Tilshman C. J.
The first reason offered for a new trial in this case, is, that the defendant was not permitted to give in evidence, the certificate of the secretary of the land office verifying the copy of an entry in a book in the rolls’ office, which together with all other papers in that office touching *297the titles of lands, had been transferred to the land office by virtue of the act of 29th March 1809. This entry was made by the recorder of deeds for the county of Philadelphia, in pursuance of the sixth section of the act of 18th March 1775. It happened that Mr. Irwin, who held the two offices of master of the rolls and recorder of the county of Philadelphia, had not kept distinct books for each office, as he ought to have done, but had blended in one book, entries belonging to both offices. In consequence of this irregularity, the entry in question found its way improperly into the land office. It was undoubtedly of a public nature, and had it been certified by the recorder to whose office it belongs, there would be much to say in favor of admitting it as evidence, although the act of 1775 does not say that it shall be evidence. But that is not the case. The master of the rolls had no authority to give such a certificate, because it did not belong to his office. It follows, that the secretary of the land office can have no authority, for his authority is confined to the papers of the rolls’ office transferred to the land office. To make the secretary authority greater than that of the ^master of the rolls’ would be to make the stream rise higher than its source. I am therefore of opinion that this certificate was properly rejected.
The second reason for a new trial, is, the rejection of certain deeds containing a recital of a deed from Richard Dennis and wife to Thomas Carpenter, and a reconveyance from Carpenter to Richard Dennis, on which the title of the defendant rests, and which are alleged by him to have been lost, or destroyed. The lost deeds bear date the 22d and 23d January 1761, and contain a conveyance of all Mrs. Dennis’s right to the estate of her father John Coates, who in the year 1760 died seised of a considerable real estate which he devised to Mrs. Dennis and his other children in common. It is to be observed that by the will of John Coates, Thomas Say was appointed a trustee with power to make partition among Coates’s children, of the lands devised to them. On the trial of this cause the defendant had given evidence of the existence and loss of the deeds in question, sufficient, in the opinion of the judge, to let the whole matter be laid before the jury, to whom he referred it. The defendant then offered the deeds containing a recital of the lost deeds, and the court rejected the evidence.
It is a general rule, that the best evidence shall be given which the nature of the case admits of. Proof by the oaths of the subscribing witnesses is the best evidence of the execution of a deed; but if they are dead, proof of their hand*298writing, and of the handwriting of the grantor will be received. I am now supposing that the deed is in existence and produced to the jury. But suppose it to be lost, to be an ancient deed, and the subscribing witnesses to have been long dead.. The rule still adapts itself to the nature of the case. In such a case it would be unreasonable to insist on proof of the handwriting of the subscribing witnesses, or of the grantor, because it may be impossible to find any person who has seen the deed. Necessity, either absolute or moral, is sufficient ground for dispensing with the usual rules of evidence. The oath of a G-entoo, administered according to the forms of his own religion, was admitted in the case of Omichund v. Barker, 1 Atk. 21, where the principles of evidence were well considered. It is on this principle of necessity, that the law pays great respect to ancient deeds, *esPecially ^ attended with possession. If such a deed is in existence, proof by the oath of witnesses is dispensed with, unless the deed carries on its face marks of suspicion. It is laid down by Lord Gilbert in his Law of Evidence 101, that the inspeximus of an ancient deed may be given in evidence though the deed need no enrolment, “ because an ancient deed may be easily supposed to be worn out or lost.” But in judging of ancient deeds, possession is a circumstance of great importance. • If the deed had not been executed, it is to be presumed, that the persons entitled to the land, would not have suffered the possession to remain out of them. But where possession has not gone along with the deed, the presumption is against it, because if the deed is genuine, it is difficult to account for the want of possession. In the present case, the deed alleged to have been lost, is fifty years old. If the possession had been held solely under this deed from the time of its date, there could be little doubt, but that a recital of it in another ancient deed would be evidence, although the general rule is, that a deed containing a recital of another deed, is not evidence of the recited deed except against a person who makes such recital and those who claim under him by title acquired afterwards. It is necessary therefore to inquire into the possession. The persons claiming under the lost deed have held the possession from its date. But it cannot be said that the possession has been always held in consequence of the deed, because Richard Dennis who lived till the year 1797, was entitled to the possession during his life as tenant by the curtesy. It is a middle kind of ease then, in which although the possession may not have been held' solely in consequence of the deed, yet there never has been any possession against it. It de*299rives considerable weight from antiquity, although not near so much as if it had been the sole cause of possession. In considering the propriety of admitting a recital as evidence, it must be recollected, that in the present instance, it was offered as a corroborating circumstance of the existence of the recited deed, a foundation having been laid satisfactory to the court, both of its existence and loss. And it is material that the persons making the recital are those most likely to be acquainted with the truth of the matter, and whose interest it was that no such deed should have existed. Thomas *Say had no interest, but was the confidential friend of Mrs. Dennis’s father, appointed by his will to make partition among his children ; and the Coates’s family were the heirs of Mrs. Dennis’s children, to whom their mother’s estate would have descended, if it had not been conveyed to their father. The assertion of such persons must make a strong impression. But it is objected, that however impressive the declaration of a man of character may be, even without his oath, yet the law admits the word of no one in evidence without oath. The general rule certainly is so; but subject to relaxation, in cases of necessity, or extreme inconvenience. How is it expected that a deed, like the present, is to be proved, when the subscribing witnesses have been dead eight and twenty years, and the deed itself is not to be found ; for in considering this question we must assume that it is lost. Is it not necessary to resort to secondary evidence, without oath ? If the deed could be produced, it might be reasonable to require proof on oath of the handwriting of the witnesses or of the grantors. But that is not the case. Why was it, that on the trial of this cause, the court admitted in evidence, a copy of the deed alleged to be lost, taken from the records of New Castle county, although it had not been recorded agreeably to law? There was no oath of the recorder that it was a true copy of the original deed. It was from the necessity of the case, and from the improbability that the recorder would have placed it on record, without having seen and carefully compared it. Why then shall not a recital be evidence? I confess, I see no good reason unless it has been adjudged to the contrary. No such adjudication has been shown in the case of an ancient lost deed ; but an authority which has not been answered, has been cited in favor of the evidence. In the report of Ford v. Grey, 1 Salk. 285, it is said, that a recital of a lease, in a deed of release, is good evidence of such lease against the releasor and those claiming under him; “but against others it is not, without proving there was such a deed, and it was lost or destroyed.” *300The same case is cited in 12 Yin. 233, pi. 17, and thus expressed. “ A bare recital of a deed is not evidence; but if it could be proved that such deed had been lost, it would do, if it were recited in another.” If this evidence were to be conclusive, there might be danger in admitting it. But *thejury are to judge of its weight. It is only submitted to them as a circumstance proper to be taken into consideration.
There is another point of view in which this matter may be considered. It is material that the nature of Richai’d Dennis’s possession should be ascertained. The bare act of holding the land does not decide this, because it is consistent either with an estate for life or in fee simple. But the conveyances offered in evidence show the quo animo the land was held. They prove the actions of Richard Dennis; they show that he exercised acts of ownership of the most public nature inconsistent with an estate by the curtesy, with the full knowledge, approbation and co.-operation of Thomas Say and the Coates’ family. Under the particular circumstances of this case, I cannot but think that these deeds are evidence to explain an equivocal possession, especially after the plaintiffs had offered evidence to show that Richard Dennis did not consider himself as entitled to an estate in fee.
Upon the whole, I am of opinion, that the deeds of conveyance of Richard Dennis in which he was joined with Thomas Say and the Coates’ family, or either of them, were competent evidence, and therefore there should be a new trial.
Yeates J.
This cause was warmly disputed on the trial. Every inch of ground was contested with zeal and ingenuity by the conflicting parties, and no less than twelve points of law respecting evidence were noted at the instance of the counsel. The verdict having passed for the plaintiff, three exceptions have been taken by the defendants, to the opinions of the court at Nisi Prius, and are assigned as grounds for awarding a new trial.
1. It has been contended, that the certificate of the secretary of the land office that the transcript of the entry made in the day book remaining in his office, was a true copy, ought to have been received in evidence.
It appeared by the oath of John M. Irwin, who was clerk in the office of his father in 1791, (then being master of the rolls, as well as recorder of deeds for Philadelphia county) that his said father had made use of the blank leaves found *301in a boob, wherein patents and powers of attorney had been recorded, for entering of deeds brought into his office *to be recorded pursuant to the directions of the act of assembly, passed on the 18th March 1773 ; that the said book was retained by the master of the rolls, when the two offices were held by different persons, on account of the patents enrolled therein; and that the copy produced had been compared by him with the original entry made in his handwriting, and found to be correct.
Upon this testimony, I admitted the copy in evidence, but rejected the certificate on four grounds. Because 1, I did not consider this a paper or entry in the land office, within the meaning of section third of the law of 9th April 1781; 1 Ball. Sfc. Laws 891. 2. The book directed to be kept by the law of 18th March 1775, was merely to secure to persons bringing deeds to be recorded, the right of having their deeds entered in regular succession. 3. If even this book of entries should be deemed a public record book, it belonged to the office of recorder of deeds, and not of the master of the rolls, and consequently was not within the provisions of the law of 29th March 1809, 9 St. Laws 122, which directs that “all the patent books, papers, and documents relating to the titles of lands shall be delivered to the secretary of the land office, and be by him deposited in his office.” And 4. That no legislative provision had been made, as was usual in all eases wherein a new department of office was erected, that copies thereof certified by the officer, should be received in evidence. I know of no other legal mode of proof of the entry made in this book, except the book itself, or a copy proved to have been compared with the original, the latter of which was pursued in the present instance.
2. It has been contended that the deed from Warwick Coates and wife, and Bichard Bennis and his last wife Mary to the Bevd. Jacob Buché and others, for the site of a church, dated 16th June 1764, should have-been admitted in evidence, to prove the existence of the deeds alleged to have been lost. This deed contained a recital therein, that the said Bichard Bennis and Hannah his wife had conveyed all the share and purpart of the said Hannah, under the will of her father John Coates, to Thomas Carpenter, by deed dated 22d January 1761, who had reconveyed the same *to the said Bichard Bennis, by deed dated on the r*ggp following day; and that both deeds had been re- “■ corded in New Castle county.
It must be observed that the said John Coates died seised *302of valuable lands, in and near the city of Philadelphia, and by his last will, duly proved on the 20th March 1760, devised his estate real and personal equally to be divided amongst his children, and appointed Thomas Say his trustee, with authority to divide his estate amongst his children in case they could not agree themselves to make partition. On the 26th January 1761, Say made a valuation of these lands in distinct parcels, amounting in the whole to 7408?. 9s. 10c?. and assigned lots to the different children, according to an original paper admitted to be in his handwriting,' and agreed to go in evidence to the jury. This division was assented to by the devisees, as appeared by a memorandum on the said paper, subscribed by all the children on the day following excepting the said Richard Dennis and Hannah his wife. Those children, with the exception of Dennis and his wife, afterwards on the 6th June following, executed a release to the said Thomas Say of their undivided shares and purparts, in order to enable him to convey to them respectively their lots in severalty which they had drawn for. Hannah Dennis died on the 16th August 1761, leaving issue by her husband, who in consequence thereof would have been entitled as tenant by the curtesy, to the possession of any lands held in her right, during the term of his natural life. The antiquity of the deeds is therefore in a great measure out of the question. Dennis died after-wards in the month of January 1797, and made his will devising the premises in question to the defendant his son. The plaintiffs made title to the ground in dispute, Hannah Garwood claiming under Hannah Dennis as her granddaughter, and John Buckley Elliot as her great grandson. These details are necessary for the understanding of the points of evidence before the Court.
The great question on the trial, was, whether Hannah Dennis had by any act in her life time, divested herself of her right to the land in controversy ? The defendants attempted to show that this had been effected by the before *re°ite(i deeds from Dennis and wife to Carpenter, and his subsequent reconveyance to Dennis himself, which were said to be lost.
What then is the rule of law as to recitals being evidence; and on what principle is the rule founded? Recitals in a deed are only evidence against the grantors and their heirs, and those claiming under them, either by descent, or purchase subsequent to the execution of the reciting deed. This appears clearly by the English cases, and by the decisions of our own courts. In Bonnet’s Lessee v. Devebaugh et al. 3 *303Binn. 175, 179, 190, it was determined, that though a recital in a warrant of acceptance, that a survey had been made in 1755, by virtue of an order from the late proprietaries, and a patent grounded thereon, were binding on the proprietaries and commonwealth, yet it was no evidence against other persons claiming the premises by actual settlement, previous to the warrant of acceptance. So in Penrose v. Griffith, Supra 231, heard last September term, in the southern district, it was decided, on great consideration of the peculiar circumstances of the case, that a patent reciting seven different mesne conveyances of the warrant right, with their dates respectively, was no evidence of those transfers, against a person claiming under an improvement prior to such warrant.
This doctrine rests on the most rational ground. The law requires the sanction of an oath,in the development of facts; but recitals in conveyances are nothing more than written assertions of acts and events, without such proof; and stand precisely on the same footing as the oral declarations or confessions of the parties. The latter will be admitted against the party himself, and those claiming under him, posterior to such declarations or confessions; but not against third persons whose pretensions are adverse to him. So also of recitals. Now here the plaintiffs claim under Hannah Dennis ; but the defendant claims under the will of Richard Dennis, which affects to dispose of the subject in dispute as his own. As to the matters in controversy, he was not their representative; nor can any act or deed of his detract from their title, who not only derive no right under him, but *whose pretensions are in direct opposition to his supposed rights. I can see no difficulty whatever in this question.
8. But it has been urged, that other deeds ought to have been permitted to go to the jury, after what had passed in the course of the trial. A deed from Richard Dennis and Mary his wife to Blair M’Clenachan, dated 26th July 1791, for 100 acres of land, in "White Clay Greek hundred in the county of New Castle, in consideration of 1545Í. with a covenant of special warranty against the children of Hannah his first wife and their heirs, had been read in evidence on the part of the plaintiffs; and Captain William Paul had also been examined as a witness on their behalf, as to certain expressions made use of by Dennis, in two conversations had with the witness, in the fall of 1775 ; in one of which Dennis declared he never had a title for the estate of his first wife (Hannah), and in the other, that he could not make a title to *304the Northern Liberty lots, which came to him by his first wife; complaining of the hardships he was subjected to in the payment of taxes for that property. One of the deeds offered in evidence by the defendant, was dated 30th April 1763, from the aforesaid Thomas Say, and the surviving children of old John Coates, to Mary Taggart for a lot in Front street above Arch street, in consideration of 6007 In this deed, it was recited that Dennis and his first wife had conveyed the lot to some person in fee, who had reconveyed it back to Dennis, the same Dennis being a party thereto. The other deed was dated 28th June 1769, from Richard Dennis and Mary his wife to John Philips for two lots on Front street, in consideration of 2007 containing the same recitals as the before-mentioned deed of 16th June 1764 to Jacob Duché and others, and was likewise offered in evidence by the defendant. His counsel insisted on the trial, that these deeds proved acts of ownership done by the defendant’s father, over the landed property derived from his first wife Hannah, inconsistent with his holding as mere tenant by the curtesy ; that if conversations, disavowing a right in himself thereto, and declaring that he could not make a title therefor, could be given in evidence against the devisee, so also might other written declarations at other times, avowing the right of Richard Dennis thereto, and the actual exercise of ^at be *given in evidence in his favor, in order to do complete justice; and that these deeds, so offered, tended to invalidate the testimony of Captain Paul, from the improbability of the fact supposed to have been related in conversation.
To these observations I have several distinct answers.
If acts of reputed ownership should be deemed to be relevant, under the special circumstances of this case, on which I profess to give no opinion at present, they are susceptible of legal proof, by witnesses upon oath or affirmation; but let not one of the most fundamental rules of evidence, that no man shall be permitted to make evidence for himself, where there cannot be a pretence of necessity, be violated.
I take it to be another established rule of evidence, that a man’s confessions or declarations may be given in evidence against him, and when so given, the whole of what passed at the time, shall be received and weighed together; but that no one will be allowed to urge, what he has declared at other times, in his own favor. This rule is founded in strong good sense, and equally prevails in civil and criminal cases. On this point, I recollect a decision, which I cited upon the trial, the impressions whereof will never be effaced from my *305memory. Galbraith’s Lessee v. Scott was tried at Nisi Prius in Lancaster in 1781, before M’Kean Chief Justice, and Bryan Justice. The lessor of the plaintiff claimed one moiety of a tract of land, conveyed to Samuel Scott by John Maese, formerly of this city; and founded his pretensions ou a supposed partnership in the purchase. The defendant held as devisee under the will of Scott. The plaintiff to establish his right, called two witnesses to prove, that Scott had declared he and Galbraith bought the lands jointly. Shippen, Bradford and myself for the defendant, offered to show by other witnesses, that the uniform declarations of Scott, both before and after the times referred to, were that they had originally intended to purchase in partnership, but from the inability of Galbraith to pay his proportion of the consideration money, the first contract had been rescinded. We inferred from thence that the plaintiff’s witnesses, must have either mistaken, or misstated what had passed between them and Scott. This testimony was excepted to by Wilson, Smith (of York Town), and Chambers, of counsel with the plaintiff; *and after a lengthy argument, the same was overruled by the court, who stated the rule of evidence to be as I have laid it down. It was no hasty decision as was suggested by the defendant’s counsel. The plaintiff recovered ; and on a new ejectment brought for the same lands, the same decision was given on the evidence by other judges. I consider that case much stronger than the present. There both parties claimed under the deed of Maese to Scott; but here the plaintiffs claim under Hannah Dennis as their ancestor, and the defendant claims under her husband.
It was certainly competent to the defendant to invalidate the testimony of Captain Paul by legal evidence. If he had spoken of these conversations at different times in a different manner, it would have gone to the jury to detract from his credit. But how can the two deeds offered produce., that effect ? How can the acts of Dennis lessen the credibility of Paul? Will any rational man consent, that his own veracity, or the accuracy of his memory, shall be tested by the conduct of persons over whom he has no control ? And as to the improbability of the conversations having happened, I will merely observe that if the deeds had been read, they would only show, what has happened to other men, that the. sayings of Dennis at one time cannot be reconciled with what he has done at another.
As to the dictum attributed to the court in Ford v. Lord Grey (reported in 6 Mod. 45) that “ we cannot hinder the reading of a deed under seal; but what use is to be made of *306it is another thing,” it has long been exploded, both here and in England. The relevancy of every deed offered in evidence, must necessarily be judged of by the court before whom the trial is had. Were it otherwise every deed entered in the recorder’s office, might be read on the trial of an ejectment.
I have stated the purposes for which these two deeds were explicitly declared to have been offered in evidence on the trial, by the defendant’s counsel; and have given such answers to the remarks made by them, as have occurred to my mind. I am very sensible that a deed may be evidence for one purpose, though not for another. When however the counsel avow the objects for which the testimony is adduced, the correctness of the court’s opinion must be tested by *that view of the subject. If no object is declared, and the evidence is admissible in any point of view, it is error if it be rejected. Such was the opinion of this Court, in a writ of error by Richardson v. The Lessee of Stewart, Supra 198, at the last term at Chambersburgh.
It has been said during the motion for the new trial, by one of the defendant’s counsel, that the recitals in the two deeds which were overruled, repelled the arguments urged on the part of the plaintiffs, as to the covenant of special warranty contained in the deed to M’Clenachan. Unquestionably they would produce that effect, and a much greater one also, if those recitals were legal evidence of the existence of the deeds recited. But this is the very question now to be decided.
By another of the counsel it has been remarked that Thomas Say’s having joined in the deed of 30th April 1763 as a trustee, is a powerful argument, to show that he must have seen the recited deeds, made at no great period of time before; and that the circumstauce of Dennis not having executed it until September following, accounts for the recited deeds not having been more fully set out therein. But the terms of that recital only relate, as it would seem, to that particular lot, and not to the general share of Hannah Dennis in her father’s lands. If this is the true construction, that deed can have no effect on the cause before the court, which respects other lands. I regard the valuation of the whole estate of John Coates, by Thomas Say, as evidence of the appropriation of the lots classed therein to the different devisees, and of the assent of the subscribing devisees thereto; though I do not view it as evidence of independent facts stated therein, such as Dennis having taken a copy of that paper,—Say’s giving a plan to all the heirs,—and *307making deeds to the heirs of John Coates for their shares on the 8th June 1761. The paper was introduced to show, that the two lots of ground in dispute, were appropriated amongst others to the share of Hannah Dennis; but the counsel on each side conceiving that strong arguments could be adduced therefrom in support of their pretensions respectively, enlarged fully on all the facts stated therein. I will only remark ^further, that if this paper should be deemed legal evidence of Say’s having conveyed to Richard Dennis and Hannah his wife in June 1761, or if David Ware is correct in what he said respecting his having seen such a conveyance to Dennis and his wife, then this deed from Say of 1763, ought not to be admitted, because it tends to invalidate the deed previously given by him. For if Dennis and his wife conveyed all her landed estate, under her father, to Carpenter, on the 22d January 1761, and Carpenter had reeouveyed to Dennis on the day following, and Say had been informed thereof, it would have been his duty to convey those lands to Dennis himself.
It possibly may be alleged, that the plaintiffs’ counsel have founded an argument on the representation of Say, considering it in the light of a written declaration by him without oath, that he had made a deed to the heirs of whom Hannah Dennis was one, and that, therefore, his deed of 1763, might be shown as a written declaration, impugning the effect of the former. But such purpose was not avowed during the course of the argument respecting the evidence; not could it be known what would be the system of reasoning to the jury, until the cause had so far progressed. If the plaintiffs’ counsel, in summing up, treated these written statements or memorandums as evidence, when in fact they were not so, the adverse counsel might have applied to the court for their opinion on this point; and in case the judge mistook the law, his errror in this particular might now be urged as a ground for a new trial. I deem it sufficient on this head to observe, that previous to overruling the deed of 1763 in evidence, it was not offered to show that it was inconsistent with the written memorandum, said to have been made by Thomas Say.
The law greatly regards the public safety and repose, and almost any thing will be presumed in favor of ancient possession. Length of time is encountered, in the present ease, by the consideration that Richard Dennis would legally hold the landed estate of his first wife Hannah, from the time of her death in 1761 until his decease in 1797, as tenant by the curtesy; and consequently it has little operation, to relax the *308general rules of evidence in the questions before us. When such circumstances are laid before the Court, as may induce *a reasonahle probability that deeds have been lost or destroyed, the inclination of my mind leads me to make every allowance, ■which the casualties attendant on human life may demand. But I do not feel myself warranted to depart from one of the most fundamental rules of evidence, “that no man shall be permitted to fabricate evidence in his own favor,” in order to obviate the supposed hardships of a particular case. It is of the highest moment, that there should, in the administration of justice, be no deviation from the established rules of evidence, which is not justified by evident necessity. After having given the subject every consideration in my power, I see no ground to alter the opinion I had formed on the trial.
My great object on the trial, was to bring all the matters of fact before the jury. I Avas satisfied that reasonable grounds had been laid before me to induce an opinion, that the recited deeds once existed, and that after using every effort in the poAver of the party, Avhose testator was once supposed to possess them, they could not be found ; but on these subjects, the jury were ultimately to decide, as questions of fact. The opinion of a majority of the court, as to the deeds overruled in evidence, supersedes the necessity of explicitly stating my sentiments as to the weight of evidence on either side. It avíII come on again, and other causes, we are told, depend on the same facts; and it is my wish that subsequent juries should be wholly unembarrassed by any expressions which may now drop from me, where there is no necessity for me to deliver my opfinion thereon.
Brackenridge J.
Where a deed is alleged to have existed and to be lost, and such evidence of this is given to the court as will justify the letting it go to the jury (in which case the party interested, and Avho may be supposed to have the custody of the deed, may be heard on oath, with other persons, as to the existence and the loss), the question will be, what evidence shall be heard by the jury ? Not the oath of the party, certainly, but that of disinterested witnesses. But the existence and contents may in that case be made out by the best evidence in the nature of the case, that, independent of the deed itself, can be produced. He may show as in this case, that having been tenant by the *curtesy, he was suffered to continue in possession after the death of the feme, for a length of time, without an adverse claim by those interested in succeeding to the estate. This of itself *309can weigh little ; but it is a circumstance inducing some presumption of the conveyance from the feme, which is alleged, and which may have been in the knowledge of those interested in succeeding, as it is only since the time of the loss alleged, that the claim on their part has been set up. But proof having been given which goes to the existence of the deed, it is assailed by testimony that the alleged grantee, the tenant by the curtesy, after the death of the feme, or after the alleged deed made, did make declarations of not having a title to the estate in question. Shall evidence be given to weaken this; that is, to weaken the persuasion of the jury in crediting the testimoiw that such declarations were made ? Doubtless; but what shall that evidence be? Shall his declarations at another' time, to other persons be evidence ? Shall this be considered as making testimony for himself? These declarations having been made at a time when he could not be supposed to be looking forward to a loss, which has subsequently happened, it would seem to me might be set up to rebut the testimony before given and to take away from the effect of it. A uniform disclaimer of property in that estate would be strong evidence against the alleged existence of a title; and claim of property on the other hand, certified by other persons, must somewhat weaken our belief of the alleged declarations that he did disclaim. It goes to the evidence of the fact that he did at any time disclaim. For though a declaration at one time is not inconsistent with a contrary declaration at another, yet it diminishes the probability that such a declaration was made. Be that as it may, shall not evidence of acts inconsistent with such a declaration be evidence, a correspondence with third persons, letters written at a time when the loss could not have been contemplated, relating to the estate in question and implying an ownership, be evidence to rebut such testimony of declarations? Be that as it may, shall not an act disposing of parcels of the estate in question to other persons, by deeds which recite the deed alleged to be lost, or refer to it, be evidence? This at a period before the time at which the deed is alleged to be lost, and respecting which *loss, proof has been already given. The recital of a deed is not evidence for him claiming under it; for the deed itself must be produced. But when the question is whether the original did exist, the being referred to in other deeds must weigh. For it would naturally occur to him contesting the deed, to ask if such deed did exist at the time you made the deed to A, B, or C, which was before the time of the alleged loss of the deed, how is it that in those deeds you did not recite or refer *310to this ? It rebuts all surmise of this nature, to show that he did recite or refer to-the deed in question. This evidence is of a secondary nature ; but the primary being out of the way, this is the best that in the nature of the case exists and can be given. It is a case where the matter alleged must be made out by circumstances, and I take it that a recital of the nature before, is evidence, taken in connection with what accompanies.
This leads to an inspection of the deeds in which the deed in question is recited ; and in which there is a reference to other matters, and by inspection some recognition of the deed in question by those who may be presumed to have had a knowledge of it. It is on this principle that a deed of the 30th April 1763, may be evidence; in which deed, the title which had been in Hannah Dennis as a devisee of John Coates is deduced, and her deed with that of the tenant by the curtes}’' Richard Dennis, of the premises, or her estate therein, to some person is recited; more especially as these devisees are parties to that deed, from whence some presumption if not a certain conclusion arises that they had knowledge of it. But what is more, when Thomas Say, who was the trustee under the will for all the children, is a party to that deed in which the recital is made of a deed to some person, and by that person to Richard Dennis, and when we hear of no other deed answering t.o that description but that to Carpenter and from him to Richard Dennis, which is the deed in question; and since we cannot but couple with this the memorandum in the handwriting of Say to the paper containing the appraisement of the real estate, which speaks of Thomas Say drawing for Richard Dennis, but when he came taking a copy himself, &c., with respect to this deed I can have no doubt but that it was rebutting evidence, if not' evidence fortifying that of the ^existence of the deed alleged to be lost, that might have been given even before any allegation of a disclaimer on the part of Richard Dennis had been offered, or testimony offered with regard to this. But on the ground of rebutting evidence to a disclaimer of title, I incline to think that the deed of 28th June 1769, in which the deed in question is recited to have been made, and to which a certain Coates, probably one of the family of Coates, is a witness, was also evidence, and ought not to have been rejected. On the same principle also, as showing the exercise of ownership, the deed of the 26th July 1791, Richard Dennis and second wife to Blair M’Clenachan, is evidence; for though this deed does not recite the deed in question, yet it contains *311a warranty against those whose claim might be adverse to it, the children of the first wife. And it is not to be presumed that Diehard Dennis, knowing that, he had not the title of the former wife in him, would undertake to warrant against the children who might claim under it; for any present consideration, to a prudent man, would not suffice for the damages that would be recoverable against him in ease of his failing to fulfil his warranty.
An argument may be deduced from the devise of Diehard Dennis of the 17th January 1784, putting his children by the first wife on a footing with the children by the second ; since had he considered them as succeeding to the estate of the mother, such provision might have been unnecessary. In this view of the case, it might have been offered for the defendant, had it not been brought forward on the part of the plaintiff.
With a view to a new trial, the opinion of the Court is called for with respect to the evidence first offered in this ease on the part of the defendant, the memorandum in the entry book of the recorder of deeds for the county where the lands lay, purporting that deeds of the nature in question, were lodged to be recorded. The evidence of this memorandum or entry, was a copy from the book sworn to be a copy, and this by the deputy who for the time being kept the books of the recorder, and made the entry. The certificate of the person in whose possession the book at the present time is, but in whose official possession it is not, for ^though in his office it is not o/his office, was not evidence, and therefore properly detached and rejected.
A more important and difficult inquiry on the former, and will be on the new trial, and on which the opinion of the Court now as at present advised would seem to be expected, is with respect to the exemplification offered of the deeds in question under the certificate of the recording office of the state of Delaware; for the real estates of the feme lying in different states, that of Pennsylvania and that of Delaware, it became necessary that., the deeds vesting the estate in the tenant by the curtesy, Diehard Dennis, should be recorded in both states, so as to warrant the giving an exemplification of the deeds. The testimony of the deputy who made the entry in the book, with a view to a record in Pennsylvania, was given as to the appearance of the papers, and what they purported to be, and that the probate was such as to entitle them to be put upon record in Pennsylvania. But it appears that such evidence of being the act of the feme did not exist, as by law did entitle the deed alleged to *312be by her made, to a recording in the state of Delaware. There was an acknowledgment or probate before a chief magistrate of Pennsylvania which entitled it to be put upon record in this state; but in the state of Delaware the acknowledgment of the feme taken in open court was a prerequisite. Hence it is that these writings on the book of the recorder cannot be known as records, because there, not under any law, they can be no more than as on the private book of him that was recorder for the time being, for which reason no statute proof can be given of them, as to their existence or contents. The exemplification and certificate are therefore out of the question. "Whatever evidence can be given respecting them must be on common law principles. The evidence offered can be no more than that a person who at the time purporting to have recorded, or correctly speaking, copied, was the recorder, and has left a book which in his handwriting or that of a deputy, contains writings copied purporting to be such deeds as are alleged to have existed, and of which copies are offered by one who had inspected this book, and taken copies. If such evidence is admissible, I take it the proof of such person having been a officer the time must be a certificate *under the state seal. But as to such book being in his office at the time when, or now found in it, and as to his bandwriting or that of a deputy, it must be on the oath of some one who shall inspect the writing. Even this would not be evidence with regard to the giving publicity, and affecting a purchaser; but as writings found in a certain place and under circumstances by which they may be presumed to be there for the purpose of being recorded, from whence may arise some presumption of their existence at that time.
On a new trial if it is ordered, it will lie with the defendant, if he shall think it necessary to support the proof of the existence and contents of these deeds, to consider what common law proof of this nature he may think it advisable to offer; and in what manner to procure it, by commission or otherwise; the distinct and particular consideration of the admissibility of which may be taken up, if on a second trial the verdict should be for the defendant.
I have used the term secondary evidence; but to explain myself, the •whole of the evidence in the case of a lost writing, is secondary in relation to the writing lost. But evidence being given on the part of the plaintiff in answer to this secondary evidence on the part of defendant, the rebutting evidence given by the defendant in answer to this, is *313secondary with respect to it. But had the defendant been the plaintiff in this case, I can have no doubt but that after the evidence which regards the recording or attempt to record, he could have given these recitals in the deeds made by himself as evidence of the existence of the deed recited ; and this as distinguished from rebutting evidence, might be called primary.
New trial awarded.
[Cited in 13 S. & R. 124; 1 R. 95; 1 Ash. 275 ; 2 id. 49 ; 10 W. 84; 8 W. & S. 192 ; 4 C. 425. In Murphy v. Lloyd, 3 Wharton 549, the court, declared that in Garwood v. Dennis, the facts were so complicated that the decision could not be regarded as an authority, except in a case precisely like it.]