The question presented by this appeal is whether the value of the corpus of an irrevocable trust created in 1929 is includible in the grantor’s estate as a transfer “intended to take effect in possession or enjoyment at or after his death” within the meaning of section 302(c) of the Revenue Act of 1926, 44 Stat. 70, 26 U.S.C.A. Int. Rev.Acts, page 227. The Tax Court decided adversely to the commissioner’s contention, and he seeks reversal of its order.

The facts were stipulated. Frank Hall, grantor of the trust, died testate, a resident of New York City on July 31, 1939. The executor under his will is Irving Trust Company. This company is also the trustee under the irrevocable trust set up by the grantor in February 1929. He was then a childless widower, 64 years of age, residing with his sister Bessie H. Walton and her two children. By the terms of the trust the net income thereof was to be paid to the grantor during his life, and upon his death the trustee was to divide the trust property into four separate trusts: one for his sister Elinor H. Blatchford during her life, one for his sister Bessie II. Walton during her life, one for his nephew Peter H. Walton during his life and one for his niece Sarah Walton during her life, the nephew and niece being given the privilege of demanding, on reaching the age of 40, the principal of his or her trust. In case any of the above named beneficiaries should die before the grantor, the property which would have been held in trust for such deceased beneficiary if living, was to be added to one of the other trusts. Upon the death of the beneficiary of any of the aforesaid separate trusts, the trustee was directed (paragraph 5) “to divide the principal thereof among his or her issue then living, if any, or, failing such issue, to transfer and pay over the principal to his or her brother or sister, if then living, or to the then living issue of such brother or sister, if then deceased, or failing all of the foregoing, to the other next of kin of the Grantor then living, to be divided among them per stirpes.” Upon the death of the grantor, “if none of the foregoing persons named or designated survived him,” the trustee was directed (paragraph 6) to divide the trust property among the next of kin of the grantor then surviving. Paragraph 7 provided that “next of kin” means “only those who are of the blood of the Grantor.” At the time the trust instrument was executed there were living twenty-two persons who were his near next of kin of the blood, the *174youngest being eight years old. When the grantor died twenty-one of such persons were living. Paragraph 8 is a spendthrift trust clause providing that the “interest of any beneficiary of a trust hereby created shall not be transferable by assignment or otherwise, and the benefits accruing under any trust to any beneficiary thereof shall not be subject to commutation or incum-brance or to legal process in any form of action or proceeding whatsoever.”

The commissioner’s argument that because the grantor reserved the net income for life, the corpus is taxable under § 302(c) cannot prevail in this court. The Joint Resolution of March 3, 1931, 46 Stat. 1516, has no retroactive application. Hassett v. Welch, 303 U.S. 303, 58 S.Ct. 559, 82 L.Ed. 858. Until otherwise instructed we shall continue to follow May v. Heiner, 281 U.S. 238, 50 S.Ct. 286, 74 L.Ed. 826, 67 A.L.R. 1244, as we have recently done in Helvering v. Proctor, 2 Cir., 140 F.2d 87, 155 A.L.R. 845; and Commissioner v. Irving Trust Co., 2 Cir., 147 F.2d 946.

The argument gains no additional strength by reference to the spendthrift clause and the contention that thereby the grantor projected his control of the trust property beyond his death. After creating the trust the grantor has nothing to do with respect to the spendthrift provisions, and his death cannot affect them. As Mr. Justice Holmes stated in Shukert v. Allen, 273 U.S. 545, 547, 47 S.Ct. 461, 71 L.Ed. 764, 49 A.L.R. 855, with respect to a trust to accumulate income, the trust is “unaffected whether he lives or dies.” With respect to the grantor’s life income, the clause was ineffective. Schenck v. Barnes, 156 N.Y. 316, 50 N.E. 967, 41 L.R.A. 395; N. Y. Personal Property Law, sec. 15, Consol. Laws N.Y. c. 41; secs. 98, 103 N. Y. Real Property Law, Consol. Laws N. Y. c. 50. With respect to the remainder interests the spendthrift provisions were inapplicable. Bergmann v. Lord, 194 N.Y. 70, 75, 86 N.E. 828; 1 Bogert, Trusts and Trustees, § 220.

The principal contention is that the grantor retained a possibility of reverter, however remote, which requires inclusion of the trust property in his gross estate. We think not. The property was conveyed to the trustee; the grantor’s life interest was followed by the life in-» terests of the named beneficiaries who survived him, with • remainders over upon their death, and, if all the foregoing gifts failed, remainders were given to “the other next of kin of the grantor” (paragraph 5). Then to preclude all possibility of reverter in case the property should fail to pass under paragraph 5, by paragraph 6 he gives remainders to his next of kin. We cannot see ’ that he retained any “string” by which he could regain possession or control of the trust property or that his death altered any of the interests created by the trust instrument. The only possibility of reverter he had was the possibility that by operation of law the property might revert to his estate in case he should survive his own next of kin — if that be conceivable.1 Such a possibility does not subject the property to an estate tax. Commissioner v. Kellogg, 3 Cir., 119 F.2d 54; Lloyd’s Estate v. Commissioner, 3 Cir., 141 F.2d 758; Commissioner v. Lasker’s Estate, 7 Cir., 141 F.2d 889. And the recent Supreme Court decisions upon which the commissioner relies — Fidelity-Philadelphia Trust Co. v. Rothensies, 324 U.S. 108, 65 S.Ct. 508; Commissioner v. Estate of Field, 324 U.S. 113, 65 S.Ct. 511; Goldstone v. United States, 325 U.S. 687, 65 S.Ct. 1323, do not require a contrary holding.

Decision affirmed.

(dissenting).

In Helvering v. Proctor, 2 Cir., 140 F.2d 87, the government urged us to hold that Helvering v. Hallock, 309 U.S. 106, 60 S.Ct. 444, 84 L.Ed. 604, 125 A.L.R. 1368, had overruled May v. Heiner, 281 U.S. 238, 50 S.Ct. 286, 74 L.Ed. 826, 67 A.L.R. 1244. A majority of our court refused to so hold; I dissented. For unexplained reasons, the government, having described Proctor to us as a test case, failed to seek certiorari. Its strategy seems to have been, instead, to make flank attacks designed to wear down May v. Heiner. That strategy seems to have been successful. See Fidelity-Philadelphia Trust Co. v. Rothensies, 324 U.S. 108, 65 S.Ct. 508; Commissioner v. Estate of Field, 324 U.S. 113, 65 S.Ct. 511; Goldstone v. United States, 325 U.S. 687, 65 S.Ct. 1323.1

*175• Ordinarily, when a case has been decided by this court against my dissent, I do not again dissent when the same question again arises.2 But the fact that in the Fidelity-Philadelphia and Field cases, Mr. Justice Douglas, in concurring, indicated that the Supreme Court had not yet decided whether Heiner survived Hallock, persuades me that it is proper here once more to dissent for the reasons set forth in my dissenting opinion in Helvering v. Proctor, supra.

I would, however, not favor reversal here but, instead, woud remand to the Tax Court for determination of the issue whether the grantor in fact relied upon May v. Heiner, i.e., failed to relinquish his life interest on the assumption that May v. Heiner would apply.

In the Proctor case, I suggested that, if Heiner has been overruled, it should nevertheless be regarded as effective in any instance wliere it had thus been relied upon. The rational basis for judicial reluctance to overrule a gravely unsound decision, with not only prospective3 but also retroactive effect, would cease to exist if stare decisis were treated, in that way, solely as a sort of estoppel doctrine.4 Except when a person has detrimentally changed his position because of his knowledge of a decision, I can see no valid reason for adhering to a precedent when, upon subsequent reflection, it shows up as ill-considered and unwise. To assume, without opportunity to prove the contrary, that men have so changed their position — have acted or been inactive because they had in mind a court’s decision — and, on that ground, to refuse to wipe it out, no matter how erroneous, does not seem to me to be an intelligent method of administering justice. For it is difficult not to agree, with reference to most of the conduct of the great majority of men, with the following comment by John Chipman Gray, a profound legal thinker and experienced practical lawyer: “Practically, in its application to actual affairs, for most of the laity, the law, except for a few crude notions of the equity involved in some of its general applications, is all ex post facto. When a man marries, or enters into a partnership, or engages in any other transaction, he has the vaguest possible idea of the law governing the situation, and, with our complicated system of jurisprudence, it is impossible it should be otherwise. If he delayed to make a contract or do an act until he understood exactly all the consequences it involved, the contract would never be made or the act done. Now the law of which a man has no knowledge is the same to him as if it did not exist.”5

In my dissent in Proctor, I suggested that the burden of showing reliance on Heiner should be on the taxpayer. But now (reverting to an earlier suggestion6) I think it would be more fair, if Heiner has been overruled, to remand with directions that the burden of proof on the issue of reliance should rest on the government.