I consider the decision of this case made by a majority of my associates plainly erroneous; and, as briefly as I am able, I will proceed to state the grounds on which my dissent is placed. On the part of the plaintiffs it is claimed that by a provision in the fourth clause or section of the will of Thomas Mackaness, the premises in question, in case his daughter Margaret should die leaving children, were devised directly to such children; in other words,-that her children were to take as purchasers, and not by descent from their mother. One of the plaintiffs is the daughter and only child of Margaret, and -they claim solely under this alleged devise. In order to maintain their title, they must show either, (1) That the devise of the same premises to the testator’s daughter Margaret' was of a mere life estate; or (2) If that devise was in fee, then that it was a fee of such a peculiar character that it was not descendible to her children, if she should, leave any; in other words, that the same fourth clause *358of the will gave one fee to Margaret, and then another by way of'executory limitation to her children. Unless Margaret took, in the primary devise to her, simply a life estate, then the estate of inheritance given to her must in some way be reconciled with the estate of inheritance supposed to be given to her children. Looking only at the particular provision in which those children are mentioned, it requires but the slightest possible mental effort to say that the testator intended to give the estate to them on the death of their mother. But this conclusion is manifestly of no value until we first examine and- determine what estate or interest was devised to Margaret herself. That point being settled, if there is then room to let in the supposed fee of her children, the plaintiffs’ title can stand; otherwise it cannot.
The very learned and discriminating counsel for the plaintiffs .has himself altogether rejected the theory that in the primary devise to Margaret, her heirs and assigns, a mere life estate was given to her. I agree with him in that respect. But some of my associates, who vote to reverse the judgment, if I understand them, think differently. That question will, therefore, be first examined; assuming for this purpose that the alternative is between a life estate in Margaret on the one hand, and a fee determinable only on the event of her leaving no children on the other.
By the clause under consideration, the testator, in the first place, devised the premises in question to his daughter Margaret, “ her heirs and assigns forever,” with a limitation over, in the event of her -dying unmarried and without issue, to her sisters- or their children. If the devise had stopped here there would not have been room for the slightest hesitation in saying that Margaret took a fee, and not a life estate merely. The devise to her and her heirs, with the limitation over, would together constitute a fee determinable on the event specified. It was once a question whether a devise of this nature did not create, or was not an attempt to create, an estate tail. But that point has long been settled (Fosdick v. Cornell, 1 John., 440; Jackson v. Anderson, 16 id., 382.) The material thing *359to be observed is, that no jurist or lawyer ever supposed that such a limitation over, engrafted upon such a primary devise, had the effect of reducing that devise to a life estate. True, the estate might last only for a single life, but it might last forever, and would last forever, unless the specified event occurred, to wit, the failure of issue at the death of the first taker. .Such an estate is a determinable fee, which becomes a fee simple absolute as soon as the condition is discharged. It is also an estate which can be sold, and the purchaser takes it with the determinable quality. But when that quality becomes extinct—in other words, when it becomes no longer possible for-the devise over to take effect— the estate is unfettered and the purchaser holds it absolutely. These views are elementary, and I think they present precisely the condition of the title in the present case. Margaret and, her husband conveyed her estate, and the defendants hold under that title. If the condition on which the premises were given over was gone when she died, leaving issue, their estate is now an absolute fee.
But the testator, after thus devising, in plain terms, the fee to Margaret and then to her sisters contingently on her failing to leave children, proceeded to make a provision concerning such children, if she should leave any. That provision has some peculiarity of phrase which has occasioned whatever there may be of doubt in the case. He declared that, if his daughter Margaret should die “ either before or after his decease leaving lawful issue,” then, not merely the particular estate mentioned in the fourth clause now in question, but the part and parts, share and shares, of real and personal estate," given to her by the whole will, should go to such child or children, their heirs, &c., if one only to him or her solely, if more than one, then “ equally to be divided between them, share and share alike.” This provision, however obscure in other respects, has one clear and intelligible purpose, about which there can be no difference of opinion. As the law was at the time the will was made, if Margaret had died before the testator, the devise of the fee to her would have lapsed entirely, *360so that her children could not take in succession to her. The premises in question, in that event, and all other interests devised or bequeathed to her, would have been undisposed of by the will. Against that event the testator very distinctly provided, by declaring that the children should take everything which the will gave to her; and, of course, in such an event, they would take by purchase as substituted devisees or legatees. And I am of opinion, notwithstanding the redundancy of expression, that this was the testator’s whole intention in making that provision. He had already, in the plainest possible language, given to his daughter the fee, and, if she survived the testator, her children would take the estate under the canons of descent, “ equally, share and share alike.” But he further intended that the devise should not lapse by her death before his own, in other words, that the children should have the estate notwithstanding such an event. In expressing this idea, and no more, he might have said, not inappropriately, “ I intend and declare that the estate herein given to my daughter Margaret shall go to her children, if she dies as well before as after my decease.” Such language would scarcely have been ambiguous, connecting it with the previous devise to her and her heirs. It could hardly be claimed that such words would change the character of a plain devise in fee, which had preceded them. But the language used does not materially differ. We may read the testator’s mind thus: “I have already given this estate in fee to my daughter, provided she has children, and such children, of course, will inherit it. But she may die before this will can take effect, and, therefore (to prevent a lapse), whether she dies before or after my decease I still intend they shall have it.” I am entirely convinced that the testator intended no more than this. If, however, he had a further design, I think it was merely to qualify the general words of inheritance contained in the primary devise, so as to make the estate descendible only to the issue of his daughter. This would not change the result, because it would be an attempt to create an estate tail which the statute converts into a fee simple.
*361A construction which concedes a fee to the children of Margaret as purchasers, and to her only a life estate, of course requires us to strike out the clear and precise words of inheritance contained in the devise to her. Gan this be done ? A glance at the other dispositions of his estate made by the testator, will show that it cannot. He had two other daughters who were married. In the first clause of the will he gave to one of them a dwelling-house and lot “ during her natural life,” on her decease to her children, who might then be living, and their “ heirs and assigns forever.” If there should be no such children, then he gave the same" house and lot to the husband “during his natural life,” and on his decease then over to the testator’s other two daughters (Margaret being one), their “ heirs and assigns foreverand there was a further limitation to their children in fee, in case they, or either of them, should be dead at the time when the preceding limitation should take effect. In the second clause other real estate was devised to trustees for the benefit of his other married daughter “ during her natural life,” and on her decease the same estate was given to the use of her children, and to the “heirs of their bodies forever.” This was plainly an attempt to create an estate tail. The second clause also contained contingent limitations in fee, in favor of the testator’s other two daughters and their children. In the third clause certain other real and personal estate was given to the testator’s wife and to his unmarried daughter Margaret “ during their joint natural lives," and the natural life of the survivor; and in case the daughter survived the wife, then the same real and personal estate was given to Margaret, “her heirs and assigns forever.” In the fifth clause he gave all his real estate in Nova Scotia to his three daughters, their “ heirs and assigns forever.” The sixth and seventh clauses contained some bequests of personal estate which need not be specified. In the eighth and last clause the testator, amongst other things, disposed of all the residue of his estate, giving to his wife, to Margaret, and to one of the married daughters, each a fourth part, declaring, in each instance, the devise and bequest to *362their “ heirs, executors, administrators and assigns forever.” The other fourth he placed in trust for the other married daughter “during her natural life,” and after her decease then for her children, “ their heirs, &c., forever.”
Now, the other parts of the will which have just been referred to, support, very decisively, the construction which the precise and technical words of inheritance contained in the fourth clause are, in themselves, entitled to receive. It cannot fail to be observed that the testator, throughout his will, expressed himself in very plain and unambiguous language, in regard to the different estates and interests which he intended to create. When he designed to give a life estate he said so in the most exact language, and when an estate of inheritance he employed the very words adapted to that purpose. It is true, that in all parts of the will there is great redundancy of expression tending to obscure the general sense, but the quantity of estate intended to be given to each devisee is, nevertheless, defined in terms, the meaning of which is free from all doubt. When we find then, that in the particular devise now in question, the testator used the language upon which the law, according to universal understanding, raises a fee, and moreover the very language employed for that purpose in nearly every other clause in the will, it would seem impossible to conclude that any other estate was intended.
And there is still another argument entitled to great weight, to be derived from a general survey of the will. In nearly every clause an estate of inheritance in the various parcels of the testator’s lands, and in some of them absolute interests in his personal property, are given directly or contingently to his daughter Margaret, and in no instance a mere life estate. Now, referring once more to the provision in the fourth clause, concerning her children, we find that it includes not only the house and lot, which is the subject of the general controversy, but whatever estate or interest is given to her by the will at large in any portion of the testator’s property. If, therefore, that provision is allowed to reduce the fee in this house and lot, so plainly devised to her, to a mere life estate, it must have *363the same effect upon all other parts of the will containing limitations in her favor; and the consequence will be that none of those limitations gave her more than a life estate. This conclusion we know to be false, and in rejecting it we must also reject a construction which seems inevitably to lead to it.
The difficulties above stated, in the way of considering the devise to Margaret as one of a mere life estate, are so great that the counsel for the plaintiffs, in the argument before us, wisely rejected that construction. He conceded that a fee was given to her, and yet claimed that her children, if she should have any, took the same fee, not by descent from her, but by a direct devise to them to take effect as soon as she died. This, too, is the view taken in the opinion of my brother Strong, so far as I am able to comprehend him. It remains to be seen how a result so curious and remarkable is attained. The process is as follows: The devise over to the sisters of Margaret and their children was contingent on her dying “unmarried and without leaving a child.” The argument for the plaintiffs is, that the word “ and,” as here used, creates a double condition, so that it must be performed in both its branches before the devise over could take effect. In other words, Margaret must die botli unmarried and without a child or else the sisters and their children were'not to have the estate. According to this construction Margaret took a fee under the devise to her determinable in one aspect, on not marrying, and in another on having children if she married. Thus, the limitation in favor of the sisters would at her death determine her fee if she never married, and the one in favor of her own children would determine it provided she married and left children. Between these two contingencies another was possible, to wit; marrying and having no children, and in this way only was it possible that the so-called fee given to her might not terminate with her life. It is conceded that she took a fee, and this is the construction on which that concession depends. Beading the will in this way, the fee would be displaced at her death in either of two events: 1st, if she never *364married; 2d, if she married and left a child, an event which actually occurred. But if she married and died without issue, then her fee would become absolute.
This construction, I suppose, is worthy of a serious consideration, or it would not have engaged the favorable attention of some of the judges of this court. But for that difference in opinion I should have thought it a proposition too plain to require an argument, that the testator designed to give the estate to the sisters of Margaret or their children, on the single contingency of her dying without leaving lawful issue. I ' ought, perhaps, to concede that his intention would have been more grammatically expressed, if the word “ or” instead of “and” had been used in expressing the condition on which the devise over was to take effect. But every lawyer knows that in wills and other instruments one of these connectives is often used where the other would express the intention with more exactness and precision. (Fairfield v. Morgan, 5 Bos. & Pull, 38; Jackson v. Blanshan, 6 John., 54; Sayward v. Sayward, 7 Greenl., 210.) The courts have uniformly adhered to the intention, even at a slight sacrifice of the grammar. In the limitation before us, the intention of the testator does not admit of a doubt. If his daughter Margaret never married she could not have lawful issue. She might marry and still have none, and that was the precise event on which the estate was to go over to other objects of the testator’s bounty.
If we adopt the construction contended for on the part of the plaintiffs, these are the results: If Margaret had no children in consequence of never marrying, the sisters or their children were to take the estate on her decease. But if she had none in consequence of sterility, they were not to have the estate. So if she married and lived in a state of vestal seclusion from her husband, that would be a performance of the condition so as to defeat the limitation over. If, on the other hand, she fulfilled the primary design of that institution, and became the mother of children, the penalty would be the loss of the fee, under the limitation over in favor of those children. Mo example of such a will is found in the books, *365and I venture to say that such a will was never made. It is very common in testamentary dispositions to devise a fee to a son or daughter, and on failure to leave issue, then over to surviving sons and daughters, or other relatives having claims on the testator’s regard. But I repeat, there is no instance of a will where the limitation over was made to depend on a double condition, so that marriage, without issue, would defeat it. The uniform and natural intention in all such cases is, to give the estate to others on the failure of issue or descendants to take the inheritance of the primary devisee. To construe this devise differently is to impute to the testator a singularity of motive, rarely if ever found to exist.
Then as to the limitation in favor of the children of Margaret, if she should leave any, I have already, in refuting the construction which gives to Margaret a mere life estate, explained the object of that provision. Let us now look at it again in view of the fee which is conceded to her under the primary devise. What kind of a fee is this, if the plaintiffs are right in saying that the children were, in any event, to take the same fee as purchasers, and not by descent from their mother ? Such a fee was never heard of before. It is very common to give a life estate to a son or daughter, with remainder in fee to issue. But a devise of a fee to the first taker, with an executory limitation in fee to his or her issue I have never seen. The fee is often given over to others by way of executory limitation, on failure to leave issue, but such a limitation over to the issue, after a fee to the ancestor, would be a disposition of property somewhat grotesque and remarkable. So there are instances, almost innumerable, of fees at the common law, and of attempts to create fees, since our statute abolishing entails, descendible only in a direct line to issue or descendants. These are known as estates tail. But a fee transmissible to remote or collateral kindred, yet incapable of descent to issue, is an estate hitherto without a name. I do not deny that such an estate can be created. But it needs at least one example of the kind in all the books in order to render plausible a construction which imputes to the testator a design *366of giving such an estate to his daughter in the will now before us.
There is another view which is decisive against this construction. The concession of an estate given to Margaret, which could endure beyond her lifetime, is entirely upon the idea that the limitation to her sisters would be defeated if she married although she left no issue. Every other possible event, as we have seen, would terminate her interest when she died, and so reduce it to a mere life estate. If she did not marry, the estate on her death was to go to her sisters. If she married and had children it was to go to them (so the plaintiffs insist) not by descent but by the direct gift of the testator. These limitations must take effect, if ever, at the time of her death, and would, of course, reduce her estate to one for life, but for a third possibility, to wit, dying without issue after being married. Now if the testator, in that event, meant that the estate should not go to the sisters, we cannot be mistaken in supposing that he meant' the same thing in respect to the sisters’ children. These children were to stand in place of their mothers, if the latter should happen to die before the decease of Margaret; in other words, before the limitation over could take effect. Let us mark, then, the language of this substituted devise. It is if in case Margaret should die “ without lawful issue” the sisters should be dead leaving children, then their (the sisters’) shares were to go to such children. Here is no double condition. The sisters’ children were, in the event of their mothers’ death, to take the estate on the single contingency of Margaret’s “dying without lawful issue,” whether she married or not. No one, I am sure, will contend that the contingency is single as to the children and double as to their mothers. The former were simply to take the place of the latter; and inasmuch as the precise language of the provision renders it certain that the former were to take on the single event of a dying without issue,, so we may be certain of the same intention in regard to the latter. I do not see how this can be made plainer.
What then is the result? If, as the plaintiffs insist, the *367children of Margaret were to take as purchasers on her death, inevitably her estate was one for life only. She would either leave issue or she would not. If she did, such issue were to have the estate by the testator’s own gift, and not by descent from their mother. If she did not, then the sisters or their children were to have it. One of these limitations would terminate the estate given to Margaret the moment she died, and such an estate is a life estate. This, I suppose, will not be doubted. But it is conceded that she took a fee, and not a life estate. The reasons for this concession I have heretofore stated. The irresistible consequence is that her fee was not determinable according to the plaintiffs’ construction on her dying leaving issue, but on the opposite event, single in its nature, of dying without issue, whether she did or did not marry. This is the defendants’ construction and it is mine, and this leaves the children of Margaret to take only under the canons of descent in case the devise did not lapse by the death of their mother before that of the testator. I adopt this construction because it is an easy and natural one, and because difficulties absolutely insurmountable are in the way of any other. I will only add that I am as confident that such was the intent of the testator as I am of the fact that he made the will.
I think the judgment should be affirmed.
Demo, J. I wish to add a few words in confirmation of the construction which Judge Comstock has given to the clause limiting the estate to the testator’s other daughters, in case his daughter Margaret should die “ unmarried and without leaving a child, her surviving.” According- to. the argument of the plaintiffs’ counsel both circumstances were essential to enable the executory limitation in favor of the other daughters to take effect, namely, that Margaret should die either without ever having been married, or without leaving a husband, as well as without leaving a child surviving her; so that if she were married, though she died childless, her sisters, or their issue, would not take as devisees, but the estate would descend, if *368she died without a will, ás in the case of the intestacy of a person seised in fee simple. If this is the true construction of the clause, it may well be that the sentence immediately succeeding the one mentioned, in which her children are referred to, should be construed as-an alternate executory limitation in their favor. By the will, as thus interpreted, the fee simple, carefully limited to Margaret in the first instance, was certainly to be defeated by force of the subsequent limitations, if she left children, and was also to be defeated in favor of her sisters, if she did not leave children, except in the event of a marriage which had not'been followed by issue which survived her; so that the only contingency which would preserve the inheritable estate would be an unfruitful marriage; In every other event it was to be defeated. I do not wish to add anything to show the improbability of such an intention, or its incongruity with other independent provisions of the will; but I think a critical examination of the fourth article alone will show that the death of Margaret without surviving issue, - simply, and without any other circumstance was the event-which- was to enable the other daughters of the testator to take. The intention of the testator was, that the issue of the other daughters should stand in the place of their mothers in case the latter should die leaving issue, before the limitation should take effect by the death of Margaret. It was plainly the intention that the issue of the other daughters should take upon the same precise event upon which their mothers would have taken if they had been living, and upon no other-contingency. The single object of the clause upon that subject was, to enable the issue to represent their mothers. But, in making the provision, it became necessary again to define the event upon which the fee simple given to Margaret was to be superseded by this executory limitation, and it was stated tobe “in case my said daughter Margaret shall die without laiuful issue;" and if, at her death, the other daughters should then be dead, &c. In order to hold that any other circumstance than simply the death of Margaret without issue was necessary to give the estate to the other daughters, we must impute to *369the testator an intention to give it to the issue of these daughters in an event in which he would have withheld it from their mothers. This would be hostile to the language as well as .to the apparent scope of the provision substituting the daughters for their mothers, for it is the parts and shares bequeathed to their respective mothers which the issue are to have, if they happen to be the parties representing these branches of the family at the death of Margaret.
If the question were to be determined upon the primary language alone, I should, I think, concur in the conclusion arrived at by Judge Comstock. The sentence seems to me to be elliptical. What the writer meant to say, I presume, was: “If my said daughter Margaret shall die unmarried and” also if she shall die “ without leaving a child her surviving, then,” &c. But however this may be, and admitting that the language here is equivocal, yet, when the testator came again to speak of the contingency upon which the property was to go to the other branches of his family, he leaves no doubt but that the event was the death of his daughter Margaret without issue surviving her, without any reference to the question whether she had ever been married. I am in favor of affirmance.
Judgment reversed.