Sedgwick, J.
This plaintiff was the county attorney for Sioux county for the years 1907, 1908, 1909, and 1910. For the first' three years of those named and for the first three-fourths of the year 1910 he was paid the salary at the rate of $500 a year, and afterwards he filed a claim with the county board for $750 balance due him for salary for the time *828named, claiming that he was entitled to $700 a year instead of $500. The county board rejected his claim, and upon appeal to the district court for Sioux county lie was allowed $150 only. From this judgment of the district court he has appealed.
1. During the time specified he filed claims for his salary quarterly at the rate of $500 a year, which were allowed, warrants duly drawn, and by him presented and paid. The first contention of the county is that, by filing these claims and accepting these warrants, he is now estopped to make any further claim for salary. The trial court, upon demurrer, held that this was no defense, and AAre think that this ruling in the trial court, under the condition of this record, was correct. Under the statute, as it then was, the salary of the county attorney depended upon the number of inhabitants of the county. In counties of 4,000 or more and less than 8,000 inhabitants the salary was fixed at $700 a year, and the record of the action of the county board upon the plaintiff’s claim recites: “The claim of M. J. O’Connell for $750 for back pay as county attorney for years 1907, 1908, 1909, and first 3 quarters of 1910, claiming that the number of inhabitants exceeded 4,000 or more, was taken up, and OAving to the fact that the board has no Avay of ascertaining that the population of Sioux county was 4,000 or more for the abcwe years mentioned, on motion the claim was rejected.” The record shows that no other issue urns presented to or determined by the county board than the issue so determined. The answer filed in the district court presented only this issue. This record shows that it was conceded by the county in the action of its county board that the justice of the plaintiff’s claim depended solely upon the number of inhabitants in the county during the years of his services. An amended answer was filed alleging the estoppel relied upon, but, upon demurrer, this issue was by the court stricken from the answer. Under these circumstances the case, ought now to be determined upon the issue so tried and determined.
*8292. The question is whether during the years named Sioux county contained 4,000 or more inhabitants.. Section 19, ch. 7, Comp. St. 1907, fixes the salaries of county attorneys in these words: “In counties having from four thousand (4,000) to eight thousand (8,000) population, seven hundred dollars ($700.)” It did not specify how the number of inhabitants should be determined. By the federal census of 1900 Sioux county contained 2,055 inhabitants, and by the federal census of 1910 it contained 5,599. In some cases, and for some purposes, the statute requires that the population of a given political subdivision can be determined only by a duly authorized census. In other cases it contemplates that the number of inhabitants may be determined by other methods. It has been several times held by this court, and uniformly so far as the writer recalls, that when, as in this case, the statute makes no provision as to the method or time of ascertaining the number of inhabitants, the latest general census of the federal government does not necessarily control. The salary of the county attorney for any year depends upon the number of inhabitants of the county during that year, and that fact is to be determined, as are other facts in litigation, by the best competent evidence that can be adduced under the circumstances. State v. Long, 17 Neb. 502; Kokes v. State, 55 Neb. 691; State v. Davis, 66 Neb. 333. In the opinions in these cases there is some peculiar discussion as to the ratio between the number of votes cast at a general election and the number of inhabitants of the voting precinct. In State v. Long, supra, Judge Maxwell, in the opinion of the court, said: “The votes polled at the election in 1883, when multiplied by the well-known ratio of population to the number of voters in a county, is evidence tending to prove the number of inhabitants.” And in Kokes v. State, supra, Harrison, C. J., in the opinion of the court, commented quite at large on this expression, and, apparently without disapproving of evidence of the number of votes polled as determining the number of inhabitants, concludes that the evidence in that *830case was not sufficient to prove that Valley county at the time under consideration had a population of 8,000. The expression of Judge Maxwell above quoted we think, if rightly considered, is not subject to criticism. There may not be a fixed ratio between the number of votes polled and the number of the inhabitants in the precinct, because there .is no fixed ratio between the number of voters in the precinct and the number of votes polled. There is, however, a ratio between the number of voters in the. precinct and the number of inhabitants therein, and if the exact ratio is not known it is uniformly found that there are at least a certain number of inhabitants for every given number of voters residing in the precinct. There is also a comparatively uniform proportion between the number of children of school age in a given precinct and the actual residents of the precinct. To say that the number of inhabitants can be accurately determined by knowing the number of children who actually attend school, or knowing the number of voters who do actually exercise the elective franchise, would, of course, be a non sequitur. There must, of course, be as many school children in the county as attend school therein, and as many voters as the votes actually polled. If the number of votes polled in the county is multiplied by the ratio of population to the number of voters residing in the county there must be as many inhabitants in the county as the number so obtained, since the number of voters in the county must always be at least as great as the number of votes polled. It is not necessary in this case to inquire whether there is a well-known and uniform ratio between the population of a county and the voters and school children therein. We have the number of inhabitants by the census of 1900, but the record does not show the census of school children in that year, and we have no basis from which to find the ratio of the number of school children to the number of inhabitants' for the years .in question. The number of school children in the county in 1910 was 1,696. This number multiplied by 3.25 was very near the population *831of the county as shown by the government census. The number of school children in 1907 was 8G3, multiplied by 3.25 is 2,804.75. The number of school children for 1908 was 931; for 1909 it was 1,34.8; for 1910 it was 1,696. In 1908 the total vote cast was 1,034, but in 1909 the vote cast was only 616, and in 1910 it was 880. If, therefore, we say that this evidence was competent and tends to show that the ratio of school children to the population during the respective years was 3.25, it shows affirmatively that the population in two of the years was less than 1,000, and the whole evidence does not establish with sufficient certainty the plaintiff’s claim for any year, except the year 1910, which was allowed by the district court.
The judgment of the district court is
Affirmed.
Letton, Fawcett and Hamer, JJ., not sitting.