(dissenting). In a will construction proceeding, testator’s statutory distributees appeal. The Surrogate’s decree, affirmed by a divided Appellate Division, accepted a “ Godfrey Nurse Fund ”, reconstituted in 1967, as the designated residuary legatee under testator’s will.
The issue is whether the Godfrey Nurse Fund established in his lifetime in 1956 by the testator expired in 1966, or was only restructured and its purposes recast by a judgment based on a stipulation in the final accounting of the “ expired ” trust. Put another way, the question is whether the 1967 judgment disposed merely of the remainder interest in an expired and frustrated trust or whether it continued unbroken the life of the 1956 trust so that it might qualify as the named residuary legatee in the 1961 will.
I dissent and would reverse the order of the Appellate Division, and would hold that the 1967 rearranged “ trust ” did not qualify as the residuary legatee.
The trust created in 1956 was intended to establish and operate the Godfrey Nurse Laboratory for Experimental Surgery at Harlem Hospital in the City of New York. It ‘ ‘ expired ’ ’ *394by the terms of the trust instrument on May 23, 1966, and the principal was then to be turned over to the hospital for “ the general purposes of such institution.” Since this was before the testator’s death on December 22, 1968, the residuary clause fails and the residuary must pass by intestacy (EPTL 3-3.7, subd. [e]). The rearranged “ trust ” created in 1967 was, by its terms and effects, a new fund, created out of the remainder of the 1956 trust, not merely an amendment of that trust. As held otherwise by the majority and by the courts below, the 1967 judicial rearrangement of the 1956 trust would constitute an open-ended and free-floating testamentary device which the law proscribes.
Dr. Godfrey Nurse was a wealthy and distinguished physician in New York City’s Harlem. Desiring to benefit the Negro physician and the community where he had made his name and fortune, he established, on May 23, 1956, an irrevocable inter vivos trust known as the 11 Dr. Godfrey Nurse Fund The sole purpose of the trust was “ to furnish and appropriate funds necessary for the operation of the Godfrey Nurse Laboratory for Experimental Surgery at Harlem Hospital ”. The trustees were to pay to the hospital, in their discretion, an annual sum up to $10,000 out of income, and, if necessary, out of principal, for the establishment and operation of the laboratory. At the end of 10 years, or sooner if the laboratory ceased to'function, the principal of the trust, as noted earlier, was to be paid over to the hospital for its general purposes. The trustees were Thomas B. Dyett and the then City Bank Farmers Trust Company.
In his last will dated June 16, 1961, Dr. Nurse bequeathed the residuary of his estate to the trustees of the inter vivos trust known as “ the Godfrey Nurse Fund ” for the benefit of the Godfrey Nurse Laboratory for Experimental Surgery. Notably, in naming the fund, its specific purpose was spelled out by the will so that there can be no mistake which fund was meant.
The purposes of the inter vivos trust were never effected. As Dr. Nurse wrote on May 4,1966 to Dr. Yerby, the then City Commissioner of Hospitals, he had wished that the laboratory in experimental surgery be set up under the aegis of his friend and colleague, Dr. Aubre de L. Maynard, the Director of *395Surgery at Harlem Hospital. When this proved impossible, the city, which owned Harlem Hospital, suggested that the laboratory be installed in a new hospital facility then under consideration; but this too was never done. Meanwhile, Dr. Nurse had approved the use of the irrevocably entrusted fund, for lack of anything better, to finance surgical projects, lectures, scientific travel and related activities sponsored by Dr. Maynard. Now, in May of 1966, with the trust perforce ending, Dr. Nurse had been advised that he could not under the trust instrument prevent the fund principal from passing to Harlem Hospital and the city. In his letter to Dr. Yerby, Dr. Nurse insisted that the remainder of the trust be used solely for the Surgical Service of Harlem Hospital with Dr. Maynard as director of the fund. When Dr. Yerby, in a reply letter of May 24, 1966, suggested that Dr. Maynard could not be the fund director once his association with Harlem Hospital should end, Dr. Nurse instructed the trustees to seek a judicial termination of the trust.
A petition to settle and conclude the trust was filed by the trustees in Supreme Court on November 1, 1966. The petition recited that the trust had terminated. During the proceeding a stipulation by the attorneys for the parties, dated March 31, 1967, agreeing to the disposition of the fund principal, was reached, with Dr. Nurse’s approval. By the stipulation, the fund principal was to be paid to the City of New York, to be held and administered by it as a fund also to be known as “ The Godfrey Nurse Fund ”. Its terms, generally for the benefit of Harlem Hospital, included establishment of lectures, research and travel grants, and a fellowship, all named after the benefactor. The directors, with power only to pass on applications for grants, were to be Dr. Maynard, two other physicians, and the Commissioner of Hospitals. The stipulation was embodied in the Supreme Court judgment entered on May 29, 1967.
(Assuming, dubiously, that posttestamentary evidence of intention is admissible, at about this time, according to the late Mr. Dyett, Dr. Nurse asked him to prepare a codicil to his will leaving the residuary of his estate to the University College of The West Indies, Kingston, Jamaica. The codicil was prepared and delivered, but never executed because Dr. *396Nurse had suffered a stroke. Dr. Nurse died on December 22,1968, not having changed his 1961 will.)
Mr. Dyett, as executor under the will, brought the present proceeding to construe the will. Despondent distributees, who are various next-of-kin, contend that since the fund, for which Dr. Nurse intended the residuary, no longer existed, the residuary -should pass to the statutory distributees by intestacy. The Surrogate properly found no general charitable intent by which to invoke the cy pres doctrine. In support of his holding the Surrogate quoted: “ Where it clearly appears that the testator intended that the property should be applied only to the particular purpose which failed, or for the -benefit of a particular association or corporation which was dissolved, it has been held that the doctrine of cy pres is not applicable and that the property reverts to the settlor or his estate.” (4 Scott, Trusts [3d ed.j, § 399.3.) But he nevertheless found “inter vivos cy pres ” in the stipulation of March 31, 1967, which he construed to amend the original trust declaration rather than to create a new fund. The Godfrey Nurse Fund, was still extant, he thus held, and therefore the beneficiary under the residuary clause of the 1961 will. The Appellate Division affirmed over a strong dissent.
A testator may by will appoint all or part of his estate to a trustee of a pre-existing trust, the terms of which are evidenced by a written, executed instrument, incorporated by reference into, the will (EPTL 3-3.7). There are, however, among others, two prerequisites for such a testamentary disposition. First, the trust instrument must have been executed and acknowledged by the parties, in accordance "with the requirements for the conveyance of real property, prior to or contemporaneously with the execution of the will (EPTL 3-3.7, subd. [a]). And second, the trust must not have been revoked or terminated before the testator’s death (EPTL 3-3.7, subd. [e]).
The Godfrey Nurse Fund “ for the benefit of the Godfrey Nurse Laboratory for Experimental Surgery at Harlem Hospital,” the trust thus designated and limited in. purpose, is the residuary beneficiary of Dr. Nurse’s will. It terminated, according to the instrument of its creation, in May, 1966, 10 *397years after its inception, and significantly, more than two years before testator’s death. A second “ trust ”, if indeed it was a trust or only a conditional gift, also known as the Godfrey Nurse Fund, but having different purposes and a different “ trustee ”, was improvised in 1967, as a compromise to dispose of the remainder of the 1956 trust. This arrangement wa,s not an amended version of the 1956 trust to which the residuary clause referred and described so definitively by purpose. The 1967 arrangement was not created by instrument prior to or contemporaneous with the execution of the will, nor was it set up by any instrument executed in the manner required by law for a “ pour-over ” trust or a testamentary instrument. "With the 1956 trust terminated prior to death and the will unchanged to permit disposition of the residuary to the newly constituted fund or any other beneficiary, the residuary clause fails.
The 1967 device was perhaps not even a trust but most likely only a conditional gift (see, generally, e.g., 7 N. Y. Jur., Charities, §§ 5, 9, 10; see, especially, St. Joseph’s Hosp. v. Bennett, 281 N. Y. 115, 118-123). That the 1967 device was not clearly a trust but perhaps a conditional gift emphasizes that the 1967 occurrence, whatever else it was, was not a continuation of the 1956 trust.
Paramount doctrine in interpreting a will requires following the intention of the testator when he executed it (Matter of Englis, 2 N Y 2d 395, 404; Matter of Hoffman, 201 N. Y. 247, 255). In his 1961 will Dr. Nurse spelled out his residuary beneficiary as “ the Godfrey Nurse Fund for the benefit of the Godfrey Nurse Laboratory for Experimental Surgery at Harlem Hospital in the City of New York ”, Obviously, the identity of name between the 1956 trust and the 1967 fund is of no consequence. The nomenclature served only the purpose of continuing the name of the original but frustrated donor. Without recourse to nontestamentary documents, the will negates the possibility that Dr. Nurse intended, despite his frustration, to leave the residuary to a trust, no longer extant and the primary purpose of which had never been fulfilled and would never be fulfilled.
When the 1956 trust fund had terminated on May 23, 1966, Dr. Nurse having retained no power to modify or revoke the *398trust, the hospital’s (or the city’s) right to the remainder was absolute. The Godfrey Nurse Fund had terminated because the happening of a specified event or the expiration of a stated period of time terminates a trust (see Restatement, Trusts, 2d, § 334; 4 Scott, Trusts [3d ed.], § 334; Bogert, Trusts and Trustees [2d ed.], § 991).
Although a second fund, also dubbed the Godfrey Nurse Fund, was ultimately set up to receive and hold the remainder of the expired trust, this was a new fund created out of the unrestricted remainder, not an amended version of the limited trust designated in the will by name and there defined by purpose as the residuary legatee. The Supreme Court judgment, born of a stipulation by the attorneys for the parties, established the substitute fund only because of a controversy over the disposition of the remainder of the expired trust. True, Dr. Nurse had acquiesced in the use of the original Godfrey Nurse Fund for purposes other than those specified in the trust instrument. A primary reason, as it emerges from his correspondence in the record, was his high regard for Dr. Maynard. When, however, Dr. Nurse realized that, with the ending of the trust and the impending passing of the remainder to the hospital and the city, neither he nor Dr. Maynard would or could continue to control the use of the trust corpus, he finally sought judicial intervention to redirect its use. It is significant that the expired frustrated trust embraced a substantially smaller sum than the anticipated residuary of the estate.
The trustees ’ petition for a judicial settlement and an accounting was to prevent the devolution of a remainder to the city unrestricted except that it had to be used for the general purposes of the hospital. The judicial proceeding was not necessary, however, to terminate the 1956 trust. The specified duration was 10 years and the trustees ’ petition stated that the trust had terminated on May 23,1966. The court, of course, except in extraordinary circumstances, lacked power to extend the trust (see, generally, Ann., Trust Term — Extension, 46 ALR 2d 907).
The stipulation signed by the attorneys for the parties with Dr. Nurse’s consent nowhere suggests, let alone states, that it was intended to amend the original unamendable trust. More*399over, the second Godfrey Nurse Fund had different purposes and a new “ trustee ”, if that it be, the City of New York, as the legal owner of Harlem Hospital, the named remainderman of the 1956 trust. It was stipulated that the fund held by the trustees be turned over to the city “ to constitute a fund ” which was also to be known as “ the Godfrey Nurse Fund ”. Although the city was to “ hold and administer ” the fund, a board of directors was designated, evidently to assist only in the allocation of grants. Their powers, in the stipulation and the judgment based on it, are only to pass on applications for grants.
With this history of the 1956 ‘ ‘ Fund ’ ’ and the radical change in its structure, management, and purpose, it is unacceptable to regard the later fund as a continuation of the first. The only strings of identity are the corpus and the honorific name of the “ Fund ”. This satisfies neither the intention of Dr. Nurse either as settlor or testator, nor does it meet the minimal requirements of the statute which authorizes a “ pour-over ” testamentary disposition into an existing trust.
Moreover, the Surrogate’s reasoning in rejecting the application of the cy pres doctrine to the residuary clause was, as noted earlier, quite correct, because the testator’s will had not evidenced a general charitable intent. It is difficult to see how a more generalized intent was applicable to the inter vivos trust, before or after the 1967 judicial surgery. Before the judicial surgery the inter vivos declaration of trust was as specific in intent as the residuary clause. After the judicial surgery, the new fund’s purpose could not satisfy the specific intent of the residuary clause. In any event, that judicial surgery could not qualify as an amendment to a “ pour-over ” trust because the form of the “ amendment ” did not satisfy thq • statute.
Indeed, the underlying stipulation for the judgment could not qualify as an amendment to a “ pour-over ” trust under EPTL 3-3.7. Paragraph (1) of subdivision (b) of the statute requires that any amendment to a trust must be “ executed and acknowledged in the manner herein provided for executing and acknowledging the instrument which it amends.” Under EPTL 3-3.7 (subd. [a]) the trust instrument must be executed and *400acknowledged 1 ‘ in the manner required by the laws of this state for the recording of a conveyance of real property ’ \ The 1967 stipulation was signed not by Dr. Nurse, but by his attorney. An authorized agent may execute a deed or conveyance for the grantor, but the authorization must be in writing (Beal Property Law, § 243; General Obligations Law, § 5-703). Assuming Dr. Nurse’s attorney was so authorized, the trust instrument would have to be acknowledged by the executing parties or proven by attestation (Beal Property Law, §§ 243, 291). Neither the stipulation nor the judgment into which it was incorporated, were formally acknowledged or attested.
Of course, the requirement of EPTL, concerned only with the manner of execution of a trust instrument, is not met by an instrument of record, even a judgment. It is met only by an instrument executed by the settlor in a manner, that is, with the formalities, that a recordable instrument must be executed. The trust instrument need not be “ recordable ”. Indeed, unless real property or a security interest is involved, it will be rare, or never, that a trust instrument or its amendment will be recorded. The statutory reference to a manner of execution is to assure authentic execution and not to produce a recorded or recordable instrument. Hence, the fact that the judgment in the 1967 proceeding is a record instrument, because rendered in a court of record, is just as immaterial as if it were a Us pendens or a pleading which would also be of record. Section 297-b of the Beal Property Law is not concerned with the manner of execution of an instrument and applies only to a judgment affecting title to real property and permits recording in the index of real property records.
The Legislature, in enacting EPTL 3-3.7, was aware of possible abuse in permitting an estate to devolve by nontestamentary documents. If pour over trusts were to be allowed, their unrestricted amendment was deemed vulnerable to fraud and mistake (Beport No. 6.1B, Temporary State Commission on the Modernization, Eevision and Simplification of the Law of Estates, N. Y. Legis. Doc., 1963, No. 19, pp. 286, 313). EPTL 3-3.7 was therefore intended to enforce the recognized requirements for valid incorporation by reference, namely, that the will refer to a then-existing document, describing it so that it *401could be unequivocally identified, and that the incorporation reflect the testator’s unambiguous intention (p. 290).
Whichever way one turns, the 1967 stipulation-judgment fails to sustain a continuing amended trust. As an amendment it does not satisfy the requirements for execution mandated by statute. In content it does not supply an amendment to an unamendable trust declaration but sets up a different fund without a trustee and with different purposes. In purpose and intention it was only to placate the frustrated benefactor by placing some limitation on the use of the otherwise largely unrestricted remainder of the inter vivos trust. The stipulation-judgment is a far cry from establishing a pour over trust into which the residuary clause can be twisted to pass the residuary of the estate.
Tested, as it must be, by these standards, the 1967 fund as interpreted by the courts below would be an open-ended, free-floating testamentary document that EPTL 3-3.7 does not and was never intended to allow. As for Dr. Nurse’s intention, the residuary clause is explicit and it is legally perverse to look elsewhere. Moreover, as the dissenters below aptly noted, it is cruel irony indeed that the residuary should pass to an improvised fund whose very existence he had been forced reluctantly to accept by way of compromise, caused only by the absence of a reverter provision in the 1956 inter vivos trust instrument. In short, the manipulation of the trust into the rearranged fund as a continuing legal entity is legally impermissible, and the violation of testator’s intention gross. Even if one knew that he would have consented to such a result, and one knows the contrary, the law of testamentary disposition and the devolution of trusts would forbid it.
The law prefers, it is true, to avoid intestacy; but it absolutely rejects rewriting another’s will after his death and prefers intestacy to the more abhorrent result (Matter of Englis, 2 N Y 2d 395, 401, 405, supra).
Accordingly, I dissent and vote to reverse the order of the Appellate Division and to remand the proceeding to the Surrogate’s Court for distribution of the residuary by intestacy.
Judges Gabbielli, Jones and Wachtleb concur with Judge Babin ; Chief Judge Bbeitel dissents and votes to reverse in a *402separate opinion in which Judge Jasen concurs; Judge Stevens taking no part.
Order affirmed, with costs to all parties appearing separately and filing separate briefs payable out of the estate.