delivered the opinion of the court.
John. Armstrong made and published his last will and testament on the 85th of March, 1851. He had been twice married, and at the date of his will had seven living children — three by his first and four by his last wife. He devised to each of these seven children a specific portion of his estate, and after making some bequests to his’grand-children and to other persons, he inserted the following residuary clause in his will, viz : “ AH the rest of my estate not hereby disposed of I wish equally divided between my three children living, and my four children by my last wife, except that in addition to the bequeathments made to Amelia, Harriet, and Salty, they be paid one thousand dollars each for the purpose of procuring furniture when they may want it for house-keeping.
On the first day of May, 1851, he made a codicil to his will, and on the 5th day of the ensuing month he added another codicil to it, and departed this life on the 12th of August, 1851. On the 21st of May, 1851, in the lifetime of the testator, and before he added the last codicil to his will, Sally, his youngest child died, being at the time of her death .about five years old. Sally was one of the four children by his last marriage, and the testator’s will contains a do-vise to her of a large amount of real and personal property. The last codicil to the will, which was made after her death, makes no allusion to that event, nor to that part of his estate which the testator had devised to her.
After making the devises already mentioned, to his seven children, the testator inserted the following clause in his will, viz : ‘‘Item 12th. It is my express desire, and so I will it, that the real estate in the foregoing bequeathments be held sacred for the support and comfort of those to whom it is given, during their natural life, and that the same descend to their children *335at thsir demise; but in the event of any of said children departing this life without issue, or such issue dying themselves, then that portion so willed to each person, thus departing this life, shall descend to the full brother or sisters who may survive, to he divided as nearly equal among them as may be; and in the' event of their all departing this life, then the whole estate, real and personal, is to be divided among my first wife’s children.”
The controversy in this case relates to the property devised by the testator to his daughter Sally. On the one side it is contended, that as she died in the lifetime of her father, the devise to her lapsed, and the property, having fallen back into the testator’s estate, passed under the residuary clause in his will to all his children living at the time of his death. On the other side, it is contended that the devise over takes effect notwithstanding her death in the lifetime of the testator, and that the property devised to her belongs exclusively to her full brother and sisters who survived her.
The only other parts of the will that have been relied upon to elucidate the question in issue between the parties, are the following: “Item VUh. Should any reverse of circumstances transpire or occur to lessen the bequeathmenfs made by thi3 instrument, then it is my will that each legatee suffer a diminution in the amount willed to them in proportion to the loss sustained.” “I find on a, survey that my daughter Amelia has a few hundred dollars less bequeathed to her —it is my will that her portion, in money, be made up to the amount given to my other children, as near as may be.”
The argument based upon the last mentioned provisions of the will is, that they show conclusively that the general and prevailing intention of the testator was to produce the most perfect equality among all his children, at the period of his death, and it would be inconsistent with this intention for this devise over to be sustained, and thereby exclude some of his *336children from a participation in the part of Ms estate so devised.
The 13th item of the will, however, is in the follow' ing language : “Whereas, many years past I made a disposition of other portions of my estate to my children by my first wife, -which is on record in the office of the clerk of the county of Mason; and whereas, the sums advanced to my children by my first wife are principally charged to them and my sons-in-law, on my large ledger — it is my will that each of their accounts, as they thereon stand, be in full of any demand on my real and personal estate, except such sums as are in this Instrument provided for.”
It might be regarded as exceedingly doubtful, considering all these provisions in the will in connection with each other, whether the solicitude manifested by the testator, that perfect equality should exist among the legatees, was not confined to his children by his last wife, who were ail young, and none of whom had been previously advanced or otherwise provided for. This inference in regard to his intention is strengthened by the fact that the calculation he made and attached to his will, of the amount of stocks and money he had bequeathed to each of these four children, shows that the amount given to Amelia, who was one of them, was less than that which was given to either of the other three, but at the same time it was much larger than the amount he bequeathed to either of his three children by his first wife. How much the latter had previously received by way of advancement does not appear in this record, but whatever it was, it, together with what was devised to them in the will, constituted all of the testator’s estate which they were to be entitled to.
But if it be conceded that the object of the testator was to produce entire equality among his living children in the distribution of Ms estate, and to continue that equality up to the period of his death, if they all lived until that time, still it is perfectly obvious that *337he did not- intend such equality to be continued, in the-event that either of the younger set of children should afterwards die without children; and, although he probably did not, when he made his will, contemplate the occurrence of such an event in his lifetime, yet if he had, the presumption is strong that he would have made the same disposition of that portion of his estate, 'which he intended for the deceased devisee, that he has actually made in Ms will.
It is apparent, from the twelfth clause in the will, that the testator intended to set apart that portion of his estate which he devised to his four younger children, exclusively for their benefit, and that, upon the death of either of them without children, the part given to such devisee should go to the survivors. Why then should his intention, thus clearly manifested, be presumed to have had reference alone to the happening of this event after his death ? He provided for it as an occurrence that was probable, and whether it happened in his lifetime or after his death, the result would be the same, if the estate of the devisee passed to the other children. The same reason that would influence liim in the one case would equally influence him in the other.
But it is contended, that the making of the codicil after the death of his daughter, without any allusion to that event, especially when the legacies therein bequeathed by him to his grand-children are taken into consideration, shows that the testator regarded the legacy to his deceased daughter as having lapsed, and in consequence thereof greatly increased, by the provisions of the codicil, the legacies that he had previously given to his grand-children.
By the estimate heretofore alluded to, which was annexed by the testator to his will, it appears, that after deducting all the legacies made by him from thp total amount of his personal estate, there still remained a balance thereof, which exceeded the total amount of the additional legacies contained in the codicil. No presumption then can arise that the testator made *338this codicil to his will for the purpose of disposing of that part of the estate which he had given to his daughter Sally, on the supposition that it had lapsed, inasmuch, as independent of it he had a portion of .his estate, which he had not specifically disposed of, more than sufficient to satisfy all these additional legacies.
1. A codicil added to a will is a re-publieation'of the will itself, and the wholo is to be taken together, as if executed at the date of the codicil. (Davis’ heirs v, Taul and wife, 6 Sana, 51.) 2. If there be any person appointed by a will to take in case of the death of the first devisee, and such person eau take in the manner appointed by the will, fhe devise will not lapse although the first devisee die before the testator. (Royer on Leg., vol. 1, page 329; Jar-man on Wills, ml. 2, 476; 2 Ath., 3 2 0; Brown's heirs v. ‘Brown’s dev. 1 Sana, 39.)A codicil is in legal effect a re-publication of a will, and the whole is to be construed together as if the will had been executed at the date of the codicil. When thus construed it is evident that the devise over, after the death of Sally, must be regarded as having been confirmed by the testator. Davis' heirs v. Taul and wife, 8 Dana, 51, But as tbs codicil under consideration was only executed in the presence of one subscribing witness, and for that reason is not a republieation of that part of the will which disposes of the real estate, it can only be relied upon to evince the intention of the testator that the devise over should take effect, and not fail because of his daughter’s death in his lifetime. So far then as the intention of the testator is to control the construction of the will, it will, in ohr opinion, operate decisively in sustaining the validity of the limitation to the surviving brother and sisters of the deceased devisee.
The rule that a devise or legacy lapses by the death of the devisee or legatee-, in the lifetime of the testator, instead of being consistent with his intention most commonly operates to defeat it; but it is a rale of necessity, because where there is no devisee there cannot be a devise. If, however, there be any person appointed by the will to take in case of the death of the first devisee, and the pei’son so appointed can, upon that event, take in the manner contemplated by the will, the devise will not lapse, although the devisee dies before the testator, but the ulterior gift takes eP feet immediately on the testator’s decease, as a direct unconditional gift. Wliere the event upon which the executory limitation is to take place has happened, the occurrence of the contingency, either before or *339after the testator’s death, is regarded as being within the terms of the will, and also according to the intention of the testator. Roper on Legacies, vol. 1, page 329; Jarman on Wills, 2 vol. 426; 2 Atk. 370; Brown's heirs v. Brown's devisees, 1 Dana, 39.
Whether then the intention of the testator is to be learned from a consideration of all the provisions in his will, or is to be arrived at according to well established rules of legal 'construction, in either case the same result ensues. The testator is to be regarded as equally intending the devise over to take effect whether the first devisee dies in his lifetime, or after his death.
Where the death of the person designated to take in the first instance occurs in the lifetime of the testator, but not under the circumstances prescribed in the will, the ulterior gift will not take effect, not however because the first devisee died before the testator, but because the contingency upon which the limitation .over was made to depend did not happen, and the ulterior gift could not have taken effect even, if the death of the first devisee had occurred after that of the testator. This principle is illustrated by the ease of Carpenter v. Heard, &c., 14 Pickering, 449. In that case there was a bequest oí $ 17,000 to the testator’s grand-daughter, Judith C. Lee, to be paid to her when she arrived at the age of twenty-one or was married; which ever might first happen; but if she died under twenty-one years of age and unmarried, then the sum hequethed to her was given over to three other grand-daughters of the testator. The granddaughter, Judith C. Lee, attained the age of twenty-one, and died unmarried in the lifetime of the testator, and the court held that the ulterior gift could not take effect. The reason is obvious. The granddaughter’s death did not occur under the prescribed circumstances; she did not die under the age of twenty-one, and therefore the contingencyupon which the limitation over depended did not happen. ■ .If her death had taken place under the same circumstances, after *340the testator’s decease, the ulterior legatee would not have been entitled to the legacy. If, then, the evénfc upon which an executory limitation depends has actually happened, it is immaterial whether it occurred before or after the testator’s death, provided it be of that description upon which a valid ulterior gift can.. b8 limited.
This brings us to a consideration of the 12th clause of the will, on which the validity of the limitation over depends, the preceding estate never having taken effect. It is evident that the four children by his Iasi wife are alone referred to in this clause by the testator, except in the ulterior limitation to his first wife’s children. The “portion willed to each person,” departing this life in the manner prescribed, “shall descend to the full brother or sisters who may survive; and in the event of their all departing this life, then the whole estate, real and personal, is to be divided among my first wife’s children.” The same persons are clearly alluded to throughout the whole of this clause, and it is only in the event that all the persons thus alluded to shall depart this life that the estate is to be divided among the children of the first wife. The first wife’s children are made the ulterior devisees, the preceding estate is given to the testator’s four last children, and is continued upon the death of either in the manner pointed out in the will, by his or her portion vesting in the survivors, and only fails “in the event of their ail departing this life,” in which case the tilterior gift to the first wife’s children is to take effect. The persons alluded to, therefore, whose portions shall descend to the full brother or sisters who may survive, are clearly the testator’s four children by his last wife. No part of the estate given to either of these children is to go to the first wife’s children, unless they have all departed this life. That such was the object and intention of the testator is conclusively manifested by the ' provisions made by him in this clause of his will.
But who does the testator refer to when he provides that in tha “went of any of said children departing *341this life without issue, or such issue dying themselves, then that portion so willed to each person, thus departing this life, shall descend to the full brother or sisters who may survive.” He had just previously, in the same clause, expressed his desire that the real estate devised should be held sacred for the support and comfort of those to whom it was given, during their natural lives, and that the same should descend to their children at their demise. In the “event of any of said children departing this life without issue,” would, according to its gramatical signification, refer to the children mentioned in the previous sentence, being the grand-children of the testator. But the whole context, as well as the language in which tbe devise over is made, demonstrates clearly and conclusively that such was not the meaning of the testator, and that he was referring to his own children, and not to his grand-children. If any of the children referred to departed this life without issue, the portion willed to such person was to descend to the full brother or sisters who might survive. His own children had a full brother and sisters, and the devise over was, when applied to them, appropriate and intelligible. But whether it ever could have any application to the children of his four last children he could not forsee; nor can it be supposed, when he gives a fixed and certain destination to the portions of any of the persons departing this life without issue, to the full brother or sisters of such person as might survive, that he intended to provide for a state of case so extremely remote, uncertain, and improbable, as the existence of full and half-brothers among grand-children then unborn. He must have meant his own children, because the terms of the devise applies peculiarly to them, and might never have any application to any set of grand-ebildren who might be subsequently born. The testator had in his mind his four children by his last wife; they were the peculiar objects of his solicitude and care; it was for them, and not for their children, that he was providing; there can be no doubt, *342that they were the persons referred to by him in the words of his' will under consideration.
3. The testator had two sets of children — by the first wife three, by a second four — after providing for the children of each, lie has this provision in his will: “It is my express desire, and so I will it, that the real estate in the foregoing b equeathments bo held sacred for the support and comfort of those to whom it is given, during their natural life, and that the same descend to their children at their demise; but in the event of any of said children depicting this life, without issue, or such issuedying themselves, then til at portion so willed to each person, so departing this life, shall descend to the full brother or slaters who may survive, to be divided as nearly equal among them as may be, and in the event of them all departing this life, then the whole estate, real and personal, is to be divided among my first wife’s children.” One of the children of the last marrlage died an infant before the testator.— Held, that the portion of such child, embracing all real and personal estate given by the ■will, did not lapse, but vested in the full brothers and sisters who survived. (Jar-man on Wills, vol. 2, chap. 5U, page 475.)*342The testator in this clause of his will provides, first, that the real estate given to his children shall be held sacred for their support, during their natural life, and that it shall descend to their children; In the next place, that if any of his children by his last wife depart this life without issue, then, that portion of bio estate which he has willed to such person, thus departing this life, shall go to the full brother or sisters who may survive. The portion thus devised over, is not merely the real estate which had been given to the child dying without issue, but the whole portion, including real and personal estate, none of which was to be divided among the first wife’s children so long ao any of the others survived. That such was the intention of the testator is placed beyond doubt by the disposition which he made of the portion given to each of these children, and by the terms of the ulterior gift, to his children by his first wife. This ulterior gift embraced all the estate, both real and personal, which had been given to his last wife’s children, but was to take effect only in the event of the last named children “all departing this life.” The language of the testator is not reasonably susceptible of any other construction. Whether bis intention can be carried into effect, to its full extent, according to the settled rules of law upon the subject, is a question not now presented for our determination.
The contingency upon which, the limitation over was to take effect, was the death of any of said children without issue, or such issue dying themselves, if there should be any! If either of the children died without issue, the first contingency upon which the limitation over depended would occur, and if the death happened in the testator’s lifetime, the executory devise would take effect at the time of his death. It is obvious that this must be the consequence of the death of one of the children in the present case, if the intention of the testator is to be consulted. His daugh*343ter Sally died in his lifetime, without issue. Upon the happening of that event, the estate devised to her was to go over to her full brother and sisters; and having made that disposition of her portion of the estate, in the event of her death without issue, in general terms, applying as well to her death in his lifetime as subsequently, he did not deem it necessary after her death, under the prescribed circumstances, to make any addition to his will, for 'the purpose of carrying his intention into effect. But it is contended, that the event upon which the limitation over depended, was not only the death of Sally, without issue, but also the death of such issue. This additional contingency was annexed, in case there should be issue, but as there was no issue, the limitation over depended alone upon the first contingency, to-wit: the death of his daughter Sally without issue, and that contingency having occurred in the lifetime of the testator, the devise to the surviving brother and sisters of the deceased devisee took effect on the death oí the testator, not as an ex-ecutory limitation, but as asimple absolute gift. The cases referred to by Jarman on Wills, 2d vol., page 475, chapter 50, seem fully to establish the position, that a gift over, to take effect in the event of the prior devisee dying under certain circumstances, applies to the event happening in the lifetime of the testator as well as subsequently thereto, unless a different construction is required by the language of the will.
' 4. The words “dying without issue'’ are novi construed as importing a failure of issue at the time of the death of thefirst devisee, andrcoi an indefinite failure of issue.*343It is argued, that the limitation in question is invalid because it was not so created, that it must necessarily have taken effect, if at all, within the period prescribed by the rule against perpetuties, consequently, if the;first devisee had not died in the lifetime of the testator, that she would have taken an absolute fee, and the limitation would have been defeated; and that her death in the lifetime of the testator cannot make valid a limitation that was void in its creation, by reason of its remoteness.
What wms the contingency upon which this limitation was to have its commencement? It was the *344death of one of the children without issue, or such issue dying themselves. The words “dying without issue,” are now construed as importing a failure of issue at the time of the death of the first devisee, and not an indefinite failure. But even under the rule by which they were construed to mean a general or indefinite failure at any time however remote, the annexed words, by which the ulterior limitation was made to depend, not only on the failure of the issue of the first taker, but also on the decease of such issue, would have been regarded as sufficiently indicating that the testator used the expression “dying without issue” in a limited and restricted sense. Lewis on Perpetuities, 52 vol. Law Library, side page, 243. And this construction would have been fortified by the gift over haying been expressly made to the surviving brother and sisters; but this is not material under tbe operation of the rule by which the expression “dying without issue,” is construed to mean a failure of issue at the time of the death of the devisee. If the limitation had depended upon the first named contingency alone, that is, the death of the first devisee without issue, there could have been no doubt of its validity. But inasmuch as should the first devisee have issue, that issue was also to die before the commencement of the limitation, this was an event which, as it is argued, might not happen within the period prescribed by the rale, and made the limitation void for remoteness.
5. A ‘ devise to several, and “if one die without issue, or- such, issue dying themselves,then that portion so willed to such person thus departing this life shall descend to the full brother or sisters who may survive, to he divided,” isa., is a valid devise, and not a perpetuity. {Attorney Gen. 8. Wallace’s dev., 7 B. Monroe, GI5.)*344What did the testator mean by the words “or such issue dying themselves ?” lie has applied terms of contingency to an event which was inevitable, and certain to happen at some time. He knew that such issue would die, and consequently he could not have intended to refer to that event, whenever it might occur, as it was certain and did not depend on any con - tingency. It is necessary then, in order to ascertain his meaning, to connect with death some circumstance, in association with which it is contingent. That circumstance naturally is the time of its hap*345pening', and such time must have been the death of the parent of 3uch issue, there being no other period to which .the words can be referred. The devise would, if this be the true interpretation of the language used, be that if the first devisee should die without first having had issue, or having had issue, that issue should die before its parent, then the limitation was to take effect. It is true that the expression “dying without issue” was of itself sufficient to indicate that, if there was no issue alive at the death of the first devisee, the contingency had happened upon which the commencement of the limitation depended, yet the testator may have supposed that dying without issue might be construed to mean dying without having had issue, and to make his meaning definite and certain, added such words as would clearly express the idea, that if the devisee had issue, yet if that issue died, and the devisee had no issue living at the time of his or her death, the limitation was then to take effect. This construction of the words in question would seem to bo sanctioned by the case of Attorney General v. Wallace's devisees, 7 B. Monroe, 615. Thus construed, the ccntingéney upon which the limitation was to take effect must have happened upon the termination of a life in being, and the gift over was a good executory devise, according to the well settled doctrine on the subject.
But the circumstance in association with which the death of the issue is made contingent, may be the death of the issue, itself, without issue, which is the only other contingency that can connect itself with the death referred to in tho context. If this be the proper construction of this expression, then the death of the issue, without issue, whenever that event happened, was the second contingency upon which the limitation -was to take effect. As this contingency might not happen within the time prescribed by the rule against perpetuities, the devise over, so far as it *346defended upon it, was clearly void for remoteJ18SS.
6. When a limitation in a devise is made to take effect oa two alternative contingencies, one of which is too remote, and ilia other valid, altnough it is void so far as it depends upon the remoto event, it -will be allowed to take effect upon the oth'>r event. (Lenin on Perpetuities, side page, 501; Jar-en on on Wills, side page, 244; 7 B.Müñrqe, 611.)It does not follow, however, even conceding that this was the meaning of the testator, that a limitation of this description cannot be valid. Jet subsequent events occur as they may. As a general rule it is no doubt true, that a limitation bad in its inception', will not be made valid by any events happening subsequently to the time of its creation. Whether this rule ought to prevail, when the limitation is created-by will, imcmuch as the death of the testator, in the case of wills, is the time at which the period cf remoteness is to be computed, as well as the period for ascertaining the objects of his bounty, we will net stop to inquire, for this rule, admitting that it should apply to limitations created by will, is still, like almost all other general rales, subject to some exceptions.
Where a limitation is made to take effect on two alternative events or contingencies with a double aspect, one of which is too remote, and the other valid, as being within the prescribed limits, although it is void so far as it depends upon the remote event, it will be allowed to take effect on the alternative one. Lewis on Perpetuities, supra, side page 562, Jarman on Wills, 1 vol. side page 244; Attorney General v. Wallace's devisees, 7 B. Monroe, 611. Such a limitation is made an exception to the general rule, that every limitation must, in its original form, be such that it will necessarily vest, if at all, within the legal boundaries of remoteness. In the creation of a limitation of this kind, two events are contemplated, distinct from each other, on ci'her of which the gift ig intended to take effect, and the remoteness of one of which, therefore, does not affect the validity of the other, in case it should be actually realized; and it so far differs from all other limitations, in the construction to bo put npqn it, in reference to the laws of remoteness, that its validity depends entirely on subsequent events. The rule allowing such an exception seems to be founded *347upon the consideration, that it is absolutely certain, at the time of the creation of the limitation to which it applies, that if it take effect at all it must take effect within the prescribed limits, and is consequently exempt from all objection on the ground of remoteness.
This exception to the general rule upon the subject is exemplified by Jarman in Ids Treaties an Wills, vol. 1, side page 24.4, by the following example: “As, in the case of a limitation to A and his heirs, and if he shall die without leaving any issue at his death, or leaving such, they shall die under the ago of twenty-three years, then to B in fee. Hero there are two distinct events, on either of which the executory devise Is to arise, namely, A’s leaving no issue at his death, and his leaving issue, who die under twcnty-lhi ae. If the first event happen, i e, if A die without leaving issue at his death, the executory limitation is cío ally good, as it would be in the ease of a simple devise over on this event; but if the second event happen, i e, if A leave issue, though they die not only under twenty-three, but under twenty-one, or any other age, actually within the prescribed limit, it would be bad.” And he cites several cases that sustain and illustrate this species of alternative limitations.
Now, in the present ease, if the true construction of tho devise be, that upen the death, of the first taker without issue, or if there should be issue, upon the death of such issue without issue, then over to the surviving brother and sisters, the gift over is evidently to arise on an alternativo event, one branch of which is within, and the other is not within the prescribed limits. And as the first event, that is, the death of the first taker without issue, has actually happened, and was such as must necessarily have happened, if at ail, within the prescribed limits, the gift ver is clearly valid. If, however, iho first taker had died leaving issue, inasmuch as the second event, that is, the death of such issue without issue, ivas one that might not have happened within the prescribed iim*348its, the gift over would have been bad, and would not have taken effect.
H. Taylor for plaintiff; Morekead & Brown, Crittenden and I'Iord for defendants.It thus appears, that according to the events that have actually happened, the executory devise in this, case is valid, and the full brother and sisters of the deceased devisee arc entitled to the portion of the estate, both real and personal, which was given to her by the testator.
Wherefore, as a different construction was given to the will by the court below, the judgment is reversed, and canse remanded that a judgment may be rendered in conformity with the principles of thio opinion.