*373The opinion of. the court was delivered by

This was an' action of ejectment for a tract of land in Cambria county, containing four hundred acres, brought by the plaintiff in error against the defendants in error, in the Court of Common Pleas. The defendants obtained a verdict and judgment, but three exceptions to evidence, and seven to the charge of the court were taken by the counsel for the plaintiff.

The plaintiff claimed under a warrant to Robert Irwin, dated January 22,1794, and a survey on the same on the 24th of May, 1794; and, having given the said warrant and survey in evidence, he offered a paper which he supposed to be official. This paper contained a draught of the survey and a copy of the field notes, and there were two indorsements on it.' One of the indorsements was, “ Fees paid T. V.;” the other, “sent to Mr. OKeiffe, deputy surveyor of Cambria county, December 8th, 1807:” (signed) J. JLnderson, D. S., B. C.” It was proved, by the oath of Isaac Proctor, deputy surveyor of Cambria county, that the paper was returned to him by William O’Keiffe, formerly deputy surveyor of the said county, and that the words, “ Fees paid,” appeared to the said Proctor, to be the handwriting of George Woods, formerly surveyor of Bedford county7, but the witness had never seen him write, and knew his handwriting only by.comparison. Proctor proved, that the rest of the indorsement was in the handwriting of Doctor John Jlnderson, formerly deputy surveyor of Bedford county. This paper was objected to by the defendants’ counsel, and rejected by the court, whereupon the counsel for the plaintiff excepted to their opinion. This paper was liable to various objections. There was no certainty of. its being an official paper. Doctor Jlnderson certified that he sent it to O’lleiffe, but he did not say where he found it, or that it had ever been in the office of the deputy surveyor of Bedford county. He should have been examined on oath, and then the defendants would have had an opportunity of cross-examining him. The handwriting of George Woods was not well proved. Comparison of hands alone is not evidence, except in case of public officers, who have been so long dead that better proof could.not be expected. But there must be living persons capable of proving the handwriting of Mr. Woods. I am of opinion, therefore, that the court was right in rejecting the evidence.

The second exception was to the rejection of the deposition of William Clark, offered in evidence by the plaintiff To this evidence the counsel for the defendants made several objections. First, that it did not appear that notice of the taking of the deposition was served on all the defendants. This objection might be got over. John M‘Donald, who proved the service, swore, that he served copies of the notice, which may be intended, that he served a copy on each defendant. The service was proved by. the affidavit of M‘Donald, and if all the notices were really not served, *374the defendants might have brought the witness into court, and examined him. But there were other, and fatal objections to this deposition. The notice was, that it would be taken “at tbe house of William Clark, in the county of Bedford, and township of Napier.” Now, it did not appear that it was taken at the house of William Clark, by the certificate of the justice before whom it was taken, or by parol evidence; either of which would have been sufficient. The justice certifies, that it was sworn and subscribed by the witness, before him, on the 28th day of June, 1821. But where the deposition was taken does not appear, except that it' was in Bedford county, which may be inferred from the words, “Bedford county, ss.,” at the head of it. This point was expressly decided, in Selin v. Snyder, 7 Serg. & Rawle, 166. There the notice was, that depositions would be taken “at the house of Jldam Weaver, inn-keeper, in the borough of Lancaster.” Depositions were taken before a justice of the peace of Lancaster county on the day appointed, but it did not appear where they were taken, except in the county of Lancaster. It was held by this coui't, that the depositions so taken, were not evidence; that it was incumbent on the party who offers a deposition, to prove that it was taken accoi’ding to the notice, unless the adverse party attended, io.whieh case the defect would be cured; that the defendants might have proved, by parol evidence, that the depositions were taken according to the. form of the notice, though not so certified by the magistrate. But no such proof having been given, it did not appear that the deposition was taken at the house of Jldam Weaver, and therefore it was not evidence. That case is not to be distinguished from the one now before us. The deposition of William Clark, therefore, was properly rejected.

Another deposition of Clark’s was also offered in evidence, which was taken at a former trial. There was no error in rejecting this, because the parties were not the same as in this cause; nor did it appear that Clark was not living, and within the jurisdiction of the court. Indeed there was no doubt that he was living, and within its jurisdiction.

The third exception was to the rejection of a connected draught of eleven tracts of land, offered in evidence by the plaintiff. This draught came from the office of the surveyor general, by whom it was certified, “that it was a connected draught .of eleven tracts of land situate on the waters of Otter creek, or little Conemaugh, in the county of Bedford, surveyed on the warrants granted to the persons whose names are written on the.plots respectively, all dated the 23d of July, 1773, except that to Robert Irwin, which is dated the 22d of July, 1794.” Such a draught would not have been evidence to make title to the land in dispute. In order to make title, the return of survey should be produced, as well as the authority under which it was made. But where the object is only illustra-*375Hon, as in the present case, to show that Robert Irwin?s survey did not interfere with the others, it has been usual to admit this kind of evidence. It is very convenient, saves considerable éxpense in obtaining a copy of each survey, and is attended with no danger. It is true that this is not a certified copy of any draught of record in the office of the surveyor general, and it was objected, that to admit it in evidence would be contrary to the act of the 31st of March, 1823, (Purd. Dig. 258, Pamph. 233,) by which “copies of all records, documents, and papers, in the office of the secretary of the commonwealth, secretary of the land office, surveyor general, auditor general, and state treasurer, when duly certified by the officers of the said office respectively, shall be received in evi-den'ce in the several courts in this commonwealth, in all.cases where the original documents and papers .would .be admitted in evidence.’.7 But it is certain, that this act, so far as concerned the office of the surveyor general and secretary of the land office, was but affirmative, because certified copies of the papers in those offices, had always been admitted in evidence before. It has been several times held by this court, that a certificate from one of those officers, that a certain paper, for example a return of survey, was not in his office, might be received in evidence. This practice had been established before I came on the bench, and therefore I yielded to it, though not strictly reconcilable with principle. It must be granted, too, that a connected draught, certified by the surveyor general, is not the best evidence the nature of the case admits of, because you might produce certified copies of each survey, and the warrant under which it was made. But we know, that, to a.void inconvenience, the rule which requires the best evidence, has been often relaxed. We. are in the constant practice of receiving as evidence, papers which require the certificate of a judge, or justice of the peace, with no other proof of the person’s being a judge or justice than his own assertion. Yet his commission, or a certified copy of it, if recorded, would be better evidence. I am aware of a decision of the Circuit Court of the United States, in the case of Griffith’s Lessee v. Tunckhouser, (1 Peters, 421,) that a connected plot of different tracts made out and p.ut together in the land office, was not evidence, because it did not profess to be a copy of any plot on record in the office. My great respect for that court raised the only difficulty I have felt on this question. I should be disposed to follow its precedent, were it not that I am satisfied it is contrary to the practice which has prevailed in our courts, and that practice, sanctioned too, as it seems, by the opinion of the legislature. For in the fee bill of the 28th of March, 1814, (Purd. Dig. 280, 6 Sm. L. 228,) a fee is allowed to the surveyor general “for connecting separate draughts into one general draught; for each separate draught therein — twenty-five cents;” and another fee of twenty-five cents, “for certifying and affixing the seal of office to the same.” I cannot help supposing *376that these fees were given, because it was known that such draughts-were often wanted, to be given in evidence in courts of justice. The expense of a certified copy of each separate survey, and the warrant or application under which it issued, would, in many cases, be very considerable; and there is little danger of the officer’s certifying a falsehood, in which he would be open to detection from his own records. I am therefore of opinion, that there was error in rejecting the evidence.

The remaining exceptions are to the charge of the court; and, to judge of them, we must keep in view some of the principal features of the evidence. Robert Irioin’s warrant, under which the plaintiff made title, dated the 22d of January, 1794, was for four hundred acres, “in the forks of the north and south fork of little Conemaugh, in the township of Quenaboring, Bedford county.” The survey on this warrant was made on the 24th of May, 1794, and a copy thereof certified on the 4th of February, 1804, by the then deputy surveyor, and addressed to the then' surveyor general, by whom it was accepted. The defendants claimed under an improvement commenced in the year 1797, or 1798. The defendants also, gave in evidence two warrants, one in the name of Hugh Rogers, the other of Randal MlJLllister, dated in the year 1773, on which surveys were made in 1784; and after showing other warrants and surveys adjoining the former, they endeavoured to prove that these two surveys comprehended great part of the land claimed by the plaintiff. They gave in evidence, too, the copy of a letter from the plaintiff to Dr. William Smith, deceased; and parol evidence wa’s given by both parties in support of these pretensions.

The President of the Court of Common Pleas charged the jury to the following effect: — “ If the warrant on which the plaintiff’s survey was executed was vague and indeseriptive, it ought to have been returned in a reasonable time, or a sufficient reason given for the delay. Here is a survey on a loose warrant, not returned by the officer to whom it was directed, but, nearly ten years after-wards, certified by his successor, on the representation of the plaintiff himself, that there were no interfering claims; although, from his letter to Dr. Smith, he knew there were, and also knew that the land was settled upon by the present defendants. It appears further, that he had relinquished his claim under the survey to Dr. Smith, and had declared to others, he had no claim to any land on that side of the creek where that survey lies. If you believe all this, the survey ought not to attach, until actual return into the office; and, in that case, it gives no right to the land in question, if, as is the undisputed fact, the defendants had six or seven years before the return and acceptance of the plaintiff’s survey, entered upon the land as vacant, and, by actual residence and substantial improvements, acquired an equitable title.”

I perceive no error in this direction. The law, with respect to *377warrants of vague description, was accurately laid down; although, in speaking of laches in returning a survey, it would have been better if the jury had been told, that the owner of the warrant is not to be prejudiced by the neglect of the deputy surveyor, leaving it to them to decide whose the neglect was. It was left to the jury; for the judge, in the conclusion of his charge told them, “that if they believed the evidence adduced for the defendants, and were satisfied that the plaintiff had unreasonably delayed having his survey returned, the verdict should be for the defendants.”

The plaintiff’s counsel complained particularly of that part of the charge in which the jury were told, “that the plaintiff had relinquished his claim to Dr. William Smith.” That was a matter of fact, of which the jury would judge from the evidence, and, if the judge was mistaken in his opinion on that point, they were not bound by it, but might correct him. Indeed the judge did directly afterwards submit this, among other facts, to the jury. It was not an error in law, and so we have freqqently determined. The observations I have made will apply to the plaintiff’s first five exceptions to the charge, with one additional remark, that several of the exceptions are not'supported in fact, because the judge is supposed to have said what he did not say, or what he said only hypothetically. But, besides these general exceptions to the. charge, the plaintiff has assigned error in the judge’s answer to' two points specially proposed to him.

1. It was charged, “ that if the late Judge Smith (then a deputy survej-or,) executed the warrants of Rogers, &c. in 1773, without returning them, new surveys might afterwards be made, on vacant lands, without a new authority, provided the warrantees agreed to accept them.” If a survey is returned,, there is an end of all authority under the warrant, and no new survey can be made, nor any alteration of the return, without new authority from the land office. But, until a return is made, the general principle is, that although a survey be made, it may be altered by consent of the warrantee. It has always been held, that a lost warrant may be laid on any vacant land, and that the owner of an application may shift his ground, and take up any vacant land, provided nevertheless that a third person, who has acquired a right of appropriation, shall not be injured. These principles will be found in the cases of Healy v. Moul, &c., 5 Serg. & Rawle, 181, Steele v. Finlay, 2 Sm. L. 255, and Hepburn v. Levy, 4 Dall. 218.

The title, under a shifted application, does not commence until the return of survey, except against one who has notice of it bef@re the return, and then it commences fronrthe time of notice. Kyle v. White, 1 Finn. 246. Bond v. Stroup, &c., 3 Binn. 66. There is little or no difference between an altered and a new survey. There is certainly no difference as to the commonwealth, and, as to third persons, they are protected. If they have appropriated the land, they will hold it against either an altered or a new suiVey. *378In the opinion of the court delivered by Judge Duncan, in Healy v. Moul, (5 Serg. & Rawle, 187,) it is said, “I do not question the right of an owner to have his lines extended, if his warrant be not filled; nor do I doubt of the right of the deputy to alter his original survey, to throw out, or take in other lands, provided this is done shortly, and does not interfere with mesne rights between the time of the original survey and the alteration before the return. But when the survey is returned, the deputy is functus officio, and it requires a new authority from the surveyor general, or the Board of Property, to warrant such alteration.” Considering, then, the principles thus positively established, and the analogy between them and the case of a new survey, I am of opinion that the charge of the court on this point was right.

The other proposition submitted to the court was this: — “The survey of Hugh Rogers, if a shifted survey, is to be considered as subsequent to the survey of Robert Irwin, not being returned before the bona fide survey of the plaintiff.” As to this, the court charged, “The survey on High Rogers’ warrant, if a shifted warrant, but of which, as has already been stated, there is no satisfactory evidence, would giveiio title until actual return.. But whether the survey on the plaintiff’s warrant be, what the question presupposes, a bona fide one, and prosecuted with due diligence, is submitted to the jury on the evidence before them.” I perceive no error in this answer, or at least none which was prejudicial to the plaintiff. The plaintiff had no right to ask the court to assume the fact of his survey being bona fide; and, without the assumption of that fact, the question of law did not arise. Neither could the plaintiff recover, if the jury should be of opinion that' his survey was not prosecuted with due diligence, because then the right of the defendants under their improvement and settlement would prevail, and it would, be immaterial whether Rogers’ survey took preference to Irwin’s or not. 1 am of opinion, on the whole, thát there was error in rejecting the connected draught offered in evidence by the defendants, and in no other part of the record.

• The judgment is to be reversed, and a venire facias de novo awarded.