-4,-77 14 — 98—17, to James Eckford for life, etc., but it appears she never owned said land. In defendants7 answer and cross bill it is averred that testatrix lived upon the land in controversy, and it was known and commonly designated as her “home farm,77 that she intended and understood that by her will- she was disposing of the real estate of which she might die seized, and none other, by express provisions and accurate description; that by said will she intended to devise the land in controversy to defendant- for life, and he claims the devise as made effectuated that intention; that in the will the land in controversy was erroneously described (as before stated) by mistake or inadvertence of the scrivener who drew the will; that she directed him to so draw her will as to devise the land in controversy to defendant for life, said directions being in parol; that the scrivener intended to properly describe the land in controversy in said will, but, by oversight *61■or mistake, he misdescribed it as being the “S. E. 1-4” instead'of the “S. W. 1-4,” that it was not her intention to devise the land in controversy as a part of the rest, residue, and remainder of her estate. The foregoing allegations were attacked by motion and demurrer, and the question of the admissibility of parol evidence to ascertain and give effect to the testatrix’s intention, as pleaded, raised. “It was further stipulated by the counsel, and consented to by the court, that said motion and demurrer should be taken up and argued concurrently ; that said demurrer shall be treated as supplemental to the motion, and with a view of properly or more fully raising the legal issues that the motion might be insufficient or improper to raise, and with a view to dispose of said cause on its legal merits, and obviate all necessity of parol evidence upon the part of either party.” This was done. The court overruled the motion, also the demurrer; to which rulings plaintiffs excepted, and elected to stand upon both, declining to plead further. Whereupon the court rendered judgment construing the will as prayed by defendants, and dismissed plaintiffs’ action.
II. So much of the will of deceased as is material to an understanding of the questions raised is here set out: “(3) I am the owner of the following described real estate, situated in Mitchell county, Iowa, to wit [describing several tracts of real estate]; also, the southeast quarter (1-4) of section number fourteen (14), township number ninety-eight (98), range number seventeen (17) west of the fifth (5) principal meridian,” which, with certain other real estate, is designated as “p arcel numb er three (3 ) of my estate. ( 4 ) It is my will, and I hereby direct,- that all the property, real and personal and mixed, owned by me at the time of my death, shall be disposed of under the provisions of this will.” “(6) It is my will, and I hereby give-.and devise to my brother James Eckford, the aforesaid parcel number *62three of my estate, to be held and used by him during his own natural life; but the said James Eekford shall not have power to sell or incumber the said premises, and the said parcel number three, upon the death of the said James Eekford, shall be and become the property of my nephew James Thomas Eekford, son of my brother Robert Eekford, for his own use and benefit, forever,’; etc. The last provision of the will is as follows: “It is my will, and I hereby give, devise, and bequeath to my sister-in-law Augusta Eekford, wife of the said John Eekford, and the children of the said Augusta Eekford and John Eekford who may be living at the time of my death, all the rest, residue, and remainder of my estate, real and personal and mixed, not herein-before disposed of; the same to be taken and held by the said Augusta and the said children, share and share alike, each for his or her own use and benefit, forever; each child taking the same share as the said Augusta.”
III. The question presented is whether extrinsic evidence is admissible to show the mistake in description, and to show the testatrix’s intention. While it is the duty of the court, in the construction of a will, to ascertain, if possible, the intention of the testator, yet the intention to be sought after is not that which existed in the mind of the testator, but that which is expressed by the language of the will itself. Whatever differences may exist in the adjudicated cases in the application of this rule, it is believed that the rule itself is unchallenged and inexorable. Bingel v. Volz, 31 N. E. Rep. (Ill. Sup.) 13; 2 Woerner, Adm’n, sections 414, 421; 1 Redf. Wills, 433; 1 Jarm. Wills, 409; 2 Jarm. Wills, 838; Mann v. Mann, 14 Johns. 1; Ryerss v. Wheeler, 22 Wend. 148; Cheyney’s Case, 5 Coke, 68; Vernon’s Case, 4 Coke, 4; Strode v. Lady Falkland, 3 Rep. Ch. 98; Martindale v. Warner, 15 Pa. St. 471; Wig. Wills, Introduction, p. 9; Id. “Construction,” 29, 30; Starkweather v. Society, 72 Ill. 50; *63Decker v. Decker, 12 N. E. Rep. 753, 121 Ill. 341; Schouler, Wills [2 Ed.], sections 421, 466. Or, as the rule is sometimes stated, “the true inquiry is not what the testator meant to express, but what the words used do express.” Burke v. Lee, 76 Va. 386; Couch v. Eastham, 3 S. E. Rep. 26, 29 W. Va. 784; Elliott v. Topp, 63 Miss. 138. So it is said extrinsic “evidence may be admitted in a proper case, when the effect of it is to merely explain or make certain what the testator has written; but such evidence is never admissible to show what the testator intended to write.” Sturgis v. Work, 22 N. E. Rep. 996, 122 Ind. 134. Again, it is said: “The will must speak for itself.” Huston v. Huston, 37 Iowa, 670. “But the question is not hs to what he dictated, but what he signed..” Chambers v. Watson, 60 Iowa, 346, 14 N. W. Rep. 336, in dissenting opinion of Adams, J. It is said in Smith v. Bell, 6 Pet. 74, by Chief Justice Marshall: “The first and great rule in the exposition of wills, to which all rules must bend, is that the intention of the testator expressed in his will shall prevail, providing it be consistent with t'he rules of law. This principle is asserted in the construction of every testamentary disposition. It is emphatically the will of the person who makes it, and is defined to be ‘the legal declaration of a man’s intentions, which he wills to be performed after his death.’ These intentions are to be collected from his words, and ought to be carried into effect, if they be consistent with law.” Kurtz v. Hibner, 55 Ill. 519; Walston v. White, 5 Md. 297; Hawman v. Thomas, 44 Md. 30; Funk v. Davis, 103 Ind. 281, 2 N. E. Rep. 739; McCauley v. Buckner, 8 S. W. Rep. 196, 87 Ky. 191.
IY. It is claimed that this is a case of latent ambiguity, and hence extrinsic evidence is admissible. A latent ambiguity is said to exist 1 ‘when the language employed is clear and intelligible, and suggests but a single meaning, but some extrinsic fact or evidence *64aliunde creates a necessity for interpretation, or a choice among two or more possible meanings.” Black, Law Diet. It-is also said to be “an ambiguity which arises, not upon the words of the will, deed, or other instrument, as looked at in themselves, but upon those words when applied to the object or to the subject which they describe.” Am. and Eng. Encyclopedia of Law, p. 530, and cases cited. Lord Bacon observes that a latent ambiguity “is that which seemeth certain, and without ambiguity, for anything that appeareth on the deed or instrument, but there is some collateral matter out of the deed that breedeth the ambiguity.” Wig. Wills, 197. These ambiguities are of two kinds: “First, where the description of the devisee, or the property devised, is clear upon the face of the will, but it turns out that there is more than one estate or more than one person to which the description applies; and, second, where the devisee or the property devised is imperfectly, or in some respects erroneously, described, so as to leave it doubtful what person or property is meant.” Patch v. White, 117 U. S. 210, 6 Sup. Ct. Rep. 617, 710 (minority opinion). In the same case, in the majority opinion, it is said a latent ambiguity may arise “when the will contains a misdescription of the object or subject, as where there is no such person or thing in existence, or, if in existence, the person is not the one intended, or (the thing does not belong to the 'testator.” The definition last given is much broader, I think, than is warranted by the great weight of authority. It would be a useless task to undertake to reconcile all the cases wherein the question of the admissibility of parol evidence to aid in the construction of wills has arisen.
I am mindful, also, of the well settled rule (additional to the one stated in division 3 of this opinion) that while, in the construction of a will, reference may be made to surrounding circumstances to ascertain the *65devisee or the thing devised, thus placing the court in a-position, so far as is practicable, where it may interpret, the language of the will from the testator’s standpoint, yet surrounding circumstances can not be resorted to to inject into the will an intention not therein expressed. Schouler, Wills [2 Ed.], secs. 465, 467, 568, and cases cited; Bingel v. Volz, 31 N. E. Rep. (Ill. Sup.) 13. The rule of construction laid down by a learned author is: ‘ ‘As the law requires wills of both real and personal estate (with inconsiderable exceptions) to be in writing, it. can not, consistently with this doctrine, permit parol evidence to be adduced to either contradict, add to, or explain the contents of such will; and the principle of this rule evidently demands an inflexible adherence to it, even where the consequence is a partial or total failure of the testator’s intended disposition; for it would have been of little avail to require that a will db origine should be in writing, or to fence a testator around with a guard of attesting witnesses, if, when the written instrument failed to make a full and explicit disclosure of his scheme of disposition, its deficiencies might be supplied, or its inaccuracies corrected, from extrinsic sources.” 1 Jarm. Wills [5.Ed.], 409; Schouler, Wills [2 Ed.], sec. 568; 2 Woerner, Adm’n,. 421.
V. Applying the rules heretofore mentioned, and the rule as to when a latent ambiguity arises, as laid down in the minority opinion of the United States supreme court in Patch v. White, supra, to the facts in this case, we are to determine whether a latent ambiguity exists, and whether extrinsic evidence of the facts pleaded is admissible. The rules seem plain, but their application to a given case is often surrounded with great difficulty. This to some extent, doubtless arises from the fact that courts, in their desire to arrive at the. testator’s intention, and to avoid seeming hardship, sometimes ignore well-settled rales of construction, and *66thus have permitted hard cases to make bad precedents. Again, no two wills are exactly alike, and, as is said by Judge Story, the cases “almost overwhelm us at every step of our progress, and any attempt even to classify them, much less to harmonize them, is full of the most perilous labor.” I am clear that, upon principle and authority, the case presented to us is not one of latent ambiguity. It is not a case where the description is uncertain. It describes land in existence. It is a perfect description; and, if the testatrix had owned the property, there is no question that the devise would have been effectual to pass the title. How, then, does the ambiguity appear? Solely from the fact that testatrix did not own the land attempted to be devised. The contention, then, of appellees, briefly stated, is that in any case where there is a perfect description of land in a will, but it appears that the testator did not own such land, evidence of the testator’s declarations as to what he intended to devise, of his instructions to the party who drew the will, of the scrivener himself, may be received to show what? That there was a mistake or oversight made in the describing of the real estate intended to be devised. If such evidence is competent under such circumstances, it is clear it is just as competent to prove by extrinsic evidence that a mistake was made in the will in naming the object of the testatrix’s bounty; that she intended to devise her property to parties not mentioned in the will; or that one devisee who, by the plain reading of the will, would take certain property, should have it taken from him, because by such evidence it appeared that the testatrix in fact intended it for another. To my mind, such a holding would avoid every safeguard which surrounds the execution and construction of wills, and subject the disposal of the testator’s property to the infirmities which always surround evidence which rests only on the recollection of the individual, as to the facts and *67circumstances which time, age, and the like tend to render uncertain. If the court may strike out of the description of the real estate in this will the “S. E.l-4” and substitute in lieu thereof the “S. W. 1-4,” upon extrinsic evidence, I can see no reason in law why it can not make any other change which such evidence might indicate to be in accord with the testatrix’s intention, either as to the devisee or devise.
It is clear to me that to permit extrinsic evidence to change a complete description of real estate undertaken to be devised in a will in a case like this, and substitute in lieu thereof another description, is not ascertaining the intent of the testatrix from the language of the will itself, as the law requires, but is in fact reforming the'will; to that extent making a new one for her. That can not be done. A reformation of a will can not be thus accomplished under the guise of a construction of it. It is alleged in the cross bill that this erroneous description “appears only by mistake and inadvertence of the person who drew the said will; that, by oversight or inadvertence, the said attorney misdescribed the said tract of land,” etc. In a recent case in Illinois, where the same question was presented and the facts were' practically identical, it was held extrinsic evidence was inadmissible. Bingel v. Volz, 31 N. E. Rep. (Ill. Sup.) 13. In Kurtz v. Hibner, 55 Ill. 514 (and which is cited with approval by this court in Fitzpatrick v. Fitzpatrick, 36 Iowa, 674), there was a devise to Elizabeth Kurtz of an eighty acre tract in section 32. It was proposed to show that the testator, when he died, owned but one eighty acre tract, which was described precisely as was the one devised to said Elizabeth,. except that it was in section 33 instead of 32; that the draughtsman of the will by mistake wrote the erroneous description. The evidence was excluded, and the court said: “The law requires that all wills of land shall be in writing,-and extrinsic evidence is never *68admissible to alter, detract from, or add to the terms of the will. * * * The devise is certain both as to the' object and subject. There are no two objects, no two subjects.” So it is said: “The courts are so strict that they will not permit the terms of a will to be altered, even when the devisor has, by mistake, misdescribed the land in a devise, by substituting that which could be clearly proved to have been intended.” Stark-weather v. Society, 72 Ill. 50. In Miller v. Travers, 8 Bing. 244, where the testator devised all his real estate in the county of Limerick and city of Limerick, and he in fact had no real estate in the county of Limerick, but had in the county of Clare, which was not mentioned in his will, and the devisees offered to show by parol evidence that the lands in Clare were inserted in the devise to them in the first draught of the will, which was sent to a conveyancer, who by mistake erased the words “county of Clare,” and the will was thus executed, it was held that such evidence was inadmissible. In Box v. Barrett, L. R. 3 Eq. 244, it is said: “ Because the testator has made a mistake, you can not afterward remodel the will, and make it that which you supposed he intended, and as he would have drawn it if he had known the incorrectness of his supposition.” In Tucker v. Society, 7 Metc. (Mass.) 188, the testator gave a legacy to said society. It was claimed by the Seaman’s Friend Society. The learned Chief Justice Shaw said it was well proved by circumstances surrounding the execution of the will, as well as by extrinsic evidence, that the testator intended to make the bequest to the Seaman’s Friend Society, but was led to make the mistake by means of erroneous information. It was held to be a case of mistake, not latent ambiguity, and extrinsic evidence was not admitted. In Rapp v. Reehling, 23 N. E. Rep. 68, 122 Ind. 255, it was held that if, by a mistake of the scrivener or translator of the will, it was not written as the testa*69tor intended, and that the property therein bequeathed ■was intended by him to be given to others, that extrinsic evidence to show such fact was inadmissible, as its effect would be “to contradict the will, or to show that the testator’s intention was different from that expressed by the will.” In Ehrman v. Hoskins, 6 South. 776, 67 Miss. 192, it was sought to show by the scrivenerthe intention of the testator to give appellee therein a certain lot, and that the scrivener made a mistake in drawing the will, in properly describing the lot. The court held that the effect of the proposed evidence was “to substitute for the will made by the testator that which the witness endeavored to show he really intended to make, and as to which he failed by mistake of the •drawer of the will.” The court below admitted extrinsic evidence to show what lot the testator really intended to devise, as against land accurately describedin the will. Except as to the initial point, the supreme court said the evidence was inadmissible, even under the majority holding in Patch v. White, supra.
If the testator devised property which he did not own, so that the description is false, not in part only, but in toto, in consequence of a mistake of the testator as to his ownership, the rule allowing a latent ambiguity to be explained by extrinsic evidence does not apply. 2 Woerner, Adm’n, p. 894; Hanner v. Moulton, 23 Fed. Rep. 5. Funk v. Davis, 2 N. E. Rep. 739, 103 Ind. 281, was a proceeding to correct an alleged mistake in the description as to lands devised in a will as “N. W. 1-4 of N. W. 1-4,” etc. It was claimed that the testator did not own the land described in the will, but did in fact own the “N. W. 1-4 of the N. E. 1-4;’’that, by mistake, the testator gave the draughtsman who prepared the will the word “northwest” when he meant “northeast.” Parol evidence was excluded. The •court said: “Evidence is admissible which in its nature and effect merely explained what the testator has writ*70ten, but no evidence can be admissible which in its nature or effect is applicable to the purpose of showing merely what he intended to have written.” In Sturgis v. Work, 22 N. E. Rep. 996, 122 Ind. 134, the testator devised to his daughter land described in his will as the “W. 1-2 of 8. W. 1-4,” etc. He never owned that land, but did own the “W. 1-2 of N. E. 1-4.” etc., of same section, township, and range; and parol evidence to show the mistake was held inadmissible. In Fitzpatrick v. Fitzpatrick, 36 Iowa, 674, it appeared that the testatrix never owned the land describedin the will, but did when she executed the will, and when she died, own another eighty acre tract. It was claimed that a mistake was made in writing the description. That was an action to quiet title, and a reformation of the will was prayed. This court, in an exhaustive opinion, held that extrinsic evidence could not be admitted, and cited many cases in support of its holding. In Sherwood v. Sherwood, 45 Wis. 357, the testator devised lot 10, in block 20, in the city of Oshkosh, which he did not own, instead of lot 9, in block 20, which he did own. This court refused to reform the will. See, also, Bishop v. Morgan, 82 Ill. 351; Heslop v. Gatton, 71 Ill. 528; Bowen v. Allen, 113 Ill. 53; Bradley v. Rees, Id. 327.
VI. “A devise of land, correct in its general description, may be established by the correction upon extrinsic testimony as to what it describes in detail, or vice versa.” Schouler, Wills [2d Ed.], section 574. And, as the same author says, “the principle in this latter class of cases is that where there is in the main a sufficient description in the will to ascertain accurately what is devised or bequeathed, a part which is inaccurate may be stricken out as surplusage, but that nothing substantial-shall be added to the will.” It. is claimed in the case at bar that the erroneous description can be stricken out, and that sufficient will remain *71to lead to the identification of the land in controversy. A few cases will illustrate this rule. Where the testator devised “thirty-six acres, more or less, in lot 37, in the second division in Barnstead, being the same I purchased of John Peavy,” and there was no such lot ■as 37 in such second division, but there was a lot 97 in it, a part of which the testator had purchased of said Peavy, extrinsic evidence was admitted. Here the lot. was identified, outside of the specific description, as the same lot he had purchased of John Peavy. In such a case, the general description clearly leads to the identification of the property. Winkley v. Kaime, 32 N. H. 268. Where testator devised his. lot on Third street (Philadelphia), in the possession of H. R., to his daughter, and it appeared he had no lot on Third ■street, but did have one on Fourth street, which otherwise answered the description, parol evidence was admitted. Allen v. Lyons, 2 Wash. C. C. 475, Fed. Cas. No. 227. A devise of “S. E. and S. W. quarters of .section 4 — 60—38, devisee of S. W. 1-4 to have access to the big spring,” parol evidence was admitted to show that testator never owned any land in section 4 — 60—38, but did own the southeast and the southwest quarters of section 4 — 59—38, the southwest quarter of which was accessible to a big spring. The reference to the big spring was a general identification of the land which, under the rule, permitted parol evidence to specifically locate it. Riggs v. Myers, 20 Mo. 239. In Emmert v. Hays, 89 Ill. 16, the testator in his will described land, some of which he did not own, but followed the specific description with the words, “being what is known as the ‘Hays farm,’ ” and it was held that parol evidence was admissible, and that the specific description might be rejected. In Patch v. White, 117 U. S. 210, 6 Sup. Ct. Rep. 617, 710, where the testator bequeathed lot 6, in square 403, together with improvements thereon *72erected, and appurtenances thereto belonging, and it appeared he owned lot 3, in square 406, the majority of the court (five judges) held the case was one of latent ambiguity, and parol evidence was admissible. They said that the will itself sufficiently identified the subject of the gift, after striking out the false description ; that it was identified by its ownership by having improvements on it, by its being in a square the number of which commenced with 400, and by its being the only lot belonging to the testator that he did not otherwise dispose of.
Every case I have cited, except the last, as well as the great majority of the adjudicated cases, hold that to come within the rule herein laid down, so as to admit of the introduction of extrinsic evidence, there must be a sufficient general description in the will to lead to an identification of the property after the particular description is stricken out. There is, however, a class of cases which hold that such expressions of ownership in a will as, “as to my real estate, I dispose of it as follows: I own,” followed by the words, “I devise the same,” etc., are sufficient as a general description to carry the land in fact owned by the testator, when it is made to appear that he did not own the land specifically described. Pocock v. Reddinger, 108 Ind. 573, 9 N. E. Rep. 473; Cleveland v. Spilman, 25 Ind. 95; Black v. Richards, 95 Ind. 184. Thus the- words “my land” have been held sufficient to carry the land then in fact owned by the testator, and an attempt to specifically describe the land did not render nugatory this general description. Judy v. Gilbert, 77 Ind. 96. So held, also, where the words “my real estate” were used in a will, and the "testator’s real estate was not otherwise devised. In Becker v. Becker, 12 N. E. Rep. 750, 121 Ill. 341, a testatrix devised her shares of the Mechanics’ Bank stock. She had no bank stock, except shares in the City Bank. It was held that *73the word “Mechanics’” should be rejected as inapplicable to any property she owned, and the bequest passed the City Bank stock. Asylum v. Emmons, 3 Bradf. (Sur.) 144. The cases cited above from 9 N. E. Rep. 473, 25 Ind. 95, 95 Ind. 184, 77 Ind. 96, 12 N. E. Rep. 750, and others, that hold such expressions as “my real estate” or “the real estate I own,” and the like, occurring in wills, constitute a good general description of the property, irrespective of an erroneous particular description, seem to me to go outside of and beyond any reasonable application of the settled rules of construction. If one devise land by particular description, and it is erroneous, but he follows it with such an expression as 4‘known as my home farm,” or “now occupied by my son John,” or “having a brick barn thereon,” or the like, it is clear such expressions would constitute a good general description of the property. But the words “my real estate,” or “my land,” or the like, can, it seems to me, in no proper sense be considered as constituting a general description of the land such as is required in such cases.
VII. As I have already stated, it is claimed that the erroneous description in the case at bar occurred by “mistake and inadvertence” of the person who drew the will, or by “oversight and inadvertence” of the. attorney drawing the will. It is the settled rule that the court of equity, even, can not reform a will. It can correct mistakes only when the error appears on the face of the will. So that both mistake and correction can be ascertained and supplied by the context from a plain interpretation' of the instrument as it stands. Schouler, Wills [2 Ed.], sec. 220; Fitzpatrick v. Fitzpatrick, 36 Iowa, 674; Yates v. Cole, 1 Jones, Eq. 110; 2 Pom. Eq. Jur., sec. 871; Bingel v. Volz, 31 N. E. Rep. (Ill. Sup.) 13; Sherwood v. Sherwood, 45 Wis. 357; Chambers v. Watson, 56 Iowa, 676, 10 N. W. Rep. 239. It is a well settled rule that parol evidence can not *74be admitted to supply omissions or defects in a will arising from a mere mistake, oversight, or inadvertence of the testator, or of the draughtsman. Schouler, Wills [2 Ed.], sec. 569; Bingel v. Volz, 31 N. E. Rep. (Ill. Sup.) 13; Kurtz v. Hibner, 55 Ill. 514; Miller v. Travers, 8 Bing. 244; Jackson v. Sill, 11 Johns. 201; Patch v. White, 117 U. S. 210, 6 Sup. Ct. Rep. 617, 710, minority opinion, and cases there cited; Pocock v. Reddinger, 9 N. E. Rep. 473, 108 Ind. 573; Judy v. Gilbert, 77 Ind. 96; Funk v. Davis, 103 Ind. 281, 2 N. E. Rep. 739; Couch v. Eastham, 27 W. Va. 796; 2 Woerner, Adm’n, sec. 421, and cases cited; In re Lyon’s Estate, 70 Iowa, 375, 30 N. W. Rep. 642; Fitzpatrick v. Fitzpatrick, 36 Iowa, 674. It is clear from the authorities, cited that evidence of the facts pleaded in the answer and cross bill was not admissible.
VIII. I will briefly refer to the cases in our own state touching the questions arising in the case at bar. In Hopkins v. Grimes, 14 Iowa, 77, the testator employed the word “homestead” in his will to describe the property he devised to his wife during widowhood, and it. was held that as such word, at the time the will was made, had not a definite and technical meaning, and there was nothing in the will to render certain the description, resort might be had to extrinsic evidence. In Alden v. Johnson, 63 Iowa, 127, 18 N. W. Rep. 696, in speaking of the intention of the testator, the court says: “We can look only to the will itself, guided by the rules of interpretation, in order to determine the intention of the testator, and can not, for that purpose, resort to other sources to discover it.” In Huston v. Huston, 37 Iowa, 670, parol evidence was offered to prove declarations of the testator respecting the bequest-to the plaintiff as being in satisfaction of debts sued on; also, respecting a devise to defendant, and as to testator’s intention that the defendant should have most of his property. The court said they were prop*75erly excluded, — “the will must speak for itself.” In Chambers v. Watson, 60 Iowa, 339, 14 N. W. Rep. 336, the majority of the court held the question involved was not one of ambiguity, but of imperfect description, and that parol evidence was admissible. It will be remembered that the description in the case at bar is perfect. In the case last cited, Adams, J., in a dissenting opinion, held the description absolutely void, and that parol evidence was inadmissible. That was a case where it was necessary to add to the description in order to identify the property claimed to have been devised.
In the ease at bar it is proposed to strike out certain parts of a perfect description, and ingraft therein other words, thereby making the language describe land not mentioned-in the will. Severson v. Severson, 68 Iowa, 656, 27 N. W. Rep. 811, was an action for the admeasurement of dower, and the court reformed an erroneous description. The will showed on its face that the testator undertook to devise to his wife the undivided two thirds of his real estate, and partly by correct description, and partly by misdescription, he gave her the whole of it. He then devised to his daughter, by correct description, an undivided one third of his realty. It was held that the wife took an undivided two thirds, and the daughter an undivided one third, of his realty, and that from the will alone it appeared that the correction in the description should be made. In re Lyon’s Estate, 70 Iowa, 380, 30 N. W. Rep. 642, it was held that parol evidence was not admissible to show that the testator intended to devise different land from that described in the will. The court said: “Now, it seems to us that this, in substance and effect, is a revocation or alteration of the provisions of the will, which are clear, certain, and definite, without a compliance with the statute. * * * The will constitutes written evidence of the testator’s intent, which can only be over*76come by some writing executed as provided by Code, secs. 2329, 2330.” Covert v. Sebern, 73 Iowa, 564, 35 N. W. Rep. 636, was a case where testatrix devised to her “stepson, H. S. Covert,” certain property. She had no stepson whose initials were “H. S.” but did have a stepson named John Harvey Covert, who was-actually called Harvey, and parol evidence was admitted to show that he was the. person intended. It is clear, inasmuch as Covert was designated as “her stepson,” that an error as to his initials would not be fatal. The case comes clearly within the rules heretofore laid down. In Christy v. Badger, 72 Iowa, 582, 34 N. W. Rep. 427, the description in the will read: “A small farm in Wayne county, Iowa, near the Missouri line.” It appeared that the land was not in Wayne county, but in Lucas county. It was held parol evidence was not admissible to identify it. The court says: “If, after the false description is discarded, there remains in the devise language sufficient to direct to the identification of the subject with sufficient certainty, an estate will pass thereby; but when the false language is eliminated, and nothing remains directing inquiry which may result in discovering the true subject of the devise, it is void. Thus, if the description by metes and bounds or by congressional subdivisions-is false, but it is aided by language declaring the ownership or possession of the land by the devisor or others, or the like, it is good.” The opinion was undoubtedly right, but I do not feel bound by all that is said in the concluding clause of the quotation, if it is to be construed as holding that a mere statement in the will that “I own” certain land, and like expressions, constitute a sufficient general description to effectuate a devise of land otherwise misdescribed. In Fitzpatrick v. Fitzpatrick, 36 Iowa, 674, it appeared that testatrix never owned the land described in the will thus: “That tract of land situated in Table Mound *77township, Dubuque county, Iowa, described as follows: The west half of the northeast quarter of section 23, in Table Mound township, Dubuque county, Iowa.” But she did own the east half of the southwest quarter of section 23, 88 — 2. It was alleged in the petition that the scrivener made a mistake in writing the description. In the course of an elaborate opinion, the court says: “In all the cases coming within the scope of our investigation of this question, where extrinsic evidence had been admitted to remove a latent ambiguity, the language of the will, after rejecting the false description, has been sufficient to show what property or what person was intended by the testator. * * * But we have seen no case where other words than the words of the devise have been allowed to be imported into the will in order to describe a devise, or to identify property to which the words of description in the will did not apply; * * * and when, by rejecting the false description, the re-, maining words do not describe the property or person to any extent, parol proof to show the testator’s intention is inadmissible.” Again, it is said: “The description of the will is complete and perfect, describing: a tract of land which answers perfectly the description. We are asked to disregard this description, and hear parol- evidence to show that the testator intended to describe another and different tract of land. Not that she intended another piece of land by the use of the words of the devise, but that she intended to use different words of description; and that, through the mistake of the person writing the will, the words intended were not inserted, but other and different language was: used. If the false description be rejected, there are then no words left in the will to describe the premises claimed under it.” It was held that parol evidence was inadmissible. In Kiene v. Gmehle, 85 Iowa, 312, 52 N. W. Rep. 232, it is said: “The intention of the *78testator, as shown by the will, construed according to established rules, must control.” I hold that extrinsic evidence was not admissible; that there are no words of general description of the property in controversy in the will whereby the erroneous part of the description could be eliminated, and leave sufficient to admit of the introduction of extrinsic evidence to apply the language of the will to the thing devised. It is true that there are able opinions holding that parol evidence is admissible in such cases, but, after a careful examination of the authorities cited by counsel in their able arguments, as well as many others, I am forced to the conclusion that to open the door in a case like this to the admission of extrinsic evidence would, to a great extent, do away with the protection afforded by the wise provisions of the law touching the execution of • wills. •
IX. The real estate in controversy, not being otherwise disposed of by the will, passes to plaintiffs under the residuary clause heretofore set out. Schouler, Wills, sec. 521; Floyd v. Carow, 88 N. Y. 560; Riker v. Cornwell, 113 N. Y. 115, 20 N. E. Rep. 604. For the reasons given, the case should be reversed.