The exceptions raise the question whether this appointment by a donee of a general power of appointment is void in whole or part as contravening the rule against perpetuities. The rule, which for centuries has had much learning and effort expended in its application, was one of the methods adopted by the judiciary, in the interest of public policy, to prohibit the undue limitation and prolongation of estates. Mr. Gray in his learned treatise “The Rule Against Perpetuities” (3d ed.), says (p. 2) that the rule is a mode “adopted by the Common Law for forwarding the circulation of property which it is its policy to promote”. Without attempting to trace the develop*417ment of the rule it will suffice to say that the limit, under the rule, for the creation of executory interests to commence is “within the period of a life or lives in being, and twenty-one years, allowing for the period of gestation”: Smith’s Appeal, 88 Pa. 492, 496. It should be noted that “It matters not how many lives there may be so that the candles are all burning at the same time, for the life of the longest liver is but a single life”: Smith’s Appeal, supra.
In testing any given case under the rule there is a distinction made between the effect of testamentary or inter vivos provisions made by the testator or settlor himself, and those made by a donee of a power of appointment under the terms of the trust. The distinction while narrow, is, nevertheless, extremely important. In cases of trusts created by a testator or settlor, the validity of the disposition is governed by possible events, not necessarily by actual events as they may have happened to materialize. For instance, if an interest could arise beyond the limit fixed by the rule, the disposition is void irrespective of the fact that such a person was in fact born within the limitation of the rule. The test is whether, under the provisions of the trust, it is possible for a person to take an interest held to be too remote under the rule. If so, the limitation is void: 1 Jarman on Wills (7th ed.), 266, 288 et seq.; Lilley’s Estate, 272 Pa. 143; Ledwith et al. v. Hurst et al., Execs., 284 Pa. 94; Scott’s Estate, 301 Pa. 509; Friday’s Estate, 313 Pa. 328.
Contradistinguished from the case of the creation of a future interest in the original will or deed is when such interest is created by the donee of a power. While the remoteness of the appointed estate is still measured from the time of the creation of the power: Lawrence’s Estate, 136 Pa. 354; it is the fact and not the possibility which rules. If the appointment as actually made does not violate the rule, such appointment is not rendered void merely because the appointee might have appointed in a manner too remote. It is well settled that the exercise of a power of appointment is not rendered void because of the fact that within its terms an estate might be created which possibly would be too remote. The determination of the validity of the appointment in such cases depends upon the facts as they actually exist at the time the appointment is made and not possibilities: Lawrence’s Estate, 136 Pa. 354; Boyd’s Estate (No. 1), 199 Pa. 487; McClellan’s Estate, 221 Pa. 261; 48 C. J. 978, §63, 979, §65, note 74.
To illustrate the distinction in the operation of the rules: Suppose A bequeaths his estate to B for life, and after B’s death to B’s children for the life of the survivor with provision that, upon the death of any of B’s children so possessing a life estate, their children should take the deceased parent’s share until the decease of the surviving child, with remainder over to the issue of all children. Obviously, the substitutionary provisions for the disposition of the income until the death of the surviving child would be void as too remote. It would be possible that one of B’s children might die over 21 years after B’s death, and then let in a new estate. Therefore, the whole provision, as to the substitutionary gift of income would be void, as would the remainder. On the other hand, if A gives his estate to B for life, with a general power of appointment, and B gives the estate to her children, who are all living at the death of A, for life of the survivor, and under exactly the same provisions as above stated, then the whole disposition is valid. The naming of these children, so living, is but “lighting another candle” which is burning during the lives of A and B. It is still necessary to look to the original will or deed, and to read into it the terms of the appointment. However, where it appears that the appointment in fact names persons in being in the lifetime of the donor, the *418disposition is valid and is not rendered void because it might have been possible to appoint otherwise.
The only case in Pennsylvania which appears contrary to the above rules is Smith’s Appeal, 88 Pa. 492. It is true that this case in terms has never been overruled. Part oí the decision, however, was expressly overruled. The remainder has been criticised by eminent text book writers, doubted by the learned justice who wrote the opinion, and has not been followed by Judge Pen-rose in this court, affirmed in a later decision of the Supreme Court. In this situation we are of opinion that Smith’s Appeal is not now a binding authority upon the principle upon which it was originally decided. A discussion of the facts and decision, in Smith’s Appeal, and its subsequent history, may be helpful in illuminating the situation as- it exists in the present case. In Smith’s Appeal the testator devised the residue of his estate to trustees to pay a share thereof inter alia to a daughter for life and upon her death to pay the principal to such persons and upon such uses as the daughter might appoint. The daughter appointed to her children for life with a general power of appointment, and in default to their heirs. A most important element, which is particularly emphasized herein, is that all the children (donees) of the daughter were living at the time of the death of testator. The lower court (the Orphans’ Court of Philadelphia) upheld the validity of the trust for the benefit of the children of the daughter for life: 36 L. I. 226. The Supreme Court, in an opinion by Mr. Justice Paxson reversed this decision and declared the entire appointment void as against the Rule against Perpetuities: 88 Pa. 492. Obviously, that part of the appointment which passed the estate to the children of the life tenant (daughter) for life ought to have been upheld because the life estates were created and commenced to operate immediately upon the ending of an existing life, which life was in existence at the time of the death of the original testator. The Supreme Court in Lawrence’s Estate, 136 Pa. 354, 355, expressly overruled this part of Smith’s Appeal, and directly decided that “When the donee of a power to devise, who was in being at the creation of the power, appoints by will, in trust for life-tenants to take at his death, with remainder over, such appointment for life will be good, whether the appointees for life were born before or after the creation of the power, and even if the estate appointed in remainder be too remote”. This of course corrected the original error and since Lawrence’s Estate this has been the uniform decision of the Supreme Court. But upon the main point, to wit, the validity of the remainders, Smith’s Appeal was not expressly overruled. Mr. Justice Paxson expressly ruled that the exercise of the power of appointment must be written into the will of the donor, and all of its provisions must be considered as if contained in the will of the donor. The learned justice then makes the test of remoteness as of the date of the death of the donor. He -treats the appointment as if made by the donor himself and subjects it to the test (p. 495) of “possible and not . . . actual events”. This of course is not the generally accepted rule. All authorities are now in accord that the true test is to ascertain what in fact the donee of the power actually did, and if that does not transgress the rule the appointment is valid.
It is interesting to note that Mr. Justice Paxson, the writer of the opinion in Smith’s Appeal, says in Coggins’ Appeal, 124 Pa. 10, 30: “Subsequent reflection has left some doubt in my mind as to the soundness of the ruling in that case [Smith’s Appeal] upon the main question involved, and as I wrote the opinion I may be allowed to criticise it.”
*419Gray on the Rule Against Perpetuities (3d ed.), which is the leading text hook upon this subject, severely criticises this decision. He says:
“§523. In Smith’s Appeal A. devised property in trust for B. for life, and on her death as she should by will appoint. B. appointed to trustees in trust for her children for life without liability for their debts, and directed that on the death of each child its share should go over. All B.’s children had been born in the lifetime of A. The gifts over were held void for remoteness. But, with submission, this was incorrect. The Court say that B. might have had children who were not born till after A.’s death. But at B.’s death, when the appointment took effect, her children were a fixed number which could neither be increased nor diminished; ‘children’ was simply a compendious designation for certain determined individuals. It will not be questioned that if B. had appointed the property to John, Thomas, Mary, &c., naming all her children, all of them having been born in the lifetime of A., the gift over on their death would have been good; for it would have been good if inserted totidem verbis in the will of A. Now that is exactly what she has done: the words of gift to all the children, as used by B., are identical in meaning with words appointing to John, Thomas, Mary, &c. They have not merely turned out to be the same (which would not be enough), but they could never mean anything else; for at her death, when the words of her will become operative, her children were John, Thomas, Mary, &c., and could never be anybody else. The decision in Smith’s Appeal makes the question turn upon whether B. used one of two expressions, which, in her will, would have precisely the same meaning; and the reason for this distinction is that the words, if used by A., would not have had the same meaning. There is no question of public policy involved. B. could confessedly have accomplished her object, if she bad used words which in her will would mean exactly the same as the words she has used, neither more nor less; not, it must be again observed, words which have turned out to mean the same, but words which at and from the time of their going into effect could never possibly mean anything else. Such verbal and arbitrary distinctions are entirely opposed to the spirit of the Rule against Perpetuities, which, though severe, is rational. Mor gamy. Granow is opposed to Smith’s Appeal.
“§523 a. Mr. Lewis’s remarks and the decision in Smith’s Appeal seem to have been prompted by confounding two separate questions: (1) What appointment is in fact made? (2) Was it certain at the time of the creation of the power that such appointment, if made, would vest within twenty-one years after lives in being? The first question is to be determined by considering what the language of the appointment means as used by the donee, at the time he uses it; and thus having found out what appointment the donee means to make, the second question is to be asked, viz.: was it certain, when the power was created, that, if the donee should ever make the appointment he has in fact made, it must vest within the required limits? Therefore all present unconditioned appointments by a donee living when the power was created are good; for they take effect, if by deed, in his lifetime; if by will, immediately on his death. So all appointments on the death of persons living when-the power was created are good. It makes no difference that the appointment is in form to a class, if the class is closed before the appointment, as in Smith’s Appeal; or that a condition is in terms attached to the appointment, if the condition is in fact fulfilled before the appointment, as in Morgan v. Gronow. The error in Smith’s Appeal was that the Court went to the creation of the power, not only to determine the validity of the appointment (which was right), but to determine the meaning of the appointing instrument (which was wrong). It considered not *420what the words of appointment meant as used by the donee, but what they would have meant if used by the creator of the power.
“§523 b. The happily turned but ambiguous phrase that an appointment should he considered ‘as if written into the original instrument’ was, probably, the source of the error. The statement is true if understood of the validity of an appointment. It is not true if understood of the meaning of the words used in making an appointment. The Court, in Smith’s Appeal, seems to have adopted and applied the expression without sufficient consideration of its true force; and consequently a wholly proper gift was rendered void by the imposition of an arbitrary requirement, not called for by the Rule against Perpetuities nor by public policy. These were perfectly protected by subjecting the appointment to the test given in the preceding section. And what was this arbitrary requirement? It was that words were not to be given the meaning which they really had when used by the person using them, but the meaning which they would have had if used at some other time by some other person. That is, a fiction was invoked, for no useful end, to defeat a lawful object. It is easier to believe that the decision passed per ineuriam than that the learned Court deliberately contemplated such a result. The history of the Rule against Perpetuities is full of slips by eminent judges, often acknowledged by themselves.”
Judge Penrose of this court regarded Smith’s Appeal as overruled when he wrote the opinion, affirmed per curiam, in Boyd’s Estate (No. 1), 199 Pa. 487. Certainly the affirmation by the Supreme Court of Judge Penrose’s opinion in that case had the effect of overruling Smith’s Appeal, which unquestionably was an erroneous statement of the law, subsequently criticised not only by its author and by such an eminent text writer as Gray, and not followed by either Judge Penrose or the Supreme Court. In our opinion the decision in Boyd’s Estate, taken in consideration with the facts, squarely settled any doubt as to the state of the law now prevailing. In Boyd’s Estate the trust was for the benefit of a daughter for life, and upon her death according to her general appointment. The daughter exercised the power of appointment in favor of her son by providing that at the age of 25 years he should receive $25,000 absolutely and as to the residue for the son for life, with remainder over. Contrary to Smith’s Appeal, Judge Penrose ruled that (p. 493) “the gift to him [the son] of $25,000 upon attaining the age of twenty-five years — if that event occurred within twenty-one years of the death of his mother, the donee of the power (but not otherwise) — do not infringe the rule against perpetuities”. In other words, Judge Penrose and the Supreme Court looked to see what the mother (the donee of the power) actually did. It appeared that the son was in fact over 4 years of age when his mother died. That being so, facts and not possibilities govern. As the son must attain 25 years within 21 years from the date of death of his mother, the bequest was held good. As to the residue, the grandson was not in being at the death of his grandfather, but survived his mother, the donee of the power. Following Lawrence’s Estate, the life estate in the son was held valid, because it took effect upon the death of the life tenant. Obviously, the remainder was bad and was held void as too remote.
Cox et al. v. Dickson, 256 Pa. 510, and Hazard’s Estate, 25 Dist. R. 225, have had our consideration. Both of them involved the same will. It is to be observed, however, that the donor died in 1865. The appointee under the donee’s appointment was not born until 1873. Tested under the above authorities, when the appointment is examined, the appointor in fact violated the rule in that she created future estates too remote. Had the appointee been born during the life *421of the original donor and survived him, the appointment would not have violated the rule. Herein lies the distinction and illustrates the rule.
We are therefore of opinion, and so decide, that a power of appointment is not rendered void because within its terms an appointment might possibly have been made which is too remote. The question depends upon the validity of the appointment which was in fact made.
With these principles as our guide we have applied them to the facts of the instant case. E. Burgess Warren placed his residuary estate in trust, inter alia, to pay over to his daughter, Caroline, one fourth of the net income therefrom for life. At the death of Caroline, his trustee was directed to hold and dispose of the fourth share in the principal or corpus “in such manner as shall be directed in and by [Caroline’s] last will and testament”. There were provisions that in case Caroline should die intestate, the trust estate should be paid to Caroline’s children and their issue. Caroline died testate and appointed the estate to her trustee, to hold for the life of her husband (who has since died) and after his decease “to divide the income of the said trust estate into as many parts or shares as at the times of half-yearly distributions there shall be children of mine then living and children of mine then dead represented by descendants then living, subdividing the share of income falling to each set of descendants of a then dead child of mine among them per stirpes, upon the principle of representation, and to pay, assign, transfer and set over to each child and descendant of a deceased child who shall thus be found entitled, his or her share of the income half-yearly. The share of income thus falling to any descendant of mine who may be a minor, to be used and applied by my Trustee in and about the maintenance, education and support of such minor during minority, directly, without the intervention of a Guardian.” The trust as respects income is to continue until “the decease of the last survivor of my children”. At this time — the death of Caroline’s surviving child — the principal is to toe distributed as follows: “to divide the principal of said trust into as many parts or shares as there shall be children of mine then dead represented by descendants then living, sub-dividing the share of principal falling to each set of descendants of a dead child of mine amongst them per stirpes, upon the principle of representation, and to pay, assign, transfer and set over to each descendant of a deceased child of mine who shall thus be found entitled, his or her share absolutely and in fee.” Caroline died leaving surviving a husband (who is now deceased) and two children (married daughters), both of whom are still living. The daughters each have issue, who are minors represented by a guardian ad litem.
The controlling fact, in our opinion, is that when Caroline died, and her appointment went into effect, her two children surviving (Margaret F. B. Purves and Barbara B. Jefferys), were both living at the death of E. Burgess Warren, the donor.
Therefore, under Lawrence’s Estate, Boyd’s Estate, and McClellan’s Estate (supra), these children being alive at the donor’s death, the appointment to them for the life of 'the survivor is clearly within the rule. As the survivor is but one life, the contingent interest in income of the issue of a deceased child until the death of the surviving child is also valid. While no one has questioned the validity of the present life estates in the children of Caroline as they presently exist, yet we emphasize the atoove situation because of the claim of these children that their life estates have merged with a void remainder. These children do not possess vested life estates. When one of them dies, their interest ceases, and it is paid to their issue until the death of the survivor. Therefore, even though the remainders should be void, and it should be determined *422that under E. Burgess Warren’s will these children have vested remainders, there can be no merger because the children do not possess both the vested life estates and remainders. As the life estates are validly appointed, it is premature for the court to pass upon the validity of the remainders: Lewis’s Estate, 29 Dist. R. 796 (affirmed in 269 Pa. 379); Lockhart’s Estate, 29 Dist. R. 179 (affirmed in 267 Pa. 390); Whitman’s Estate, 248 Pa. 285; Ewalt v. Davenhill et al., 257 Pa. 385; Jones’s Trust Estate, 284 Pa. 90, and Hecht’s Estate, 316 Pa. 12.
It is argued that because Caroline had a life estate and a general power of appointment, she possessed the equivalent of a fee and hence the validity of her appointment should be measured commencing as of her death. But Caroline did not possess a fee. It is to be noted that there was a gift over in default of appointment. The alternative gift over precludes an absolute ownership: Lyon et al. v. Alexander, 304 Pa. 288; Rehr v. Fidelity-Philadelphia Trust Co., 310 Pa. 301; Roberts’ Estate, 18 D. & C. 463. Curran’s Estate, 312 Pa. 416, and Perkins’s Trust Estate, 314 Pa. 49, were surcharge cases where appointees under a general power of appointment were held to be bound by acts of the donee with respect to consenting to investments. But, irrespective of this, Caroline, the donee of the power, expressly chose to exercise it — npt as her own estate but as the estate of the donor over which she possessed the right of appointment.
Neither do we regard the trust void in its entirety as a mere scheme or design to circumvent the rule. There are cases, like Feeney’s Estate, 293 Pa. 273, and Johnston’s Estate, 185 Pa. 179, where such a plan was attempted. In those cases the life estates were held merely incidental to and for the purpose of supporting void remainders. Mr. Warren gave his daughter a life estate, and she, under her appointment, gave her children (grandchildren of Warren living at the time of his death) life estates. Irrespective of how remote the remainders may have been, in such circumstances, the life estates are valid: Whitman’s Estate, 248 Pa. 285.
We are therefore of opinion that 'Caroline’s appointment of income to her two children (who were living at the decease of the donor) and their issue for the life of the survivor of them was a valid exercise of the power; that while unquestionably presumptively the remainders are valid, yet because there is no merger, and no general invalidity of the whole scheme, we refrain from passing upon the validity of such remainder. We agree with the auditing judge that the fund must be awarded back to the trustees for the purposes of the trust. We refrain from passing upon the validity of the remainders until the termination of the life estates.
Exceptions of Mrs. Jefferys and Mrs. Purves are dismissed. The exceptions of the guardian ad litem, nos. 1, 3 and 4, are sustained, and no. 2 is dismissed. The adjudication as thus amended is confirmed absolutely.