delivered the opinion of the Court.
This was an action of ejectment, in which the verdict and judgment were in favor of the defendant. William A. Harris was allowed by the Court to be examined'as a witness for the defendant, to prove the boundary line in dispute to be as contended for by the defendant, against the objection of the plaintiff, on the ground of interest. He was a very material and important witness on the question in controversy.
*53The witness on the 6th August, 1840, conveyed, by deed of that date, this, with other lands, to Massey Hill, and warranted the title. On his examination he says, “he making the deed to Massey ITill, I conveyed to the line as claimed by Senter, as shown on the plat running from double index south to the pine.” Hill conveyed to W. P. Hayes, he to Wm. Hayes, and he to the defendant, Senter, all with general warranty.
Now, the question is, Was the witness competent, upon this state of facts ? A decision either way would not be without authorities to support it. It is not proposed now to review them, as that is sufficiently done by all the recent writers on evidence.
In Burke vs. Clarke, 2 Swan, 310, it was decided that the warrantor of the soundness of a slave could not be a witness for his vendor in a suit against the latter for the unsoundness of the slave, by a purchaser from him with a similar warranty. In Elliot vs. Boren, 2 Sneed, 662, it was held that the vendor of land, with a covenant that he “was lawfully possessed of the land,” was incompetent as a witness for the vendee in an action of unlawful detainer to prove the possession.
But this is a different case from either of those, and requires us to go a step farther, or, at least, to extend the same principle.
It is laid down in 1 Grreenleaf on Ev., § 397, that, “ if the title is in controversy, the person who is bound to make it good to one of the litigating parties against the claim of the other, is identified in interest with that party, and therefore cannot testify in his favor.” That is doubtless the law; but it is not exactly the case before us. Here the witness is a remote, and not an immediate war-*54rantor; and in this differs from both of our own decisions, and the case stated in Greenleaf. At the close of the section cited it is stated that, to constitute incompetency, the witness’s “liability must be direct and immediate to the party calling him, and not circuitous, and to some other persons; as if a remote vendor, with warranty, is called by the defendant as a witness, when the article has been successively sold by several persons, with the same warranty, before it came to the defendant.” Without stopping to examine the cases referred to by the author, to sustain this position, we contest its application to cases of successive warranties of land, and hold that in such cases, inasmuch as the witness is interested in the establishment of the title of the person by whom he is called, and in case of its failure would be liable to his action, he is incompetent. It is not a remote liability, although he is not an immediate warrantor. It was held in Hopkins vs. Lane, 9 Yerg. 79-85, and well supported by authorities cited, that the covenant of warranty runs with the land, and that the last vendee, with warranty, may maintain an action of covenant against the first or any other warrantor. So the witness, Harris, was immediately liable to the action of defendant, if his title failed to any part of the land embraced in his deed. For some purposes he might be a competent witness, but not upon any question in relation to which he has made a covenant from which his evidence would tend to relieve him. Such may have been the nature of the question in the unreported case to which we are referred, or the competency may have been restored by a release.
The judgment will be reversed, and new trial granted on this single ground.