Cowen, J.
By the Court, It is too late to contend that a devise to the survivor or survivors of another, after his death without lawful issue, or any other clause of equivalent import, is void as being a limitation upon an in*518definite failure of issue. The word survivor, or the 'like, qualifies the technical or primary meaing of the words dying without issue: being considered the same as if the testator had added, living at the time of his death. Paterson, v. Ellis, 11 Wend. 280, 292, et seq. and the cases there cited by Savage, C. J.
The question to be decided arises on the last clause in the will; and the objection started upon the remoteness of the event on which the limitation of Barbara’s share was to pass to the survivors, being out of the way, the only remaining one is, whether the words used to designate the persons who were to take by survivorship, comprehend either Margaret, the deceased mother of the plaintiffs, or the plaintiffs themselves. If the former, then the share in question vested in her on the death of Barbara without issue, and came to the plaintiffs by descent. If the latter, then they take by direct devise.
The testator devised to his grand children in fee, his farm, viz.: So much to the children of his step-daughter Margaret, so much to the children of his daughter Susannah, and so much to the children of his other daughters respectively, and of his sou. All the interest he gave to any of his children in the real property, was life estates to Margaret and her husband in her children’s share, and the same to Susannah and her husband in the share of their children; with a provision in trust, securing the income to the other children for their lives, in the shares which were devised to their children respectively. The residuary shares of the personal estate, except the bequests made directly to Margaret and Susannah, were disposed of very nearly in the *same form [ *519 J as the shares of the real estate. And the final clause of the will, so far as it may be construed to reach both kinds of estate, carries them over on the same event, to the same objects. On the death of any of the testator’s children or his step daughter, without issue, then the share or portion which otherwise would have gone to the issue, was to be equally divided among the survivors of his children or grand-children., The event happened on the death of Barbara. Did any interest in that share of the farm intended for her children, pass, either to Margaret or her children ? She was the testator’s step-daughter; neither she nor her children were of his blood ; and it is not contended that either come within the primary meaning of the words under which they claim. Nor is Margaret called the testator’s child in any part of the will; but always his step-daughter, whenever she is named, being some five or six times in different parts of the instrument. It is said she is provided for as a child ; but the same thing might be said of any one for whom the testator had made a similar provision, a sister or a niece, or an adopted child. There is more plausibility in saying that she was considered by the testator as one of his children, from the circumstance that, in the devise which 'distributes the farm, her children are grouped with those of Susannah, John, &c., under the common title of grandchildren. But in that very clause she is styled, as elsewhere, the step*519daughter; and it would seem to me a most violent departure from the language of the law, of the will, and of common parlance, were we to allow so slight an implication to prevail. Suppose that, for the purpose of being explicit, he had classed the children of a cousin in the same way, for the same purpose : “ I give to my grand children in fee, viz.: the children of my cousin Margaret;” and then going on with his real grand children, having before alluded to and provided for her as his cousin in four or five different'parts of his will; — would any one, therefore, think of including her as one of the objects, in a subsequent devise to his children ?
Beside, admitting Margaret to rank as one of the children, it [ *520 ] can hardly be maintained that she would have taken *a fee. The original devise of the fee in the farm shares was to the grand children, the parents taking no more than life estates respectively, either legal or equitable ; and the will would totally depart from the scheme with which it set out, unless the contingent clause meant that the parents should still be confined to the same limited interests — grand children taking the fee by direct devise.
In either view, the claim of the plaintiffs, as heirs of Margaret, cannot be sustained.
We next come to the question, whether her children are, in this particular case, to be considered as the testator’s grand children. I say in this particular case: for the counsel have no difficulty in agreeing that the law, when left to its own operation, denies that they could take as being included in the class of grand children, any more than Margaret, by a devise to children. The counsel also agree that the testator may make his own glossary in the will itself; and that if he is to be understood in any part of it as declaring that for all the purposes of his will, the children of his step-daughter were to bo considered as his grand children,, that overcomes the law. This brings the question down to one of mere construction on the main devising clause. The glossary is to be found there if any where. In this he says, “ I give to my grand children my farm, as follows, to wit: to the children of my step-daughter Margaret, lot No. 1; to the children of my daughter Susannah, lot No. 3,” &c. apportioning his farm. He would not have been understood in this devise, which was quite special, either as to persons or parcels, had he stopped with the words grand children and farm. He therefore adds, to wit, that is, meaning Margaret’s children to come in for so much of the farm, and Susannah’s for so much, &c. And there can be no doubt, that for the purposes of this particular devise, he is to be understood as calling Margaret’s children his grandchildren. But was this merely for the sake of explicitness in this clause, or did he mean to define the term grand children for the purposes of the last clause. That he meant it for the sake of being clear in the first clause, we know. He then goes [ *521 ] on as I have mentioned, contradistinguishing *Margaret as his *521step-daughter, not once again recurring to her children as his grandchildren, and omitting the former explanation in the only clause where the word again occurs, although in that very clause we find him still harping on Margaret as his step-daughter. I admit he may have intended here to use the word in the same sense as before ; but that is not enough to warrant us in departing from the primary meaning. We can but guess that he intended his former explanation to apply.beyond the particular place where it was used, “ to wit,” here I mean her children. Persons answering to the primary meaning both of children and grand children were in esse. The words can be perfectly satisfied in their legal and natural sense, while we accord due scope to the rule which allows an artificial sense, though we confine it to the clause where that was obviously intended. No question is made that the words children and grand children cannot of themselves mean such by marriage, but only by blood— as tho word father, when used in a statute, means a father by blood, not a step-father. 2 Nolan’s Poor Laws, 231, 2, and the cases there cited. This was settled as to the children and grand children so long ago as Hussey v. Dillon, Ambler, 603 ; 2 Eden, 194, S. C., nom. Hussey v. Berkeley, decided by Lord Northington. The same case holds that children mean those in the degree of relationship in the first descent, while the word grand children takes in those of the second descent, and even the third, viz. great grand children ; for these are still grand children, though of a more remote degree. This, says the lord chancellor, accords with common parlance, which is the true way of interpreting words in a will. A residue was there bequeathed to grand children, after various bequests in the will and codicils had been made, both to grand children and great grand children, and one of the latter, Miss Hussey, had been mentioned as a grand child. This is the strongest application I have seen of the rule now set up by the plaintiffs. The word was nearly, if not quite sufficient, to take in great grand children ; but if there were any doubt of this, yet Miss Hussey, one of the latter, being mentioned as a grand child in the context, the lord 'chancellor said it was decisive. It [ *522 ] does not follow from this, nor any other case, that because persons happen in one part of a will to be indicated by an improper word as explained in that part, therefore the use of the same word in another part must necessarily or plainly mean the same persons. I have looked into several cases which let in the words in their artificial latitude, and I find that most of them begin and end with the particular clause where the word is used. Such are several cases where the word children was held to comprehend grand children, Royle v. Hamilton, 4 Vesey, 437; and even great grand children, Withy v. Thurliston. Ambl. 555. In these it was obvious, from the explanations given in the clause, that all the issue of the legatees were intended to be provided for under the word children, and so the court went down to the remotest of the living descendants. Another case, Gale v. *522Bennett, Amhl. 681, was nearly of the same character; but, beside the meaning apparent on the face of the will, the devise was to the children of one who had none, but only grand children. So there was a necessity to look out of the will and extend the word in order to prevent its failing of all effect. And by several cases it is agreed, that, to warrant a departure from the clear and settled meaning of words, there must he either such strong necessity, or at least a clear explanation. A devise to the children of A. shall not let in his grand children, though A. had only one child and several grand children. Crooke v. Brookeing, 2 Vern. 106. This case has often been approved. Radcliffe v. Buckley, 10 Vesey, 195, 201. Here the master of the rolls recognizes what Lord Alvanley said in a previous case, Reeves v. Brymer, 4 Ves. 698, viz. that children may mean grand, children where there can be no other construction ; but not otherwise. The remark may bo too strong for a case where you are searching out a secondary meaning on the face of the will; but there is no case which'warrants us in taking up such a meaning, where the primary sense may be literally satisfied, and there is nothing in the will to raise any thing more than a doubt. This doubt on the face of the will is very properly ranked by Mr. Wigram [ *523 ] among explainable ambiguities. *Wigr. on Extr. Ev. 16, 17, 30 ; and at the pages cited, he lays down the process .by which the mind is to extricate itself, and arrive at certainty. The rule at p. 16, 17, is thus: “ Although the words son, child, grandchild, &c. may be construed in a secondary sense, where the will would be insensible if the primary meaning of the words were adhered to, yet it is only where that is the case, that a departure from the strict sense of the words is permitted.” Here, to the cases I have already cited from Vesey, he adds Shelly v. Bryer, 1 Jac. 207, and Hart v. Durand, 3 Anstr. 684. The latter case is parallel in principle with Crooke v. Brookeing. Shelly v. Bryer is more material to the case in hand. There the testator gave the general residue to his nephews and nieces, and afterwards by a codicil he called Harriet Shelley his niece, and recognized her as entitled to the residue in his will under the description of niece, although she was in fact his grand, niece. Upon this it was contended that the testator had so explained the residuary clause as to let in grand nieces ; but 'it appearing that there were nieces in esse who could take under that clause, it was answered by Sir Thomas Plumer, M. R. thus : “ It appears by the cases that where there are, or may be at the time when the distribution is to take place, persons answering the description, the court is not at liberty to include any not within the terms.” Accordingly, Mr. Wigram adds atp. 30, before cited: “ So, son means an immediate descendant; where, however, with reference to extrinsic facts, it is impossible that the word can be used in such its proper sense, that construction of, the word is, of absolute necessity, excluded ; and the necessary inference, that the testator used the word in some improper or inaccurate *523sense, lets in the inquiry, in what sense the testator used it.” He refers to Steede v. Ferrier, 1 Freem. 292, 477. There it is said to have been agreed, “ that if a man hath Robert a grand son and no son, and he deviseth to his son Robert, the grand son should take, for he is a son, though it be with an addition of grand." The case itself was this: a man having both a son and grand son Robert, devised to his son of that name ; but, he dying, the testator republished his will, and this being equivalent to *making an original will, a majority of the court inclined that [ *524 ] the grand son should take. Such are some of the cases illustrating the ground on which courts have departed from the primary sense of words; such the rules deducible from them. If we depart in the case at bar, we must do so on mere conjecture ; and without the 'least necessity. There are, and were at the time of the will and the testator’s death, both children and grand children of his blood ; and there was nothing in the character of the devise over, which prevented their'taking in the manner, proportion and extent of interest indicated. I know it was contended in argument, that the moral beauty and harmony of the devise over may be thus marred, there being six original takers, but only five remainder men. The case was compared to a Grecian temple with six columns in front and only five in the rear. If it were impossible to build such a temple, the consequence might follow. But- the difficulty being one of mere taste, the answer is gustibus non. The moral appearance will be very different, as the temple may happen to be viewed by the successful or unsuccessful party. Another figure was adopted. It was said that by including Margaret’s children in the first clause among the testator’s grand children, he baptized them by that name. This can mean nothing less than that he clearly intended they should be considered his grand children to all intents and purposes, not- only in that clause, but in other parts of his will. I have already shown that this by no means followed, and I think the cases quite clear that unless it be apparent, nay necessary, in order to give effect to the subsequent clause, the secondary sense cannot be imported into it.
I have not looked into this case the 'less, because the vice chancellor of the first circuit adhered to the primary sense when appealed to for a distribution of the personal estate of Barbara, which was carried over by virtue of the words in question, surrounded with the same circumstances now again insisted on as qualifying them. Barnes v. Greenzebach, 1 Edw. Ch. R. 41. The result is, to my mind, what I had supposed it would be from the known learning and industry of that judge ; a conviction that his decision was *not only warranted, but required by the spirit of the [ *525 ] adjudged cases, and the only safe principles of construction.
The result is, that judgment upon the special verdict must be rendered in favor of the defendants.