Fowler, S.
The question arising upon the accounting involves the. determination whether or not three bequests to trustees, aggregating $140,000, are charged upon the real estate of which the testator died seized and possessed. The answer depends primarily on the terms of the will of William Lummis.
Paragraph tenth of the will provides for a trust fund of $50,000, to be held in trust during two lives. The income is given to the testator’s son William Maxwell Lummis, during his life, and upon his death [259]*259to his children,. if any, and if none the trustees are directed to divide the. annual income amongst the testator’s living children, or their issue. Upon the termination of the trust the principal sum is bequeathed to testator’s children then living. In paragraph eleventh of the will a similar trust is provided for in the sum of $40,000 for the benefit of testator’s daughter, Harriet Lummis. In paragraph twelfth another similar trust is created in the sum of $50,000 in favor of the testator’s son John. Paragraph thirteenth is the residuary clause. This clause is similar to the preceding clauses except that the sum placed in trust is not specified, but is stated to be the residuary estate. In all four of the paragraphs a trust is created for two lives, and upon the termination of the trust the remainder interest is given to the testator’s children. In addition to the foregoing provisions there are pecuniary legacies bequeathed by the will which aggregate in all $3,500.
As the net personal estate is considerably less than the amount necessary for the payment of the legacies, the question arises whether or not the residuary estate is to be encroached upon and whether or not there is a charge upon the real estate therein devised. The judgment depends upon the intention of the testator as determined primarily from the provisions of the will, and possibly from the circumstances surrounding him when he made the will. Carley v. Harper, 219 N. Y. 295.
On the assumption that parol evidence was admissible, some parol testimony of extrinsic facts, in addition to those derivable from an inspection of the account, was offered in evidence at the hearing. This evidence was all received subject to the surrogate’s final decision upon its admissibility." On the part of the widow there was given in evidence a statement [260]*260showing that four months prior to the date of the will the personal estate of Mr. Lummis amounted to $209,995. There was given in evidence by the trustees the statement of the testator’s personalty as it existed at the time he made his will. At that time it appears that the personalty had decreased to about $175,000. From the testimony of Mr. Ratigan, a friend of the testator, it appeared that Mr. Lummis had complained about the extravagance of his family. This testimony was offered by the trustees for the purpose of showing that the testator knew that his estate was dwindling. Further testimony to the same effect was offered from the mouth of Mr. Peck, the attorney who drew the testator’s will. Mr. Peck testified to a conversation he had with testator the day before the will was executed. At that time the testator stated to him in substance that the extravagance of his family was compelling him to live beyond his means and use up his principal. It also appears from the affidavit submitted by the widow, which was accepted in lieu of her testimony, that she has practically no estate of her own. The account shows a gross personal estate valuation of $174,858.24. The amount actually realized is $5,914.04 less. The funeral expenses and expenses of administration and debts paid amount to $28,732.06. The transfer tax was $1,100. Thus the general estate for distribution amounts to $139,112.14. As that sum includes $6,170 as a valuation of specific legacies, the net general estate is really $132,942.14, subject to commissions and expenses. That is the amount out of which provisions for legatees, amounting to $143,500, are to be satisfied, unless the deficiency is charged against the realty.
Before passing upon the competency of the extrinsic evidence offered on the hearing in this matter, it may be well to examine the principles generally controlling [261]*261the admissibility of extrinsic evidence in cases of construction, as there is very often some misapprehension concerning them. The rule requiring a last will to be in writing is not a rule of the law of evidence. The law of evidence requires no writing for any transaction. It is the substantive law, the Statute of Frauds, or, in this instance the Statute of Wills, which requires a testamentary disposition to be contained in a writing duly attested. This being so, the ordinary question submitted in courts of construction in regard to wills is: How far the express command of the state, that the transaction shall be in writing, may be departed from by the introduction of evidence not in writing, if such evidence is calculated to supplement the written expression of the intent of the testator? The departures from the express command of the statute are determined by no principle of the law of evidence, but by principles of an equitable character, excogitated by the judges themselves from those ideal juristic conceptions, known roughly to us as equity, and once as the ‘1 law of nature. ’ ’ But neither these departures, nor the conceptions, are to be found in the pure law of evidence, for the common law of evidence is only a combination of rules regulating the admissibility of evidence in trials in certain courts of law. When extrinsic evidence is not permitted to vary a will in writing, it is not a rule of evidence which excludes it, but the Statute of Wills.
When a court of probate sits as a court of construction (except as to construction of legacies) it is not sitting in its natural forum, and outside of its peculiar forum, such as matters of construction for example, it is bound by the ordinary rules of courts of construction, or,' in other words, by those adopted in courts of equity. In probate matters these rules are not a criterion. Hurst v. Beach, 5 Madd. 351; Guardhouse v. [262]*262Blackburne, L. R. (1 P. & D.) 109. In this matter the Surrogate’s Court sits not as a probate court, but as a court of construction, and consequently the rules adopted in other courts of construction are now applicable here.
In courts of construction the right of a party to introduce extrinsic evidence not in writing, for the purpose of aiding the construction or intepretation of a document required by positive law to be wholly in writing, is involved in some obscurity. Before taking up, for brief consideration, the rules applicable to such extrinsic evidence, let me advert to a corollary of the primary rule, generally laid down by the judges in courts of construction, in reference to the interpretation or construction of wills; the intent of the testator, inferable if possible from the will, is to govern construction ; and such intent is.to be deduced, if possible, from his written expression. This rule, it will be observed, is not peculiar to wills; it applies to all written unilateral instruments of a dispositive character. It applies a fortiori when an instrument is required by statute to be in writing. The difficulty with the principle last stated is that it is provisional. It does not profess to include any case where the intent is not inferable to some extent from the writing itself. Hence the common-law courts have laid down certain other principles regulating exceptions to the general rule, and these .subordinate rules so regulating the exceptions are far more complicated, and even less well understood, than the general principle itself. Volumes have been written on the exceptions, some of them entirely devoid of method, some of them faulty in logic, and very few, if any, altogether scientific or philosophical in treatment.
The reputed disregard of the science of interpretation, in common-law countries, is most apparent when [263]*263we think of the Federal Constitution. No document extant has ever been subjected to more intelligent and critical interpretation than the Federal Constitution of the United States, and yet there is not one published text book which deals scientifically with the fundamental principles governing such interpretation and construction. There are volumes enough stating topically the rules laid down empirically, their application and their cross-application, but none attempting to refer the rules themselves, abstractly and scientifically, to the metaphysical or general principles governing the interpretation of written documents. Indeed, which of the rules are a priori and incontrovertible nowhere appears. That this lack of scientific interest in the general science of interpretation is bound to be remedied in course of time must be apparent. But to recur to the principles governing extrinsic evidence:
When we come to inquire, “ When is the intent of the testator not inferable wholly from his written will?” we find it a most difficult resolution. But it may be affirmed generally that it is whenever the written expression, with its purely verbal definitions of values, affixed to them by lexicographers or by custom, is not deemed by the court to be consistent with the real intent — then the door is opened, sub modo, to evidence beyond the writing. The interpretation or construction, in other words, then invokes the judicial processes known as equitable. Construction is nothing other than an attempt by the judge in equity to bring the text and the intent of a dubious writing into precise harmony with equity, with equitable probabilities, and with the higher principles of justice and reason. But a resort to this extraordinary process cannot be invoked arbitrarily. Otherwise the Statute of Wills would be practically nullified by the reception of evi[264]*264dence not in writing. There must be a ease shown for equitable interference. The written expression must be at least dubious. If the will on its face is absolutely clear and not dubious in itself or with reference to any external circumstances, then no matter how inequitable the will may be, at common law the door is not to be opened for its amelioration beyond that possible from the written expression itself. This same general principle was fully stated in the Roman law: “ Ubi in verbis nulla ambiguitas, ibi nulla occurrit voluntatis, quaestio.” Vattel states the same general principle: “ Non licet interpretatari quae interpretation non egent.” Our courts and most courts of the common-law system act on the same principle. Champlin v. Champlin, 1 Sheld. 355; Higgins v. Dawson, 1902 App. Cas. 1.
But it has been said by an American jurisprudent of celebrity that “ written expression has no meaning until translated into things.” On this theory it is sometimes claimed, and I venture to think erroneously, that in every case where a will or other dispositive writing is up for construction or interpretation certain extrinsic things may be given in evidence in aid of interpretation or construction, viz., the situation of the maker of the instrument, or the testator, the nature of his property and the family surroundings when the will was made. Hawkins Wills, 14; Wolfe v. Van Nostrand, 2 N. Y. 440. I think this does not state accurately the common law, unless it is first made to appear that the intent of the testator or the true operation of the instrument is not deducible wholly from the written expression. Evidence of anything outside of the writing is always a resort to extrinsic evidence, and this is never permitted by the common law until some doubt is shown prima facie. The principle, as I understand it, in our common law is that when it is [265]*265made to appear on the hearing that there is something dubious or ambiguous in the written expression with reference to its future operation or action on persons and things named in the will, then, and not until then, the court will look at the state of testator’s property, his family and the like at the time the will was made. The old cases on this head are collected by Phillimore. Principles of Jurisp. p. 364. But where there is no ambiguity the rule is otherwise. Higgins v. Dawson, 1902 App. Cas. 1; N. E. R. Co. v. Hastings, 1900 id. 260; Shore v. Wilson, 9 C. & F. 355, 365; Matter of Raab, 79 Misc. Rep. 185,188; Matter of Farmer, 99 id. 445; Tole v. Hardy, 6 Cow. 341; Ritch v. Hauxhurst, 114 N. Y. 512; Champlin v. Champlin, 1 Sheld. 355; Underhill Wills, § 456, and cases cited; Morris v. Sickly, 133 N. Y. 459; Fries v. Osborn, 190 id. 35. I have now alluded, very briefly, it is true, to practically all of the substantive law regarding the receipt of extrinsic evidence in the course of the interpretation of written documents. All the residue of the principles stated in the modern books are mere inferences, corollaries or exceptions, sometimes collected under the head of Evidence, but really having little to do with the substance of the law of evidence.
That a writing is to be construed according to the intention of the writer certainly is not new, for I find that it was stated to be settled law 2,000 years ago (Quintiliani Oritoriae Institutionis, lib. VII, caput V et caput VI), and this fact the Corpus Juris abundantly confirms (D. 50, 17, 96.) The Civilians are of great authority in courts of this character. There is, I may observe, much of interest on the law of wills in such out of the way Latin books as those of Quintilian or Valerius Maximus, devoted to entirely different subjects. The Digest is often clarified by matter found in Latin literature of a general nature. In this [266]*266court and courts of like character the Latins and the Civilians are frequently consulted (Hurst v. Beach, 5 Madd. 321), as in processes of interpretation the Roman Law and the Civilians have greatly influenced the common law. Williams Wills, 94. But I am bound in this matter wholly by the principles laid down by our courts of construction, although occasionally, in the absence of precedent, the civil law may be here controlling. Matter of Van Ness, 78 Misc. Rep. 600; Matter of Swarts, 79 id. 388.
In this cause now before the surrogate there is no need to go beyond the modern reports, as it seems to be a cause where there is something dubious to- consider. Consequently the extrinsic evidence of the circumstances surrounding the testator at the time he made his will is here admissible to prove or disprove an intention to charge lands. McCorn v. McCorn, 100 N. Y. 511, 513; Brill v. Wright, 112 id. 129, 135; Irwin v. Teller, 188 id. 29, 33; Ely v. Megie, 219 id. 112, 127.
But what is most generally meant in modern law by extrinsic evidence in aid of interpretation of a will is not the circumstances surrounding testator, but direct evidence of the testator’s intention, i. e., his declarations made outside of the written expression contained in the will itself. Now, I apprehend that it is well settled in our common law, whatever the principle may be in other systems of law, that no direct evidence of a testator’s intention, as evidenced by his declarations, made either before or after the execution of his will, is admissible in our system, in a case of construction, except where there is a latent ambiguity or where a person or thing is described in the will in terms equally applicable to two or more. That is the only instance when such evidence is admissible. Thayer Ev., 482; Matter of Vosseler, 89 Misc. Rep. 674, 676; [267]*267Williams v. Freeman, 83 N. Y. 569; Strahan, Law of Wills, 122. At this point we may observe that at common law there are two schools of thought dealing with interpretation; the liberal school, which in cases of ambiguity would admit any evidence whatever provided it was relevant, and the strict school, which would confine the admission of extrinsic evidence within the narrowest possible limits. Most common lawyers belong to the latter school. The civilians generally belong to the more liberal school. But I apprehend that in my seat I am to go not by my personal preference, which would lean to the liberal side, but by the adjudications. Now our adjudications on this head are express.
In this matter the extrinsic evidence was offered and taken subject to the objection to testator’s oral declarations offered to show the intent of his will. As there was no jury in this matter I permitted such evidence to be taken, subject to all objections to its final reception. It made no real difference in a case before the surrogate alone, if evidence was taken subject to further consideration, even if it proved in the end to be incompetent, for it might then be disregarded. Meanwhile its reception or rejection had been the subject of a more careful consideration than the exigencies of the ordinary trial at bar permitted.
After due consideration I conclude that the parol evidence referred to as, taken on the hearing is properly before the court in this case. Such evidence is explanatory of the circumstances surrounding the testator at the time the will was made, and it is well settled that such evidence is receivable for the purpose of determining the testator’s intention to charge legacies upon his realty whenever a doubt is disclosed. Though the use of parol evidence in the construction of wills is very limited, it is permissible to examine [268]*268into the circumstances surrounding the execution of the will when a doubt is disclosed. Matter of Turner, dissenting opinion of Miller, J., 142 App. Div. 645; affd., 206 N. Y. 93; Matter of Farmer, supra. The declarations of the testator, which have been testified to, showing that the testator appreciated the extravagance of his family, were not offered as declarations of testamentary intention, but for the purpose of showing one of the surrounding circumstances attending the execution of the will. Such declarations of a testator are not received as his expressions of his intention either to charge or not to charge legacies upon the land. They aré received in order to show testator’s knowledge and appreciation of the extent of his property. This is deemed a proper circumstance to be considered on the question now before the court. Even viewing them as declarations of intention, it would seem that such evidence might be admissible on the theory of “ rebutting ati equity.” Wigm. Ev. § 2475.
The rule which makes legacies primarily payable out of personalty may be modified in its application by proof of declarations of a testator that he believed the legacies to be payable out of realty as well as personalty. But the theory for receiving declarations of intention in such a case as this is that declarations of testator’s intention are admissible where there is a latent ambiguity. Matter of Turner, supra. Whether or not legacies are to be paid out of realty is a latent ambiguity. Where there is nothing in the will to determine out of what property the legacies are to be paid, and there is shown to be insufficient personalty for that purpose, an ambiguity immediately arises dehors the will.
After full consideration of all the evidence in the case, I believe there is not" sufficient upon which to base a finding that the legacies are to be paid out of the [269]*269real estate. As is the ease with all questions of intention, no precedents are absolutely controlling in other cases, for all differ. The rule of law which induces courts of this .state to charge legacies upon land has been so frequently reiterated that the principles involved are clear and have been exhaustively stated. Only recently in Carley v. Harper, supra, the Court of Appeals, through his Honor Judge Pound, very succinctly stated and summarized the guiding principles most commonly invoked. It has been stated that the burden is upon the legatees claiming the charge upon land. Brill v. Wright, 112 N. Y. 129. But even if the language of the court in that case is susceptible of the meaning, that such burden rests, both as a matter of fact and of law, on the legatees, the more recent decisions of the Court of Appeals, culminating in Carley v. Harper, seem to lighten such burden, if not to remove it entirely.
The situation of the rule at the present time seems to be this: the distinction between realty and personalty in wills is not to be overlooked. The courts, however, seem to take notice of the failure of the average testator to appreciate any such difference. If the language of the will is susceptible of the meaning, that realty and personalty are to be used alike in satisfaction of the legacies, a charge upon the realty will be imposed. Simultaneously with the apparent inclination of courts to lighten the presumption against imposing charges upon land in favor of legatees, another presumption — that the testator intends payment in full of the legacies bequeathed by him—has gained in force. • The real question in each case of construction seems to be whether or not the will, when construed in the light of extrinsic circumstances, evinces an intention that the legatees should be paid at all events. It is for that reason that throughout this case [270]*270before me so much stress has been placed upon the amount of personalty owned by the testator at the time of his testament. Where the personalty is grossly out of proportion to the legacies the courts are inclined to impose a charge upon the land; otherwise the will would be a vain and a useless act on the part of the testator. See McGoldrick v. Bodkin, 140 App. Div. 196.
The blending of realty and personalty in a residuary clause, the existence of a power of sale and the direction to pay inheritance taxes out of the residuary estate, while worthy of consideration, are never absolutely controlling in determining charges on land. Those circumstances are frequently referred to in the opinions of appellate courts, but the dominating factor has always been the amount of personalty on hand at the time the will was executed. In the case now before the surrogate the usual argument is predicated of the existence of a power of sale, the blending of the realty and personalty in the residuary clause and the direction to pay the inheritance taxes. The contrary argument, in favor of the contention of the widow, is just as strong, based upon these same premises. A power of sale is often put in a will because the precedents will have it so and because such a power may in some way at some time prove valuable. The express direction to pay the taxes out of the residuary estate may be regarded as an indication that the legacies themselves were not so payable, especially where the will was drawn by an attorney. The blending of realty and personalty in a residuary clause was stated in Brill v. Wright, supra, of itself not to be a dominant consideration. Irwin v. Teller, 188 N. Y. 25.
The learned counsel for the trustees argued that the conjunction of the word “ devise ” with the word “ bequeath,” and “ rents ” with “ income and [271]*271profits,” in the paragraphs of the will creating the three trusts indicates testator’s intention to establish .these trust funds out of the realty. This conclusion does not necessarily follow. The testator merely referred to a sum of money. The use of the words “ devise ” and “ rents ” seems to be the result of inartificial expression and the too servile following of some familiar precedent or formula. The argument of the trustees is not fortified by looking at the power of sale in this connection. Were the power to be exercised it could only operate on personal property, and there was no occasion whatever for the use of the words “ devise ” or “rents.” It would seem, then, that the testator did not contemplate the exercise of the power of sale so far as the creation of the three trust funds is concerned.
In many wills, where a charge has been implied, children or close relatives were aided at the expense of strangers. In this case nothing would be left for the widow were the contentions advanced in favor of the children to prevail. Moreover, the testator has not made even a suggestion that the children. are to be preferred beneficiaries. The three trusts for the latter are in similar form to the one for Mrs. Lummis. The legacies are of particular amounts of money, whereas the widow is given the income of the residue of both the realty and the personalty. The will has every appearance of an intention to give the specific amounts to the children, solely because the testator then believed his fortune large enough for all his testamentary purposes. Two 3rears elapsed between the making of the will and its probate. During that time the evidence does not show an intention to resort to the realty for the purpose of discharging the legacies. But subsequent events are not admissible evidence on this point. Morris v. Sickly, [272]*272133 N. Y. 456. The Court of Appeals, however, in Carley v. Harper, supra, said that' subsequent reduction of personalty and the increase of realty have a bearing upon the question. The court cites Scott v. Stebbins, 91 N. Y. 605, apparently overlooking Morris v. Sickly to the contrary.
It is very probable that the testator never conceived of the situation which now presents the real difficulty. It is often the fact that a court of construction must resolve hypothetical intentions which were really not present when an instrument was drawn and executed. This is notably so in the instance of instruments operative inter vivos as well as with last wills and testaments. In a certain class of written contracts subsequent uncertainties of meaning frequently arise from the use of careless expressions of intention, or by reason of the incompleteness of comprehension on the part of the parties to the writing. One well-marked illustration is a contract where the court must imply a condition as matter of law. The reason assigned for such an equitable interpolation is the failure of the parties to foresee a contingency which, when it later arises, must be construed as to its legal effect upon the rights and the obligations of the parties. When the contract refers to the sale of the product of a particular factory, for example, the destruction of the factory during the vendor’s performance excuses further deliveries, even though the contract is silent concerning such an accidental interruption. The law will then imply a condition. This is so not because any particular language of the writing suggests such an excuse for the vendor, but because a court of equity assumes that the entire scope and subject matter of the contract points unmistakably to the inference that the parties really intended such a provision in the contract. Though the court of construction appreciates the absence of any expressed [273]*273intention, one is nevertheless generally supplied. This is not an objectionable interpolation of language into the contract, but the interpolation of an intention gathered and construed from the intention expressed and the extrinsic facts. So with cases of wills where the testator fails to state what should be the legal consequences of the instrument, in certain contingencies, a court of construction may place itself in his position and read as a portion of the will what the language requires. The court then merely attributes to the testator the effect which, placing itself in his position when he made the will, it can see he must have intended.
The real question is often disguised in courts of construction. It is, What is the proper construction when there is no evidence of intention either inside or outside of the will? A few observations on this- point will do no harm. The usual method of dealing with such exceptional cases does hot violate any of the cardinal rules relating to the construction of wills. It is the identical process and basis upon which courts of equity have always proceeded to construe wills. An admission that the testator had no real intention upon the question which the court is resolving by construction is only a frank avowal of what is plainly manifest in too many cases on wills. The construction of a will in such a case is then purely a matter for equity. It is different from a literal interpretation of the express language. Matter of Catlin, 97 Misc. Rep. 223. The effect of certain provisions in a will alone is to be resolved.
WTien precedents must be relied upon exclusively for extracting the intended legal effect from the language contained in the instrument, a court of construction should not make any pretense of ascertaining the real intention when none exists. It is not then ascertain[274]*274ing the intention, but expounding equitably an intention which equity assumes to be the real intention. The intention is but the court of construction’s understanding of the effect of the language of the instrument as applied to the circumstances given in evidence. The prohibition against constructing a will (Herzog v. Title Guarantee & Trust Co., 177 N. Y. 86) is not then violated. The court need not then indulge in conjecture nor enter upon speculation nor offend an established rule of the courts by supplying language. Dreyer v. Reisman, 202 N. Y. 476. All that properly need be done is to apply equitably the inference raised from the express language as courts usually apply it.
A frank appreciation that, in many cases on wills, particularly those where such a question as that now here is up for consideration, the court is not really interpreting the language of the testator, but is equitably expounding the effect of the provisions of the instrument, would be far more conducive to precision and system in the law of construction than to assume that the court must gather the testator’s intention when it is plain that none in fact ever existed. Subordinate rules of construction will then not be so often misapplied. The paramount desiderandum in every case on wills should be an equitable determination. Particular provisions and claims of the will should not be used as a fanciful structure upon which to rear an intention which the whole language of the testament shows that the testator never entertained.
But be this as it may, the contention that the legacies by way of trusts in paragraphs ten, eleven and twelve of the will of Mr. Lummis were charges upon the realty ought not to prevail in this case.
Decreed accordingly.