Opinion by

If it were made clearly to appear that the legacy to Dr. McCollin was cancelled by the testator under the impression that the legatee was dead, the doctrine of Campbell v. French, 3 Vesey, Jr., 321; Mendinhall’s Appeal, 124 Pa. 387; and kindred cases, would save the bequest because it was revoked under a mistaken apprehension, but we agree with the conclusion of the court below that it not only does not appear that the cancellation of the legacy was made under such misapprehension but that the evidence shows that it was made for a different reason. The legacy was annulled in accordance with one of the methods permitted by the Act of April 8,1833, P. L. 249, independently of the superimposed words of cancellation, and the presumption is that such obliteration was made either before the will was signed, or after it was executed, with the intention of revoking the particular bequest. The. words written across the face of the bequest do not explain themselves, but in the light of extraneous facts disclose the testator’s motive with reference to that part of the will to which they refer. There is nothing on the face of the will which necessarily relates these words to the legacy to Dr. McCollin: It is only because they extend from that portion of the will containing the Service legacies on to the part con*584taining the bequest to the appellant that' the fact is assumed that the testator had reference not only to the Service but the McCollin bequests. It is evident that these words were not written by the testator to effect a cancellation but rather to give her reason for her action, for the bequests to Dr. McCollin and to Elizabeth Service and Mary Service are effectually stricken from the will by defacing and obliterating the names of these legatees. It is to be noted that the word “cancelled” in the superimposed words is written across the Service bequest, and that the clause of the will relating to Dr. McCollin is only affected by an extension of the remainder of the sentence and a part of the testator’s name. It could hardly be contended that in the absence of the defacing marks on the McCollin bequest the words in question would have the effect to destroy the bequest to him. When resort is had to other evidence than the will the significance of the inscribed words becomes apparent, for the legatees named in the Service bequests died before the testator and therefore could no longer be objects of her bounty. This fact makes what was written across the will intelligible and reasonable in explanation of the testator’s conduct with reference to the cancellation. The burden is on the appellant to show that the legacy to him was cancelled on the mistaken supposition that he was dead. This he undertakes to do by showing that he is alive and contending that the explanatory writing relates to the legacy to him as well as to the Misses Service. As there is no evidence that the testator believed the appellant to be dead, except that inferable from these words, and as it appears without controversy that they truthfully refer to the Service legacies, we think there is not such a necessary reference to the bequest to the appellant as to justify the conclusion that the testator cancelled his legacy because she thought he was dead. If there is any doubt about this, however, the evidence offered shows that the cancellation of the McCollin legacy was for another reason. The testimony of Mr. Gilfilian is that in a conversation with the testator about a year before her death' she told him that she had cancelled the bequest to Dr. McCollin in her will, and gave as a reason for *585such action that she did not wish Dr. McCollin to receive any of her money. It is necessarily implied in this statement of the testator- that she knew that Dr. McCollin was living at the time of her conversation with the witness, and at that time she referred to the legacy as having been cancelled. There is further evidence that the family of the testator and the family of Dr. McCollin were intimate, and exchanged visits, and presumably lived in the same vicinity, all of which is inconsistent with the theory that the testator believed at the time of the alteration of her will that Dr. McCollin was dead.

The suggestion was made at the argument that Mr. Gilfillan is an interested witness, being a legatee under the will, and therefore not competent, but that point does not seem to be pressed. Any objection to his testimony is rather to its weight than its competency. It was credited by the court below and no reason has been presented why it should not be now. It is not unreasonable or improbable, and is in no way contradicted. It was not offered, as seems to be supposed by the learned counsel for the appellant, to reform an instrument on the ground of mistake, but to identify the legacy to which the inscribed words refer. Without explanation they would be meaningless, and it was competent to show by parol that they had reference to the Service bequests. The decree is in accordance with the evidence and the law applicable to the case and is therefore affirmed.