Posted on

Technical descriptions of English windmills



Smock mill, standing today


Rex Wailes considered Ripple mill to have been one of the most unusual smock mills in Kent. It had a double layer of 9” weatherboards and diagonal bracing to the framing. The eight pine cant posts were given additional strength by cast iron angles at each corner of the base. They {the cant posts} were wedge-shaped and all had been renewed at some point during the mill’s working life except one. This had been whitewashed and sawn down the middle; it measured 7½” by 5”. The corners of the smock were each protected with three ounces of lead. A mound of earth had been raised to act as a stage, with short tunnels through it giving access to the doorways in the base.(1) The front cap gable had at some point been rebuilt and the breast beam drawn in “by a timber lying on a small casting which begins on the cap sheers, between the neck {of the windshaft} and the brakewheel”.(2) There were five truck wheels; four were of solid wood with iron tyres, and each roughly 18” in diameter, while the fifth, mounted on the breast beam, was smaller and wholly of cast iron.(3) The girders supporting the fantail measured 6” X 9”. The winding worm was of wood, and there was provision for hand-cranking the cap.

 The sails had a span of 66 feet, the smallest recorded by Rex Wailes in Kent. They were patents with 3ft 5” by 10” canvas shutters. There were forty-two in each sail, thirty on the drive side and twelve smaller ones, with a leading board taking up the rest of the space, on the lead side. 

 The brakewheel had been altered, the arms being shortened(4). It had six iron arms and a square hub and a wooden rim and cogs, the latter measuring 16” with 4½” pitch. It was keyed onto a 7” square cast-iron windshaft which had a special casting made to carry the tail bearing. The brake and brake lever were wood as was the hook for the latter. Onto the wooden(5) upright shaft was wedged a solid 4’ iron wallower with 5” wood cogs and a friction ring bearing on the bevel wheel on the chain barrel of the sack hoist. The upright shaft was supported by a cast-iron sprattle beam with a separate bridging box fitted. The all-wood c8ft diameter great spur wheel had six compass arms. It was originally about 2ft higher up on the shaft; fitted-up mortices indicated its former position.

 There were three pairs of underdriven stones (it is possible a fourth was fitted originally): two pairs of peaks one 4ft and the other 4ft 2”, and one burr(6). They were in octagonal cases with wood horses and hoppers. All the nuts were iron, approximately 18” in diameter, and with four arms. Two of them were raised by jack ring, rack and pinion, the ring moving up and down a slightly circular cone which had a feather cast on it to secure the nut. The fourth nut did not have this arrangement, though fitted up for it, but was held out of gear by two turned rods with plates under the stone nut cogs.

 There were an oat crusher and a grindstone on the first floor(7), and a 14 h.p. gas engine supplied auxiliary power(8).         

(1) RW 1955

(2) PD 4/1946, in HESS

(3) Ibid

(4) HESS

(5) HRH in HESS

(6) HRH and PD in HESS

(7) HRH in HESS

(8) HRH and PD in HESS


Smock mill, gone

One pair of stones was 3ft in diameter. The governors were driven by belt from the upright shaft.(1)

(1) RW in HESS


Post mill, standing today


Rolvenden mill bears the date 1772 on one of the timbers of the side frame(1). It is a good example of an eighteenth-century post mill which has remained unmodernised, probably because it did not continue working into the twentieth century. Another rare survivor, it still retains its relatively ancient character and the atmosphere that goes with same.  It stands on a slight mound within a pleasant pastoral setting. A document exists from 1556 concerning the “mill house”, and a mill is shown here on Symonson’s map of 1596. In the mound can be seen the remains of a flight of stone steps, intended to make access to the mill easier.  The tarred buck is relatively tall and narrow, and thus typical of the Kent-Sussex border region. The high curved roof and pointed breast are also characteristic of it. For windows there are small doors which can be opened to allow light to penetrate the interior. The mill was painted white during its working life, but the use of tar is excusable as it is a better preservative. There was once a lantern in the roof to accommodate the brakewheel(2), though this is not really evident today. The tail was extended at some point. The main corner posts have gunstock heads, a sign of comparative antiquity.

 I did not examine the framing on the stone floor, for reasons given below; my thanks to those of our party who did make the ascent – Duncan Breckels, David Plunkett and Peter Filby – and took photographs of the machinery there and on the bin floor, which have proved invaluable for the purposes of this survey.

 The original roundhouse, shown in old photographs, was apparently of wood, since it was taken for firewood during the Great War(3). A new wooden structure was later built to replace it but this was superseded by the current brick one as part of the  restoration – a very good one for the period – carried out in 1956 in memory of John Nicholas Barham of Hole Park, Rolvenden, son of Harold and Peggy Barham, who was killed in 1955 aged 18 in a motorbike accident in Goudhurst. His mother purchased a 99-year lease of the derelict mill from Mrs Whitehead, whose descendants the Georgetti family still own Mill Farm. The Rolvenden Windmill Trust (Registered Charity No.293203) was given the task of rebuilding and afterwards maintaining the mill in perpetuity. The restoration was carried out by the millwrights Thompsons of Alford in Lincolnshire with advice from the then Wind and Watermill Section of the Society for the Protection of Ancient Buildings. A stone plaque was placed in the northern wall of the roundhouse dedicating the restored mill to John Barham.(4) 

 Though inevitably showing its age the mill has been kept in good order ever since, with further repairs undertaken whenever necessary. The tailpole, replaced c1998 using a larch tree grown on the Hole Park Estate(5), remains in place but the ladder to the buck has not been put back, the aim being to deter unauthorised access, and a portable one is required for the purpose. At present there is also no ladder from the spout floor to the stone floor, excepting a number of battens nailed to two vertical members of the side framing. I baulked at climbing this!  Within the roundhouse, which has a tarred timber roof, one is confronted by the main post and trestle, which are clearly old but still sound; the post, along with other main timbers such as the crowntree and side girts, is of massively thick and solid construction, a factor which helps to explain the mill’s survival even though it was out of use for over seventy years from c1882 until the restoration of the 1950s. Due to its having stood with buck fixed facing in the same direction, south-west, since stopping work the stone floor framing including the upper side rail is noticeably distorted there from buffeting by the wind, as is evident both internally and externally. The buck also has a backwards lean. There are however no serious threats to the structure’s stability. Some internal strengthening with iron tie rods has been carried out over the years and four posts inserted within the roundhouse to support the buck at the corners(6). 

 The post was measured at 30 inches square at the base by Rex Wailes (1955). It bears the carved inscriptions “EW” and “H Allen 1828”.

 Regrettably there is as of yet no photograph showing the mill in full working order, but it finished its active life with at least one pair of common sails, and has been restored with four replica commons, painted white to contrast with the buck and mounted on steel stocks fitted after the 1987 hurricane during which the sails broke free and were seen turning at speed before one stock fortunately broke, thus bringing them to a standstill and preventing further damage.  New sweeps were fitted during the same restoration, one of which was subsequently replaced in c.2005. This work, as with much of the maintenance that has been required over time, was carried out by Anthony Hole of Dorothea Restoration.(7)

 There is relatively little use of iron in the machinery of the mill. The wooden windshaft, albeit with an iron poll end, carries all-wood clasp-arm brake- and tailwheels both of the same design, though as at Stocks Mill, Wittersham, the tailwheel has no separate rim, the cants being shaped to form its circumference.  The wheels are secured to the shaft with wedges and iron straps. The shaft is 21 inches in diameter(8) and squarish in section but with 16 sides.  It has mortices for a compass arm brakewheel, probably driving a single pair of stones. These are towards the rear of the shaft, indicating that it was reversed at the same time the iron canister and the clasp-arm wheels driving head- and tailstones were fitted.(9) The brake is of iron above and wood below the approximate centre line of the brakewheel, again as at Wittersham mill. The brake lever is all-wooden.

 The auxiliary machinery is difficult to interpret, some details having evidently been changed during the 1950s restoration. Wailes, who inspected the mill before that date, records that the sack hoist was driven by a friction gear in contact with the outside edge of the tailwheel, and a jog scry from the face of the wheel. A wire machine and “screen” were driven off the face of the brakewheel by a succession of three spur pinions and then by belt from the tail (presumably meaning the tailwheel). Within the apex of the roof towards the tail, their bearings located in cross-timbers between the rafters and the members of the rear gable, are two horizontal wooden shafts; one, above and to the left of the other, carries a wooden drum in the form of two discs between which are mounted a number of staves, as in a lantern pinion wallower or stone nut; a belt presumably went from this to somewhere. A section of the shaft is straked with iron as with a sack hoist bollard, so this appears to have been its function, but such is not necessarily the case and its position does not tally with Wailes’ description. It is altogether more likely to be a machine drive though where its power take-off came from isn’t clear. The other, shorter shaft has a solid bevelled wood drum which is, in fact, more or less touching the rim of the tailwheel, but it is otherwise blind. Mounted horizontally between vertical stub timbers from the rear sprattle beam, and in roughly a position where, formerly. it would have meshed with the rear projections of the tailwheel cogs is an iron-bound solid wood nut whose spindle is fitted with a wheel and handle for manual adjustment. This is in all probability one of the spur pinions which formed part of the wire machine drive; the others have evidently gone, as have the various machines themselves. It is therefore not easy to relate Wailes’ notes to what one sees today.


Both pairs of stones remain, though without furniture. The head stone nut is a finely made, solid wood affair of the “cartwheel hub” type, that for the tailstones all-iron. The latter has been displaced from its glut box and now rests in the runner stone at an angle.

 Behind the tail beam and abutting it is what appears to be a grain bin, of taller and narrower type than usual, and formed by boarding over a structure of four vertical posts.


As with many ageing post mills this has dipped, and there is a slight slope up to the post on all sides.

Left side framing: There are four uprights between the main  corner posts, and a diagonal and some intermediary timberwork between the rear main and extension corner posts. The third upright, looking from the tail towards the breast, appears to depend from the crowntree. Short members are fixed between the second, third and fourth uprights at varying heights and there is also one between the fourth upright and the corner post, from which the timber on which the head governor is mounted projects.

Right side framing: There are three uprights between the main and extension corner posts. The first is actually two uprights staggered with a diagonal between them. The second and third have battens nailed to their outer faces to form a ladder, as noted above. The main corner post bears a nicely carved date. The breast framing on this floor consists of the prick post and two diagonals; across these timbers are fixed an upper and a lower transverse beam, with further down a horizontal timber on each side of the prick post and tenoned into it. There are four uprights within each of the panels thus formed.

  On all sides there appears to be an absence of intermediary framing between the members described above, except where mentioned further short connecting timbers. The rear wall is completely without any apart from that forming the door posts and lintels; otherwise it is simply boarded over. It is possible that some alterations were made to the construction of the buck framework, among other things, during the 1956 repairs (I did not examine the stone and bin floor framing).

 The bridgetrees run fore-aft. The head bridgetree is tenoned at one end into the lower transverse beam in the breast, and at the other into a hanger from a stone bearer. The stone bearers are tenoned into the upper transverse beam. The hangers for the tail bridgetree and brayer depend from the stone bearers.

 The governor for the headstones is in the front left-hand corner of the mill; as observed, its spindle is mounted on a stub timber which in turn is fixed to a block of wood fixed between the corner post and the final upright member of the intermediary framing; this is at an angle since the two timbers do not seem to be quite in alignment with one another. This is most likely to be a product of structural distortion after the mill ceased work (the bridgetrees do not quite line up either).

 From the governor a long steelyard runs east, looking towards the tail, to a link on the transverse beam into which the forward end of the head bridgetree is tenoned. The left steelyard link appears not to go to the bridge tree but to descend through the transverse beam to another timber, parallel to the transverse beam and projecting from the side framing, through which it in turn passes to end in a handle for manual adjustment.

 The governor for the tailstones is unfortunately now missing, but appears to have been mounted on a timber fixed between the second and third upright of the left side framing, to which the steelyard runs diagonally from the aft end of the tail bridgetree. It had wooden arms and links, a feature unique to Kent where it was also found at Ash mill. The belt drums on the stone and governor spindles are solid wood flanged discs.

Based on an inspection carried out on a Mills Research Group visit on 4th October 2009. Thanks to the Rolvenden Windmill Trust and in particular Mr Edward Barham for allowing access.

(1) Jenny West, The Windmills of Kent (Skilton 1972, 1979)

(2) RW 1955

(3) W Coles Finch, Watermills and Windmills, C W Daniel 1933

(4) Mr Edward Barham 28/7/2016

(5) Ibid

(6) Ibid

(7) Ibid

(8) RW 1955

(9) Ibid


Smock mill, gone (inaccurate replica built on existing base in recent years)

As at Benenden the curb was of shot type. Here there were 24 rollers 3” diameter by 4” long. Originally they were of cast iron but were later replaced with wrought iron as this was better able to bear the weight of the five sweeps. The late Mr William B Horton said that the larger mill in Windmill Street, Hythe, had a shot curb. Shot curbs turned more easily than dead but they gave a lot of trouble. Hill of Ashford disliked and never recommended live curbs.

 The windshaft was octagonal at head and tail and very neatly fashioned. It had corrugations cast in it to assist lubrication. There were four pairs of stones, which are said to have sometimes been all driven at once.(1)

(1) RW 1955

SANDWICH, White Mill

Smock mill, standing today


Built c1760, White Mill is a particularly interesting example because of the amount of original wooden machinery which has survived, in contrast to later mills such as Cranbrook, Meopham, Margate and Willesborough where many parts were iron from the start or have been renewed in that material. It ceased work by wind in 1926 and was then engine-driven until 1953(1). Subsequently much restoration work was carried out by the legendary millwright Vincent Pargeter, and the building is now the hub of a folk museum run as the White Mill Heritage Centre. It is an octagonal smock mill, white-painted, on a single-storey brick base enclosed by a wooden building, open on one side, whose roof, supported on iron columns, serves as a stage. Inside the smock tower is plastered (this at one time included the cap(2)) and the framing painted white, except for a couple of panels on the dust floor.

 There are four double-shuttered sails with quarter-elliptic springs. The fan drives down to a large wooden winding worm on an iron spindle which carries an iron wheel with attachment for a crank for hand winding. The cap is of the broad, high-roofed type generic to smock mills of this age (the traditional Kentish variety having not at that time developed). It turns on an iron dead curb on a wooden rack. There are four truck wheels, one at each end of the lateral and longitudinal timbers of the cap frame. At the left rear of the cap is a vertical flanged wooden pulley with a rope wound around it; I am not sure of the purpose of this apparatus.

 The brake wheel has an iron spider and wooden cant and rims and is mounted on a slender iron windshaft (the slimmest I have so far seen in Kent). The solid wood bevelled wallower is partly rotted away on one side. It is not clear how it drove the sack hoist (as presumably it did), but there are holes in its underside where a friction ring would have been fixed on. The upright shaft is also of wood; it has a pot and pintle thrust bearing(3).

 The bin and dust floor are combined, with most of the space being taken up by the bins and the walkway between them. In the centre  of the floor is a sunken area inside which can be seen a flanged iron drum on a layshaft; this may be part of the secondary dresser drive but I could not ascertain what it was itself driven from. The transoms on this floor are staggered, i.e. not at the same level all the way round.

 The sack hoist is a cylindrical wooden shaft on which is a six-armed iron pulley. One end of the shaft turns in a beam fixed between two of the cant posts, the other in a casting on a horizontal timber which is hinged in a slot in a vertical post. At its other end the timber is suspended by a rope from a horizontal shaft on which is a flanged four-armed iron pulley, worked by friction by a rope which travels the width of the floor to a small roller mounted on the inward projection of one of the cogs of the rack and then down to a twist peg on one of the diagonals of the intermediate framing. Pulling on the rope turned the pulley and shaft and raised the timber to throw the sack hoist in or out of gear. The shaft of the pulley rotates at one end in a cheek piece bolted to the neighbouring cant post, and at the other in a diagonal timber descending from the rack, which is braced, by a second, horizontal timber at an angle to it, to the right hand vertical of the intermediate framing on that panel.


The great spur wheel is a wooden clasp arm, in contrast to later mills in Kent which generally had iron spurs. The lower end of the short upright shaft is mounted in a horizontal timber whose ends are attached, via cheek pieces, to two other horizontal timbers at right angles to it and spanning the width of the mill, the whole forming an H shape. In the northern of the two longitudinal timbers a semicircular recess has been cut to create space for a now bygone wheel to turn in. Running parallel to the timber from west to east and terminating in the lateral beam is a windlass for raising and lowering the stones, in the form of a long wooden roller with mortices in it where an iron bar was inserted.

 The two pairs of stones, in an east-west line on either side of the great spur wheel, are overdriven from the latter via large solid wooden stone nuts of what I call the “tambourine” type. Also engaging with this wheel is the auxiliary drive nut, moved in and out of gear by an iron lever. It is mounted on a secondary upright shaft, around which the forked end of the lever fits. On this floor the shaft is encased in trunking.

 The hopper of the eastern pair of stones is fed by a large spout, which served as a sieve, suspended from the southern of the horizontal timbers of the supporting frame for the upright shaft, and itself fed by a short spout from one of the bins on the floor above. It is not clear how the sieve was agitated.


In the centre of the floor is a third pair of stones driven by the engine. The spindle of these continues above them to terminate in a bearing on a ceiling joist.

 North of the engine stones is a flour machine mounted on a very large bin. Hangers are suspended from a beam in the ceiling running east-west and support a curved bridge tree for the spindle of the western of the two pairs of wind-driven stones on the stone floor. The governors and the spindles for the other stones are mounted on the brayers. Unusually, all the steelyards are sheathed in wood. The action of the eastern governor lowers rather than raises the collar on the spindle, requiring a weight to be hung on a timber attached to the steelyard.

 The secondary upright shaft descends, unguarded, to the floor and carries a downward-bevelled iron mortice nut, the cogs of which are missing, which engaged with a solid bevelled wood nut on a layshaft carrying three solid wood bollards of increasing thickness from left to right. From the easternmost bollard a belt goes to drive the flour machine via a flanged drum on the latter’s spindle. The end of the layshaft is located in the bridge tree for the eastern pair of wind stones, which can be adjusted to engage/disengage the dresser. There is a second drive to the machine, the origin of which is not quite clear. A belt appears to pass down through the large spout feeding the dresser from the apparatus on the bin/dust floor mentioned above, which is partly hidden by the structure of the bins; the flange visible on the spindle of this apparatus would itself have taken a belt from somewhere or other, the first stage of the process, but I couldn’t work out where. The second belt goes to a small pulley on a short layshaft which carries at its other end a toothed gear engaging with a worm on the dresser spindle.


Here is to be found the engine drive, which is the most elaborate I have seen on an English windmill. The drive belt entered the mill through a hatch in the wall of the base, going to a double pulley on a horizontal shaft which carries a further, single pulley whose purpose was unclear, before the drive is continued by a pair of large iron bevel gears and a vertical shaft. The whole apparatus is mounted within a wooden frame. On the vertical shaft is a horizontal pulley with curved arms, from which a belt goes to a corresponding pulley on the spindle of the engine stones. The spindle rests on an iron bridge tree with a wheel for adjustment by hand. A governor and steelyard are provided.

 The adjacent engine house currently contains a Lister petrol engine which is not connected to the wind machinery but can be used to drive a plate mill to grind corn.

(Description based on surveys by Guy Blythman on 13th July 2008 and 24th July 2016)

(1) West 1972, 1979

(2) R Hawksley 1947, in HESS

(3) RW 1955


Smock mill, standing today


Sarre mill, built c1820, is a fairly typical Kentish smock mill, with an octagonal tarred body standing on a two-storey brick base. Former owner Malcolm Hobbs, who carried out the restoration of the mill, believes the millwright was Holmans of Canterbury though Rob Cumming, Kentish windmill expert, disagrees. It is a well-proportioned specimen, one of the best-looking I have ever seen, and in this respect similar to Herne and Draper’s Mill, Margate. I have also heard commented on the strong similarity between this mill and that at Ripple, recently restored. Although the brick base is quite high – as at Ripple – it appears no stage was thought necessary. Probably the mill carried common sails originally and was also much lower, but the base was added or raised in height from one storey to two at the same time that patents were fitted. At present the mill carries four double-shuttered patents with a guide pole, braced to the cap by diagonal timbers, for the striking chain, and a fantail.

 The mill was extensively restored, amounting almost to a rebuild, in 1985-91 during which some modifications were made. It is unclear how much of the fabric was renewed but at least one original cant post survives, plated on each side to strengthen it.

 The cap, of typical Kentish design, turns on a dead curb which has been renewed in steel. It is braced to the tops of the cant posts on each side of each post with nice iron brackets. This bracketing occurs throughout the mill, being used to brace the transoms to the cant posts. The truck wheels, which are solid wood and of broad diameter, are mounted at the ends of the lateral member of the cap frame – the sprattle beam – and at the rear end of each of the longitudinal members, with a fifth wheel mounted on an iron bracket bolted to a block of wood projecting from the breast beam. At present the brakewheel is chained to this so it cannot move when not desired. The sheers are doubled, the lower timbers having slots in them in which the truck wheels are mounted. The diagonal timbers of the fanstage begin just behind the sprattle beam, their ends being flush with it.

 The brakewheel has a wooden rim and cants bolted to a fine 8-armed iron spider which itself is secured to the windshaft by iron wedges(1). The rim and cants must be a reconstruction as only the spider remained when Jenny West surveyed the mill for her 1972 book The Windmills of Kent. A ring of iron cogs is bolted on with a section left out so that the wheel does not engage with the wallower for the whole of its circumference, allowing the engine drive to be engaged without turning the sails. A section of the rim of the original wheel, with several of the cog mortices numbered (in Roman numerals) so that the cogs could be removed allowing the wind drive to likewise be disconnected, is preserved in the base.

 Behind the wheel the octagonal iron windshaft, said to be dated 1833 and thus probably a replacement for the original wooden shaft, tapers inward towards the tail. In front of the wheel the taper is reversed, the shaft, which here is ribbed, broadening out to finally end at the neck in a circular section. This feature is not common among Kentish windmills but is also found at Stelling Minnis and I believe one or two others.

 I was not able to enter the cap but looking up from the dust floor I could make out a bevelled 6-armed horizontal iron pulley on a vertical iron spindle, part of the fantail gearing. According to Mr Malcolm Hobbs this is of an unusual design.

 Mounted on the wooden upright shaft is a bevelled iron wallower (original, and not a new casting as has been stated elsewhere) on an iron boss with six short horns which are wedged and bolted to the upright shaft. Beneath it is a wooden friction ring for driving the sackhoist. This ring was originally in six sections, of elm, and bolted to the wallower arms. Having rotted away it was replaced in the restoration with a doubled ring in four sections.(2) It drives an iron pulley with six curved spokes on an iron spindle. Between two iron flanges the bollard is built up with a number of wooden battens. Around it is wound a chain which passes over a guide roller in a bracket on a short vertical timber fixed between the curb and the transom beneath it. The roller was put in during the restoration and is not an original feature(3). One end of the sack hoist spindle is located in a bearing on a casting on a beam hinged in a mortice in a vertical post. This beam serves as a lever for raising the hoist in and out of gear with the wallower. The other end of the spindle turns in a block of wood mounted at an angle on a transom, in the space between the right diagonal member of the intermediate framing and the adjacent cant post. The hinged beam is suspended from another beam fixed between one of the cant posts and a batten bolted between the two diagonals on the opposite panel of the smock frame, and is also supported by the vertical post. Bolted to its end, and at a right angle to it, is a lever worked by a rope which passes over several pulleys and then down to the bin floor below. The lever raises the hinged beam and throws the sack hoist in/out of gear; the sack hoist pulley turns in a groove in a horizontal timber which is tenoned into the vertical post of the sack hoist supporting frame and braced to it by a diagonal member. (M Hobbs October 2008: the horizontal timber acts as a brake when the sack hoist is out of contact with the friction ring on the wallower. It is a very simple but efficient arrangement. When the pulley is between the two, the weight of the chain returns down to the bottom of the mill. It can be stopped at whatever position required, even with a 2¼ cwt sack of grain on the chain. So one has complete control from any floor.)

 The upright shaft is square and chamfered at the top. Below it is encased in circular trunking so I could not see how many sides it had.

 There are a number (five in all) of boarded-over square openings in the floor; all except one would appear to be former sack traps, now no longer in use, for they are in two halves with a semicircular section cut away on the inner edge of each to form a hole where they meet.

 The transoms on the dust floor are at the same level, with the exception of one panel where the timber is higher to allow for a window.

 The bin floor is almost entirely devoted to the bins. On the stone floor are two pairs of stones; their alignment and the presence of a wooden governor bearing on the spout floor indicates there had formerly been three(4). The place of the third pair is now taken by a Bamfords oat roller. The stones are overdriven from an 8-armed iron great spur wheel with teeth cast integrally. The upright shaft is footed in a lateral beam between two longitudinal beams which span the mill, forming an H shape. The ends of the longitudinal beams rest on vertical posts adjacent to and within the cant posts on either side.

 There is a grindstone in one corner. The quants are very long, though this is not unusual with overdriven stones.

 The stone nuts for the surviving stones are iron mortice and of broad diameter, with four short iron arms each. The one for the engine drive is a solid all-wood affair of “tambourine” type, on a vertical iron shaft encased in wooden trunking. All these nuts are thrown in and out of gear by forked levers.

 The present governors were salvaged from the site of Barham smock mill, destroyed by fire in 1970, the originals having been donated to the restoration of Margate mill(5). There is one for each of the surviving pairs of stones. On the spindle of each governor is a 4-armed flanged iron pulley from which a belt travels to a similar pulley on the stone spindle. The governors are mounted in simple wooden frames consisting of a horizontal timber and two hangers suspended from the beam in the ceiling above.

 The bridge trees and hangers are of iron. On the vertical shaft for the engine drive is a large diameter iron pulley with two sets of five curved iron spokes. It is not clear what this pulley actually drove but below it on the shaft is a bevelled  wood-cogged iron pinion meshing with another on a layshaft which carries four further pulleys, all of iron, which decrease in diameter from east-west; the largest has six curved spokes, the two smallest four. First and second pullies: the two are so close together that you cannot tell how many arms the second has. The first and second pulleys, the former loose and the latter fixed, are very close together and a flat belt, operated from the cellar floor with a hand-operated belt shift, is moved across from one to the other to take the mill in and out of drive from the gas engine which latterly provided power here(6). On the west side is a modern milling machine with elevator, and to the north-east of it a further elevator in wooden trunking. The space at the centre of the floor is taken up by spouts and a large bin.

 Semicircular sections cut away from the longitudinal timbers bearing one of the stones indicate the former presence of a now vanished gearwheel, but the timbers, though original, are thought, probably along with other parts, to have come from another mill(7).

 The gas engine which latterly provided power here is preserved within the adjacent granary.

 A beautiful set of measured drawings of the mill, commissioned by Malcolm Hobbs, are to be seen at the Mills Archive Trust, Reading.

(Description based on visit 24th July 2008)

(1) M Hobbs October 2008

(2) M Hobbs 18th November 2008

(3) M Hobbs October 2008

(4) Ibid

(5) M Hobbs 18th November 2008

(6) M Hobbs October 2008

(7) Mr D Adams, who helped in the restoration of the mill, 2008


Smock mill, demolished 1953

The white-painted three-storey smock stood on a brick base 18” thick, 9ft high and a little less than 20ft across (outside measurement)(1). The floor space within the mill was very small(2). From the fan spindle a pair of small iron bevels transmitted the drive to an inclined shaft passing inside the cap, where a second pair of bevels drove a horizontal shaft passing to the side of the cap to carry a small iron pinion which meshed with an iron spur gear about 2ft diameter, carried on a second horizontal shaft below the first; an iron worm which meshed with the wooden teeth of the curb was mounted on this second shaft.(3) The curb was all-wood, not iron-faced like others in Kent, with wood teeth on the outside(4). The centring wheels were large(5). The very finely made brakewheel had six iron arms and wood cants, rim, brake and cogs(6). One of the spouts from the bin floor had a sieve feeding the hopper of the stone(7).

(1) Denis Sanders 11th June 1947, in HESS

(2) HESS

(3) DS, as above

(4) RW 1955

(5) HESS

(6) HESS, DS in same

(7) RW 1955


Smock mill, demolished c1950

The smock, latterly covered with tarred felt while the cap was painted white, stood on a single-storey brick base 8’ high by 8’ 3” across the octagon(1). The cant posts were wedge-shaped(2). The curb was of wood with the teeth outside(3). The cast-iron windshaft was to the pattern of Hill, the Ashford millwright, as were the cant posts, so the mill was probably built by him. In 1946 the windshaft was on the ground beside the mill and lying across it was the very heavy square casting made to fill up the opening left in the brakewheel when the shaft replaced a wooden one(4).

(1) HESS 5th August 1946

(2) PD 19th April 1946, in HESS

(3) HESS

(4) PD in HESS

Post mill, collapsed 1953


This was a white mill with a tarred(1) breast. It stood above a single-storey roundhouse with tongued and grooved boarding(2). The main post was 2ft 5” square at the base, becoming octagonal halfway to the roundhouse roof and “tapering 1ft 7in smooth round at first floor level and 1ft 3in at crowntree”. The crosstrees measured 1ft square.(3) A new main post was put in in c1885 and it was perhaps at the same time that spur gearing was put in and wooden brake- and tail wheels and bridgetrees renewed in iron(4).

 Four spring sails were carried on an iron windshaft 8½” square tapering to 7” round at tail. The all-wooden brakewheel was 10ft in diameter with a six-segment rim and gripe arms, and was mounted on a 2ft 6” square iron casting on the windshaft. The brake and brake lever were wood. The wallower and upright shaft were both iron, the former being 2ft 6” diameter and the latter 3¾ in round, and the 3’ great spur wheel iron mortice. The tailwheel, with rear facing wooden teeth, was 8ft diameter with an iron mortice rim in four segments, and six iron arms cast integrally with the hub in two sections. The whole was secured to a 7½” square boss on the windshaft. There were two pairs of stones in the breast, peaks on the left and 3ft 6in burrs on the right, and one pair of burrs in the tail. The headstones were in octagonal casings with wood horses and hoppers. The nuts were iron, 15” in diameter on 3” round spindles, and were lifted out of gear by fork and lever. The tail stone nut was also iron, and 14” in diameter on a 2½” square quant. There was one governor to each pair of stones; those for the headstones were placed behind them and against the walls, and the tail governor on the left side. The sack hoist was in the form of a 2ft 6” wood pulley, friction-driven from the top of the tailwheel, on a shaft 8” square at its mounting and 8” round beyond and extending to the tail of the mill where it carried an iron-straked wood bollard. For the drive to the dresser, a 16” wood nut meshed with the brakewheel cogs on the right, and was on a 2” square iron shaft extending laterally across the mill behind the brakewheel to terminate against the left-hand wall in a 2ft wood belt pulley.(5)

(1) HRH in HESS

(2) RW 1955

(3) HESS 5/8/1946

(4) PD in HESS

(5) HESS


Tower mill, standing today


Stanford tower mill is the sole remaining example in the county of Kent with the potential for full restoration, the majority of the machinery still being present, and as such is of particular importance. The large black mill was built in 1851 for John Hogben, as a very nice inscription carved on a ceiling timber on the spout floor testifies(1), and worked by wind until 1946. It then continued until 1969 under first an oil engine and then an electric motor(2); this probably assisted the survival of the internal plant. This being a relatively late mill, there is very heavy use of iron, sometimes less attractively than elsewhere, in the machinery, making it a good example of a modern (relatively speaking) de luxe windmill. It was recently listed Grade II*, and the present owner has been carrying out repairs with the aim of conserving its fabric and machinery and improving the appearance of the building. 

 The brick tower, which was originally tarred, has five floors. The lower section, comprised of the ground and spout floors, is cylindrical, as at St Martin’s Hill, Canterbury, and Wingham (now demolished) mills. The builder was Hill of Ashford, who was more accustomed to smock mills and copied the shape one would have, i.e. a battered tower on a base which was straight-sided if more probably octagonal than round. Rob Cumming has described the mill as “Willesborough {another Hill mill} in brick”. Hill probably also built Canterbury (St Martin’s Hill) and Wingham tower mills although there is so far no firm evidence. The “drum” has buttresses built out from it to support the horizontal platform of the stage at second-floor level. The iron brackets serving as footings for the diagonal timbers still remain, but the present stage is positioned higher. The interior is lighted by elegant iron-framed windows, for which the walls are deeply recessed on the inside, in the form of rounded arches. 

 The rounded form of the cap (basically of the usual Kent “post mill roof” shape) also identifies this as a Hill mill, since it is found at Willesborough. Here it was tarred along with the tower. It had a small hatch in its side on the left, besides that giving access to the chainwheel. There were four patent sails and a fantail, with striking by rocking lever, and a chainwheel was provided for hand winding when necessary.(3) No trace of any of this apparatus remains. The original cap roof was removed at some point, being replaced initially by an unsightly asbestos and corrugated iron version, but has been reconstructed on the original base frame timbers, which still support the iron windshaft, whose shape varies through circular at the neck bearing, then octagonal, then square at the boss on which the brakewheel is mounted, and finally octagonal again to the tail. The brakewheel, has eight flanged iron arms and a wooden rim and felloes. The (wood) cogs are missing. Braking is by a segmented wooden brake and wooden lever.

 The cap frame consists of breast beam (with a short lateral timber on top of it supporting the windshaft neck bearing), a transverse beam in front of the brakewheel, sheers, sprattle beam, and two longitudinal timbers connecting the latter to the tailbeam. The post in which the brake lever is hinged is mounted not on the right sheer, as usual, but on a longitudinal timber, resting on the three lateral timbers to the left of and parallel to it. The cap turns on an iron dead curb which is in two layers, the lower being stepped inwards and resting on a wooden ring on top of the tower. The six-spoked truck wheels, carried in large iron brackets, bear on the side face of the upper layer of the curb.  It was not possible to count them all, those on the breast and tail beams not being visible due to the construction of the present cap, but there is one on each of the ends of the timbers bracing the sprattle beam to the tailbeam, and of the short timbers projecting from the sheers on a line with the sprattle beam; the latter are bolted to the side faces of the timbers rather than their undersides. Likewise the top bearing of the upright shaft is bolted to the underside of the sprattle beam, another departure from usual practice.


This is very shallow, so that the wallower and sack hoist are at more or less knee height. The upright shaft is square (Wailes  measures it at 7 inches(4)) and of iron with four ribs. An all-iron 8-armed bevelled wallower, like the great spur wheel a fine piece of work, is fitted with a wooden friction rim to drive the iron sack hoist on the north-west side, which is in the form of a pulley with six curved spokes mounted on the flanged bollard. The pulley rests on a block of wood on the floor which is shaped so as to allow it to turn. At the wallower end the bearing is located in a timber hinged in a mortice in a vertical post at one end and at the other mounted between two posts united at their tops in a single block of wood fixed to a wooden plate in the tower wall. The other bearing turns in a vertical timber bolted directly to the wall.


This is partly missing. The upright shaft is encased here in wooden trunking one side of which has been removed, exposing a coupling in the shaft about halfway down. There are windows on the north and west sides, with a bin beneath the latter, and a sack trap in a sunken section of the floor on the south side. 


There are windows on the north, east and west sides. The stairwell is on the southwest side, with the ladder to the bin floor adjacent to it, and there is a bin on the southeast side fed from above. The bridge beam of the upright shaft is bolted to the undersides of the usual pair of long timbers, here running east-west, which together with it make up the dummy floor serving as the support frame for the shaft. On the west side a large meal spout from the bin floor spans the distance between them. The three pairs of stones are overdriven from the 8-arm iron mortice great spur wheel. Two remain, without their furniture, on the  northwest and southwest sides. They rest on raised wooden rectangular plinths with segmented skirting around the latter. The northwestern (peak) runner stone has a very large all-iron stone nut, without quant, resting on it; the nut is cast with four arms radiating from the corners of a square centre. Its own nut is of the same type; the longer, lower section of the quant, which was two-piece, has been unbolted and removed. The southwest (French) runner has two weights let into its surface, each bearing the inscription “HUGHES DOVER ROAD LONDON”. The nut for this pair is missing but is presumably the one lying on the northern stone. The quant has disappeared.

 The third pair of stones was on the north-east side. On the south-east is the engine drive stone nut, all-iron with four arms at compass points and a central boss, on a square iron quant encased for the lower half of its height on this floor in wooden trunking.

 According to Wailes the nuts were thrown out of gear using the lever and wedge method(5). 

SPOUT FLOOR (top floor of lower vertical section of tower)

There are windows on the east and west sides. The stairwell is on the southwest side with the ladder to the stone floor just to the right of it. Doors open on to the stage on the north and south sides. There is a large meal ark in the centre of the floor with a chute going to it on the left-hand side from one of the stones. The main ceiling beams run east-west and from them are suspended the iron bridge trees, running parallel to the beams with links at their eastern ends. The extension to the quant of the southwest stones remains, but without its belt drum.

 It is not clear whether there was one governor for each set of  stones. Two remain, lying on the floor; they have pear-shaped weights and four-spoked grooved iron belt drums. On the far left of the southern main ceiling beam is the curved iron bracket which formerly supported the spindle of the governor for the southwest stones. One steelyard remains, curved with a forked end where it fitted around the stone spindle; it has been removed from its former position and placed inside the meal ark. 

 On the southeast side the quant, or secondary upright shaft, for the engine drive is footed in an iron bracket, bolted to the side face of the southern main ceiling beam and stayed to the ceiling joists. An all-iron four-spoked upward-bevelled gearwheel engages on its east side with an iron-bound wood bevel nut on a cylindrical iron layshaft carrying an 8-spoked iron pulley. The neck of the layshaft turns in a bearing bolted to the iron bracket. The other end of the shaft is mounted in a timber running north off the southern main ceiling beam to a wooden hanger from the ceiling.

 There is a spout on the north side and another on the south-east adjacent to the engine drive gear.

 A new vertical post supports the northern main ceiling beam beneath the northwest pair of stones.


There are windows on the north and south sides. The door on the west side opens onto a new balcony from which the ground is reached via a short flight of stairs. This was not the original arrangement(6). The two main ceiling beams run north-south and are each supported by two equidistant vertical posts. Some of the posts are new, the others not.

Survey based on visit by G Blythman and R Cumming 27/9/2014


(2) Jenny West, “The Windmills Of Kent”, Charles Skilton 1972,   


(3) Photographs in Mills Archive

(4) “Windmills in Kent”, Transactions of the Newcomen Society             vol.29, 1953-5, p221-39)

(5) Ibid

(6) Photograph in Mills Archive


Smock mill, standing today


Davison’s mill was a late addition to the Kentish and English windmill scene, built in 1866. The last miller, Alec Davison, continued to work it until his final retirement in the autumn of 1970. It was then acquired by the County Council, although its maintenance is now the responsibility of the Stelling Minnis Windmill and Museum Trust, set up in 2010. Repairs were carried out in 2003.

 The tarred octagonal smock rests on a very shallow brick base only fourteen inches high. There is a stage, on vertical supports with the horizontal members tenoned into the cant posts at first floor level. The mill is not a large one. There are four floors: combined bin and dust floor, stone floor, spout floor and ground floor.

 The cap, which measures 11ft 6” long by 9ft 9” high(1) is of a traditional Kentish style with vertical sides for about two-thirds of its height. It turns on a dead curb. There are four double-shuttered patent sails, with striking by means of a rocking lever protruding through the rear gable of the cap, and winding is by a six-bladed fantail.

 The windshaft is of iron with a circular flange at the neck, aft of which there is a short tapered section before the poll end, as at Sarre mill. The wooden clasp-arm brakewheel, 7ft 10” diameter with eight cants and single (not doubled) arms(2) has an iron band around the rim and 76(3) iron teeth. It is mounted on an octagonal casting staked to the windshaft(4). The brake is iron. The bevelled iron wallower has a wood friction rim bearing on an iron wheel with four short spokes and a large central boss on the wooden sack hoist bollard. The hoist is mounted on the usual wooden frame, with the bollard resting in a hinged lever and the timber beneath it grooved to allow the wheel to turn. Mounted below the hoist just above the floor is a solid wood pulley on a short layshaft; I could not discern the purpose of this apparatus.

 The upright shaft is in two sections; on the dust/bin floor it is iron and circular, with a universal joint at floor level, and below it is square and wooden.

 There are two pairs of underdriven stones, one of which is dismantled with the runner placed on stands. The bedstone remains embedded in the floor. The runner bears a maker’s nameplate “LUCKINS LONDON”. The other pair of stones retains what appears to be its original furniture.

 The nice iron mortice great spur wheel, with eight ribbed arms, is 8ft in diameter, with above it on the upright shaft (instead of below which is the usual practice) a large solid flanged wooden belt drum for the governor, which has wooden arms and controls both pairs of stones. The stone nuts are large and also of iron mortice type. The bridgetrees and brayers are of iron. The upright shaft is footed in a horizontal beam between two vertical posts.

 Latterly wind power was supplemented here by a 1923 Ruston Hornsby paraffin engine, which remains in its shed beside the mill. There is a second cog ring on the underside of the rim of the great spur wheel, for the drive. This engages with a bevelled iron pinion on a layshaft carrying a large double pulley with two sets of twelve spokes. 

 The lower ends of the cant posts, and the sills, being so close to the ground they have tended to be affected by rising damp and rats, leading to their renewal in recent years.

Description based on visit by G Blythman made 17th August 2008.

(1) RW 1955

(2) Ibid

(3) Ibid

(4) Ibid


Hollow-post windpump, standing today partially collapsed

Stodmarsh windpump (TR 233604, OS Sheet no. 179), situated on the east bank of the Little Stour river the marshland around which it once drained, is a rare and remarkable survivor. I have heard it suggested that it was built in the 1850s; the 1872 Ordnance Survey marks it “in ruins” but I am not inclined to think it has been out of action for longer than the early years of the last century. The local inhabitants are unable to say how long the pump has stood for or when it was last used; indeed virtually nothing seems to known about its history, making it something of an enigma.

 Although built of wood the pump gives the impression of having survived for an astonishingly long time in its derelict condition. How much longer it is likely to do so is anyone’s guess but it is leaning at an ever more precarious angle and will soon fall into the river or the marshy ground surrounding it. Urgent action is needed to save it or at least record what remains for posterity. (In 2010 it was observed that the superstructure had collapsed into the river, but it was possible to haul it out using a winch. The main post is still standing, so far as is known (2019) and possibly reusable, being solidly built of oak.) 

 The following survey is based on a study of photographs in “The Windmills of Kent” by Jenny West, on the Windmill World website, “The Windmills of England” by R J de Little and in my own personal collection, and on a visit to the pump carried out in the spring of 1994. Due to the absence of photographs showing the pump in working order, some details must be conjectural.

 The design of the windpump is simple, indeed crude, but it is solidly constructed, and would have to be to endure the sometimes harsh environmental conditions in this remote, marshy rural spot. The superstructure, which is not enclosed, consists of two horizontal pieces of wood located one above the other – the upper being roughly equivalent to the crowntree, and the lower to collar and sheers, in a conventional full size post mill – and connected fore and aft by a vertical timber at a point halfway along their breadth. These timbers extend a short distance above the upper horizontal member and the windshaft rests between them, turning in sockets in the wood. The main post passes through circular holes in the horizontal members with, presumably, a bearing at each of the points where it does so. The upper horizontal member is located just below the top of the post where the “wallower” is.

 In being of open frame design, with no attempt to protect the machinery from the elements, Stodmarsh resembles another hollow-post pump which formerly stood at Broomfleet in the East Riding of Yorkshire. In other ways it is similar to that from Pevensey Level in Sussex, now reconstructed at the Weald and Downland Open Air Museum, Singleton, but whereas Singleton drives a lift pump through eccentrics Stodmarsh had instead an Archimedean screw, positioned at an angle to the river bank.

  At its lower end the post broadens, tapering slightly, and is chamfered. The trestle is high and the quarterbars steeply inclined and in this respect the pump resembles that at Singleton. Triangulation is achieved by a pair of “crosstrees”, horizontal timbers at right angles to and crossing over one another where the post rests on them, each connecting an opposite pair of quarterbars. The latter do not rest on the crosstrees but continue downwards to ground level, each being braced either to the ground or to some further structure at a point on its inner face near its foot by a short vertical post. Unfortunately decay and consequent displacement of timbers, along with overgrowth and the danger that physical contact may cause the structure to collapse, makes it difficult to determine the precise arrangement and in particular what the pump ultimately rests on and how it is anchored in the ground. Suspended from the lower crosstree and passing down either side of the lower bevel pinion (see below) are two vertical timbers, between which is fixed a horizontal timber in which the lower bearing of the pump shaft is footed. They seem to continue to ground level, and if footed there would have helped provide additional stability. 

 The pump is winded by a weathervane attached to the rear vertical member, consisting of a long horizontal timber going from the member just above the “crowntree” and a shorter one below, which must at one time have been attached to it but has now broken away. The basic nature of this apparatus would not be out of keeping with the overall rough simplicity of the structure but it is possible there were other components which have fallen off (and become buried in the marshy ground). The timbers appear to be angled downwards but this may only be because decay has caused them to fall slightly out of place.

 The sails are mounted on an iron cross, and in the absence of striking gear it seems likely they were spread with canvas. The cross is ribbed on the backs of the arms.

 All the gearing is of iron. On the short windshaft is a four-armed toothed wheel meshing with a four-armed bevel pinion from which the drive is transferred by an upright shaft down the hollow centre of the post to a second bevel pinion, again with four arms, just below the level of the crosstrees, which presumably engaged with a third on the shaft of the Archimedean screw. The screw today is unfortunately missing having been accidentally broken off during clearing of the surrounding reed bed and lost beneath the boggy ground. (During the 2010 inspection of the site a buried object which might be it was located using a metal detector and a hole in the ground approximately 15” deep beneath the lower bevel gear may indicate its former position. A timber was found with an aperture in it through which the shaft of the screw could have passed.) A large piece has broken out of the rim of the “brakewheel”.

The upper bevel gear has 28 teeth, the lower 36.

 There is no sign of a platform from which maintenance of the structure could be carried out; most probably if repairs were needed the person charged with the task stood on the crosstrees, which provided a platform of sorts, or used a ladder.

 The windpump is charming in its simplicity and deserves a better fate than to be allowed to collapse and rot, which is what seems likely to happen unless something is done. Unfortunately all efforts at preservation have so far failed.

 The remains of the pump were inspected by a group of enthusiasts on 30th May 2010, and the following details noted/dimensions taken:

Corner post 5ft long, 6 X 3 in section. Mortice in it is 4 inches wide and an inch and a half deep.

Windshaft 2” square, 3ft 9” long; 4” from former position of brakewheel to head.

Neck bearing 9” from head, 3” across.

Neck bearing unit 7” deep, 6” wide, top piece three and a quarter inches. Plate itself is one inch thick.

Dimensions of tail bearing unit same as for neck bearing unit.

One sail bar examined: 11” long, 1” X one and a quarter inches section.

Cross: each arm 21” long and approximately 5” wide.

Span of cross: 40 inches.

Stocks: 7ft long (could be slightly longer as a bit of the one examined may have rotted away). 3” wide and three and a half inches deep.

Mortice in stock for sail bar: one and a quarter inches long, one inch deep, one inch wide.

Prick post four and a half inches deep and 6” wide; length 5ft 5”.

Lower (chamfered) section of main post 1ft square (initial estimate by Michael Roots).

Crosstrees (paired) 8ft 4” long, 5” by 3” section.

Quarterbars 10ft 4” long, four and a half inches by 4” section.

Main post 1ft across down to segmented collar.

Main post 14” wide at bottom.

Main post 10” wide at top of crosstrees.

Main post 4ft high from segmented collar up to ring at top.

Segmented collar is 2” diameter.

Middle section of post (from segmented wooden collar to chamfered section) 68” high and 1ft thick.

Bottom (chamfered) section of post 14” high and 13” wide.

Total height of main post 128 inches (10ft 8”).

Mortice in post for quarterbar to take tenon: 8” high (approx.), 4” deep and 18” wide.

Corner post 5ft long and 6” X 3” section.

Mortice in corner post 4” wide, one and a half inches deep.

Timber in which pump shaft footed 33 and a half inches long and 5” X 4” in section. Tongue (tenon) at end of it 6 and a half inches section.

“Cap” width 40” measured from tailbeam.

Upper bevel gear 16” diameter.

Lower bevel gear 25” diameter.

Seven sail bar mortices in each stock (possibly eight; one may have rotted away in the stock examined)

 The timbers, both partly rotted away, with semicircular rebates in them are probably not the weather and tail beams as I originally thought but the boards which fitted around the post and are shown in a photograph of mine from 2008, the rebates meeting to form a socket through which the post passes. The timbers were unfortunately not measured.

 The lateral timber in which the pump was footed was tenoned into two horizontal timbers which must have been fixed between the quarterbars. These were not measured.

 One pair of stocks had a metal reinforcing plate attached to each stock in addition to the cross arm. Each stock had two bolts for fixing the cross arm.

 Rob Cumming believes the neck and tail bearing units were cast later than c1850.

Measurements were taken of the pump in the 1960s, when it was more complete, by the millwright Vincent Pargeter and are reproduced here:

POST 12” X 12” at base

     10” X 10” below collar, 10” diameter above

BODY 3’6” long, 1’6” wide

SAIL WHIP 8’ long, 3” X 3″


SAILS Drive side 1’9” wide, leading side 5” wide

SHEERS 9” X 3”

UPRIGHTS, HEAD AND TAIL: 5 and a half inches X half an inch

QUARTERBARS Four and a half inches X four and a half inches

(VP to G Blythman 29th June 1911)

I am indebted to Patricia Parr for some of the information contained in this article.