* I wish to thank Professor Sheldon Rothblatt of the University of Calfornia at Berkeley and Professor Daniel Calhoun of the University of California at Davis for their advice in the preparation of this article.

¹ Speck, W. A., Gray, W. A. and Hopkinson, R., ‘Computer analysis of poll books: a further report’, Bulletin of the Institute of Historical Research, XLVIII (1975)Google Scholar; Speck, W. A. and Gray, W. A., ‘Computer analysis of poll books: an initial report’, Bulletin of the Institute of Historical Research, XLIII (1970)Google Scholar; Speck, W. A., Tory and Whig, the struggle in the constituencies, 1701–1715 (London, 1970), chap. 6 and p. 114Google Scholar; Plumb, J. H., The growth of political stability in England 1675–1725 (London, 1967), pp. 55–7CrossRefGoogle Scholar; Plumb, J. H., ‘The growth of the electorate in England from 1660 to 1715’, Past and Present, no. 45 (11 1969), 115–16CrossRefGoogle Scholar; Holmes, G. S., The electorate and the national will in the first age of party (Lancaster, 1976), pp. 2, 7, 32Google Scholar.

² Speck, Tory and Whig, p. 77

³ This article is based on copies of these poll books: Kent A.O., Q/RPei for the 1713 election, and Bodleian Library, Rawlinson MS B338 for the 1715 election.

⁴ Sedgwick, R. (ed.), The history of parliament: the House of Commons 1715–1754 (2 vols. New York, 1970), 1, 609Google Scholar.

⁵ The poll book entries were checked and corrected for variant spellings. Both poll books recorded each voter's name, his parish of residence, and the freehold on which he based his right to vote, but this information was not always identical. For instance, the 1713 poll book might identify one voter as resident in Bethersden and voting for a freeehold located in Biddenden, while the 1715 poll book might have the same voter resident in Biddenden and voting for a freehold located in Bethersden. The following rules were used to decide whether any names which appeared in both poll books, but for which each poll book recorded different information, represented one voter or two different voters. If voter John Doe was shown as resident in the same parish in both elections, but as basing his right to vote on freeholds located in different parishes at each election, Doe is considered a single voter who voted in both elections. If voter John Brown was shown as claiming a freehold in the same parish in two elections, but as resident in two different parishes, one of which is the same as that in which his freehold is located, Brown is also categorized as a single voter who voted in both elections. If voter John James was recorded as resident in one parish with his freehold in a second in 1713, and in 1715 as resident in the second parish with his freehold in the parish in which he was resident in 1713, James is accounted a single voter, voting in both elections. Detailed analysis of voting patterns requires that each voter be assigned to a single parish. Since the two elections were only seventeen months apart, it is unlikely that there was any significant migration of voters between elections. Most probably, the information in the poll books which might indicate change of parish of residence exaggerates the migration which did take place. Therefore, the parish of residence assigned Brown and James is that noted as their residence by the 1713 poll book. Finally, let us consider the case of one last voter, John Smith. The 1713 and 1715 poll books both record Smith as claiming a freehold in the same parish, but each poll book identifies him as a resident of a different parish, in neither of which is his freehold located. Smith is categorized as two different voters, one of whom voted in the first election but did not appear at the second, the other voting in 1715 but not in 1713. Obviously, any set of rules for the collation of names in poll books which were produced with less than a slavish regard for factual accuracy results in occasional misidentification, but this set of rules allows for the identification of most of those voters who voted at both elections for whom the poll books provide similar but not identical entries, and prevents the false conflation of two different voters.

⁶ The calculation of the results of split and single votes counted each single vote as half a vote and the loss or gain of one of the two votes cast by a split voter as the loss or gain of half a vote.

⁷ Hasted, E., The history and topographical survey of the county of Kent, 2nd edn (12 vols., Canterbury, 1797–1801; reprint 1972), III, 322–3Google Scholar.

⁸ Hasted, ibid., IV, 193, 199, 227.

⁹ It might be argued that the relatively large number of voters in these six parishes indicates an electorate more able to float as less bound by the ties of deference than the electorate of the average Kentish parish. Yet in Canterbury there were 6 floating voters (4 T/W and 2 W/T) in an electorate of 310, and in Maidstone there were 13 floating voters (7 T/W and 6 W/T) in an electorate of 224.

¹⁰ Hasted, History of Kent, X, 9, II.

¹¹ Sedgwick, House of Commons, I, sub Rochester and Queenborough.

¹² Hasted, History of Kent, VIII, 450, 177.

¹³ Sedgwick, House of Commons, I, sub Hythe.

¹⁴ SirNamier, L., The structure of politics at the accession of George III, 2nd edn (London, 1961), p. 121Google Scholar.

¹⁵ Only in Sheerness, a ville dominated by a fort, are these trends as marked as in the boroughs. Sheerness marshalled a total electorate of ten in these two elections. In 1713 only five voters appeared and all supported the Tories. In 1715, again, only five voters went to the polls, but these were five different voters, all of whom supported the Whigs.

¹⁶ Speck, Tory and Whig, pp. 76–7, 94–5.

¹⁷ Speck, ibid. p. 77; Holmes, The electorate, p. 29; Namier, Structure of politics, p. 67.

¹⁸ Holmes, The electorate, pp. 21–2; Speck, , Gray, and Hopkinson, , ‘Computer analysis of poll books: a further report’, Bulletin of the Institute of Historical Research, XLVIII (1975), especially pp. 75, 84, 86Google Scholar. It should be noted that while the 1975 article emphasizes the importance of the casual voter, it also confuses the question of the relative importance of the floating vote. In Speck, and Gray, , ‘Computer analysis of poll books: an initial report’, Bulletin of the Institute of Historical Research, XLIII (1970), 111–12Google Scholar, the term is defined as the voter who completely switches his party allegiance. In the 1975 article the term is also used to denote split voters who move to straight party voting and vice versa.

¹⁹ All of the six parishes and seven boroughs except New Romney voted overwhelmingly for the government's candidates The poll for … the county of Kent (London, 1734).

²⁰ While there has not been an intensive modern study of the distribution of land in eighteenth-century Kent, A. Everitt, in The community of Kent and the great rebellion 1640–1660 (Leicester, 1966), pp. 35–37Google Scholar, concluded that seventeenth-century Kent was dominated by gentry families rather than by members of the aristocracy, and Thompson, F. M. L., English landed society in the nineteenth century (London, 1963), pp. 32, 113–15Google Scholar, reveals that nineteenth-century Kent was both free of aristocratic control and rich in gentry. See also SirNamier, L. and Brooke, J. (eds.), The history of parliament: the House of Commons 1754–90 (3 vols. London, 1964), IGoogle Scholar, sub Kent.

²¹ Kent A.O., Q/JC 36, 37. Those in the honorary portion of the commission, including the sons of peers, are not considered in this article nor included in the numeration of justices. Eighteen of the non-voting justices lived out of the county and an additional eight were dead. Another seven of the justices who did not vote in 1713 were dead by 1715. Neither of the Tory candidates, both justices, voted at either election.

²² Five justices who had not voted in 1713 were also removed by the 1714 commission. Neither the commission of 14 July 1713 nor that of 19 October 1714 were published at the Assizes immediately preceding the elections. However, by November 1714 Kent's clerk of the peace knew the details of the alterations made in the 1713 commission by that of October 1714, and was arranging for those who remained in the commission to take the oath to the new sovereign. Justices were being sworn in January, and probably even earlier than January (Kent A.O., Q/JLI, John Heath to Mr David Fuller at Maidstone, 6 November 1714; British Library, Add. MS 42598, fo. 118, 11 Jan. 1714–15). News of the alterations in the composition of the bench by the commission of July 1713 had reached Kent's justices by 16 August (British Library, Add. MS 42598, fo. 104; Add. MS 42588, fo. 37, Tho. Morris to W. Brockman, 16 Aug. 1713).

²³ Including three who had died since 1713.

²⁴ Beside the seven boroughs examined earlier there were several other places within Kent over which the justices did not exercise jurisdiction. Of these, 17 parishes and 7 boroughs dispatched 944 voters to the polls, who in 1713 awarded the Whigs a margin of 10½ votes and in 1715 increased the margin to 35. Another 471 voters resided out of the county, and it likely that the 72 voters whose place of residence cannot be identified were also not inhabitants of Kent. In 1713 the balance of this electorate favoured the Whigs by 10 votes, and in 1715 the Whigs led by 20½.

²⁵ Four of the nine Tory justices did not live in the administrative county; one was an inhabitant of Deptford, and two more resided in parishes where there were other politically active justices whose influence would predominate in 1715, if judicial office entailed persuasive power. That leaves only two parishes – a population too small to yield a statisically significant result. Of the two justices who switched party, one did not live in the administrative county, and the other lived in Greenwich.

²⁶ In one more parish the influence of two justices was summed. Teston was governed by Sir Philip Boteler, who supported the Tories in 1713 only, and by his brother, who supported the Tories in 1715 only. The electorate of Teston was classified as influenced by a justice who supported the Tories at both elections.

²⁷ As the table compares the voting patterns of the justices with that of the electorate in their parishes, both the figures for the total electorate and those recording the voting behaviour of the electorate exclude the justices. Split voters who cast split votes at both elections or voted in only one election did not affect the balance of votes. Neither did voters who plumped for one candidate of the same party at both elections. Therefore, these voters are not enumerated in the portion of the table displaying change in split and single votes, but they are included in the figures for the total electorate.

²⁸ The levels of significance yielded by the application of the chi square test to the comparison of judicial categories are displayed in Appendix I, Table 6.

²⁹ The results obtained by the application of Cramèr's V to the voting patterns of judicial categories are displayed in Appendix II, Table 7.

³⁰ For Tory optimism and fervor in the 1715 election, see Speck, W. A., ‘The general election of 1715’, English Historical Review, xc (1975), 507–16CrossRefGoogle Scholar.

³¹ For more detailed analysis of change in the composition of the bench in the early eighteenth century, see Landau, N., ‘Gentry and gentlemen: the justices of the peace, 1680–1760’ (unpublished Ph.D. dissertation, University of California at Berkeley, 1974)Google Scholar, chap. 3, and Glassey, L., ‘The commission of the peace, 1675–1720’ (unpublished D.Phil, thesis, Oxford University, 1972)Google Scholar, chaps. 5–8. Similarly, the political activity of Kent's justices and the care lavished by the Lord Chancellor on the appointment of justices reveal the political importance of the mid-century justice. Fifty-eight per cent of Kent's justices voted in 1734 and 64% in 1754. Five volumes of Hardwicke's papers are filled with correspondence on the appointment of justices (British Library, Add. MS 35600–35604). For further analysis of the importance of political affiliation in the appointment of mid-century justices, see N. Landau, ibid. chaps. 4, 6.

³² Speck, Tory and Whig, chap. 8; Speck, , Gray, and Hopkinson, , ‘Computer analysis of poll books: a further report’, Bulletin of the Institute of Historical Research, XLVIII (1975), 64–5, 86–7CrossRefGoogle Scholar; Plumb, The growth of political stability, pp. 53–7; Plumb, , ‘The growth of the electorate from 1660–1715’, Past and Present, no. 45 (11 1969), 92, 115–16Google Scholar; Holmes, The electorate, pp. 14–18, 28–33.

³³ von den Steinen, K., ‘The fabric of an interest: the first duke of Dorset and Kentish and Sussex politics, 1705–1765’ (unpublished Ph.D. dissertation, University of California at Los Angeles, 1969), p. 171Google Scholar.

³⁴ Plumb, The growth of political stability, especially chap. 4.