But the elector may render his meaning ambiguous and uncertain by putting other marks on its face. Where I have found such marks I have not been moved to reject the paper if the preference has still been left clear. "Of course if it is upon the face of the ballot-paper left in doubt whether the man intended to vote for one candidate or the other, the weight of the objection that the vote is uncertain is obvious, for the simple reason, that one candidate has just as much right to claim the vote as the other, and so it ought not to be counted for either." Cirencester Case[2]. But, to quote the very apt words of Hawkins J. in that case[3], I have done my best to discover whether "there existed positive indications on the part of the voter of an intention to vote without a thought of leaving behind a trace to enable him to be identified," and wherever I have found that positive indication on the one hand, and the absence of real clue to identification on the other, I have allowed the vote. The English Act (sec. 2) avoids any ballot-paper on which "anything ... is written or marked by which the voter can be identified." Comparing with these the words of our own Act, already quoted from sec. 158 (d), as well as the rest of the Act in each case, I am convinced that this provision in the Australian Statute is intended to bear the same liberal construction as had before its passage been judicially placed on the corresponding and similar language of the English Act. I am therefore of opinion that it is not enough, in the absence of all extraneous evidence, to urge that there are marks which might by some possibility enable some one to identify the voter. The case is palpable where voters have placed their names or initials on the ballot-paper, and tolerably plain where there occur names or initials evidently not written by the presiding officer and therefore in all probability written by voters. The question is one of opinion formed on very varying facts, and the Court must be guided to its conclusion merely by applying its common sense to the consideration of the marks or writings, and treating each case on its merits. In one instance I have felt myself justified in allowing a paper which has in addition to the cross the word "yes," both within the square; and another paper which reads thus: to the left of a candidate's name a cross was properly placed in the printed square, to the right and opposite there was a cross, the word "yes," and a pencilled square. The other candidate's name was struck through, and opposite was written the word "no." But the handwriting of the "yes" and the "no" is like a thousand others, and so far from the marks justifying an opinion that they "will enable" anyone to identify the voter, there is scarcely room for even surmise on the point, in the absence of a shred of evidence pointing to any improper practice or plan. It is not my intention to multiply instances of papers accepted or rejected. I have examined all the papers with great care and have dealt with each one. With the Judges in the Cirencester Case[4], I do not think it would serve any good purpose to discuss in detail the various votes and objections, nor to announce my judgment upon each particular case which I have had to consider. On this part of the case I find as a fact that the petitioner's lead of 29 has been increased, by the admission of parcels 6 to 10, to 30 votes, and by further examination of the votes claimed for either candidate, or challenged for reasons other than the mistakes of officers, the number is increased to 35.

I now come to consider the class of papers in which the votes, if thrown away, have been lost by official mistake. These papers are contained in parcels 1 to 5, 12 to 15 inclusive, 19, 20, 21 and 38. In the 8 parcels last named there were 31 papers, and of these I allowed 6 for the petitioner and 4 for the respondent. Of the 6 allowed to the petitioner 3 were admitted by respondent's counsel. Of the 4 allowed to the respondent 2 were admitted on behalf of the petitioner. Thus the petitioner's majority is increased by 2, and of the votes actually admitted and counted as good, the petitioner has a majority of 37.

I may mention here a case of a vote which I have disallowed after much contest at the bar. It was on a postal ballot-paper. By applying the detached counterfoil to the ballot-paper it is evident that the whole of a number is disclosed. In tearing off the counterfoil at the time of putting the paper into the box, the officer has lost the line of the perforation, and has left part of the counterfoil adhering to the ballot-paper. That adherent part of the counterfoil contains the greater part of the official figures originally placed on the counterfoil. Thus by applying the rest of the counterfoil to the ballot-paper, which retains the part detached, anyone can discover the number with ease. Sir John Quick argues that these figures are not such a "mark or writing" as are aimed at by sec. 158 (d), because they are not on the ballot-paper. They are certainly on that which the officer has made the ballot-paper for the purpose of the election: they are clearly within the mischief of the section, and I think the officer's mistake has caused the vote to be thrown away.

One of the absentee voters' papers had not been quite completed by the filling in of the name of the Division - Echuca - on its back. In other respects it was in due form as marked by the voter. In its original state, it had attached to it a counterfoil, on which the voter had declared that his name appeared on the Electoral Roll for the Division of Echuca. That declaration had been witnessed, while still attached to the ballot-paper, by the Returning Officer, and after the elector had marked his vote on the paper it had been handed to the Returning Officer with the counterfoil still attached, so that he knew the voter belonged to the Echuca Division. In fact he must have known that to be able to send it to the Divisional Returning Officer for that Division, as he did. That officer, when he received it, had under the Regulations of 14th September 1906, reg. 25 (1) to examine it and satisfy himself that the person signing the counterfoil was on the roll for the Division he declared himself to belong to. Until he did that the Deputy Returning Officer had no right to tear off the counterfoil, as he has done, and I have not the slightest doubt he did his duty and accepted this as a perfectly well-authenticated ballot-paper. I have therefore no hesitation in accepting it. I have devoted some attention to two or three cases because they are in a measure typical, and are likely to be the subject of decisions by Returning Officers at future elections.

I come now to the papers challenged by reason of the alleged mistakes of the officers and contained in the parcels I have mentioned. There are first parcels 1 to 5, amounting in number to 179, exclusive of parcels 6 to 10, already dealt with. It appears from the evidence that the respective Returning Officers who issued them, and who were called, numbered them on the backs from the rolls. The votes were good but for this numbering. Under some electoral systems - those of Victoria and Queensland were instanced - the officers are directed by law to number the ballot-papers in the manner here adopted. But under the Australian system, not only is there no such direction, but there is an express provision vitiating a ballot-paper which when used "has upon it any mark or writing not authorized by this Act to be put upon it which in the opinion of the Returning Officer will enable any person to identify the voter" (sec. 158) (d), and in the Court of Disputed Returns the opinion of the Court must be substituted for that of the Returning Officer. It is obvious that the comparison with the rolls of the numbers on the back of the ballot-papers affords the very best and most convenient means of finding out the name of the voter and how he voted, and where the system does not make such knowledge the property of the officers for the purposes of the election, papers containing such writings must of necessity be vitiated. It was very difficult for Sir John Quick to contest this position. To contend that the enactment is merely directory is to say that the most necessary precaution for the completeness of the secrecy, if absolute completeness is, as here, insisted on, is not an obligation upon the officers charged with the maintenance of that secrecy. That can not be. Nor is it more convincing to be told that the objecting candidate must challenge such papers as these through his scrutineers at the booths, and if he does not take that course, is afterwards concluded of his right. True, it is the Returning Officer's opinion that in the first instance has to decide the matter. But the decision is one that is within sec. 157, and is therefore subject to reversal by this Court. Sir John Quick also took a distinction between ballot-papers and votes of which I need only say that in this connection there is nothing to sustain it, if he means that a bad paper can carry a good vote. These papers then are every one of them informal and cannot be counted. But had the officers attended to their duty and refrained from marking numbers on them, the votes upon them would have been counted, and of these the petitioner would have received 60 while the respondent would have received 119. That would have turned the petitioner's majority into a minority, as the respondent would on parcels 1 to 5, if not numbered, have led by 59, as against the petitioner's former lead of 37. That is, the respondent would have had then a majority of 22 on the aggregate.

But there were other papers marred by the errors of officers. Of these two are not sufficiently marked for authentication (parcels 14 and 19). They would have been for the petitioner, and would have brought the respondent's majority down to 20, and in parcel 38, consisting wholly of "absentee voters'" papers, 7 papers were bad because the officers issued them without the names of the candidates upon them, and 8 bad because they bore no officer's initials either on back or face. Of these 15 the petitioner would have had 7 and the respondent 6 (the remaining two being votes thrown away even had the papers been officially correct), so that in the final result the respondent would have had a majority of 19. There remain 4 absentee voters' papers. As these appeared by the counterfoils to be incurably bad for want of attestation, they were properly rejected unopened. I do not intend to examine them. Being bad for official error, they could only be considered on the question whether the aggregate of such errors had affected the result. That question has been determined apart from them, and they could not alter that determination even had they all appeared to have been cast for the petitioner. Moreover, I think these papers are within the decision of Griffith C.J. in Chanter v. Blackwood (No. 2)[5]. In these circumstances I cannot declare the petitioner entitled to the seat, as it is only the mistakes of the officers, proved to have affected the result (sec. 200), which have deprived the respondent of it. Nor can I declare the respondent entitled, inasmuch as the petitioner has polled a majority of the valid votes cast. It is clearly a case in which the will of the electors has been frustrated, and in which they are entitled to express it effectively upon the issue of another writ. I therefore declare the election absolutely void under sec. 205 (III.) of the Acts, and as there commanded, a new election must be held.

Solicitors, for petitioner, Strongman & Crouch, Melbourne.

Solicitors, for respondent, Quick, Hyett & Rymer, Bendigo.

[1] [1904] HCA 2; 1 C.L.R., 39.

[3] Day's Elec. Cas., 155, at p. 159.

[5] [1904] HCA 48; 1 C.L.R., 121, at p. 128.