delivered the opinion oe the Court.
At the regular city election in the city of Charleston on *620the third Tuesday of April, 1895, William R. Highland and Henry A. Heal were candidates for the office of mayor. The city was divided into four wards. Appellants, who composed the city council, at a meeting held as provided by ordinance for the purpose of canvassing the returns, referred the matter to a committee of three of their number, of whom a majority found and reported the result as to that office to be that Highland received 628 votes and Heal 625. A motion being thereupon made and seconded that the report be received and the candidates as named be declared elected, an amendment was offered to receive it on all except Highland, who was defeated by a tie vote—the mayor, Patton, not voting—and the original motion was then adopted by a vote of four to three, one of the aldermen who had voted for the amendment not voting. A motion then made, that the candidates respectively reported as having received a majority of the votes be declared elected, was carried unanimously as to each except Highland for mayor, and as to him by a vote of four to three. These proceedings appear more fully in the journal of the city council.
The only question in the case arises upon the returns of the election in the second ward. As to them the evidence is that the poll books, tally list and judges’ certificate—which are the constituents of the returns (The People, etc., ex rel. v. Ruyle, 91 Ill. 528) seem not to agree. The list of voters in the poll book numbers 347; the certificate of the judges, attested by the clerks of the election, is that “ W. R. Highland received 188 votes for mayor,” and “ II. A. Heal received 158 votes for mayor,” making only 346 in all; the tally list shows 193 lines opposite the name of Highland and 158 opposite that of Heal (making 351) and yet the footing of the tallies for Highland are carried out as 188, being five less than actually and certainly appear.
The number of lines shown being taken to be the number of votes cast for Highland would give him a majority of three, while the number as carried out, which is also the number certified by the judges, would give Heal a majority of two.
*621Appellants having adopted these lines as showing the number of votes cast in the second ward for Highland and declared the result accordingly, the relator filed the petition herein for mandamus commanding them “ forthwith to examine and canvass the returns of said election for mayor, and, as a part of such canvass, to include therein the returns from said second ward as certified by said judges and clerks of election, showing 188 votes cast for William E. Highland, and 158 votes cast for your petitioner.” Several pleas were filed and issues joined, but all of the facts in the case and all the evidence are above sufficiently set forth; upon which the court, refusing all the instructions asked on behalf of appellants, directed the jury to find the issues for the plaintiff, and upon such finding, overruled a motion to set it aside, and rendered judgment awarding the writ of mandamus as prayed for and for costs. To these rulings and this judgment, exceptions were duly taken, and this appeal was prayed and allowed.
The petition, verified by affidavit of the relator, alleged, among other things, that he was present with his counsel at the meeting of appellants held to canvass the returns, and then and there insisted that they should canvass the returns from the second ward as certified to, and declare the result accordingly; which allegation was not in any way denied. Besides the city clerk, who identified all the records offered in evidence, the only witness examined or offered on either side was the petitioner, who testified only to facts showing his eligibility to the office for which he was a candidate at the election in question.
In their proceedings in the premises the city council were not acting as judges of the election or qualification of their own members, but as canvassers of the returns of the election of incoming city officers, in pursuance of the duty imposed by Sec. 10, Art. 4, Ch. 24 of R. S. (Hurd’s Ed. of 1893, p. 254), which was to “ examine and canvass the returns, and declare the result of the election, and cause a statement thereof to be entered upon its journals.”
The duty thus imposed is purely ministerial. The Peo*622ple ex rel. v. Head, 25 Ill. 325; Same v. Hilliard, 29 Id. 420. They determine nothing judicially. Their duty, and their whole duty, is prescribed by the statute, and that they are to do. It is “ a mere mechanical, or, rather, mathematical duty. They may probably judge whether the returns are in due form, but after that they can only compute the votes cast for the several candidates and declare the result,” said the court in the Head case. And in The People ex rel. v. Nordheim, 99 Ill. 561, it was said that “the canvassing board, from the very nature of the transaction, must in every case determine for itself whether the papers transmitted to it are, within the meaning of the law, returns of an election. The canvassers must be satisfied that they are genuine, and that they purport on their face to give the result of the election.” The board determines these questions “ for itself,” that is, ministerially; not judicially, that is, for the parties in interest. This sort of judgment is not judicial or discretionary. Hor is it peculiar to these ministerial officers. The unavoidable necessity for such judgment arises in the performance of almost every duty, which is, nevertheless, regarded as purely ministerial. The sheriff’s duty, imposed by his writ, is to serve it upon A B, or levy upon his property. There may be a question about his identity or thatof the property upon which the officer levies. He must decide it, but if he mistakes the fact, or the law which determines the fact, however honestly, he fails to perform the act required of him; and so in almost every such case. These election returns come from different officers in different election districts. Some one body must be appointed to canvass them all and declare the result of the whole. Very doubtful and difficult questions may arise in the attempt to ascertain whether they are in proper form or just what they purport lo give as the result of the election. But however difficult, it is a part of the ministerial duty of the canvassing board to solve them for itself. But they must solve them correctly, both as to the fact and the law, or they do not perform the ministerial act which the law commands, and their action may be controlled by mandamus.
*623Here the doubt is, which is correct as to the number of votes cast for Highland in the second ward—that shown by the tally list or that shown by the certificate of the judges % These differ. Both can not be correct. NTeither tends to explain the other or to show how the difference occurred. Neither is, in itself, at all uncertain. The carrying out or footing of the tally list does not make it uncertain, as it appears of itself, but is itself plainly an error. The record shows 193 lines, in squares of five each, excepting the last, which contains only three. They may have been counted by squares; and these appearing alike and being consecutively adjacent, one may have been inadvertently overlooked; or, possibly, one may have been filled by mistake or fraud, after the count Avas correctly made and the footing set down. But hoAv this was is matter of mere surmise. There is neither proof nor presumption as to how the difference betAveen the tally list and the judge’s certificate occurred. Having occurred and plainly appearing, the can Amassing board was to determine ministerially Avhich should be accepted as correct. It is to be regretted that they did not call in the judges and afford them an opportunity for explanations and correction, if they were prepared to make them, before the result was declared, as was suggested in some of the cases cited.
Which, then, if either, is the better evidence of the number of votes cast for Highland ?
The Avriter is inclined to the opinion that the law, in case of irreconcilable difference between the certificate of the judges and either or both of the other “constituents ” of the returns, must intend which is to govern the canvassing board. Otherwise it must give them some judicial power of determination, which is not claimed; or if claimed, is not warranted by the decision of the Supreme Court already referred to. He holds that in such cases the judges’ certificate controls. That is Avhat authenticates the other constituents, and they do not authenticate it; that where it is in proper form and without any uncertainty on its face, as it here appears, he apprehends it must control, though where *624it is informal or its import is uncertain, it may be aided or explained by either or both of the others, which are inferior in character, and of force for that purpose only.
A majority of the court, however, are unwilling to go so far; they deny to either constituent a controlling force, holding that all are to be considered and the result to be found from a consideration of all, which alike and together constitute the returns to be canvassed. They agree that in this case the poll books contradict the tally list and aid the certificate. If there were only 341 votes, as the poll books show, there could not have been 351 votes legally cast, as shown by the tally list, but there might have been 346, as shown by the footing and judges’ certificate. The tally list, therefore, was probably incorrect, and consequently also the declaration of the result of the election based upon it.
Then the returns in this case have not been canvassed and the result declared, as intended by the statute, according to either view of it, and mandamus is the appropriate remedy. Ruyle v. The People, 91 Ill. 525, and cases cited, supra.
The judgment of the Circuit Court will therefore be affirmed. • -