delivered the opinion of the court.

That the plaintiff in ejectment is entitled to recover upon full proof of title, and an adverse possession by the defendant, at the *388time of commencing the suit is a legal proposition not susceptiblé of controversy. It is not every refusal, however, to state legal principles to a jury, however clear in themselves as abstract doctrines of the law, which will be decided to be error. It is sometimes very hazardous to lay general principles of law before the jury, where they cannot be directly applied to the facts in proof. It has been uniformly decided, therefore, that it is not error in the judge to refuse instructions unless the party shows the existence of proof to which they can be fairly applied. Before we can decide that the court below was wrong in refusing this instruction, it will be necessary to examine, in connection with it, the facts of the case, and this will be done in determining upon the propriety of the second instruction asked for and refused. The questions embraced in this instruction are essentially all that arise out of the whole record, and in disposing of them we shall necessarily decide this cause. ,The principles involved in the above charge asked for by the plaintiff, are within the established rule upon the subject of boundary. The act of congress of the 11th of February, 1805, provides, that the boundary lines of sections which shall not have been actually run and marked as required, shall be ascertained by running straight lines from the established corners to the opposite corresponding corners, but in those portions of the fractional townships where no such opposite corresponding corners have been or can be fixed, the said boundary line shall be ascertained by running from the established corners, due north and south, or east and west lines as the casé may be, &c. If there was, then, no proof before the jury, that the division line between the sections had been actually run and marked on the ground by the original survey, the line could be ascertained in no other legal mode than that pointed out in the act of congress referred to. Whether it had been run and marked so as to close all further inquiries into tbé question of boundary, was certainly a question for the jury. And it might very properly have been submitted for their determination. With a view, however to determine, fully and satisfactorily the whole question involved in this assignment of the errors, we will consider it in connection with the decision made by the judge, upon the legal effect of the dotted line of partition in the original plat of survey. *389That decision was, that this line being part of the map and so appearing upon its face, could not be corrected after a sale of the land, though made in mistake. The rule on this subject is well settled, and is uniform in all the cases which have been adjudicated where boundary was to be ascertained. The survey is to be taken as part of the patent. It is the source of title, is a matter of record, and may, therefore, be resorted to in order to control the calls of the patent. A consequence of this principle is, that if the plat and certificate of survey show an artificial or natural boundaries, though they may vary from the course or distance called for, they will nevertheless be taken as the true boundaries of the tract, if they can be well ascertained as described in the grant. 1 Marshall, 96; Pirtle’s Digest, 125. The reason and policy of this doctrine is well explained in the case of Hubert and Wife v. Wise and Others, 3 Call. Rep. 238. Judge Pendleton says, “ the marked trees upon the land remain invariable, and are to govern as to the boundary. Such lines, therefore, when proved, are never suffered to be departed from. If the true line according to course and distance, called for in the plat and certificate of survey, depart from the line proved to be actually run, and evidenced by marked trees or other natural or artificial monuments, the latter must prevail;” and he gives as the reason, the liabilities to mistakes by the surveyor, sometimes putting north for south, east for west, or in copying the descriptions into the patent. It would be highly detrimental, therefore, if a mistake in the calls of a patent might not be corrected by reference to the plat and certificate of survey. In the case of Throop v. Cheesman, 16 Johns. Rep. 264, it is said by the court that the corner of lots as fixed by the surveyor general cannot be disregarded. In the case of Lyon v. Ross and Wife, 1 Bibb’s Rep. 467, the same doctrine is fully stated and illustrated. The court say, when a line has been ascertained and marked by natural or artificial objects, it is to be considered as the proper boundary, though found to deviate from a rectilinear or mathematical line. That case was precisely analogous to the one before the court. Both parties claimed under deeds from the same grantor. The farms lay adjoining, and the question was purely one of boundary. The partition line called to run from corner to corner without any intervening object. By a plat *390filed in the cause, the spring which was claimed to be on the land of.the defendant was represented to be one pole from a straight line according to the calls in the deed. But the proof was full that the line of partition which was run by the grantor, passed through the middle of the spring and was so run purposely to give both farms the use of the water. And on this proof the court decreed in favor of the complainants to have the quiet and undisturbed use of the spring.

The general rule is as stated and relied on by the counsel for the appellees, that the actual or visible boundary, whether natural or artificial called for in a certificate of survey is to prevail so long as it can be found or proved. The legal presumption is, that the surveyor has done his duty by marking and bounding'the survey. And though this presumption can be destroyed by undoubted testimony, yet, as this was the fault of the officer of the government and not of the owner of the survey, he ought not to be injured. This being the case, the present inquiry is satisfied by determining whether the line of partition between the sections in this case, was actually run and marked by the original survey. The question of boundary is to be decided, like all others, by proof, and parol evidence is as much admissible to prove boundary as :any other fact. We will, therefore, consider this subject in reference to the testimony furnished by the orginal plat of the survey, and by the witness who testified in regard to it. Does the map per se, furnish us with conclusive evidence that the line was run? If so, the controversy is at an end, for it cannot be disregarded, nor can it now be corrected. What are the indications on the plat, that this was done? The dotted line from the corner which runs northwest and intersects the Flowers tract, is the evidence relied on by the defendants. Is this to prevail in contravention of the established and admitted fact that it is a wide deviation from the true line as ascertained by actual survey? Is it to control the acknowledgments of the parties, and the proof of witnesses? It may be answered that it certainly must, and according to settled rules should do so, if it is to be considered as a part of the plat and certificate of survey. Is every trace or line upon the map to be regarded as an essential and constituent part of the survey, in the absence of any collateral or concurrent testimony of the sur*391veyor to designate its office? Does every lineament of the plat speak a certain and well defined language? What is spoken by the dotted line in this plat? We are not authorised to say what office it performs, because there is nothing in the certificate of the survey itself, which explains it. But it may be said in answer to all this, that the purchaser is not to be prejudiced by difficulties of this character. That be bought on the faith of the survey, and that the line of boundary indicated on the plat is to be preserved at all events. But this would only lead us directly back to the question of the intrinsic proof furnished by such a line. Looking at it as a naked trace of the pen, one who is not conversant with the structure and figure of maps of this kind and the uses of the various lines of the diagram would be able to form no opinion perhaps. But if it should so attract attention as to excite a curiosity to find out its use and object, an inquiry would be very naturally made of the surveyor himself to explain it. Its meaning would, therefore, depend on his explanation. As a technical question of science, its design upon the map is to be determined by the opinion of those whose skill in the science enables them to decide. The testimony of M’Fail is, therefore, to be regarded on this subject. He was one of the deputy surveyors of the United States. He says that this dotted line from the corner at B is not designed to fix the boundary. That it is no line. That the dotting of it is evidence that the partition or boundary between the two sections never was run and marked on the ground as required by law. And this is so understood, he tells us, among all surveyors. This must also have been the view taken of it by the surveyor general, who ordered M’Fail to make the survey returned by him and approved by the surveyor general. We take it, therefore, that the dotted line upon the face of the original map contains no intrinsic proof that the boundary indicated by it was ever made. Is there then any extrinsic proof that it was ever run? There is none furnished us. On the contrary all the evidence on the record shows that it never was run. The surveyor M’Fail, states that he searched diligently for the line but could find none, and that none had ever been run and marked on the ground by any one until he did it. This proof is fully calculated to destroy the force of the dotted *392line, if it did not carry its own explanation upon its face. For even in cases where the line has been run, if there were no objects on its direction to give it locality, or where those that existed have become extinct, it must then be ascertained by the course called for.

If, therefore, we adopt the interpretation of the dotted line, which is contended for by the appellees, and take it to say that the boundary was closed by the original survey, by what criterion is its identity and locality to be ascertained? Where are the natural or artificial objects by which it was ascertained and marked? By what monument can it now be traced? How is it to be fixed? Where are the marked trees to indicate its direction? M’Fail tells us that there are none. It is true that the line intersects the Flowers tract, but there are no monuments or objects, on any part of the boundary of that tract, by which we can fix the point of intersection. If then the line was run, we have no means of ascertaining its locality. It is, therefore, in reason, necessity and principle, the same as if it never had been run. This distinction is fully sanctioned by the decision of the court in the case before noticed of Lyon v. Ross and Wife, 1 Bibb, 238. The line of division must not only be run, but it must be marked on the ground, to bring it within the general rule of a closed survey. We are therefore clearly of the opinion that the court below erred -in holding this line to be conclusive evidence of boundary, and also in refusing to give the first and second instructions asked for. This disposes of all of the errors assigned except that which refers to the allowance of the amended consent rule. The consent rule as it is technically termed, was adopted in the court-of king’s bench and common pleas in England, for the sake of convenience, . and was a consequence of the fictions allowed in the action of ejectment. By its terms the defendant simply consents to waive proof of the lease, entry and ouster, stated in the declaration, and then the cause is submitted to the.jury upon the question of title. It could never work any prejudice to the party. The practice for some time varied in the two courts, the rule being special in the common pleas, and general in the king’s bench. The court of common picas required the defendant to specify the particular part of the premises to which he claimed title. But at this time, *393the practice on this subject is the same in both courts, and it is now necessary in all cases to give evidence of the defendant’s possession of the disputed premises at the time of the commencement of the suit. Tillinghast’s Adams on Eject. 235; 9 Cowen, 661. The only case now in which a special consent rule can be necessary is, where an actual entry is necessary to be made upon the land previous to suit brought. Thus, where an ejectment is brought by one joint tenant, tenant in common or coparcener against his companion, as' an actual ouster is necessary, the defendant may apply to the court upon affidavit for leave to file a special consent rule, to confess lease and entry, but not ouster. 9 Cowen, 236. And this will always be granted. 2 Cowen, 442. The effect of these rules of practice is to render nugatory the amendment of the consent rule in this cause. Since it can have no influence upon the rights of the parties or the quality and measure of proof required in the cause. It did not prejudice the plaintiff, and he cannot, therefore, assign it for error. But, at any rate the allowance of amendments is not a matter for which error will lie. 3 Comyn’s Digest, 566; 11 Wheaton.

For the reasons, however, which have been stated in relation to the other assignment of errors, the judgment of the court below must be reversed, the. cause remanded, and a venire de novo awarded,