Bill in equity by the United States to annul patents issued May 10, 1895, and January 6, 1896, to the Northern Pacific Railroad Company, and March 5, 1901, and January 4, 1904, to its successor, the Northern Pacific Railwary Company, for certain described lands. The foundation of the bill is that the patents were issued by mistake as public lands granted to the railroad company under the act of Congress dated July 2, 1864 (13 Stat. at L. 365, chap. 217), the lands actually being, it is alleged by the government, part of the Yakima Indian Reservation under a treaty with the Yakimas of June 9, 1855 (12 Stat. at L. 951), ratified March 8, 1859, and proclaimed by the President April 18, 1859.

There is no question made of the title of the railroad and railway companies, or of their respective vendees, other than as the lands fall within or without the reservation. If they were within the boundaries of the reservation, they were lands of the Indians; otherwise, public lands of the United States, and passed to the companies, respectively, under the act of Congress and the patents issued in pursuance thereof.

The question then is, What were the boundaries of the reservation? or—to use the present tense as the more convenient what what are the boundaries of the reservation?

By article 1 of the treaty the Indians ceded, relinquished, and conveyed to the United States a tract of land which was explicitly described, reserving by article 2, from the tract, the land included within the following boundaries: Commencing on the Yakima river, at the mouth of the Attahnam river; thence westerly along said Attahnam river to the forks; thence along the southern tributary to the Cascade mountains; thence southerly along the main ridge of said mountains, passing south and east of Mount Adams, to the spur whence flows the waters of the Klickitat and Pisco rivers; thence down said spur to the divide between the waters of said rivers; thence along said divide to the divide separating the waters of the Satass river from those flowing into the Columbia river; thence along said divide to the main Yakima, 8 miles below the mouth of the Satass river; and thence up the Yakima river to the place of beginning.

All of this tract, it is provided, 'shall be set apart, and, so far as necessary, surveyed and marked out, for the exclusive use and benefit' of the Indians, as an Indian reservation.

It will be observed that the calls in the description of the tract reserved are very confident, and seem to assure certainty by prominent and unmistakable natural monuments. Controversies, however, almost immediately arose, the Indians contending for one location of the calls and enterprising settlers contending for another. The Interior Department ordered a survey, which was made, and which is known in this record as the Schwartz survey. Upon this the title of appellants depends. The discontent of the Indians continued and another survey was ordered by the Interior Department to be made by E. C. Barnard. This survey is the foundation of the bill and of the contention of the government. It was made and reported to the Interior Department with a map delineating the exterior boundaries of the reservation. This report was transmitted to the Speaker of the House of Representatives with a draft of a bill granting authority for the detail by the Secretary of the Interior of an Indian inspector to negotiate an agreement with the Indians for the adjustment of their claim for the lands embraced in the tract claimed by them, containing 293,837 acres, as shown by the Barnard report; that is, for lands without the Schwartz, but within the Barnard, survey.

In pursuance of the recommendation of the Secretary of the Interior, Congress, on December 21, 1904, enacted the statute quoted in the margin.

After the passage of the act the government demanded a reconveyance of the lands, which was refused. This suit was then brought.

The controversy in the case, therefore, turns upon which of the surveys, Schwartz's or Barnard's, correctly marks the boundaries of the reservation. The difference in the surveys amounts to 293,837 acres. The and entered an decree canceling the patents. circuit court accepted the Barnard survey The decree was affirmed by the circuit court of appeals. 112 C. C. A. 359, 191 Fed. 947.

Sec. 1. That the Secretary of the Interior be, and he is hereby, authorized and directed, as hereinafter provided, to sell or dispose of unallotted lands embraced in the Yakima Indian Reservation proper, in the state of Washington, set aside and established by treaty with the Yakima Nation of Indians, dated June nine, eighteen hundred and fifty-five: Provided, That the claim of said Indians to the tract of land adjoining their present reservation on the west, excluded by erroneous boundary survey, and containing approximately two hundred and ninety-three thousand, eight hundred and thirty-seven acres, according to the findings, after examination of Mr. E. C. Barnard, topographer of the Geological Survey, approved by the Secretary of the Interior April seventh, nineteen hundred, is hereby recognized, and the said tract shall be regarded as a part of the Yakima Indian Reservation for the purposes of this act: Provided further, That where valid rights have been acquired prior to March fifth, nineteen hundred and four, to lands within said tract by bona fide settlers or purchasers under the public land laws, such rights shall not be abridged, and any claim of said Indians to these lands is hereby declared to be fully compensated for by the expenditure of money heretofore made for their benefit, and in the construction of irrigation works on the Yakima Indian Reservation. [33 Stat. at L. 595, chap. 22.]

The special controversy in the case is the location of the western boundary of the reservation. But, as partly determinative of that, the western point of the northern boundary must be considered. The northern boundary of the reservation commences at the junction of the Yakima and Attahnam rivers and proceeds to the forks of the latter, and along its southern tributary to the 'Cascade mountains.' What constitutes the Cascade mountains is the first serious dispute in the case. The appellants contend that the mountains are given location by the termination of the southern tributary of the Attahnam river. In other words, the headwaters of that tributary mark the Cascade mountains. But the next call is to be considered. By that call the line is to run 'southerly along the main ridge of said mountains;' and, as said by the circuit court, the line must reach the main ridge to run southerly along it. The court erred, appellants contend, by assuming that the treaty makers meant to designate the main ridge of the mountains instead of a ridge of the mountains. We cannot, of course, reproduce all of the argument of counsel. It is, in effect, that the treaty makers meant what they said, that their knowledge was not imperfect, that they knew where the waters of the Attahnam river terminated, and they turned south from there along 'that ridge of those mountains' in which they found themselves. Assuming this, it is said, 'every difficulty in following the calls of the treaty at once disppears.' But the difficulties do not disappear; they multiply, and mountains and rivers appear to conflict in their testimony. The next call must be changed to be accommodated to counsels' view. That call, in full, is this: 'Thence southerly along the main ridge of said mountains [Cascade mountains], passing south and east of Mount Adams, to the spur whence flows the waters of the Klickitat and Pisco rivers. Counsel would strike out the comma after the word 'mountains' and the comma after the word 'Adams,' asserting then the main ridge to be that which passes (passing) south and east of Mount Adams to the spurwhence whence flows the waters of the Klickitat and Pisco rivers. In other words, the call primarily locates and defines the ridge, and not the boundary line. And so change the call, it is further said, and there is intelligible continuity between it and the next call, which reads, 'thence down said spur whence flows the waters of the Klickitat and Pisco rivers) to the divide between the waters of said rivers.' Punctuation, it may be admitted, is a fallible standard of the meaning of a statute (Ewing v. Burnet, 11 Pet. 41, 54, 9 L. ed. 624, 630; Hammock v. Farmers' Loan & T. Co. 105 U. S. 77, 84, 85, 26 L. ed. 1111, 1113, 1114). It is, however, not without force, and in the present case the location of the commas is consistent with the purpose of simply marking the course of the boundary line. But even without changing the punctuation, counsel contend 'that the words 'passing south and east of Mount Adams' qualify the word 'mountains,' and indicate which ridge was intended; namely, a main ridge (as distinguished from spurs or 'subdivides') which should pass south and east of Mount Adams.' We cannot assume a plurality of main ridges, and that the treaty meant to distinguish one from the others. The main ridge necessarily had a definite and conspicuous individuality and needed no identification. It is used in article 1 of the treaty to mark the course of the boundary his line to it because he considered other the United States. The Indians always claimed it as the western boundary of the reservation, and the earliest maps confirmed the claim. Schwartz had no difficulty in determining it. He did not run his lime to it because he considered other calls were more controlling. He was in no uncertainty as to its location. It was and is a natural and conspicuous landmark, and was selected to define the immense area of land ceded by the Indians to the United States, and the lesser though extensive tract reserved by them for their own use. We must keep in mind their situation,—what they gave and what they reserved. They were not deeding, as the government forcibly says, acres or even townships. They gave up a principality. They reserved, it is true, a much lesser tract, but it was natural and inevitable that 'the greater boundaries of nature' should be selected to define both. These the Indians could understand and estimate. 'The inferior ridges or spurs, connected with, but leading away from, the main ridge,' could not be so definitely intelligible. The Indians had to be satisfied. They entered into negotiations with the representative of the government reluctantly, their chief testified. They feared the encroachments of the white man. Their fears were allayed by adapting the treaty to their understanding, by delineating the land they conveyed and the land they reserved by great and commanding objects. They have never indicated by word or act that the main ridge was not single and distinct in their minds, or that it was at any time confounded by them with lesser ridges. They never have wavered in the expression of their understanding and their insistence that it constituted the western boundary of the reservation, and that it extended to the base of Mount Adams on the south. They always had, as we shall see, and intelligible conception of the western boundary and its definition by natural objects. It is only by regarding this understanding and the more prominent natural objects that the more prominent natural accommodated to the topography of the country.

Some of the natural objects, considered by themselves, it may be admitted support the contention of appellants. The most important of these is that mentioned in the 5th call of the treaty. According to the 4th call the line runs southerly along the main ridge to the spur whence flow the waters of the Klickitat and Pisco rivers, and (5th) 'thence down said spur to the divide between the waters of said rivers Schwartz's survey. He knew that the main ridge of the Cascades is west of the tributary of the Attahnam river, but he put it out of consideration or effect. He regarded what he conceived to be the divide between the waters of the Klickitat and Pisco rivers as dominating all other calls, although he was directed to confer with the agent at the Yakima Agency, with other white persons, and with Indians familiar with the country, and obtain all the information possible and that would tend to a proper location and establishment, according to the provisions of the treaty, of the section of the boundary line he was directed to survey. He did not run his line to the main ridge of the mountains, because, as he said, he 'could not do it without crossing the Klickitat river, and the treaty did not call for that.' This was his error. He gave too much strength to some of the calls of the treaty and against other calls, without attempting to give them all effect from a consideration of the topography of the country and the testimony he was directed to take. In this attitude of mind he made his survey and seems to have rejected everything which would disturb it.

We realize that there is confusion in the calls, irreconcilability, it may be, from some points of view; but our effort must be to ascertain and execute the intention of the treaty makers; and as an element in the effort we have declared that concession must be made to the understanding of the Indians in redress of the differences in the power and intelligence of the contracting parties. United States v. Winans, 198 U. S. 371, 49 L. ed. 1089, 25 Sup. Ct. Rep. 662. The present case invokes in special degree the principle.

As we have seen, there were certain conspicuous landmarks which would attract the attention and be intelligible to the understanding of the Indians. Lesser marks would be given no significance. We have already observed the importance in this regard of the main ridge of the mountains, and it was given emphasis besides by such a conspicuous object as Mount Adams. Mr. Barnard testified that Goat rocks are prominent points on the main ridge, and that Indian Chief Spencer told him that the northern line extended westward from the head of the Attahnam river to a sharp point east of Goat rocks, which point was plainly visible and a well-marked feature in the landscape, and that the boundary line extended to a conical hump on the southeast slope of Mount Adams, which is well defined and plainly visible. The map made by the direction of Governor Stevens in 1857, to show the Indian reservations in Washington territory at that time, and also the White Swan map, show that the northern boundary runs to the main ridge of the mountains.

The Stevens map, though vouched for by him to be accurate, has many inaccuracies, as now demonstrated by a better knowledge of the country, and adds to the confusion if we seek to extend its testimony beyond a confirmation of the Indians' claim that the main ridge of the mountains is the western line of the reservation. By it the south fork of the Attahnam river is made to reach the summit of the Cascade far west of Mount Adams, and the line is run thence for some distance south on the ridge; thence southeasterly to the divide between the Satass and Columbia rivers. The tract delineated is relatively narrow from north to south, due probably, as the government says, to a misunderstanding of the true situation of the Satass-Columbia divide, and a failure to bring the west line down the main ridge to the southeasterly slope of Mount Adams, as required by the treaty. There is another inaccuracy. The map shows the Klickitat river as heading west of the spur upon which Mount Adams is represented as rising. The mistake, now known to be such, shows how imperfect knowledge of the country was, and the importance of giving effect to the more commanding features of the landscape.

Schwartz turned from the 51st mile post sharply north, deeming, as we have seen, the divide between the waters of the Klickitat and Pisco rivers as controlling. But to the west of the 51st mile post there is a mountain called Grayback, which the Indians claim was on the boundary line of the reservation. Schwartz disregarded it, although he testified that there was a ridge running westerly from a point a little south of the 51st mile post, terminating in the Grayback mountain. He did not follow that ridge, he says, because it formed the divide between the waters of the Klickitat and Columbia rivers and did not form the watershed of the waters flowing into the Satass river. And yet Barnard, considering the calls of the treaty, and in adaptation of them to the topography of the country, followed that ridge as part of the southern boundary, and in 1861 it was surveyed as part of the southern boundary. The survey is called the Berry & Lodge survey, and was made by the direction of the superintendent of Indian affairs for the territory of Washington. He directed them to proceed from the Yakima river westerly along the divide between the Satass and Columbia rivers and along the divide between the Klickitat and Pisco rivers until they arrived at the source of either the latter or the former, where they should terminate the survey. He added: 'Should you find before arriving at the source of either of these rivers that the 'divide' has assumed the character of a perfect natural boundary, you will terminate your survey at the point where this description of boundary is attained.' The plat of the survey indicates that the south boundary was run to a point on or near the Klickitat river, and marks that stream as originating on the south slope of Mount Adams, and flowing thence southwesterly. It also shows a tributary of the Pisco river as headed near the east side of the mountain, and a spur of hills projecting between them southeasterly to meet the ridge constituting the Satass-Columbia divide. The field notes of the survey are attached to the government's brief and have this note: 'South boundary only was surveyed, in accordance with the instructions of the superintendent. The other boundaries are defined naturally.' Some of the marks and posts of this survey were found by Barnard.

One other piece of evidence needs only to be adduced. Two Indians, one of them Chief Spencer, told him that in 1860 they accompanied certain government agents of Governor Stevens along the southern boundary of the reservation, proceeding along a well-defined ridge to Grayback peak, upon the summit of which a marked wooden post was found set in the ground. From there, the agents told them, after sighting through an instrument pointed at a conical hump on the southeast slope of Mount Adams, that the line went straight to that point. This account was subsequently repeated. Chief Spencer (it was to this chief that Governor Stevens addressed himself in regard to the Indians removing to the reservation) testified that Governor Stevens promised to stake out the reservation, and that some government men, while standing with him at the junction of an Indian trail on a road called the Goldendale road, and which is marked on the Barnard map as being between Mount Adams and Grayback, told him that the line ran from one to the other, and that Goat rocks would be the northwest corner. He further testified that at the forks of the road and the trail there was a blazed tree on one side and a pile of rocks on the other. The statement received corroboration from Barnard, who testified that he discovered a blaze forty years old upon one of two large pine trees at the place indicated, both of which had been anciently blazed.

There is evidence which may be adduced in corroboration of the testimony of the respective witnesses, but we have referred to enough to indicate the character and relative strength of that which makes for or against the contentions of the parties, and, considerably weighing it, we think it establishes the correctness of the Barnard survey. And we have arrived at and announce this conclusion with full sense of the weight which should be given to the action of the Land Department in approving the Schwartz survey and the issue of the patents. The action of the Land Department is necessarily a strong consideration. But it is opposed by later action and also by congressional action. At any rate, the action of the Department has been brought in controversy, and because it may be supported by plausible or even strong arguments, it does not follow that the opposing claim becomes immediately so doubtful as to determine judgment against it. On the contrary, the questiom must be examined and decided with due regard to the entire situation, keeping in mind the action of the Department as an element to be considered, and applying the rule of the cases that it should not be disturbed except for reasons that are clear and convincing; assuming, without deciding, that the rule applies to a case in which the government is proceeding in the right of the Indians.

The court of appeals expressed the view that the rule that resolves doubts in favor of the patent issued by the United States does not apply in such case, citing Leavenworth, L. & G. R. Co. v. United States, 92 U. S. 733, 23 L. ed. 634; Stewart v. United States, 206 U. S. 185, 51 L. ed. 1017, 27 Sup. Ct. Rep. 631; Minnesota v. Hitchcock, 185 U. S. 373, 46 L. ed. 954, 22 Sup. Ct. Rep. 650. Much can be said in support of that view. It must be borne in mind that the Indians had the primary right. The rights the government has are derived through the cession from the Indians. If the government may control the cession and control the survey, and by the action of its agents foreclose inquiry or determine it, an easy means of rapacity is afforded, much quieter, but as effectual, as fraud. We should hesitate to put the government in that attitude. It rejects that attitude and accepts a greater responsibility. It yields to the rule which this court has declared—that it 'will construe a treaty with the Indians as 'that unlettered people' understood it, and 'as justice and reason demand in all cases where power is exerted by the strong over those to whom they owe care and protection,' and counterpoise the inequality 'by the superior justice which looks only to the substance of the right, without regard to technical rules,' Choctaw Nation v. United States, 119 U. S. 1, 30 L. ed. 306, 7 Sup. Ct. Rep. 75; Jones v. Meehan, 175 U. S. 1, 44 L. ed. 49, 20 Sup. Ct. Rep. 1.' United States v. Winans, supra.

It is contended that the Northern Pacific Railway Company and the individual appellants are bona fide purchasers, and, as such, entitled to protection under the act of March 2, 1896 (29 Stat. at L. 42, chap. 39, U. S. Comp. Stat. 1901, p. 1603). Section 1 of that act provides that suits brought by the United States to vacate and annul any patent to lands theretofore erroneously issued under a railroad or wagon road grant should only be brought within five years from the passage of the act, and suits brought to annul patents issued after the passage of the act should be brought within six years. And it is provided 'that no patent to any lands held by a bona fide purchaser shall be vacated or annulled, but the right of such purchaser is hereby confirmend.' The act was one of a series of acts, and manifestly applies only to the public lands of the United States subject to acquisition under the laws enacted for the disposition of the public domain.

We have seen that the act of December 21, 1904, protects rights acquired prior to March 5, 1904, to lands within the Barnard survey 'by bona fide settlers or purchasers under the public-land laws.'

The appellants are not within that class, nor for the reasons we have stated can they avail themselves of the defense of the statute of limitations under § 8 of the act of March, 3, 1891, 26 Stat. at L. 1093, chap. 559, as amended, 26 Stat. at L. 1099, chap. 561, U. S. Comp. Stat. 1901, p. 1531.

Decree affirmed.