Opinion by
Mr. Chief Justice Mitchell,The statute requires that a will shall be in writing, and signed by the testator “at the end thereof.” The end meant by this provision is the logical end of the language used, which shows that the testamentary purpose has been fully expressed. The position of the signature with regard to the bottom or end of the page is only evidence on the question whether the testator has completed the expression of his intention. Prima facie that is the natural place for the signature to be placed to show the full expression of the testator’s wishes and therefore is presumptively the right place for it, but it is only evidence and must give way to evidence of a different intent.
In Hays v. Harden, 6 Pa. 409, Chief Justice Gibson says: “Signing at the end of a will was required by the statute to prevent the evasion of its provisions that followed the English statute of frauds, which the judges held to be satisfied wherever the testator’s name, in his own handwriting, was found in the introductory or any other part of the instrument.” In Heise v. Heise, 31 Pa. 246, Strong, J., says: “Nor should we lose sight of the mischiefs which existed at the time when it (the statute) was enacted; mischiefs which it was designed to remedy. Among these, none was more serious than the facility with which unfinished papers, mere inchoate expressions of intention, were admitted to probate as valid wills of decedents. Letters, memoranda, mere notes unsigned, which were entirely consistent with a half formed purpose, and which may have been thrown aside, and never intended to be operative, were rescued from their abandonment, proven *192'as wills, and allowed to prevail as dispositions of property which there was much reason to believe the decedent never intended. It was to remedy this mischief that the act of 1833 provided, that every will should be signed at the “end thereof,” that thus, by his signature in that place, the testator should show that his testamentary purpose was consummated, and that the instrument was complete.” And in Knox’s Estate, 131 Pa. 220, it was said: “The purposes of the act of 1833 were accuracy in the transmission of the testator’s wishes, the authentication of the instrument transmitting them, the identification of the testator, and certainty as to his completed testamentary purpose. The first was attained by requiring writing instead of mere memory of witnesses, the second and third by the signature of testator, and the last by placing the signature at the end of the instrument. The first two requirements were derived from the English statute; the third was new (since followed by the act of 1 Vict., c. 26), and was the result of experience of the dangers of having mere memoranda or incomplete directions taken for the expression of final intention: Baker’s App., 107 Pa. 381; Vernon v. Kirk, 30 Pa. 218.”
In the present case the connected sense of the text is entirely clear, though it does not follow the usual order of arrangement. But it does not deviate from it more than many letters written in the style of the present day where the writing jumps from the first to the third page and then back to the second. The "full substance of the testatrix’s intent and its expression are there, and the signature is at what she intended and regarded as the end of her will. Where that is manifest the continuity of sense and not the mere position on the page must determine the statutory “end thereof” as the place for the signature. The unusual, and as it might be called irregular arrangement of the several parts of the will is not so great as in Wikoff’s Appeal, 15 Pa. 281. There is' no Pennsylvania case which conflicts with this view, though care must be taken to distinguish cases like Hays v. Harden, 6 Pa. 409, where an addition after the signature was held to be testamentary and therefore to invalidate the will; Wikoff’s *193Appeal, 15 Pa. 281, where the converse was held, that a mere memorandum not testamentary in character did not have that effect; Heise v. Heise, 31 Pa. 246, where a testamentary clause after the signature but not signed, and shown not to have been on the will when the latter was executed, was held not to affect its validity. This was, however, a fully executed will and plainly within the protection of the clause of the statute that no will shall be repealed otherwise than by some other will or codicil duly executed and proved, or by burning, canceling, obliterating or destroying the same, by the testator or by someone in his presence and by his express directions.
We are not unmindful of possible danger in allowing marginal writing to be counted as part of the text of the will. Certainly parol testimony to that effect where there is no reference in the will, to identify and incorporate it, must be received' with caution. But to some extent parol testimony must always be admissible, as to prove signatures, show identity, etc. The exact point at which it must stop cannot be laid down in any hard and fast terms, but must depend on the necessity of the case. It will be observed that in Wikoff’s Appeal, 15 Pa. 281, and Heise v. Heise, 31 Pa. 246, the testimony was admitted apparently without objection. In the present case the question is purely theoretical as the good faith of the entire transaction is admitted.
Decree affirmed.