This is an appeal from an order admitting to probate as the last will and testament of said deceased three documents written in his handwriting. The three documents in question follow:

It is contended by appellants that the first document is not a will because it is not signed within the meaning of the code requirements of a holographic will. It is not essential that such a will be signed at the end thereof. “The true rule, as we conceive it to be, is that, wherever placed, the fact that it was intended as an executing signature must satisfactorily appear on the face of the document itself. If it is at the end of the document, the universal custom of mankind forces the conclusion that it was appended as an execution, if nothing to the contrary appears. If placed elsewhere, it is for the court to say, from ‘ an inspection of the whole document, its language as well as its form, and the relative position of its parts, whether or not there is a positive and satisfactory inference from the document itself that the signature was so placed with the intent that it should there serve as a token of execution.” {Estate of Manchester, 174 Cal. 417, 421 [Ann. Cas. 1918B, 227, L. R. A. 1917D, 629, 163 Pac. 358, 360 ].) This rule has in substance been many times repeated by the supreme court.

We are of the opinion that the signature in the document of July 14, 1926, meets these requirements. Appellants claim that in no California decision has a signature in the exordium clause been held sufficient and that in many cases it has been held insufficient, such as Estate of Manchester, supra, Estate of Hurley, 178 Cal. 713 [ 174 Pac. 669 ], Estate of Bernard, 197 Cal. 36 [ 239 Pac. 404 ], and Estate of Devlin, 198 Cal. 721 [ 247 Pac. 577 ]. In each of these cases, however, there was also evidence on the face of the will of incompleteness indicating that the signature found at the beginning of the will could not have been intended as the token of execution and could not be regarded as proof that the execution of the will was a complete and concluded act. On the other hand there are several authorities which uphold a signature at the top or beginning of a will. (Estate of McMahon, 174 Cal. 425 [L. R. A. 1917D, 778, 163 Pac. 669 ]; Estate of Streeton, 183 Cal. 284 [191 Pac. 16]; Estate of Morgan, 200 Cal. 400 [ 253 Pac. 702 ]; Estate of England, 85 Cal. App. 486 [ 259 Pac. 956 ].)

In the last two cited cases the signature appeared at the very top of the document in a phrase independent of the will proper as though it were a label or description of the will to follow, thus—Last will or last will and testament of -. We are unable to distinguish the ease at bar from these two authorities. The mind which writes at the top of the will “Last will of -” is the same mind which writes in the first paragraph “I-do hereby publish and declare this my last will” and the two expressions convey substantially the same idea. Whether the signature appearing within either one of these expressions amounts to a signing of the will must be determined from the entire will. The Sullivan will of. July 14, 1926, ends- with a period and appears to be complete. After listing the legacies, appointing an executrix and inserting a residuary clause, the testator’s mind returned to the bequest to his stepson found in about the middle of the will and added what appears to be a finishing touch by giving directions as to.how that particular bequest should be treated. Completeness or lack of completeness appearing on the face of the purported will has been an important consideration entering into the decision of all the cases referred to herein. We think the Sullivan will of July 14, 192'6, shows by itself that the testator adopted his signature in the first clause of the document as the token of execution and that it is therefore signed within the meaning of the code requirements.

The conclusion upon the point is arrived at, as of course it must be, without regard to what the testator afterward did with the will and without regard to what he afterward thought about it or wrote about it. However, the codicil of October 13, 1926, is admittedly validly executed according to law. Its terms are such as to incorporate into and as a part of itself the document of July 14, 1926. In Estate of Plumel, 151 Cal. 77 [ 121 Am. St. Rep. 100 , 90 Pac. 192 ], a codicil to an invalidly executed will was held to make the defective will a part of the testator’s last will and testament. See, also, 26 Cal, Jur. 849, 850, 851, and cases cited. Appellants do not contend that a preceding validly executed will is essential for the rule of incorporation by reference to apply, but contend that the requirements of that rule are not met in the instant case. They quote these requirements from Estate of Young, 123 Cal. 337 [ 55 Pac. 1011 ], to the effect that the description (of the incorporated document) in the will itself must be so clear, explicit and unambiguous as to leave its identity free from doubt. The description of the July document in the October document is claimed to be fatally defective because it refers to the July document as a codicil will. It is said that the document is not a will because not signed, that it is obviously not a codicil, that there is no such thing as a codicil will and the July document itself is not labeled as a codicil will. For these reasons it is said that the precise document referred to is not proven to be in existence, if in existence it is not the July document produced and probated, and that the description is too uncertain and vague to be capable of being applied to any instrument in particular.

Absolute certainty in any description is not possible. The description 'here is not perfect because it refers to the July document .both as a “codicil will” and as “my last will and testament.” The date, however, is exactly correct and the description of the portion of its contents referred to is also accurate. An inspection of the two documents without recourse to other facts in evidence can leave no reasonable doubt that the testator referred to the July document probated. The fact that an incorporated document is upon the same sheet of paper or attached to the same sheet of paper containing a valid codicil has been held to constitute a sufficiently certain description for incorpo ration by reference without other description. (Estate of Sherrett, 67 Cal. 585 [ 8 Pac. 181 ]; Estate of Plumel, supra; Estate of Henderson, 196 Cal. 623 [ 238 Pac. 938 ].) These authorities will suffice to indicate that absolute certainty is not required of a description of an incorporated document. The fact that two documents are" each on different sides of a single sheet of paper can be nothing more than a strong inference. It does not amount to a complete description in words in the incorporating document especially where the latter does not refer in words to the incorporated document as being on the obverse side of the paper.

The third document is the envelope in which the first two documents were found sealed after the death of the testator and the writing thereon. In determining the sufficiency of the description in the second document to incorporate the first by reference, logic directs that the sealing of the two documents in one envelope is just as eloquent as the attaching together of the two documents by other mechanical means. This disregards the inscriptions upon the envelope.

We do not agree, however, with respondent’s contention that the writing on this envelope constitutes by itself a will incorporating by reference the two documents sealed within the envelope. If it should be conceded that the date of October 13, 1926, where it secondly appears is the date of the inscriptions and not merely a statement of the date upon which the testator opened and resealed the envelope, still the inscriptions do not appear to have been written animo testandi. This document seems only to be a simple description of the contents of the envelope and a statement that it had been opened and resealed by the testator. For the same reason it should not be regarded as a part of the will within the envelope. (Estate of Manchester, supra.) In view of our conclusion upon the self-sufficiency of the two documents inclosed we need not determine the precise value of indorsements upon the envelope. This document does not add any substance to the terms of the will and codicil and its admission to probate together with the first and second document is at the most harmless.

The judgment appealed from is, therefore, affirmed.

Sturtevant, J., and Nourse, J., concurred.

A petition for a rehearing of this cause was denied by the district court of appeal on December 5, 1928, and a petition by appellants to have the cause heard in the supreme court, after judgment in the district court of appeal, was denied by the supreme court on January 3, 1929.

All the Justices present concurred.