This case turns upon the construction of the will of Benedict Middleton, which bears date in 1782. The division which was made of Jane Wroe's estate is assailed as illegal, so far as respects that part *313of her slaves which she derived from her father, under the limitations of his will. The bill, indeed, demands a repartition of the lands and slaves held by Ann Garner under the will, as well as of those held by Jane Wroe in the same manner. I understand the counsel, however, as confining the question to the case of Jane Wroe's slaves, as the statute of limitations has secured the title under the partition of Ann Garner's; and I shall accordingly consider that question only.

It is conceded on all hands (hat the limitation over is good, and the true question is whether under that limitation the parties are to take per capita or per stirpes ? I am of opinion that they take per stirpes.

There can be no question that the devise in the first clause of the will indicates the intention that the devisees and legatees should take per stirpes. The estate is given to Benedict Lamlcin, the representative of a deceased daughter, and to the five living daughters, each to have one sixth; and it is most distinctly declared that Elizabeth Lewis is to have only one sixth, which sixth is to pass in remainder to her children. As original takers, then, under the will, the children of Elizabeth Lewis took by stocks, and not per capita. Is there any thing in the will from which it is necessarily to be inferred that the testator abandoned this distinctly defined intention of giving per stirpes instead of per capita? I think not. Indeed, such a construction would involve inconsistency in the will. For let us suppose that Hannah, the first decedent among the devisees except Elizabeth Lewis, had died before Elizabeth Lewis instead of after her. In that event, according to the will, Elizabeth Lewis would have taken one fifth of Hannah's portion (either for life or in fee, it matters not which) unless we suppose her to be excluded, or her children to take with her. But she cannot be excluded ; for she is one of the survivors, and the share of the deceased daughter is to be divided among the survivors. Neither can *314it be believed that the testator intended to give her one , ' , portion of Hannah's share, and her four children each one equal portion, which must be the case if they take per capita; for thus one stock, viz. Elizabeth and her family, would take five times as much as either of the other stocks. Had Elizabeth survived Hannah, therefore, it would seem clear that she would have taken one fifth in absolute property, or for life only, with remainder to her children. In either case the construction contended for would be rejected, and the estate would go to the stock, instead of the descendants per capita.

If such then would have been the effect of the will if Elizabeth had died after her sisters, can it be seriously contended that it is to receive a different construction, because she died before them ? If We stick to the letter, there is as much reason for applying the partition per capita to the one case as the other. But in truth there is no absolute necessity for its application in either. The will furnishes another among the innumerable instances in which unskilful draughtsmen have used one •set of words to provide'for many shifting contingencies, and have welded together what ought to have been kept separate and distinct. It declares that “ if either of the daughters or grandchildren die without issue, his or her share shall be equally divided among the survivors of them and their heirs.” I incline to think that this is to be understood distributively, thus: “if either of my daughters or my grandson B. Eamlcin should die without issue, his or her share [that is, one sixth of the whole] shall be equally divided among my surviving daughters and grandson aforesaid; and if either of my other grandchildren (children of Elizabeth Lewis) die without issue, his or her share [namely, ¿ of -g-th] shall be equally divided .among the said surviving grandchildren.” If this be the just construction, then if Hannah had died before Elizabeth, the latter would have taken one fifth of her portion, which upon her *315death would pass to her children as distributees, — not as purchasers; but as she died before her sisters, she and her children took no portion of their estates. This seems indeed the more reasonable, as upon her death her sisters took no portion of hers, but it passed to her children by virtue of the limitation over. The parties, however, have acquiesced in a partition which has treated them as entitled jure representationis, and as they have not sought to set it aside, I am not disposed to disturb it; the less so, as in cases like this it is altogether impossible to devise a projet which is not liable to objection, and I am very sensible there may be objections to this. Perhaps, indeed, the safest ground on which we can rest is, that as the construction is difficult, and the proper interpretation of the will, to say the least, uncertain and obscure, it is best to adhere to that which was placed upon it by the partition of Ann Garner's estate in 1820, and apparently acquiesced in by all concerned until the filing of this bill in 1827. This, of itself, is a sufficient justification of the decree of the chancery court; which should therefore be affirmed.

Cabell and Brockenbkoüg-h, J. concurred.

Decree affirmed.