Agriculture since 1870

AGRICULTURE SINCE 1870

Between 1870 and the present day English agriculture, and Wiltshire agriculture along with it, has experienced severe fluctuations of fortune. (fn. 1) In general the causes of these fluctuations have been in no way peculiar to Wiltshire, and the fundamental explanation of the county's agricultural history therefore lies outside the scope of this chapter. Certain local peculiarities in the way the county reacted to changes in the comparative costs of agricultural production in England and in overseas exporting countries are, however, discernible. Strictly speaking, these peculiarities pertain not to the county, but to agricultural regions which cut across county boundaries. Thus the fortunes of southern, Chalk, Wiltshire have been linked to those of the similar parts of Berkshire, Hampshire, and Dorset; while the history of the Clay Vale is in many respects that of Somerset as well.

Some of the bare statistics of change may be deceptive. The number of male farmers and graziers in the county, as given in the Census Reports, having fallen from 1851 until 1871, rose in every decade thereafter, from 2,977 in 1871 to 4,261 in 1921. It fell in 1931, but seems to have risen again in 1951. (fn. 2) The steep rise from 3,351 in 1911 to 4,261 in 1921 may possibly reflect an influx of postwar smallholders, although there was in fact a drop of 500 in the number of separate holdings between those dates. (fn. 3) Otherwise there is no evidence of a subdivision and multiplication of farms between 1871 and 1911, which would produce an increasing number of farmers. Since 1921 there may have been a consolidation of holdings which would have led to a decline in the number of farmers. The figures of separate holdings, however, decline continuously from 7,848 in the first return of 1875, to 5,343 in 1940, and this suggests that at most the decline was an acceleration of an existing tendency, and that ever since 1875 the reaction to altered circumstances had been a reduction in the number of individual farms. (fn. 4) The fluctuations in the number of farmers given in the Census Reports probably reflects changes in habits of self-description rather than real changes.

The figures of holdings are themselves deceptive, for they deal with administrative and legal units, rather than with units of agricultural production. The decline in numbers of holdings down to 1940 understates the real decline in numbers of production units, because since the eighties the practice was spreading whereby one man operated several distinct farms. (fn. 5) The rise in numbers of holdings since 1940 does not necessarily reflect an increase in the number of production units, since the tendency to operate multiple farms has in fact been growing vigorously, and the bulk of the increase has come in the holdings of less than 5 acres apiece, not classified as farms.

The decline in the labour force has, however, been certain, dramatic, and almost uninterrupted. In the census of 1871 there were 27,026 'male workers', while in 1954 there were 10,744 'regular male workers'. In the same period the number of female workers declined from 5,432 to 513. (fn. 6) The total cultivated acreage has ranged from 722,551 acres in 1870, to 759,412 acres in 1887. It then declined without pause to 593,636 acres in 1944, and has risen again to 622,009 acres in 1954. Meanwhile the county from being 58 per cent. arable in 1870–4 dwindled to a bare 26 per cent. arable in 1935–9, and has since then once more become predominantly arable with 52 per cent. of the cultivated area under the plough in 1950–4. (fn. 7) An examination of changes in livestock shows that the county has been denuded of sheep, the 773,986 of 1870 having only 73,481 successors in 1954. It has, however, become thronged with cattle, for while in 1870 there were 83,475, in 1954 there were 217,950. The farm horse has, of course, become a rarity in that time. (fn. 8) Such figures clearly tell of enormous changes.

In retrospect the early eighteen seventies came to be regarded, as do most times before any great change, as a period of great prosperity. Agricultural labourers have never been likely to regard the times before 1878–9 as very happy ones. In Wiltshire, partly as a result of a visit by Joseph Arch to stimulate union organization, the labourers obtained wage increases between 1873 and 1877 of 1s. or 2s. a week; these brought their ordinary weekly wages in winter to between 10s. and 12s. (fn. 9) Their wages and conditions, however, still remained greatly inferior to those of urban and industrial workers, and their standard of living was barely above subsistence level. The seventies, indeed, quite apart from the position of the labourers, were not at the time regarded as a period of bright agricultural prospects and serene progress.

Institutional problems in the landlord-tenant relationship remained unsettled, adversely affecting agricultural efficiency and output. Damage to crops by game, especially by rabbits, continued to cause loss of output, sometimes serious, some financial loss to tenants, and a certain amount of ill feeling. If the tenants' grievance was not more loudly voiced, this was because so many preferred occasional monetary compensation, and a chance of sport for themselves, to maximum crop yields. It was a problem which, on the less enlightened estates, and those managed with strict preservation of game, did not begin to be solved until the 1880 Ground Game Act, and by then the balance of bargaining power between landlord and tenant was in any case already shifting decisively in favour of the tenant. (fn. 10)

Of greater importance was the question of security for tenants' outlay in farming: as long as this was felt to be inadequate, as it still was in the seventies, it acted as a discouragement to good farming. A committee of the Central Chamber of Agriculture found in 1874 that in Wiltshire, as in most other counties, there had been considerable improvements in the previous 25 years in the general practices governing the compensation paid to outgoing tenants for unexhausted improvements. The range of compensation, however, was still partical, ill defined, and far from uniform; and important categories of expenditure on feeding stuffs and fertilizers were totally, or frequently, omitted. (fn. 11) The existence of this state of affairs was confirmed by the Wiltshire witnesses before the Royal Commission on Agriculture in 1881. (fn. 12)

The insecurity was most felt where most capital was committed in the best farming. Not surprisingly, it was at Wilton that the largest occupier used the opportunity of an audit dinner in 1871 to 'throw out an intimation that as yearly tenants they were scarcely justified in making the large outlay which high farming required—and that they ought to have better security, of course meaning that they ought to have leases'. (fn. 13) The agent, Lord Pembroke's guardians, and Lord Pembroke himself, when he came of age shortly afterwards, all objected to leases, none were granted, and the source of dissatisfaction remained until the unpredictably changing circumstances of the depression put leases quite out of favour.

Alongside the technical and economic limitations of the seventies, however, the institutional imperfections were of secondary importance. The detailed report in 1867 on the condition of the Savernake estate by a Shropshire land agent, H. W. Keny, concluded that it was 'with some few exceptions, very badly farmed, and that the produce obtained from the soil is very far below what it ought to be'. For this he blamed a slovenly crop rotation, overstocking with breeding ewes, inadequate use of purchased manures, poor management of both horse and human labour, bad and insufficient farm buildings, and the lack of any compensation for unexhausted improvements; defects for which poor estate management, in the selection and control of tenants and in the provision of finance, was largely responsible. (fn. 14) The years after 1878 revealed many other areas where tenants had been farming with insufficient capital, and with methods too backward and inefficient to lend themselves to yielding the largest output at the lowest cost as the new price levels demanded.

In the south, at any rate, even the technically efficient farmers of the large corn and sheep farms were in an economically weak position in the seventies; for the north there are no data. In the decade before 1877 the farmer's income seems to have formed only 20 per cent. of the net farm product, while the landlord took 45 per cent. In other words the farmer's income was still at or below the one-half of his rent at which it had been estimated for income-tax purposes in 1842. In the interval rents had of course increased, so that the farmer's money income had presumably increased also, roughly in step. Since, however, between the forties and the seventies farming operations had been changing, and now required considerably more capital per acre from the farmer, this constant rate of income was only maintained through an increase in the interest and a decline in the profit element. The farmer's income, that is profit and interest, was in fact 8 per cent. on his capital. (fn. 15)

To put it in another way, the large corn and sheep farmers of southern Wiltshire appear, by the seventies, to have evolved a high-cost farming system technically excellent no doubt, but economically unsound. Their farming had advanced so far along the path of diminishing returns that their profit margins were uneconomic when compared either with other types of farming or with other kinds of business. The farmers themselves were doubtless more or less content so long as they managed to win a livelihood from their operations, and would perhaps have agreed that an agreeable and pleasant life carrying a desirable social status was adequate compensation for low profits. (fn. 16) The real cost of this state of affairs, however, was an undue and excessive vulnerability to the hazards of bad seasons and unfavourable price movements.

It was a symptom of this condition that in June 1871 several of the largest tenants on the Wilton estate expressed 'a very gloomy view of the future prospects of farming', and asked for rent reductions to enable them to carry on: two of the complainants indeed quitted the following year. (fn. 17) This came after a run of three good harvests, after only two years of moderately depressed corn prices in 1869 and 1870, and while wool and lamb prices remained steady: conditions which should have caused no trouble to soundly based farmers. (fn. 18)

If some of the larger farmers felt themselves to be on uncertain ground, some of the great landowners were also anxious about their financial situation. On receiving the tenants' complaints in 1871, Lord Pembroke's senior guardian remarked, 'These claims for a reduction of rent are most serious—coming as they do after the very large expenditure on the farms. I always understood from Mr. Robson [the agent] that the rent would be increased so as to cover the rent charge placed on the property in consideration of the outlay, and I certainly never contemplated a demand for a reduction of rent.' (fn. 19) In fact the financial results on the Wilton estate of the large outlay on improvements made since 1846 appeared to be moderately satisfactory. The increase in rents between 1846 and 1871, deducting allowances made to tenants in both years, was just over £10,000. The annual cost of the improvements, financed by 30-year loans from the General Land Drainage and Improvement Co., was, on the completion of the final works contract in 1871, £9,300; a sum which included interest payments and redemption instalments. (fn. 20) As, however, part of this rent increase was due to the policy of allowing lifeholds to fall in and reletting them at rack rents in place of the former nominal rents, the amount by which rents increased on a constant acreage was probably not sufficient to balance the improvement charge. (fn. 21) In any case the rent increases were spread over the estate as a whole, and were largely the result of general factors such as price movements and intensified competition for farms. The improvements were not similarly spread, and it is almost certain that the additional rents of the farms on which improvements were carried out were not enough to meet their cost. The experience of Lord Pembroke's minority, 1862–71, in this respect was even more clearly unsatisfactory: in this period the improvement charge increased by £5,000 a year while rents rose by only £3,100. (fn. 22)

Much the same sort of thing had been happening on the Longleat estate, and towards the end of the sixties the Marquess of Bath became concerned lest his outlay on the estate was not producing an adequate return. Between 1859 and 1867 over £48,000 was spent on the estate. In the same period rents increased by about £4,000 a year, but the marquess discounted most of the increase: 'This is to be accounted for in the increase of property rented and relet, in lands purchased, in life leases that have expired, and partly no doubt in the improved value of some farms by the outlay on them by draining . . . and buildings, but the latter is I fear a small item.' His anxiety led him to bring in John Clutton, the agent for much Crown and Church property, to report on the finances of the estate. Although Clutton's report established that there had been no great extravagance in the outlay, in that the farm buildings upon which he reported had been erected at reasonable cost, it showed that the net increase in rental was equivalent to a yield of about 3 per cent. on the outlay. The main effect of improvement outlay was simply to maintain the rent producing power of the estate, not to increase it. (fn. 23)

At the other end of the scale was the Savernake estate, where the heavy burden of debts and family charges had kept improvement outlay to a minimum. There, between 1846 and 1867, the gross rental less allowances had risen by 11 per cent., whereas on the Wilton estate in the same period it had risen by about 30 per cent. But at Savernake, on the limited improvement outlay, this increase was equivalent to a return of 22 per cent., while at Wilton it was equivalent to a return of only 6½ per cent. (fn. 24) The great improving landlords had their estates neatly farmed by the most capable tenantry available. They had, however, been investing their money uneconomically, and in doing so were largely responsible for creating an over-capitalized agriculture, for the new buildings and the drainage they provided led to an economically precarious system of farming. On the other hand, to an unimproving landlord like the Marquess of Ailesbury, the rent increases were obviously to a large extent an unearned increment, and the too small scale of physical improvements showed its effect in the bad farming and secondrate tenantry which were noted as characteristic of the estate in 1867.

In 1873 estates of over 10,000 acres covered 23 per cent. of the area of Wiltshire. If it is assumed that south, or Chalk, Wiltshire covered two-thirds of the county, and that all the great estates thus defined lay within it, then they covered 34 per cent. of south Wiltshire. The three largest estates which have been examined, Wilton with 47,000 acres, Savernake with 37,993 acres, and Longleat with 19,977 acres in Wiltshire, had between them in 1873 more than half the area of all the Wiltshire great estates thus defined. They covered 12.6 per cent. of the area of the county, and 19 per cent. of south Wiltshire. (fn. 25) Their farm acreages at the same time covered similar proportions of the total cultivated acreage reported in the Agricultural Returns. (fn. 26) Hence it is reasonable to conclude from the history of these three estates that the proper balance of landlord's investment requisite for a healthy condition of agriculture had not been reached on the great estates of Wiltshire.

Some landowners, like the Marquess of Bath, felt further that their financial position was unpleasantly and unnecessarily weak, because total estate expenses absorbed too high a proportion of gross income. To some extent this may have been the case. Certainly net income was a considerably smaller proportion of gross income than it had been 30 years earlier, and in spite of the increases in gross rents, net income per acre had also declined. The decline was, however, in effect a sign that changes in the distribution of the total agricultural product had been taking place, and that the margin available to repay the outlay on agricultural production was in danger of severe limitation, at least in the arable farming which predominated on these estates; as such it did not necessarily mean that agricultural prosperity had been declining, nor that agriculture had reached an unhealthy condition.

The seventies were, therefore, far from being a Golden Age in Wiltshire, and the structure of agriculture was not free from blemishes and weaknesses. They were nevertheless a period of expansion. Rents continued to rise. They reached their peak on the Longleat estate in 1878. On the Wilton estate, which lay wholly within corn-and-sheep Wiltshire, and where a ceiling for rents was fixed by the sliding scale of 1852, the peak was passed in 1873. (fn. 27) More important was the increase in the area under cultivation. It is true that the maximum recorded cultivated acreage, crops and grass, was not reached until 1887. It was then 759,412 acres, out of 859,303 acres in the administrative county. Between 1869 and 1879, however, the total cultivated acreage had increased by 4 per cent., while between 1879 and the peak year, 1887, it increased by only 0.8 per cent. The maintenance of this slight rate of increase after 1879, due to the growth in the recorded area of permanent grass, was almost certainly a statistical illusion, caused by individual farmers upgrading into the category of permanent grass land which they had earlier regarded as rough grazing and therefore outside the scope of their returns. More eloquent of the increasing acreage was the fact that the area under arable crops of all kinds reached its peak in 1875, at 430,783 acres, having increased by 3½ per cent. since 1869. The same was true of the area of corn crops, which, at 227,363 acres, had increased by only 1 per cent. since 1869. The area under rotation grasses reached a peak in 1876 of 85,391 acres, which was not surpassed until 1945; it had increased by 29 per cent. since 1869. (fn. 28)

The annual agricultural statistics only began in 1866, and owing to their novelty and to the fact that they were compiled from voluntary returns it is natural that the accuracy of the figures in the early years of the series should be suspect, and that increases in acreage should be attributed to increasing efficiency in collection of returns rather than to real agricultural changes. (fn. 29) Nevertheless there is a very strong suggestion that the Wiltshire figures depict real changes. The maximum recorded area under wheat was 103,522 acres in 1868; thereafter the wheat acreage fluctuated, falling until 1872, rising until 1874, falling until 1877, and rising once more in 1878. In 1878, however, it was 8 per cent. smaller than in 1868. The largest area under green crops ever recorded was in the first year, 1866, when there were 110,809 acres. The figure fell sharply in 1868, the land no doubt being switched from roots to wheat, recovered and remained steady until 1875, after which a decline set in which has never been halted. In 1878 it was also 8 per cent. below its peak level. (fn. 30) Overall acreage increases, which masked declines of this order in particular important crops, are unlikely to have been exclusively, or mainly, due to increasing efficiency in the collection of the statistics.

Some confirmation of real expansion in the arable acreage is suggested by the grubbing of parts of the large Black Dog Woods, Chapmanslade, in 1868–9. (fn. 31) From Aldbourne it was reported in 1878 that 'on many farms the steam plough has taken the place of the flock, and corn growing has superseded the making of meat—a state of things which cannot last long'. (fn. 32) Wider but less precise confirmation is provided by general statements that the arable area was being increased, most rapidly in the fifties, but continuously until the depression, mainly by putting high thin-soiled downlands, the 'bakelands', under the plough, but also by ploughing up old pastures on claylands, as in the Swindon district. (fn. 33)

The expansion of the arable area was no doubt a contributory cause of the ensuing depression, overlooked by contemporaries in their preoccupation with wet seasons and American competition. It was generally agreed that lean times for the corn and sheep farmers dated from 1874 or 1875. (fn. 34) Yet in Wiltshire until the harvest of 1878 yields of all grains, of hay, and of roots, appear to have been above average, and harvest conditions generally to have been very favourable. (fn. 35) Although sheep, lamb, and wool prices remained steady or rising, the prices realized for cereals, particularly for barley, were low and unsatisfactory. (fn. 36) Here an inflated domestic supply, partly from the additional acres, no doubt played its part in depressing prices, alongside the growth in imports. Between 1878 and 1880, however, years of poor or disastrous harvests and steeply falling grain prices, the paramount influence on the supply side was plainly that of the new American cornlands.

The preceding arable expansion exerted its greatest influence upon the effects of the depression rather than upon its causes. In the first place the period of high total grain output but unremunerative prices, from 1874 to 1878, had meant that farmers had much lower profits than they expected, or even small losses. Their resources were, therefore, so depleted that many more were unable to tide themselves over the really intense depression of 1879 and 1880 than would otherwise have been the case. (fn. 37) For the farmers who did survive it meant an attenuation of their resources, so that their endeavours were excessively concentrated upon absolute economy in outlay, and thus proper adaptation to the changed economic situation was delayed. Secondly, the arable expansion imposed an additional cost on both owners and farmers in the period after 1880. One of the most obvious reactions to the depression was to lay arable land down to grass; and this operation entailed not only investment in grass seed and in proper tillage preparation, but also the sacrifice of several years' income from the crops which were forgone. Where the land thus laid down to grass was land which had only recently been ploughed up, two considerable acts of investment were required, in relatively quick succession, both of which were in a sense unnecessary. Moreover, reconversion to grass in the eighties and nineties made demands on what had become very limited resources, so that the correction of past mistakes took place at the expense of adequate finance for positive agricultural development. In very many cases, principally on the 'bakelands', but also in such lowlands as Pewsey Vale and on some of the heavy clays, arable land was not properly laid away to grass, but was simply abandoned to tumble down, i.e. left to revert to nature. (fn. 38) In such cases the final product, ridden with weed and couch grass, was greatly inferior as livestock pasturage to the original grass, whether this had been downland or valley pasture, and the loss consequent on the breaking-up operation was thus multiplied.

The depression was associated with a prolonged run of disastrously unfavourable seasons, with delayed growth in the spring, and with wet and late harvests, which themselves had a cumulatively adverse effect by producing late and difficult ploughing and sowing conditions. Moreover, the excessive wetness was a prime factor in causing the sheep rot, which ravaged the flocks of southern Wiltshire, and in causing a great deterioration in the condition of the pastures on the retentive subsoils of the Clay Vales.

Many witnesses, including the commissioners themselves in their report on agriculture in 1881, saw the wet seasons as the primary cause of the depression. E. Lywood, a large corn and sheep farmer of Maddington, near Shrewton, was not, however, of their number. He rightly pointed out that wheat and barley were not the only crops which should be considered, since they formed only a portion of the cash receipts of Wiltshire arable farms. On such a farm, in the decade 1868–77, receipts from all types of cereals formed 59 per cent., and from all types of livestock, 40 per cent. of total revenue: receipts from wheat and from sheep were equal, each being 27 per cent. of the total. (fn. 39) Lywood went on to show that of the important feed crops, the yield of turnips had been good every year, and the hay crop was good throughout the wet seasons except in 1879. He seems not to have suffered much on his farm from the sheep rot, since the number of store lambs sold each year did not drop. Nevertheless, the price of store lambs took a downward turn in 1879 and 1880, as did that of wool. He laid the blame for the depression clearly on falling prices for all kinds of agricultural produce.

Lywood obscured the simplicity of this diagnosis by attemping to show that increasing costs of production were an equally important cause, eating into the farmer's income at the top, while falling prices ate into it from the bottom. To attempt to demonstrate increases in the unit costs of cereals, from 3s. a bushel of wheat in 1868 to 6s. a bushel in 1878, was futile, since the cost of production of an acre is more or less constant irrespective of yields, and therefore the lower the yield the higher the unit cost. Lywood's figures of costs were, indeed, simply another way of saying that there had been some bad harvests. In the main, however, he adhered to the central point that falling prices were responsible, and that prices were falling because 'the full effects of free trade, which was enacted in 1842 [sic], have not been felt until the present time'. (fn. 40)

The fortunes of the Clay Vale, dairy Wiltshire, illustrate clearly the limitations of the meteorological interpretation. Here the soil and topography dictated an overwhelming predominance of permanent grass. In 1880, for example, in the Melksham Union, in the Clay Vale, permanent grass occupied over 86 per cent. of the cultivated area and corn crops under 7 per cent., compared with 31 per cent. permanent grass and 30 per cent. corn crops for the Amesbury Union, selected as typical of the chalk district, and 45 per cent. permanent grass and 27 per cent. corn crops for the county as a whole. (fn. 41) The farmers of the vale depended on dairying, cheese being at this time their chief market product. They were indifferent to the level of cereal prices, or even stood to gain from falling and low grain prices in so far as they were already in the seventies and eighties purchasers of feeding stuffs.

Yet these farmers too suffered from depression in 1879 and 1880. Certainly the wet seasons, by affecting the quality of the pastures and by aggravating liver rot among cattle, caused a fall in the annual make of cheese. The crisis, however, was confined to 1879 and did not persist throughout the wet years; and it was a crisis of over-production. The high and rising cheese prices of the boom years of the early seventies had stimulated an increase in the American production of cheese for export. The American cheese saturated the market, prices fell by 30 to 50 per cent. in 1879 to a level too low for even low cost American cheese to find profitable. Merchants curtailed their imports of American cheese, and in the course of 1880 prices righted themselves, returning to what was accepted as a normal level. (fn. 42)

This sharp, but brief, crisis left no really deep marks on the agriculture of the Clay Vale. The sudden fall in income in 1879, due to low prices and an output of cheese said to be one-third below normal, undoubtedly caused distress, and led to increased indebtedness to cheese factors. (fn. 43) This drain on tenants' working capital may well have assisted the subsequent switch from cheese to liquid-milk production, for the cash return for liquid milk was almost immediate, while cheese had to be stored for several months before it was marketable. Wiltshire dairy farmers often sold an entire year's make at a time to the factors, and thus had to provide a running year's expenses. The immediate effects were illustrated by rent reductions of not more than 5 per cent.; and no reports came in of dairy farms falling into hand; on the contrary they were said to remain eminently lettable. (fn. 44)

In the longer run dairy-farming Wiltshire had to make many adjustments. Some of them were under way before 1879, starting perhaps with the cattle plague of 1868, which ravaged the dairy herds of the London cowhouses, and gave the first powerful stimulus to the development of a rail-borne London milk supply. They ended just before 1914 with the virtual extinction of Wiltshire cheese-making. The long period over which they occurred suggests that they took place at the pace of normal economic change and were not the outcome of anything which can be called a depression.

In the decade following 1868 many north Wiltshire dairy farms which were close to railway stations, and had no difficult or expensive local transport problems, entered the liquid milk market. The farms astride the G.W.R. between Swindon and Melksham are an example. (fn. 45) Even then the dairy farmers probably varied their product as seasonal shortage or plenty dictated. For instance, on West Leaze Farm, 1½ mile west of Swindon, and one of the prize farms in the Royal Agricultural Society's competition of 1878, it was reported that: 'At our first visit (January) the milk was sold to the Aylesbury Dairy Company; but when we were there in May Beaven [the tenant] had recommenced making the produce of his cows into cheese.' (fn. 46)

Of greater moment in the seventies were changes taking place within the business of cheese-making itself. The traditional products of north Wiltshire were Gloucester cheeses, either single or double according to size, and North Wiltshire loaf cheese, a derived and modified version of the Gloucester. The best qualities of these cheeses were highly esteemed and in the seventies sold in the same price range of 60s. to 70s. a cwt. as good Cheddar, though the very best Cheddar was worth more than this. As both the Gloucester and Wiltshire were skim-milk cheeses, their manufacture produced a certain amount of butter as a by-product, which was sold locally at 1s. to 2s. a lb. Cheddar on the contrary was a full-milk cheese, and it was held that any difference in the quantity or value of the Gloucester or Wiltshire cheese was more than counterbalanced by the value of the butter. (fn. 47) In an increasingly competitive and selective market, however, inferior qualities of all kinds of cheese were forced down to unremunerative prices. Here Cheddar had a great advantage, for its production had, by the seventies, been reduced to a standard process capable of accurate control and virtual elimination of poor quality produce. For making Gloucester or Wiltshire cheeses, on the other hand, no system existed, all depended on rule of thumb, and the degree of experience and skill of the individual cheesemaker, generally the farmer's wife. This always remained so in the case of the Wiltshire loaf cheese. (fn. 48)

One way out of this situation was to try to introduce more precision into the Wiltshire process. For example, by 1878 nearly all cheesemakers were using purchased liquid rennet, of standard quality, in place of home-made rennet produced from dried vells from young calves. (fn. 49) But this effort was not carried far: not until 1912 was it reported that thermometers were widely used in what Wiltshire cheesemaking then remained. (fn. 50) A simpler step to take was to abandon Wiltshire cheese altogether and adopt the already standardized Cheddar process. Cheddar-making, whether in dairy or factory, meant the end of butter-making. The 'butter money', just as 'egg money' often was, had been a perquisite of the farmer's wife, a sort of earned pin-money. Here was a powerful influence against change. Nevertheless the economic advantages were such that by 1875 the Frome district of Somerset, once a Wiltshire cheese area, had been conquered by Cheddar, and Cheddar was effectively penetrating Wiltshire. By 1878 most of the local makes were fast disappearing from the London cheese-warerooms in Tooley Street, in face of competition from Cheddar and from American and Canadian cheeses. But whereas Derbyshire, Leicestershire, Blue Vinny, Ayrshire, or Dunlop cheeses were nowhere to be seen, a few Wiltshire loaves were still coming to the London market. (fn. 51) A steady demand remained, however, in Leicester and the Midlands, sufficient to absorb the dwindling output of Wiltshire loaf cheeses over the next 40 years.

Thus, just as total cheese production in the county was beginning to drop sharply, Cheddar cheese was in any case beginning to drive the native product out of existence. Sir Gabriel Goldney, who farmed part of his estate near Chippenham, stated in 1894 that he had 'tried making Cheddar cheese, but after a couple of years I found it not so profitable, taking labour into account, as selling milk'. He also stated that all the farmers of the area sold their milk raw if they possibly could. (fn. 52) By 1912 very little North Wiltshire cheese was being made. Before 1914 it had dropped out of the categories for which prizes were offered at the annual shows of the Bath and West of England Society. When the Ministry of Agriculture came to inquire into agricultural markets in the 1920's Chippenham was no longer a cheese market; indeed by then 'practically all the milk produced is sent in liquid form to London or to factories and depots in the county. Little dairy produce is made on the farms.' (fn. 53) Recently some cheese-eaters have shown an aversion to the very uniformity of the factory product, which was the main appeal of the factory system to the producer, and have been prepared to back their preference for farmhouse varieties by willingness to pay for it. Coming, however, in 1957, this change in demand is probably 50 years too late to resurrect the North Wiltshire loaf cheese.

Changes in the purposes of dairying gradually produced changes in the type of cattle kept. At one time crosses of the 'north country Longhorn' had been the favourite cheese-cows in Wiltshire, but these had been replaced in the course of the 19th century by Shorthorns, which gave larger milk yields. (fn. 54) When the judges visited the dairy farms entered for the competition of 1878 'nearly all the cattle we saw were either Shorthorns or crosses of that breed, and all the bulls used were Shorthorns'. (fn. 55) The imprecision of this language conveys a strong impression that even on farms entered for competition, and therefore felt by their tenants to be examples of the very best practice, no clearly defined and carefully preserved breeds or crosses of dairy cattle had been established. When the judges remarked further that 'the use of pure bred males cannot be too strongly impressed on cow owners in important dairy districts', they came as close to blunt criticism as the decorum of the Royal Agricultural Society ever permitted. The remark meant in effect that in the ordinary practice of the district, dairy herds were nondescript and haphazard collections of cows, most of which perhaps bore some traces of Shorthorn ancestry. (fn. 56)

From this time onwards, no doubt, attention to breed of cattle increased, under the influence of competition rather than of exhortation. Change undoubtedly proceeded slowly, with one particular type of Dairy Shorthorn making headway, and perhaps becoming the dominant breed by 1914. Thereafter came diversity, until by 1956 the numbers of Ayrshires and Friesians may each have been as great as the number of Dairy Shorthorns. (fn. 57)

The cow was not, of course, the only beast on the farms of the Clay Vale. In the cheese days the pig played a part in the dairy economy. It consumed the whey, it helped to provide manure for the hay meadows and for the small area of corn grown, and it was an important source of income: it was said that 'the pig was the gentleman who paid the rent'. (fn. 58) Between the years 1870 and 1874 there were on the average 66,630 of them. (fn. 59) The growth of cheese factories and the liquid-milk trade removed the whey from the farms, however, and there gradually ceased to be any compelling technical reason why pig-keeping should form part of dairy farming. Nevertheless, the movements in the total head of pigs shown in Table 4 below, indicate that the normal dairy farmer has continued to keep approximately similar numbers of pigs, although these have now come to be fed on imported feeding stuffs.

The development of consumer demand for lean, mild-cured bacon, has been continuous since the seventies. It led the local curers to popularize the Large WhiteBerkshire cross as the ideal bacon pig, particularly after the nineties. The local curers also co-operated with the Wiltshire County Council in testing a wide range of pig diets. (fn. 60) Success in these efforts produced pigs which matured more quickly, and hence resulted in economies in working capital, and in lower costs of production. It also meant that a herd of pigs could be made ready for market more rapidly, so that a fall in total numbers kept at any one time did not necessarily mean a decline in annual output. Comparison of total numbers between five-year periods, as given in Table 4, roughly eliminates the violent short period fluctuations of the pig cycle; but it does not allow for this second factor. It is likely that in 1950–4, with a rising trend in numbers, annual output of pigs was considerably higher than in 1870–4. Certainly in both the First and Second World Wars, when imports of feeding stuffs were severely cut, annual output fell steeply and took some years to recover afterwards. Possibly in the 20 years prior to 1914, when the breeds of pig were changing most rapidly, output remained constant though total numbers fell.

It can be argued, therefore, that the chief effect of Danish competition has been to prevent output rising, and thus force Wiltshire producers to accept a declining share of an expanding market. Certainly Wiltshire farmers do not seem to have exploited the presence of bacon factories in the county to the fullest extent. They have never provided more than a fraction of the pig supply required by these factories, and in 1956 it was estimated that they supplied between 10 and 15 per cent. of the pigs handled in local factories. (fn. 61) Overwhelmingly the pig output of Wiltshire has continued to come from farmers to whom pig keeping is a subsidiary activity. In a sample of farms in the Clay Vale, 24 per cent. of total receipts came from pigs in 1939, and only 12 per cent. in 1955. (fn. 62) In consequence, the economies of specialization and of scale have not been explored, and costs have not been reduced to their fullest extent.

All these broad changes in the farming practices of the Clay Vale occurred gradually over the 80 years after 1870. The farmers of north Wiltshire have indeed experienced sharp and unpleasant crises, as in 1879–80, or in 1921–2 when milk and pig prices fell sharply, or again in 1930–1 when this happened again. These have been general economic crises, however, by no means confined to agriculture. Between crises there is little sign that these farmers suffered from depression.

Though the dairying portion of the Longleat estate lies mainly in Somerset, and includes only a small part in the Wiltshire Clay Vale, south-west of Westbury, the fortunes of the estate serve to indicate the contrast between trends in the Clay Vale generally, and those in Chalk Wiltshire. On the Longleat estate in 1894–5 rents actually received, after deducting remissions from the nominal rental, were on a constant acreage only 6 per cent. below their average amount in the peak years 1874–8, and were 6 per cent. above their level of the period from 1869 to 1873. (fn. 63) In the interval the landlord had taken over the actual payment of an increasing amount of the tithe rent charge, until in 1891 it had become compulsory for him to pay it all. Deducting the landlord's tithe payments, the rent in 1894–5 was 11 per cent. below that of 1874–8, but exactly the same as that of 1869–73. A fall in rents of 11 per cent. from peak to trough was indeed less than half the amount by which the general price level fell in the same period, and implies that the purchasing power of the Marquess of Bath's rent per acre was rising. (fn. 64) The fall in money income did produce some inconveniences when it came up against fixed charges or costs which were not easily reducible, so that the income after paying estate expenses and taxation fell from 54 per cent. of gross receipts in 1874–8 to 48 per cent. in 1894–5. Nevertheless, the £22,082 which was the annual average of this net income in 1894–5 undoubtedly had greater purchasing power than the £24,736 which had been the average net income for 1874–8. Nor was this achieved at the expense of any drastic economies in the upkeep of the estate, for the £6,670 a year spent on repairs and £828 on improvements in 1894–5 represented physical works not markedly smaller than the £7,954 and £1,135 annual averages under these heads in 1874–8. (fn. 65) Not surprisingly the consulting agents felt that the estate was coming through the 'depression' very lightly. (fn. 66)

Beneath this buoyancy in the landlord's position may be inferred the existence of a tenantry well able to maintain payment of stable real rents, a tenantry whose own real incomes can therefore hardly have suffered even if their money incomes did fall. Moreover, the figures apply to the Longleat estate as a whole, no analysis by parishes or other subdivisions being possible. Nearly half the estate by annual value lay in the Chalk district, the area of Horningsham and the Deverills, and here there was certainly acute distress with steeply declining rents and farms falling into hand. (fn. 67) In the dairying part of the estate, therefore, rents must have been even higher and the tenants who paid them even more prosperous than the bare figures suggest. The course of arrears on the Longleat estate bears out this impression of a well-sustained capacity to pay rents. Only in the crisis year 1879, and the two following years, was a large amount of fresh arrears incurred, averaging however only 4 per cent. of the gross rental in the three years, and reaching no higher than 5½ per cent. of the gross rental in 1879 itself. After this no significant arrears appeared until 1893, and then they amounted to under 2 per cent. of the gross rental. (fn. 68)

In contrast, on the Wilton estate serious arrears had begun to appear from 1874. In 1876 alone they amounted to over 5 per cent. of the gross rental, and in 1879 to over 11 per cent., while in every year from 1884 to 1888 they were very close to 10 per cent. of a shrinking rental. At their summit, in 1889, the estate was carrying accumulated arrears of £20,468 on a gross rental of £24,659. The Longleat peak of accumulated arrears had come in 1882 and was only £7,123 on a gross rental of £40,512. (fn. 69) At Wilton gross rents, calculated on the same basis as that used for Longleat, reached their peak in the period 1869–73. By 1890–4 they were 42 per cent. below this peak on a constant acreage, and by 1895–9 they reached their lowest point at 43 per cent. below the peak. These figures for 1890–4 and 1895–9 were 39 and 41 per cent. respectively below the level of 1874–8. Deducting landlord's tithe payments from the gross rents, the rental of 1890–4 was 50 per cent. and that of 1895–9 55 per cent. below that of of 1869–73, and 47 and 53 per cent. respectively below that of 1874–8. (fn. 70) Such a drastic fall in gross income could not fail to produce an effect on net income. Owing to the large improvement rent charge, a fixed annual payment, the burden of which rose as rents fell, net income was in any case low on the Wilton estate. In 1869–73 it was 22 per cent. of gross income, in 1874–8 it was 26 per cent. Sufficient evidence survives to show that net income had fallen to 17 per cent. of gross income in 1889–91, and had become a negative quantity in 1896, when estate expenditure exceeded gross income by 8 per cent. The net income was £11,138 in 1874–8, and £4,907 in 1889–91; the loss in 1896 was £2,122. This fall was accompanied, moreover, by real cuts in estate maintenance. In 1869–73 £12,112 a year was spent on repairs, but only £6,077 in 1889–91, which meant that the physical equipment of the estate was being allowed to deteriorate. (fn. 71) Here the real decline in the fortunes of the landowner indicates the real decline in the position of the tenantry, just as the course of the Wilton arrears shows the prolongation of the tenants' distress.

The Savernake estate lies in a different part of Chalk Wiltshire, and includes a large area in the Pewsey Vale; it also illustrates the fortunes of the corn and sheep-farming area. The gap in the estate accounts between 1867 and 1895 means that comparisons with the other estates cannot be made over exactly similar periods, and that there is no information for the critical years of rapid decline. By 1895 the rent reductions were complete, and arrears had been almost eliminated. No doubt they were mainly written off. The figures which can be compiled, however, tell a very similar story to those of the Wilton estate. Rental, reduced to a constant acreage basis, and with tithe payments deducted, was for 1895–9 48 per cent. below the level of 1863–7. On this estate the utmost economy in estate expenses was imperative. By 1895 not only was the Savernake estate the Marquess of Ailesbury's sole source of support, his Yorkshire estate having been sold in 1886–7, (fn. 72) but also the estate had to carry annual charges of £3,500 for interest on inherited debt, and of £5,000 for jointures, out of a total income of just over £23,000. (fn. 73) The agent could only bemoan 'the terrible jointures', and look for no financial relief 'until the Dowagers die'. Meanwhile he pruned expenditure 'as far as I possibly can consistently with the renovation and maintenance of the estate'. (fn. 74) As a gesture, in token of his sympathy with the difficulties of the neglected and impoverished estate, which the 5th Marquess had inherited from his spendthrift nephew in 1894, the new agent charged no salary for himself throughout his agency. (fn. 75) In contrast, expenditure on the Wilton estate could be fairly liberal, since the Earl of Pembroke had the large income of his Dublin house property to draw on for personal and family expenses. Thus at Savernake net income after deduction of estate expenses and taxation was maintained at a fairly high proportion of gross income: 46 per cent. in 1863–7, and 41 per cent. in 1895–9. Moreover, it remained steady at between 40 and 41 per cent. throughout the subsequent period down to 1918, while at Wilton the proportion fluctuated, falling to 12 per cent. from 1907 to 1914, and rising to 28 per cent. in 1915–19, when the war forced a reduction in outlay on repairs. (fn. 76) In spite of the differences in the personal position of the two owners, however, the fall in Savernake net income from £20,252 a year in 1863–7, to £10,528 in 1895–9, and the fall in Savernake repair outlay from £11,130 to £4,834, was very much of the same degree as the fall in the Wilton figures. Indeed, not only was the movement in average annual repair outlay per acre similar, but so also was its actual amount. This may be seen in Table 1.

Table 1. Average Annual Outlay on Repairs, per 100 Farm Acres, in Pounds Sterling

The foregoing suggests that the rate of decline in gross rents, net incomes, and estate outlay was unaffected by differences in the condition of particular estates at the onset of the depression, for Wilton was then a model of good, if over-capitalized, management, while Savernake was an example of neglect.

The close correspondence between the amounts of rent reduction from peak to trough at Wilton and Savernake may have been no more than coincidental. When the Royal Commission was collecting evidence in 1894 and 1895 it was clear that there had been no uniform reduction of rents in south, or Chalk Wiltshire. E. P. Squarey (fn. 79) thought that the fall was higher on small estates 'which were screwed up to the extremest rack rent', and lower on large estates where rents had never been forced to their maximum. (fn. 80) In principle no doubt this ought to have been the case. Squarey, however, went on to speak of an average reduction since 1878 of between 35 and 45 per cent. and an extreme reduction of 55 per cent. (fn. 81) Since the reduction on the two largest estates in Wiltshire was greater than his upper limit, the point is scarcely proved. Moreover, on more medium-sized estates, the reductions also appear to have been at or above his upper limit. On Sir Henry Meux's estate, which comprised 11,895 acres at Dauntsey in 1873 and 20,461 acres in 1892, the reduction in rent between these two dates, after making adjustments for the intervening purchases, was 41 per cent. On Sir J. Dickson Poynder's estate of some 3,500 acres at Hillmarton the reduction between 1878 and 1892 was 40 per cent. (fn. 82) It was also 40 per cent. on Sir Michael Hicks Beach's estate of 7,800 acres at Netheravon, between 1873 and 1898. (fn. 83) On Sir John Dugdale Astley's estate of 7,887 acres at Everleigh in 1873, and 4,553 acres in 1894, the reduction on a constant acreage was 48 per cent. (fn. 84) By way of comparison, the fragmentary accounts of one small estate which chance to have survived, namely the 1,307 acres of the Revd. Edward W. Northey at Box, indicate a reduction of 52 per cent. between 1873 and 1894. (fn. 85)

Squarey was on firmer ground in stating that rents had fallen more on poor and inferior land than on good land. (fn. 86) It was precisely for this reason, because as very large estates they included poor as well as good land, that the reductions at Wilton and Savernake may well have been typical of the whole area. At any rate they accorded well with the opinion of R. H. Rew, the assistant commissioner reporting on south Wiltshire, who concluded after receiving reports of rent reductions varying from 28 to 70 per cent., that about 50 per cent. was the average reduction on corn and sheep farms. (fn. 87)

Rent reductions of this order were patently not sufficient to protect all farmers from distress, loss, and bankruptcy. Nevertheless, for the efficient farmer they were adequate. Rew collected figures of the costs of production per acre of one large farmer over the whole period since 1854. Between 1874–8 and 1889–92 his rent was reduced by just over 50 per cent., and his costs as a whole fell by 42 per cent. (fn. 88) In the same period the price of wheat fell by 38 per cent., barley by 32 per cent., and wool by about 50 per cent. (fn. 89) In such a situation the farmer's real income should not have suffered. It was the farmer that had lost all his capital in the crisis years, that was inefficient, or was rigid in his methods, who went under.

This process was painful, but by no means universal. The comment of the new farmers of 1911 that 'most of the older occupiers had been ruined or had retired in the depression', ignored the normal rate of change among farmers. (fn. 90) Table 2 shows that the depression years did severely curtail the average length of tenancies, but that in any case an occupancy of as much as 20 years was uncommon, and has remained so.

It has sometimes been felt that more really large farmers had been ruined than any others, that large farms had ceased to be workable, and that if only landowners had been able to afford the capital cost of subdividing large farms, and replacing one set of farm buildings by two or three, it would have been economically desirable for them to have done so. (fn. 91) The evidence hardly supports this view. Of the tenants on the Wilton estate who occupied farms of over 900 acres in 1875, 71 per cent. survived for 10 or more years, and 57 per cent. for 20 or more years. For tenants of all farms of over 500 acres the figures were 55 and 41 per cent. Of the tenants who occupied in 1885, 71 per cent. on farms of over 900 acres, and 68 per cent. on farms of over 500 acres, survived for 10 or more years. On this estate, therefore, the rapid turnover of farms in the depression was among the farmers of under 500 acres. (fn. 92) Experience in the Pewsey Vale might seem to point in the opposite direction. Here, however, the wholesale departure of large farmers in 1880 was caused by the obstinacy of a single large landowner who refused to grant a permanent reduction of rent, preferring to have his farms thrown up and left uncultivated, the farmhouses and cottages uninhabited. (fn. 93) No doubt these formed the bulk of the 3,893 acres of farmland reported as unoccupied and derelict in 1881. (fn. 94) But by 1887 these farms were let once more, and formed part of the vast farming enterprise built up by S. W. Farmer, who by 1895 was farming 14,000 acres in the vale. He then held five farms on the Savernake estate, but only one of these was large, comprising, as it did, 1,458 acres. (fn. 95)

Table 2

Duration of Tenancies

Wilton Estate 1855–1916 (fn. 96)

Percentage of total tenantry that occupied for:

Percentage of total tenantry that had occupied for:

Wiltshire 1941–3 (fn. 97)

Percentage of farmers that entered present occupancy in

S. W. Farmer, or rather the firm of Frank Stratton &Co., in which Farmer provided the finance and Stratton the managerial ability, was more than a freak, it was a portent. Not only did it introduce dairying and milk production for the London market into the Vale of Pewsey, where previously cattle had been 'universally banished from the district', (fn. 98) but it demonstrated, on an exaggerated scale, the economic answer to the problems of the depression. The way to cut costs to match falling prices, where the costs themselves in the shape of feeding stuffs and fertilizers did not automatically fall, was not to attempt to concentrate a given amount of labour and capital on a smaller area of land, as the idea of subdividing large farms implied. The solution was rather to introduce new products, of which milk was by far the most important, using cows as a factor of production in place of men, horses, sheep, and seed: in general terms to increase the amount of land in relation to the other factors of production. This could be done either by the Farmer method of creating larger units of production, or by allowing the level of labour and capital used in high farming to fall somewhat within existing farm boundaries. In either case the result was a lower output per acre, but a higher output per man and per unit of capital employed, and, provided that the cost of the land fell sufficiently, as it did, a lower unit cost.

Until 1914 the chief movement, in these terms, was to reduce the amount of labour and the amount of working capital used by reducing the scale of farming operations. The self-binder, reckoned to have been in very general use by 1901, (fn. 99) was an important means of replacing labour, particularly casual labour, by capital. At the same time opinion was emphatic that total abandonment of high farming, that is the adoption of extensive farming with a minimum outlay on labour, cake, and fertilizers, did not pay. The successful farmers were rather those who contracted the arable areas of their farms by turning over their thin 'bakeland' to sheep runs, but continued a modified form of the sheep and corn system, and the crop rotation which went with it, and who produced some milk where possible on their more fertile lower and 'bottom' land. (fn. 100)

It was only after the crash of 1922, when money wages were not greatly reducible, and the labour element in costs could only be reduced by drastic cuts in employment, that extensive farming came into its own. Even then it could take one of two forms. There was the ranching, 'dog and stick' type of farming, which then characterized the farms of Stratton & Co. and many of the south Wiltshire farms. On these there was wide abandonment of all cereal growing, and concentration upon the production of milk. Then there was the continuous corn-growing, on large mechanized farms, which was also to be found in the area in the twenties and thirties. Here occasional bare fallows were used to clean the land, and superphosphates employed to maintain fertility. (fn. 101) Both developments dispensed with hurdle sheep, with water-meadows, and with much labour, and thus marked the real break with the high farming systems. Still later, since 1940, the absolute shortage of labour, rather than its cost, has encouraged investment in labour-saving machinery, so that while the amount of labour employed in relation to land has continued to decline, the ratio of capital to land has risen sharply.

Until 1914 Stratton & Co. perhaps found few imitators. When at its peak of prosperity, this highly successful firm was farming 25,000 acres in Wiltshire and Berkshire, with 2,000 milch cows, and operations on such a large scale were not easily copied. (fn. 102) But in 1893 a tendency to run two or three farms together was noted; and in 1922 an occupier of 3,000 acres had apparently held that amount of land since before the First World War. (fn. 103) Large farm units were not rendered out of date and unworkable by conditions after 1878, they were on the contrary required by those conditions. The opposite impression arose from the undoubted fact that large farms were difficult to let in the eighties and nineties. (fn. 104) This was because the crisis years had caused such losses of tenants' capital, that when a large farm became vacant it was difficult or impossible to find a farmer with the necessary capital to take it on. The men who moved in from outside the county, from Devon, Cornwall, and 'the North', not apparently in a very large stream, were importing hard personal work and lower standards of living, not large capitals. (fn. 105) Banks were not disposed to lend liberally to farmers, and complete newcomers to the industry were not numerous.

In fact the very large holdings of over 1,000 acres held their own in the period 1875– 95, both in number and in the proportion of the total cultivated acreage of the county which they occupied. This was 15 per cent. in 1875 and 16 per cent. in 1895. All hold ings of 500 acres and over, which occupied 43 per cent. of the farm land in 1875 and 42 per cent. in 1895, (fn. 106) also maintained their position. After 1895 these size categories were no longer used in the returns. When they reappeared in 1934 there were only 29 holdings of over 1,000 acres against 87 in 1895; and when acreage figures reappeared in 1944, holdings of 500 acres and over occupied only 27 per cent. of the county. (fn. 107) It is possible that the largest farms, having lain on the poorest land, had been taken over by the War Office between 1895 and 1934. Apart from this, the decrease may well be fictitious: for as the downland of the large farms was regraded as 'rough grazing' it disappeared from the returns.

For the inter-war years there is little evidence for the size of farming units, though Baylis's great corn farm of 12,140 acres, mainly in Berkshire, extended onto the Marlborough Downs. (fn. 108) Since 1945 with the old incentive of cutting costs, and with the new one of providing units large enough for the economic use of such expensive singlepurpose machinery as combine harvesters, the trend towards larger farm units has flourished. In 1956 several men were reported as farming 8,000 to 9,000 acres or more, in production units each composed of perhaps a dozen separate farms. (fn. 109)

Given the overwhelming predominance of permanent pasture in the Clay Vale in 1879, and the gradual pace of change there since then, the major changes in crop acreages and in land utilization in the county as a whole, as shown in Table 3, clearly occurred in Chalk Wiltshire and in the small fringe along the north-western boundary of Cornbrash land, a Wiltshire extension of Cotswold conditions. Similarly, of the changes in livestock numbers shown in Table 4, those in sheep concerned Chalk Wiltshire almost exclusively, those in cattle concerned it very largely, those in pigs concerned the Clay Vale much more, while those in horses affected all sections of the county, although the displacement of horses certainly came earliest and has been most complete in Chalk Wiltshire.

In their heyday the arable, or hurdle, sheep had scarcely been regarded as a paying proposition in themselves. They were machines for producing and distributing manure from water-meadow and green crops, to maintain the fertility of the land which grew the cash crops, wheat and barley: machines which paid for a part only of their upkeep, with wool and store lambs. Naturally if the area of corn crops declined, the number of arable sheep required fell in proportion. Broadly this is what happened down to 1914. The disproportionate fall in wool prices, however, upset the delicate costing of manure production, and led to some displacement of over costly hurdle sheep by grass sheep which required little labour. While the hurdle sheep, the Hampshire Downs, had still been 'the almost universal breed' in 1895, several flocks of Cheviots and Blackfaces were noted in 1911. (fn. 110) Typical landlord's improvements from the eighties onwards were the provision of wire fencing for the hitherto unfenced downs, and the construction of dewponds on the downs. These were necessary both to cut down the labour cost of the hurdle sheep, and to provide for the grass sheep which remained permanently on the downs. (fn. 111) The shortage of labour and feeding stuffs in the First World War, combined with the unremunerative level of controlled prices for lambs and wool, dealt the hurdle sheep a blow from which they never recovered. The flocks of Hampshire Downs then killed off were not replaced, for the basic alteration in labour costs persisted into the peace. In their place came more grass sheep, so that in a sample sale in 1925 well over half the sheep were of this type. (fn. 112) Within a declining total of sheep the proportion of Hampshire Downs has since dwindled, until by 1956 only one or two flocks remained. (fn. 113) From 1920 the trends in corn acreage and numbers of sheep have diverged sharply, revealing the revolution in farming methods.

Table 3

Crop Acreage and Distribution 1870–4 to 1950–4 (fn. 114)

(i) Movements in crop acreages (1870–4 = 100)

(ii) Crop distribution per 1,000 cultivated acres

(iii) Crop distribution per 1,000 arable acres

(iv) Movements in crop distribution per 1,000 cultivated acres (1870–4 = 100)

(v) Movements in crop distribution per 1,000 arable acres (1870–4 = 100)

The numbers of cattle rose as those of sheep fell. Overwhelmingly the cattle have been dairy cows kept for the sale of liquid milk from the start of their widespread introduction into Chalk Wiltshire. Grazing herds, for beef production, have remained what they were in 1878, the province of a handful of exceptional farms. (fn. 115) Just as the land, as a rule, was not strong and rich enough to fatten lambs, which were sold off at late summer fairs mainly to Midland farmers, so it was not strong enough to fatten cattle, and surplus calves and barren cows were sold off in large numbers to rich grazing districts such as those in Somerset. (fn. 116) Already by 1895 milk came third after corn and sheep as a staple cash product of the Chalk district, the dairy cows having appeared in numbers not only in the Pewsey Vale, but also in the valleys which intersect the Chalk. Thus on the farms which fell into hand on the Wilton estate in Bower Chalke, Broad Chalke, and Bishopstone, milk sold to a Salisbury dairy company (fn. 117) whose owner, S. J. Taunton, was also a large tenant on the estate, was an increasingly important product in the nineties, forming over a quarter of total receipts in 1896. (fn. 118) By 1911 in the Pewsey Vale it was said that 'the production of milk for the London market [had] become almost the mainstay of the industry'. (fn. 119)

Table 4

Livestock 1870–4 to 1950–4 (fn. 120)

(i) Movements in total head of livestock (1870–4 = 100)

(ii) Livestock units per 1,000 cultivated acres

Milk production increased more rapidly between the wars, in response to growing demand, and as a necessary counterbalance to a costs-prices relationship which reduced corn-growing to a yet smaller area. With exaggeration and misplaced logic, the Milk Marketing Board was said to have been the salvation of Wiltshire farmers. (fn. 122) It was the milk production, rather than the marketing arrangements, which rescued farmers from the eclipse of corn growing. Those arrangements, if anything, favoured the more remote dairy areas which were not already within the liquid milk market, at the expense of those, like Wiltshire, which were. By 1939, on a sample of farms in the Chalk district, milk provided 46 per cent. of the total receipts, against 54.5 per cent. in the Clay Vale; in 1955 the proportions were 45.5 per cent. in the Chalk, 46 per cent. in the Cornbrash district, and 65.5 per cent. in the Clay Vale. (fn. 123)

The expansion of milk production between the First and Second World Wars was assisted by the Hosier bail system, which brought lands remote from farm steadings into the range of low-cost milking. It was widely used in the thirties, but its disadvantages of incomplete dependability led to a decline in its use after 1945, once absolute economy in outlay ceased to be paramount. Nevertheless it has left a direct descendant in the 'milk parlour', more elaborate and less frequently moved than the bail, but also mobile. (fn. 124) The growth of dairy herds brought problems of watering unknown in the days of sheep, and the provision of piped field water, often with government assistance, was one of the landlord's major improvements of the thirties. (fn. 125) By 1941 only 9 per cent. of Wiltshire holdings had no field water-supply, and 27 per cent. had a piped field supply, the county being in both respects better provided than England as a whole. (fn. 126)

Down to 1914, as the old farming system contracted, the area of green crops declined at the same pace as the area devoted to corn-growing and sheep-grazing. Since 1920, however, the decline in the one area has borne no fixed relation to the decline in the other. On the one hand turnips and mangolds, most suitable for folding by hurdle sheep, but also labour-consuming crops, have formed a dwindling proportion of all green crops, while cabbage, kale, kohlrabi, and rape, suitable for cattle feed, have formed an increasing proportion. (fn. 127) On the other hand green crops as a whole have not retained their place in crop rotations, which indicates that dairy cattle have not been integrated into a balanced, self-contained, system in the way that hurdle sheep were. Ley farming, that is the popularization, and since 1945 successful extension, of the old practice of including grasses in the rotation, has provided a partial substitute for green crops. Cattle consume the produce of the leys, either by grazing or as hay, and the general purpose of the leys, as of the folded sheep, is to conserve and enhance the fertility of the land for subsequent cash crops. Nevertheless since the twenties corngrowing has come to depend on large amounts of artificial fertilizers, superphosphates, and latterly potashes, a modern chemical rediscovery of the traditional paring and burning practice, (fn. 128) and to that extent is no longer dependent on the rest of the farming system.

A further change has been the decline and extinction of water-meadows, in retrospect linked technically with the decline of arable sheep, on the grounds that the heavier cattle were unable to graze the soft water-meadows. (fn. 129) Cows might do more damage to the channels and drains of a water-meadow than nimble-footed sheep, but this was reparable, and in their heyday water-meadows were indeed frequently grazed by cattle. Water-meadows have been abandoned for economic, not technical, reasons. Since the twenties the extra value of their earlier and more vigorous growth, and of their more frequent croppings or mowings, has not been worth their high labour cost. They have been allowed to revert to dry meadow, or to naturally flooded land, until in 1956 only one or two were maintained as curiosities, and the disused hatchwork and remnants of field channels were visible in every level valley bottom. (fn. 130)

Within a declining corn acreage the changes have been rung between the three main cereals broadly in accordance with movements in their relative prices. Thus until 1895 barley declined less than wheat, but then entered into a relative decline which became a plunge after 1922, only halted in 1940. Since then it has increased in relation to wheat, and for the first time barley acreage has exceeded wheat acreage in several years since 1949. Oats, the most prolific and the least costly crop per acre, even though the least valuable per bushel, of the three, showed their great suitability for low-cost farming. Its attraction no doubt derived from its utility as either fodder or a cash crop, as well as from its cost, and though its production was affected by the shrinking farm and urban horse demand in the inter-war years, nevertheless its decline in relation to other grains was only confirmed with the passing of the need for low-cost farming after 1945.

The erratic nature of variations in the estimates of grain yields per acre over the entire period 1885–1940, in which it is impossible to discern any long-term trend other than that of stability, confirms that the intensity of cultivation was gradually relaxed. (fn. 131) Otherwise as the grain area contracted, and production was ever more concentrated on the best corn lands, application of undiminished amounts of labour and capital per acre could not have failed to raise yields. If a constant output per acre from a declining area was produced at lower real costs per acre, the productivity of labour and capital must have risen.

Until 1939 at least mechanization was limited, and the economy in labour which had been achieved was only in part due to the substitution of machines for men, and the contraction of the area under the plough. A part represented a real increase in the efficiency of manual labour. Before 1914 not only were some of the labour-consuming elements of high farming cut out, but also Wiltshire farmers were stated to be at last learning the lesson that a smaller number of better-paid workers provided more efficient and cheaper labour than a larger number of ill-paid ones. (fn. 132)

Certainly between 1870 and 1914 the real wages of agricultural labourers improved considerably, any fall in money wages after 1878 having been generally restored by the mid-nineties, while the price level had fallen considerably. From about 1895 money rates gradually rose. In 1898 normal weekly wages were 12s. in summer and 11s. in winter, or 11s. 9d. over the year; in 1902 they were 12s. 9d. over the year; and by 1914 they were 14s. At the same time piece-work rates were increased, and taking these and the value of perquisites into account, average weekly earnings were 15s. in 1898 and 15s. 8d. in 1902. (fn. 133) There was a severe setback in the First World War, with wages lagging far behind prices, until the establishment of the Agricultural Wages Board. Before its establishment weekly wages had risen to about 20s. in the summer of 1917; thereafter they were advanced rapidly to a peak of 46s. in 1920. The pre-1914 living standard was rather more than restored, and Wiltshire ceased to be an extremely low-wage area. (fn. 134) Nevertheless not until 1921, after the price collapse had begun, had wages risen as much as the wholesale prices of farm produce since 1911–13, and not until after 1925 was the discrepancy, in favour of the worker and against the farmer, at all marked. (fn. 135) The abolition of wage-fixing machinery in 1921 was accompanied by a sudden drop in weekly wages to 30s. This rate was held until about 1932, and was followed by increases which brought it up to 38s. in 1939. (fn. 136) A sharp fall in real wages in 1922 was, in other words, made good by the general fall in prices. The inter-war period as a whole, in spite of unemployment, was one of reasonably maintained real wages, with general living standards, through such things as improved housing, rising well away from the pre-1914 level. The improvement, however, was strictly relative to the agricultural labourer's own historical position: not until after 1945 did his living standards begin to become comparable with those of other workers.

Farm mechanization, as a cost-reducing and labour-displacing process, made some headway between the wars, and by 1939 had no doubt progressed well beyond the 0.98 tractors per 1,000 cultivated acres of 1925. (fn. 137) Development, however, became rapid only after 1939, when an increase in output regardless of cost was the aim. It has continued since 1945 with the object of meeting labour shortage, that is of replacing non-existent labour, rather than of directly cutting costs. Tractors per 1,000 cultivated acres have risen from 3.8 in 1942 and 7 in 1946 to 13 in 1954. (fn. 138) The existence of subsidies makes it impossible to tell how much of this investment has been economic. That it has resulted in some increase in output per man, though possibly a not very large one, is clear.

Over the period as a whole, individual labourers have become more costly. The farmer, therefore, has employed fewer of them so that the labour element in total costs has been kept fairly stable. The 28 per cent. of total costs of production spent on labour in the Chalk district in 1955, and the 30 per cent. in 1939, are not greatly different proportions from earlier figures of 29 per cent. in 1909, 22 per cent. in 1888–93, or 27 per cent. in 1868–76 found on individual farms in the same area. The significant change has been the decline of rent from 36 per cent. of total costs in 1868–76 to 6 per cent. in 1955. Similarly, the landlord's share of the net farm product has declined from 45 per cent. in 1868–76, to 21 per cent. in 1936–8, and to 9 per cent. in 1955. (fn. 139) To some extent this decline reflects the declining importance of landlord's capital and the growing importance of tenant's capital in the shape of machinery. On the less mechanized Clay Vale farms the landlord has retained a higher proportion of the farm product. (fn. 140) To some extent the decline simply reflects the rigidity of money rents whose real value has been whittled away by inflation since 1939.

Fundamentally, however, it is a proper reflection of the radically altered relations between farmer and landlord, and of the fact that rent has fulfilled its economic function of absorbing changes in the marginal profitability of cultivation almost to the point of its own extinction. Farmers threw off the agricultural control of landowners in the eighties and nineties, and became free to farm as they pleased. (fn. 141) Many landowners, prodded by the unprofitability of owning, no less than by taxation and fears of taxation, freed themselves from their land, especially between 1919 and 1922. (fn. 142) Owner-farmers, who in 1914 had perhaps less than 10 per cent. of the county's holdings, occupied 37 per cent. of the holdings and 36 per cent. of the farmed area in 1941. (fn. 143) Even apart from those who had become owners, farmers in general stood on their own feet between the wars, with landlord's control and financial support largely dormant. Many found the effects of such independence unwelcome. It has yet to be seen whether the replacement of the old landlord by the state, as general provider and protector, will or will not lead farmers to return to systems as unsound as those of the early seventies.