Evans, P. J.
The action was in ejectment on the several demises of Eobley D. Smith, executor of G. P. Culverhouse, and of A. J. McAfee, executor of J. G. Colbert, against M. J. Newberry, J. P. Spillers, and L. D. Moore, administrator of Emma M. Eay, tenants in possession, to recover the south half of lot of land number 30 in the 6th district of Crawford county. The plaintiff sustained a nonsuit, and brings error.
1. The plaintiff introduced in evidence a deed from the administrator of G. P. Culverhouse to J. G. Colbert, dated January 5, 1877, purporting to convey, in addition to other lands: “Also lots and parts of lots number twenty-nine (29), thirty (30), and thirty-five (35) in the sixth district of said county, containing four hundred and fifty acres, more or less, with mill-site and water privileges thereon; said tract being known as the old Grant Mill. place.” The deed did not specifically designate all, of the lots or any definite part of them as composing the Grant Mill place, and the burden was on the plaintiff to show that there was a definitely located place known as the old Grant Mill place, which embraced the locus in quo. 2 Devlin on Deeds, § 1013a. The plaintiff offered to testify that from talking with the older citizens in the neighborhood he was able to identify the locus as being included within the boundaries of the old Grant Mill place. The court ruled that the plaintiff should be permitted to prove the tradition with respect to the location of the old Grant Mill place as it existed in 1877 (the date of the deed from Culverhouse’s administrator to his testator), but not to testify that he knew that the locus in quo was located in the old Grant Mill place from information obtained from the older citizens of the neighborhood, who knew the boundaries. Traditionary evidence as to ancient boundaries and landmarks is admissible. Civil Code (1910), § 5772. But the reputation in the neighborhood at the present date is not admissible unless it be traditionary or derived from ancient sources or from those who have peculiar means of knowing what the reputation of the boundary was at an ancient date. Shutte v. Thompson, 82 U. S. 151 (21 L. ed. 123). There was no preliminary proof showing the source of information of the persons from whom the witness obtained his information. If his informants were acquainted with the land lines, they were competent witnesses to the fact to which the plaintiff proposed to testify, *475and his testimony clearly would not have come within the exception to the general rule of hearsay. Before reputation as to boundaries will be received in evidence, not only must it appear that the tradition is ancient and did not arise after any controversy respecting the title to the land, but the tradition must have something definite to which it can adhere, or be supported by corresponding enjoyment and acquiescence. For instance, a tree may be shown to have been pointed out by persons of a bygone generation as the corner referred to in an old grant or deed. A tree is something to which the tradition can adhere. Mendenhall v. Cassells, 3 Dev. & Bat. (20 N. C.) 50. 'It is one thing to locate the ancient landmarks and boundaries by traditionary evidence, but it is quite a different thing to show by such evidence that an unlocated larger tract embraces a smaller one. In the former instance the definite thing upon which the reputation hangs is something which existed as a landmark or boundary, and relating to which there would be common knowledge in the neighborhood.’ But it would not be competent to prove by tradition a general reputation that a smaller tract of land is covered by a deed to a particular person, where the land is described by a particular name, and is not definitely located by any evidence. Toole v. Patterson, 31 N. C. (9 Ired.) 180. To state the proposition a little differently, title to land can not be proved by parol. Thus, testimony that a lot “was known in the neighborhood as John Hardy’s land” is inadmissible to show that it was John Hardy’s land. Testimony that the witness had heard the land in dispute “called the Bailey lot by settlers around there” is inadmissible, because hearsay. Shingler v. Bailey, 135 Ga. 666 (8) (70 S. E. 563); Heatley v. Long, 135 Ga. 153 (8), 154 (68 S. E. 783).
2. McAfee, the executor of Colbert, testified that he was not familiar with the land during his testator’s lifetime, but after his qualification as executor in 1884 the land came into his hands. His testator had paid the taxes thereon during his life, and-since his death the taxes had been paid by the witness as executor. In 1891 or 1892 he leased it to Mr. Powell for five years, who operated it for turpentine, and at the end of Powell’s lease one Busk entered as his tenant and took the second “dippings” from the trees and cut cross-ties. Busk was in possession for two years, and was followed by Bissell under contract to cut cross-ties and saw timber *476on the land and also to use the turpentine. Afterwards, in 1905, he leased the land to B. Ii. Ray, the husband of Emma M. Ray, who entered as his tenant. Ray and Doctor Newberry entered into possession under the lease contract of the witness with Ray, and Ray sawed the timber on the land. The tenancy of New-berry and Ray continued under the contract from July 1, 1907, for two years. During the tenancy of Ray the land was cleared by Newberry, one of the defendants. During the time that he had the land in possession as executor of Colbert, claiming it under the administrator’s deed, neither of the defendants asserted adverse title against the plaintiff, nor until about 1908, when one of them claimed the land, and later the others asserted a claim to it. The witness further testified that the locus was included in the Grant Mill place; but on cross-examination he said that he did not have any actual knowledge of the location of the locus until about twenty-five years ago, when he went over the mill tract, and, even then, he was not certain that he rode over the locus; and he only definitely ascertained the boundaries to the south half of lot 30 when he had it surveyed about a year before the suit. “The Grant Mill place is a certain lot, lot 29; we do not claim that. I think the Grant Mill place comprised lot 29; this was right adjoining and ran down to the mill property. . . . I don’t know what Grant had there; he died or got killed during the war. . . My idea is that everything he owned was a part of the mill place. . . I don’t know whether Grant ever owned 30 or not. . . I know thirty was in the mill place, because we bought in a body.” A witness by the name of Mathews testified: “I can’t say whether in 1877 what is known as the Grant Mill place embraced the south half of lot 30. . . I hear, that is the tradition here, that the south half of lot 30 was included in the land known in 1877 as the Grant Mill place.” On cross-examination he said that he did not have that information in 1877; although living in the community he did not know the boundaries nor on which lot the Grant Mill was located, but it was either on the locus or the adjoining lot. W. T. Peterson testified: “I know the tradition, that is what people said as to the Grant Mill property and the place which was included in that. It was always my understanding that the south half of lot 30 in the sixth district was included in what was generally known as the Grant *477Mill place. I am not able to answer you directly as to whether there is any understanding in that neighborhood, or tradition or general knowledge as to what the lines of the land is known as the old Grant Mill place. I presume some people know, that is better acquainted with the lines than I am. I can not answer directly as to what is the general tradition as to where' those lines are of the old Grant Mill place. I don’t know.” We think this evidence sufficient to withstand a nonsuit. Witnesses were allowed to testify that the general tradition was that the location of the boundaries of the Grant Mill tract would include the locus. It is clear that the witnesses were not testifying to a present-day tradition, but a common reputation as to the location of the boundaries in 1877, though their knowledge was acquired at a subsequent date. This testimony should have been restricted to traditionary evidence of the landmarks and the boundaries; but as it was received, under the ruling of the court, as having probative force to establish the locus as being included within the boundaries of the larger tract, according to the common reputation of the neighborhood at the time the administrator of Oulverhouse made his deed to the plaintiff’s testator, we do not think that it can be said that there is absolutely no evidence that the old Grant Mill place did not embrace within its boundaries the 'premises in dispute. We think further that the evidence was sufficient to authorize an inference of seven years possession under this deed as color of title, so as to make out a prima facie case of prescription. Then, again, the plaintiff testified as to his actual possession of the locus, to the knowledge of one of the defendants and the husband of the intestate of another. This evidence would be sufficient to permit a recovery, if nothing more appeared, on the plaintiff’s prior possession.
3. Smith as administrator of Oulverhouse brought a suit to collect Colbert’s bid at the sale. Colbert defended the suit on the ground that Smith’s intestate had no title to the land, and especially to the locus in quo. The suit terminated in a judgment for the administrator for the purchase-price. This record was introduced in evidence; and it is the contention of the defendant in error that Colbert’s executor is estopped from now asserting that his testator had title to the land. The judgment in that case is not conclusive on the subject of title; for the reason that the doc*478trine of caveat emptor applies to administrator’s sales, and the purchaser thereat, in the absence of fraud, is liable on his bid, notwithstanding the administrator’s intestate may not have title to the land. The statement in the plea in the former case that the administrator’s intestate did not have title to the land was no disclaimer that the locus was included within the Grant Mill place, claimed to have been owned by the administrator’s intestate, and sold as his property. This admission was nothing more than a declaration of the vendor’s defective title, and in no sense was it an admission that the plaintiff’s testator purchased in bad faith. We do not think that the fact that he challenged the sufficiency of the intestate’s title, in resistance to a suit for the purchase-money, is such evidence of bad faith as would prevent the purchaser from acquiring prescriptive title based upon adverse possession under that deed.
Judgment reversed.
All the Justices concur.