18th century land use statistics
The investigation of land use and agriculture in Jersey during the 18th and earlier centuries is hampered by a lack of precise information. Contemporary observers employed a style and method very different from those found in present-day accounts, where the statistical summary is at least equal in importance to the commentary. The earliest accounts in which reliable figures are included seem to be those of Quayle, in the well-known report of 1815, and the crop returns for six parishes in 1801. This material has already been made use of, in a description of land use during the 19th century, ie from the end of the Napoleonic Wars onwards, and in an account of certain distributions at the opening of that century, when the first phase of the Wars was coming to an end. It is now suggested that statistics can be obtained for land use in the late 18th century, so that the sum effect of changes prior to the Wars can be judged, and something of wartime changes can be measured.
1795 Richmond map
The source material is that Map of Jersey published in I795. It was made to the order of the Duke of Richmond, by a team of surveyors headed by Gardner, on a scale of six inches to one mile. The team included Yeakell, with whom Gardner had been associated in some notable work of topographical survey in England. The large scale and the wide range of symbols are sufficient to make the map one of land use, for all field boundaries are shown. But the I795 map cannot be used simply as it stands. Careful comparison with the modern Ordnance Survey 1/5,OOO map suggests that, although Gardner's party fixed the course of roads with commendable accuracy, field boundaries were sketched in by eye. Hence it becomes necessary to correct the distributions by transferring them, field by field, to the OS map, after which individual distributions can be taken off and areas determined. The whole process is rather laborious, but succeeds in providing the statistical information desired. Absolute precision cannot be claimed, but it is certain that no serious errors are introduced.
In practice it is found that field boundaries changed little in the interval between the two surveys, so that the operation of transfer from one map to the other presents few difficulties. A more serious problem is that of interpreting the symbols employed on the map of 1795, for no key is given. Some of the symbols are self-explanatory, in part because they closely resemble the symbols now in use by the Ordnance Survey. But the core of the problem is whether, in the areas with no land use symbol, the land was wholly arable in 1795, or whether it included some kind of grassland.
The same difficulty has been encountered in previous studies of maps by Yeakell and Gardner. Judgement has usually been reserved, but research now in progress is expected to provide the necessary information which will permit a full analysis of the 1795 map at a later date. The present paper will be confined to a discussion of three distributions which are unequivocally recorded; it may serve also to illustrate the method of working and the kind of results which may be expected.
The map on the left shows the extent of unenclosed land (hatched) and of orchards (solid black) according to the map of 1795, corrected on the OS 1/5,000 map as a base and with parish boundaries added from the OS 1/31,680 map. The general facts of distribution are clear, but no more so than if they had been taken directly from the original map, without correction. When the areas of open land, enclosed land, and of orchards are separately determined for each parish, as shown in the table, precise description becomes possible.
The proportion of land already enclosed by 1795 - 80.5% of the whole island - is remarkably high. Not all was agriculturally productive, for the figures include roads and buildings, in addition to a limited extent of heathland or rough grazing where enclosures had not been improved, or had been allowed to revert. The unenclosed portions included narrow sandy strips fringing the bays in the south and south-east, and the cliffs and immediate cliff-tops of the north-east and north. In the west less progress had been made in the face of greater difficulties.
The exposed extreme north-western shoulder of the island and a portion of the low-lying sandy ground behind St Ouen's Bay both remained open, reducing the percentage of enclosed land in St Ouen to 66%. A beginning had been made with the improvement of the western crescent of lowland in the extreme north, but efforts to this end had proved more effective in the west of St Peter, where the old bed of the lagoon which is still present in a shrunken state had already been largely reclaimed. The more favourable condition of this patch of ground, by comparison with the adjacent sands, is reflected in the high proportion (87%) of enclosure in St Peter. St Brelade included dune sand on the low ground, as well as the sandy Quennevais on the plateau top and the difficult southern headlands, and recorded by far the highest proportion of unenclosed land - 63.5%.
The number and extent of orchards is the most impressive single item of information to be had from the map of 1795. Orchard land accounted for no less than 15.5% of the whole island. It was widely distributed over the plateau, although much more thinly in the west than elsewhere, with a slight concentration in the west centre and a very dense distribution in the east. There were also patches of orchard on the slopes which bound the plateau, as for example in the south of St Lawrence, in Grouville and in St Clement; such sites enjoyed, among other advantages, a favourable aspect.
More than 20% of the total land area was under orchards in St Saviour (36%), Grouville, St Lawrence, and St Martin. The only parishes showing under 10% were St Brelade and St Ouen, with 4% and 3% respectively. When the area of orchards is expressed as a percentage of enclosed land, the figures are even more striking. Since the whole of St Saviour was already enclosed, the figure is again 36%, but it is seen in addition that Grouville, St Martin, St Lawrence, and Trinity had more than one-fifth of their enclosed land under orchards. The lowest proportions naturally occur in the west, but only St Ouen had less than 10%.
It is unnecessary at this juncture to comment on the relation of the extent of orchards to the production of cider, which has been mentioned on previous occasions. For present purposes, the most significant fact is that, before the end of the 18th century, Jersey had nearly seven square miles of land devoted to a specialised cash crop, which required a fifth of the improved and settled land. Thus the history of land use and of cultivation during the 19th century and in later years is seen to continue an earlier theme, instead of representing a revolutionary departure from precedent. In this respect Jersey differs signally from much of the English mainland, where subsistence farming continued in some parts well into the 19th century.
Measurement of areas
The measurement of a large number of small, irregular areas such as those present in the distribution of orchards, is a wholly technical problem. The use of a squared transparent overlay, and that of a planimeter, were both rejected as being difficult in practice and possibly open to serious error in the circumstances. Instead, orchards and unenclosed land were transferred from the 1795 map to a reduction of the OS map with a scale of 4 inches to one mile, and were traced off to give a map on a single sheet. A photographic copy of this on a scale of approximately 2 inches to one mile was cut up with a scalpel, and the proportions of open land, orchards, and enclosed land other than orchards determined separately for each parish by weighing in a standard laboratory balance. The results (correct to 0.005 gm) were then converted to percentages, and these further converted to vergees, by a mechanical computing machine. Some small errors must inevitably have crept in, for example errors of cutting, or errors due to paper stretch and shrinkage, but a check of selected results suggests that the cumulative error in any single case is not likely to be greater than five per cent of the areas stated.