Joseph Dominici died testate. Upon distribution question arose as to the interpretation of the provisions of his will and of a codicil thereto, and from the decree ■as ordered Marie Kohler appeals. The only matter here in ■controversy is a dispute over the conflicting claims of Marie Kohler and of Christiane Arndt.

The conflict arises under the following state of facts: By his will Dominici provided as follows: “Lastly, I give, bequeath and devise all the rest, residue and remainder of my property, real and personal, of whatsoever kind or character, ■or "wheresoever situated, share and share alike, unto my sister, Louise Jahnke, and unto my nephew, Heinrich Schluter, and Lis sister, my niece, all residing in Luchow, Hanover, Germany.” A codicil, made about three months thereafter, is as follows: “Whereas, it has come to my knowledge that my ■sister, Louise Jahnke, one of the residuary legatees therein námed, is dead. I hereby bequeath, give and devise the share ■of my said estate in said will given, bequeathed and devised to my said sister to the other two residuary legatees therein named, Heinrich Schluter, and to his sister, my niece, whose name is Marie Kohler, and whose residence is Salzwedel, Alt-mark, Germany, share and share alike.”

No doubt, ambiguity or uncertainty arises upon the face of these provisions. No conflict between them is disclosed, for, while the residence of the testator’s niece, Schluter’s sister, is given in the will as being Luchow, and in the codicil as Salzwedel, so far as appears from the face of the instrument, the facts might have been that her earlier residence was in Luchow, and her residence at the time of the codicil in Salzwedel. Nor does the name of Marie Kohler, described as the sister of Schluter and the niece of the testator, complicate matters upon the face of the writing, since Marie Kohler might have been the name of a married sister of Schluter. But when it was attempted to make application of the residuary devises and bequests to the persons, designated, extrinsic evidence at once disclosed a difficulty. Heinrich Schluter did have a sister, and one sister only, who, at the time of the-making of the will and at the time of the making of the codicil was a resident of Luchow, Hanover. She never resided in Salzwedel. She was unmarried, and her name was Christiane Schluter. Upon the other hand, the deceased did leave a niece, the- married daughter of another sister, whose name is Marie Kohler, and whose residence is Salzwedel, Altmark,. Germany. There is thus disclosed a typical ease of a latent ambiguity. It is not a patent ambiguity, contemplated by section 1318 of the Civil Code, but a latent ambiguity, to-which section 1340 of that code is directed. That section provides as follows: “When applying a will, it is found that there is an imperfect description, or that no person or property exactly answers the description, mistakes and omissions must be corrected, if the error appears from the context of the will or from extrinsic evidence; but evidence of the declarations of the testator as to his intentions cannot, be received.”

In its effort, therefore, to determine the true intent of the testator, upon the disclosure of this latent mistake in description, the trial court took extrinsic evidence, as it was proper to do. It concluded that the testator intended to name three residuary legatees, namely Heinrich Schluter, his sister Christiane Schluter, and the testator’s other niece, Marie Kohler, and so decreed distribution of the residuary third of the estate-in equal portions to these three. But this does manifest violence to the expressed intention of Dominici. In terms the-codicil declares that the residuary one third originally left to Louise Jahnke is devised “to the other two residuary legatees”' in the will named. There can be but two persons whom the- testator had in mind, and, as Heinrich Schluter is indisputably one of these, it is to be determined whom he intended as the other. For even in cases where extrinsic evidence is admitted to explain latent ambiguities, and to perfect imperfect descriptions of beneficiaries, or the subject-matter of devises or bequests, no evidence is admissible to change or vary the testator’s expressed intent. This must always be deduced from the will itself, assisted by such extrinsic evidence. The proofs afforded by such evidence are to be employed merely as an aid to the court in determining what in fact was the expressed intent of the testator, and if that intent may not be so found, the devise or bequest must lapse, since it is not within the power of a court to make a new will. (Estate of Young, 123 Cal. 337 , [ 55 Pac. 1011 ].)

The section of the code above quoted forbids in the giving of such evidence the reception of mere declarations of the testator as to his intentions. This is a limitation upon the general rule which, as stated in American and English Encyclopedia of Law, vol. 2, p. 298, is that in case of a latent ambiguity, where there are two persons or things equally answering the description, such ambiguity may be removed “by any evidence either of circumstances or declarations of the testator. ’ ’ The text of this citation will be found abundantly supported by authority, and our code provision limiting the scope of the evidence which may be received appears to be at variance to the general rule. It will not be extended, therefore, beyond its- actual language, and will be held to apply to the mere incidental fugitive utterances or declarations of intent, as distinguished from specific instructions as to testamentary disposition which it may be proved were given. There is, of course, a broad difference between testimony of the casual remark of a man as to his intention to leave a portion of his estate to this person or the other, and the positive instructions which he has given to his attorney in the very performance of the testamentary act, and in any proper case such instructions may be received without doing violence to section 1340 of the Civil Code. Thus, in Vernor v. Henry, 3 Watts, 585 , an ambiguity arose from a discrepancy between the name and the description of a devisee, as here, and it was held that explanatory declarations made by the testator at the time of the execution of his will, indicative of his design to give his property in a particular way to a particular person, were admissible in evidence to explain the ambiguity. (See, also, In re Gregory Settlement and Will, 34 Beav. 600.) It should, of course, be unnecessary to add, that where a will without latent ambiguity expresses a clear intent, such evidence would never be permitted in an# effort to show that the intent expressed was a mistaken one. It is only where the description of the person or property by extrinsic evidence is shown to be doubtful or imperfect, that such evidence may be employed in the effort of the court to arrive at and declare the meaning and intent which the testator, by the terms of his will, attempted, however irregularly, to declare. This under the familiar rule of evidence that a doubt arising from extrinsic evidence may be removed by extrinsic evidence.

In this case, with other testimony, that of the attorney who drew the will was taken upon the matter. By respondent it is objected, though the objection is not pressed with any earnestness, that it was error to receive the testimony of the attorney because it was the disclosure of privileged communications, forbidden by subdivision 2 of section 1881 of the Code of Civil Procedure. Doubtless the argument is not strongly urged, because it appears from the record that though respondent originally objected to the taking of the attorney’s testimony, he withdrew his objection and asked the court to admit the testimony, subject to his subsequent motion to strike out. The testimony was so admitted, and respondent did not move to strike it out. Hence the evidence appears in the ease, not only without objection upon her part, but absolutely by her invitation. In the second place, the reason for the rule prescribed in section 1881 ceases in a case such as this, and with the reason the rule itself ceases. Thus, in Doherty v. O’Callaghan, 157 Mass. 90 , [ 34 Am. St. Rep. 258 , 31 N. E. 727 ], it is said: “Undoubtedly, while the testator lives the attorney drawing his will would not be allowed without consent of the testator to testify to communications made to him concerning it, or to the contents of the will itself; but after his .death, and when the will is presented for probate, we see no reason why, as a matter of public policy, the attorney should not be allowed to testify as to directions given to him by the testator, so that it may appear whether the instrument presented for probate is or is not the will of the alleged testator.” (See, also, In re Wax, 106 Cal. 348 , [ 39 Pac. 624 ] ; Estate of Nelson, 132 Cal. 187 , [ 64 Pac. 294 ].) As said in Weeks on Attorneys at Law (sec. 165), “The privilege does not apply when the question arises as to the intention of a deceased person in respect to the disposition of his estate. . . . In the case of testamentary dispositions, the very foundation of the rule appears to be wanting.”

Coming now to consider the„evidence in the case, it appears that Louise Jahnke was a sister of the deceased, and that Heinrich Schluter and his sister, Christiane Schluter, were children of another sister, and that Marie Kohler was a married daughter of a third sister. It further appears that the deceased had been on terms of intimacy with the Schluter family, corresponding with them and knowing them well; that he knew the name of Christiane Schluter as well as that of her brother Heinrich, and it is in evidence that he wrote to Christiane Schluter shortly before the making of the will. The testimony shows that he was in constant and regular communication with the members of the Schluter family, and that he was kept fully informed of their family affairs. The Schluter family resided at the birthplace of the deceased, in the house he had helped to build before he left Germany. It is in evidence, moreover, that he frequently talked about the Schluter family, calling the individual members of it by name, frequently speaking by name of his niece, Christiane Schluter. As respondent expresses it, “the evidence shows that the Schluters were in his mind all the time.” There was an absence of intimacy between him and Marie Kohler and her parents, and there had been little communication between the deceased and the family to which Marie Kohler belongs in the last twenty years of his life. So much for the proofs of relationship and the attendant circumstances. Mr. Dennett, who drew the will and the codicil thereto, testified that at the time he drew the will, Louise Jahnke and Heinrich Schluter, the other two residuary legatees, were named; .the name of the third residuary legatee was omitted because deceased could not remember her name; that it was understood between the attorney and Dominici at the time the will was drawn that at some future time Dominici would bring his attorney the name of the person to whom he desired the remaining one third of his property to go, and that when Dominici obtained the name, a codicil should be written, or such change made in the will as would give the name of such third residuary legatee; that in accordance with this understanding, Dominici subsequently came to Dennett’s office, and the codicil was drawn; that the name and description of Marie Kohler, Salzwedel, Altmark, Germany, was furnished by Dominici to the attorney; that Dominici explained that he had received a letter from Germany giving the required name and residence. Mr. Dennett’s recollection also was that Dominici wrote the name and address of Marie Kohler first in German and then in English; that the name ‘ ‘ Marie Kohler, residing at Salzwedel, Altmark, Germany,” was undoubtedly written in the codicil in accordance with Dominici’s instructions; that Dominici never at any time during the execution of the will or of the codicil, mentioned the name of Christiane Schluter. (It may be noted in this connection that the name “Christiane Schluter” nowhere appears in the will.) In the discussion of the family relationships the attorney obtained the idea that the niece to whom Dominici desired to leave one third of his property was Heinrich Schluter’s sister, and he supposed that Marie Kohler was in fact Heinrich Schluter’s sister.

The intimacy between the deceased and the Sehluters, and the fact that the Sehluters were always in the deceased’s mind is, as a weapon in the hands of respondent, a two-edged sword. If the deceased really intended Christiane Schluter because of the family intimacy, there is no reason from the evidence why he should not have named her, and every reason why he should. Upon the other hand, if he really intended Marie Kohler, by reason of the estrangement between the families and the infrequent communications between them, it is reasonable to suppose that he might not recall the name and residence of the niece whom he actually designed and intended should share his bounty. There does not appear to have been any hostility between the deceased and Marie Kohler, and it is at least as natural that, in the prospect of death, he should have desired each of his sister’s families to have shared in his estate, as to have supposed that he designed, for no real reason, to exclude one of them. If Marie Kohler be understood as a residuary devisee and legatee, then by his will he would have made provision that the residuary portion of Ms estate go in equal shares to the representatives of the families of each of his three sisters. Of even more significance is the fact that he brings to his lawyer and formally gives to him in writing the omitted name and address. If we can _ conceive that by some lapse of memory, he had at the time of the making of the will forgotten the name of Christiane Schluter, while remembering the name of Heinrich Schluter, her brother, it is not to be supposed that he would have come back, three months after, with the accurate name and address of one niece, under the mistaken supposition that he was really describing Christiane Schluter, who did live in Luchow, and who had never resided in Salzwedel, where Marie Kohler in fact did reside. From what has been said, it is thus apparent that in the codicil either the words “his sister” is an error in description and should be rejected, or the phrase “whose name is Marie Kohler and whose residence is Salzwedel, Alt-mark, Germany,” is a misdescription and should be rejected. The circumstances and the evidence as heretofore set forth leave little room for doubt. The consequence attaching to the formal presentation of the name and address of Marie Kohler must be held to be the controlling and dominating circumstance, and, so far as the codicil is concerned, the words “his sister” and “above named” must be rejected as erroneous.

This conclusion, however, affects and can affect the codicil alone. The residuary clause in the body of the will is free from all doubt and uncertainty, latent or patent, and extrinsic evidence, therefore, cannot be admitted to change its terms. However strong the presumption may be that Marie KoMer was also intended in the body of the will, the presumption is one which may not be indulged and in support of which no evidence would be admissible, since the language of the will itself is free from doubt and from the need of construction. It leaves clearly and distinctly, and with no possibility for the play of construction, one residuary third of his estate to the sister of Heinrich Schluter, niece of the deceased, who resides in Luchow, Hanover, Germany. Christiane Schluter, now Christiane Arndt, completely and perfectly fills this description, and is entitled to take, and tMs regardless of any disturbing effect which it may be thought that the codicil has upon it. It is a rule of construction that dispositions made by a will are not to be disturbed by a codicil further than is absolutely-necessary in order to give that codicil effect, and a clear disposition made by a will is not revoked by a doubtful expression or inconsistent disposition in the codicil. (In re Ladd, 94 Cal. 674 , [ 30 Pac. 99 ] ; Kane v. Astor’s Executors, 5 Sandf. 533 .) Moreover, it is to be noted that the codicil itself does not attempt to overthrow the disposition made by the will, excepting as to the one third, of the residuum of the estate by the will devised and bequeathed to Louise Jahnke. It is upon this one third alone that the codicil operates, and by its terms one half of that one third vests .indisputably in Heinrich Sehluter, and the other moiety of the one third vests in Marie Kohler. The residuary disposition then should be, to Heinrich Sehluter one third plus one half of one third, or three sixths; to Christiane Arndt one third, or two sixths; and to Marie Kohler one half of one third, or one sixth. '

It is ordered that the decree appealed from be reversed," and distribution ordered in accordance herewith.

McFarland, J., Shaw, J., Sloss, J., Angellotti, J., Lorigan, J., and Beatty, C. J., concurred.