When the votes were at the respondent Vardon's instance counted again before the declaration of poll (Act 1905, sec. 161 (a)) variations were found between the original count and the second count in 66 counting centres out of 95. The ballot-papers rejected at the first count as informal were in some cases reviewed, but in the majority of cases that was not done.

The position of a petitioner applying to the Court of Disputed Returns may be thus described. It is on him to prove the allegations of the petition so far as they are not admitted. As to all things in connection with the ballot, except matters of open conduct, it is manifestly difficult, if not impossible, for him to prove a case for a recount, except by a judicial examination of the ballot-papers. He is, in such circumstances, almost, if not entirely confined to this means of proving that enough valid votes to give him a seat or to entitle him to have been declared elected, have been cast in his favour. The order for a recount is thus the means adopted by the Court to open the sources of proof to him, by enabling him to adduce the only, or almost the only, attainable evidence. Ordinarily, therefore, the Court will not be astute to resist a recount, especially as that course cannot prejudice a respondent where the election has been efficiently and accurately conducted, except so far as he may be in a sense prejudiced by the doing of justice. The fact that in the present case the votes were counted a second time under sec. 161 (a) before the declaration of the poll does not, in my opinion, stand in the petitioner's way, even supposing what I may call the mechanical conduct of that process to have been correct. The recount by this Court is a totally different matter. It is a recourse to judicial methods for the purpose largely of ascertaining whether votes have been allowed or rejected according to the law of elections; that is to say, for the determination of questions of law as applicable to the polling, by what Parliament deems to be the best constituted authority. The effort to remove mistakes, mainly arithmetical, solely by a computation conducted by the officers who made the first calculation, can by no means be considered a bar to the interposition of the Court for the determination of disputed questions of law arising out of decisions of these officers, complained of as grievances by candidates who may not have been really defeated.

I should not have hesitated to grant a recount in the present case on the evidence and arguments adduced but for a remarkable occurrence deposed to by one of the electoral officers. Mr. Croft, the Returning Officer for the Division of Angas, was unable to produce the ballot-papers for his Division, though the returns of the voting are all available. The Court has to consider the effect of this unfortunate loss on the application for a recount. All the ballot-papers of the other six Divisions of this State are forthcoming and available.

After careful consideration - the case being, so far as I know, wholly without precedent - I have come to the conclusion that the recount ought to be granted, notwithstanding the destruction of the Angas papers. The petitioner comes to the Court to upset a count which was primâ facie correct. It must stand, so far as he does not impeach it with success. The means of doing this have disappeared so far as the Angas votes, nearly one-eighth of those recorded, are concerned. It does not follow that he is to be deprived of the remaining means of proof because of the misfortune which has occurred. His chance of proving his case may or may not be lessened. The Court cannot possibly form an opinion on that question. But it is clear that the Angas votes must be accepted as they stand on the return, and to my mind equally clear that as to the remainder of the votes the recount ought to be had, and the matter is not affected - this being a Senate election - by the fact that the votes which cannot be challenged constitute the whole of those polled for one of the seven Divisions which each return one member to the House of Representatives. There may be an exception to the acceptance of all the Angas votes. It is alleged for the petitioner that he can prove that some of the papers were improperly rejected. It may possibly be that he will be entitled to give specific secondary evidence of that fact in such form as may secure the reception of those votes. On that I will not pronounce any opinion at this stage, as the proper time for its discussion will be when the case comes before me again after the recount.

It remains to appoint an officer who will conduct the recount and report to the Court; to secure him such assistance as may be necessary; to allow the parties to appoint enumerators, if desired, who may attend as well as the legal representatives; and to fix a date for the continuance of this hearing. All questions of costs are reserved meantime. Mutatis mutandis, the order will be in the form adopted in Melbourne recently in the case of Kennedy v. Palmer ; the Deputy-Marshal will conduct the recount and will report to me as to the assistance which should be granted him. The recount is a grant of the prayer paragraph (a), and therefore includes the whole of the ballot-papers, including those hitherto held informal.

I hold those ordinary ballot-papers to be good votes as to which I find that, as a matter of fact, they were so folded that the initials were visible without disclosing the votes. I hold the absentee voting papers good votes. The signature of the presiding officer on the counterfoil is before him when handed in by the voter, and also before the Divisional Returning Officer when he tears off the counterfoil; and when the gummed paper is torn open to count the votes, the presiding officer's initials are there; its authenticity is therefore complete.

These ballot-papers must be rejected; initialling is a condition precedent to the validity of every ballot-paper. The suggestion that the initials may have been torn off with the margin is too speculative, and is quite unsupported by evidence.

I do not withdraw the opinion expressed by me in Chanter v. Blackwood [1]. Without relying too much on the "Directions" to electors, I must hold these votes good, as the intention of the voter is clear. I prefer in this class of questions to rely on the opinions expressed in the Cirencester Case [2]: the question is, has the voter "clearly indicated the person for whom he wished or intended to vote? We ought to interpret the Ballot Act liberally, and, subject to other objections, to give effect to any mark on the face of the paper which in our opinion clearly indicates the intention of the voter ... whether made with pen and ink, pencil, or even an indentation made on the paper, and whether on the right or the left hand of the candidate's name or elsewhere within his compartment on the voting paper."

One vote at Angas was the subject of argument on an important point. In the turn the case has taken it is not necessary to decide the question raised on this vote, whether a recount if directed under sec. 161 (a) authorizes a revision of the decisions of the officers on the scrutiny as to the reception or rejection of votes. In view of the terms of sec. 161 (a), sec. 160 (c) (d), sec. 119 (e), (where the word "count" was clearly applied to an operation involving decision on a class of ballot-papers up to that time unopened), and Regulation 28, the question is a difficult one, and I should not, as at present advised, be prepared to hold that the decisions of the officers at a first count or "scrutiny" cannot be re-opened at a "recount." The scrutiny is spoken of in the Act as a count, or a counting of the votes, and the recount may be confined to the ballot-papers contained in any "parcel," while the votes rejected as informal have, like the votes allowed, to be put up in parcels. But I think the question should await decision on a fuller argument than was devoted to it here.

As to the question in connection with parcel A., the 21 ballot-papers initialled on the face instead of the back, and as to which it does not appear that they were so folded that the initials could be seen by the Presiding Officer. This is a different case from that dealt with in Chanter v. Blackwood [3] of the 91 persons being allowed to vote who had no lawful title to vote at all. They were strangers to the election entirely, and had their ballot-papers been traceable, their only effect on the election would have been that each such false vote would have been deducted from the score of the candidate for whom it purported to have been give. As it was impossible to trace them, and as they of themselves outnumbered the petitioner's majority on the judicial recount, it could not be said that their intrusion had not affected the result. Clearly it might well have done so.

In the present instance the 21 votes in question were unquestionably rightly rejected at the poll, and I have confirmed the rejection. That they were wrongly initialled was the fault of the presiding officer. Had they been allowed, would Mr. Vardon's majority have been converted into a minority? On a rectification of the votes allowed, Mr. Vardon has a majority of 2. Now there can be no breach of the secrecy of the ballot in apportioning these 21 votes. This is not like the application refused by Griffith C.J. in the case last cited, where it was sought to refer to a private scrutiny the applications for, and the counterfoils to, the postal ballot-papers there held to be invalid, so as to find out for whom the holders of these papers voted. To yield to that application would have disclosed for whom certain named individuals voted, but it was refused as an attempt to infringe the secrecy of the ballot. No such danger exists here. We already know how each one of these votes was cast, but we know nothing at all, and cannot know anything, of the identity of the voters. I therefore think we are entitled to look again at these papers to see whether they would have affected present results had the presiding officer done as the Act directed him and initialled the papers on their backs. On inspection I find that only 2 of them were for Vardon and none for Crosby. So far, then, from affecting the result they would have increased Vardon's majority to four had they been allowed.

Now comes the very serious question whether we can look at all at the totally uninitialled papers, 179 of which were the papers of what are called absentee voters, and 6 were postal ballot-papers, and whether we are to consider how their faces show the result would have been affected, or are to take the whole 185 en bloc and say of them that they necessarily avoid the election, as they may have affected its result. The question whether sec. 200 applies, that is, whether the error of the officers is proved to have affected the result, is also raised.

I am satisfied that I must look at this as another set of instances in which votes have been thrown away through the fault of officers in not doing as the Act ordains. Here there were no initials at all, in face of the express and unmistakeable requirement in the Statute. Mr. Piper's argument as to authenticity is ingenious, but I cannot help remembering that there is no suggestion that even one person who was not a qualified elector was allowed to cast any of the votes under review. As they were all honest attempts to vote, it is clear I can look at them for the purpose of seeing whether they would have affected the result had they been valid.

It is manifest that, but for the omission of initials, Vardon's majority of 2 over Crosby would have been converted into a minority of 4. Therefore, unlike the 21 votes in parcel A, they certainly must have affected the result, and so I cannot apply sec. 200. It is deplorable that all this litigation and expense should have been caused to perfectly innocent parties by the failure of officials to attend to duties clearly laid down in Statutes, Regulations, and printed instructions. Had I a certain power the Court possesses in England, I should consider whether it was not my duty to order an official contribution, at least to the costs.

As it is, having regard to sec. 205, sub-sec. (III.), I must declare the election of Vardon to have been absolutely void, and under all the circumstances, as the proceedings are due to the default of the officers, I do not order costs. Each party will bear his own. I should add that the inquiry has shown that the seats of Senator Sir Josiah Symon and Senator Russell could not in any circumstances have been successfully challenged.

The election of the respondent Joseph Vardon declared absolutely void.

Solicitors, for the respondent (Joseph Vardon), Bakewell, Stow & Piper .

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

[2] 4 O'M. & H., 194, at pp. 196-7; Day's Elec. Cas., 155.

[3] [1904] HCA 48; 1 C.L.R., 121.