lated, it is true, to personal property; but the principle is the same in a devise of the realty.

The word children does not ordinarily include grandchildren, or any others than the immediate descendants in the first degree, of the person named as the ancestor. But it may include them when there were no children in existence at the time of the making of the will; or when there could not be any children at the time, or in the event contemplated by the testator; or when the testator has clearly shown by the use of other words, that he used the word children as synonymous with descendants, or issue, or to designate or include illegitimate offspring, grandchildren, or step-grandchildren. (Mowatt v. Carow, 7 Paige, 328.) When there is nothing in the will to show that the testator intended to use the word children in a different sense, it will not be held to include illegitimate offspring, step-children, children by marriage only, grandchildren, or more remote descendants. (Cramer v. Pinckney, 3 Barb. Ch. 475.) In the case last mentioned several other questions arose with respect to the persons intended by the will. And it was held that the words nephews and nieces, in their primary and ordinary sense, mean the immediate descendants of the brothers and sisters. of the person named; and do not include grand-nephews and grandnieces, or more remote descendants. (Falkner v. Butler, Ambler, 514.)

But the peculiar circumstances of the case, and the structure of the will, may show that the testator used the terms nephews and nieces in an enlarged sense, so as to include all the grand-nephews and nieces whose parents were dead. In the same case the testator, by one clause of his will, gave a legacy unto each of his nephews and nieces except J. C., who was not a nephew, but one of the children of a deceased nephew; and by another clause he gave to the children of his nephew J. C. $500-it was held that the brothers and sisters of J. C., and other grand-nephews and nieces whose ancestors were dead at the time of the making of the will, were entitled to the legacies. It was also held that parents and children could not both take, under the description of the testator's nephews and nieces, but only the parents who were living, and those grandnephews and nieces whose parents were dead. (Cramer v. Pinckney, supra; and see Hone v. Van Schaick, 3 Barb. Ch. 488, and S. C. 3 Comet. 538.)

As no person can in strictness be said to be the heir of a person

now living, it is necessary that if a devise be given to a person by that designation, it must be shown by the will that heir apparent was intended, or it will be void. If the devise be to the heirs of the body of B. now living, it has been held to be a good description of the person. It shows that heir apparent is the person meant. But a devise to the heirs of B., who was in truth living, but that fact not stated in the will, is void. (Heard v. Horton, 1 Den. 165.)

A limitation by means of an executory devise, may be made to any number of persons for life successively, if in esse at the death of the testator,-to infants in ventre sa mere, and to persons unborn. Such was formerly the law, and is still, except that by our revised statutes (1 R. S. 723, §§ 15-17) successive estates for life shall not be limited to more than two persons in being at the creation thereof; and if limited to more than two, all the life estates subsequent to the two first entitled, shall be void. It was at one time doubted whether a limitation for life to an unborn person was good; but it is now well settled that it is, and also that an estate limited to the issue of such unborn person to take as purchasers, would be void, being a possibility upon a possibility, which the law will not admit. (Jackson v. Brown, 13 Wend. 441, 442. Stewart v. Nicoll, 3 Johr. Cas. 18. Chapman v. Brown, 3 Burr. 1635, per Wilmot, J.)

The word issue is a sufficient designation of a person in a devise. It comprises children and grandchildren. (Merest v. James, 1 Brod. & Bing. 484. Kingsland v. Rapelye, 3 Edw. 1.)

A devise to the testator's wife during her natural life; and at her decease to be equally divided amongst the "relations on his side," has been held good, and to belong to those persons to whom the personal estate of the testator would go under the statute of distributions. (Doe v. Over, 1 Taunt. 263.) Unless the word "relations" was thus restricted, it would embrace an almost boundless range of subjects; for it would comprehend every degree of consanguinity however remote. (See 2 Jarman on Wills, 25-68, various cases collected.)

WILL.-33

SECTION III.

Of Devises Void for Uncertainty, and of the Remedy when the Will is of Doubtful Construction.

From what has been said in the foregoing section it would seem that a devise will be void, if there be so much uncertainty, either in the subject of it, or of the person intended, as to be incapable of any clear meaning. (Mason v. Robinson, 2 Sim. & Stu. 295.) Although, in the construction of wills, great indulgence is shown to the ignorance, unskillfulness and negligence of the testator; and no testamentary disposition of property will be rendered invalid by a failure to comply with mere technical rules and forms of expression, or by grammatical or orthographical errors, nor by a confused collocation of sentences, if the intention of the testator can be discovered from the whole will. But if this cannot be ascertained, the intended disposition will fail. Conjecture is not permitted to supply what the testator has failed to indicate. The law has provided a definite successor to the property of its dying owner, in the absence of a legal disposition of it. The law must therefore take its course, if the testator from any cause fails to make a disposition of it in language that can be understood. (1 Jarman on Wills, 322,

Perkins' ed.)

There are, in modern times, fewer instances of devises void for uncertainty than in an earlier state of the law. This may be owing to a better understanding of the rules of construction, which have given a determinate meaning to many words and phrases once considered vague and insensible, or to greater skill in the courts in the application of these rules; or to both those causes.

A few examples of each kind of defects will be given. In the early case of Bowman v. Millbanke, (1 Lev. 130,) .the words of the will were, "I give all to my mother, all to my mother." The question was whether this was sufficient to carry to the mother the testator's real estate. Here it was uncertain to what the word "all" referred. It might mean all his real property; it might be all his personal estate; it might be all his estate of whatsoever kind; or it might be all of a particular portion of his estate. It was adjudged by the court to be void absolutely for this uncertainty.

In the later case before Sir Thomas Plumer, M. R. (Mohun v.

Mohun, (1 Swans. 201,) the language of the will was: "I leave and bequeath to all my grandchildren, and share and share alike." Here no property was mentioned as the subject of the gift. It did not appear what he left to his grandchildren. It was contended that the difficulty would be removed by transposing the word "all,” so that it should follow the word bequeath. But that, according to the preceding case, would leave it still uncertain. Besides, it was not a case in which the transposition of words is allowable. These words as they are situated were not inconsistent with the context. The word "all," though inoperative where the testator placed it, was not repugnant. The court held that there was an uncertainty, both in the subject and object of the bequest, and that it was therefore void.

In Jubber v. Jubber, (9 Sim. 504,) the testator, after making his will added a codicil, which was as follows: "I request a handsome gratuity to be given to each of my executors." Here no definite sum is bequeathed, nor is it said by whom the amount shall be ascertained. The will contained a provision that if any dispute should occur it should be settled by arbitration, which should be final without appeal and without reference to the law. It was held that this legacy was absolutely void for uncertainty. And the vice chancellor (Shadwell) said he should not do what Sir Joseph Jekyl did in Peck v. Halsey, as he conceived that he had no power so to do. In Peck v. Halsey, (2 P. Wms. 387,) the testatrix bequeathed to one of her grandchildren by name, "some of her best linen." This was held to be void for uncertainty; but still the master of the rolls, Sir Joseph Jekyl, recommended to the residuary legatee to give some of the best linen of the testatrix to the legatee. Whether the recommendation was followed does not appear; but it is quite clear the court had no power to enforce it. In both the above cases the gift was entirely indefinite, as to quantity.

The indefiniteness of the gift constitutes no objection, if it be of the residue after satisfying previous legacies. (Gibbs v. Tart, 8 Sim. 132. Surman v. Surman, 5 Madd. 123.)

The same principles apply to the object of testamentary gifts. It is enough that the devisee or legatee is so designated as to be distinguished from every other person, and the inaptitude of some of the particulars introduced into the testator's description is immaterial; and this whether the object of the gift be a natural or an

artificial person. A mistake in the name of the devisee, or an erroneous description of him, will not vitiate, if from other parts of the will, or the surrounding facts, there is no reasonable doubt as to the person intended. But when the entire name of the devisee is left blank, no parol evidence is admissible to show who the testator intended. (Baylis v. The Att'y Gen. 2 Atk. 239.) In Clayton v. Lord Nugent, (13 M. & Wels. 200,) the case was this: The testator wrote his will on various pages of a book at different times, part of it being executed and attested in 1820, and the remainder in 1827. No devisees were mentioned by name; but the testator's real estates were devised, "first to K., then to then to L., then to M.," &c. On a slip of paper, pasted into the book, and forming part of the will at the time of the attestation, in 1820, the testator stated that a "key and index to the letter, initials, &c. was in a writing case in the drawer of his writing desk, on a card." The testator died on the 11th of December, 1828, and on that day a card, in his handwriting, and signed by him, was found in the above writing desk, dated January 30, 1828, as follows: "K. signifies Eleanor Mary East. L. signifies Gilbert East Clayton. M. signifies second son of William Robert Clayton. N. signifies eldest son of Richard Rice Clayton," &c. Two years before the testator's death, a card with writing on it had been seen by a person lying before the testator, together with the book containing the will, which appeared to be similar to the card and writing thereon found after his death. It was held that the card found after the testator's death was not admissible in evidence, as a declaration of the testator, to show who were the persons meant to be designated in his will by the letters K. L. M. &c.

In Miller v. Travis, (8 Bing. 254,) it was said by Tindall, Ch. J. that it was a well established principle that where a complete blank is left for the name of the legatee or devisee, no parol evidence, however strong, will be allowed to fill it up; as intended by the testator.

Where, however, the blank was left for the christian name only, parol evidence has been admitted to prove the individual intended. (Price v. Page, 4 Ves. 680.) So in the case of a legacy to Mrs. G., it was referred by Lord Loughborough to the master to receive evidence to show the person intended. (Abbot v. Massie, 3 id. 148.) From the remarks of the judge who delivered the opinion of the court in Clayton v. Lord Nugent, (supra,) it would seem that the