delivered the opinion of the court.

On the trial of this case, the plaintiffs counsel objected to the introduction of testimony to limit the operation of a deed to a tract of land. The descriptive words in the deed are as follows: « All that tract, or parcel of land, containing 240 acres, situate, lying and being in the county of Wilkinson, on the Bayou Sarah waters, bounded on the north by lands of Reuben Dunham; east by lands of said Patriot Foley, south by lands of said John F. Carmichael, and west by the west branch of Bayou Sarah Creek.”

For the plaintiff it is contended that he is entitled to the 240 acres of land, and that that quantity is as consistent with the boundaries called for by the deed, as is less; but the defendant *592wishes to confine the operation of the grant to a tract of land containing 155 acres, previously purchased by Foley from Reukin Jelks, and for that purpose introduced on the trial the map of the original survey, and Spanish' patent to Jelks, together with Jelks’s petition to the Spanish government on which the grant was made. Declarations of Carmichael were also proved, stating, “ that at the time the deed was executed, he supposed the Jelks tract contained 240 acres, and did not ascertain to the contrary, until a survey was made 'subsequent to his taking the deed.’7 The introduction of this testimony was the ground of the bill of exceptions, and presents the principal question to be determined. If there is no latent ambiguity in the deed, the evidence was improperly admitted.

Latent ambiguities relate to the subject of the deed, and can only be made appear by evidence of something extrinsic arising in the application of an instrument of clear and definite intrinsic meaning to doubtful subject matter, and although they can only be raised by extrinsic evidence, either parol or written, yet they must be clearly shown before they become the subjects of explanation. We must therefore in the first place inquire whether the evidence given raised a latent ambiguity. A comparison of the boundaries and description, called for by- the deed, with those shown by the evidence, becomes necessary.

By an examination of the survey taken, it seems difficult to decide this case without a violation of some principle, whether we regard the. deed as conclusive, or admit the evidence ip^ex-planation. In the’ descriptions given in deeds, it is a rule that what is most certain and material shall control that which is less certain and less material. 5 Cowen, 373. And fixed monuments control both course and distance. 7 Wheaton, 10; 6 Mass. Rep. 33. When land is definitely described by boundaries, and the quantity of acres is given, it will be considered merely as a part of the description, and being less certain than the boundaries, must be controlled by them. 5 Mass. Rep. 357; 2 Johns. Rep. 37. At least this appears to be the rule, where the quantity.is not the subject of a positive covenant. If therefore the boundaries in this instance had been so .accurately described as to leave no *593doubt, the quantity of acres mentioned would only be deemed a false or mistaken circumstance not essential to the operation of the grant. But on actual survey, the boundaries do not correspond with the description. According to the rules laid down, Carmichael would be entitled to all the land bounded on the north by Dunham, unless there is some other part of the description that limits the extent of his boundary, and there is nothing unless it be the quantity of acres.

He has the same right to contend for the whole length of this boundary that the defendant has to contend that he is limited by the southern boundary, which is much shorter than the northern, and much too short to give the quantity of acres called for. The call on the north, or Dunham’s line, is much longer than the call on the south. To run the whole length of the northern boundary as called for by the deed, would include considerably more than 240 acres, by extending the southern boundary to an equal length; but if the northern boundary is to be no longer than the line of Carmichael on the south, then only 150 acres would be included in the survey. Taking either position it would still be bounded on the east by the lands of Foley. The boundaries as called for, therefore, make but an imperfect description. By taking the quantity of acres the southern boundary or line is extended east beyond Carmichael’s line, and on the north it stops short of the termination of Dunham’s line, and thus varies from the boundaries in the deed. This shows most manifestly that there was no actual survey of the land before sale, or the parties would have been more accurate in the description. But would parol evidence remove this difficulty? It would not. By admitting it two parts of the description given in the deed would be rejected, to wit, all the northern boundary east of Jelks’s tract, and the quantity of acres, and I am not aware that parol evidence is admissible to explain an ambiguity where the subject matter does not meet the description given in the deed in the most material points. The most usual instances given in the books of latent' ambiguities which admit of explanation are, where there are two subjects or objects which answer the description; as if a person grant his manor of S. generally, and has two manors of that name, parol *594evidence is admissible to show which was intended. Or when a devisee is uncertain, in consequence of there being two of the same name, parol evidence is admissible to show which was intended. And it is also true, that where the description applies partially to two objects, but to neither entirely, so that a doubt arises what was meant, parol evidence may be received, and the same rule seems to prevail as to the extent of the matter sold; 3 Stark. Ev. 1025-6; but I apprehend that this rule cannot be extended so far as to defeat a description which is more applicable and material, by adopting one which is neither so applicable nor material to the subject matter of the deed. As in this instance, if parol evidence were admissible it would defeat the call for the quantity and part of the northern boundary, and retain no part of the description complete except the southern boundary; the eastern boundary being equally applicable either way. It is held to be a good rule, that parol evidence should not be admitted when the deed can have an effective operation without it; Phil. Ev. 469; and in this instance I think there is no difficulty in giving effective operation to the deed without extrinsic aid. Although it may be true in general that boundary should control quantity, yet quantity is equally a part of the description, and when it is more certain and more material than boundary, and appears from the deed to have been the leading object of the grant, I see no good reason why it should not be retained as a material part of the description, especially as the boundaries are irreconcilable. The creek on the west is a natural object which would govern every thing else, and 240 acres east of it was equally as consistent with the grantor’s power to convey as a less quantity, and that number of acres is made the leading member in the description of the tract of land. He sold “ all that tract of land containing two hundred and forty acres:” the description by boundary follows the call for quantity, and it is remarkable that if no more than the Jelks tract was intended to be conveyed, the deed did not call for it by name, or that declarations had not been made specifying that tract as the subject of the grant. Such declarations would have been much better than the testimony introduced, as in that case it would have appeared distinctly what was sold; but *595the evidence introduced is by no means certain in its character. By receiving it the matter is left entirely to speculation, it being nothing more than a circumstance tending to create a probability that Foley did not intend to convey more than the Jelks tract, that being more consistent with .his interest.

There is nothing in the evidence, setting aside the supposed interest of Foley, which would- lead certainly to the conclusion that the Jelks tract was the subject matter of contract. If boundary is to control quantity, I am not entirely sure that Carmichael is not entitled to the full extent of the known boundaries called for by the deed, running as far as it is bounded on the north by Dunham’s land. If that was a known and fixed line, I see nothing, according to the authorities, which could restrict it, unless it be the quantity. Deeds are invariably to be construed most strongly against the grantor, and Í conceive that this rule of law must have its weight. I think, therefore, that the testimony does not satisfactorily raise a latent ambiguity, and of course there was none to be explained, and that it was improperly admitted.

The judgment must be reversed and a venire de novo awarded.