A single, narrow question is involved in this dispute. Did Evelyn Chew Gifford, settlor, effectively revoke the revocable deed of trust which she created in 1938?
The answer turns on the quantum of proof necessary to establish the revocation of a formally prepared written trust instrument.
Our research, and that of learned counsel, reveal that there is a dearth of authority on the subject. We are in agreement with the ruling of the auditing judge that where no method or procedure for revocation is prescribed in a revocable trust, it may be revoked informally by parol without consulting the trustee. This conclusion is supported by the following statements found in A. L. I. Restatement of the Law of Trusts 2d, §330, Comment (i), p. 138:
“If the settlor reserves a power to revoke the trust but does not specify any mode of revocation, the power can be exercised in any manner which sufficiently manifests the intention of settlor to revoke the trust.
“A revocation may be effective even though the settlor does not attempt to communicate his decision to the trustee. It may be sufficient that he manifest his decision to revoke the trust by communicating it to the beneficiaries or to third parties. . . .”
A question with respect to the amendment of a revocable trust arose in Ronsor Trust, 6 Fiduc. Rep. 371 (1956), an adjudication by our colleague, Judge Shoyer, to which no exceptions were filed. Although Judge Shoyer’s enunciation of the applicable principles of law in that case is clearly correct, we think the case is readily distinguishable from the present case by a substantial difference in facts.
*771The law is settled beyond question that to prove a trust by parol, the evidence must" be clear, precise and indubitable: Danner v. Danner, 366 Pa. 178 (1950). In Moffitt v. Moffitt, 340 Pa. 107, 110 (1940), Mr. Chief Justice Schaffer, speaking for a unanimous court, was even more emphatic. He said: “Evidence to support a parol trust must be direct, positive, express, unambiguous and convincing: Dorr v. Leippe, 286 Pa. 17, 132 A. 806; Kirk v. Ford, 330 Pa. 579, 200 A. 26.” To the same effect, see also Williams Estate, 349 Pa. 568 (1944); Rocks v. Sheppard, 302 Pa. 46 (1930); Washington’s Estate, 220 Pa. 204 (1908); Roth Estate, 15 D. & C. 2d 473 (1959).
Although it is apparent that the nature and extent of parol evidence required to establish the revocation of a written revocable trust instrument need not be as conclusive as that which must be produced to establish the creation of the trust by parol, we are firmly of the opinion that public policy demands that the proof of revocation b'e clear and convincing.’ Any other rule would be an open invitation to chicanery and fraud, especially where, as in the present case, settlor is dead and the trust instrument being challenged is a formal document.
Settlor’s letter of April 2,1957, sent to Girard Trust Corn Exchange Bank, the trustee, standing alone, was clearly not, in itself, sufficient to revoke the trust. In this letter she stated:
“It is my wish to revoke the trust under the deed which I made on January 8,1938, together with amendments thereto.
“Will you therefore, as soon as possible, send me the necessary papers to accomplish' this.”
The burden was, therefore, on exceptants to establish that settlor had effectively revoked the trust by producing additional competent and convincing evidence. The *772learned auditing judge properly decided that exceptants have failed to meet this burden.
Whether or not settlor effectively revoked this trust, depends in large measure upon the credibility of the witnesses called by the parties. In such matters, the conclusions of the auditing judge, who saw and heard the witnesses, must be given the same effect as the verdict of a jury and will not be disturbed unless there is manifest error or clear mistake. See cases cited in 4 Hunter, Pennsylvania Orphans’ Court Commonplace Book (Smith and Aker, 2nd Ed.), §17(a), p. 145. We are not only in full accord with the auditing judge’s factual findings, but believe that no other conclusion could have been reached, reasonably, under the existing circumstances.
Exceptants’ case must, in the last analysis, stand or fall on the testimony of George DuManoir, who would benefit most if the trust was held to be revoked. His relationship to settlor was most unusual. By his own testimony, like “The Man Who Came to Dinner” in the Kaufman and Hart play of the same name, he went to the Giffords for a short visit and stayed 25 years. Apparently, he was some sort of a social companion to settlor and her husband, playing tennis and bridge with them, and making himself generally agreeable. There is no suggestion in the record that he paid for his board or expected compensation for his services. Nevertheless, after settlor’s death, he made claim against her estate for $50,000 for services rendered. Under the circumstances, the auditing judge properly placed little reliance upon his testimony.
The testimony of the other witnesses called by Mr. DuManoir was all so flimsy and casual that it must be disregarded as trivial.
On the other hand, the testimony of W. Sanders Framling, who had been settlor’s attorney for many years, and of her nephew, John Upshur Moorehead, *773one of her financial advisors, both of whom were completely disinterested and trustworthy witnesses, was most persuasive and satisfactorily convinces us that settlor did not intend to revoke the trust in any manner except by executing a formal writing prepared under the direction of the trustee.
Our conclusion is strengthened by the fact that the trust instrument was amended by settlor on three occasions and partially revoked on two others, in each instance by formal writings duly executed by her before a notary public.
The exceptions are therefore dismissed and the adjudication is confirmed absolutely.