of the whole, is a frequent means of successful tyranny. It was the way the first French convention frequently managed things, and Danton knew it well. He acknowledged it.

As to the second and subsequent conditions which have been enumerated, the following observations may prove of interest. Numerous and extensive inquiries, referring to the United States as well as to Europe, and some of which I propose to give to the reader, have proved to me certain instructive facts relating to the statistics of popular elections. I do not treat in this paper of the voting in assemblies of trustees, of representatives, or boards.

I must also remark that I shall always use the term election for direct elections, in which the voter votes directly upon the question at issue, and not for a person who will have the ultimate right of the direct vote; either for a person or on a measure. The election of our presidents was intended to be a double election, and in form it continues to be such; for we elect electors. But it is well known that the election has long since become virtually a direct one, so far as the individual votes express the desire of the voters, because the persons voted for as electors declare beforehand for whom they shall vote in case they are made electors, and after being elected electors they do not become members of a deliberative body in which the question of the presidential election is discussed."

Where the double election is introduced as an active principle, it deprives elections of much, and often of all interest, and is frequently resorted to for this very purpose, by governments which do not feel sufficiently strong to refuse the claims of the people to a share in the government, yet desire to defeat the reality of such a share.

3 This knowledge of the vote which an elector will give does, of course, not affect the result. Each elector represents a majority and a minority, but his vote can only be cast for one candidate. Nevertheless, that which is called the popular vote indicates a proportion between the presidential candidates very different from that which appears from the official votes of the electors. For instance, the popular vote at the last presidential election stood:

-

The following, then, are the positions which experience seems fully to bear out :—

The more exclusive the privilege of voting is, the smaller is the number of qualified voters who abstain from voting; and the largest number of abstinents occurs where universal suffrage is freely left to itself, and not interfered with by the executive.

The smaller the number of qualified voters, the smaller is also the number of abstinents.

So soon as the number of qualified voters exceeds five or six hundred, the number of abstinents will be at least twenty-five per

centum.

The larger the number of qualified voters, voting upon the same question or persons, and under one and the same electoral system, the larger is also the number of abstinents.

The larger the area over which one and the same election or voting extends, the larger is the proportion of abstainers.

When there are three fairly supported candidates, the total number of votes polled is larger than when there are but two candidates, all other things being equal.

The whole number of polled votes, compared to the number of qualified voters, does not necessarily indicate the interest a community may take in a measure or person. Whenever people feel perfectly sure of the issue, there are many who abstain because their votes will not defeat the opponent; and many others abstain, because their candidate will be elected at any rate.

If the number of qualified voters (voting exactly upon the same question or person) exceeds several thousands, one-half of it is generally a fair number for the actual voters; two-thirds show an animated state of things, and three-fourths are evidence of great excitement. It will be observed that the words: Voting exactly upon the same question or person—are a necessary qualification of these positions. Although an election all over England may turn upon free-trade or protection, yet, if it be a parliamentary election, so that these questions appear only represented in the respective candidates, it is clear that this would not be an election extending over the area of England, in the sense in which the term is taken here, or in which we take it when we speak of our presidential election.

Voting upon men generally draws out more votes than voting upon measures themselves.

Popular votes upon measures to be expressed by yes or no are

wholly fallacious, unless this vote be the last act of a long and organic process; for instance, if a new constitution has been prepared by a variety of successive acts, and is ultimately laid before the people with the question, Will you, or will you not have it?

Popular votes in a country with an ample bureaucracy of a centralized government, on questions concerning measures or persons in which the government takes a deep interest, and by elections the primary arrangements of which are under the direction of the government, that is, under the executive, must always be received with great suspicion. It is a fact well worthy of remembrance, that the French people have never voted no, when a question similar to that which was settled, as it is called, by the election of December, 1851, was placed before them. In the year 1793, in the years III. VIII. and XIII. similar appeals were made, and the answer was always yes, by majorities even greater than that on which Louis Napoleon Bonaparte rests his absolutism. When a senatus consultum raised Napoleon the First to the imperial dignity, and the people were appealed to, there were in the city of Paris 70 noes and 120,947 ayes, and in all France, 2,500 noes against 3,572,329 ayes. A vote of yes or no becomes especially unmeaning when the executive seizes the power by a military conspiracy, and then pretends to ask the people whether they approve of the act or not.

From the best authorities on the Athenian government, for instance Bockh's Political Economy of Athens, and Tittman's Political Constitutions of Greece, under the head of Ostracism, we see that the common vote, polled by the Athenians, was about 5,000 (Thucydides, viii. 72) out of from 20,000 to 25,000 qualified voters. Six thousand votes were considered the largest amount. They were required, therefore, for extraordinary cases, such as ostracism, or for anything that was against established law, or related to individuals only. Six thousand Athenian votes thus practically corresponded to our two-thirds of votes requisite for some peculiar cases, purposely removed beyond the pale of a simple majority, that is, at least one more than one-half of the voters. Here, then, we have one-fourth of qualified voters, usually voting, although the voting took place in one and the same city by voters, the great majority of whom lived in the city.

Some writers have doubted whether six thousand votes, upon the whole, were necessary for ostracism and other peculiar cases, or six thousand votes in favour of the measure. I have no doubt that the first was the case. Plutarch distinctly says that one of

the persons proposed was always exostracised, provided six thousand votes had been cast. (Aristides, i. 7.) The same passage seems to prove that, if six thousand votes, altogether, had been cast, he who had the plurality of votes was banished; for, there were frequently several persons proposed for ostracism, or citizens knew that they were prominent, and therefore liable to fall within the ostracophory. Ostracism was a purely political institution, resorted to by democratic absolutism to clip prominences, and keep the hedge on a level. It was no punishment, and until Hyperbolus, a low fellow, was exostracised, it added to the reputation of a citizen.

That there were many abstainers from voting in Athens, we know from the fact, that on the one hand the lexiarchi sent their toxotes before them to mark with red-powdered cords the white garments of those who tarried, so that the thirty judges, presided over by the lexiarchi, might properly fine them. In this, then, the Athenians resembled the early inhabitants of New England, who punished abstaining from voting, or neglecting to send a written note.1

On the other hand, we know that every Athenian of lawful age, (viz. twenty or eighteen,) received three oboli for attending a popular assembly. This reward was called ecclesiasticon.

Why there should have been at Athens so many more abstainers than generally in modern times, may be explained, probably, on the ground that many citizens were habitually absent as soldiers, and that Athens was a direct, untempered democracy. Where the democratic absolutism visibly appears every day in the market, people get tired of it. Besides, the reason which frequently induces so many of our best people from voting, the unwillingness to leave business, must have operated very strongly in Athens, when voting was so frequent and common. Let us imagine Boston or New York as an unmitigated democratic city-state, calling every other day for the meeting of the citizens; does any one believe that the most constant voters would come from the workshops and the ship-wharves, rather than from the tippling shops and filthy lanes of vice?

I have stated already that I have directed my inquiries to election statistics for many years, and over a very large and variegated

* See the Laws of New Plymouth, published by Authority, Boston, 1836 pp. 41 and 128.

space. only.

The reader will admit that I can give a few instances

In the year 1834, there were in France no more than 171,015 electors; yet 129,211 only were polled at the different electoral colleges, that is, only 75 out of 100 qualified voters availed themselves of their privilege. So there were in 1837 in the same country 198,836 qualified voters, and 151,720 votes were polled, which makes 76 of 100.

It will be remembered how small a number of citizens compared to the whole population were entitled to vote. The number of qualified voters at each electoral college was very restricted, and the voters formed a privileged class, compared to the other citizens.

The January number of the Edinburgh Review of 1852, contains a list of sixty-four English election districts, with the numbers of registered or qualified voters, and of the actually polled votes in each, at the last general election. The districts, whose qualified voters amount to less than one thousand, have been separated by me from those which possess more than one thousand. The average number of voters of the first class were 500, and 25 per centum on an average abstained from voting. The average number of qualified voters of the other class was between 2 and 3,000, and of them 42 per centum abstained. So that, if there be about 500 voters, only 75 in a hundred go to the poll; if there be about 2,500, only 58 in a hundred do so.

This is the more striking, if it be considered that one thousand entitled voters is, after all, a very small number compared to those to which we are accustomed, and that far the greater part of the elections given in the mentioned table are town elections, or elections with the most easily accessible polls.

After the chief part of this paper had been written, a very striking fact came to corroborate the results at which I had arrived. The Edinburgh Review for October, 1852, contains an article on Representative Reform, in which there is "A Table showing the Number of Counties and Boroughs in England, Wales and Scotland, in which Contested Elections have taken place in the year 1852." Where an election afterwards contested takes place, it will be allowed that generally there must be great excitement. All voters are brought up over whom the candidates or their agents have any influence. Yet it appears from this table, "that the registered voters in all the contested places reached 507,192, while