H.—4l

3

Llixcept where otherwise indicated, the references in what follows are to the election of Councillors.] The card labels in a candidate's pigeonhole at any time would show his total at that stage. For instance, after count 2 in Beanland's pigeonhole there were no voting-papers, but the card label A given below, which shows the votes left to him after the distribution of his surplus (viz., 982, or the quota). At the same stage there were in Climie's pigeonhole two labels, one (similar to A) showing his first choices, 357, and the other (B) showing 680 votes transferred to him from Beanland. After the eighth count, when Climie's surplus (55) was distributed, and the entries at the foot of B had been made, his labels showed him to have 357 and 625, or 982 votes. The voting-papers sliowing the first choices would be attached to the card label, but the 1,103 voting-papers would not be attached to card label B, having been distributed at the eighth count to the next available choices. The card labels for the first count were printed in red ; those for all other counts in black.

At the trial count full sets of directions were issued to computers, supervisors, sorters, and counters : it is much easier to guard against error if the same routine is followed throughout the scrutiny of the votes. Copies of these directions, revised in the light of the experience gained, are attached hereto (Appendix II) ; they may prove of service in future elections, and their reprint here makes it unnecessary to give any more details of the process adopted. Tables 1 and 2 are shown in Appendix I. Table 1 shows the counting of the choices, as indicated on the voting-papers ; Table 2 shows the distribution of the votes in the manner proscribed in the Second Schedule to the Act. Table 2is thus, in effect, the score-sheet for the election, while Table 1 gives the results of the several counts and forms the basis of Table 2. Herein the method adopted in the Tasmanian general election of 1909 was used. Tables 1 and 2 are given in full for the first eleven counts, and a summary of Table 2 alone is given for the remaining counts. There were 174 effective counts—that is, counts in which votes were transferred from Table I to Table 2, but in fjhirty-two of these no transfers of Votes were made to continuing candidates, the small number of votes being either exhausted or " lost by fractions " ; other counts were also made of the papers on which no votes were transferred, to act as checks on previous sortings ; but the process affords so many natural checks against error that these non-effective counts are quite unnecessary. It might be a question for consideration as to whether it would not be expedient to quicken the process of counting by allowing the parcels of votes transferred from excluded candidates to be grouped for that purpose. For instance, the transfer of Butterfield's votes, which took thirty-seven counts, might have been done in three counts ; the continuing candidates would have received three more votes, and three votes less would have been " lost by fractions." The method of grouping transfers to which I refer is to transfer in three several lots, viz. : (a) All. papers with, transfer value 1 ; (b) papers with transfer values less than 1, but not less than |; (c) papers with transfer values less than |. [In the case of (b) and (c) it would be necessary to find a new transfer value by dividing the number of votes to be transferred by the number of papers included in the bundles thus grouped.] Clause 8, paragraph (d), of the Second Schedule prevents this from being done at present; but if it had been allowable the same candidates would have been elected in the same order, and the last count would have appeared thus in respect of the continuing candidates : — Jen-kin .. .. 745 ~j f Jenkin .. .. 720 Peek .. .. 957 | | Peek .. ..944 Wells .. .. 786 y instead of <{ Wells .. ._. 770 Exhausted votes .. 298 | | Exhausted votes « 291 Lost by fractions .. 143 J [_ Lost by fractions _ 204 The total number of counts would have been 51 instead of 174 ; and the total time of counting might have been reduced from thirty-three hours (time actually taken) probably to less than twenty hours. It may be said with truth that the suggestion involves a departure from the strict principles of the Clark system ; but the worst that could happen (provided that the voting was on party lines and that the amount of cross-voting was negligible) would, be the substitution for the last candidate of. another candidate of the same party.

CARD LABEL A. CARD LABEL B. Label of Bundles of all Papers in First Count. Label of Bundle of all Papers in Count named (any except First) on which Candidate named is the. Next Available Choice. Eikst Count. (a) Beanland was tho first choice on 2,563 papers. Count 2. SUBSEQUENT DISPOSAL (lE ANY) OE THE ABOVE VOTES. On the papers of Beanland as at count 1 the next available choice after him is (a) Climie as shown on .. .. .. (b) 1,103 papers. The transfer value is .. (c) .Jgf J The number of votes transferred to (a) is (d) 080 Total number of these votes transferred from (a) to other candidates, exhausted, and lost by omission of fractions, at count 2 ... .. 1,581 Tho number of above votes romaining for (a) .. 082 Subsequent Disposal (ie any) oe the above. Votes. Total number of these votes transferred from (a) to other candidates, exhausted, and lost by omission of fractions, at count 8 .. (e) 55 The number of above votes romaining for (a) .. (/) 625