Concurring Opinion by
Mr. Justice Roberts:I join in the opinion of the Chief Justice. An examination of the record facts, particularly of the voting machines which were set up as illustrated in the diagram, which will be found in the reporter’s state*341ment of the case, convinces me of the correctness of the result here reached. However, I do feel that I must add these observations.
This opinion is in no way a reflection upon any of the litigants. It is rather an expression of regret for the unfortunate positions in which they have been placed by the inexcusable conduct of the County Board of Elections. Neither of the reasons offered by the Board for its failure to prepare properly the voting machines by locking levers over blank spaces in rows C and D, as legally commanded, even approaches an acceptable explanation.1 The Board improperly disregarded mandatory provisions of the Election Code and failed to perform clearly directed legal duties.
It is erroneous to assume that those voters who pulled levers in row C intended to vote for candidates whose names appeared above those levers but actually in row B. Such an assumption would be founded on nothing more than the exercise of retroactive clairvoyance as to what was intended on April 28 by 5,624 lever markings over blank spaces recorded on 908 voting machines in 512 election districts in 13 Philadelphia wards.
It is unmistakably evident from the ballot faces of the voting machines, as illustrated in the diagram, that votes for Democratic candidates whose names appear in the second horizontal row, marked “B”, may be recorded only by the operation of levers above their names, as the hand pointer in that row clearly indicates. Moreover, there runs through the center of the blanks in row C (and also D) a heavy black line from column 1 through column 39 where an arrow points to the special election in column 40.
*342Likewise, the sample ballot contains four specific instructions which state with unmistakable clarity that to vote for a candidate, “Turn down a pointer over the name of each candidate in your own party row that you wish to vote for.” (Emphasis in the original.) This instruction also advises that “the Democratic Party” occupies “the second horizontal row” (here row B), and in at least two other places the voter is cautioned that he may vote only for candidates of his party in the primary.
It is an undisputed fact that levers in row C were operated in columns 1, 17, 18, 19 and others in which no names appeared either above in row B or below the levers in row C. Of particular interest is the use of columns 17, 18 and 19 in row C for the election of delegates to the Republican National Convention. If the “votes” recorded in columns 2, 3 and 4 of row C were validly cast for candidates in row B, so also were the “votes” in the remainder of row C, an obviously absurd conclusion.
It is conceded that all voters who entered the 908 voting machines in question could, irrespective of their party registrations, turn down levers in any of the columns in row C from column 1 through column 39 as well as the lever in column 40 for the special election. It is likewise undisputed that voters other than registered Democrats could and did operate levers in row C.2 It would be fallacious to assume that since more Democrats appeared at the polls than the total *343votes cast for all Democratic senatorial candidates in rows B and C (assuming that the votes in C were for those candidates), only registered Democrats operated levers in row C.3 Those voters who pulled levers in row C may have done so for the same reason as those Democrats (7,168) who failed or refused to vote for any of the senatorial candidates.
Nor can it be assumed that only registered Democrats voted in row C because it is unlawful to cross party lines in the primary, although the machines did not preclude such crossovers.4 It may just as easily be assumed that the Democratic voters who intended to vote for the candidates followed the explicit voting instructions and pulled levers only in row B. Otherwise, the Legislature engaged in meaningless enactments of precautionary requirements to prevent crossover voting. The Legislature would have needed only to rely on a presumption that no one acts or votes contrary to the instructions and the Election Code.
To hold that the votes cast below the names of the candidates are valid would require the interpretation of “upon or adjacent” in §1216 (d) of the Election Code as “either upon or adjacent.” Under such a view, voters would be given the choice of pulling down the lever over the candidate’s name, as the instructions provide, or pulling down the lever under the name, as appellants would have us hold. Appellants contend that the levers in row C, with no names ap*344pearing below them, are “adjacent” to the names in row B and would have us credit row C candidates with votes both above and below their names. The appearance of names in row C below the levers in row C would not make those levers any less “adjacent” to the names in row B. Thus, the absurd situation would result that in a general election, for example, the candidate in row A might also claim votes cast below in row B as “intended” for his name in row A and “adjacent” to it.5
Some concern has been voiced that by not counting the markings over the blanks in columns 2, 3 and 4 in row C and by not attributing those votes to candidates whose names appear in row B, 5,624 voters will be disfranchised. The application of this kind of paternalism would ignore the fact that 24 out of every 25 voters6 who entered the voting booths in the districts in question voted correctly, regularly and in accordance with the instructions to voters. Also, no one knows who pulled the levers in row C, their party registrations, and the reasons for preferring to pull levers in row C rather than those in either rows A or B (it being admitted that all voters could operate levers in row C). Moreover, no one knows whether those who operated the levers in row C did so by mistake or by design, just as no one knows why 7,168 Democratic voters who entered the voting booths chose not to cast votes for any of the senatorial candidates.
Realistically, those voters who pulled improper *345levers in row C, columns 2, 3 and 4, are no more disfranchised than those who operated levers in other columns in that row where there were no names above or below. Thus, the voter who undertook to operate levers contrary to instructions and to established custom or who chose not to pull any lever at all is no more entitled to have his “vote” counted than the voter who failed to appear at the polls.
To give so much weight to so conjectural an expression of a minute minority (less than one out of 25) indeed would be an extreme overemphasis of the possible disfranchisement of 5,624 voters to the utter disregard of the more than 120,000 voters (both Republican and Democratic) in these districts who voted properly and in accord with the instructions on use of voting machines. This becomes even more apparent in light of the more than 1,000,000 votes properly cast without irregularity in the Democratic primary throughout the state.
Furthermore, it cannot truthfully be said that these 5,624 voters would be disfranchised since many may have voted correctly for other candidates, on the loan question, and in the special election.7
Voting machines have been used in this Commonwealth for more than thirty years in the conduct of thousands of local, state and national election contests. No one has ever successfully contended that levers in one row may be utilized to vote for candidates whose names appear in another row.8
*346Under the circumstances here presented, our Court may not order row C “votes” to be counted for candidates whose names appear in row B.
Neither Gray v. Sanders, 372 U.S. 368, 83 S. Ct. 801 (1963), urged at argument, nor any earlier or more recent decision, compels, supports or even suggests a result contrary to that reached here. The Supreme Court did not say nor imply and surely did not intend that invalidly cast votes be counted under the doctrine of “one man, one vote,” or any other theory just as that Court would not demand that each voter who appeared at the polls cast a vote for each office.9
We must not be misguided or misled by what is asserted to be possible disfranchisement of those relatively few voters who operated levers in a manner other than in accordance with the clear instructions and the election laws. Our concern must lie primarily with the preservation of the sanctity of the election process by giving effect to those votes properly cast. To do *347otherwise is to ignore the prescribed rules of the contest. Obviously, such action destroys not only the recognized machinery for expressing and recording the preference of voters but also imperils the whole orderly electoral process.
Another look at the diagram (of the face of the voting machine ballot) reinforces the correctness of the determinations of the Board and the court below. That additional look also further re-emphasizes the propriety and wisdom of affirmance here.
The initial reason given for the failure to block all levers in rows O and D except for those to be used in the special election was that the Board did not wish to spend the $3,000 such action would require. At oral argument, the Board asserted that it had *342attempted to obtain covers for those levers approximately ten days prior to the election but that the manufacturer could not supply them in time.
In those divisions where separate machines were used for the Republican and Democratic primaries, respectively, votes were cast beneath the names of the Democratic candidates for the Senate in row O on the Republican voting machines. The proportion of such votes in row 0 to the total Republican vote on these machines was approximately the same as the proportion of row O votes to *343total votes on machines used for both parties. It should also be noted that some “votes” even appeared in columns 2, 3 and 4 of row D.
In the 15 wards in question (including two using the Shoup machines), 80,412 Democrats signed voters’ certificates, whereas a total of only 73,244 levers were pulled down in both rows B and O combined in columns 2, 3 and 4. (Figures supplied by Democratic City Committee.)
An assumption that no crossovers took place is consistent with the “unwise” pre-election position of the Board of Elections that it would be unnecessary to lock rows C and D. However, fol*344lowing the election, the Board was satisfied that voters other than Democrats voted in row O and so found as a fact.
Even more confusion and uncertainty would result in a primary election in which the number of candidates for any office requires more than one row for listing their names.
Thus, the suggested interpretation of “upon or adjacent” would render useless thousands of voting machines in the Commonwealth.
More than 120,000 voters appeared at the polls in the wards here involved.
The total of votes cast in columns 2, 3 and 4 of row O was much greater than the totals in columns further down the row.
In Application of Lester, 127 N.Y.S. 2d 272 (S. Ct. 1953), as in the instant case (except for the special election), the Republican candidates were on row A, Democratic candidates on row B, and no candidates on row O. There, as here, the law required that each lever on row O should be locked. In both cases, this was not done and constituted an absolute failure to perform *346a duty specifically commanded by statute. There, as here, when the machines were opened, it was found that some levers had been pulled in row C. It was maintained that since the votes cast in row O were in closest proximity to row B, it was to be inferred that those who pulled levers in row O did so with the intention of voting for the candidates in row B. The New York court rejected this contention, just as did the Philadelphia County Board of Elections, the court below, and now this Court. The New York court held that the inference there urged “. . . is not the only inference to be drawn. No one knows what voters cast their votes on Bow C. If it were done by error, rather than by design, then not even the voters who did so, could now so state, since they are probably without knowledge that they did so. There is no way of telling whether those who cast their votes on Bow C did so designedly, because they had no intention of voting for either of the two major parties’ nominees.” Id. at 276.
In this proceeding no claim was made by a voter or litigant that any voter was denied the right to vote or refused the opportunity to cast his ballot. So, too, no federal question — constitutional or otherwise — was properly or precisely raised, submitted or argued in this appeal. The record fails to reveal the presence of any such question.