676 METHODS FOR COUNTING IN ELECTIONS. 



In Column ] 1 we have four lines headed 2nd, 3rd, 4th, 

 6th, of different grades of choice; and the 12 whole votes 

 of the live candidates unsuccessful at the first count 

 are tentatively distributed as second-choice votes. J gets 

 2, but needs only .5; so .25 of each is retained, and 2 

 votes, II J C and H j M, are left, each of the value of .75, 

 Columns III and VI. M gets i and needs .5 ; so the votes 

 H M C is left with the value .5, Column IV. A gets 2 votes, 

 B A, H A, he needs 1.5; so they are each left with an unused 

 or transfer value of .25, Columns V and X. C gets and keeps 

 I vote DC. I gets i vote, 21, D I M A C B, non-effectice as a 

 second-choice vote for I, but which becomes effective as a 

 sixth for B ; and one vote, 23, E I A J M C, is non-eff'ective for 

 I, and also at the sixth count for C. D gets one vote ; vote 

 47, I 1) E B, which becomes available as a fourth-choice vote 

 for B. 



In Columns III, IV, V, third-choices are dealt with; one 

 non-effective goes to M, who is already elected, vote 45, 

 H J M C A B. .7S, and three H J C, H M C, B A, go to C, of 

 the values .75, .5, .2^, or 1.5 in all. C needs .5, so retains one- 

 third of each, leaving transfer values to two places of decimals 

 of .5. Column \'TI ; .34, Column VIIl ; and .16, Column IX. 



Ai the distribution of whole third-grade votes, vote 47. 

 ID E H goes to E as a non-eft"ective third-choice vote, and vote 

 22 goes to L, E B L. 



The votes that go at these distributions to elected members 

 are passed on for distril)ution at tlie next coinit. All others are 

 entered, so that we can see how many votes each continuing 

 member or unsuccessful candidate can get. At the fourth count 

 only the vote 47, I D E B, goes as an eft'ective vote to B. No 

 transfer is made at the fifth count. 



At the sixth count. Column II. J, M, C. D, get non-effective 

 whole votes, and B gets a vote. Columns VI and VIII give B 

 .75 and .34, so he can get 7.09. I gets .16, Column IX; C gets 

 .25, Column X ; and D .5, Column VII. 



Note 3. — \'ote 10. B A C M j 1, gives ./^ to A, Column II, 

 2nd, and .09 to C, Column V, and . 16 to i, Column IX == i, none 

 of which should be counted ; for in Column I it is counted to 

 B. So the real tcjtal effective votes are 76 - .84, or 75.16, and 

 the non-eff'ective are 9 - . 1 6 = 8.84 ; 75. 16 + 8 . 84 = 84. The vote 

 E B L is counted to B, so L has only his original 7 in Column XI. 



As H can get 7.09 votes and E has only 7, he displaces F^ 

 and a new second count must be made with B retained as a con- 

 tinuing member. When E B L goes to B, not E, E, too, has only, 

 like F, seven original votes. vSo the lot nuist be cast to show 

 whether L or F have their votes distributed. It falls on E, whose 

 second choices must be distributed with those of H, E, D, I, 18 in 

 all at the first distribution. Column II, second on the final result 

 sheet. 



