This controversy arises out of a local option election which was held in the City of Middlesboro on January 27, 1953. The question presented to the voters was: “Are you in favor of the sale of alcoholic beverages in the City of Middlesboro, Kentucky?” The election commissioners found as a result of the voting that there were 2,048 “No” votes and 1,977 “Yes” votes. There was a recount of the votes, as a result of which the circuit court adjudged that there were 2,044 “No" votes and 1,965 “Yes” votes, thus resulting in a “No” or “Dry” majority of 79 votes.
This contest action was filed by appellant in February, 1953. In March appellees filed a counter-contest. In September, the circuit court entered a judgment upholding the result of the election without filing a written opinion. Hence, we are without the benefit of the basis used by the court in reaching its decision. From that judgment this appeal is prosecuted.
The first question we must dispose of is whether KRS 117.655 and KRS 117.745 make it mandatory, or merely directory, that the voter- sign the comparative signature book before casting his ballot in an election governed by these statutes. KRS 117.745 is made applicable to all counties as well as to cities of the first class by virtue of the provisions of KRS 117.655.
KRS 117.745, our comparative signature statute, reads in pertinent parts as follows:
“Prior to every primary, special, or regular election in a city of the first class, the county clerk shall have prepared for each precinct of the city a comparative signature book * * *. Before a vote is cast by any person, at amg primary, special, or regular election in a city of the first class, he shall sign his name in the comparative signature book, unless he signed his registration record by making his mark, and the officers of election shall compare the signature of every such person with the signature of that person on the registration record form in the precinct register. * * * When the polls close, the precinct sheriff and judge of the opposite political faith shall forthwith convey the comparative signature book to the place where the county clerk receives other election paraphernalia. The board of registration commissioners shall have at that place plainly- identified, a duly authorized representative to whom the sheriff and judge aforesaid shall deliver the comparative signature book and secure a receipt therefor. The board shall retain the comparative signature books for at least the next two succeeding primary and general elections. * * ”
In only one of the ten precincts of Mid-dlesboro did all of the persons who voted in the election sign the .comparative signature book. In the other nine precincts the number of voters who failed to comply with the provisions of KRS 117.745 varied from only a few voters in. some precincts to nearly all the voters in one precinct.
In Beauchamp v. Willis, 300 Ky. 630, 189 S.W.2d 938, we expressed the view that the requirement in KRS 117.310, similar to KRS 117.745, that a voter sign his name on *69the signature book before He cast his vote, was mandatory.
It is our opinion that the primary purpose of KRS 117.745 is to prevent fraudulent manipulation of elections at the polling places. If the statute is followed it will be difficult for anyone to “stuff the ballot box” or for one person to cast votes in the names of others several times on one election day. The signature books properly signed provide a check list which can be used to detect these and other illegal practices. When the statute is ignored by the voters or by the election officers the door may be left open for manipulation of the election and the consequent destruction of the will of the electorate.
A determination that KRS 117.745 is merely directory would render it wholly ineffectual for the accomplishment of the purpose for which it was enacted. We think it is clear from the language of the statute that it was the legislative intent to make the signing of the comparative signature book a condition precedent to the right of the voter to cast a ballot.
Upon our review of the record we find that it was proven there were 108 votes which were illegal because the persons who cast them failed to sign the comparative signature books. Of these votes 98 were “No” and 10 were “Yes.” Consequently we must deduct 88 votes from the “No” majority, resulting in a “Yes” majority of 9 votes. In all but one of the precincts, these votes were proven to be “Yes” or “No” in each case by the testimony of the persons who cast them. In one precinct in which only three persons signed the comparative signature book, all but those three legal votes were thrown out and since appellant had the burden of proving how the illegal votes were cast, he must concede that the three legal votes were “No” votes.
Appellees urge that the procedure provided in KRS 117.745 for preserving the integrity of the comparative signature books was not followed and that therefore the books should not have been admitted in evidence. Without deciding the question of the admissibility of comparative' signature books in evidence where their integrity is not established, we must reject appellees’ contention because the books were offered and admitted in evidence without any objection by appellees.
In arriving at our total of illegal votes we have not considered testimony of voters which appellant improperly introduced on rebuttal when no testimony concerning their votes had been previously introduced. KRS 122.080(2) permits the introduction of rebuttal evidence in election contests within seven days after contestee has concluded. This statute cannot be construed to permit the contestant to introduce, by way of rebuttal, testimony concerning the votes of certain persons about whose votes no testimony has been offered in chief.
Five married women who voted in the election were still registered under their maiden names in violation of KRS 117.-640(2) which requires re-registration in such cases. Their votes were void for this reason. Of these votes, 4 were “No” and one “Yes”, resulting in a further deduction of 3 “No” votes, thus increasing the “Yes” majority to 12 votes.
Two felons who have not been restored to citizenship testified that they voted “No.” Their votes must be deducted from the “No” votes, making the “Yes” majority 14 votes.
Two votes were illegal for the reason that the persons who cast them voted openly without being sworn. One of these votes was “No” and the other was “Yes,” hence no deduction was made from either total for this reason.
Appellees contend that of the total of 86 absentee ballots cast in the election 55 were illegal for various reasons. Most of these votes are challenged on the ground that the persons in whose names they were cast were shown to have lived in places other than Middlesboro for some *70time prior to the election. Upon review of the testimony concerning these persons we find that 19 of these votes were illegal. In addition we find that 5 absentee votes were void as a result of irregularities in either the signature on the application or on the envelope, or in the attestation on the envelope. . Of the total absentee ballots cast, 81 were “Yes” and 5 were “No.” Since appellee has the burden of showing on which side of the question the illegal votes were cast, and since he introduced no testimony to that effect, he must concede that the five dry votes were among the illegal ballots in order to have them thrown out. Therefore we deduct 19 “Yes” votes from the “Yes” majority, resulting in a “No” 'majority of 5 votes.
Appellant contends that there were a large number of persons whose-votes were illegal because they voted in precincts ¡ in which they were not properly registered. We have examined the testimony, concerning each of these persons. Out of the votes' in this group we hold that there were 43. illegal “No” ballots and 7 illegal “Yes” bal- . lots. Consequently we have deducted 36. “No” votes from the “No” majority of 5 votes referred to above. -In arriving at the. results of this election, we have found it convenient to deal with the “No” majority of 79 votes, rather than the totals, as a starting point for our calculation. The final figure resulting from our computation is a “Yes” majority of 31 votes.
The record in this case reflects that the officials charged with conducting the election were apparently not correctly advised of their duties under KRS 117.745. And under this opinion we are declaring void many votes which were cast by voters who testified that they asked to sign the comparative signature book before voting but were informed by the precinct election officers it was unnecessary. While our interpretation of the involved statute may appear harsh, we cannot obey and give force to the statute in any other manner. However, there is much lurking in the shadows of this record which impel us to take a second look at certain testimony which we have approved that affects our computation of the ballots. Numerous persons, who were identified with “wet” interests, testified in this case that they voted “dry.” In considering similar testimony given in a contest of a “wet and dry” election held in this same county in 1947, we said:
“The reading of the testimony of the three above witnesses creates a sort of squeamish feeling. To accept such testimony would be credulity indeed, or obliging readiness to believe what was said by these witnesses without submitting their testimony to the test of critical judgment. True, one should not go so far as to be unable to hear and examine what cannot be tolerated, but here we have men operating a beer and "whiskey emporium who would come into court and say that they voted for prohibition because they can make more money by operating their business in local option territory. Do they mean to say that they plan and intend openly to flaunt and violate the law? There is no necessity of writing a lot of sentimental twaddle about this situation, but it would be difficult obligingly to conclude from their testimony that they really voted for prohibition. It is not necessary to pursue logical refinement in arriving at a conclusion, but we rather unhesitatingly state that these votes should have been charged against the cause of contestants. * * *” Howard v. Jones, 306 Ky. 402, 208 S.W.2d 47, 49.
In view of what we have said, and the fact that numerous voters are being disfranchised because of their failure to sign the comparative signature book, we think the so-called election did not permit a fair expression of the will of the electorate. Therefore, we decide that the election was void.
Judgment reversed, with directions to enter a judgment not inconsistent with this opinion.