### Mr STEVENS:

My hon. friend is in

error there. We do not deny that you can have groups under the present system, but what we do assert is that proportional representation not only tends to it but almost makes it compulsory; it multiplies groups. I do not think there is any denial of that; I do not think it is disputed even by those who support the principle. In fact they advance that as a reason in favour of the adoption of proportional representation.

My hon. friend a moment ago said that this system was, to use his own words, mathematically perfect. I would like to have some one demonstrate its mathematical accur-

acy. I am not in a position to-night to go into any detail in regard to that; I shall simply give one or two illustrations. Take _ a group or multiple riding where, we will say, five members are to be elected-a city like Toronto or Montreal, or any other large urban centre. The experience is-and I am talking now of experience-that you will have perhaps twenty or thirty candidates-at least twenty or thirty where there are five to be elected. I have seen cases in Vancouver where there were thirty nominations and only five or six were to be elected, the ballot being of tremendous length, a very complex document, Now, they take the number of individuals to be elected and divide that into the total number of votes to be polled. We will assume that there are a thousand votes to be polled; they divide that by five, the number to be elected, and add one, making 201 necessary for election. When the ballots are counted, one man who is very popular receives on the first count, 300 number one votes, and they credit him with 201. Then they arbitrarily take 99 votes and divide that into the number of seconds that are then available for distribution, and they allocate the result to each individual on this long list. My point is this: what is the mathematical reasoning of so applying these 99 votes, simply because this man happened by chance to have a larger number of first choices than the others'.

Now, the next step is this: the man at the bottom of the list, number thirty-one or whatever he may be, drops out, and his seconds are taken in a somewhat similar manner. But why should his seconds be taken any more than the seconds of number fifteen, number sixteen or number eighteen? Where does the fairness come in? I do not say it is unfair, but there cannot be any claim for accuracy or fairness in the thing; it is simply an arbitrary action. And here is the way it frequently works out; I have seen cases of this kind. Two individuals had been running along together and after fifteen or sixteen allocations of this arbitrary character it became necessary to allocate the votes of one of these individuals, each of whom represented a very small but particularly virile group in the community. It happened that they both ran first and second right through the whole count, and because it chanced to come to the turn to take one of these, the one of the two who was first was immediately placed two or three ahead of others who had held a lead over them up to that point. My point is that there is just as much of an element of chance in the allocation of these -votes as there is under the present system.

Proportional Representation

**Topic:**PROPORTIONAL REPRESENTATION IN MULTIMEMBER CONSTITUENCIES