Ejectment. The plaintiff claims through a United States patent issued to the Central Pacific Railroad Company, as successor in interest of the Western Pacific Railroad Company, under the act of July 1, 1862, and the amendatory acts. The defendant was in possession at the commencement of the action, claiming that the land was public land, and that he had complied with the pre-emption laws. .The court below gave judgment for the defendant, and the plaintiff appeals. Inasmuch as the plaintiff relies solely upon a paper title, he must recover upon the strength of that title. It is therefore immaterial whether the defendant’s pre-emption proceedings were regular or not. The question is as to the validity of the patent.

Upon the former appeal, the judgment was reversed, because the court below rejected evidence to the effect that, at the time the grant from Congress to the company took effect, the land was within the limits of a Mexican grant then sub judiee: Carr v. Quigley, 57 Cal. 395. The necessary result of this decision (although it is not stated in terms) is that, *825if at said time the land was within the exterior limits of a Mexican grant then sub judice, the patent was void, and could be attacked collaterally. The case does not appear to have been taken to the supreme court of the United States, and consequently the decision became the law of the case, and is not now subject to review. This is a familiar rule in this state, and, as we understand, it prevails in the federal court: Ex parte Sibbald, 12 Pet. 491, 9 L. Ed. 1169; Bridge Co. v. Stewart, 3 How. (U. S.) 413, 11 L. Ed. 658.

The record before us recites that, upon the retrial in the court below, the defendant “proved by testimony and established,” among other things, that the Mexican government, in 1839, granted to Jose Noriega and Robert Livermore a tract of land known as “Las Pocitas,” and described as follows, viz.: ‘ ‘ Bounded on the north by the Lomas de las Cuevas; • on the east by the Sierra de Buenos Ayres; on the south by the dividing line of the establishment of San Jose, and on the west by the rancho of Don Jose Dolores Pacheco; containing in all two square leagues, a little more or less, provided that quantity be contained within the said boundaries; and, if less than that quantity be found to be contained therein, then that less quantity, and all of said described tract of land.” This grant was confirmed by the same description by the board of land commissioners on February 14, 1854, and by the United States district court on February 18, 1859. Among the things which defendant ‘' proved by testimony and established ’ ’ was that the tract in controversy was ‘1 within the boundaries designated and set forth in the said decree of the board of land commissioners and the United States district court.” And this was expressly stipulated on the motion for new trial. Was this grant sub judice at the time of the location of the line of the road, viz., on April 16, 1868 ? The decree of the district court was affirmed by the supreme court in 1861, the mandate being filed in the lower court in October, 1865. In that year a “final” survey was made by one Dyer, a deputy in the office of the surveyor general. By this survey it would seem that nearly ten leagues were included in the description above given. This survey was approved by the surveyor general in 1867, and by the commissioner of the land office in 1868, but was set aside by the Secretary of the Interior on July 30, 1868, on the ground that it included a *826great deal more land than.the two leagues which the grantees were entitled to. A new survey was made in 1869, and approved by the Secretary of the Interior in 1871, and a patent for the two leagues was issued upon this last survey on August 20, 1872.

We think that the grant must be considered sub judice until the final survey was approved by the Secretary of the Interior. The underlying idea of the decision on the former appeal, and of the case which it followed (Newhall v. Sanger, 92 U. S. 761, 23 L. Ed. 769), was that Congress had not by its grant undertaken to prejudge the validity of claims then pending, but had intended to except the land covered by such claims from the operation of its grant to the company. Now, to say that the congressional grant covered land within the exterior limits of a Mexican grant at a time when it could not be known what portion within those limits the Mexican grant would take, is to say that Congress had prejudged the matter. Until a final survey was made, it could not be known what portion within the exterior limits the two leagues would cover; and until such survey was approved by the Secretary of the Interior it was subject to be set aside by him, as was the case with the Dyer survey, made in 1865.

If, therefore, the case stood as it did at the close of the defendant’s evidence, it would, we think, be perfectly clear that the judgment should be affirmed. But, in order to overcome the effect of the defendant’s evidence, the plaintiff proved that a “preliminary” survey had been made in 1854 by one Lewis, a deputy in the surveyor general’s office. This survey included about four leagues, but did not take in the land in controversy here. It was “approved” by the surveyor general, but the date of such approval does not appear. The plaintiff also put in evidence proceedings in the land office in contests between other parties, and the opinions therein of the commissioner of the land office and the Secretary of the Interior, to the effect that land outside of the Lewis survey was not a part of the ranch; in other words, that the Lewis survey established the exterior limits of the Mexican grant. As a matter of course, the defendant here, not having been a party to such proceedings, cannot be affected by them. They were probably offered by counsel to show that, in the view of the authorities of the land office, the Lewis survey established *827the exterior boundaries of the rancho, and were doubtless admitted by the court below upon the somewhat loose principle that, inasmuch as there was no jury, they could do no harm.

It is a very serious question whether the appellant is not precluded by his stipulation from maintaining that the Lewis survey established the exterior boundaries of the rancho. He stipulated in the court below that the tract in controversy is outside of the Lewis survey, and his argument here is based upon that fact, which is necessary to his position. Taking that fact as established, it follows that, if the Lewis survey established the exterior boundaries, the tract in controversy must be outside of them. But the appellant further stipulated in the court below that the tract in controversy is within the exterior boundaries. Can he, in the face of this stipulation, be permitted to affirm on appeal that the tract is not within those boundaries? This argument does not cut both ways. It is true that the respondent stipulated that the tract is outside of the Lewis survey; but he has not stipulated, and does not affirm, that the Lewis survey established the exterior boundaries of the rancho. If, however, we assume, in favor of the appellant, that he is not precluded by the stipulation from maintaining that the Lewis survey established the exterior boundaries of the rancho, we must, with submission to the learned commissioner of the land office and the learned Secretary of the Interior, be permitted to doubt whether this survey had any such effect. We are unable to see how a “preliminary” survey, made before the decree of the district court, could be final, or how, if it was not final, it could 11 establish” the exterior boundaries of the rancho. The time to appeal to the district court had not expired, and therefore the case was then pending; and that is equivalent to saying that it was sub judice. To say that a pending case is not sub judice is a contradiction in terms. It is equally a contradiction in terms to say that a “preliminary” survey could effect a final segregation. For this reason it seems to us very clear that this survey did not “establish” anything, and we should dismiss the subject without further consideration were it not for the deference due to the opinion of the learned heads of the land department. Their opinion seems to us to rest upon several unfounded assumptions, which we will examine.

*8281. It is assumed that the Lewis survey was intended to be of the exterior boundaries of the rancho, as distinct from the tract claimed by the grantees; and that, although the survey did not stand as a delineation of the tract to which the grantees were entitled, it might stand as a delineation of a larger tract within which the tract claimed by the grantees was to be located. We see nothing in the record before us which justifies this assumption. In the instructions of the surveyor general to Lewis, it is stated that the attorneys for one of the claimants had applied for a survey of Las Pocitas, “a tract of land claimed by the latter”; that “the claim” had been confirmed, as per the description above quoted, by the board of land commissioners; and that the survey made was to “conform to the above decree.” Lewis, who was called as a witness, testified that what he made out was “the initiatory survey of the Rancho Las Pocitas claimed by Robert Liver-more.” We see nothing to indicate that the survey was intended to be of a tract from which the claim of Livermore was to be segregated. It purported to be what he claimed. It did not stand for this; and we do not see how it can stand for something else which it did not purport to be.

2. It is assumed that the survey was made in pursuance of a statute of the United States. The opinion of Commissioner Drummond says that the survey was “evidently made pursuant to act of Congress approved August 31, 1852.” The only act of this date which has any relation to the matter is the general appropriation bill, which contains the following provision : “For surveying private land claims in California which may have been presented in good faith to the board of land commissioners, t-wenty-two thousand five hundred dollars : provided, that the authority hereby conferred upon the surveyor general shall apply only to such unconfirmed cases as in the gradual extension of the lines of the public surveys he shall find within the immediate sphere of his operations, and which he is satisfied ought to be respected and actually surveyed in advance of confirmation”: 10 St. U. S. 91. This claim was not “unconfirmed” by the land commission; it does not appear to have been “within the immediate sphere of his operations” upon the public surveys; and, what seems absolutely conclusive, it was not a survey made at public expense, and hence could not come within the purview of a pro*829vision appropriating public money. The surveyor general, in his instructions to Lewis, expressly says to him: “As this is a preliminary survey, for your compensation you will make arrangements with the claimant. I shall add that under no circumstances will this office be responsible for any surveys connected with the execution of these instructions.” The survey, then, was not made under the above-mentioned act of 1852, as supposed by the learned commissioner. He refers for illustration to the act of March 3, 1853; but it is apparent that this act had nothing to do with the Lewis survey. It provides that the surveyor general “shall also cause all private claims to be surveyed after they have been confirmed, so far as may be necessary to complete the surveys of the public lands”; and that “for surveying the base and meridian lines, and private claims, and meandering navigable waters, the deputy surveyor shall be allowed not exceeding sixteen dollars per mile”: 10 St. U. S. 245. This provision is for a survey •to be made at the public expense where necessary “to complete the surveys of the public lands.” The Lewis survey does not appear to have been necessary for that purpose, and was made at the expense of private parties.

It may be added, although it is not suggested by the learned commissioner, that it was not made under the act of 1860, and hence is not within the decision of Southern Pac. R. R. Co. v. Garcia, 64 Cal. 515, 2 Pac. 397. The act of 1860, upon which that case was decided, was passed more than five years after the Lewis survey was made. If it be assumed that it could have any relation to the Lewis survey, there is nothing in the record to show that the publication required by the act was had. We see nothing to indicate that the survey had any official character or stamp. It was expressly said to be merely “preliminary,” and was made at private expense; the surveyor general instructing his deputy to “run any lines which any contending parties may desire you to run, if they are willing adequately to compensate you for such work”; and “to qualify as if performing public work.” The probability is that the claimants Imew that no final survey could be had until the matter had passed through the courts, or the time to appeal had expired, but wanted for their own information to get an approximate idea of where their lines would run; hence this “preliminary” survey.

*8303. The opinion of the learned commissioner seems to assume that this preliminary survey, which had no force of itself, could derive force from the acts in pais of the claimants; for he dwells at some length upon the fact that the survey was made at the instance of the claimants, who assisted therein, and consequently knew all about where the lines were run, but made no sort of protest or objection thereto. Of this it may be remarked that the record here does not show any such condition of affairs, the statements in the opinion of the commissioner not being evidence against the defendant. But, in addition to this, we do not think that the word “claim,” as-used in the legislation on the subject, was intended to be controlled or defined by the conduct of the parties. That would be too vague and uncertain a test. The “claims” referred to must have been founded upon writings. And where, as here, the writings were presented to and confirmed by the land commission, the description in the decree could not be added to or varied by the acts in pais of the parties. The principle of estoppel can have no application. In the first place, no action was taken by the government, or by anybody else, upon the faith of this survey; and, in the second place, subsequent preemptioners do not in any sense claim through the claimants under the Mexican grant.

We think the Lewis survey does not overcome the effect of the case made by the defendant, and we therefore advise that the judgment and order denying a new trial be affirmed.

We concur: Belcher, 0. C.; Foote, C.

By the COURT.—For the reasons given in the foregoing opinion the judgment and order denying a new trial are affirmed.

I dissent: Paterson, J.