MR. CHIEF JUSTICE GAINES delivered the opinion of the court.

This is a certified question from the Court of Civil Appeals for the First Supreme Judicial District. The statement and questions are as follows:

"In this cause the appellees sued the appellants to recover a tract of land alleged to be situated just north of the Thos. Cartwright league in Colorado County, and adjoining same. The defendants, claiming under the Cartwright grant, defended on the theory that the land was within the boundaries of the Cartwright league and the suit thus became one of boundary, the only issue in the case being the true location of the north line of the Cartwright league.

"The plaintiff prevailed in the court below and the cause is here for the second time on appeal. At a former day of this term we reversed the judgment and remanded the cause for reasons not germane to this certificate. The cause is now pending before us on motion for rehearing. As the point we certify involves the proper construction of the field notes of the Cartwright league and is necessary to be determined in view of another trial, we set them out in full.

"`Begin at the southwest corner of the Rabb survey No. 9 on the north fork of the Colorado river. Thence north 20 deg. east 13,400 varas to the northeast corner on stake in prairie. Thence north 70 west 2000 varas to the northwest corner on stake in prairie. Thence south 20 west 12,500 varas to the southwest corner on a cotton-wood mkd. No. 10 . . . Thence down the river to place of beginning.'

"The date of the survey was 1824 and the river corners, which were well marked, have been destroyed by changes in the river. The extent of these changes is not known by any living witness.

"To allow the contention of defendants will extend the length of the side lines of the league 1200 varas and the evidence renders it probable, if it does not conclusively show, that at no time in its history has the river bed been so far north as to be within 13,400 varas of the north line contended for by defendants.

"There is, however, circumstantial evidence tending to show that the original surveyor actually carried the lines of the survey to the corners contended for by defendants.

"The land in controversy, or such parts of it as may be found to lie south of the Cartwright north line, belongs to defendants. The land, or such parts of it as may be found to lie north of that north line, belongs to the plaintiffs.

"On the former appeal (76 S.W. Rep., 61) we held that the field notes of the Cartwright called for open unmarked prairie corners at its northern end, and that evidence was inadmissible to show that the surveyor actually went beyond his call for distance and set an undescribed stake. That the length of the line could not be extended by the call for the stake. On this appeal we adhered to that holding.

"By the 18th assignment of error the appellant again presents the question for adjudication. In view of the importance of the question generally and to this litigation, and the unsettled state of the law upon the point, we respectfully certify for your decision the questions:

"First. Did we err in our construction of the field notes?

"Second. Did we err in holding that evidence was inadmissible to extend the calls for course and distance?

"We call attention to Mann v. Taylor, 69 Am. Dec., 750, and Devlin on Deeds, Vol. 2, Sec. 1011."

The lines as actually run and the corners as actually established when consistent with other locative calls fix the true boundaries of the survey. The location of such corners may be proved by any admissible evidence sufficient to lead to a belief of the fact. It may be that surveyors are careless in setting stakes at corners in a prairie where there are no natural objects to mark their exact locality, or that it is customary with them to use a stake for making corners of a very unstable character. Yet in the absence of proof to the contrary, it must be presumed that they have done their duty and have marked the corners with some object of reasonable permanence. Where a stake is once placed, it fixes the corner as conclusively as if marked by natural objects. Owing to the fact, that it may be removed or obliterated, its location may be more difficult of proof; but if proved it fixes the corner with the same certainty as where it is marked by a permanent object.

The Court of Civil Appeals in their certificate cite us to the case of Mann v. Taylor (69 Am. Dec., 750; 4 Jones' Law 272.) In that case the Supreme Court of North Carolina hold that the description — "Beginning at a stake, running thence north five hundred chains, thence west two hundred and fifty chains, thence south five hundred chains, thence east two hundred and fifty chains to the first station" — described no land and that the pretended deed was a nullity. The decision is doubtless correct. There being nothing in the description to mark the location of the stake, the land intended to be conveyed by the deed was wholly uncertain. But when the court says "a stake is an imaginary point" we do not concur. When the surveyor says a stake it must be presumed he means an actual stake — which is a very real object.

In Bolton v. Lann ( 16 Tex. 96), the court speaking through Mr. Justice Wheeler holds that "where the lines of a survey have been run and can be found, they constitute the true boundaries, which must not be departed from or made to yield to course and distance," etc. The same doctrine is announced in Dalby v. Booth ( 16 Tex. 564). In other words, the footsteps of the surveyor must be followed. In Stafford v. King ( 30 Tex. 257), Mr. Justice Smith in a well considered and very elaborate opinion lays down the rules which should govern in a case of this character, as follows: In case of conflicting calls, the order of dignity and control are: 1st, Natural objects, 2d, Artificial objects, and 3rd, Course and distance. That is to say calls for course and distance must, in case of conflict, yield to calls for natural objects and to artificial objects, among which he expressly mentions "stakes." The principles so announced have since prevailed in this court without any important modification. Gerald v. Freeman ( 68 Tex. 201), is not in conflict with the principles laid down in the cases previously cited. There it was held that a call for distance will not yield to a call for an unmarked line in the prairie which cannot itself be ascertained except by running the boundaries of another survey according to course and distance. The surveyor's testimony in the case showed that in making the survey in question he did not run all of the lines of the survey which was supposed to join it, but merely estimated its distance from maps in the land office. That is he merely conjectured the location of the line without knowing that the distance called for reached it. The decision was clearly correct.

The Court of Civil Appeals seem to have treated the case as if the call had been 13,400 to a corner without mentioning a stake. If such had been the case their ruling would probably have been sound, but a stake is an artificial object and its mention cannot be disregarded. If the place where it was originally located can be established, the call for distance should yield to it.

We answer both questions in the affirmative.