For more than a century, Bennett and Elton’s four-volume work has been the main source for the history of corn milling (2). (Note, in line with Bennett and Elton and common English practice, throughout this article corn milling refers to the milling of wheat). More than 100 years later, a comprehensive and up-to-date review of the topic was published online by Watts and Watts (19). Both works provide useful background for this article, which is written from the perspective of a millstone miller with more than 30 years of experience (7).
Peacock (16) reviewed evidence that stones have been used to crush and grind seeds, plants, and minerals for many thousands of years. Millstones were first introduced around 2,000 years ago, and they dominated grain milling until, as David (9) wrote, “with the invention of the roller mill…the long search for a means of producing uniformly white flour was over.”
Detailed accounts of the transition from rubbing stones to rotary devices such as millstones, which enabled the use of animal power in milling, have been published. Rubbing stones crush the grain between the stones, whereas the dressed faces of rotating millstones cut the grain (15,18). Early rotary mills were first driven by animal power and then by waterpower during Roman times. Wind-driven mills were not introduced until more than a thousand years later (Fig. 1).
With all of these milling methods, the resultant product requires further refinement to separate the bran from other components. In contrast, a central principle of roller milling is the gradual reduction of grain by breaking it open and reducing the particle size between successive pairs of rollers (4). At every stage of the process some flour is produced and extracted to obtain the maximum amount of white (refined) flour. The efficiency of the roller milling process ensured the demise of millstone milling, particularly in urban areas where there was a demand for high volumes of flour. Because rollers were essential for grinding harder wheat varieties, the move away from millstones was very rapid. In addition, the introduction of steam power enabled rollers to handle the huge increase in flour production required to meet growing demand.
Muscle Power, the Romans, and Waterpower
Traditionally, grinding grain using a saddle quern was almost always the responsibility of women. This type of quern was named after the characteristic flat or dish-shaped lower stone, and its use spread west from the Near East, reaching Britain ca. 4000 B.C. (Fig. 2). The saddle quern became the main means of grinding grain until the introduction of the rotary quern ca. 400 B.C. Rotary querns grind grain between two stacked circular stones; a projecting handle is used to rotate the upper stone above the lower stone. The grain is fed into a central “eye” through the upper stone, making the process more efficient and easier to use than a saddle quern. The rotary quern remained in use in Ireland and the Scottish isles well into the twentieth century (18).
The use of querns for milling required a lot of effort, and they could only supply enough flour for a small household. Bennett and Elton (2) note that when slavery was abolished in Rome, the motive power (energy) was provided by free men, and “slave mills” were renamed “servant mills.” The introduction of animal power enabled the use of larger millstones, with a corresponding increase in efficiency and output. Animals were first used to turn millstones during classical times in Greece and Italy. The earliest and best known form of animal-powered corn mill is the hourglass type, which is sometimes referred to as the Pompeiian mill (third century B.C.). It was turned by a donkey or horse and was introduced into Britain by the Romans during the first century A.D. (Fig. 3).
Horse-powered mills were particularly useful in towns and castles (8), where a steady water supply could be problematic. Remnants can still be found on farms, but these mills were never as commonly used as water mills or windmills. Their numbers did increase, however, due to labor shortages following the Black Death during the fourteenth century. Water mills, which are more powerful and efficient, were the dominant mills used in rural areas by that time.
The earliest water-powered mills are thought to date from the middle of the third century B.C., at about the same time as use of the Pompeiian animal-powered mill was becoming more widespread around the central Mediterranean region. The earliest Romano-British millstones had sloping grinding faces. The upper stones were concave, and the lower stones were convex, and they sometimes had furrow dressings laid out in a manner similar to those found on more modern stones. Later Roman millstones were larger in diameter and had flatter grinding surfaces, a development also seen in quernstones. Grain was fed through a central eye in the upper, rotating “runner” stone to be milled between it and the lower, stationary “bedstone.
Dressing and Design of Millstones
The grinding surface of a millstone is divided by furrows into separate flat areas called “lands.” From the furrows, smaller “feathering” or “stitching” grooves on the lands provide further cutting edges and help to channel the ground meal out from the stones. An experienced millstone dresser is able to create more stiches to the inch, which improves the quality of the flour. The furrows and lands are arranged in repeating patterns called “harps.” Typically, a millstone will have 6, 8, or 10 harps, with the same pattern being used on both the runner and the bedstone. When the millstones are laid face to face, the patterns create a scissoring effect, progressively cutting the grain more finely as it moves outward between the stones. Well-dressed stones remove the bran from the endosperm in large flakes that are easier to sieve out, enabling a whiter flour to be extracted from the resulting meal.
The skill to dress stones, traditionally performed by itinerant stone dressers, was an important tool in the skill set of the traditional miller. I used the traditional wooden thrift and pick, with up to 15 steel mill bills, for a full dress. These are clearly shown in Figure 4: in the photo, I am holding the mill thrift, a turned wooden handle usually of ash, in which the mill bill or picks are held and wedged. The traditional mill bill is a hard steel, double-ended wedge-shaped chisel, although before I retired I moved to using bills with harder tungsten carbide tips, which meant I could do the entire process with just one mill bill. I took the stones apart every three months to redo the stitching and once a year performed a full dress of the lands and furrows (7).
The ways in which millstones were used evolved steadily for two millennia, with improvements such as the balance rynd being introduced. The balance rynd is a two-armed, flexible support for the runner stone, comprising an iron bar or “bridge” secured across the eye of the stone, that enables the upper millstone to be hung on the spindle more easily than with the older, fixed rynd. Millstone dressing machines were introduced in the nineteenth century. These machines were perhaps essential in very large mills with 16 or more pairs of stones, but are anathema to me. I am saddened by the progressive loss of craft skills in traditional corn milling, which now provides less than 0.05% of the 4 million tonnes of flour produced per year in the United Kingdom (N. Jones and J. Cook, personal communication, from report to the U.K. government, Traditional Cornmillers’ Guild [https://tcmg.org.uk]).
Efficient use of millstones became a significant issue, partly due to the needs of the British Admiralty during the Napoleonic Wars of the early nineteenth century. To meet these needs a more systematic approach was developed to ensure flour quality. These quality standards demanded careful control of grain purchases and drying, as well as milling and grading (14). Improvements were made to both the grinding performance and the output of millstones, as well as to ancillary machinery used for cleaning the grain prior to milling and for dressing the ground meal to produce white flour, the demand for which was rising steadily. Key issues encountered with millstones were the milling rate (clearly profit related) and the heat generated in the meal by friction during milling. The introduction of millstone ventilation solved the latter problem, but at the price of generating copious amounts of dust.
A millstone ventilation system, similar to a French design, was installed by Potto Brown at Houghton Mill, Cambridgeshire, U.K. The mill ran 16 pairs of millstones, and Brown was described as “a slow grinder,” constantly conducting experiments with a view to improving the manufacture of flour. He carefully selected the best millstones, paying top prices, and he preferred to add more millstones rather than move to faster grinding. Brown believed that “the art of milling consisted in making good flour from inferior samples of wheat.” He relied on his experiments to aid in adopting the best machinery and selecting most judicious mixture of wheat (1).
The millstone ventilation system was patented by George Hinton Bovill in 1846, and his various patents were subject to many law suits (14). One patent mandated that the bedstone should rotate under a stationary upper stone as part of a system to collect the large amounts of dust in a “stive chamber.” Bovill’s process was formally tested in official trials by the British Admiralty, which concluded that the process ground the grain twice as fast. The millstones also could grind four times more grain before they needed dressing. There was no stive or dust released into the mill, and the process produced a better quality of flour because the meal was 20 degrees Fahrenheit cooler (14). The latter point was significant because meal from unventilated millstones must be allowed to cool before it can be sent through a bolter or dressing machine to separate the bran and other fractions from the white flour. If the meal is not cooled, the wire or silk mesh used in the separation rapidly becomes blocked by the warm flour.
Wind Power and Use of Gravity
The growth both in population and crop yields led to a corresponding need for increased grain processing output; however, at the same time, there was increasing competition from other industries, particularly the textile industry, for water-powered sites. Consequently, the later eighteenth and early nineteenth centuries saw a notable rise in the number of windmills operating in Britain, clearly mapped out for Kent by Pelham (17).
Traditional millers were well used to manually handling large sacks of grain and flour, but both windmills and water mills adopted wind- or water-powered sack hoists to raise sacks of grain. Various types of hoists were developed after the first written description was published by Diderot in 1765 (5). Although I was used to handling heavy sacks, the sack hoist made the process much easier, to the extent that a fellow miller even suggested we could only manage by using a sack hoist. I am sure all millers would have welcomed the ability to harness the motive power of the mill to raise the grain to the top of the mill, either by sack hoist or elevator.
The flour obtained directly from the millstones (the meal) provided the basis for wholemeal bread, but the increasing demand for white (refined flour) bread required the meal to be “dressed,” separating the ground product into white flour, semolina (i.e., farina in the United States), and bran. In a typical two-story water mill, bags of grain would be raised to the upper floor, and if a grain cleaner was employed, the cleaned wheat had to be raised again to the floor above the millstones so it could be tipped into the hopper to feed the process. Finally, the meal had to be raised a third time to be fed into the dressing machine (Fig. 5).
In contrast, a windmill tower offers the height needed for gravity to power a direct feed from the cleaner to the millstones and, finally, to the dressing machine on the floors below, reducing the labor involved and increasing efficiency. The arrangement of the first “American-style” windmill in Germany, constructed ca. 1836 in Breslau, was described in detail by Kozmin (13) (Fig. 6). The city was renamed Wroclaw and became part of Poland after the Second World War.
As development of water mills and windmills progressed, it became more essential to warn the miller when the grain in the hopper was about to run out. An audible warning reduced the prospect of two millstones, one revolving at 120 rpm, coming into contact as the supply of grain ran out, possibly causing a fire. Various types of bell alarms were developed (5).
Other improvements to windmills were made to counter the effects of changes in wind force as well as direction. For example, Edmund Lee’s 1745 patent for a fantail, although flawed, was eventually developed further to allow windmills to be turned automatically into the wind (3). The design of sails was much improved, both in efficiency and operation, and the introduction of the governor, probably in the 1760s, allowed automatic regulation of the gap between the stones, enabling the consistency of the ground product to be maintained automatically regardless of wind speed (Fig. 7).
A governor, generally a centrifugal-style regulator, detects the speed of the machinery and is used to apply its output to automatically adjust the gap between the millstones to maintain consistent product quality despite any changes in speed that might occur. Governors have also been used to control the shutters and, thereby, the speed of windmill sails, and, similarly, the water flow and, thereby, the speed of the waterwheel.
The Advantages of French Millstones
The relative merits and uses of millstones obtained from different sources have been described in detail by Collins (6). In medieval Europe grains such as barley, oats, rye, maize, millet, and buckwheat were most commonly grown and were milled satisfactorily using granite “Peak” stones from Derbyshire or lava stones from the Rhineland introduced by the Romans. When wheat began to dominate grain crops grown in the mid-eighteenth century, millers needed harder and more durable millstones that did not lose their cutting edges as easily. In the early nineteenth century English millers specializing in milling wheat began importing French “Burr” stones from La Ferté-sous-Jouarre near Paris (Fig. 8).
French Burr stones had a harder, roughened surface, which when dressed was able to break the grain more finely. Because they were harder and denser than English or German stones, dressing could be more precise, and the stones could be set closer together. Burr stones lasted up to four times longer, justifying prices that were up to four times higher than those for the best English stones. They dominated traditional wheat milling in the latter half of the nineteenth century as harder North American wheat varieties started to replace softer English varieties.
During the Napoleonic Wars the supply to England of French Burr stones from La Ferté was problematic, and the London Society for the Encouragement of Arts Manufactures and Commerce offered a gold medal or £100 to any person who discovered a quarry in Britain that could provide millstones good enough to replace French Burr stones (8).
Most British mills had at least one pair of French stones, sometimes in combination with a mixed pair, such as a French runner stone working on a Welsh (conglomerate) bedstone, and a pair of Millstone Grit Peaks or greys for general milling, such as for animal feed or coarser flours.
Practical Limitations of Millstones
Even the best dressed millstones still suffered from the need for a separate stage to remove the bran from the meal and the need to allow freshly ground meal to cool before it could be mechanically sifted. The need for cooling and sifting steps made milling with millstones a batch process, increasing the labor involved and reducing efficiency.
The flour bolter was introduced in the sixteenth century to mechanically separate the flour from the bran by beating it through a rotating cylinder of cloth. First wool was used for the cloth, then calico, and later silk. Flour bolters only became common in mills during the eighteenth century, replacing a hand-powered machine used for sifting.
In 1765 John Milne, a wire worker from Manchester, took out a patent for a flour-dressing machine. The machine encompassed an inclined cylindrical drum covered with different sizes of wire mesh that enabled several grades of flour to be extracted (Fig. 9). By the mid-nineteenth century, reels—long, hexagonal frames covered with fine silk cloth for sifting fine flour and wire mesh for separating coarser products—were coming into use in British mills. In Scotland, Wales, and parts of upland England, specialized machinery for preparing pearl or pot barley and for shelling and grinding oats was also developed.
In the United States a shortage of skilled labor was a driver of innovation, and Oliver Evans, an American inventor, engineer, and businessman, experimented with various processes to make milling less labor-intensive (11). Moving wheat from the bottom to the top of the mill to begin the process was the most onerous task in contemporary mills, so Evans’ first innovation was a bucket elevator to facilitate this process. He modified Roman techniques for moving water so that, with careful engineering, he could build a series of bucket elevators around a mill to move grain and flour from one process to the next.
He also improved the labor-intensive task of spreading warm, moist meal from the millstones so it could cool and dry. He replaced manual shoveling of meal across large floors with a “hopper boy.” The hopper gathered meal from a bucket elevator and, using a mechanical rake revolving around the floor, spread it evenly on the drying floor. The process was designed so the rake slowly moved the flour toward central chutes from which it could then be sifted.
The publication of Evans’ detailed designs for an “automatic” mill (10) caused a revolution in American milling. This concept allowed mills to be built on industrial scales with far greater efficiency. Before the introduction of roller milling, the production and maintenance of millstones were key economic issues because they were extraordinarily expensive. Sources of millstone supply from within the United States were essential, and Hockensmith (12), with his 20 years of research, made an invaluable contribution to our knowledge of milling history, particularly in the United States.
The automated milling revolution was even more far-reaching in Europe, where the “American system” was quickly adopted by the milling industry and triggered major increases in food production that were sorely needed during a period of almost continual warfare at the turn of the eighteenth century.
The world’s first flour factory was probably the Albion Mill built by the Thames at Blackfriars Bridge in London in 1780. The machinery, designed by John Rennie, comprised two steam engines driving 20 pairs of millstones through cast-iron gearing. Although a comparatively short-lived venture (it burnt to the ground in 1791), Albion Mill marked the beginning of the shift in the structure and location of the milling industry toward large mills at ports and urban centers on navigable waterways, and in doing so signaled the beginning of the demise of the small country corn mill.
In tandem with improvements in waterwheel and windmill sail technology came the development of steam power. Introduced and developed because of competition for water resources and the ready availability of transportation and coal in major industrial towns and ports, steam power gained an additional boost when it became clear that neither wind nor water could provide sufficient power to grind harder wheat varieties (17).
Experiments in Switzerland resulted in the development of the world’s first successful roller mill in ca. 1830. Gradually, the manufacture of rolls improved sufficiently to threaten the previous dominance of millstones. Emblematic of this transition, Radford’s Albert Mill in Liverpool replaced all of its stones with rollers in 1870, making it the first complete roller mill without stones in the United Kingdom (Fig. 10). There were no roller mill installations in the United States, so Albert Mill produced the first roller-made flour to cross the Atlantic (2).
As the technology and efficiency of roller systems rapidly improved, flour milling underwent a revolution and advanced at a pace comparable to today’s digital technology, with machines being rapidly superseded by newer versions. Millers refitted their mills and, in many cases, erected new buildings designed to take advantage of steam power and endure the stresses created by the banks of roller machines and associated equipment working together. Large new mills also were attractive because they could be built at ports, near railway lines, and on navigable rivers and canals, locations well-placed to receive deliveries of imported wheat.
Typical of the large new roller flour mills was the Westminster Bridge Mill (Fig. 11). At one time the mill had worked with 18 pairs of millstones; by 1885, 11 of these pairs had been thrown out and replaced by a roller plant. The new roller mill, powered by steam, employed 14 horizontal double roller mills installed by WR Dell & Son of Mark Lane, London.
This transformation resulted in a modern version of the Evans concept of grain milling, but without millstones. After two millennia, millstones were discarded because the gradual reduction method employed by the new roller mills was better suited to milling hard wheats from North America, Russia, Australia, and India. They were also able to extract a greater proportion of fine white flour, which by then was much in demand.
A century of decline in the fabric and function of traditional wind- and water-powered mills has been followed by a reprieve for some mills with the growth of the health food industry and the revival of natural, wholesome foods bringing with it a demand for wholemeal bread. Many small country corn mills have been repaired to working order, some on a small-scale commercial basis, and are producing a wide range of stone-ground flours, including those milled from organically grown grains. Perhaps this is the future of milling, with some larger firms producing stone-ground flours in addition to their roller-milled flours.
Mildred Cookson was a traditional stone miller at Mapledurham water mill in Oxfordshire for 30 years. She maintained and ran the mill commercially, milling about 40 tonnes each year. She is a founding trustee of the Mills Archive Trust, with particular responsibility for developing their roller flour mill initiatives. She is a trustee of the Society for the Protection of Ancient Buildings and has served twice as the chair of the SPAB Mills Section.