This is an action of trespass quare clausum fregit, and was tried on the general issue. The plaintiff claimed under a deed to himself from William S. Rowland, dated December 13th, 1828. In it the land is thus described: “ A certain tract of land in the county of *284Beaufort, on the south side of Pamptico river, and west side of D. creek, and on or near the upper Cypress Swamp, a branch, of South D. creek, it being part of a tract of land formerly belonging to the estate of John Rowland, deceased, which fell to me by heirship by the death of my mother.”
For the purpose of identifying the land conveyed, and shewing that the locus in quo was a part of it, the plaintiff gave in evidence a patent to John Rowland, issued in 1794, for a tract of land described thus: “ Two hundred acres lying in our county of Beaufort, on the south side of Pamptico river, beginning at .a sweet-gum in Douge’s third line on or near the Cypress swamp, a branch of South Dividing creek; thence west 100 poles to a gum; thence south 320 poles to a holly; thence east 100 poles to a red bay; thence to the beginning.” And also, proved, that John Rowland died, leaving two children, William S. and Mary; to whom the granted land descended as the heirs at Law of the patentee.
Upon that evidence the defendant insisted, that, although the locus in quo was a part of the tract described in the patent, yet that it did not appear that the deed to the plaintiff covered it. Of that opinion was the Court; and the plaintiff submitted to a nonsuit and appealed.
It is possible the deed was intended to cover the granted land or an undivided moiety of it, and that the plaintiff might have shewn facts which would have authorised that construction. But, with his Honor, we think Joe did not do so on the trial. The onus was on the plaintiff to prove such facts as would point the description to the land he claims under it. He insists, that he has done so, although he does not shew that the locus in quo descended to W. S. Rowland from his mother, or upon her death; because, as he says, the land is sufficiently déscribed without that, and it may, therefore, be struck out.- It is admitted to be a sound rule of construction, that a perfect description, which fully ascertains the corpus, is not to be defeated by the addition of a farther and false description. But we doubt whether there js in this case any §uph perfect description as will identify the subject of the deed with that of the patent. The deed says, the *285land formerly belonged to John Rowland; but it does not refer to the patent as the means by which he acquired title, and the plaintiff gave no evidence that the patentee owned no other land that would answer to the other calls, namely, as lying on the south side of Pamptico and on the swamp. The deed gives no corners, courses, nor-distances, nor qnantity even; but only the streams, and the channels through which the bargainor derived his title. The reference merely to the water courses is too uncertain as a description, to identify this as the land, unless there be evidence to render it at least probable that John Rowland has no other land on those streams. But even in the reference to the streams, the two instruments do not agree. The grant says, that the beginning sweet-gum is “ on or near the upper Cypress swamp, a branch of South Dividing creek,” without designating on which side of that creek the land lies:, whereas the deed, after mentioning that the land to be conveyed by it lies on the south side of Pamptico river, interpolates in the description the words “ and on the west side of the D. creek.” Evidence might have made it appear that the land mentioned in the patent is on the west side of Dividing creek; but no such evidence was given, and it may lie on the east side. The difference in the descriptions shews, that the two tracts may be different; and their identity cannot be presumed, when there is such a dissrepancy in the terms of description.
It is thus seen that the plaintiff was under the necessity of resorting to the residue of the description in his deed, namely, that the land had belonged to the bargainor’s father and fell to him by descent from his mother or at her death. Now, we have no right to strike out one part of this description more than another part, unless, as has been already said, the part retained completely fits the subject claimed, and the rejected part does not; and unless, further, it appear that the whole description, including the part sought to be rejected, is applicable to no other thing. For it may be that one part of the description will suit several distinct things; while the whole will embrace only a single thing. That single thing is the subject of the conveyance, if it can be found; and no other can be deemed the subject, until it be shewn, at least *286to the degree of moral probability, that there is no corpus that will answer the description in every particular. Proctor v. Pool, 4 Dev. 370. Belle v. Love, 1 Dev. & Bat. 65. Now, it may be, that John Rowland owned other land, and made a will devising it to his wife, who died intestate, and wh°m h descended to the son and daughter; or that the father devised for life to the wife, without more, and that, upon her death, the land fell into possession of the children by descent from the father. It laid on the plaintiff to offer some to rebut such probabilities: not that, in themselves, they are probabilities, but that they are created, as applicable to this case, by the language of the deed; and, therefore, they ought, in some degree at least, to have been answered. In other words, we cannot strike out of the deed any part of the description which is not made to appear false: and while these words are retained in this deed, we must say, upon the evidence given, that this locus in quo does not appear to be embraced by it,
The cases vs Dev. 370* vs Love, i ^v-^at and
Per Curiam. Judgment below affirmed.