But a single question needs consideration. The title of the plaintiff to the property described in his complaint is not challenged, but the contention of the defendants is that the land which confessedly they occupy is not a part of the land so described. In other words, the only question involved is one of description and boundary.

Plaintiff's title rests on a patent from the United States dated March 20, 1885, conveying 'lot numbered seven of section twenty-three, and the lots numbered one and two of section twenty-six, in township twenty-nine south, of range thirty-eight east of Tallathassee meridian, in Florida, containing one hundred and seventy acres and forty-two hundredths of an acre, according to the official plat of the survey of the said lands returned to the general land office by the surveyor general.' The official plat of township 29 was in evidence, which showed that sections 23 and 26 were fractional sections bordering on the Indian river. On this plat a meander line runs through the sections from north to south, the Indian river being on the west thereof. The east line of the sections is, so far as these lots are concerned, the ordinary straight line of government surveys. In the S. 1/2 of the S. E. 1/4 of section 23 is lot 7. The area of that lot is given as 73.06 acres. The N. E. 1/4 of section 26 is divided into lots 1 and 2. The area of lot 1 is 54.90 acres; and of lot 2, 42.53 acres. The boundary lines of these three lots are all straight, with the exception of the meander line on the west. The length of the section line between lot 7 and lot 1, extending from the east section line to the meander line on the west, is stated to be 30.55 chains. Along the course of this meander line, as shown on the plat, runs, according to the testimony, a bayou or savannah opening into Indian river; and west of this bayou, and between it and the main waters of the river, is a body of land extending in width a distance of a mile, or a mile and a quarter, and amounting to some 600 acres. This is a body of low land, in some places, however, from four to six feet above the level of the river, and covered with a growth of live oak trees, many of them three and four feet in diameter. It was not land formed by accretion since the survey.

The contention of the plaintiff is that inasmuch as this body of land is not shown upon the official plat, and although the boundaries and areas of the three lots are given, the latter aggregating only 170 acres, the patent for the lots conveys all the land to the main body of the river. In other words, a patent for 170 acres conveys over 700. The basis of this contention is the familiar rule that a meander line is not a line of boundary, and that a patent for a tract of land bordering on a river conveys the land, not simply to the meander line, but to the water line, and hence, as claimed in this case, carries it to the water line of the main body of the river. The testimony is apparently not all in the record, nor are all the instructions, but this presents the ruling of the court: 'It is the rule that the meander line is not the boundary line. They are run, not as boundaries of the tract, but for the purpose of finding the sinuosities of the bank of the stream. Fractional divisions, made so by the water, are designated and sold by the numbers attached to, and reference is always had to, the notes and maps of the survey. The water, in the notes, is the boundary, and when there exists a difference between the meander line, as run, and the actual margin of the stream or lake, the water is the true boundary; but the rule has its limitations, as, for instance, a case in Polk county, with which I am familiar, where there is fifteen miles intervening between the meander line and the margin of a lake. This breaks the rule, and I charge you that when, as in this case, there is from three-fourths of a mile to a mile and a quarter between the meander line and the actual margin of the river, and when for half a mile in width this land has upon it oak trees, some of which are from three to four feet in diameter, especially where the waters of the river make up, forming a bayou which conforms substantially to the meander line of the government survey, this is not within the rule.'

Whatever criticisms may be placed upon this instruction, we think that, as applied to the facts of this case, the ruling of the court was substantially correct. It is undoubtedly true that official surveys are not open to collateral attack in an action at law. Stoneroad v. Stoneroad, 15 Sup. Ct. 822; Russell v. Land-Grant Co., 15 Sup. Ct. 827. It is also true that the meander line is not a line of boundary, but one designed to point out the sinuosities of the bank of the stream, and as a means of ascertaining the quantity of land in the fraction which is to be paid for by the purchaser. Railroad Co. v. Schurmeir, 7 Wall. 272; Hardin v. Jordan, 140 U. S. 371, 380, 11 Sup. Ct. 808, 838. It is also true that metes and bounds in the description of premises control distances and quantities, when there is any inconsistency between them. Morrow v. Whitney, 95 U. S. 551, 555.

But the question in this case is whether the boundary of these lots is the bayou, or the main body of the river. That a water line runs along the course of the meander line cannot, of course, in the face of the plat and survey, be questioned; but that the meander line of the plat is the water line of the bayou, rather than that of the main body of the river, is evident from these facts. In the first place, the area of the lots is given, and, when that area is stated to be 170 acres, it is obvious that no survey was intended of over 700 acres. In the second place, the meander line, as shown on the plat, is, so far as these lots are concerned, wholly within the east half of sections 23 and 26, while the water line of the main body of the river is a mile or a mile and a quarter west thereof, in sections 22 and 27. Again, the distance from the east line of the section to the meander line is given, which is less than a quarter of a mile, while the distance from such east line to the main body of the river must be in the neighborhood of a mile and a half. Further, the description in the patent is of certain lots in sections 23 and 26, and manifestly that was not intended to include land in sections 22 and 27.

These considerations are conclusive that the water line which was surveyed, and made the boundary of the lots, was the water line of the bayou or savannah, and there has been simply an omission to make any survey of the tract west of the bayou, and between it and the main body of the Indian river. It is unnecessary to speculate why it was that it was not surveyed. It may have been a mere oversight, or it may have been because the surveyors thought that the action of the water would soon wash the low land away; but, whatever the reason, the fact is obvious that no survey was made of that body of land, and the boundary line fixed was the water line of the bayou.

The rule of public surveys, as prescribed by chapter 9, tit. 32, Rev. St. (page 438, and following pages), requires that they be surveyed into townships of six miles square, with subsequent subdivisions into 36 sections of a mile square, except where the line of an Indian reservation or of tracts of land theretofore surveyed or patented, or the course of navigable rivers, renders this impracticable, with a proviso that 'in that case this rule must be departed from no further than such particular circumstances require.' Now, if this tract west of the bayou, and between it and the Indian river, was intended to be surveyed, obviously all the lines of sections 23 and 26 would have been run along straight lines, and so as to make complete sections and quarter sections. But such lines—at least, those on the west side were not run, and, whatever the reason, the survey stopped at the water line of the bayou, and left this body of land west thereof wholly unsurveyed.

Although it was unsurveyed, it does not follow that a patent for the surveyed tract adjoining carries with it the land which, perhaps, ought to have been, but which was not in fact, surveyed. The patent conveys only the land which is surveyed, and, when it is clear from the plat and the surveys that the tract surveyed terminated at a particular body of water, the patent carries no land beyond it. Cases of this nature are naturally few in number. Lammers v. Nissen, 4 Neb. 245, is somewhat in point. In that case it appeared that between the meander line, as run, and the Missouri river, was a tract of several hundred acres; and the court held that, as that body of land had not been surveyed, it did not pass by a patent of a lot which, on the government plat, extended to the meander line. A similar ruling was made in Glenn v. Jeffrey, 75 Iowa, 20, 39 N. W. 160. Whitney v. Lumber Co., 78 Wis. 240, 47 N. W. 425, was a case in which the meander line shown in government surveys was half a mile or more from the real borders of a lake; and the court, in a very careful opinion, discusses the law of official surveys, and holds that as the meander line was a mistake the patent did not carry the land to the actual boundary of the lake, but only to the straight line which would have been the boundary of the quarter section if accurately surveyed. And the same doctrine is reaffirmed in Lally v. Rossman, 82 Wis. 147, 51 N. W. 1132.

But it is said that, because the water mentioned on the plat is called 'Indian River,' the boundary must be taken as the water line of the river, and cannot be that of any intermediate bayou. Bates v. Railroad Co., 1 Black, 204, is instructive upon this. In that case a patent had been granted for 102.29 acres lying north of the Chicago river, bounded by it on the south, and by Lake Michigan on the east. The contention was that the main channel of the river entered the lake much below the line shown on the plat, and so the patent carried a larger tract than that described therein. It appeared that there were two channels of the river, and the court said, in reference to this:

'The mouth of the river, being found, establishes the southeast corner of the tract. The plat of the survey, and a call for the mouth of the river in the field notes, show that the survey made in 1821 recognized the entrance of the river into the lake through the sand bar in an almost direct line easterly, disregarding the channel west of the sand bar, where the river most usually flowed before the piers were erected. It is immaterial where the most usual mouth of the river was in 1821; nor whether this northern mouth was occasional, or the flow of the water only temporary at particular times, and this flow produced to some extent by artificial means, by a cut through the bar, leaving the water to wash out an enlarged channel in seasons of freshets. The public had the option to declare the true mouth of the river, for the purposes of a survey and sale of the public land.'

So, in the case before us, obviously, the surveyors surveyed only to this bayou, and called that the river. The plaintiff has no right to challenge the correctness of their action, or claim that the bayou was not Indian river, or a proper water line upon which to bound the lots.

We are of the opinion, therefore, that no substantial error was committed by the circuit court, and the judgment is affirmed.