The corn mill of the late mediaeval period followed the Vitruvian pattern in which each pair of millstones was served by a separate waterwheel. At Dowrich Mill, near Crediton in Devon, this mediaeval arrangement can be seen. There are two holes for the shafts of two waterwheels, each of which served a pair of millstones; these have been lost, and have been replaced by a conventional arrangement of two pairs of millstones driven by stone nuts off a single great spur wheel and a single waterwheel. The water-driven corn mill at Barr Pool in Warwickshire, shown in an illustration in the Universal Magazine published in 1729 (Figure 4.1), shows how the Vitruvian arrangement worked in the case of the pair of millstones over the shaft. The same illustration shows a variant on the Vitruvian mill in which a second pair of millstones was driven off a lay shaft, and not by a great spur wheel. In the Barr Pool example it is clear that at the beginning of the eighteenth century the millwrights were still working entirely in wood, the only metal parts being the bearings and gudgeons.
In France, Germany and the Netherlands, the beginning of the eighteenth century saw an upsurge in the study of millwrighting and mechanical engineering. The professional millwright was becoming an engineer and he was approaching millwork design scientifically rather than empirically. In France, in 1737, Bernard Forest de Belidor produced his classic volume Architecture Hydraulique in which he showed designs for improved waterwheels. Buckets in overshot waterwheels, though still made of wood with wooden soles to the back of the bucket, were angled so that the water would flow in more smoothly, and so that the water was held in the bucket for longer, therefore giving an increased efficiency to the waterwheel. He worked out designs for all forms of floats and buckets for the waterwheels, and he improved the way in which the water was led from the mill race through hatches, or launders, on to the waterwheels. It is thought, too, that Belidor first formulated the idea that the wheel would be better if the buckets were built between the rims so that the water did not spill out at the side. He was working towards a greater efficiency in the use of water power by also improving the design of dams and water controls.
One particularly important use of water power which grew in scale in the seventeenth and eighteenth centuries was the supply of water for drinking purposes in towns. In Paris, waterworks had been erected on the Pont Neuf about 1600 and these were rebuilt by Belidor in 1714. In London, a similar series of waterwheels was built under the northern arches of London Bridge by George Sorocold about 1700, to replace an earlier set inserted by Peter Morice, a Dutch engineer, in 1582. Sorocold had been responsible for the installation of several other water-driven water supply systems in English towns, including Derby, Doncaster and Leeds. The system at Pont Neuf was known as a ‘moulin pendant’. The Seine rises and falls quite severely and so the waterwheel has to rise and fall with it. Since the moulin pendant is a stream wheel, which is turned only by the flow of the water, it is important that the floats retain the same relationship to the flow of the water at all levels of the river.
At Pont Neuf the whole body of the pumps and waterwheel was raised on four large screws as the water rose, so that the pumps could continue to work. The waterwheels at London Bridge had a slightly different set of conditions to deal with. The bridge spanned the tidal Thames and the starlings (foundations) of the bridge piers reduced the water passage to 50 per cent of the river’s width. At high water the difference in level across the width of the bridge was 25cm (1ft) and at low water 1.38m (4ft 6in). To meet these differences in level the shafts of the waterwheels moved up and down on hinged levers and the gears continued to be engaged with the pumps since they moved about the centre of the shaft on the hinged beams. Later waterwheel-driven pumps were installed at Windsor (The King’s Engine), Eton and Reading and these continued in use, in some cases, until the beginning of the twentieth century.
In Germany there were similar pumps for pumping the town water at Lüneburg, but more important examples existed to pump water for the fountains in the Nymphenburg gardens, near Munich. The idea of the water supply of formal gardens being raised from nearby rivers was developed to its fullest extent in the Machine de Marly, built about 1680 to supply water to the gardens and fountains of Versailles. Fourteen waterwheels were built below a dam on the River Seine and the water was brought on to these wheels through separate mill races. These undershot waterwheels were 11m (36ft) in diameter and 1.4m (4ft 6in) wide. In addition to 64 pumps adjacent to the wheels, connecting rods took a further drive a distance of 195m (600ft) to a second group of 49 pumps which lifted the water a further 57m (175ft) above the first reservoir. In all, the fourteen waterwheels operated 221 pumps and lifted the water 163m (502ft) above the level of the river.
While the waterwheels of the period, such as those at Marly, appear to have behaved well, they were clearly cumbersome and inefficient. The millwrights were able to build large wheels with high falls, such as those in the mountainous metal-mining areas, but these were still empirical in design. The designs shown in the books of this period, such as Jacob Leupold’s Schauplatz der MeuhlenBau-Kunst published in Leipzig in 1735, bridge the gap between the apprentice system of training millwrights and the scholarly, scientific approach which came later in the eighteenth century. Leupold’s book shows, by means of copperplate engravings, exactly how water-driven mills could be built. The plans are very accurately set out, with useable scales, so that the millwright could build his mill.
The associated explanations, to be read with key letters on the plans, explain every step which has to be taken. The illustrations show grain mills, ‘panster’ mills with rising and falling wheels, mills with horizontal waterwheels, boat mills, paper mills, oil mills, fulling mills and saw mills. They are a design guide to every conceivable form of mill which the millwright could be asked to construct. There are tables showing how the lantern gears and pinions should be set out, so that the millwright could almost work with the book propped up in front of him. There are other German text-books of a similar character which produce even greater detail for the millwright. A good example is the text-book on water-driven saw mills Von der Wasser-Muehlen und von dem inwendigen Werke der Schneide-Muehlen by Andreas Kaovenhofer, which was published in Riga in 1770. This book details all the joints and fastenings required in a waterwheel, for example, and as in Leupold’s book details of dam and watercourse construction are also included.
In France, the great encyclopaedia of Diderot, with its associated eleven volumes of plates, was published between 1751 and 1772. These plates, like Leupold’s, showed the methods of construction and manufacture of every trade. Thus, in the chapter on the making of black powder or gunpowder, the nineteen plates show not only the various machines required and the stages to be undertaken, but also the way in which that machinery was driven. While the purchase of a set of Diderot volumes would have been beyond the purse of a master craftsman, enough copies would exist in manor houses and stately homes for these plates to have been seen, and used, by the millwrights of the locality. Indeed, it may well have been that the landowner as client would show such books to his millwright. Modern understanding, based on mass communication and transportation, finds it hard to realize how much craftsmen moved about, and equally how the intelligent gentry absorbed everything they could see and find on their travels abroad or on their ‘Grand Tours’ with their tutors. If they had a mechanical bent they would follow this up in the workshops and libraries of the countries they visited.