Opinion.— The proof shows that the location of the north line of the Strunk three hundred and twenty acre survey can be fixed with absolute accuracy by running the lines of that survey according to its field-notes and the marked corners on the ground. Its northeast corner is at a willow which is still standing. Its southwest corner is well known,' and its bearing trees; two of them yet stand, and the stump of the third remains. By running a6 line from this southwest corner north, nineteen degrees east, the northeast corner of the Gilmore survey is passed at six hundred and seventy-five varas; this corner of the Gilmore survey is well known, and fixed by its bearing trees, a cottonwood marked XH, and a willow. Continuing this line to a point from which a line drawn according to the calls of the patent from the northeast corner of the Strunk survey will intersect it determines with certainty the northwest corner of the survey, and this line is the true south boundary of the Cervantes survey. The distance from this Gilmore corner on the Strunk west line to the corner is but one hundred and seventy-four varas.
If we disregard the calls of the Cervantes patent for the lines and corners of other surveys and assume that other *479surveyors who have since surveyed the land have measured it correctly, then it is found that by measurement the lines in the field-notes would never reach the line of the Strunk three hundred and twenty acre survey, nor extend to the Christopher east line, nor to the northwest corner of that survey, but a vacancy would be left of very irregular shape, and apparently without reason for it, between the Strunk, Christopher and Cervantes surveys. It is only by attributing to these measurements controlling weight in the determination of boundary that any land would be left to satisfy the pre-emption file and location under which Marshall claims.
The proof tends to show that, in fact, the surveyor did not run all the lines of this Cervantes survey separately for it upon the ground, but that he platted his field-notes from other surveys made. He doubtless determined the distances by calculation, and manifestly designed to close the surveys upon each other.
The difference in distances and the excess of some sixty seventy acres, if so much, between the calls of the patent for distance and quantity, and the calls for lines and corners in the patent, and the quantity thereby included in this six hundred and forty acre survey, are too slight to raise any presumption against the intention that the boundaries of these adjoining surveys should be the boundaries of the Cervantes survey.
But we think the calls of the patent for the Strunk north line and the Cervantes lines, as these can be definitely fixed upon the ground by artificial and well known marks, must control the determination of these boundaries. The line of a tract of land may as well be the subject of a call as any other object, and in this case, as we have seen, where that line is can be very readily found upon the ground by the calls for its corners, which are well known and established. Bolton v. Lann, 16 Tex., 96; Johns v. Shultz, 47 Tex., 579; Booth v. Strippleman, 26 Tex., 442; Buford v. Gray, 51 Tex., 335.
*480The judgment of the court upon the facts ivas, in our opinion, fully supported by the testimony.
It is assigned for error that the court refused to permit the defendants to introduce proof of improvements made by them in good faith, but the right to recover for improvements was dependent upon whether the defendant had possession of the premises at least one year next before the commencement of the suit. Here the defendant had not alleged such possession, and he had, therefore, set out no right to recover for his improvements. His pleadings were not merely defective. There was an utter absence of any foundation laid to warrant a judgment in his favor. It was not error, therefore, to reject the testimony offered. E. S., art. 4813.
Affiemed.