Posted on

Technical descriptions of English windmills


HALSTEAD, Greenstead Green

Smock mill, gone


{According to 1832 and 1833 sale notices the mill had} brickwork 3ft thick in the 10ft high base, an internal diameter tapering from 22ft below to 14ft at the curb, and a smock frame “35ft upon the stand”. There were six floors. The patent sails, 33ft in length, were automatically turned to wind and drove two pairs of French stones through a “cast iron shaft and nut”. Two more pairs of stones could be fitted on the stone floor if required. {The mill had a domed cap with gallery; there was no stage}.

 Rex Wailes records a few details of the machinery. The brake wheel, with iron segment teeth, had an iron brake, the only example noted by him in Essex. The spur gear, with iron hub and arms, but a wooden rim, was enclosed in a cog box on the first floor; it drove one pair of stones overdrift through an idler gear, and two pairs underdrift on the second floor. The former was doubtless a late-added pair. As at Frost’s Mill, the engine drive was to a bevel ring below the great spur gear. R Hawksley additionally noted that a two-part wooden upright shaft with an iron mortice wallower had been employed. The “cast iron shaft and nut” of the 1832 sale notice probably referred to the windshaft and wallower.

Frost’s Mill

Smock mill, base survives


External photographs abound, stressing the harmonious proportions of this mill, which, seen in repose with four double-shuttered sails over a slender body, conveys an impression of pent-up energy. Upright shaft had a winged gudgeon. Domed cap with gallery and finial, winded by fantail. C1867 a steam mill was erected and its drive applied to a ring built on the underside of the great spur wheel in the windmill.

 The windmill was about 45ft high to the finial set on the domed cap, had an 8-bladed fantail, and stood on a single-storied brick base measuring 11ft from ground to sill level and 17ft 6in across the octagonal flat faces internally. The curb had an inside diameter estimated at 14ft. The brickwork of the base is stepped at five courses outside, recalling the smock bases at Bulmer and Southminster. Above ground level there were three floors: stone at 12ft, forming the first floor, bin at 10ft higher, and dust at 9ft above that, leaving 6ft to the curb. There were three pairs of underdriven stones, two French burr and one peak, operated by two pairs of anti-clockwise double-shuttered patent sails transmitting power through an iron mortice brake wheel, an iron wallower, a two-part upright shaft, iron above and wood below, an iron mortice great spur wheel made in two parts by Whitmore and Binyon, and  all-iron stone nuts bearing on iron bridgetrees. The iron wind and other shafts and wheels were sold as scrap metal after removal, but some of the timbers were retained, including the eight curb segments. These last form an interesting contrast to a survivor of the re-towering of 1820. At that time eight discarded elm curb segments were built into a house in Halstead as lintels over the windows; some 150 years later one of them was removed for a reconstruction – and was brought back to its original home. It has large skew mortices for former wood cogs of 4½ in pitch and 3in face to take the worm drive. The later set, in oak, have iron segment teeth of 2½ in pitch and face bolted onto the outside, there being no old cog mortices. Both rested over the cant posts centrally, but were secured to each other by different methods. In the older system, wooden bridge pieces were housed in and bolted from below, while the “modern” segments were brought tightly into a ring by 11in handrail (concealed) bolts with threaded ends tapered to aid location of the nuts, special mortices being cut to allow the nuts to be introduced and coaxed and tapped into position. These segments were also bridged with iron straps. They may represent a third generation of curb construction dating from the tailwinding episode of 1867, as they appear rather sophisticated for 1820. The curb was live, the rollers running on a plain flat track; the inside track for the centring wheels appears also to have been a flat strap of iron without flange.

 The floor beams, transoms and intermediate studs and diagonals were arranged according to the best millwrighting practice, both diagonals and transoms having staggered joints at points of convergence to avoid undue weakening of the main beams; the pairs of floor beams ran in different directions at their various levels, locking the structure into a rigid unit. Four horizontal beams, in two pairs, formed an open “box” centred on the great spur wheel, and carried the vertical pine beams for the bridgetrees.

 The steam drive for the windmill was by rope across the yard. The driven pulley outside the base was mounted on a large iron bracket shaped to the canted angle of a corner of the brickwork – still in situ – a small but attractive example of bespoke ironfounding so characteristic of later mill work. From this support the shaft ran to the underside of the great spur, having initially brought in its train the inevitable butchery of existing timbers and the improvisation of existing supports.


Tower mill, gone


A water colour of the windmill shows a red brick tower, with domed cap, four double-shuttered sails and a fantail. This painting is complemented by several photographs in one of which the fantail is seen to be 8-bladed and right-handed, the sails being right-handed. On both these photographs appears the wooden shed giving cover to the portable steam engine.


Smock mill, gone


In about 1900 the canvas shutters were removed from the sail frames. In April 1940 some fine interior photographs were taken by Donald W Muggeridge. The mill had a boat-shaped cap, but with a ridge falling to the rear, whence projected a fan stage with a 7-bladed fly above. The cap was covered with galvanised iron sheets about 2ft by 3ft, and the upper part of the body with corrugated iron; this for long impervious covering went far to explain the survival of the mill some fifty years after being left to hold itself together as best it could.

 The four double-shuttered anti-clockwise patent sails, each with eight bays, were struck from an external wheel below the fanstage operating through a pinion and rack. The windshaft was of iron, the brake of wood. The fan drive appears to have been via the initial two bevel pinions to a long inclined shaft, which entered the cap to transmit the drive via two more bevel pinions to a horizontal shaft with worm gear. This turned a worm ring horizontally and so the final spur pinion directly below in mesh with the rack outside the curb. The wooden upright shaft had evidently carried a compass-arm wallower, later replaced by an iron gear. On the floor below was a solid wooden spur gear converted to a down-turned face gear; this drove a wooden layshaft and gear with a belt drum attached. The three pairs of stones were underdrift and enclosed in octagonal vats. The stone nuts were raised out of gear off a taper by the use of forked iron levers and chains. The great spur wheel was an iron mortice wheel, probably not that originally used.

 Stanley Freese refers to the “large frame as in watermills” which carried the bridgetrees, probably footed at ground level, as at Bulmer, to bear much of the weight of the stone assemblies. The last miller thought most of the timberwork to be pine, as indeed were the stone floor beams as seen in a photograph. That the miller kept a grindstone at both ground and curb levels testifies to the length of the cant posts {the brick base was very shallow, little more than a plinth}.


Post mill, gone


The mill last worked in 1915, using spring sails and two pairs of French burrs of 4ft 6in diameter, overdriven side by side in the breast. There was no bolter, oat crusher or fantail. The mill had an all-wooden windshaft, with a clasp-arm brakewheel and no tailwheel, features drawn clearly in a sketch by Thomas Hennell, in which the sack hoist drive is seen to be from a built-up wooden ring towards the rear of the shaft. Graham Wilson, the restorer-miller of Over Mill, Cambs, took several “interior” shots of the machinery at a late date (1952). These show the stone nuts, spur wheel and wallower, all solid wood-planked in two discs, and the spur most unusually mounted on the iron upright shaft above the wallower. The mill roof was over-sailed above the straight-boarded body, and the lower weatherboarding was trimmed closely to the roundhouse roof. There were jowled corner posts to stress the pre-1800 date. The mill turned clockwise and had at least one oak middling, in its last working years, passing through the wooden poll.

 Donald Smith comments on the very short windshaft without tail wheel, and states that the mill body was originally only 15½ ft long by 11½ ft wide, and that a 2½ ft section was added at the rear to take the bolter. Whether the mill was originally designed for one or two pairs of stones in the breast cannot now be resolved, but the likelihood is that there were two. Reference should be made to the account of the new-built post mill of 1765 at Belchamp Otten, in the ownership of William Stammers of Sudbury, which had indirect drive to two pairs in the head.

Lamb Lane, tower mill



The Metsons followed the Eleys, last working the mill by wind in 1916 when Arthur Metson was using also an auxiliary 8 hp steam engine. The pairs of 4ft 6in and 4ft 10in French burrs were underdriven on the second floor, there being five floors in all, including ground level. There was no stage, but a neatly railed gallery ran round the ogee cap, and there was an ample fan staging. The 8-bladed fan turned four large double-shuttered clockwise patent sails to wind, which, according to one photograph, had three sails of ten bays and one of eight, probably with some variation in the number of shutters per bay. The petticoat was strikingly finished with a scalloped dog-tooth design, as often formerly seen on railway platform shelters.


Post mill, gone


In 1871 the mill was employed for grist production, using patent sails, which in an undated painting are represented as four anti-clockwise single-shuttered sails with their striking gear external to the mill. These details are confirmed in the single known photograph, which gives a rear view and shows a “blister” in the weatherboarding on the roof to accommodate either the brakewheel or the sack-raising gear.


Smock mill, gone


The four double-shuttered anti-clockwise patent sails and the 6-bladed fantail were taken down in c1923; enquiries show 1952 to be the year when the elegantly proportioned white-painted smock frame was truncated and roofed over at just above mid-stone floor level {the stump was destroyed by fire in 1985}. The mill had an ogee cap and a gallery of Whitmore pattern, probably dating from the tailwinding of 1863. Altogether there were six floors, and the four pairs of stones, three of 4ft and one of 4ft 2in diameter, were underdriven on the fourth floor reckoned from ground level. It was a “modern” mill, having wind and upright shafts, brake wheel, wallower and great spur, also the stone nuts, cast in iron, though the bridgetrees were wooden. When the steam drive was used the brakewheel was disconnected, probably by removing slip cogs, and the spur wheel and upright shaft were made to revolve, as at Stock tower mill, driving such stone nuts as were put in gear.

 Limited mill remains can have the virtue of concentrating attention on features which might have passed unnoticed in a complete windmill in full regalia. At Great Holland there are two features of particular interest in the surviving lower frame; the sawn-down cant posts and the mode of support for the former four pairs of stones and machinery. The cant posts, like all other main timbers now seen, are in pine; they have been halved in the radial sense and bolted together without intermediate packing. At their feet are bolted heavy cast-iron webbed brackets, suitably angled to rest upon the sills and against the inclined sides of the posts, which have one bracket on each side. The posts vary in section, but approximate to 11in square. The transoms and intermediate verticals are slender for their length, being 6in square. The brick base on which this frame rests measures 21ft across the flats internally (23ft 6in inside the weatherboarding at sill level), and the height of the present structure is about 34ft, or approximately half that formerly attained by the ball finial.

 On the second or spout floor, now the uppermost which is complete, crossed diagonals have been added in the five windowless wall panels, no doubt strained and spiked into place; they are also bolted to the intermediate verticals. One supposes they were added to offset distortion of the tower by twisting. The basic framing already included normal diagonals. The floor beams alternate in alignment, being north-south for the first and stone floors and east-west for the spout floor. Beneath the stone floor are its two main supporting beams, 12in square on 7ft 4in centres, and parallel to these run two more north-south beams of similar section on 4ft 8in centres forming a “box” with them, wholly or partly within which ran the great spur gear and nuts. The support for the two sets of beams, having 18in of clear space in the vertical sense between them, was elaborately contrived to spread the load and transmit it either to the brick base or directly to the ground. The lower box beams were supported at mid-length by the bearer for the upright shaft, 5ft 8in long by 13½ in sq, itself supported by a pine stanchion centrally placed beneath, nearly 10in square, and extending down to the ground floor about 20ft below. This would have carried the thrust of the upright shaft, spur and wallower, and much of the weight of the runner stones, stone nuts and spindles. At their ends, the lower box beams sat on the transoms and lay against the inside faces of the cant posts, resting also on horizontal members bolted direct across the adjacent cant posts carrying the beams. These horizontal members were themselves further sustained by shaped, solid wooden blocks screwed into the cant posts beneath. The lower box beams were eventually made to contribute near their ends to the support of the upper pair as described below.

 The upper box (or floor) beams were so spaced as to come exactly in line with two opposing pairs of cant posts – those against the inside faces of which the lower beams were set. Their initial method of support was by abutment in an angled cut in the rear face of the cant post. The inner face of the beam (i.e. that facing the parallel member) overlapped and projected past the cant post by a width of 2in only at its narrowest part, but widening towards the outer end to accommodate itself to the cant post shape, continuing thus for the full depth of the beam to a point just short of the weatherboarding. This vertical “half-dovetail” rested on a transom and was single-bolted to the cant post. Of the four such original joists, only one has been left unaided; the others have been supplemented by a system of wooden bridge pieces, 6ft long by 4 X 5in, which run between intermediate verticals and sit on the lower box beams. Substantial upright supports, 16-18 in high, run between the upper box beams and the bridge pieces where they cross. Below the spout floor, its comparatively light-duty beams are given added support by offset iron brackets on the cant posts, not unlike those which anchor cant post to sill; otherwise they bear upon suitably placed transoms and are attached to the cant posts.


Post mill, gone


Photographs show the mill to have been attractively designed with three windows at the rear and a porch with a decorative finish over the spout floor door. A sack slide ran down the left side of the tail ladder. The four single-shuttered anti-clockwise spring patent sails were winded by tailpole and regulated from inside the mill body, which topped a substantial roundhouse.


Post mill, some remains survive


Photographs of the mill show a substantial structure of C19 type. The stringers of the external ladder entered unusually at the top of the rear door, and the roundhouse had a mansard roof and loading door built in over crosstree level. There were four single-shuttered anti-clockwise patent sails and tailpole winding. The mill body was taken down in 1917, leaving the roundhouse to be converted into a workroom. This in turn fell into disrepair and in 1972, when the property lay empty, the exposed brick piers and crosstrees were festooned with ivy. The oak crosstrees represent an odd pair; the lower one carries chamfers and thickens towards the centre to give a large abutment to the main post, while the upper member has neither of these features, but has had one end renewed and held by an iron-strapped scarf joint. The brick piers were tall enough to give adequate head clearance to the miller below the crosstrees.


Rebuilt post mill, standing today


Farries 2: The Anatomy of a Post Mill: Mill Green, Ingatestone

 The detailed analysis of the construction of Mill Green windmill set out below was prompted by several opportune circumstances. The mill was rebuilt in 1759 by Lord Petre. In 1959-60, on the occasion of the second and modern rebuild by Mr R F Collinson, the owner, an opportunity was presented to the author to measure in detail the old and discarded timbers comprising the whole of the mill body. A record of all the major and many of the minor joints was made, and a model of the framing, capable of being dismantled and reassembled, was built to a 1 to 8 scale in oak, which task compelled a close examination of the three-dimensional relationships, the purpose of each timber, and the problems of erection. As was to be expected the mill showed certain peculiarities, probably stemming from the idiosyncrasies of the millwright, and contained some overweight or even superfluous timber which has been faithfully replaced in the recent construction. The substructure from the post downwards was sound enough to remain, so that precise details of the joints could not be fully ascertained, but these plainly depart little form normal practice, as their visible form and a little probing suggest.

 It will be noted that {in 1759} certain work was contracted out to specialists, as by time-honoured custom still observed, and old materials were probably salvaged for use where possible, to reduce costs. It appears that the mill was rebuilt on the same site as then – and now – occupied, for work was done on the roundhouse, which the account shows was repaired and not new-built. The two crosstrees have each had a substantial end section replaced and scarfed onto the remaining sound timber. In the terminology used by Cecil Hewett these scarfs would be inescapably categorised as “stop-splayed scarfs with under-squinted square butts and transverse key”. This repair feature probably dates from 1759, as carbon dating might confirm. The form of key or wooden peg used used would give tremendous compression to the notched butts, and the joints, encompassed by two iron straps, still hold firmly. The crosstrees stand entirely within 3ft 4in of ground level, as is the case at Ashdon; this suggests an earlier date for the substructure than that for most post mills still in being, in which the crosstrees are often found to be above head height to give ample storage and freedom of movement, as at High Easter. The roundhouse at Mill Green was added between the sides of the existing brick piers, though whether or not immediately after the erection of the mill is difficult to establish, there being no obvious difference in the character of the bricks. The wall is not bonded into the piers, which project a little outside and carry wooden louvers at the crosstree ends for ventilation. Behind these the roundhouse wall is shaped into small embrasures, being splayed over the templates. The roof slopes up to a vertical collar, about 1ft high, boarded across two horizontal rings of scarfed timbers as formerly used at Billericay; this gives good weather protection to the upper ends of the quarterbars.

 Most of the superstructure is in oak, but the post is thought to be of sweet chestnut. The crosstrees are about 24ft long, and in section 12in wide by 11in deep for some 8ft from their ends; both then thicken out at top and sides to give a section of 15in wide by 13in deep from which to derive abutments to the centre post which is itself cut suitably into four horns projecting downwards. The thickened sections of the crosstrees have an overall length of nearly 7ft and imply the tedium of dressing down three surfaces over the remaining 17ft from each crosstree. The lower crosstree is naturally bowed downwards to the centre, a feature seen at Birch mill also. At their intersection at the post the crosstrees have no cuts on their undersides, but are rebated both on the side and upper faces. The side cuts are 2in deep and the top cuts descend to a depth of 4in on the lower and 3in on the upper member. There is a clearance of 1in between the crosstrees as they pass under the post, and presumably a small gap, not visible, above the upper crosstree, following accepted practice. If the mill body were permitted to bear via the post on the centre of the crosstree, thereby reducing the reaction from the ends of those members, it would lose stability; Jesse Wightman, in his boyhood, was told by old millers that there must not be any weight on the crosstrees at the centre, nor the crosstrees touch one another where they cross, otherwise the post would “wring off at the waist”.

 The main post is 25in square at the base and extends upwards through nearly 23ft. It is steadied by the crosstree abutments, as determined by the rebates noted above, which fit snugly against the horns of the post, no wedges being used. The whole substructure is very accurately framed. Should an adjustment have been necessary to bring the post to the vertical, it could have been effected by jacking up one crosstree, or three if need be, and compensating with thicker templates or with mortar applied beneath them. The oak templates laid over the piers are about 3in thick; those on which the bowed crosstree is laid are necessarily tapered and the brickwork follows suit to a slight extent.

 The quarterbars approximate to 10in wide by 13in deep and are let into the post and crosstrees as illustrated. At the foot the normal birdsmouth joint is somewhat modified. At each end of each quarterbar the hidden tenons have a locating function, holding the members to their task of transmitting the great weight above to the crosstrees and piers via the haunches of the joints, and in this the full width of the quarterbars was in part employed. The crosstrees have the function of ties in addition to locators for the post base; they hold the quarterbars to their work, which members, as an added assurance, carry retaining irons to hold each to its crosstree housing. The iron straps also offset a tendency for the horizontal component of the thrust from the quarterbar to split the crowntree end; at Mill Green the crosstrees were made to extend 18in beyond the joint. In the course of time it was possible for visible or hidden rot to attack the joint, progressively reducing its strength, and in some post mills pairs of wooden clamps were eventually bolted through on both flanks of the two mating timbers, giving a triple assembly resembling a limb in splints to keep the mill in being; such was to be seen at Woolpit, Suffolk, and is present in Ashdon mill today.

 At Mill Green a four-piece iron collar is set round the post just over the upper crosstree, above which the post has very large shakes on its west face. Although the depth of the quarterbar mortices into the main post was not checked, a figure of 6in is to be expected, and since there were four such mortices at the same level the post was considerably weakened. Here we have one among several of the reasons for the magnitude of the large oak butt normally selected for a main post. Not a few posts snapped at this point of weakness, examples in Essex being Hatfield Broad Oak (1881) and Waltham Holy Cross (1911). The members of the substructure are suitably chamfered on the arrises for visual effect and to spare the miller and his sacks from sharp impacts. The chamfers descending the post are ornamentally stopped. The templates on the piers are also chamfered.

 The tops of the quarterbars are made flat at 8in over their lower point of entry into the post to form a crude table over which was formerly set the collar, missing at Mill Green in 1959 and therefore not present in the mill today. The collar normally consisted of four oak pieces of section 8in to 10in square; one pair having double tendons at each end, and the other the corresponding mortises, the joints being fixed by dowels, or by pegs inserted into the projecting tongues of the tenons. There were other modes of assembly of this part, which in Essex was known as the girdle. Over the collar rode the sheers and two transverse linking members in the same plane, shaped to the circular section of the post at this level; this assembly we may refer to as the “live collar”. At Mill Green the underside of this tells an interesting story. The rear transverse piece is there heavily scored in a circular sense, and the marks are continued for a few inches on the undersides of the sheers, but are entirely lacking on the forward surfaces. It is clear that the mill was tailsick, bearing hard down upon the rear of the collar. This may account for the late absence of that member, since its removal may have eased the task of winding. Another set of score marks confirm – if confirmation were needed – that the mill took most of its wind from the SW quarter, towards which it has faced probably for the last 75 years {this was written in the late 1970s}. Each quarterbar shoulder close to the main post is scored by the passage of ropes or chains let down through the sack traps, but those on the NE side are channelled more deeply than the rest.

 At Mill Green there are three iron bands, each 2½ in wide by ¾in thick, within 10in of the top of the post. Another, stouter, band, driven down over the taper for some distance until very tight, was applied to prevent further splitting, for a major shake has developed running downwards and measuring 1¾ in wide at the circumference and 8in deep.

 Bolted up to the underside of the crowntree is an iron plate 3ft long and occupying the full width of the timber; from this an iron pintle projects downwards into an iron pot sunk into the top of the post and there secured by four wings, 1½ in wide by 9in deep, cast integrally, and encircled by the three above-mentioned bands.

 The old crowntree, now preserved in the roundhouse, has no cavity for an upward projecting pintle. The plate bolted up to the crowntree turned over a metal ring cast round the pot and heavily greased. The nature of these castings, the absence of a pintle trimmed out of the main post timber, and of the corresponding socket in the crowntree, indicate the much later replacement of the crowntree which Barker presumably fixed, new or re-used, in 1759. This represents an interesting repair operation in which the usual pintle-to-socket relationship between post and crowntree was reversed very effectively. The now discarded but preserved crowntree (1959) is in one piece, and quite sound, except for the ends, where exposure during dereliction made the timber unusable for its normal purpose. The plate by which the crowntree was shod had two transverse flanges let into the timber to hold it captive (as repeated in 1959); as seen today it has towards each extremity one large central bolt hole and some smaller ones, the larger probably intended for the use of iron ties running down to the sheers, if required, since they are appropriately spaced. As the reused sheers now present have no drilled holes in corresponding positions, one might conjecture that their replacement had also at some time been made, which is supported by their very good condition and the almost total lack of scoring at collar level. The real or imagined permutations of structural changes as inferred from visible evidence can become most complex, and must in this direction be left at the point where the very probable grades into the merely possible.

 Without a collar below the sheers, the whole weight of the mill would be borne by the top of the centre post; with a collar over which the sheers rode in part, some weight would be transferred directly to the quarterbars, and sagging of the sheers discouraged. Further, the sheers would afford some support to the forward and rear framing, including the weatherbeam and thus the windshaft, whereas in a mill without a collar the sheers would hang entirely from the structure above. In Mill Green mill as at present constituted, the sheers are held by bolts from the three transverse members resting in part over them and set between the three pairs of corner posts. In the context of sheer-collar relationship, the reversal of the notching of the forward sill and of the meal beam above into the forward corner posts is instructive. The collar could well be fitted last to suit the gap between the sheers and quarterbars, and was generally so constructed as to render this an easy operation. With greater support the mill would be partly insured against head or tail sickness, but whether balanced or not the collarless mill would be likely to respond more readily to a push on the tailpole.

 The Mill Green sheers, 20½ feet by 12in square, on 3ft centres, were tied laterally with three cross members in all, of section roughly equal to that of the sheers, apart from the members housed over them. Two were shaped to follow the curvature of the post, about which they rotated as a steady bearing or live collar, as described above. They are sometimes named “woodwears”. One transverse member fully at sheer level, and behind the sack traps, was mortised on its rear side to take the inner end of the tailpole. The sheers and the three members tenoned in between them could have been assembled off the mill and lowered over the tapered post before the addition of the crowntree.

 The old side girts, of section 19 X 10in, of length 15ft exclusive of the tenons, and 16ft 3in overall, were locked over the ends of the crowntree using lap-dovetails with over-squinted shoulders. From the centre line of the joint to the forward corner post (rear face) was 6ft 7in; to the rear, 8ft 5in., a disposition aiming for the ultimate balance of the fully equipped mill about the crowntree. No lateral movement of the girts was possible, and these members would need to be lowered vertically onto their housings. The side-girt housings into the corner posts bore a refinement not noted elsewhere and not reproduced in 1959. The full 19 in depth of the girt was shouldered into the corner posts and suitably angled to accept the downward thrust, and was notched at the top to prevent withdrawal. To unite the girt and post it was necessary to tilt the post over the girt and hook it into position before bringing to the right angle, and the tenon, which extended to a depth of 6in in the post, was cut short at the bottom to give clearance for this operation.    Three strong pegs held the joint rigid, probably subjected to the “draw-boring” technique, whereby the holes in the tenon were slightly off-centre from those in the post to force the post against the girt shoulders.

 The front and side frames require separate consideration before priorities of assembly of the body are reviewed. The front frame, apart from the two corner posts, consists of four transverse members, each differently housed into the posts, a central prow or prick post, and four diagonals converging upwards into the prick post in pairs to triangulate the structure and prevent distortion. The transverse members measured 9ft 6in between the inside faces of the posts, and the forward sill, meal and weather beams were shaped out forward to the prow to the extent of 7in at centre, “streamlining” the body in the traditional style to release the “expended” air. These members were also shaped forwards on their rear faces by some 3in. The forward sill, 12in wide by 7in deep at centre, was housed to a depth of 1¾ in over dovetails formed in the upper face of the sheers but wholly concealed, the dovetail ends being stopped at 1¾ in behind the forward edge of the sill. At its ends, this beam was notched full width on its upper face into the posts, in addition to offering tenons so angled as to enter the posts at right angles. It therefore gave support to the posts in conjunction with the side girts. The second transverse beam in ascending order, the meal or forward stone beam, on 8ft 9in centres above the sill, was supported by the frame, being notched full width on the under face into the corner posts and tenoned in the vertical sense. Its main purpose was to afford support to the front stone bearers, and it was therefore set opposite the crowntree, which performed a similar function on its rear and front faces where support pieces were applied. Deferring observations on the third cross member until later, we come to the fourth and uppermost tie beam which was the weatherbeam. This was housed over the top side rails by lap-dovetails similar to those at the ends of the crowntree and the front of the sheers, but not identical. The corner posts were tenoned into the top side rails. Since the main and heavy duty of the weatherbeam (13½ in wide by 10in deep and 11ft overall) was to carry the windshaft and sails, it was given additional support at centre by the prick post, which was tenoned in. The prick post, 6½ in wide by 6in deep, was halved into the forward sill, below which it extended to carry the weatherboarding well down over the roundhouse roof, and was supported from the sheers via the sill. It was also halved at the meal beam. The forward shaping of corner posts and weatherbeam allowed the prick post to support the latter to better effect, with a margin of 1-2 in on the forward side of the soffit.

 To advance the position of the weatherbeam, the corner posts were thickened out in front from their basic depth of 8½ in fore-aft to 12in at the top; this was developed through the upper 5ft, giving the “gunstock head” feature seen in many post mills and other timber structures as an alternative to wooden knees. The main rear corner posts at Mill Green, hung on the side girts, were and are not so thickened, but of section 11in fore-aft by 8½ in throughout, while for the rear (false) corner posts a section of 7in X 5in suffices. The forward placing of the weatherbeam was sustained in some mills – Thaxted Fox and Hounds for instance – by short diagonal supports from the corner posts; in many this over-sailing was barely present. To trim the gunstock head by removal of unwanted thickness from the remaining 18½ ft of a post would have been a laborious and wasteful exercise, and a naturally shaped piece was doubtless sought for the purpose. An inverted butt might approximate closely to requirements. Thanks to this practice the side rails would be better bedded at the weakened section where they received the front corner post tenon from below and the dovetail of the weatherbeam from above. The ends of the rails could be conveniently boarded against the weather. In cases where the rails were set on unthickened posts and continued beyond to leave body at the joint, their ends were often left exposed outside the cladding. At Moreton, forward-facing gunstock heads on the front corner posts were also present, while at Aythorpe Roding were six turned inwards.

 A forward position of the neck bearing of the windshaft was of benefit in obtaining clearance of the base or roundhouse by the inner pair of sails, and of some import in the case of a small body on a large substructure, but the angle required could of course be arrived at by adjustment of the pillow blocks resting on the weather beam. At Mill Green the windshaft measured 16ft between tail bearing and the centre of the rear middling, and was inclined at seven degrees. The sails were approximately 29ft long from canister to tip and passed very close to the roundhouse and within four feet of the ground. To cause the tips of the inner sails to pass an extra foot out from the roundhouse wall, it would be necessary to raise the neck bearing by about 6in, then giving the windshaft a tilt of nine degrees. To replace the windshaft with one of different size or modify the angle of the existing one would necessitate a chain-reaction of readjustments, involving the brake- and tailwheels, stone spindle beams, striking gear and sack tackle, and perhaps even the stones themselves, with tentering gear, and was not to be lightly undertaken.

 Of interest at Mill Green is the third cross member in the prow in order upwards, apparently superfluous, uniformly 9in square, and passing behind and quite independently of the prick post. It lies 27in on centres slightly off the vertical below the weatherbeam, and is notched full width into the posts so as to sit on them. The original was not examined in 1960, and its points of entry in the corner posts had very badly rotted. It was probably an addition to the original structure, let in from the rear as at Aythorpe Roding mill, and would be unlikely therefore to have carried tenons. The beam could have accepted considerable pressure from above, and this gives a clue to its purpose, which must have been to provide a base for the propping of a sagging weatherbeam, its undoubted function at Aythorpe Roding (removed in the 1976 rebuild). It may have proved useful for the raising of that beam, or of the windshaft, for bearing-renewal or for repair. Any tell-tale marks, holes or cuts which the lost beam may have shown to indicate its use – in workshop parlance – as a tooling beam, are now lost for ever.

 It is difficult to imagine that the front intermediate framing would have been assembled up aloft. This frame is three-dimensional, and the diagonals are inclined in two directions relative to the prick post; their upper tenons are therefore splayed to suit and they must be introduced to their mortises before the prick posts is set back on the transverse members.

 The side frame members comprise side girt and corner posts, bottom and top side rails, two intermediate uprights, one on either side of the crowntree, and two main diagonals, the last tenoned at their foot into the intermediate posts, which were spaced with a straining piece 10in by 4½ in at this level horizontally. The diagonals were notched full width at either end and tenoned also, giving a very robust construction.

 As above explained the joints between side girt and corner post were such that the corner posts had to be hooked over the girts, after which the full depth of the girt end was sunk bodily into the post with angled abutments. Before the two posts were brought perpendicular to the girt, the bottom side rail would be inserted into the prepared housings. The various members bore chiselled Roman numerals for identification and checking in these operations, each main joint being a unique match of timbers. The intermediate posts, 7in by 4½ in and on 34in centres continued right through from the base of the weatherboarding below the bottom rail to the top rail, into which they were tenoned. They were halved into the lower rail and set in flush with the outer face of the girt. In the latter case, however, was a refinement in that they ran full section for 1½ in upwards into the girt before their cutaway section was reached, which left 2in intact out of the 4½ in dimension. Hence the housings in the girt were stepped to increase stability and support, and the intermediate posts could accept some of the load on the crowntree end; the lower side rail was notched in at its base into the corner posts to carry the logic of this construction to its conclusion. The prolongation of the uprights through the entire width of the side girt would also discourage “hogging” of that member. In many mills, the intermediate uprights in the side framing terminated upwards in the underside of the girt. When all was pressed home, the diagonals below the girt, tenoned and notched in full width suitably at both ends, gave further support to the girt very close to its extremities; despite all these precautions, one had badly cracked and the crowntree junction at the last, but largely owing to age and decay. Thus the girts, functioning as major supporting members, were themselves generously supported in a subtle feed-back scheme involving all the major timbers back to the sheers, and representing a most ingenious and harmonious conception.

 At Mill Green the top side rails are 9in by 8in. The rear beam is of very heavy section, 12in square, and spans the extreme rear of the body. Usually a much lighter member is employed for this purpose. Within the various frame sections is fixed studding to carry the weatherboarding, made out of 3in to 4in stuff on 12in to 15in centres, tenoned into the nearest horizontal member above or below, but spiked with iron nails into the diagonals. In the upper panels of the side frames were fitted two short diagonals springing from the side girts up and out to the corner posts; these were set into mortises cut large to allow them entry and subsequently blocked up, for the hooking manoeuvre inhibited their fixing in any other way.

 The relationship of mortices cut in the corner posts for transverse and side members is also of interest; it was desirable to stagger them to preserve strength. The bottom side rails entered 3½ in on centres below the forward sill and corresponded in height of top surface with the sheers to take the spout floor boarding. The meal beam was set 24in on centres below the side girts; here there was no problem, for it was required to lie opposite the crowntree. The rear (main) transverse members followed a similar pattern.

 Behind the main corner posts are the false corner posts supported by a cross member or sill 7½ in by 7in set on the ends of the sheers, and also by three horizontal pieces, 6in by 5in approximately, set between the main and false corner posts and braced by diagonals inclined downwards into the main corner posts. At the rear the door is suitably framed, and an intermediate cross member 8in by 7in forms the base for the two diagonals inclined up and out to the false corner posts. Whether or not this section was added is considered later.

 The doubtless original forward spindle beam was cut away at its extremities and tenoned into and wedged against the top side rails with angled cuts made in the traditional style, allowing fine adjustment by various tapered wedges, or “throat” wedges, in the two senses required. It was 11in wide by 10in deep, 10ft long overall, and had one tenon at each end to enter the side rail mortises and a stub end to lie over the rails. In the fore and aft sense the beam was adjusted by the use of two pairs of folding wedges introduced horizontally into the mortices in the rails. In the new construction mortised blocks have been coach-screwed to the inside faces of the rails, allowing the old spindle beam to be reused in shortened form with ends re-cut. The tail beam for the windshaft was of similar heavy section, and similarly applied to the rails, but was carried over them in substantial section to ensure strong support for the rear of the windshaft, originally of wood but subsequently of iron. The striking rod passed through this tail beam and out to the rear of the mill, but the exterior mechanism was missing in 1959. The tenons of both these transverse beams were accommodated in mortises 15in by 2 ½ in and 5in deep cut into the top side rails, giving therefore the play needed for adjustment fore and aft, while the rebates in the under faces of the beams allowed lateral adjustment by wedges against the side rail edge faces. In the vertical sense no resetting was required for the glut boxes, used in this mill for the disengagement of the stone nuts, could be fixed as required to take out the stone spindles. The positioning of the glut box for the rear stone spindle was arrived at unusually. A block with a suitable square-cut section removed from its upper part was applied to the underside of the tail beam, allowing the through passage of a rebated section of a longitudinal spindle beam passing from the rear framing forward to a point above the centre of the stones. Here also wedging adjustment was used between angled cuts in the spindle beam and the rear and forward faces of the applied block. Inclined channels had been cut in the forward face of the front spindle beam to allow the passage of the iron teeth segments which had been bolted over the sawn-off cogs of the brakewheel. The tail wheel also had iron segment teeth, both wheels being true face gears. These were part of the usual train of modifications found in old mills.

 The roof is framed between two principal pairs of rafters 5in square, curved in single pieces to give the required shaping sufficient to accommodate the brake wheel and sack hoist, and tenoned into the side rails below. At the front an inclined pair of corner spars, footed on the weatherbeam and tapered to abut the main uprights, gives the forward shaping. Between these lies one cross member 5in square above the weather hatch. At 6ft 9in above the side rails are purlins 6in deep by 4in wide tenoned into the principal rafters, between which stand twelve sets of spars let into the purlins and pegged together at the top, no ridge piece being used.

 The rear ladder is hinged to the sill above the ends of the sheers for raising when the mill is turned. The present stringers are in pine 11in by 6in and on 5ft centres with 22 treads between. Six treads are trimmed at centre to allow free passage of the oak tailpole, which projects some 30ft beyond the rear of the mill. This is tenoned into the cross member between the sheers behind the sack traps, and is held by tie rod from the main rear sill and would formerly have been secured near its extremity by an iron stay from a cross member at the top rear of the body to enable it to carry the very heavy ladder when lifted.

 Within the mill body the stones, bridgetrees and brays were the remaining heavy items awaiting assembly. To carry the forward stones the foundation was formed by two members 10in wide by 9½ in deep on 3ft centres running fore and aft from the meal beam to a 4½ in square piece fixed to the forward side of the crowntree, into both of which supports they were sunk about ½in, losing no strength themselves. Two other subsidiary bearers 10in wide by 7in deep on 4ft centres were set between the main bearers to form an approximate square from which the bedstone could be packed up on wooden wedges and levelled as necessary, due regard being had to the preservation of strength in the main fore-aft pair when setting in the cross members. From the rear cross member depended a 14in by 4½ in piece in oak, descending 2ft 9in with twin mortices to carry the bridgetree tenons at this, the pivoted end. This hanger was secured in the beam above by three stout 1¼ in pegs. The 5ft long oak bridgetree, nicely shaped and chamfered and with iron bridging box, to which was assigned the duty of carrying the runner stone, was held at its forward end on a bray 9½ in deep by 3½ in which was pivoted at the rear face of the left corner post and ran right across the prow of the mill to the lighter irons. The steelyard returned to the opposite (left) side of the mill to the centrifugal governor driven by belt from the stone spindle.

 The rear stone arrangement was similar with regard to size and layout except that the bridgetree was pivoted at the forward end, near the crowntree. Both governors were mounted on wooden brackets fixed to the left side framing. The to-and-fro of the tentering gear contrasts notably with later designs where the bray was either very short or even dispensed with altogether as at Stock tower mill.

 The stones could conveniently be hoisted up from the base before the floors were laid; there then remained secondary items such as bins, chutes, sack hoist, taps, studs and weatherboarding. On the spout floor no joists were employed, the boards running across the mill over bottom rails and sheers, but on the stone floor they ran fore and aft, slightly lower below the crowntree, as is usual. A third (part) floor was set to the rear under the roof at about 3ft over the top side rails, comprising a short walkway between bins. The need for this would have become more pressing with the “laying-up” of a second pair of stones and a dresser in the tail, which may have occurred many years after 1759. This prompts the question of the extension of the mill body rearwards beyond the main or primary rear corner posts to give the necessary space. In some old post mills such as Ashdon, Essex, and formerly Chinnor, Oxon, evidence of this modification was to be had. This includes obviously extended top side rails, blocked-in mortices for door posts in the rear sill at spout floor level between the main corner posts, where formerly weatherboarded. Instances of extended sheers have not been noted, but it would appear from old prints and watercolours that a common practice was to carry them beyond the rear face of the mill body to form a platform with porch over, so that in many cases they offered a ready-made base over which to frame an extension. At Mill Green the tell-tale signs of later extension are either not present today or have vanished unnoticed with the destruction of the old timbers. The presumed original transverse member at spout floor level between the main corner posts has been retained and its upper surface is exposed. It would appear to have started life with some form of built-out section, perhaps similar to that depicted in a drawing of Little Chishill mill on an estate plan dated 1769.

 A post mill would shed its weatherboarding a number of times during its long years like a snake its skin, and the building of a rear extension could be conveniently combined at some juncture with the fitting of a new roof and top rails, together perhaps with the introduction of a tail wheel, as well as a flour dresser. The Mill Green annex could be so accounted for, being of an overall depth of a little over 3ft, and carried mainly by the sheers. Some thrust is imposed on the main corner posts by the use of the two diagonals on each side, which triangulate the support of two intermediate transoms 5½ in by 4½in in section bridging the interval between the “primary” and “secondary” (or false) rear posts. The abutments of the timbers in the extension are out of character with the original mill framing in that they are not notched in where expected; this fact is still observable in an old re-used false corner post laid as a joist at stone floor level, as well as in the modern replica.

 The iron windshaft has an overall length of about 18ft and is square in section, tapering from a 10in side behind the neck bearing to 6½ in behind the tailwheel, and has chamfered edges. The neck and tail journals are of diameters 9in and 5in respectively. Both the 9ft brake wheel and 7ft tail wheel are of original clasp-arm construction and, in the case of the brake wheel, having 33in square iron plates bolted on at front and rear over the clasp arms to enable the whole to be fixed to the windshaft with iron wedges. Four circular holes of 3in diameter have been left in each plate, creating – in the superannuated mill – an excellent and well-patronised nesting box. The wear from braking had caused the rim of the wheel to be worn down to the outer ring of dowels holding the felloes to the cants; it had accordingly been lined by wooden segments 6½ in wide by 1½ in thick as now seen, butted together and dowelled to the rim. The wooden brake was in four segments varying in length from 5ft 8in and tenoned end-wise together as well as linked by paired iron straps. Both brake and brakewheel appear to be in elm.

 The brakewheel has four cants, and the tailwheel six. In both cases iron segment teeth have been bolted on, and the wheels are thereby face gears. The brakewheel has 10 segments of 12 teeth apiece, making 120, which drove an iron mortice nut with 20 wood cogs. A revolution of the sails thus gave 6 tuns to the forward runner stone, a 52in burr. The original wooden brake wheel cogs numbered only 69, and were of 4½ in pitch. The rear nut was also of iron mortice type, having 15 cogs driven by the 100 in the tailwheel, thus giving a slightly faster speed to the rear runner. The old cog mortices in the tailwheel, if present, are not visible, as the wheel was wood-plated at front and rear before the iron segments were bolted on. To the arms on the forward, non-working face, is fixed the wooden sack hoist pulley, of 3½ feet effective drive. Curiously enough, this has blocked-in compass-arm beds. The sack traps are behind the crowntree. To raise the sacks the miller pulled a rope from any level, which brought up the rear end of the hoist bollard by lever control to take up the slack in a chain between the driving pulley on the tailwheel and a driven pulley on the bollard. The sack hoist assembly was, of course, built across the central roof spars.

 Reference has been made to the bridgetrees and the associated brays, all of which survive, both sets being similar in design. We may take the forward stone control as an example of the regulation obtainable. The fulcrum around which the steelyard pivoted could be set under any one of a row of notches near the end of the yard arm, thereby altering the mechanical advantage of the governor control. If it were placed at 6in from the end, and therefore, in this event, at some 84in from the governor, a unit rise at the latter end would cause a depression of 1/14 at the free end of the bray. This would lower the bridgetree end by 1/25 units, the point of support of the bridgetree being at 5/9 of the length of the bray from its pivoted end. The footstep bearing of the stone spindle was at 2/5 of the bridgetree length from its pivot, so that the depression of the runner stone would amount to 1/60 of the rise of the governor collar. If the fulcrum were now moved to a position 3in from the end of the steelyard, the stone would move down through only 1/125 of the rise in the governor. Thus the system was capable of very sensitive adjustment through a considerable range of settings, and for a given position of the steelyard setting the gap between runner and bedstone could be altered by operating the screw control.

 It has been suggested above that the crowntree discarded in the late rebuild of 1959 was itself the replacement of an earlier one, and the fact that the main stone bearers were not tenoned into it but merely rested on applied timbers, supports this conjecture. However, mill builders generally considered it bad practice to weaken the crowntree in any way, and preferred to add “slabs” for the stone “trimmers” or joists (Wightman). Major repairs to post mills often called for extensive propping of the body frame from ground level, and {Vincent} Pargeter considers that recesses cut in the outside faces of the main vertical components generally originate from such operations. Each of the four main corner posts at Mill Green carried angled recesses which were plugged with wood before the weatherboarding was applied. On the right forward corner post there were three such shallow notches, each about 1 ½ in deep, full width: two on the forward face (one near the top and the other a little over spout floor level) and the third on the inside face close to the second. On the rear main corner posts was a single recess a few inches below the top. Thus provision was made to prop the forward and side direction. These recesses may have dated from Barker’s rebuild of 1759.

(Farries Ch.10 (Ch.1 of Vol II) gives a list of the main structural members with their dimensions)


Smock mill, gone


There are numerous photographs, which show four single-shuttered, anti-clockwise patent sails and a “planted” fantail, which was cut down to suit from the dismantled Bentley mill at South Weald. The mill was tarred, and probably had four floors above the ground. It dated from the heyday of Essex smock mill building and was handsomely proportioned.


Composite mill, gone


The first Stephen Roast died in 1797 and it is claimed that the erection of the ultimate windmill, set up as an ordinary post mill, was financed from his legacies. In about 1850 the mill body was jacked up and a circular brick tower constructed beneath to create a composite mill. The motive is stated to have been the need to increase storage space, but a compelling reason was purportedly a defective main post. The heights of the slightly battered round brick tower and the post mill body above were each about 24 ft, making an overall total of nearly 50ft. The tower had a diameter of 21ft and contained three floors, the uppermost of which had four porthole windows framed by double circles of tapered bricks. Directly below were further windows or loading doors. The framing of the post mill body, which measured 19½ ft by 10½ ft, was on familiar lines. To hold the crowntree in position when the main post had become redundant, wooden bearing pieces were angled into the pair of intermediate verticals which rose fore and aft of the crowntree ends. These verticals also gave support to the ends of the side girts through the usual long diagonals, and in other framing panels, also, the structure was well triangulated. Below, strong cross members were secured to the base of the corner posts and rear intermediate verticals to form the basis of the table below which the structure rotated on the tower.

 The body ran on rollers spaced 3-4ft apart round an iron track on the wall top, and was centred by iron wheels braced by tie rods, one to another, as is still to be seen at White Roding. To wind the mill, an 8-bladed fantail was set over the top rear as at Ramsey, but here the shaft ran down to the left of the rear door to rack level, whence the drive was taken to a central point under the mill body, then through bevel gears and two shaft, one on either side, suitably angled, to drive two other short shafts on which the final spur pinions geared with the rack. These pinions acted on opposite sides of the circle, and rode over some 450 iron teeth before the mill was once turned. Hand winding gear was provided on the rear landing. The rollers ran outside the rack and were hidden from view below by a wooden petticoat and from above by a walkway guarded by a railing. The mill body was braced externally by stout T-section irons running down diagonally from mid-side girts to the extremities of one of the heavy central transverse beams over curb level.

 Much of the ironwork, including the rounded windshaft with two squares cast on it for the head and tail wheels, is said to have been made by Christy and Norris of Chelmsford near the end of the last century. The wooden brake and tail wheels were both built up as clasp-arm wheels, the first with eight cants, the second with six. The well-worn brakewheel rim was lined with wood to which was

applied an iron brake. This wheel drove two pairs of French stones underdrift, while the tail wheel drove a pair of peaks overdrift. A fourth pair in the first floor of the tower was operated by steam. Most of the smaller gearing was of iron. R Hawksley gives the forward stones as 4ft 4in driven through gear ratios of 80-28 and 72-20, and the tail stones as 4ft running at 118-15. The stone nuts were raised out of gear by crank and ratchets, and the bridgetrees were wooden. Immediately over the wallower was an iron bevel gear for drive to the sack hoist and bolter. The last working sails were four single-shuttered anti-clockwise spring patents, having a spring cross.


Post mill, gone


A new iron windshaft was fitted by John Bryant of Colchester; a photograph of the operation of hoisting it by the use of a sail middling and tackle has survived. A generous photographic coverage of the mill shows two pairs of double-shuttered patent sails in use, turning clockwise; later there were single-shuttered sails, and the striking gear was outside the rear of the mill body. There was an iron brake on a wooden brake wheel and both pairs of French stones were in the breast, underdriven and controlled by one governor. There was no tailwheel. The 6-bladed fantail over the rear ladder was a late addition. The mill was of attractive appearance, with neatly finished weatherboarding and wood palings on one side of the platform behind the spout-floor door. At the back of the roof ridge was mounted a weather vane.


Post mill, gone


The only known view of this windmill {is} a photograph with rather poor definition. It had four clockwise single-shuttered sails, winded by tailpole, and apparently set in an all-wooden windshaft, and there was a black stripe below the poll down the breast of the mill as at Great Easton. The stones were arranged head and tail.


Smock mill, gone


It was a white-painted smock mill with a fantail, domed cap with vertical weatherboarding {and finial}, and four shuttered sails with canvas on wooden frames. There were three pairs of overdriven stones. The cap was preserved as a garden shed close by, but by 1981 had all but disintegrated. It was roughly 8ft in height over a base diameter of about 12-14ft.