Posted on

Watermills and Windmills of Middlesex (Second Edition)


Generally speaking, the development of the windmill in Middlesex followed the same pattern as in other parts of England. Windmills first appeared in Europe during the twelfth century; the earliest reference to one in Middlesex, at Harrow, occurs in 1236. It is possible that a mill existed at Clerkenwell as early as 1134, but the evidence is far from conclusive.

 The mediaeval windmills would mostly have been “post” mills – those in which the boxlike main structure is mounted on bearings on top of a post which serves as a pivot, around which it can be turned by means of a beam called a tailpole projecting from the rear of the superstructure through the access ladder, to keep the sails facing into the wind. The post was surrounded and supported by a trestle structure, originally exposed but from the eighteenth century usually enclosed within a circular building called a roundhouse, which served as extra storage space for grain waiting to be ground. The other two main types of windmill, the smock mill and the tower mill, did not appear until much later. In both a stationary tower (of wood in a smock mill, so called because of a fancied resemblance to the garment once worn by rural labourers, and of brick or stone in a tower mill) contains the machinery and millstones and only the roof section or “cap”, whose base supports the axle or windshaft that carries the sails, is rotatable.

 Judging from those depicted in contemporary carvings and manuscripts, the early post mills would have been very small with room for only one pair of millstones (later post mills might have two or three pairs, one in the head and one in the tail, or two side-by-side in the front or breast with perhaps a third in the tail), and not big enough to accommodate more than one person.  Being at the same time constructed of wood, they must have been quite fragile. Like the watermills they were initially all manorial mills, owned by the lord of the manor, who took a share of the financial proceeds and wouldn’t allow corn to be ground anywhere else, and operated by a miller who was usually one of his tenants.

 As a technical novelty, largely untried, the windmill was regarded with some suspicion by contemporaries. It was perceived as a supplement to the watermill, which had an ancestry in England going back to at least Saxon times, rather than a replacement for it. As researches have shown, it tended to be built in areas where there was already an established watermill; the intention was that if the latter were to become unusable for any reason there would be a windmill in fairly close proximity which could take over from it until it was working again.  Unfortunately this supplementary role created problems – it meant that the mill might not be used, or receive maintenance, very often and this gave decay a chance to creep in. In addition, as has been mentioned above the mills were fragile structures, easily damaged or destroyed by the elements. The chequered career of that at Ealing illustrates their vulnerability. Another fate which might be suffered by Middlesex windmills in this period was to be pulled down by the poor for firewood in times of hardship. This was feared likely to happen at Pinner, though largely because the mill was in a decayed state.

 These factors made upkeep a costly business, which may have deterred prospective patrons. Repairs had to be postponed and sometimes seem not to have been carried out at all. One gets the feeling that in many cases the lord grew tired of a troublesome responsibility and simply left the mill to fall to pieces. In addition to the general problems that windmills faced at this stage of their development, a mill’s chances of survival could be further jeopardised if it were built at a time of economic hardship, like that at Upper Halliford, Shepperton, in the 1380s. At the other end of the scale, particular local circumstances at Harrow gave the windmill an advantage it would not otherwise have had, and it was able to hold its own against the watermill and maybe even supersede it.

 In time, the windmill became an accepted part of the rural economy. The crucial factor ensuring its success in the long run was population growth, which increased the number of mouths to be fed and made it more essential that where there was no source of water in reasonable proximity to be harnessed for the grinding of corn a windmill should be built instead.

 Nevertheless it is possible to say that in Middlesex the windmill was never as common or successful, compared to the watermill, as in say East Anglia, Sussex or Kent. Although just over 100 windmills are known to have existed in the county between the twelfth century and the twentieth, the total number standing at any one time is unlikely to have been more than 30. The reason for this was the dominance of the watermill, made possible by the existence of the Thames along with lesser rivers such as the Lea and Brent. The distribution of windmill sites throughout the county was fairly even, as the map shows; Middlesex is generally low-lying, with few natural obstacles to the free flow of wind and thus excellent terrain for windmills which are not normally built on hills, as many people think; although this would give them a plentiful supply of wind, it also means they are more exposed to the elements, with the consequent likelihood that repairs will be more frequently needed. There were fewer geographical restrictions on their location than with the watermills, which could only be built where there was an adequate source of water. 

 The period of the Industrial Revolution saw windmills undergo considerable, if gradual, technological change. Tower and smock mills became common and some were much larger than many post mills, Brentford being a good example, with room for two, three or even four pairs of stones and consequently able to produce a greater amount of flour. Due to the the mills being larger and taller a platform known as a gallery or stage was often built out from the supporting brickwork to give access to common and spring sails for reefing the former or adjusting the shutters of the latter. Where cost permitted wooden components were replaced by iron ones, making for greater durability and efficiency. Certain other technical refinements were also made. Common sails – the earliest type, in which the wooden framework was covered with canvas which performed a similar function to the sail of a ship – might be replaced with spring or patent sails where a number of shutters were mounted in the framework, these being opened or closed as necessary to lessen wind resistance, by means of springs in the former and a system of self-adjusting weights in the latter. Many mills, like Enfield and Willesden (Shoot-Up Hill) in the area of our study had one pair of spring or patent sails and one pair of commons as a cost-saving exercise.

 Instead of being turned by hand via a tailpole or a rope or chain – a laborious task – the caps of smock and tower mills were often “winded” by a device called a fantail, patented by Edmund Lee in the 1740s and resembling a miniature windmill. When the wind changed direction it tended to strike the fantail at an angle, causing it to rotate and through gears meshing with a toothed rack on top of the tower turn the cap and sails to face the wind. The fantail was usually located at the rear of the cap, in the position previously occupied by the manual gear. A few post mills also had fantails, usually mounted either on the end of the tailpole or on the access ladder.

 In a windmill the drive to the stones is from above, in that the sails and main gearing are at the top of the structure and the stones further down – the reverse, in effect, of the layout found in watermills. Arrangement of the drive to the stones differed between post mills and smock/tower mills. In post mills it could take one of several forms. (1) There might be one pair of stones in the front or breast of the mill driven from a brake wheel – so called because a brake could be applied to it to stop the mill – mounted on the windshaft, and a second pair in the rear driven from a tail wheel. (2) Two pairs of stones side by side in the breast were driven from the brake wheel via a wallower and great spur wheel, and a third in the tail by the tail wheel via a stone nut. This method was a comparatively late innovation. (3) Some mills had two pairs only, positioned in the breast and driven from the brakewheel via the wallower and spur wheel, and no tail wheel. We do not know which arrangement tended to be used most in Middlesex. 

 In smock and tower mills there was usually a wallower, upright shaft and great spur wheel driving the stones through stone nuts. Ancillary equipment was driven via intermediary gearing from the brakewheel or tailwheel in a post mill and either the great spur wheel or a crownwheel on the upright shaft in a smock or tower mill.  

Initially the effect of economic progress is not so much to create new technologies as to make the existing ones more efficient. Although the first large steam-powered flour mills began to appear towards the end of the eighteenth century, in most parts of England windmills and watermills, with improvements, continued to be built for another 100 years, some remaining in use well into the twentieth century. The comparatively early decline of the windmill in Middlesex was due not to technological change but to the growth of London. As the nation’s capital expanded in size during the eighteenth and nineteenth centuries it absorbed parts of neighbouring counties – Essex and Kent in the east, Surrey in the south, Hertfordshire in the north and Middlesex in the west. It was Middlesex which bore the brunt of this expansion, to the extent of ceasing altogether to be a separate entity. An entire county, largely rural in character as most counties were originally, disappeared under bricks and concrete and at the same time lost its identity sociologically and politically. There has been nothing comparable to it in English administrative or social history. Today it tends to be forgotten that most of what is now London north of the Thames, including the one-time seat of government of the British Empire at Westminster, was once Middlesex.

 As farmland was sold off to developers and the rural working class moved to the expanding towns to take up jobs in factories and workshops, the way of life of which the windmill was essentially a part disappeared. Now obsolete, it was swept away by the new housing and pushed further and further back towards the borders of the county, which urbanisation had not yet touched and where the march of progress was slower. It did not disappear from the capital entirely; although windmills for grinding corn were now more or less a thing of the past, there appeared an assortment of mills for other purposes, some erected on the roofs of other buildings in order to catch the wind of which the new development would otherwise have robbed them. The builders were often foreign engineers or carpenters in a city which had always been highly cosmopolitan. As late as 1892 a windmill was erected on the roof of a building in Great Eastern Street, Shoreditch, to drive machinery within it for the production of flour. However these examples all appear to have been short-lived.

 By 1850 there remained at the most half-a-dozen mills, most of them in the rural part of the county. In time these too succumbed, either to continuing urbanisation or the advance of new sources of power such as steam, oil and gas, which were becoming cheaper to use and therefore more widespread. Hounslow and Edmonton mills worked latterly in partnership with, or as supplements to, steam-powered mills, which helped to prolong their useful lives. The last mills to cease grinding were those at Enfield and Edmonton; the former may well have carried on into the early 1900s.

 The following table shows how the number of windmills in the county changed over the 800 or so years covered by this book:

C12              none

C13                 7

C14                14

C15                 1 (what explains this drop?)

C16                 6

C17                26

C18 (up to 1780)   17

C19                25

C20                 2

Total number of recorded sites = 98

 The figures given are for the number of new mills built and do not take into account existing ones. They show increases during the thirteenth and fourteenth centuries and from the latter part of the eighteenth century. The second increase may be regarded as part of a national trend.

 Whether the windmill has, apart from one or two relatively minor remains, entirely disappeared from the Middlesex scene is actually a matter for conjecture as it’s sometimes hard to draw the line between what is a windmill and what isn’t. In the 1950s a correspondent of Martin T Mason’s wrote: “It is not quite correct to suggest that all forms of windmills have ceased to operate in London. There are smallish propeller-like machines working at times on the roofs of certain tall buildings in Central London, one of which is the Cumberland Hotel at Marble Arch. They work electric dynamos for emergency current. They are, nevertheless, London windmills! Their coming {is} due to the post-war power cuts.”(1) This finds an echo in the recent proliferation everywhere of wind turbines, some huge and some very small, as concerns over climate change lead us to at least experiment with natural, non-polluting methods of generating electricity. How successful they will prove in the long run remains to be seen. They are expensive and not always reliable – much like their traditional counterparts, in fact, except in the sense that they are considered visually intrusive in a way the old windmills certainly are not.


The windmills of one area tended to differ to some extent from those of others in their architectural and technical features. It is not easy to say what the typical Middlesex windmill – if there ever was such a thing – looked like. The problem applies to internal features as well as external. There are no interior photographs of Middlesex windmills in existence, nor do any  sketches or plans survive showing their workings. About their internal arrangements we know only what we can glean from sale notices. So far there is nothing to suggest they differed in any way from those of windmills elsewhere.

 Here again the problem is the early disappearance of most of the mills. One should also take into account that trends in windmill design were really the creation of individual millwrights or their firms, and in the case of Middlesex we have even less information about these than we do the mills. We know the names of some of them and can occasionally connect them with particular mills, but their blueprints, account books etc have not survived.

 In order to identify the main design trends among a county’s windmills, one really needs a good selection of photographs of different mills. Although it is not quite true to say that the camera never lies, photographs are generally a more accurate record of what a mill looked like during its working life than paintings or sketches. The latter, being subject to artistic licence, are to be treated with caution – the artist would sometimes get the mill’s proportions wrong, or give it features it never actually possessed – although it is possible to regard some as more authentic than others.

 At present we have photographs of six windmills, excluding contemporary ones of the windpump base at Clerkenwell and the possible windmill tower in Crane Park, Twickenham, and reliable non-photographic illustrations of several others. The photographs are of Edmonton, Enfield, Hampton, Hounslow, Monken Hadley and Willesden mills. That of Hampton has been heavily retouched, although I believe it nonetheless to be authentic, while Monken Hadley is seen as little more than an outline on the horizon.  Sunbury, Southall (smock mill), Acton and Pinner mills would also have survived long enough for photographs to have been taken, though so far none have come to light. It is also possible that there may be photos of some of the Millwall mills and that at Stoke Newington.

 This is not quite enough to give us a representative sample.  Another problem is the tendency of mills near the borders of a county to be similar to those in the adjoining areas of neighbouring ones, the regional trend becoming more marked as one gets closer to the centre. Not only had the majority of windmills in Middlesex disappeared before the advent of photography in the 1840s and 1850s, due to the expansion of London, but those which remained were in these border areas where regional differences were blurred. Edmonton and Enfield mills, for example, had certain features also found in the Essex, Bedfordshire and Hertfordshire region. We cannot be sure that the mills of which we have photographs are really typical of Middlesex as a whole. Kilburn mill stood fairly near the centre of the county, but without many other mills to contrast it with we cannot say whether it was typical.

 There is one feature of the smock mills which may have been characteristic. Four – Acton, Enfield, Hampton and Sunbury – are known to have had little or no brickwork beneath the wooden tower. It was usually the custom for smock mills to rest on a brick or stone base of at least one storey. It is true that in Enfield’s case this feature can be explained by the mill’s closeness to areas where it is known to have occurred, but nevertheless the presence of this characteristic in three mills, each a reasonable distance from one another, does suggest that it may have been a regular feature of smock mills within the county. The similarity in cap shape between Hampton and Willesden smock mills might also indicate a regional peculiarity.

 It is possible that the “typical Middlesex windmill” after which we have here been chasing is little more than a wild goose. It should be borne in mind that there were never very many windmills in the county at any one point in its history, which may have tended against the appearance of a particular Middlesex millwrighting style. Taken by itself the pictorial evidence suggests the mills were varied in character, and I suspect this was indeed the case. Several, such as the White Lead Mills at Islington, the Hoxton mill which copied them, and perhaps the post mill at Hounslow, were of a style clearly imported from another part of the country. In London itself, where the growth of that vast, cosmopolitan city had obliterated the traditional character and way of life of rural Middlesex, there seems to have arisen an assortment of mills which conformed to no standard pattern, and in some cases could be quite bizarre in appearance. As noted above they frequently had to be erected on the roofs of other structures whose height would have obstructed their wind. All the farmland having of course been built over, they tended to be for purposes other than grinding corn. Pumping, for example at Chelsea, Clerkenwell, Soho and Tyburn, was perhaps the most popular but “industrial” uses such as the production of white lead – whether by conventional millstones or edge runner stones is unknown – as at Islington and Millbank, or sawing wood as at Limehouse, were not uncommon on either bank of the Thames, and at least two mills operated ventilation systems at prisons and hospitals.

(1) Mr C V M Smart, date of letter unknown