Meantime, the general government had placed in the market the lands so withheld from the first certification and they have been disposed of under the general land laws to settlers and purchasers, or they have been granted, in some part, to the State to aid in the construction of railroads. Thus it happens that, at the present time, the lands claimed by the State, upon this basis, are generally in the hands of private owners by direct conveyance from the government, or through grants to aid in internal improvements; and the intervention of the Department in favor of the State, would, in most cases, operate only as the expression of the opinion of the Department that the State possesses a better title than has been conveyed to the grantees of the government, and attempt to invest the State with a title by patent after other patents have been already issued. The contestants of the right of the State to the particular tracts now under discussion, are persons who have since 1880, made entry of them from the government; although arguments have been heard on behalf of a much greater num ber indirectly interested in the subject.

It may be added that no evidence is furnished to show that the lands claimed now are, or at the date of the grant were, in fact, swamp or overflowed; and the presumption must be allowed, therefore, that they. were not, since such is the testimony of the reliable survey.

The points of law involved may now be considered without further de tail of the facts, except, perhaps, as connected with some special features.

I.

So far as any obligation to follow the original surveys is asserted to spring from the alleged agreement between the Land Office and the State to adopt the field-notes as the basis for designating swamp lands, three reasons appear to be cach sufficient to answer the claim.

In the first place, no power existed in the Commissioner, or in the Secretary himself, to so bind the government. The act of Congress made the grant and defined the subject granted. The Department could not add to its terms or impose an obligation to patent lands not swamp and overflowed and unfit for cultivation. One Secretary might adopt for himself a practice or rule to govern his performance of the duty to designate and certify lands granted, which the act devolved upon him; but he could not impose that rule or practice upon another, nor even oblige himself, if he found it to be incorrect in operation. The supreme court has determined that the ascertainment of the specific tracts granted by the act of Congress, is a question of fact, to be settled by the Secretary upon evidence, or, upon failure of a determination by him, by a jury. Railroad Co. v. Fremont County (9 Wall., 89); Railroad Co. v. Smith (9 Wall., 95); Buena Vista County v. Railroad Co. (112 U. S., 165). What, therefore, will be accepted as sufficient to establish the fact is a question only of evidence, which every Secretary must decide for himself. The act of the Legislature of Michigan amounts

only to the authoritative declaration on the part of the State, of its acceptance of the field-notes as a rule of designation if the Secretary pur sues it. The Department cannot be held bound by that act, or by any understanding to pursue that rule, until, at least, after certification and patent.

Secondly, no agreement of the kind can be asserted upon the basis of the facts narrated. The instructions to the surveyors-general do not amount to a contract with the State. They are properly to be regarded only as the declaration of a rule of evidence which the Department was willing to pursue on condition that the State was willing to accept it; but no more obligatory, nor less open to change, than other instructions of the Department to officers under its jurisdiction. The assent of the State, which was the condition of the instructions being operative, did not give them the character of inflexible law or binding obligation. And, although the Department has usually pursued the course so outlined, it has repeatedly expressed the reservation of the right and purpose to proceed upon more satisfactory evidence when the surveys are shown to be so incorrect as to be unreliable. La Chance v. Minnesota (4 L. D., 479).

Finally, so far as this case is concerned, the obligation of the sup posed understanding turns upon the application to be given to the terms employed, "the notes of the surveys". For, if this phrase means the notes of the corrected and permanent surveys, which was its interpre tation from beginning to end of the certifications made to Michigan as already shown by the facts narrated, the force of the claim is turned against the State. The value of contemporary construction is universally acknowledged; and, in this case, irrespective of any question of estoppel, the action of the Department and its acceptance by the State during so many years, can leave little doubt that the field-notes referred to were intended to be those which furnished the best evidence of the fact. The force of this contemporaneous construction is aug. mented by the fact that at the time when the phrase was first employed in the adjustment of the grant with the State of Michigan, many resurveys had already been made, some of them at the request of the State itself, and that others were in progress under specific appropria tions therefor by Congress; appropriations which were, in terms, "for correction of erroneous and defective surveys," "for re-surveying and correcting erroneous surveys" and the like. It cannot be presumed that when corrected surveys already existed, or were in progress, reference by this phrase was intended to those which were, or should prove to be, erroneous and defective, instead of those which were correct and reliable. If, therefore, the meaning of the phrase, as applied to surveys already made or in progress, attached to the notes of corrected surveys, instead of the original defective ones, it cannot be doubted that it equally applied to the notes of such surveys as should be subsequently, by authority of Congress, likewise corrected. And since this

obvious conclusion was, in fact, recognized and acted on by both parties when the re-surveys were afterwards made, the meaning of the phrase in the supposed agreement must be accepted accordingly; and so far as obligation attends it, the consequence is unfavorable to the present claim of the State.

II.

The effect to be awarded to the certifications first made upon the basis of the original surveys presents the most serious question upon this appeal. The swamp land act was a grant in præ senti, passing title to the lands of the described character from its date. Wright v. Roseberry (121 U. S., 488). But identification of these lands was essential to this title to any specific tract, and the act left it to the judgment of the Secretary of the Interior to conclusively make that identification. It is only where he has failed to act, that resort to another tribunal is admissible. Wright v. Roseberry, supra.

Upon the same authority, it must also be admitted, perhaps, that when the Secretary has made this identification by a certification not open to valid question, the duty to patent upon the request of the Governor so essentially follows, that the title may be deemed perfect to the specific tracts thus determined to be granted even before the act of patenting.

When, however, the certification made by the Secretary is, before patenting, challenged for fraud or mistake, I think the right and duty remain with him to correct the identification according to the facts, so that the patent shall issue only for lands which were, in truth, granted by the act of Congress. To put the case strongly, is it to be presumed that any court would, by mandamus, compel the Secretary to issue the patent, when by false and fraudulent evidence he has been deceived into affixing his signature to a certification which embraced, among the tracts described, a large body of land beyond all dispute high and dry, in no part wet, swamp or overflowed, or unfit for cultiva tion? Or would a writ go to compel the patenting of such lands which, through accident or gross mistake, had been so included in a certified list? To suppose such a judgment possible, would impute to a court the violation not only of the act of Congress, but of plain and acknowledged principles of justice. It necessarily follows that when such a case occurs, the Secretary retains the power, until a patent issues, to correct his error in the attempt to identify the granted lands, by conforming his certification to the truth as discovered after it was made. It equally follows, that when the Secretary has, before patent, made discovery of fraud, accident or mistake, in consequence of which he has included in his certification lands not granted by the act of Congress, and has corrected the certification so as truly and rightly to identify the lands actually granted, the only right to a patent existing in the State, or which any court would enforce, attaches to the land actually

granted, and correctly identified by the amended certification. Other wise, the singular consequence would attach to the performance of this act which is no where else suffered by the law, that fraud, mistake or accident becomes, openly and undeniably, the basis of title, and the government is defrauded of its lands by means which would be a sufficient foundation to relieve another grantor from a conveyance actually made. This reasoning seems, to my mind, indisputably to demonstrate that a right of correction in a proper case, remains in the Secretary, even after a certification has been made, if the issuance of patent has not dispossessed him of jurisdiction.

Applied to the facts of this case, this line of reasoning also demonstrates that the State is not entitled to the lands now claimed. Because it has been shown that these lands were erroneously embraced in the certifications made upon the basis of the original surveys, that the evidence which led to their inclusion in the lists was false, that the lands are, in fact not swamp or overflowed and were not granted by the act of Congress, and that the mistake of the Secretary was corrected by another list, in accordance with the facts. The correction was made in the exercise of a rightful jurisdiction remaining in the Department, as I think I have shown; and, had the State resorted immediately thereafter to a court for the writ of mandamus, it can not be supposed that its judgment would have obliged the government to part by patent with its rightful title to lands which had not been granted, simply be cause of a mistake of its officers induced by false evidence and corrected before judicial aid was invoked, while patent was, at the time, proffered to every tract to which the State was entitled by the grant.

But the case of the State now stands much more unfavorably to its demand. All the lands which were, in fact, granted to it--at least within the township now in question-were long ago identified and patented; the correction of the lists of identification, and all the cir cumstances upon which the correction proceeded, were fully made known to its officers; in the course of the correction, lands not embraced in the first list, were added to those certified, while none were excluded except such as were not subject to the grant; the patents were accepted by the State and no claim asserted for the residue until after the lapse of many years. Meantime, the lands omitted from the original certifications in order to conform them to the lands actually granted, have been sold or otherwise disposed of by the government, and are held by persons entitled to be regarded as innocent purchasers in good faith. Under such circumstances the inequity of the claim to lands which are not in fact such as were described in the grant, the title which did not pass in præsenti by the act itself, which never has passed by patent, and to which no other right exists than can be drawn from a mistake in supposing them within the grant, a mistake arising from false evidence, a mistake corrected without objection by the officers of the State, becomes, to my mind, so clear and unmistakable that the

duty of this Department is to lend no aid to the demand. If, upon such a foundation, there be any principle of law which will award to the State these lands not granted, it must be discovered and enforced by others.

III.

It is further claimed on behalf of the State, that the act of March 3d, 1857, confirmed these certifications, and directed the approval and patenting of the lands therein described to the State, irrespective of the correctness of the identification. That act declared:

That the selection of swamp and overflowed lands granted to the several States .. heretofore made and reported to the Commissioner of the General Land Office so far as the same shall remain vacant and unappropriated, and not interfered with by an actual settlement under any existing law of the United States, be and the same are hereby confirmed, and shall be approved and patented to the said several States, as soon as may be practicable after the passage of this law.

This theory of the act necessarily compels counsel for the State to assert that all the lists reported by the surveyor-general and certified by the Secretary, both the original and the corrected lists, were thereby confirmed. But it must be presumed, I think, that Congress knew the facts, and its language must be applied to the facts as they were understood and interpreted by the government at the time.. All the corrections in the certification had been previously made. Each list of identification in respect to which any claim is now asserted had been amended by striking out the lands which, upon the better evidence of the re-surveys, were shown not to be swamp or overflowed, and by the addition of those which were of that character but not embraced in the original list. In point of fact, the lands erroneously included in the first certifications, had been actually erased from those certifications by lines of ink drawn over them and annotations showing them to be expunged from the lists. In each case of an erroneous certification another amended list had been prepared as a substitute, made in accordance with the facts. In legal contemplation, this amended list stood in the stead of the other, and the case must be regarded to be the same in law as if the correctious had been made actually upon the original list, by erasure of the lands improperly included in it and addition there of such as ought to be included but had not been. In other words, there was but one selection remaining, the amended and corrected list which stood in place of the preceding. I have shown this to have been done in the exercise of a rightful jurisdiction in the Department. It must be presumed that this was the list which Congress had in mind when passing the confirmatory act of 1857.

This view is re-enforced by the contemporaneous construction of the Department and officers of the State, already sufficiently stated; and by the equitable considerations to some extent presented, sufficient to operate an estoppel against a different interpretation now.