The present contest lies within narrow limits; but it is proper that the jury should be intrusted in a few plain legal principles. A plaintiff in ejectment must show a good right, or that he is entitled to the possession. The rules as to warrants and applications conferring a title, I take to be these. If the description of the land contained therein point out specially and exclusively certain lands with accuracy, the right vests immediately, provided it is duly followed up. But if it be general, and may equally suit several tracts, the right only vests from the time of making the survey, which reduces the pretensions of the party to certainty. If the location be shifted or removed, namely, if the survey is made on lands different from those applied for, the title takes effect from the acceptance of the survey in the surveyor general’s office in general, and as to those in particular who had knowledge of the survey, from the time of such knowledge received. These rules are founded in justice and sound reason.

The late proprietaries were the owners of the soil, and disposed of their lands by their public agents. Having once sold certain lands, their rights therein ceased, and while the vendee honestly performed the conditions of sale, neither they nor their agents could sell them a second time. Hence arises the respect due to priority of title. When the contract between the public agents and an individual is reduced to absolute certainty, the right becomes fixed and the lands are no longer vacant as to other appliers. But in the case of a shifted survey, the officer having no regular authority to do the act, the proprietaries were not bound b}7 the survey, until their agents recognized it by it in their *office, and then the contract became complete. No one was injured thereby. An intervening right was not affected in the smallest degree. It united the interests of the proprietaries and the people, it saved expense, and subsequent appliers, by using due diligence, could knpw what lands had been previously appropriated.

Apply these rules to the case before us. The plaintiff’s warrants were earliest, and so were the surveys thereon, if really and in truth they were made. John Hubley’s warrant calls for land on the waters of Spring creek within two or three miles of Nittany mountain. Michael Hubley’s calls for land adjoining John. The former was the leading one, and falls under the class of a general, indescriptive warrant, and not a shifted one. The defendant’s is more properly a shifted one, because Joseph Cooper’s calls for the black oak *51ridge, which lies more than three miles west of the end of Nittany mountain. Lippincot't’s warrant adjoins Cooper, and Gill’s joins Lippincott; but both were to include part of the black oak ridge, which must have been a notorious place at the time.

The real question here turns, not upon the locality of the different warrants, but whether surveys were really made on the plaintiff’s warrants. If in fact they were executed on the lands in question, the plaintiff is entitled to recover. Now what more evidence can reasonably be expected of the reality of a survey, than we now have? Here is a return of survey, which is prima facie evidence of a survey, though not conclusive.- It may be impugned by other evidence. But here we have the certain language of nature, that the two courses N. 56° E. 281 perches,-and N. 38W. 281 perches and 240 perches, present regular marked lines, corresponding at the time when the trees were blocked, with the surveys in 1774. It is also sworn by M’Kinney, that ten years ago, he saw some pointers near the black oak now gone, the common corner of the plaintiffs survey, and Ludwig Karacher. Ve have still more. The official return of survey made by Charles Lukens for Karacher, said to have been made in 1775, calls for John Hubley on the course S. 37° E. 240 perches from the post to the black oak. These form a strong mass of evidence to prove a survey on the ground. A considerable part of the wood has been cut down for the iron works of Benner, and the land is cleared where part of the Nines ran. Ve cannot therefore be surprised that all the lines marked have not been discovered. The sur- ^ veyor and chain carriers have probably paid the common debt of nature.

It was the duty of the surveyor to make return of his survey ; but in the view I have taken of this matter, it is wholly immaterial when the returns of survey were made. The contracts were complete when the surveys were made, and the proprietaries as well as the appliers were bound thereby. They could not sell to others, lands which were no longer their own to sell. I cannot regard Benner as a bona fide purchaser without notice. Exclusively of the marked trees apparent upon the grounds, the patents to Ludwig Lauman were issued on the 12th January 1792, and the deed from Matlack to Benner was dated as a subsequent day, on the 2d May 1792. But the rule does not apply here, because each of the parties claims a strictly legal estate by patent.

The doctrine of settlement and improvement has nothing *52to do with this case. No such claim can hold on lands appropriated to individuals ; and all this land was cleared after the date of the patents issued to Lauman. The building of Benner’s house was in 1793, and this suit was commenced to August term 1802.

The sole point to be determined by the jury, is whether the plaintiff’s warrants were really and in fact executed by surveys thereon ; and if the jury are satisfied in the affirmative, they ought to find for the plaintiff. But if they have solid grounds to decide their judgments in the negative, then they should find for the defendant.

The jury found a verdict for the defendant; and his Honor having granted a new trial upon the plaintiff’s motion, the defendant appealed to this Court, where the case was now argued by the same counsel who were concerned in the Circuit Court.

The plaintiff claims under a warrant for 300 acres “ on the waters of Spring creek within two or three miles of Nittany mountain,” dated 27th April 1774. The purchase money was paid to the receiver general, 29th April 1774, and the plaintiff gave evidence of a survey made 21st May 1774. A patent was not obtained till 12th January *1792. At what time the survey was returned, does not certainly appear. It is unfortunate, that in the proprietary offices, there was no record of the time of return of surveys. The only evidence to be had of it, is in a book of accounts kept by the surveyor general, in which it was his custom at the time the return came to his hands, to chai’ge his deputy who made the survey, with a certain fee due from the deputy to himself. Thex-e is reason to think, that this book contains very impei’fect evidence,pax’ticulaxdy with respect to surveys retunxed by Charles and Jesse Lu-kens, sons of John Lukens the surveyor genei’al. The survey in question was made by Charles Lukens, and there is no mention of it in his father’s book. There is an indorsement on the original return, from which the defendant infex’s that it was made 7th December 1791. The indorsement is “fees paid 7th December 1791and undexmeath there is another indorsement without date, “acceptance fees paid by P. Miller, 5s. 6c?.” Nothing conclusive can be drawn from the indox’sement, because it does not appear at what time P. Miller paid the acceptance fee.

The defendant makes title under a warx’ant 5th December 1774, for “ 300 acres about two and a half miles from the end *53of Nittany mountain, to adjoin a survey of Retíben Haines, to include part of the black oak -ridge in Northumberland county, in Bald Eagle or Potter township.” A survey was made by Charles Lukeus in October 1775, which was returned 25th March 1-776. A patent was issued on this survey, 13th June 1776.

The plaintiff gave strong evidence of a survey made in 1774. Two of the lines were traced by marked trees on the lines, and by one corner tree; the other lines ran chiefly through land which is now cleared. On another line the plaintiff’s survey is proved by an adjoining survey calling for it, in the year 1775. On the whole, the evidence was so satisfactory as to leave no reasonable doubt of the survey having been made in the year 1774.

The judge of the Circuit Court, before whom this cause was tried, directed the jury to find for the plaintiff, in case they should be of opinion that the survey was actually made ; but if they were not satisfied of that fact, to find for the defendant. The jury found for the defendant. The judge *ordered a new trial, and from that order there was an appeal to this Court.

The weight of evidence in favor of the survey was so great, that there ought to be a new trial, unless it clearly appears, that there is some principle of law destructive of the plaintiff’s title. The defendant’s counsel assume it as a fact, that the plaintiff’s survey was not returned till the year 1791; and contend that in point of law this survey should have no effect till the time of its acceptance by the surveyor general. They state their case too strongly, in assuming the year 1791 as the time of return of survey; for in truth, the time of the return is unknown. Nor can I agree with them in their legal inference, supposing they were right as to the time of the return. Warrants or locations may be divided into several classes. The first are descriptive of the land intended to be surveyed, either precisely or with such reasonable certainty, as is sufficient to designate it. If due diligence is used in obtaining a survey, the title under these warrants attaches from the date. The second class comprehends those, which give .but a loose description of the land applied for. I take the warrant in question to be of this nature. It fixes no precise spot, but allows a scope of several miles. The •title under such a warrant cannot attach until the survey is made, for till then it cannot be certainly known where the land lies. I believe that a very large proportion of the locations entered on the opening of the land office in the year 1769, was of this class. Great allowance must be made for *54looseness of description in a country very little explored. The third class consists of what are called shifted warrants or locations; that is to say, where the survey is made on land different from that described. These surveys have no validity, except against persons having notice of them, until the time of their acceptance in the office of the surveyor general, for very obvious reasons. Being an alteration of the contract, they can have no force, till accepted, because the proprietary was under no obligation to keep the land vacant, and may have suffered some other person to appropriate it before the return of the shifted survey. Indeed it was only by courtesy and the custom of the land office, that such surveys were accepted at all. It is evident, that loose warrants and shifted warrants .are of a very different nature. If the *proprietary thought proper to sell a man 800 acres of land, to be surveyed on a certain water within a compass of two or three miles, in such place as the purchaser should direct, the appropriation should be considered as complete from the time of the survey made, unless something improper afterwards took place on the part of the purchaser. The survey was made by the officer of the proprietary, whose duty it was to return it to the office of the surveyor general, for so he was commanded by the warrant. If he neglected his duty, the purchaser is not to be injured by that. Much has been said by the defendant’s counsel, about an implied abandonment by the plaintiff. But to presume that a man has abandoned a tract of land, which he has paid for, is against all probability. I do not say, that there is no instance in which a title may be affected by the non-return of a survey on a loose warrant. It is possible that one, who has trusted the surveyor with the selection of the spot, may dislike it, after the survey made, and refuse to pay the reasonable expenses, which may be the cause of the delay of the return. A case may arise, where a man knowing that his survey is not returned, permits the land to be surveyed for a third person, and suffers him to settle on it, and improve it, without giving notice of his survey. In short the general rule admits of many exceptions, founded on some default or improper conduct of the warrantee. But I see nothing in the present case to make it an exception. When I speak of a general rule, it is implied, that this is not a new case. I heard the law laid down in this manner at Nisi Prius in the case of Hubley’s Lessee v. Chew and others, Northumberland county, October 1796, in which I was of counsel for the defendants; and I find from the manuscript notes of the late Judge Smith, which are in my possession, that it *55was ruled in the same way by the late Chief Justice M’Xean and Judge Smith in Dixon’s Lessee v. Moorehead, Westmoreland county, May 1798, and by Judges Yeates and Smith, in Bogg’s Lessee v. Silvas, Westmoreland county, May 1799. That being the case, I am not for departing from it. The only argument against the rule I have mentioned is, that third persons may be affected by it. But that cannot often happen, and scarcely ever, without dishonesty and negligence in the deputy for if land has been surveyed on a loose warrant, it must or *ought to be known to him. But if the deputy surveyor is dishonest, very great mischief may flow from making the return of survey the commencement of the title; for after the warrantee has done every thing in his power, by putting the warrant in the hands of the deputy, and procuring a survey to be made, he may be cut out of his land by the roguery of the officer in omitting to make his return. Establish the rule how you will, inconveniences will arise in particular cases. It has been settled on deliberate reflection, by persons well acquainted with land affairs both in theory and practice. Very little hardship can arise from adhering to it, but great evil may be introduced by disturbing it. I am therefore of opinion that the judge of the Circuit Court was right, in ordering a new trial.

It would seem to me, that it may not be unadvisable that it be submitted to another jury, to consider the fact of the survey in question having been made upon the ground. At the same time, I cannot say that I am prepared to acquiesce in the sentiments of the judge before whom the cause was tried, that this was the only point to be submitted to the consideration of the jury. The return of survey, I take to be material; for the warrant, in this case, was not of that precise nature, that it could attach to a particular spot, exclusive of all others. At least, this is a matter which might also be left to the consideration of a jury. It is possible that more than one survey of 800 acres might lie on the waters of Spring creek within two or three miles of Nittany mountain; and if so, there could not be such an agreement of the survey with the warrant, as to give notice to all the world of a survey made under that warrant, upon that ground, without a draught of that survey returned into the office, and attached to the warrant, as in contemplation of law it ought to be. Notice to the proprietary office, and notice to the subsequent purchaser, is necessary ; because otherwise, the office being kept open, and supposed to be for *56the purpose of examining, great deception must ensue to both. I cannot therefore sanction the idea that where actual notice is not brought home to a subsequent purchaser, of the prior equity of a survey, constructive notice by a return is not material; in this case at least, where the warrant would *not seem to exclude a survey under another warrant on the waters of Spring creek within two or three mies of Nittany mountain. But with this exception to the charge of the judge who granted the new trial, I am contented to affirm his judgment.

Judgment affirmed.

[Approved in 2 S. & R. 398, 400. Cited in 6 S. & R. 133 ; 10 id. 216 ; 12 id. 129; 1 P. & W. 78; 5 W. 524; 8 id. 120 ; 4 W. & S. 74; 9 C. 479 ; 7 Wr. 203.]