delivered the opinion of the court:
F. E. Roland, appellee, and U. L. Walker, appellant, were opposing candidates for the office of mayor of the. city of Herrin, in Williamson county, at the city election held on April 20, 1909. Appellee was a candidate on the “Labor ticket” and appellant was a candidate for mayor on the “Good Government ticket.” The city of Flerrin is divided into four wards, numbered from-1 to 4, inclusive. By the returns, duly certified to by the judg'es and clerks of the election, appellee received in ward No. 1, 157 votes; No. 2, 234; No. 3, 131 and in ward No. 4, 66 votes,—a total of 588 votes. By the same returns appellant received^ 119 votes in ward No. 1; 137 in No. 2; 190 in No. 3, and 146 in ward No. 4,—making a total of 592 votes. The returns were canvassed by the city council of Herrin on April 23, 1909, and Walker was declared duly elected mayor of the city. On May 19, following, appellee filed his petition for a contest of this election in the circuit court of Williamson county. Upon the trial of the case the ballots were re-courtted, with the result that' appellee was given 148 votes in ward No. 1, 246 in'No. 2, 131 in No. 3 and 67 in ward No. 4, which made a total of 592 votes. Appellant was given 111 in ward No. 1, 122 in No. 2, 190 in No. 3 and 146 in ward No. 4, making his total number of votes 569. The re-count gave appellee 12 votes more in ward No, 2 and appellant 15 votes less in that ward than was shown by the certified returns. The changes made by the re-count in this ward changed the result of the election, and accordingly the circuit court declared appellee duly elected. From this decree Walker has appealed to this court.
It is thus apparent that the only question involved which requires consideration is, which is the better evidence of the result of the election in ward No. 2?—the certified returns or the ballots themselves? While there *131are other minor questions involved, the determination of this question settles the whole controversy.
Where the ballots have been properly preserved they are the best evidence of the result of the election. (Bonney v. Finch, 180 Ill. 133; Caldwell v. McElvain, 184 id. 552.) But in order that the ballots should be controlling as evidence, it must affirmatively appear that they have been preserved in the manner and by the officers required by the statute. (Beall v. Albert, 159 Ill. 127; Jeter v. Headley, 186 id. 34.) If the evidence discredits both the ballots and the returns, the true result of the election must be determined by a consideration of both, and of all other circumstances which will aid in determining the truth of the matter at issue. Dooley v. VanHohenstein, 170 Ill. 630; Smith v. Reid, 223 id. 493.
The appellant contends that the ballots have not been properly preserved, and appellee contends that there were irregularities tending to impeach the returns in the manner of counting the ballots and declaring the result by the election officers. Both of these contentions find some support in the evidence. After the election was closed in ward No. 2 a count of the ballots in the box was made and the number was found to agree with the number of voters shown to have voted by the poll-books. Harland, Ellis and York were the judges, and DeArmond and Neely were the clerks of the election in ward No'. 2. After the ballots had been counted and found to correspond with the poll-books, the judges, in the presence of three challengers,—one representing the Labor party, another the Good Government party and another the Socialist party,— proceeded to separate the ballots into four piles,—one representing the “straight” tickets for each of the three parties represented in the election and one pile containing the “mixed” or “scratched” ballots. In separating the ballots into these four piles all of the judges did not examine and verify all of the ballots, but Mr. Harland, one of the *132judges, and Mr-. Humphrey, a challenger for the Good Government ticket, went over and examined all of the ballots in the different piles to determine whether they had been properly classified. The other judges and clerks did not examine any of the ballots except such as they personally handled in classifying them into these piles. Each of the persons who handled these ballots, or any portion of them, in separating them into piles, testifies that he did not put any ballot marked with a cross in the circle of the Good Government ticket and a cross in the square to title left of the name of appellee in a Good Government pile of straight tickets, but that all of the ballots placed in that pile by them were straight ballots. The' result of this classification of the ballots was that there were 178 straight Labor ballots, 84 straight Good Government ballots, 28 straight Socialist ballots, 126 mixed or scratched ballots and about-9 blanks. After the ballots were separated into piles the straight ballots were counted, and Mr. Harland announced the number and the clerks tallied up these numbers on the tally-sheets. After they were thus tallied the straight ballots were strung by Ellis on a wire in bunches without being folded, a knife being used to cut a slit through the bunches of ballots to admit the wire passing through. The 178 straight Labor ballots were first counted and strung, and then a piece of blotting paper, on which one of the clerks wrote the words “178' Labor,” was strung on the wire to separate the Labor ballots from the next batch to be strung on the wire. The 84 straight Good Government ballots were tallied and strung on the wire by Ellis in the same way, and another piece of blotting paper, on which was written “84 Good Gover,” was put on the wire. Then'the 28 straight Socialist ballots were tallied and strung on the wire in the same way, and another piece of blotting paper, on which was written “28 Sos,” was strung on the wire. Then the mixed or scratched ballots were taken up, examined one by one. and *133tallied, and strung on the wire by the election judges. After the ballots were thus strung on the wire, the wire was brought around the ballots and fastened but was not sealed. The ballots were then put in a canvas sack with a draw-string at the top, the top drawn together with the and a knot tied, and sealing wax was melted and poured on the knot but no impression was made on the wax. There is some controversy in the evidence as to the color of the sealing wax, some of the witnesses saying that it was a grayish-green, while others remember it as red sealing wax. After the ballots were thus sealed up, Mr. Harland, one of the judges, took them home with him and locked them in his trunk and kept them until he delivered them to the city clerk at his office, which was on the afternoon o°f the next day, being April 21, 1909. Appellee, Roland, was then the city clerk and the ballots were delivered to him by election judge Harland. Roland placed the ballots in a large roll-top desk in his office, where they remained until April 23. The city clerk’s office, in which the ballots were kept from April 21 until the 23d, was occupied by a firm of insurance agents. On April 23 Roland made two bunches of the ballots, poll-books and tally-sheets and took them to the City National Bank in Herrin, where they were thrown on the floor in the corner of a fire-proof vault. The book-keeper of the bank, Mr. Frazier, was present when the packages were placed in the vault. In this vault the bank maintained safety deposit boxes for the use of their customers, and some twenty or more persons had free access to the vault at all times when the bank was open for business, besides the officers and employees of the bank. The ballots remained in this condition, on the floor of the bank vault, until about the first of June, following, when Harry Wood, the successor to appellee as city clerk, secured the ballots and placed them in a large wooden box, locked it, and sealed it by pasting papers over the top and sides of the *134box in such way that the opening of the box would tear the papers. When the ballots from the Second ward were opened and re-counted in the circuit court it was found that 13 of the 84 Good Government ballots which were .marked with a cross in the circle and had been counted as straight ballots and for appellant, had a cross in the square before the name of F. F,. Roland and also a cross by the name of another candidate on the Labor ticket. All of the judges and clerks who handled any of the ballots in ward No. 2, and Harland and Humphrey, both of whom examined all of them, testified that no ballots were placed among the 84 straight Good Government ballots marked as these 13 ballots are marked, being appellant’s Exhibits 13 to 25, inclusive, and that no such ballots were placed in the Good Government pile of straight ballots.
Appellee seeks to account for the 13 ballots in question being counted for appellant through a careless oversight on the part of the election judges. These ballots, together with all others voted in ward No. 2, have been certified to this court by the trial judge and we have carefully inspected them. The ballots are printed on clear white paper and are five inches by nine in size. The headings of the three tickets are printed in large letters to the right of the circle, running from left to right,—“Labor,” “Socialist” and “Good Government.” The circles are one-half inch in diameter. The squares opposite the names of the candidates are one-quarter of an inch on the side. There were only five offices to be filled, as follows: Mayor, city clerk, city treasurer, city attorney and alderman. Each of the three parties had a candidate running for each office to be filled. In other words, there were no blank places for names to be written on the ballot. The crosses in the square opposite the appellee’s- name are all very distinctly made. All appear to have been made with an indelible purple pencil. They are so distinct that by actual test we have ascertained that the crosses opposite appellee’s name, *135and all other markings on the ballots, can be distinctly seen-.by a person of normal sight, in ordinary daylight, a distance of twenty feet. Considering the small size of these ballots and the distinct character of their markings it is practically impossible that the judges and clerks who canvassed them should make a mistake on 13 ballots out of 84. If the judges and clerks and challengers were so careless as to overlook 13 ballots out of the 84 Good Government ballots, it is singular that like mistakes did not occur in reference to the Labor, Socialist and mixed tickets.
It is also a matter of importance to be noted, that while the marks in the squares on the 13 ballots in question show that they were marked for first one and then another candidate on the Labor ticket, they are all carefully marked in the square opposite appellee’s name. Some of these ballots are marked for appellee and one other candidate on the Labor ticket and also for one candidate on the Socialist ticket. They are all marked in the circle of the Good Government ticket. On the ballots which are thus voted for appellee and for one of his associates on the Labor ticket and for one person on the Socialist ticket, only two persons are left for subordinate offices, which are voted for by the cross in the circle of the Good Government ticket. It will thus be seen that the ballots show that the voter, in order to vote for two persons for subordinate offices on the Good Government ticket, marked a cross in the circle of that ticket, while on the same ballot, in order to vote for the mayor, which was. the principal office to be filled and manifestly the office which would have most influence in determining the voter’s party ciioice, the voter marked a cross in the squares. This, to say the least, is a very unusual and awkward way to vote a mixed ticket, but it is a legal method of voting, and it is not improbable that a voter, now and then, might resort to this method of expressing his choice; but when 13 ballots are found out of 84 all marked in the circle of the Good Government *136ticket and all marked in the square opposite the name of the candidate for mayor on the Labor ticket, and all of them marked in the square opposite some other candidate’s name on the Labor ticket, it is ground for serious suspicion. We cannot refrain from remarking how strange it is that none of the thirteen voters marked the circle of the Labor ticket in order to express their choice for mayor and the other candidate on that ticket for whom they desired to vote, instead of marking the circle always on the Good Government ticket.
Exhibit No. 14, which is one of these ballots in question, is marked in the square for mayor on the Labor ticket and also in the square of .that ticket for city attorney, and is marked in the square for city treasurer on the Socialist ticket and for alderman on the Socialist ticket. This only left the office of city clerk to be voted for on the Good Government ticket, and yet, in order to vote for the city clerk alone on the Good Government ticket, the voter appears to have marked the circle when, in fact, he voted for more men on each of the other tickets than he did on the Good Government ticket.
In addition to the suspicious circumstances already pointed out, a close, critical examination of the crosses made in these 13 ballots in the Good Government circle and the other crosses made on the same ballots opposite the squares indicates that they were made by different persons. Thus, on Exhibit No. 25, which is one of these 13 ballots, the cross in the circle is poorly executed and the lines are slightly irregular, indicating that the person who marked it was nervous and made a wavy line, while the crosses in the squares on the same ballot show no such indications but the lines in the squares are uniform and steady. . Again, in Exhibits Nos. 15 and 17 the cross in the circle appears to have been made with a blunt pencil while the ballot was evidently lying upon a rough surface, and the pressure of the pencil so indents the cross that it *137is plainly visible from the opposite side of the paper, while the marks in the squares on the Labor ticket on these ballots appear to have been made when the ballot was lying upon a smooth surface, and the markings do not show on the opposite side of the paper as they do in the circle. These suspicious circumstances, when considered in connection with the illegal and careless manner in which the ballots were preserved after they went into "the hands of appellee, so completely impeached the integrity of the ballots that they cannot be said to overcome the official • returns made by the election officers.
With the facts disclosed by this record before us, our conclusion is that the-ballots were so exposed as to give an opportunity for unauthorized persons to tamper with them, and the circumstances tend strongly to satisfy us that they were, in fact, tampered with by someone after they left the hands of the election officers. There is no charge that the election officers were guilty of any intentional fraud or misconduct of any kind. The only criticism that is made is that they were not as careful as they might have been in examining these ballots. This criticism is not without some justification in the evidence. The election officers should have followed strictly the direction of the statute in canvassing these returns, but, notwithstanding their failure to observe the statute in every particular, we are satisfied that their returns are entitled to more weight in determining the issue here involved than the ballots. The court erred in holding that the ballots were the best evidence and in rejecting the returns of the election officers. Upon the face of the returns appellant was elected mayor of the city of Herrin.
The judgment of the circuit court of Williamson county is reversed and the cause remanded, with directions to enter a judgment in favor of appellant and against appellee.
Reversed and remanded, with directions.