By the Court,
Harris, J.The judge, at the circuit, was undoubtedly right in holding that the settlement deed, executed as it was by only two trustees, was void, if the third trustee was in fact living at the time of execution. The power of trustees over the subject matter of the trust, is equal and undivided. They can not, like executors, act separately — all must join, both *536in receipts and conveyances. Thus, where a testator gave his estate to two trustees for certain purposes, and, in case one died, authorized the,survivor to execute the trust, the refusal of one to act prevented the other from acting during their joint lives. So where a power of sale was given to three trustees, without authority to the survivors, in case of the death of one, it was held that such survivors were incompetent to act without a new appointment. (Willis on Trustees, 136, and cases there cited.) In this case the patent contained authority for the surviving trustees to execute the trusts, in case of the death of one; so that if Col. Jacob Hornbeck was dead, on the 13th of February, 1778, the settlement deed might legally have been executed by the survivors ; but if he,was alive, his execution of the deed was as necessary as theirs, to constitute a valid conveyance. It became, therefore, important to determine upon the trial, whether Col. Hornbeck was alive when the deed was executed. Upon this question I think the burden of proof rested upon the plaintiff. The trustees were to be chosen annually. Col. Hornbeck had been elected one of the trustees, in June, previous to the date of the deed. He was, of course, then alive. The deed, upon its face, assumes that he was yet alive, and makes him a party to the conveyance. Under these circumstances it was necessary for the plaintiff, in order to avail herself of the benefit of the deed, to show that Hornbeck was really dead when the other trustees executed it. This she attempted to do. For this, purpose she offered in evidence a memorandum, found on the fly leaf of the book of records of the town of Rochester, which speaks of the wife of Col. Hornbeck as his widow, and refers to an examination which the other trustees had made in respect to the trust property which had been in the hands of Hornbeck. This memorandum bears date the 10th of April, 1778. There is no evidence to show by whom, or under what circumstances it was made. It formed no part of the record properly made by any person, whose duty it was to make entries in the book. Indeed, it was not entered as any part of the record. The fact of its being found on the fly leaf imparts- to the memorandum no more effect as evidence than it would have had if found upon *537any other paper in the hands of the officers of the town. All that can be said of it is, that somebody, no one knows who, at some time, no one knows when, for some purpose, no one knows what, not as a record, not in the course of business, nor in the discharge of any duty, has seen fit to make this memorandum. I know of no rule of evidence, nor any precedent which would allow such an entry to be made use of as proof, to show that, on the 13th of February, 1778, Col. Hornbeck was not alive. I think it should not .have been admitted.
hior do I think the memorandum indorsed by Coekburn on the field book admissible evidence to prove the death of Horn-beck. It is nothing more than the written declaration of a third person, in respect to a matter with which he had nothing to do. The general rule is, that such declarations, though he by whom they were made is dead, can not be given in evidence. It can not be known that he was under any strong motive to speak the truth. It is true that what a man has himself actually done, and committed to writing, he being under an obligation to do the act, and it being done in the discharge of his duty, may be submitted to a jury as evidence that the act, thus committed to writing, was in fact performed. “ The general principle,” says Story, J. in Nichols v. Webb, (8 Wheat. 326,) “ is, that memorandums made by a person, in the ordinary course of his business, of acts or matters which his duty, in such business, requires him to do for others, in case of his death, are admissible evidence of the acts and matters so done.” (See also Welsh v. Barrett, 15 Mass. Rep. 381; Halliday v. Martinet, 20 John. 172.) Thus, the entries of clerks, of what they have done in the usual course of their business, have been received to prove the facts stated in those entries, when the clerks who made them are dead. (1 Stark. Ev. 315.) An entry made by a notary, of what he has done in the customary business of his office, may be evidence of his acts when he is dead. So, the memorandum made by a bank messenger, in the usual course of his employment, stating a demand of payment and notice to an indorser, is admissible as evidence of such demand and notice, when he is dead. Other cases might be mentioned, in which such entries have been *538received in evidence; but, in every instance, it has been where the entries related to some act performed by the party making them, in the discharge of his duty, and in the usual course of his business. To extend the rule beyond this limit, would be contrary to the most obvious principles of justice.
It is true, the learned judge instructed the jury to disregard the evidence, if they should be of opinion that the memorandum was made after the deed was executed. How the jury were to determine this question I am unable to perceive. There is nothing in the case to show when Cockburn made this memorandum. He might have made it at the time the deed was executed, and he might equally well have made it twenty years afterwards. But, whenever made, it was not competent evidence to prove the death of Hornbeck. It was not made in the discharge of any duty. It related to no act which Cockburn himself had performed. The same memorandum, made by any other man, would have been equally admissible.
Rejecting the evidence derived from the memorandum made by Cockburn, and the entry upon the fly leaf of the town records, I do not perceive that any evidence is left, to overcome the presumption that Hornbeck was alive when the deed was executed. The deed, therefore, was not shown to have been duly executed. Upon the principles stated by the learned judge himself, in his charge to the jury, and with entire accuracy, the deed itself should not have been submitted to the jury as furnishing any evidence of a conveyance by the trustees of Rochester to the proprietors of the Hardenburgh patent. It was void, for the Want of due execution.
I think, too, that the settlement agreement of the 21st of June, 1776, ought not to have been received without proof. It was admitted as an ancient deed or muniment of title. The mere existence of any instrument for more than thirty years is not enough, in any case, to authorize it to be read in evidence. Kent, Ch. J. in Johnson v. Blanshaw, (3 John. 292,) says, " It is the accompanying possession alone which establishes the presumption of authenticity in the ancient deed. Where possession fails, the presumption in its favor fails also. The length of the *539date will not help the deed, for if that was sufficient a knave would have nothing to do but to forge a deed with a very ancient date. (See also Healy v. Moule, 5 Serg. & Rawle, 185; McGinnis v. Allison, 10 Id. 197.) The theory upon which such evidence is allowed is stated by Starkie with remarkable clearness and felicity of language as follows: “ Presumptions are frequently founded upon, or at least confirmed by ancient deeds and muniments, found in their proper legitimate repositories, although, from lapse of time, no direct evidence can be given of their execution, or of their having been acted upon. It seems, however, that in order to the reception of such evidence, or at least to warrant a court in giving any weight to it, a foundation should be first laid for its admission by proof of acts, possession or enjoyment, of which the document may be considered as explanatory.” (1 Stark. Ev. 66.) So Gilbert says, “If possession has not gone along with it there should be some account of the deed, because the presumption fails where there is no possession, for it is no more than old parchment, if no account be given of its execution.” (Gilb. Ev. 103. See also Norris’ Peake, 163; Jackson v. Laraway, 3 John. Cas. 283; Hunt v. Luquere, 5 Cowen, 221.)
Middleton v. Mass, (2 Nott & McCord, 55,) was, like this, an action of trespass to’ try the title to a tract of land. The plaintiff produced a deed, dated in 1739, which had been proved before a magistrate and recorded in the auditor’s office a few days after its execution. He offered no proof of its execution, nor did he prove any possession of the land, or any acts of ownership over it, so that the question was whether it was admissible as an ancient deed, without proof of its execution. The presiding judge being of the opinion that it was not, the plaintiff offered to prove that the deed had been in possession of himself and those under whom he claimed, for more than thirty years, and contended that it ought to be admitted on this proof. But the judge thought otherwise, and the plaintiff was non-suited. A motion was made to set aside the nonsuit, on the ground that the deed ought to have been received in evidence as an ancient deed, on the proof of the possession of the deed *540alone for the time mentioned. The motion was denied. Johnson, J. in delivering the opinion of the court said, “ It is not the place only where an ancient deed is found that always makes it evidence, but it is when the possession is according to the provisions of the deed."
[Albany General Term, Sept. 1, 1851.Harris, Watson and Wright, Justices.]
In this case, as I understand the facts, it is not pretended that there has been such a possession as would furnish a presumption of the authenticity of the instrument from that source alone, even if it were an instrument which, if proved, would sustain a claim of title. It was necessary, therefore, before it could be read in evidence, that some proof of its execution should be adduced. It was proved that the subscribing witness to the agreement was still alive and residing in this state. No reason was given why he was not produced. I think the omission was a fatal negligence on the part of the plaintiff.
I am inclined to think that the instrument was also objectionable on account of the alterations appearing upon its face, but I do not sufficiently understand the materiality of those alterations to express an opinion upon this question. The rule on this subject, as stated by Phillips, is that, “ if there is any blemish in the deed by rasure or interlineation, the deed ought to be proved, though above thirty years old, and the blemish satisfactorily explained. (1 Phil. Ev. 477. See also 1 Stark. Ev. 344. Butter's N. P. 255. Jackson v. Osborn, 2 Wend. 555. McMicken v. Beauchamp, 2 Miller's Lou. Rep. 290.) In the latter case the court take the broad ground that writings erased or interlined are presumed to be false. (Provost v. Gratz, 1 Pel. C. C. R. 364. Morris’ Lessee v. Vanderen, 1 Dall. 64.)
For these errors in the admission of evidence, a new trial must be awarded. If upon that trial the plaintiff shall succeed in overcoming the difficulties already noticed, other questions of very grave importance, may arise, some of which were discussed with great learning and diligence by the very able counsel who argued this motion, but the decision of which is neither necessary nor appropriate to the proper disposition of the application. New trial granted.