¶1Opinion by
¶2We are of opinion that our decisions in the two cases of Derr v. Greenawalt, 76 Pa. 239, and Charles v. Huber, 78 Pa. 448, control the determination of the present contention. In Derr v. Greenawalt the will was executed by the testator in the presence of the two attesting witnesses, with an unfilled blank left in it for the name of the residuary legatee. One witness testified that the name was afterwards inserted in the presence of the testator, and by his direction. The person who inserted the name testified that he wrote it, but could not recollect whether he did so at the direction of the testator, or in his presence. The will continued in the possession of the testator a long time after, and it was proved that subsequent declarations were made by the testator that the person named was his legatee. We held that these facts were insufficient to establish the residuary bequest. Mr. Justice Sharswood, stating that there was no difficulty in regard to the rule applicable to the subject, quoted with approbation the language of Chief Justice Gibson in Hock v. Hock, 6 S. & R. 47, as follows: “ Proof of execution must be made by two witnesses, each of whom must separately depose as to all facts necessary to complete the chain of evidence, so that no link of it may depend on the credibility of but one. When the evidence is positive there can be no difficulty, for the witnesses then attest the simple fact of execution itself, but where the evidence of one or both is circumstantial, each must make proof complete in itself, so that, if the act of assembly were out of the question, the case would be well made out by the evidence of either. Circumstantial proof cannot therefore be made by two or more witnesses alternating with each other, as to the different parts of the aggregate of circumstances which are to make up the necessary sum of proof; the evidence of each not going to the whole.”
¶3The facts in Derr v. Greenawalt were that Mrs. Huber testified that her son George Rise wrote the name in the blank in the presence of the testator, and by his express direction, but George Rise was unable to testify that he wrote the name in the blank in the presence of the testator or by his direction, though he said that he did write the name. There was, therefore, but the testimony of one witness as to the essential fact that the name was inserted in the testator’s presence and by his authority.
¶4*608In the present case there was the positive testimony- of Mrs. McDonald that she herself made the erasures after the will was executed and attested, and that she and the testatrix were alone when it was done. She said she herself proposed the erasures, but that her mother, the testatrix, wished it done. The subscribing witnesses testified that the will did not have the erasures in it when they attested it. There was no republication or re-execution of the will after the erasures were made, and there is but one witness who testifies to the fact of the erasures being made by authority of the testatrix. It is impossible to presume either that the erasures were made before execution, or that they were made by the executrix, because the positive and uncontradicted testimony is that they were made after execution and not by the testatrix herself. The statutory proof is therefore lacking and the will with the erasures in it cannot stand as the will of the testatrix.
¶5In the case of Charles v. Huber, supra, the testator made a will devising land to his two sons at a valuation. It was attested by two witnesses. When presented for probate the will showed an erasure of the valuations and other valuations interlined. One of the witnesses having died, his writing was proved and the will admitted to probate. In an issue on appeal the surviving witness testified that the alteration was made after the execution, by direction of the testator, afterwards reacknowledged before the same two witnesses; the handwriting of the deceased witness was proved on the trial of the issue. It was held that the paper as altered was not proved as a will. In the opinion, this court said: “ The will as it stood before alteration was not proved, for the deceased subscribing witness had not proved the will as it had originally stood, but proved it only in its altered state, which left the original text proved by a single witness only. Then the alteration being proved by a single witness only, the paper in neither form was legally proved. The consequence is that the paper fell as a will.”
¶6Both of the cases cited go much further in the direction of the invalidity of the wills in question than it is necessary to go in the present case. At the utmost, under the testimony, the alteration was made by one who was not a subscribing witness, by authority of the testatrix. Both of the subscribing witnesses were examined and one of them said the erasures were *609not in when he signed as a witness, and the other did not know* Buf Mrs. McDonald testified that she was present when the will was executed and saw the testatrix and the attesting witnesses sign, and that the erasures were made by her after that time and within a few days before the death of the testatrix. There was no pretence of republication or acknowledgment, and the proof of the will in its altered state depends only upon the testimony of one witness. Under all the authorities such a will cannot stand.
¶7As the will in its original state, without the erasures, was fully proved by the oaths of the two subscribing witnesses, it ' follows that it was the last will and testament of the testatrix, and in that condition, and as such it must be admitted to probate.'
¶8The order of the court below is reversed, and it is ordered and decreed that the paper writing purporting to be the last will and testament of Phcebe Ann- Simrell, as it was in its original state before any of the erasures therein were made, is her last will and testament, and should be admitted to probate as such, all the costs to be paid out of the estate of the testatrix. The record is remitted for further proceedings.