In conducting their visitation surveys, bishops throughout the country were invariably concerned with establishing the number of inhabitants of each parish within their diocese. As in the Exeter returns, this was usually expressed in terms of the number of families residing within a parish. For the period from the mid-seventeenth century until the first censuses of the early nineteenth century visitation returns provide the only extensive demographic data other than parochial registers on which local population estimates can be based. The most complete population listing of the period, the 1723 oath of allegiance rolls do not provide a comprehensive listing of inhabitants comparable to the Protestation Oath of 1641-42.56 For the archdeaconries in the county of Devon the four most complete extant sets of returns allow population changes to be traced from the mid- to late- eighteenth century. To date, no attempt has been made to subject the Devon returns to detailed analysis, although the Cornish replies have been studied.57 However, some comments can be made on the usefulness of visitation returns as a source of population history.58
Leaving aside the obvious problems of the omission of some parishes from each of the separate surveys, the visitation returns do have several shortcomings as sources for reconstructing local population levels. Foremost amongst these is that the bishops of Exeter only ever enquired after the number of families residing in each parish and not the number of individuals. The only figures in the returns that refer to individuals are the number of communicants, and these are universally low and so of no value for determining population sizes. In using the returns to estimate the population of towns in Devon and Cornwall in the mid-eighteenth century, Jonathan Barry adopted a standard multiplier of 4.25 for converting numbers of families into population totals.59 This figure provides a useful mechanism for using the visitation returns to produce relatively quick estimates of population size for a particular parish or group of parishes. However, the average household size is likely to have varied between parishes and the criteria employed by the clergy for assessing the number of families in their parish were not consistent.60 For example, in 1744 the Rector of Aveton Gifford estimated that the parish contained 125 familes, 'Including the smal families, consisting of two or three persons only'. Making his return in the same year, Robert Wright of St Mary Arches noted that the number of inhabitants 'cannot be well computed from the no. of Families' as over fifty consisted of a single person 'Many of whom rent single Rooms and particularly in One House there are Ten such families'.
The estimates of numbers of families provided by the clergy are often rather vague, with little evidence as to how the figures were obtained and how much care had been taken in arriving at them. In 1744 the Rector of Hemyock complained that it was difficult to determine the exact number of families in his parish, 'it comprehending a very large Tract of Hills and waste Ground, on which are scattered up and down a great many Cottages', before offering a conservative estimate of 200 families. His successor in 1779, John Land, offered a different estimate of 140 families 'upon a moderate Computation'. No doubt this problem is more acute for parishes with particularly large populations or those which covered a substantial geographical area. The population of Totnes in 1744 was estimated at 500 families, and by 1779 the figure offered had risen to 700. In the Plymouth area the Reverend Barlow of Stoke Damerel in 1744 thought that there were 1,000 families in his parish. According to his successor thirty five years later the number had risen to 2,500. In 1779 John Gandy of St Andrew, Plymouth thought that there were 800 households containing approximately 1,000 families on account of the number of houses containing more than one family. Such vague figures suggest a certain degree of guesswork on the part of the incumbent.
Only occasionally is a more precise total offered. In 1744 the vicar of Cullompton described how he had been from house to house some eight years previously. At that date he had calculated that his parish contained 760 families and 3,586 'souls'. Unfortunately his successor in 1779 John Veryard Brutton was rather less precise, offering a figure of between 3,000 and 4,000 inhabitants. Incumbents of smaller parishes were often able to offer figures that were apparently more accurate. Thus Nathaniel Dewdney of Clyst St Lawrence was able to state that his parish contained 33 families, a figure that had fallen to no more than 15 by 1779. It is to be assumed that the incumbents of West Ogwell were responding with some confidence when they both returned the number of families in their parish as seven. In general, the visitation returns provide the most accessible guide to local population levels during the eighteenth century. However, comparison with other sources, in particular baptism and burial registers, should be undertaken to produce a more complete picture at the level of the parish.