Kentucky Coal Lands Co. v. Mineral Development Co.

DENISON, Circuit Judge.

The plaintiff in error brought what was called an action of ejectment in the state court against the defendant in error. The párties will be named herein as they were called below. The defendant removed the case into the court below upon the ground of diverse citizenship; its petition for removal showing that plaintiff was a citizen of New York and defendant was a citizen of Virginia. The plaintiff moved to remand; its motion being based on the ground that neither party was a resident of the Eastern District of Kentucky. The motion to remand was denied, and plaintiff assigns error thereon.

When the case came on for trial on the merits before a jury, it developed that the Kentucky patent on which plaintiff’s title depended was the same one which had been involved and had been considered by this court in Mineral Co. v. Tuggle Co., 151 Fed. 450, 81 C. C. A. 34; and the District Judge, thinking that the question here was only the construction of the patent and so was a question of law, and that it had been decided by this court in the former case, directed a verdict for defendant. Error is also assigned upon this ruling.

Having in mind that a suit cannot be removed from a' state court to the United States District Court of a district in which it could not have been brought (Ex parte Wisner, 203 U. S. 449, 27 Sup. Ct. 150, 51 L. Ed. 264; In re Moore, 209 U. S. 490, 28 Sup. Ct. 585, 706, 52 L. Ed. 904, 14 Ann. Cas. 1164), it is apparent that the right of removal in this case depended on the answer to this question:

“Was the general grant of jurisdiction found in the earlier part of section 1 of the act of March 3, 1887, covering cases where the matter in controversy exceeds the jurisdictional amount, and is between citizens of different states, so limited by the later part of the same section that such an action as this could not be brought at all in the federal courts, unless the land involved lay in the district of the residence of the plaintiff or the defendant?”

[1] The common-law action of ejectment was classified as mixed rather than as either real or personal; but the actions which are commonly called by that name, when brought under Codes like that of Kentucky, are purely actions to recover the property. As is said of the Code action in Pomeroy on Code Remedies (3d Ed.) 294:

“It far more nearly resembles, in all of its essential features, the ancient real actions which were displaced in use by ejectment. * * * It does not bear the slightest resemblance to the action of ejectment as that was contrived by the old judges and lawyers, and only confusion and misconception result from applying to it that name.”

While the petition in this case prays both that plaintiff be adjudged the owner of the described tracts and that it recover damages for the *47unlawful detention, yet the action to recover the land itself is none the less a real action because there is united with it, under the permission of the Code, another action which is personal; and it is enough for the present inquiry if there is disclosed on the record one separable controversy properly removable to the federal courts, since in that case it is immaterial that another nonremovable controversy is joined. Barney v. Latham, 103 U. S. 205, 210, 26 L. Ed. 514. The distinctive character of the action makes it clear that it is a local and not a transitory action, because that is true even of a common-law ejectment. Northern Indiana Co. v. Michigan Central Co., 56 U. S. (15 How.) 233, 242, 14 L. Ed. 674.

[2] Assuming then, as we must, that the present action is local, not transitory, we meet the question whether it is affected by the limitation as to residence found in the latter part of the section, or whether that limitation pertains only to transitory actions. The Supreme Court has several times said that this clause or its predecessor in the statutes is not one pertaining to jurisdiction, but rather to venue or to territorial distribution (In re Hohorst, 150 U. S. 653, 660, 14 Sup. Ct. 221, 37 L. Ed. 1211; Sweeney v. Carter Co., 199 U. S. 252, 256, 26 Sup. Ct. 55, 50 L. Ed. 178); and yet, as to those classes of actions to which the restriction does pertain, it is hard to see why the failure to distribute the action anywhere is not as fatal as would be the failure to grant jurisdiction at all; so the real inquiry must be whether the limitation does pertain to local actions.

The decision of the Supreme Court in Casey v. Adams, 102 U. S. 66, 26 L. Ed. 52, goes far towards answering the inquiry; perhaps it is a complete answer. One section of the National Banking Law provided that suits against a bank might be brought in any court of the United States within the district in which the bank was established, or in any state court held in the county or city where the bank was located. This provision was broad and general, and upon its face applied to any and every action, personal or real, local or transitory. It differed from the residence limitation clause of section 1 of the act of 1887 only in that the prohibition against suits in other territorial divisions was implied instead of expressed; yet, under the rule that whatever is not named in such enumeration is excluded, it would seem clear enough that a transitory action against such a bank could not have been brought in any place not specified in the permission; and the court seems to have assumed that, if the statute applied at all to the action there involved, the suit must be dismissed. However, the court had no difficulty in finding, from the necessity of the case, an exception not hinted at in the language of the statute. Chief Justice Waite said (at page 67 of 102 U. S. [26 L. Ed. 52]):

“This, we think, relates to transitory actions only, and not to such actions as are by law local in their character. Section 5136 [Comp. St. 1913, § 9661} subjects the banks to suits at law or in equity as fully as natural persons, and we see nowhere in the Banking Act any evidence of an intention on the part of Congress to exempt banks from the ordinary rules of law affecting the locality of actions founded on local things. The distinction between local and *48transitory actions is as old as actions themselves, and no one has ever supposed that laws, which prescribed generally where one should be sued, included such suits as were local in' their character, either by statute or the common law, unless it was expressly so declared. Local actions are in the nature of suits in rem, and are to be prosecuted where the thing on which they are founded is situated. To give the act of Congress the construction now contended for would be in effect to declare that a national bank could not be sued at all in a local action where the thing about which the suit was brought was nothin the judicial district of the United States, within which the bank was located. Such a result could never have been contemplated by Congress.”

If Casey v. Adams may be distinguished because of the less positive character of the limitation in the Banking Act as compared with that in the act of 1887, it becomes necessary to make further study of the latter.

As matters stood under the Revised Statutes, jurisdiction of cases of diverse citizenship was conferred by section 629, and there was no reference to the residence of either party. The distribution of the jurisdiction (that is, the venue of the action) was taken up generally by section 739, which said that, except in the cases provided in the preceding section and in the following three sections, suits must be brought in the district whereof the defendant was an inhabitant or was found. The preceding section (738) refers to suits “to enforce any legal or equitable * _,* * ..claim” against property in the district, and provides for service by publication. Section 740 refers to suits not of a local nature, where different defendants live in different districts of the same state; section 741, to suits of a local nature, where the defendant resides in another district in the same state; and section 742 to suits of a local nature, where the “land or other subject-matter of a fixed character lies partly in one district and partly in another” — providing that in such catee suit may be brought in either district, and process be executed “as fully as if the said subject-matter were wholly within the district” in which the suit is brought.

Considering together these sections 629 and 738-742, the arrangement is clear and logical. Section 629 conferred jurisdiction. Section 739 (by itself and by virtue of the. reference therein contained to sections 741 and 742) distributed the cases by one method as to local actions and by another method as to other actions; and sections 738, 740, and 741 provided methods of service, where the defendant was not found within the district. ;

The act of 1875, by section 1, revised section 629, and in the midst of it inserted the substance of section 739. By section 8 of this act there was a-revision and re-enactment of. section 738. Sections 740, 741, and 742 were neither re-enacted nor expressly repealed. Whether they, and especially section 742, have been repealed by implication, was a question which the Supreme Court suggested but left undecided. Greely v. Lowe, 155 U. S. 58, 15 Sup. Ct. 24, 39 L. Ed. 69; Petri v. Creelman, 199 U. S. 487, 493, 26 Sup. Ct. 133, 50 L. Ed. 281; and Galveston Co. v. Gonzalez, 151 U. S. 496, 499, 14 Sup. Ct. 401, 38 L. Ed. 248. That there was no such repeal, but that the sections, or some of them, continued in force, was held by Judge Speer, at the Circuit *49(East Tennessee R. R. Co. v. Atlanta Ry. Co. [C. C.] 49 Fed. 608), by Judge Coxe in Goddard v. Mailler (C. C.) 80 Fed. 422, and by Judge-later Justice — burton in Horn v. Pere Marquette Ry. Co. (C. C.) 151 Fed. 626. The reasoning of these cases is forceful, and, if they are rightly decided, they are persuasive that the jurisdiction over local actions continued after 1875 as before. Section 738, as above stated, was substantially re-enacted by section 8 of the act of 1875, and by section 5 of the act of 1887, and has continuously been the law. It has been spoken of as if it directly conferred jurisdiction of the cases to which it refers. Mellen v. Moline, 131 U. S. 352, 365, 9 Sup. Ct. 781, 33 L. Ed. 178; Jellenik v. Huron Co., 177 U. S. 1, 11, 20 Sup. Ct. 559, 44 L. Ed. 647; Citizens’ Co. v. Illinois Central R. Co., 205 U. S. 46, 59, 27 Sup. Ct. 425, 51 L. Ed. 703.

[5 ] Such references are appropriate to the cases where made, because they involved the getting of jurisdiction (i. e., power to proceed) by authorized service; but in the broader sense of the term “jurisdiction,” it is quite plain that this section 8 only provides a method of obtaining service in those cases of which it assumes the court already has general jurisdiction. Greely v. Lowe, 155 U. S. 58, 15 Sup. Ct. 24, 39 L. Ed. 69; Tug River Coal Co. v. Brigel (C. C. A. 6) 67 Fed. 625, 629, 14 C. C. A. 577. Its true office in the present inquiry is not to create jurisdiction, but to raise the necessary implication that jurisdiction exists by other statutes, and so to determine the interpetation of those other statutes. Taking for illustration a bill to remove a cloud from real estate — a case clearly within this section — it would be useless to make and carefully preserve this section for serving process in such a case on a defendant who did not live in the district, if no such action could be brought in that district; and the existence of this provision compels us to conclude that when section 1 required suit to be brought only in the district where the defendant lived, it did not refer to a bill to remove cloud. While this reasoning, as applied since 1887, stops short of being conclusive, because section 8 might still have occasional application in those instances where the land was situated in the district of plaintiff’s residence, and so the suit could be there brought for that reason, yet considering the history of the various sections, and the exceptional character of this supposed instance, the argument is strongly convincing. It seems hardly necessary to say that section 8 cannot create or imply jurisdiction in a particular district court in cases where resort to substituted service is necessary, and fail to have the same effect in the same case, if by chance the defendant is found in the district. For these reasons section 8 alone strongly implies that the general plan of the Revised Statutes is not intended to be disturbed, and that after the acts of 1875 and 1887, as before, the federal courts had full jurisdiction of local actions where the necessary diverse citizenship existed, and without regard to the residence of the parties. If section 742 continued unrepealed, this inference would seemingly be inevitable.

It must be conceded that some doubt is cast upon this broad implication by Ladew v. Tennessee Co., 218 U. S. 357, 31 Sup. Ct. 81, 54 L. *50Ed. 1069, and Wetmore v. Tennessee Co., 218 U. S. 369, 31 Sup. Ct. 84, 54 L. Ed. 1073. In the latter case, the plaintiff’s land, continuing injury to which was alleged, was. in the Eastern District of Tennessee —the district where the federal court was sitting — and the action was of a character often called local. If it was, for the purpose of that case, rightfully to be so considered, then we must infer from the decision that the District Court does not take jurisdiction of a local action, qua local, and probably that section 742 was repealed, and that Horn v. Pere Marquette, supra, is, to that extent, overruled. On the other hand, such an action was surely transitory under the Tennessee rule (Ducktown Co. v. Barnes [Tenn.] 60 S. W. 593), and the federal courts follow the state rule on this subject (Huntington v. Attrill, 146 U. S. 657, 669, 13 Sup. Ct. 224, 36 L. Ed. 1123; Peyton v. Desmond [C. C. A. 8] 129 Fed. 1, 4, 63 C. C. A. 651). However this may be, we need not go to this extent and say that jurisdiction over all local actions is preserved to the local district. It certainly is so continued as to the actions described in section 8; and the present action — at least if we omit the separable demand for damages — we deem clearly within the description of section 8. It has so been held even with reference to the common-law ejectment action in two well-reasoned cases. Spencer v. Kansas City Co. (C. C.) 56 Fed. 741; Elk Garden Co. v. Thayer Co. (C. C.) 179 Fed. 556. The opinions of the Supreme Court in the Wet-more and Eadew Cases, supra, are not to the contrary. It is as hard to see how Wetmore’s claim for damages to his land in the district, by the distant operation of defendant’s smelter, could be thought within the description of section 8, as it is difficult to understand wherein the present suit to have title adjudicated and to get possession falls short of being a suit- “to enforce any legal * * * claim to * * * real or personal property within the district.” We, therefore, must conclude that section 8 provides a method of getting service upon the defendant in such an action as this, though the defendant does not live in the district, and whether or not the plaintiff does, and so necessarily implies that the jurisdiction apparently existing because of diverse citizenship is not cut off because neither party lives in the district.1

This action was brought before January 1, 1912, when the present Judicial Code went into effect. It is significant that this Code, while retaining the language of the latter part of section 1 of the act of 1887. to the effect that these suits can be brought only in the residence district of one party or the other, goes back to the arrangement of the Revised Statutes. This limiting or distributing clause is no longer found in subdivision 1 of section 24, but takes its place in section 51, and is a complete restoration of section 739 of the Revised Statutes. It is expressly said that this residence limitation shall apply, “except as provided in the six succeeding sections,” and one of these succeeding sections (57) is the same as R. S. 738, and section 8 of the act of 1875, *51and another (55) is the same as R. S. 742. The very enactment of this Code, so clearing up all the confusion that had arisen since 1875 on this point, would be of some force as a legislative declaration of the unbroken meaning of these provisions; but this declaration seems very clear when we consider also sections 294 and 295 of the Code, and the *52revisers’ notes stating that section 51 “re-enacts existing law.” If there was any doubt remaining about the question which has been discussed, it would be removed by the implications which must be drawn from these sections of the Code. United States v. Freeman, 3 How. 556, 564, 11 L. Ed. 724.

*51

*52[3] Having concluded that the court acquired jurisdiction, it becomes necessary to determine whether it properly directed a verdict for defendant. This depends upbn the construction and location upon the ground of the 12,000-acre grant to Isom Stamper, made by the state of Kentucky in 1848, pursuant to the survey in 1846. According as we adopt the contentions of the plaintiff or the defendant, now presented, we will include or exclude from the grant the upper half of the course of Turkey creek and its watershed on both sides. This will be a tract of approximately 2,500 acres. The present suit involves the title to two subtracts out' of this 2,500 acres. The trees upon these two subtracts were, in 1903, sold by the plaintiff, then holding and claiming the same title as now, to the Tuggle Land Company; and to prevent interference with such timber, the Tuggle Land Company brought, in the court below, an action in equity against the defendant, then holding and claiming the same title as now. That action reached this court, and its opinion is reported in 151 Fed. 450. It was decided that the Isom Stamper patent in question did not cover these tracts, and the defendant’s title thereto was approved. Obviously, if the parties in the present case were the same as in the Tuggle Case, the question now presented would be res judicata; but since there is neither identity nor privity between the plaintiffs there and here, the court below rightly considered that the controlling principle was one of stare decisis — in other words, the court below thought that the former decision of this court amounted to a construction of the contract or a declaration of the legal effect of the terms of the grant and their undisputed environment — and so it held that, as matter of law, the grant was to be located on the ground as the defendant claimed then and now and so as to exclude the land in dispute. This brings us necessarily to the terms of the grant; and since the previous case was in equity, in which this court heard and decided facts as well as law, it also is to be considered how much of the decision was matter of law, which must (at least presumptively) now be followed, and how much was matter of fact, which cannot control the action of a jury in a subsequent case between other parties.

The controversy cannot be well comprehended, except by reference to the map reproduced on the preceding page. This discloses three theories, or contentions, as to how the terms of the survey should be applied upon the ground. As shown by the legend, one line indicates the proper location according to the theory of defendant, another according to the theory of plaintiff, and another according to the theory of the Tuggle Land Company, plaintiff in the former suit. It should be said that all the controversy in the former suit and in this has been with reference to the southern line, and that the true locations of the northern and western lines have not been and are not now involved. The language of the grant is as -follows:

“Beginning on three chestnuts and chestnut oak near the head of the Pigeon fork thence running the dividing ridge between Turkey creek and the *53Line fork to the Defeated branch thence the dividing ridge between the Defeated branch and Turkey creek S. 10 W. 68 poles to a chestnut, S. 10 E, 98 poles to a chestnut oak and sarvis S. 68 W. 125 poles to a stake, thence S. 5 E. 100 poles to a stake, thence S. 68 W. 2,000 poles to a stake, thence S. 40 W. 320 poles to a stake, thence N. 1,328 poles to a stake, thence N. 77 E. 1,900 poles to a stake, thence S. 33 E. 600 poles to the beginning.”

The beginning corner is not in dispute. Aside from such complications as might come from the uncertain location of the stake calls, difficulties arise from the presence of the language “thence running the dividing ridge between Turkey creek and the Line fork to the Defeated branch, thence the dividing ridge between the Defeated branch and Turkey creek.” This may refer to two distinct courses, first along one ridge and then along another, or it may refer to one course along a continuous ridge described in terms of its two parts. Under either of these views, doubt arises whether this language is intended of itself to describe and constitute a boundary or boundaries, or whether it is only descriptive of the locality upon which the other boundaries described by course and distance should be located. If the latter, then we see at once that there is no certainty as to when the boundary of stated courses and distances leaves the ridge or ridges, and is to be located independently of that aid. Manifestly, it must leave at some time, but, on the face of the description, there is no indication which stake call is the last one located on the ridge.

If these described ridges constitute of themselves two independent boundaries, then the description has 11 distinct sides, and is an 11-call survey; but if neither of these ridges is itself a boundary, and they constitute only monuments by the aid of which the stake calls are to be located, it is a 9-call survey.

It was the contention of plaintiff Tuggle, in the former case, that the survey had 11 sides; the defendant in that case contended for a 9-sided figure, as shown on the map. This court decided, under the case there presented (and from which the testimony in the present record does not differ, so far as it affects the point now mentioned), that the theory of 11 sides was wholly untenable. The court thereupon adopted the only alternative theory then suggested, viz., the defendant’s theory that there were nine boundaries, and that they were located exactly according to the stated courses and distances. ,By this theory the first two boundaries substantially followed the ridge; the third and fourth followed it approximately; but the fifth, which was a line about six, miles long, departed entirely from the ridge and, crossing Turkey creek and leaving out a large part of its valley, led ■off into the higher mountains to the west.

In the; present case, plaintiff insists upon a third theory of construction and locatio, which adopts a figure of nine boundaries, but makes a different location. The first and second boundaries are located along the top of the ridge, just as done by defendant; the third and fourth courses, instead of being run “S. 10 W. 125 poles to a stake,” and “thence S. 10 E. 100 poles to a stake,” thereby carrying these two stakes partly down the slope of the ridge, are proposed to be run south about 45° west about 125 poles to a distinct bend in the ridge, and thence south about 45° east about 160 poles, thereby *54keeping the line approximately upon the summit of the ridge, and bringing the last-named stake to a point where the ridge makes a very sharp, almost right-angled, turn. Instead of then running the fifth boundary, as defendant does, wholly away from the ridges, across the valley and up the opposite mountains on its stated course, plaintiff proposes to run this course winding along the summit of the ridges until it reaches the end of the second ridge, at the point marked on the map “Eagle Gap.” The first mile of this course, and after a southerly bend, then again about three miles, will average not far from 68° west, the stated course. The intermediate and the final half miles (approximately) swing sharply south, so that the supposed terminus, Eagle Gap, would be south about 45° west from its supposed starting point, at a distance of about 1,200 poles in a straight line, and about 1,800 poles along the course. The remaining four boundaries, plaintiff proposes to run according to the stated courses and distances, with only such variations as may be thought proper to compensate for the changes made in the first five lines and appropriate to close the figure. Under any method, starting from Eagle Gap and getting to the beginning point, they will inclose all the land in controversy.2

It is entirely clear to us, even if it had not been formerly decided, that these ridges were not intended to .be of themselves, boundaries, but that they were intended, as said in our former opinion, to serve as aids in locating the opening calls of the survey. Natural objects, thus collaterally named, are not monuments in the sense in which that term is used with reference to termini, but they have no less positive effect. It is settled in Kentucky that where a boundary runs in a stated compass course, and is also said to be along a river or along a ridge, the compass course must yield to the course described by reference to the natural objects, so far as there is necessary inconsistency between the two courses, and following the sinuosities of the stream or ridge or running in a straight line as other considerations may determine. Bramblette v. Howard (Ky.) 93 S. W. 902, and cases cited. So much, viz., that the ridges are not boundaries but are aids in the nature of monuments, is clear upon the face of the grant, when once the court is informed as to the general topography of the.country. It is equally sure that at this point certainty ends and ambiguity begins. The ridges are to be used for locating the “opening calls.” Does that mean the first two, or the first four, or the first five? The terms of the grant suggest no answer.' If it is once conceded, as it is (plaintiff and defendant agree that the first two boundaries were marked' on the ground, and the third comer established on or near the summit), that this “aid” applies to more than the first course, it cannot be certainly known that the aid should be abandoned until it has been exhausted by reaching the end of the *55ridges. Whether it should leave the ridges at the end of the second course, as defendant contends, or at the end of the fifth course, as plaintiff contends, or at the beginning of the fifth course (a construction satisfactory to neither), must depend upon the question of intent— what land did the surveyor intend to include within the stated boundaries — when did he intend to leave the ridges? Of course, we do not refer to any secret intent, or one not inferable from his chosen words when properly interpreted, and it would be more accurate to say “What did he intend his language to mean?” Doubtless, where the description is clear upon the face of the instrument and is not varied by proof of actual and inconsistent location upon the ground, it will be of no importance that the surveyor intended to include something else; but, where the description is fairly capable of being understood in more than one way, the liberal rule of Kentucky does not pronounce it void, but ascertains the real intent of the language. It is only because intent is, in such cases, the controlling element that a monument controls an inconsistent course or distance, and that the actual running on the ground, when proved, controls the described line. That such intent is thus, in such cases, the ultimate inquiry, is the clear rule in Kentucky.

In Mercer v. Bate, 4 J. J. Marsh, 334, 344, the court said:

“But there is, in this respect, a palpable and essential difference betwixt an actual and an ideal line, or a marked and open line. “And as in the one case Madison might he bounded by the marked line wheresoever it might be (if he made no mistake), so in the other he must be restricted to the line as it appeared to be, and as he believed it was when he called to adjoin it. In the first case, he would have a right to the marked line, because, being visible, he knew where it was, and therefore intended that, as marked, ifr should be his boundary. In the last case, for the very same reason, wherever he supposed the invisible lines to run, he must be bounded, because be intended when he made his survey to be, and therefore was bounded by it.”

In Ralston v. McClurg, 39 Ky. (9 Dana) 338, it was said:

“And the cardinal object is to ascertain what the surveyor would have done if he had gone on to complete the work. Beckley v. Bryan, etc. [Ky. Dec. 91; Preston’s Heirs v. Bowman] 2 Bibb, 493. This is to be ascertained, not by vague conjecture, but by rational deductions from his report, as compared with the existing facts.”

Obviously, such intent will be a question of law for the court, if there is no dispute of fact (though such conflict may lie, not in the primary facts, but rather in the inferenial facts to be deduced therefrom (Watkins v. King [C. C. A. 4] 118 Fed. 524, 534, 55 C. C. A. 290; Kentucky v. Hamilton [C. C. A. 6] 63 Fed. 93, 97, 11 C. C. A. 42; International Co. v. Stadler [C. C. A. 6] 212 Fed. 378, 382, 129 C. C. A. 54); otherwise, the intent will be a question of fact.

On behalf of plaintiff, it is said that the boundaries can be made to follow the ridges to Eagle Gap without any greater variations from stated courses and distances, shape of plat or acreage, than was very common at that time in that locality; that the naming of the two ridges shows that it could not have been intended to leave the ridges before the second one was reached at all; that the sixth course, and the resulting curious form of that part of the plat figure, can be explained *56upon plaintiff’s theory, but are otherwise unintelligible; that Stamper intended to take in all of this Turkey creek watershed, and supposed he had; that he conveyed various portions by deeds, using the summit of the ridge as a southern boundary, and that his grantees occupied thereunder; that it was generally understood in the community that this was his land, and that such understanding (at least as to most of the land) was never disputed until 1903, when some one made the locations under which defendant mainly claims; and so on. Defendant either disputes these claims or, to destroy their force, offers something else in explanation or rebuttal. Without balancing the force of these things, or deciding which ones are or are not legally pertinent, we think they leave the ultimate fact not clear enough to be decided by .the court, as matter of law; they make an issue peculiarly appropriate for decision by a local jury.

In addition to these considerations, most, if not all, of which were urged in the former case, but in the different light caused by the different theory then under discussion, the present case presents some entirely new testimony which is claimed to show that the third, fourth, and fifth courses, as well as the first two, were actually marked on the ground, and are evidenced even yet by a certain line of blazed trees running the summit of the ridge, all the way to Eagle Gap. Defendant meets this by claiming that, so far as such blazed trees exist, they were marked at a much later period and in connection with the making of other surveys. It is, of course, conceded that if, in fact, this line was so run and marked by or under direction of the surveyor and at the time of the survey, plaintiff is right in its theory. The evidence offered for plaintiff on this issue is by no means absolutely convincing. If we were trying the fact, we might or might not think it at all persuasive ; about that, we express no opinion; we are satisfied that it so far tended to establish plaintiff's claim that the court could not withdraw that issue from the jury; and we think it follows, for both reasons' stated, that the issue of what lines the surveyor actually intended to describe as his southern boundary by the language he used should have been submitted to the jury.

[4] The previous decision of this court did not prevent this result. This is for two reasons: First, because the court was then deciding in; part an issue of fact, and on the present trial there is additional, pertinent evidence which must be considered; and, second, because the theory of construction now urged by plaintiff was not presented to, or considered by, the court in the former opinion. True, the facts were there which would have raised the theory, if any one had thought of it; but in applying the rule of stare decisis, as distinguished from res judicata, the subsequent court is bound only in these matters of law which the prior court did in fact consider and decide. It is not bound by any implied holdings dependent on legal questions which lurked in the record but were not observed. New v. Oklahoma, 195 U. S. 252, 256, 25 Sup. Ct. 68, 49 L. Ed. 182. What we intend to hold regarding the present effect of the former decision is this: We consider it to decide,, as matter of law, that the language of the survey, read in connection with the facts then and now undisputed, did not call for a survey of *5711 sides, but for one of 9, and that the ridges were to serve as aids in locating the course and distance calls. To this extent, it is both entitled to be followed under the rule of stare decisis, and its rightfulness is confirmed by the further study the question has now had. So far as it can be thought by implication to have declared, as matter of law, that defendant’s theory of construction and application was right, this declaration must be treated like any other ruling or assumption of law, made when the parties were silent regarding what later turns out to be a controlling consideration, and which therefore practically was taken for granted and by tacit consent, and must yield to a later conclusion to he made in a case between other parties and after actual study of the conflicting arguments. So far as it is a conclusion of fact based upon disputable inferences from the evidence, it pertains to what is, in the present' case, a question for the jury, and it is immaterial whether we think.it rightly decided.

We are not required to say what the situation would be if such an implied holding or declaration of a matter of law had become a rule of property. No title or right acquired on the faith of the former decision is here involved.

It results that the judgment must be reversed, with costs, and the case remanded for a new trial. Some questions of admissibility and force of evidence are presented upon other assignments of error, but they will not arise again under the same surroundings as on the first trial, and they cannot, with advantage, be now considered.

Reference should be had to the exhaustive discussion by Judge Cochran, in the court below, on the motion to remand (C. C.) 191 Fed. 899. See, also, cases cited by Judge Sanford, in Western Union Co. v. Louisville Co. (D. C.) 201 Fed. 932, 943.

Our statements as to the courses and distances of the southern boundaries, when run according to plaintiff’s theory, we have estimated by reference to the map made by plaintiff’s surveyor. They should therefore be taken only as stating what plaintiff's evidence tended to show, and they are not intended to be- mathematically accurate.