This suit involves primarily the legal interpretation of the will of John Lander, deceased, a resident of Phillipsburg, New Jersey.
At the date of the will there were living testator's wife and his two daughters (Elizabeth Petrie and Mary Stewart), and also six grandchildren, children of his daughter Mary. His daughter Elizabeth had no children. To state its general scheme and summarize its contents briefly, it gives his entire estate first to his wife for life; then to a trustee in trust for his two daughters during their lives; finally to his grandchildren.
The language of the gift to the grandchildren is as follows:
"Fourth. After the decease of my wife, and my daughters, Elizabeth and Mary, I give devise and bequeath all my said estate then remaining — real personal and mixed unto my dear Grandchildren, children of my daughter Mary L. Stewart, and of my daughter Elizabeth L. Petrie if she have any living at the time of her decease and to their heirs, share and shair alike; and I hereby direct the said Jacob S. Stewart, to execute and deliver all conveyances or assignments that may be necessary to cary out the provisions of this fourth section of my will."
Testator's wife and two daughters survived him. The wife's life estate ended in 1901; the daughter Elizabeth died *Page 569 — still childless — in 1928; the other daughter, Mary, died in 1930. Of the latter's six children, one predeceased testator; of the five who survived him, two predeceased their mother (and also predeceased Elizabeth), and the other three survived their mother.
Let me interpolate here to say that the fourth paragraph of the will, as quoted above, is exactly as it appears in the original will. An examination of the latter, in the office of the registrar of the prerogative court, discloses an amazing number of errors in the certified copy which was offered in evidence. There are no less than eighteen errors in the copy of this one paragraph. None of them, as it happens, would lead to a different interpretation in this case, although a number of them are of a kind which might in some cases have been disastrous — such as changes in spelling, omission of words, omissions and insertions of punctuation, and omission of errors and corrections made by testator appearing on the face of the will (which might be of great significance and importance in discovering what had actually been in a testator's mind). The advisability of the use of photostatic copies in evidence, in such cases as the present, is clearly indicated. One brief in this case bases an argument on a word used by testator which was entirely omitted from the certified copy.
Two questions must be determined: (a) Is the gift of the remainder interest a present gift or a future gift? and (b) To whom is the gift made? — is it made only to testator's grandchildren, or substitutionally to the grandchildren or their "heirs?" The two questions are interwoven and interdependent; the answer to question "A" depends largely upon what is the correct answer to question "B", and vice versa.
Reading the fourth paragraph by itself; giving to its phraseology the natural, ordinary, grammatical construction; giving to its non-technical words their usual and ordinary meanings, and giving the definitely established technical meaning to such technical words and phrases as have such definitely established meaning — it would seem at first impression that the gift of the remainder is a gift in futuro — *Page 570 operating immediately "after" the death of the survivor of the life tenants, and that it is made only to testator's grandchildren (or certain of them). Further consideration, however, especially in connection with other provisions of the will and of testator's surrounding circumstances and natural impulses (all of which are required to be taken into consideration: Torrey v. Torrey, 70 N.J. Law 672;59 Atl. Rep. 450; Coyle v. Donaldson, 91 N.J. Eq. 138; 108 Atl. Rep. 308;Garrabrant v. Callaway, 113 N.J. Eq. 424 — at p. 428) — brings substantial doubt on both points.
Taking up question "A" for first consideration: ordinarily a gift "at" or "when" or "after" a particular event occurs, means that the gift is not to take effect until that event occurs, and the gift does not vest until then. Howell, Exr., v. Green,,31 N.J. Law 570 (at p. 572). However it is also determined in that case (and repeatedly thereafter) that where it appears from the entire will that the only reason for postponement of the future gift is to let in some other interest, the gift is deemed presently vested — the ownership passes at once though the time of enjoyment is postponed. Especially is it settled law that where there is a gift of a life estate and a gift of the remainder interest after the termination of the life estate, the latter will be deemed vested immediately (if there is nothing to indicate a different intent). Redmond v. Gummere, 94 N.J. Eq. 216 (at p. 217); 119 Atl. Rep. 631. So, if in the instant case the gift of the remainder interest had been made to one or more definite and presently existing beneficiaries — e.g., to six specifically named persons, or "to my grandchildren living at my death") — there would be no doubt but that the gift vested immediately.
On the other hand if the will clearly indicated that the remainder was intended not to vest until a later time, it would not vest at testator's death. Such for instance would be the case if the gift were "to such persons as shall be named by both of my daughters in their respective last wills;" obviously, the remainder could not vest in the designated class until termination of the life estate in the two daughters, because the identity of the persons who would compose that *Page 571 class could not be ascertained prior to the death of the survivor of the two daughters. Such also would be the case if the gift were "to John Doe, if and when he survives both daughters;" the gift to him would be clearly conditioned upon the contingency of his surviving the two daughters — not to become operativeunless that event happens, nor until that event happens (if it does happen).
There are "in between" cases, however, where the testator's intent is not so clear. Suppose the gift of remainder, after the life estate in the daughters, is "to my grandchildren then living" — does the remainder vest at once in the grandchildren living at the death of testator subject to divesting in the case of any grandchild who dies during the life estate, or is it contingent and not vested in any of them until the termination of the life estate? By definite and specific expression in that behalf the testator may of course accomplish whichever of those two results he desires; but in the absence of any such definite and specific expression how can one be sure of which result he intended? He may have intended the one or he may have intended the other — or he may have given no thought to the point at all, and hence in fact had no intent one way or the other. If litigation arises involving conflict between two sets of claimants, one claiming under one theory and the other set claiming under the other theory, the point necessarily must be determined. In such cases the only possible guide or indication is the usual meaning of the words actually used, and theprobability that the testator intended the result usually accorded to the use of such words; and an artificial rule has had to be adopted that since the gift is in words usually indicating a gift in future and to persons who cannot be definitely known until the happening of the future event, no present gift is intended and no vesting occurs until the happening of the future event. Gifford, Admr., v. Thorn, 9 N.J. Eq. 702 (at p.732); Post v. Herbert's Exrs., 27 N.J. Eq. 540 (at p.543). See, also, Voorhies v. Otterson, 66 N.J. Eq. 172;57 Atl. Rep. 428; Teets v. Weise, 47 N.J. Law 154.
Again, suppose the gift of remainder after the life estate *Page 572 in the daughters, is simply "to my grandchildren" — without saying "now living" or "then living;" which is meant, a present gift of the future remainder to the grandchildren already living, or a future gift to those who shall then be living? Here again an artificial rule has of necessity had to be adopted (operative if there is nothing to indicate a different intent), based upon the probability of what testator would have definitely specified if his attention had been called to the fact that his actual expression was ambiguous. The gift being of the remainder interest after the life estate and there being grandchildren already living, the gift is held to vest at testator's death in the grandchildren then living (Cranstoun v. Westendorf,91 N.J. Eq. 34; 108 Atl. Rep. 776; Trenton Trust Co. v. Sibbits,62 N.J. Eq. 131; 49 Atl. Rep. 530), but with the further provision that if any other grandchild be born during the life estate it will immediately become vested of a share likewise.Heater v. Van Auken, 14 N.J. Eq. 159; Woolston v. Beck,34 N.J. Eq. 74; Budd v. Haines, 52 N.J. Eq. 488; 29 Atl. Rep. 170;Clark v. Morehous, 74 N.J. Eq. 658; 70 Atl. Rep. 307, and authorities therein cited.
Compare, also, Moore v. Dorey, 87 N.J. Eq. 325;100 Atl. Rep. 207 (where, however, the question as to vesting was not involved).
The rules as laid down by the authorities are then — first, the well established, primary, general rule that a gift in future language, without more, means a gift to take effect in future; that where a will says "I give * * * to John Doe after" a future event (and there is nothing to indicate a different intention), the donee takes nothing until the happening of that event, and not then unless he be then living. To this primary rule there is inter alia an equally well-established exception — if the subject of the gift is a remainder interest and the future event is the termination of a prior life estate, and the donee of the remainder is a person or a class of persons living at testator's death (and there is nothing to indicate a different intention), the donee will take a vested interest at testator's death (enjoyment only being *Page 573 deferred); this on the theory that the natural presumption would be that testator intended (no reason the contrary appearing) to make a present gift of the entire interest in the estate — the life estate to one and the remainder (subject to the prior enjoyment by the life tenant) to the remainderman or remaindermen. In re Buzby, 94 N.J. Eq. 151 (at p. 153);118 Atl. Rep. 835. But if the donee of the remainder interest is a class whose members are not to be ascertained until the future event, then the case does not come within the exception, and the exception does not operate; the general primary rule operates and the gift does not vest on testator's death.
In the instant case the gift is of the remainder interest, after a life estate in the two daughters and the survivor, and is to testator's grandchildren — of whom there were six then in existence. Upon these facts, without more — if there were nothing else in the will and surrounding circumstances — there can be no question but that the gift should be interpreted as vesting at testator's death in the grandchildren then living, subject to be opened in favor of each grandchild (if any) subsequently born during the life estate. The situation would be similar to that inHeater v. Van Auken, supra. There are, however, a number of other things in the instant case which have an apparent or possible bearing upon the question of testator's intent, and which must therefore be carefully considered.
The first of these is the fact that the gift is made not simply to testator's grandchildren, but to "my dear grandchildren, children of my daughter Mary L. Stewart, and of my daughter Elizabeth Petrie if she have any living at the time of her decease and to their heirs, share and share alike," which brings up question B: To whom is the gift made? The interrelationship and interdependence of the two questions is readily apparent — for if it were certain that testator intended a class determinable only in future as the donees, that would tend to indicate an intention that the gift should not vest immediately (pursuant to the rules already discussed); and on the other hand, if it were certain whether *Page 574 testator intended an immediate vesting or a future vesting, light would thereby be cast on testator's intent as to just what persons should take as beneficiaries.
From a reading of the clause above quoted in accordance with the natural and ordinary meaning and construction of its language and phrasing, it would seem reasonably clear that the gift is to the "Grandchildren * * * and to their heirs, share and share alike;" and that the intervening words — "children of my daughter Mary * * * and of my daughter Elizabeth * * * if she have any living at * * * her decease," are parenthetical, so to speak, and explanatory of just what persons he intended as included in the class "grandchildren;" expressing his intent that the gift is not to go only to his then present grandchildren to the exclusion of all future grandchildren if any there should be (as might otherwise be inferred from his use of the adjective "dear" — an unusual word to apply to grandchildren not in existence).
If this be so, and if to the phrase "and to their heirs" is accorded (pursuant to the usual rule) its well established legal meaning, then it is clear that the gift is to none other than grandchildren. The gift being to grandchildren "and their heirs," the word "heirs" is a word of purchase and not of substitution. Kutschinski v. Bourginyon, 102 N.J. Eq. 89;139 Atl. Rep. 596, and cases cited.
It is contended, however, that the testator intended the word "heirs" as substitutionary, and that the word "and" should be read as meaning "or;" and there is much to support such a contention. The word "and" has so frequently been construed as "or," and vice versa, where it is evident that such was testator's actual intent, as to need no citation of cases. This will was drawn by testator himself (in his own handwriting), and while it is largely made up of legal phraseology, it is nevertheless evident from the whole that testator was a layman, lacking in the knowledge of an experienced draftsman, having but scanty education and of no high intelligence; it seems probable that the will was a "revamping" by the testator of an earlier will written by an *Page 575 experienced draftsman. His spelling is woefully deficient, likewise his punctuation, and he is obviously careless. In several instances he mis-spells words which he spells correctly in other places; he uses commas where they do not belong and omits them where they should be used; and in at least two instances he has indisputably made unintentional omissions of words and clauses. (At the end of the first half of item third of the will, some such clause as "shall be paid from Elizabeth's share" or "shall be repaid by Elizabeth" must be supplied to complete the obviously unfinished sentence; and in the sixth paragraph, where he speaks of "my said children, Elizabeth or the husbands of either of them," clearly he has accidentally omitted Mary's name).
It is quite possible therefore that although expressing the gift of remainder to his grandchildren "and to their heirs" he may have meant to give it to his grandchildren living at the termination of the life estate and the heirs of grandchildren theretofore deceased — to each grandchild if it survived the life estate or to its heirs if it did not. Cf. Shannon v. Shannon,101 N.J. Eq. 816; 139 Atl. Rep. 173. Such a phrase may be interpreted as substitutionary, if such an intention is to be gathered from the entire will and the surrounding circumstances.Zabriskie v. Huyler, 62 N.J. Eq. 697; 51 Atl. Rep. 197;affirmed, 64 N.J. Eq. 794; 56 Atl. Rep. 1133; McKiernan v.Beardslee, 72 N.J. Eq. 283; 73 Atl. Rep. 815. These same authorities, however, equally establish it as the law, that a mere possibility is insufficient, even if it be of such character as to be deemed an equal possibility. There must be something found elsewhere in the will indicating at least a probability of such intention, in order to warrant giving to the clause a meaning different from its usual and ordinary meaning. The possibility of a lapse and intestacy (even if that possibility has ultimately come to be an actual fact) is not alone sufficient to warrant a departure from the usual meaning.
In the instant case it is impossible to find anything else in the will tending to support the conclusion that testator intended the clause as substituting beneficiaries instead of *Page 576 being simply descriptive of the nature or extent of the estate given; and it must be concluded, therefore, that this will expresses no gift, contingent or otherwise to any "heirs" of any of testator's grandchildren — nor to any one other than testator's grandchildren. It still remains to be determined however to which of the grandchildren the gift is made (as well as whether the gift vested in any of them prior to the termination of the life estate).
The testator has defined his meaning of the word "grandchildren" as being "the children of my daughter Mary * * * and of my daughter Eliabeth * * * if she have any living at * * * her decease." (Italics mine.) The italicized phrase may be construed as referring only to the daughter Elizabeth; it may also be viewed as quite reasonably possible (especially under the circumstances hereinbefore mentioned as to testator's lack both of testamentary skill and common education and the evidences of carelessness in the language of the will), to interpret it as referring to each of the daughters; and it must also be deemed possible, for the same reasons, that the testator may have accidentally omitted a similar phrase after Mary's name (just as he accidentally omitted words and clauses in the third and sixth paragraphs, as hereinbefore pointed out): if such omission be supposed and be corrected, the reading would lead to the same result as the second interpretation.
In favor of the second or third interpretation (which are in practical effect the same) is the strong natural probability that his impulse and intent would be not to differentiate between the children of one daughter and those of the other, but to treat them both alike, and the fact that if the clause be read according to the first interpretation he does not treat them both alike but differentiates between them and apparently, or possibly, discriminates in favor of Mary's children as against Elizabeth's.
This discrimination would be quite marked and real if it should be deemed that Mary's children took a vested interest at testator's death because testator's language very definitely makes the vesting of any interest in any of Elizabeth's children *Page 577 deferred until, and contingent upon, Elizabeth's death leaving them surviving her. The fact that Mary already had six children whom testator knew and for whom he evidently had affection (evidenced by his words, "dear grandchildren;" and in item second of the will he also refers to "my dear grandchildren, children of my daughter Mary"), whereas Elizabeth's children were not an actuality but a mere non-existent, theoretical, future possibility, might doubtless be deemed a sufficiently rational reason for differentiating between Mary's then living children and Elizabeth's possible future children, but there would seem no rational or logical reason for differentiating between the possible future children of Elizabeth and the possible future children that might be born thereafter to Mary.
Such discrimination and differentiation, however, practically if not entirely, disappears (as will be shown later), if the remainder is deemed not to vest at testator's death, in any of the grandchildren; and the argument would in such case be of little if any weight. In any event it cannot be deemed controlling. It is possible that he may have had reason to believe that Mary would have no more children (she did not, as a matter of fact) — and hence was not differentiating between Mary's possible future children and Elizabeth's possible future children, but only between Mary's actual, known and loved children and Elizabeth's theoretically possible children — a differentiation reasonably understandable. Much more important still is the fact that in item second of the will he makes a clear and unmistakable differentiation between Mary's children and Elizabeth's children (where he provides that if Elizabeth survives Mary, Mary's children shall take the share of income theretofore enjoyed by Mary, but that if Mary survives Elizabeth, Mary shall take the whole of the income — although he necessarily had then in mind the possibility that Elizabeth might then have children, for he mentions it in the very next sentence). This, of course, completely negatives the argument that he could not have intended to differentiate between the two sets of children. *Page 578
It is quite evident and certain therefore that the first interpretation, which is the natural and ordinary (and thereforeprima facie probable), interpretation of the clause, must be accorded judicial sanction. There is nothing to indicate any other intention as more probable; nothing against it but mere possibilities. The gift is to Mary's children and to Elizabeth's children if Elizabeth have any children living at her decease.
But, by Mary's children, did testator mean those living at his death, or those living at the termination of the life estate?
Clearly Elizabeth's possible children were not intended to take any vested interest at testator's death, even if any were then in existence, because their interest is expressly made contingent upon their surviving Elizabeth; but can it, and should it, be deemed that Mary's children took any vested interests on testator's death?
As has been seen — if the gift were only to "Mary's children," then under the established rules (and the circumstances of this case) it would (in the absence of indications of a different intention) vest at testator's death in those children of Mary living at that time, subject to being opened to let in any after-born children; and if the gift had been only to "the children of Elizabeth living at the end of the life estate," it would be contingent and not vest until the occurrence of that event — in the absence of evidences of a different intention. It is possible that testator may have had in mind that Mary's living children should take an immediate vested interest; that he may have known that under the ordinary rule in such case any children born to Mary during the life estate would immediately take vested interests, and that he had in mind that the possible future children of Elizabeth could and should if they survived Elizabeth, likewise thus in future enter into and share with Mary's children the vested ownership of the remainder; and if that were testator's intent there would be no difficulty in according effect to it.
Here again, however, is only a possibility — not a probability; *Page 579 the probability rather is to the contrary. The gift is not made in words naturally importing an immediate gift, nor to "Mary's present children" with a provision for future contingent interest to future possible children of Mary and Elizabeth. As has already been mentioned, the natural and probable impulse and intent of testator would be not to make any material differentiation between Elizabeth's children and Mary's children (especially since in his provisions for Elizabeth and Mary themselves he has made no material difference); so that, other things being equal, a construction which places the children of both daughters upon substantially the same footing, is to be favored while one which materially differentiates between them is improbable.
The gift as actually expressed in the will is a gift not (either separately or jointly) to "Mary's children" and "Elizabeth's children," but to a single class, testator's "Grandchildren." This is important to be kept in mind, for while the technical wording of the gift will give way to a different intent by testator if such be found from other parts of the will, still it naturally must and will control if no such different intent can be said to be at least probable. The membership of this single class is explained and defined by testator as being made up of Mary's children (or certain of them) and certain of Elizabeth's (possible) children, nevertheless the gift is to the single class, grandchildren. (It is doubtful that the capital "G" has any real significance; but if it does it emphasizes this point.) The membership of that class, as a whole class, cannot be ascertained (under testator's specifications), at his death — nor until the death of the life tenants (or at least one of them, Elizabeth). Therefore, under the established rule earlier mentioned herein (applicable in the absence of satisfactory evidence of a different intent), the gift of the remainder would not vest at testator's death but would be contingent until the happening of the future event (the fee being meanwhile in the trustee).
There is (as has been shown) nothing to be found in this will from which it can be said that it is probable that testator intended something different from that which he actually *Page 580 said, in making the gift to the single class, "grandchildren;" nor from which it can be said that it is probable that testator intended the interest to vest immediately upon his death. The application of the usual rules therefore leads to the conclusion that the remainder was contingent and not vested — that it did not vest until the death of the two daughters or at least until the death of Elizabeth. (Under the circumstances of this case it makes no difference upon which of these dates the vesting occurred, because, while Elizabeth predeceased Mary by two years, she died without ever having any children and those children of Mary who predeceased Mary also predeceased Elizabeth.)
This conclusion is in fact, in the view of this court, even more clearly indicated, or at least corroborated, by certain probabilities naturally deducible from some expressions in the will, and the circumstances.
As already noted, the words of gift naturally and ordinarily import a gift not at testator's death, but in future — "After the death of Mary and Elizabeth I give." So, likewise, in addition, do the words descriptive of the thing given — "all my said estate then remaining." These words evince an intent that the life tenants should have not merely the life enjoyment of the estate, but also the power of disposal (Weaver v. Patterson,92 N.J. Eq. 170; 111 Atl. Rep. 506), and hence an intent to give to the remaindermen only that which should remain undisposed of at the death of the life tenants. Finally, so does his designation of the donees. His words express a gift at a future time, of a thing which can be ascertained only at that future time, to a class of persons, some at least of which can be ascertained only at a future time (probably at that same future time). From all this it would seem that there can be little, if any, doubt, but that this unlettered, untutored man — unfamiliar with the legal words and phrases he (unwisely) assumed to employ himself, and unacquainted with the legal results to be anticipated from the use thereof — had (and was expressing) the conception that this gift was to take effect at that future time, and that he had only that conception. There is nothing *Page 581 from which it appears that he had any knowledge or realization of a gift passing the right of ownership in the present while deferring the right of enjoyment until a future time.
It is true that in Naundorf v. Schumann, 41 N.J. Eq. 14;2 Atl. Rep. 609, where there was a gift to the widow and "in the event of her death" "what shall remain" was to be divided equally among named brothers and sisters, it was held that the widow took a life estate without any power of disposal of corpus, and that the brothers and sisters took a remainder which vested at testator's death. That determination was, however, shortly thereafter overruled in the case of Rodenfels v. Schumann,45 N.J. Eq. 383; 17 Atl. Rep. 688 (although the precise point here at issue was not passed upon. See, however, Weaver v.Patterson, supra). In any event, it would not be dispositive of the instant case, because in this case the remainder is not given to named existing beneficiaries, but to a class of which at least some members could not be definitely known at testator's death (and in the Naundorf Case there is neither determination nor expression as to the effect which words actually indicating a power of disposal would have on the question of the vesting of the remainder interest).
Moreover, it would seem clear that the words "my estate then remaining" could be deemed not sufficient to indicate an intent to give to the life tenant a power of disposal, without greatly impairing their effect as indicating an intent that the remainder be given at the future time instead of at testator's death.
In further corroboration there is the testator's expression in the fifth item of the will — where he gives the trustee power of sale during the life estate and directs that the proceeds be kept invested "until the expiration of the trust hereinbefore created" (the life estate) "when it shall pass as I have above directed;" and again in the sixth item — where he provides that if any creditor of Elizabeth attempts to take Elizabeth's income, the trust for her shall be immediately revoked and "her shareshall then pass as directed *Page 582 in case of her decease." This language is of course not conclusive — it could refer to a passing of right of possession or enjoyment, but as used by this unlearned man the probability is that he meant the passing of ownership. His direction to the trustee to execute conveyances to carry the gift into effect is another similar indication.
Still further, and perhaps of even greater importance, is the fact that the interpretation against an immediate vesting practically eliminates the discrimination between the children of the two daughters, which otherwise would exist. Testator's intent being that ownership would pass to the grandchildren at the death of the two daughters, his conception would correspondingly have been that it would go to those of Mary's children who should then still be living. The addition of the clause "if she have any living at the time of her decease," after the mention of Elizabeth's children, may very well be — indeed probably is — explained by the fact that he was speaking of children who were to him at that time not an actuality but a mere possibility.
The mental processes of this testator, in the determination, and formulation of his testamentary plan, under the circumstances surrounding him at the time, would naturally have been — and therefore probably were — to the following effect: "I want my estate to go first to my wife as long as she lives. I do not need to concern myself about providing for our daughters during her life, for she will naturally want to help them if necessary. After my wife is dead I want it to go to my daughters for their lives. They are both women and not so capable of taking care of property safely or wisely — I'll give it to a trustee to pay them the income as long as they live. Then after they are both dead, I'll give what then remains to Mary's children; some of them will still be living, if not all. Elizabeth has no children now, but she may possibly have some before she dies. If she does have any living when she dies, I ought to include them too. They will be my grandchildren just as much as Mary's children, and all grandchildren ought to share alike."
This does not account for the lack of mention of possible *Page 583 future children of Mary; nor for the fact that the contingency attached to the gift to Elizabeth's possible children is their surviving Elizabeth (instead of surviving the termination of the trust at the death of both daughters). As to the former, it is quite possible that he may have been informed that Mary's future children if any would be included by law — or that he may have felt certain that she would not have any future children (she in fact did not have any). As to the latter, it is probable (considering his mental equipment and characteristics) that he simply made a careless slip which he failed to observe or realize.
It is concluded therefore that the remainder interest did not vest until the death of Mary, the survivor of testator's two daughters; that it then vested in the three grandchildren (children of Mary), then surviving, namely Grace Bergstrom, Beulah Sadler and John Lander Stewart; a conclusion based both on the application of those established rules of construction pertinent to the language of this testamentary gift, and as well also on what appears from the whole will probably to have been testator's intent.
In view of this result, it becomes unnecessary to express a determination of the questions raised as to the transmission of interests from those of Mary's children who predeceased the termination of the life estate.