Fowler, J.

An instrument purporting on its face to be the last will of Alex J. Johnston was denied probate. The only points involved are, (1) whether the purported testator had signed the instrument at the time the witnesses thereto signed it, and if so, (2) whether it was duly attested.

(1) It has been declared by this court that “in the absence of clear proof that the witnesses to a will signed it before the testator did, it will be presumed that the testator signed it first.” Flood v. Kerwin, 113 Wis. 673, 679, 89 N. W. 845, citing Allen v. Griffin, 69 Wis. 529, 35 N. W. 21, and Will of O'Hagan, 73 Wis. 78, 40 N. W. 649. It is suggested that these cases are not in point because the wills involved therein had attestation clauses while the instant will has none. But in Will of Griffith, 165 Wis. 601, 606, 163 N. W. 138, there was no attestation clause, and the will was signed by mark. It is said in the opinion in that case:

“The appearance, on the face, of regularity should be regarded as involving a presumption of due execution so strong as to prevail, in the absence of clear and satisfactory evidence to the contrary.”

Moreover, the instant will is in the handwriting of the testator, and that would seem to be of equal force with an attestation clause in support of the presumption of due execution. The instant will “on its face” has every appearance of regularity. Another point of difference between the instant case and those above referred to is that what is said in those cases is said in reference to signing by'the testator at the 'same time 'the witnesses signed, so that the signing by the testator and the witnesses constituted a continuous transaction. But it is held in the Griffith Case, supra, that where *144the witnesses and the testator all sign as part of a single continuous transaction, it is immaterial whether the testator signs before or after the- witnesses. The statements in the cases first cited are therefore pointless, except they apply to cases where the testator and witnesses signed at different times. The presumption stated is not limited to such cases.

“It is well settled that an instrument in writing, signed by the testator, and subscribed in his presence and at his request, which may be implied from circumstances, by two competent witnesses, is prima facie, and so far as formality of execution goes, a valid will.” Skinner v. American Bible Society, 92 Wis. 209, 213, 65 N. W. 1037.

“Such attestation [of a will by two witnesses] is of itself not only prima facie evidence that the instrument was properly executed, but it raises a strong presumption that it was so executed. Had the witnesses deceased before the probate of the instrument, mere proof that the attesting signatures were their handwriting would have established the will. And the rule would have been the same although the signa-. tures of the witnesses were not preceded by any attesting clause or certificate. To defeat probate, the strong presumption of regularity thus appearing- upon the face of the instrument must be overcome by proof. . . . Hence, in the present case, if the fact that the witnesses subscribed the instrument before the testator defeats the probate thereof as the will of the testator, the fact should not be.found, against the presumption of regularity, without very clear and convincing proof.” Will of Lewis, 51 Wis. 101, 113, 7 N. W. 829.

In the instant case there is no proof tending to show that the testator signed after the subscribing witnesses. Had the witnesses been dead when proof of the will was made, the will would have been admitted to probate on proof of the signatures of the testator and the witnesses. Proof of these signatures is absolute and unquestioned. Such proof, and the appearance of regularity, raised a presumption, and in the absence of evidence to the contrary, required a finding that the testator had signed the will when he presented it to *145the witnesses for attestation. Moreover, the statement of facts preceding the opinion contains much besides proof of signatures and appearance of regularity to support such finding.

(2) It is not necessary tO' validate a will that the testator sign it in the presence of the witnesses. Where not so signed, it is sufficient if he has signed it before it is attested by the witnesses. Skinner v. American Bible Society, supra; Flood v. Kerwin, supra; Estate of Lagershausen, 224 Wis. 479, 483, 272 N. W. 469.

In the instant case neither of the attesting witnesses saw the testator’s signature on the will. The adjudicated cases are not in accord as to whether a will is properly attested when none of the witnesses either saw the testator sign or saw his signature. An extensive note in 38 L. R. A. (N. S.) 161, cites and states the cases bearing upon the point. In Dougherty v. Crandall, 168 Mich. 281, 134 N. W. 24, in connection with which the above note is inserted, a holograph will was held properly attested under the circumstances above stated. In Nunn v. Ehlert, 218 Mass. 471, 106 N. E. 163, 167, under the same circumstances, probate was denied. The statutory provisions in both Massachusetts and Michigan are the same as our statute (sec. 238.06) in requiring that wills must be in writing, signed by the testator and subscribed and attested by the witnesses in the presence of the testator. There were attestation clauses in the wills involved in both the Dougherty and Nunn Cases. The holding of the Nunn Case seems out of line with that in Ela v. Edwards, 16 Gray (Mass.), 91. The rule of the Ela Case is said in the Nunn Case opinion to apply only where the witnesses are dead or for some other reason cannot testify. None of the witnesses was dead in the Ela Case. Three witnesses were required under the Massachusetts statute. Two of them testified, and both testified that they did not see the testator’s signature, and that he did not sign in their pres*146ence. The will was held properly attested and admitted for probate. The third necessary witness to the will was out of the state when proof of the will was made. The real basis of the holding in the Nunn Case seems to be that the word “attestation” in the statute implies that the witness must attest, that is, must testify that the signature is the signature of the testator, and that he cannot do this unless he sees the signature or sees the testator sign. But the language of the statute is .that the will must be “subscribed and attested to by the witnesses.” It is the will, not the signature of the testator, that must be “attested.” The case holds that—

“. . . When the [testator] hides from the subscribing witnesses the signature which is upon the instrument previously signed by him and goes no further than to ask the subscribing witnesses to sign the paper placed before them, even if that request be accompanied by a statement that the paper is his will, there is no acknowledgment by the [testator] of his signature, and so no valid attestation of his signature by the subscribing witnesses. All that is acknowledged by the [testator] 'in that case is that the paper is his will. In such a case there is no acknowledgment by the [testator] that the signature on the paper (if there be a signature upon it) is his signature.”

This seems false reasoning. There is no will if there is no signature of the testator. Therefore, if the testator acknowledges “the paper is his will,” he acknowledges that his signature has been thereto attached. If there in fact be a signature, the attestation is true. If in fact there be no signature, the attestation is false. But the fact that the signature was attached is presumed in the instant case, for the reasons above stated. There was therefore a proper attestation of the will. This is the idea of attestation taken in the leading case of White v. Trustees of British Museum, 6 Bing. 310, cited in Flood v. Kerwin, supra, as authority for the rule above stated that “it is not essential that the testator sign in the presence of the witnesses, if he has in fact signed *147prior to the time they do.” In that case the testator did not sign in the presence of the witnesses, nor did two of the necessary three witnesses see his signature thereon. The opinion states, page 318, that—

“. . . Any acknowledgment before the witnesses that it [the signature to the will] is his signature, or any declaration before them that it [the will] is his will, is equivalent to an actual signature in their presence, and makes the attestation and subscription of the witnesses, complete.”

On page 320 it is stated that the court found that—

“. . . The testator knew this instrument to be his will; that he produced it to the three persons and asked them to sign the same; that he intended them to sign it as witnesses ; that they subscribed their names in his presence and returned the same identical instrument to him.”

Later on the page last cited it is stated:

“We think . . . [the facts next above stated] place the testator and the witnesses in the same situation as they stood where such oral acknowledgment of signature has been made, and we do therefore, upon the principle of those decisions, hold the execution of the will in question to be good within the statute.”

One of the three witnesses in the White Case saw the signature to the will; but this does not render the case inapplicable. For the will had to be attested by all three witnesses, and if it was not properly attested by two of them, it was just as invalid as if none of them had properly attested it. The fact that one of the witnesses saw the signature, while the others did not, only proves the fact here found pursuant to presumption that the will was signed before it was witnessed.

By the Court. — The judgment of the county court is reversed, and the record is remanded with instructions to enter judgment admitting the will to probate.