The opinion of the court was delivered by
Depue, J.The controversy in this case relates to the equal .undivided one-fourth part of the fourteen and seventy-three hundredths acres, which the testator purchased of Lippencott. It is conceded that the premises in question passed under the residuary devise, if they were not specifically devised to 'Thomas by the first paragraph of the fourth subdivision of the will; and that, if the fourteen and seventy-three hundredths .acres passed under the residuary devise to the testator’s four children, the deed from Samuel conveyed the one-fourth part .thereof to the plaintiff.
The question, therefore, is whether the fourteen and seventy-three hundredths acres were devised in entirety to Thomas by •the first paragraph of the fourth subdivision, or whether only the equal undivided fourth part passed to him under the last paragraph of that subdivision.
*406Thomas claims the whole estate in these lands, under the description .of “ all that my farm and plantation near Crop-well, conveyed to me by the heirs of my deceased wife, and where my son, Thomas Evens, now resides, containing about eighty-five acres, more or less.” Thomas resided in the mansion house on the parcel which had been conveyed to the testator by his children as the heirs of his deceased wife, but used and cultivated both parcels, and had done so from the year-1850 down to the testator’s death.
The problem to be solved is, whether, under the circumstances, the fourteen and seventy-three hundredths acres, which the testator purchased of Lippencott, passed to Thomas under the description last mentioned.
It being proved that these two parcels of land had been rented and occupied together as one tract since the year 1833, the judge received the testimony of the scrivener who wrote the will, that the testator, on the occasion of drawing the will, camelo his house “with items on a piece of paper for each son;” that he had these premises marked down on it as “ my Cropwell farm, containing eighty-five acres;” and that the words “ conveyed to me by the heirs of my deceased wife,” were not on that paper, but were inserted in the will by the scrivener as his own language, which he used as an additional description, to distinguish the premises from the testator’s other property. The import of this testimony was to expunge from the will words which the law regards as the language of the testator, and to alter the terms of the devise. This evidence was illegal.
By the statute, a writing is made indispensable to the existence of a will, and what has been Avritten cannot be added to, detracted from, or altered by extrinsic evidence. The functions of the court are to ascertain the intention of the testator from the language of the will. Extrinsic evidence may be resorted to, which, in its nature and effect, is simply explanatory of what the testator has written, but no eAÚdence can be received for the purpose of showing what he intended to have written. Wigram on Wills, § 9. In every case of a contro*407verted construction, the sole question is non quod voluit sed quod dixit. Extrinsic evidence of the circumstances, situation and ^surroundings of the testator, and of his property, is legitimate to place the court which expounds the will, in the situation of the testator who made it, and thus enable the court to understand the meaning and application of the language he has adopted; but the testator’s intention must ultimately be determined from the language of the instrument, as explained by such extrinsic evidence, and no proof, however conclusive in its nature, can be admitted with a view of setting up an intention inconsistent with the writing itself. 2 Taylor on Ev., § 1082. No difficulty, however great, in deciphering the obscure language of the devise, or in unraveling the intricacies in the testator’s descriptions of the person or property to which his testamentary disposition should apply, will justify resort to such evidence. With the assistance of such explanatory evidence as has been mentioned, the court may be called upon to harmonize the conflicting provisions of the will, and to reject such descriptive parts as are only false demonstrations; but if, with the aid of such evidence, the testator’s meaning cannot be ascertained, the will will be simply void for uncertainty. Wigram on Wills, Prop. VI.
The only exception to this legal rule is that the declarations of the testator may be resorted to in cases of a latent ambiguity, which arises where there are two or more persons or things, each answering exactly to the person or thing described in the will. In such an event, parol evidence of what the testator said may be lawfully adduced to show which of them he intended; but such evidence will not be allowed to show that he meant a thing different from that disclosed in the will. Den v. Cubberly, 7 Halst. 308; 2 Taylor on Ev., §§ 1092, 1093; Hawkins on Wills 9.
The testimony of the scrivener, with regard to his instructions for the will, was not admissible within this exception. The description of the premises intended was not applicable to the two parcels of land devised by the testator under the several clauses of his will. It may be uncertain or inaccurate, *408as applied to either. But an uncertainty arising from mere inaccuracy of description is not such an ambiguity as authorizes the admission of evidence of intention. The difficulty is to be solved by construction—by the rejection of such of the words of description as appear to be surplusage, or a mere false demonstration. Wigram on Wills, § 203.
The cases on this subject so prolific of judicial decision, show how inflexible the foregoing rule of construction is regarded. In Den v. Cubberly, 7 Halst. 308, the testator devised to the defendant a lot of land, beginning in Joseph Wall’s line, at the corner between him and R. Chambers; thence “as the line runs ” till it comes to the middle of the road, &c. The testimony of the scrivener was offered to show that the testator intended the line between Wall and Chambers, and in his instructions for the will designated the spot where the line would strike the road. This evidence was excluded, and it was held that the question was simply one of construction to ascertain the meaning of the testator from the language he employed. In Cleaveland v. Haven, 2 Beas. 101, evidence that the will was not drawn in pursuance of the written instructions given by the testatrix, and that both the scrivener and the testatrix understood the will differently from its legal import as deduced from its language, was held to be inadmissible ; and in Nevius v. Martin, 1 Vroom 465, testimony of the scrivener, that by the instructions given to him by the testator at the time of writing the will, he was directed to devise the lot in question in fee, and that he believed such intention was expressed by the language of the will, was rejected as incompetent. In Conover v. Wardell, 7 C. E. Green 492, extrinsic evidence of the meaning of descriptive words and names used in the executed papers of the parties, was held inadmissible by the Court of Errors and Appeals. These cases settle the law in this state. They are in full accord with the English cases. Miller v. Travers, 8 Bing. 244, is a leading case on the subject. It was a case in chancery, heard before the Chief Justice of the Court of Common Pleas, the Chief Baron and the Lord Chancellor. The devise *409was of “all the testator’s freehold and real estate in the'county of Limerick and city of Limerick.” The testator had no lands in the county of Limerick, and only a small estate in the ‘city of Limerick, which was inadequate to meet the charges in the will; but he had lands in the county of Clare, not mentioned in the will. It was proposed to be shown in evidence that in a copy of the will submitted to the testator and approved by him, the devise in question stood thus: “ All my freehold and real estate in the counties of Clare, Limerick, and in the city of Limerick,” and that the scrivener did, by mistake, and without any authority, strike out the words “ counties of Clare,” and substitute the words “ county of ” in lieu thereof, and that the will was executed by the testator without adverting to this alteration. It was admitted that the evidence, if competent, would have been sufficient to establish beyond contradiction the intention of the testator to include his estates in Clare in the devise. But the court excluded the evidence as inadmissible, and gave effect to the will according to the language used in it. In Bernasconi v. Atkinson, 10 Hare 345, the bequest was to the testator’s first cousin, Vincent B., the son of his late uncle, Peter B. The testator had no cousin named Vincent B., who was the son of his late uncle, Peter B.; but he had a first cousin George Vincent B., who was commonly called Vincent, and who was the son of a deceased uncle, named Joseph. The evidence of the solicitor who drew the will, that the testator had, by his instructions, expressly indicated that George Vincent B. was the person for whom the legacy was intended, and that he (the solicitor,) had inserted the description without the direction of the testator, and in a mistaken belief of the parentage of the legatee, was declared to be inadmissible. Other cases holding adversely to the admissibility of such evidence, except in cases of a latent ambiguity as above defined, will be found in the following citations: Doe v. His-cocks, 5 M. & W. 362; Doe v. Hubbard, 15 Q. B. 227; Webber v. Stanley, 16 C. B. (N. S.) 698; Drake v. Drake, 8 H. of L. Cas. 172; Charter v. Charter, Law Rep., 7 H. of L. Cas. 364.
*410The judge also admitted evidence that after the purchase of the fourteen and seventy-three hundredths acres, the two parcels were treated by the testator as one tract, and were rented and occupied, and cultivated as such. The tendency of this proof was to show that by the words my farm, the testator meant the whole body of his land near Cropwell.
Testimony of this character is admitted or rejected, according to the circumstances of the particular case. It was excluded in Doe v. Oxenden, 3 Taunt. 147; S. C. on Error, 4 Dow 65, and in Hand v. Hoffman, 3 Halst. 71; and was admitted in Webber v. Stanley, 16 C. B. (N. S.) 698, and in Ricketts v. Turquand, 1 H. of L. Cas. 472. In the first two cases the evidence was excluded for the reason that its effect was to enlarge or abridge the natural import of the language used by the testator. In Webber v. Stanley, the testimony was received to show what the testator meant by my Ted-worth estate,” and in Ricketts v. Turquand, to ascertain what was intended by “ all my estate in Shropshire, called Ashford Hall.” Goodtitle v. Southern, 1 M. & S. 299, and other cases hereafter cited are in this latter class of cases.
It is sometimes said that this sort of evidence is admitted only when there is a latent ambiguity. The use of the expression latent ambiguity, in this connection, has given rise to some confusion in the cases. The office of such testimony is that of interpretation merely, to ascertain the sense in which words have been used by the testator. Where in a deed, will or written contract, general words of indefinite signification are used, and there is nothing on the face of the instrument to qualify them, or limit and apply them to a particular subject matter, evidence of extrinsic circumstances—matters of fact as distinguished from mere declarations of intention—is admissible for the purpose of ascertaining in what sense such indefinite language was used. The effect of such evidence is not to vary the language employed, but merely to explain the sense in which the writer understood it. 2 Taylor on Ev., § 1085. This whole subject has been so fully discussed in the opinions of the judges in Shore v. Wilson, 9 Cl. & Fin. 355, *411that an extended citation of the authorities would be superfluous. An instance of the admission of such evidence, under the circumstances above mentioned, will be found in Sandford v. N. & H. R. R. Co., 8 Vroom 1.
Though evidence of extrinsic circumstances be admitted to ascertain the meaning of vague and general terms, the intention of the testator must ultimately be determined by the language o'f the entire instrument, after the sense in which such general language was used has been ascertained by extrinsic proof. The office of such evidence is that of interpretation—to find out the true sense of the written words as the parties used them; then follows the process of construction, which is performed by subjecting the instrument to the established rules of law for the construction of written instruments. 2 Taylor on Ev., § 1087. When such evidence is once admitted the construction of the devise is still for the court. “ Facts, extrinsic of the will,” says Erle, C. J., in Webber v. Stanley, "must be ascertained for the court in the usual manner, either by the admission of parties,- or by a jury. When they have been ascertained, the operation of construction is to be performed by the court.” 16 C. B. (N. S.) 751.
In Skull v. Glenister, 16 C. B. (N. S.) 102, the same learned judge said: "The construction of a deed or other written instrument is for the court; it is almost always necessary that the court should be informed of the surrounding circumstances under which the deed is executed: if there be any disputes, these must be ascertained by the jury; and when that has been done, the court is to apply the description in the deed.” In the same case, Williams, J., said: " The construction of a deed is for the court, though it is the proper function of the jury to assist the judge as to any controversy of fact which may arise; when the facts are ascertained, it is the duty of the judge to say what is the proper construction and effect of the deed.”
The defendant’s counsel contend that on a construction of the testator’s will in conformity with the above principles, the *412devise in the fourth subdivision comprised the entire tract of land he owned near Cropwell, including the land he purchased of Lippencott, as well as that conveyed to him by the heirs of his deceased wife. This contention is rested on the maxim, falsa demonstratio non nocet. They cite, in support of their argument, Goodtitle v. Southern, 1 M. & S. 299; Down v. Down, 7 Taunt. 343; and Doe v. Earl of Jersey, 1 B. & Ald. 550. These cases were decided upon the peculiar language of the respective-devises. In Goodtitle v. Southern, the devise was of “ all that my farm lands and hereditaments called Trogues-farm, situate, &c., now in the occupation of H. Clay.” The testator was seized of a tract of land which the proof showed he called Trogues-farm. It had been devised to him by the will of one Houghton, by a description as three several parcels, the first of which was called Trogues’, or Trough’s pasture, and the others called Dale’s closes, described as two closes situate at the bottom of the pasture called Trogues’, otherwise Trough’s pasture. The first of these parcels was in the occupation of Clay : the others were not. It appearing, by extrinsic evidence, that the testator treated all these lands as constituting Trogues-farm, the court rejected the additional words, “ in the occupation of H. Clay,” as an erroneous description, and held that all passed under general description of Trogues-farm. In Down v. Down, the testator devised to his wife “ his messuage, farm, and lands, called Coltsfoot farm, situate,, &c., then on lease to one F.” In fact, in the lease to F. of the Coltsfoot farm, seven acres of the farm were excepted thereout. The controversy was as to the title of the seven acres. The court rejected the words of additional description, “ then on lease, &c.,” and held that the entire farm passed. In Doe v. Earl of Jersey, the devise was of “ all my Briton Ferry estate, and all the land, &c., of which it consists.” In a subsequent part of his will, in a devise to another person, the testator mentioned his Briton Ferry estate as situate in the county of Glamorgan. In fact, the Briton Ferry estate was part in the county of Glamorgan and part in the county of Brecon. It *413was held that all the lands which were parcel of the Briton Ferry estate in both counties, passed by the devise. In each of these cases the premises had a known name, and the court, seeing on - the face of the will an intention to give the whole' under the designation of the name, rejected the additional-description as a false demonstration. Indeed the rule may be said to be a general rule that, whenever the testator’s intention to give the whole as an entirety clearly appears from the language of the will, whether such intention is expressed by a designation, by a name, or by abuttals, or other descriptive words, additional words of description which prove "to be only partially true, will be rejected as a mistaken description. Doe v. Galloway, 5 B. & Ad. 43 ; Doe v. Carpenter, 16 Q. B. 181; Drew v. Drew, 8 Foster (N. H.) 489, and cases cited.
But it is an erroneous application of this maxim to suppose that an enumeration of particulars in the description will, in all cases, be overruled by general language in the devise. On the contrary, where there is a clear enumeration of particulars, purporting on their face to be designed as qualifications or restrictions of a preceding general description, words of general devise must yield. West v. Lawday, 11 H. of L. Cas. 375; Drew v. Drew, supra.
The question whether language purporting to be descriptive is a false demonstration, and to be rejected as such, or a true restriction of the more comprehensive words of general devise, is one of construction from the language of the devise, to-ascertain at what period in the descriptive words the description of the premises really intended, is complete, and what phrases or expressions are merely superfluous. Counsel insist that in this case the words “ conveyed to me by the heirs of my deceased wife,” should be rejected. These are important words, capable in themselves of giving certainty to the subject matter of the devise. They follow right after, and are immediately connected with that member of the sentence containing the general words of description, and are part and parcel of the description itself. Great reliance was placed on the words *414“ all that my farm and plantation.” All what farm and plantation ? The testator answers the interrogatory for himself, in the immediately succeeding and connected words, “ conveyed to me by the heirs of my deceased wife.” With such a plain exposition of the intention of the testator in his own language to disregard what he has said, and construe his will on general notions of what with propriety he might have done, would be to substitute conjecture for construction.
The evidence admitted may be sufficient to raise a presumption that the testator intended that the two parcels should be kept together as they had been united for years before; but we are not at liberty to act upon such presumptions in the face of the plain language the testator has adopted. Voluntas stet pro ratione. In Doe v. Lyford, 4 M. & S. 550, the testator Avas seized of a messuage lying in the íavo hamlets of Sutton Wick and Sutton Courtney, which he had purchased of one Lovibond. Sutton Courtney Avas a rarish comprising three hamlets, of Avhich the hamlet of Sutton Courtney was one, and the hamlet of Sutton Wick was another. The testator devised to I. L. A. “ all that his messuage, farm lands, Ac., at Sutton Wick, in the parish of Sutton Courtney, which he purchased of Lovibond.” The court held that the Avords “ at Sutton Wick ” Avere Avords of restriction, and that only that part of the farm Avhich lay in that hamlet passed by the devise, notivithstanding the inconvenience of separating the estate which had been held by unity of title, and let at one entire rent. Doe v. Greathed, 8 East 91, is also a case very much in point. In Slingsby v. Grainger, heard first by Vice Chancellor Stuart, (2 Jurist (N. S.) 276,) and next by the Lord Justices of Appeal (Id. 1176); and finally in the House of Lords (7 H. of L. Cas. 272,) the gift Avas of “ the whole of my fortune iioav standing in the funds,” and it Avas decided in all these courts that the gift carried only that part of the testator’s personal estate Avhich Avas invested in the funds. In the House of Lords, Lord Cramvorth, adverting to the maxim falsa demonstrado non noeet, and speaking of the argument that the words “standing in the funds,” might be *415rejected as a mistaken description, says : “ I certainly should have acceded to that at once, if the expression, ‘ my fortune/ had not been so connected with ‘ now standing in the funds ’ as to make the latter a part of the description of the former.” And, after referring to Goodtitle v. Southern, he continues: “ The distinction is between those cases in which there has been a complete description of the thing given, and a subsequent misdescription as to some particular connected with it, and cases in which that which is subsequently connected with the description is so connected as to' form part of the description of the thing given.” Remarks of a similar import were, made in the same case by Lord Wensleydale. These observations are a correct statement of tlie principle which uniformly guides the court in the application of the maxim falsa demonstrado non noeet.
Another rule of construction equally well settled is, that if there be a subject matter wherein all the demonstrations in a written instrument are true, and another wherein part are true and part false, the words then shall be intended words of true limitation, to pass only those lands wherein all these circumstances are true. This, in effect, is one of Lord Bacon’s maxims. Bac. Max., Reg. XIII. The operation of this rule is to qualify words of a general or more extensive signification by the additional words of limitation. Ex prceeedentibus et eonsequentibus optima fiat interpretado. Thus in Doe v. Bower, 3 B. & Ad. 453, the testator devised “ all my messuages situate at, in, or near a street commonly called Snig Hill, in Sheffield, which I lately purchased of and from His Grace, Charles Duke of Norfolk.” The testator owned six houses in Sheffield which he had purchased of the duke. He had purchased all of them by the same conveyance, and had redeemed the land tax on all by one contract. Four of the houses were about twenty yards from Snig Hill; the other two were about four hundred yards distant, in a place called Gibralter street. The court having concluded that the words “ at, in, or near Snig Hill” were words of restriction, held that only the four houses passed, for the reason that these four answered all the terms *416of the devise, and that' it must be understood that the testator meant to pass these only, although he had purchased all of the six of the Duke of Norfolk. Examples of a similar application of this maxim -will be seen in the following cases: Doe v. Bell, 8 T. R. 579; Doe v. Parkin, 5 Taunt. 321; Griffith v. Penson, 9 Jurist (N. S.) 385; Barton v. Dawes, 10 C. B. 261; Doe v. Meyrick, 2 C. & J. 223 ; Webber v. Stanley, 16 C. B. (N. S.) 698 ; Smith v. Ridgway, L. R., 1 Exch. 331; Homer v. Homer, 8 Ch. Div. 758. It is also a necessary sequence from this rule that where the description consists of several particulars of different degrees of importance, that subject matter will be selected to which those particulars apply which are superior in number or importance rather than that which corresponds with those of a lesser number, or of minor consequence.
Applying the several rules of construction which have been considered, to this will, I think it quite evident that by the language of the will the premises in question did not pass to the defendant under the first paragraph of the devise to him. Separating the description in that part of the devise into its component parts, they are these: (1) all that my farm and plantation near Cropwell, (2) conveyed to me by the heirs of my deceased wife, (3) and where my son Thomas Evens now resides, (4) containing about eighty-five acres more or less. The connection of the words “conveyed to me by the heirs of my deceased wife,” is so immediate and direct with the words “ all that my farm and plantation near Cropwell,” as' to be part of the same description, and words of true limitation, without which the description of the thing intended is not complete. Thomas, in fact, resided on the premises conveyed' to the testator by the heirs of his deceased wife. He merely cultivated the other premises, which had been conveyed to the testator by Lippencott, in connection with the place on which he resided. The whole description applies to, and correctly describes that part of the testator’s lands which had been conveyed to him by the heirs of his deceased wife, except the designation of quantity. A description by a statement of *417quantity is always regarded as the most unreliable of the methods of description. It may help to ascertain the premises intended, but will not control other and more certain means of description, and will readily be regarded as a false demonstration, when the premises intended are ascertained with convenient certainty by other descriptions. Llewellyn v. Earl of Jersey, 11 M. & W. 183; Doe v. Ashley, 10 Q. B. 663.
The fact that there is a devise to Thomas in the succeeding part of the same subdivision, of the one-fourth part of the residue of the testator’s real and personal estate, also supports this construction in some measure. The evidence leaves it in doubt whether the testator owned any other lands not specifically devised, on which this residuary gift could operate; if he did, they consisted of marl pits which had been exhausted, and were of insignificant value.
The testimony of the scrivener was improperly admitted, and the judge should have instructed the jury that on the whole case, the fourteen and seventy-three hundredths acres— the premises in controversy—did not pass to the defendant in entirety, under the will of his deceased father, but passed under the residuary clause, and therefore the verdict in favor of the defendant should be set aside.