The only question for decision is whether or not the description of land in a certain deed is void absolutely, for uncertainty. The description is in words and figures as follows:
“Twenty-five acres more or less in fractional sections 21 and 22, T. 19, R. 18, Elmore county, Alabama, on the West side of the Coosa river, bounded on the south as follmvs: By Cohn & Goldberg; on the north by lands OAvned by the state of Alabama; on the west by the undersigned, the west line to be established by survey, bounded on the east by the Coosa river.”
The geometrical proposition involved in the description in the deed in question, abstractly speaking, is this: Given, the length and direction of one side of a quadrangle containing “25 acres more or less,” and the direction of tAvo other sides which are parallel, each to the other, *547construct the fourth side so that one of the parallel lines shall not exceed 396 feet, and the other shall not exceed one mile in length.
This we think can be easily done. In fact, we think an indefinite number of figures may be constructed, any one of which will equally meet the requirements of the proposition.
This is the second appeal in this case. The report of the former appeal may be found in 177 Ala. 85, 58 South, at page 417. The concrete question involved on the first appeal, and the only question involved on this appeal, is whether or not the deed under which appellant claims title was void for uncertainty of description of the lands attempted to be conveyed. It was decided on the former appeal that the deed was void. An application for a rehearing was made, on the hearing of which another full consideration of the case was had by the entire court; each judge giving it a thorough consideration, and the unanimous conclusion of the court, upon the application, was the same as that reached on the original hearing. The case has been twice fully and ably argued orally, as well as in a number of written and printed briefs, by able and eminent counsel for appellant, and, as strange as it may appear to counsel, we are yet of the opinion that the deed in question-is absolutely void on account of uncertainty in the description of the lands attempted to be conveyed.
But for the ability, the learning, and the eminence of counsel who represents the appellant on this appeal, and the earnestness with which he now insists that we were in error in the opinion and decision on the former appeal, we would merely affirm, upon the opinion rendered on the former appeal.
The opinion on the former appeal did not answer all the argument made on this appeal, for the all-sufficient *548reason that some of the argument now made was not advanced on the first appeal, but a different argument was then made and a different reason assigned, by counsel, to show why the deed was not void. We do think, however, that the opinion did sufficiently answer the argument made on that appeal.
The body of land described in the former complaint, which was then alleged'and claimed to be the land described in the deed, is entirely different in area, and in every dimension except its eastern boundary, from that described in the present complaint, and now insisted to be the land described in the deed. The description in the present complaint is materially, if not entirely, changed. Indeed, no one reading the two complaints would ever suspect that they were intended to describe the same body of land, or that either described the identical body of land attempted to be described in the deed.
But, as was said in the opinion on the former appeal, the description in the deed is so uncertain that any number of bodies of land could be carved out of the fractions of sections 21 and 22, any one of which would, with equal certainty, answer to every call in the deed. It was for this reason that the deed was then, and is now, held to be void.
It appears from an inspection of the records on the two appeals that a number of surveyors have attempted to locate the land described in the deed, and that no two fix upon the same body. It also appears that the same surveyor, who made the surveys for the appellant, located different bodies of land on the two appeals. There were exhibited, on both of the trials in the lower court, maps, made by the appellant’s surveyor, designed and intended to locate and describe the same lands described in the deed; and the two bodies of land located by these ■ maps are greatly, if not entirely, different. It is true *549that each contains some land embraced in the other, bnt each contains some land which the other did not.
These maps, together with the original and the amended complaints, the reporter will set ont, as near as may be, in the report of this decision.
Appellant has had a number of counsel engaged in the trial of this case, and all of his counsel have not agreed as to the particular body of land described in the deed; nor has each attorney always had the same opinion as to the exact body of land described in the deed. Counsel for appellant on this appeal attaches to his brief a diagram and plat of the body of land which he now claims is described in the deed, which is practically the same as that made by the engineer of surveyor, and attached to the transcript as an exhibit. The same counsel, on his application for a rehearing on the other appeal, not content with the diagram and plat made by the surveyor and attached to the transcript as an exhibit, made a map and plat entirely different from either of those made by the surveyor and attached the same to his brief. And this diagram is very different from the one he attaches to his brief on this hearing, as well as from the first maps and plats made by the engineer, contended by other counsel to correctly locate the land described in the deed.
The reporter will set out, as near as may be, a copy of the map or plat contained in counsel’s brief on his said application for a rehearing.
In the original brief filed on the first appeal in this case (page 8 of the printed copy), in attempting to distinguish this case from cases relied on by opposing counsel, and to describe the body of-land sued for, and described in the deed in question, counsel for appellant said:
*550“In the case at bar the Coosa river is the east boundary line and the only line not definitely fixed by the evidence is the west line. Under the decisions above cited, the west line must be a line exactly parallel with the Coosa river following its meanders, and located at such a distance as to include twenty-five acres.”
It should be said that this was different counsel from the one who now insists that the west line is a straight line, without any regard to the meanders of the river. But while it is true that counsel, who now claims that the west line is a straight line, was counsel on the application for a rehearing and made the same claim, yet the straight line contended for on the application is not the same straight line he now insists upon. The two lines connected at different points, were of different lengths, and established different areas of land.
On the first appeal, this same counsel contended that the deed conveyed 25 acres exactly, if that area was to be had in fractions of sections 21 and 22, and there has never been any dispute that these fractions contain much more than 25 acres.
On the first appeal this same counsel contended that the phrase “more or less,” contained in the deed, was meaningless if 25 acres could be found in fractions 21 and 22, north of Cohn & Goldberg’s land and west of the Coosa river. Counsel in his brief in the application for a rehearing (pages 9 and 10), among other things, said:
“Now, here, the first sentence in the deed shows a clear contract to sell and convey by the acre at the price of fifty dollars per acre, and there is no lump sum to support á sale in gross. Therefore the sale is one by the acre.
“Can we construct the plat on that theory? We have shown that we can. Then the words 'more or less’ are senseless, and must be rejected or held to have been put *551in as precautionary words, to prevent the grantor from being held liable for the precise number of acres if not to be found in the plat as constructed.”
“Here the land was to be measured out of the grantor’s land, and therefore 25 acres ‘more or less’ meant 25 acres absolutely, if that quantity was in the sections.
“And here the gross sum to be paid was $50 per acre, which could only mean $50 multiplied by 25 acres; there being no other process of obtaining an aggregate, therefore ‘more or less’ must be rejected or applied as suggested above to cover a possible deficiency of land in the section 21.
“Again, the western line is to be surveyed, which could only be done by talcing 25 acres absolutely if in the section. And still again ‘other parts of the deed’ confirm this view, for the first sentence expressly says the whole consideration is $5 paid and $50 per acre to be paid ‘for 25 acres,’ and not for 25 acres ‘more or less,’ which Avould be nonsense.
“All this shows that ‘more or less’ must be rejected or held to refer to a possible deficiency of land in the section to fill out the call for 25 acres.”
Counsel noxv, not only concedes, but contends, that effect must be given to the phrase “more or less,” and that the quantity conveyed was not 25 acres absolutely, but Avas considerably less. Exactly hoAV much less than 25 acres is yet left in uncertainty, from the pleadings, the argument, and the evidence, as we shall show.
The complainant claims 20.06 acres, and the argument in part contends for that amount, but in the complaint he describes a body of land Avhich is by the evidence shown to contain much less than 20 acres.
The diagram introduced by appellant and attached as an exhibit to the transcript, and a copy of which is attached to appellant’s brief, shows a parallelogram, A, *552B, C, D, the opposite sides of Avhich are 8,020 and 390 feet, respectively. This parallelogram, of course, contains 20.06 acres, which is the amount of land claimed in the complaint and contended for in argument, and which appellant seeks to be put in possession of, by the judgment and process of the courts; but this same diagram, and the other evidence, shows that 4.44 acres of that 20.06 acres lies east of the water’s edge, and under the waters of the Coosa river, which, of course, there Avas no attempt or intention to convey. The diagram also shows a strip of 1.59 acres east of the eastern line of the parallelogram, but west of the Avater’s edge of the Coosa river. The meanders described in the complaint are not those of the thread of the river, of its water’s edge, nor of the river generally, as referred to in the deed; but are those of the “banks” of the river. If this bank should be held to be the high-water mark, and not the low-water mark, then the complaint, as explained by the evidence and the diagram, would describe much less than 20.06 acres as claimed; it would, under such construction and diagram, contain less than 15 acres. The meaning of the word “bank,” when used in describing boundaries on navigable and nonnavigable streams, and the difference between the meanings of the Avords, “bank,” “stream,” “river,” and “waters,” when used in the description of boundaries, may be found in Words & Phrases (volume 1, p. 689), under the heading “Banks;” but it is not necessary to a decision of this case for us to define the word “bank,” for the word is not used in the description contained in the deed, but only in that contained in the complaint.
We are unable to understand, even by the aid of the argument of counsel, how it is that the description in this deed authorizes the construction of a parallelogram, A, B, C, D, on the base A, B, which is 390 feet, *553from which the land sold and conveyed is to be selected. We must confess that we are unable to do so, though,' as counsel argues, “schoolboys” would have no difficulty in locating the land sold and conveyed, from the description in the deed, and though we are shown, in minutest detail, how the schoolboys Avould proceed to construct the parallelogram and to thereby locate 20.06 acres of land which would answer to every call in the deed. In the first place, we cannot understand why we should arbitrarily adopt the base, A, B, 390 feet, upon Avhich to construct the parallelogram, if a parallelogram must be constructed; notAAdthstanding counsel tells us it is because this is the parallelogram which comes nearest to containing 25 acres, within the calls of the deed. The deed says that the land conveyed is bounded on the south by Cohn & Goldberg — meaning, of course, Cohn & Goldberg’s land. The northern boundary of Cohn & Goldberg’s land, referred to as the southern boundary of the land in question, is shown to be 396 feet in length, by actual measurement — at least, the deed from Holt to Cohn & Goldberg describes it as having been actually measured, and as being six chains in length. Now, when Ave Avere schoolboys and studied surveying, a surveyor’s or “Gunter’s” chain (we infer this is the kind of chain meant) was 66 feet in length; and according to the multiplication table in use at that time 66X6=396, and not 390. If the length of this chain, or the multiplication table, has ever been changed, we have no judicial or other knowledge of the alteration. And we cannot accept or adopt the argument of counsel, predicated on a change in either- of these standards, notwithstanding we recognize the great ability and learning of counsel making the argument, and notwithstanding he bears the same name as the honored inventor of the chain. We likewise cannot understand how it is that this particular *554parallelogram, A, B, C, D, constructed on the base A, B, which is 390 feet, and its other side 3,020 feet, of all possible parallelograms comes nearest to giving the exact area of 25 acres, the calls of which shall correspond to those of the deed. If this particular parallelogram contains only 20.06 acres, a similar parallelogram, constructed on a base of 396 feet, the full length of the Cohn & Goldberg line, would certainly come nearer to containing 25 acres than does the one actually constructed by counsel. Such figure similarly constructed on the base of 396 feet would also contain less than 25 acres, but more than 20.06 acres, and therefore nearer 25 acres.
We likewise cannot understand the necessity or the propriety of constructing any parallelogram, from which the land in question is to be taken, if all the land in the parallelogram is not to be taken or is not described in the deed. If the land described is described so as to show that it forms a parallelogram, a square, a rectangle, or a circle, we can then understand the necessity and the propriety of constructing, marking out, and ascertaining the boundaries of such figure, whatever it may be; but we know of no rule of law, of mathematics, or of surveying, which authorizes, much less requires, the construction of a parallelogram from which must be selected a part or all of another and different quantity <of land described and' shown not to be a parallelogram, but to be a figure one or more sides of which are made up of crooked or broken lines.
We cannot apprehend why we should not, just as well and just as appropriately, construct a square or a circle, including all or any part of the land mentioned in the deed, from which the lands conveyed should be selected, as construct the parallelogram constructed and contended for by counsel in this case.
*555The parallelogram, A, B, C, D, is not claimed to inclose the exact land conveyed, bnt tbe claim is that it incloses the quantity of land conveyed by tbe deed. Tbe diagram shows, on its face that it contains some land not conveyed, and that some of tbe land conveyed lies outside of it, and to tbe east of it. How, therefore, can it be said to contain tbe quantity conveyed?
Tbe deed says in terms that “tbe west line is to be established by a survey.” If counsel’s theory is correct, there would be no necessity to establish this west line by a survey, as tbe one be claims is merely a straight line between tbe points A and D.
As before stated, it was first contended in this court, though not by counsel now appearing, that tbe west line bad to be surveyed, and was a crooked, or broken line, corresponding to tbe east boundary which followed tbe meanders of tbe river. While we could not agree to that contention, we think it more tenable than tbe one now presented.
If anything is made certain by tbe description in tbe deed, it is that tbe western boundary of tbe land intended to be conveyed was not then known, but was “to- be established by a survey,” and was to be a crooked or broken line; yet, under counsel’s theory, this is tbe one line or boundary made definite, fixed, and certain, by tbe other calls in tbe deed, and tbe only one that would not have to be measured to know its length.
We cannot -ratify the theory and contention of counsel for appellant, and tbe judgment of tbe lower court must be affirmed.
Affirmed.
All tbe Justices concur, except' de Graffenried, J., who dissents.