Cones, Not Domes: John Nash and Regency Structural Innovation Author(S): Jonathan Clarke Source: Construction History, Vol

The Construction History Society

Cones, not Domes: John Nash and Regency structural innovation Author(s): Jonathan Clarke Source: Construction History, Vol. 21 (2005-6), pp. 43-63 Published by: The Construction History Society Stable URL: http://www.jstor.org/stable/41613894 . Accessed: 06/04/2013 16:45

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp

. JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

The Construction History Society is collaborating with JSTOR to digitize, preserve and extend access to Construction History.

http://www.jstor.org

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions ConstructionHistory Vol . 21. 2005-6

Cones, not Domes: John Nash and Regency structural innovation

JonathanClarke

TheMuseum of Artillery inthe Rotunda is a littleknown building within John Nash's considerable oeuvre,but it is perhapshis most structurally ambitious. Situated on the edge of Woolwich Common in southeastLondon, where it has stood since 1 820, its distinctive tented roof is immediatelyrecognisable, evokingboth battlefield and festival (Fig. 1). Inside,the tent-like feel of thebuilding is evenmore pronounced,with canvas lining seemingly draped from a hugeDoric column. This post, analogous to thecentral pole of a militarybell tent, helps support an elegant,yet fully concealed timber-framed roof structurewhose outer ends rest on a circularcolonnade enclosed by polygonal brick walls (Fig. 2). The roofstructure, 116ft in diameter and described in 1830as having'no equalbut the dome of St. Paul's Cathedral',is all themore remarkable for the fact that both the brick walls and the central post are secondaryfeatures.1 Originally clad in boardsand oilcloth, not lead, and encircled by timber-framed walls,the Rotunda roof and colonnade were, for the first few years, an entirelyself-supporting spatial enclosure.The Rotundawas designedlyunlike St Paul'sand other domical roof structures, and its catenoidalform demanded innovation instructural technique. It wasan extraordinary building born of unusualcircumstance. It is theaim of thispaper to examinethe Rotunda's structural design and influences,as well as tolook at how, in turn, its structural form influenced one of Nash's better known works,the Brighton Pavilion.

Fig.l and2. TheWoolwich Rotunda (Jonathan Clarke; PSA Rotunda G (CN) 211561).

Festivebeginnings: Carlton House, 1814 Nash'sRotunda began life as a hugepolygonal ballroom erected in thegrounds of Carlton House, London,in June 1814. It wasoriginally the centrepiece ofa suiteof temporary buildings erected for a grandfête held in honourof the Duke of Wellington, and is thesole surviving structure ofthe whole CarltonHouse complex.2 The Prince Regent delighted in stagingcelebratory events in the gardens of hishouse, and over the previous 30 yearstents and marquees had become familiar sights in the summer months.The scale of thesecelebrations increased, and to maintainsome control over the costs of 43

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions Cones, not Domes: JohnNash and Regencystructural innovation annuallyacquiring tents, the Prince had begun commissioning temporary, demountable buildings that had greaterlongevity and morescope for architectural character. Such buildings were abundant, 'immense,and admirably contrived'3 in 1811when they were designed under the direction of James Wyatt,following whose death in 1813,responsibility forCarlton House and its stock of semi-permanent marqueestransferred toNash. The fêtes of 1813 were relatively low-key, but the abdication ofNapoleon inApril 1814 gave the Prince enormous cause for celebration, and Nash, then his favoured architect, scopeto impress. He expeditedthe design and erection of as manyas 15 to20 newtemporary rooms intendedfor three celebrations inJune, July and August (Fig. 3). 4 Onlytwo were erected in time for the receptionofthe Allied Sovereigns inJune, and it was not until the following month that the full panoply wasready. This was the occasion of the Duke of Wellington's fête on 21 July,the most extravagant ever stagedby the Prince at Carlton House.5

Fig.3. Woodcutshowing Nash 's Chinesepagoda and bridge erected over the canal at StJames 's Park inhonour of the Grand National Jubilee of 1 August1814. Thepagoda proved more temporary than anticipated,catching fire before midnight and toppling into the water. Published by J Pitts on 16 September1814 ( ScienceMuseum/Science & Society Picture Library, Picture Reference: 10411213).

ThePolygon Room described

The 'PolygonRoom', or 'TentRoom', as theRotunda was originally called, was the showpiece of theensemble of interconnecting temporary structures which were designed collectively toaccommodate 2, 500 guests,including royalty, nobility, foreign ambassadors, ministers and officers of state. The temporarybuildings were laid outin an H-formationto the south of CarltonHouse, and included refreshmentrooms, promenades, giant supper rooms, a botanicalarbour and a Corinthiantemple to Wellington.Atthe centre of the whole arrangement was the Polygon Room, with three apartments tothe east,west and north.6 The recent discovery of the original designs for the Polygon Room testify to the skilland ingenuity that went into their design and rapid construction (Fig. 4); no drawingssurvive detailingthe form or construction ofthe other temporary buildings at CarltonHouse. The drawings are

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions JonathanClarke datedMay 1814,suggesting that the Polygon Room was assembledand erected in tenweeks or less. Nashwas responsiblefor the overall design, but on structuralmatters it seemscertain that William Nixon,his chief carpenter, aided him considerably, as will be shown.

Fig.4. Drawingsof the 'Tent Room, Carlton House': Section (top), elevation (inset), roof plan (left) andground plan (right). All are unsigned, but dated May 1814. (TheNational Archives (TN A): PRO WORK43/575, 43/574, 43/573).

Thedrawings show a twenty-foursided structure inthe form of a hugebell-tent, 120 feet across and 80fthigh with a distinctiveconcave roof and match-boarded walls pierced by doors and large windows onalternate faces. The timber roof structure, carrying outer and inner sheets of canvas, is formedfrom an arrayof 24 taperinghalf-ribs that fan downwards and outwards from the apex. Eachhalf-rib is a trussedassembly of memberswith curvilinear laminated timber upper and lowerchords shaped as catenarycurves. They are strapped together atthe crown, their bevelled upright members - analogous to kingposts - combiningtoform a cylinderrather like the staves of a barrel.This hollow timber cylinder carrieda louvered turret, which helped ventilate the building, and may have improved the acoustics. The lowerends of the ribs rested on two concentric circular wall plates joined together inthe horizontal plane bydouble-diagonal timber braces. The outer wall plate was carried on andbraced to 24 timberposts, whichalso framedthe panel walls. The innerwall plate was carriedon reededtimber columns set slightlyinside the outer posts. The ground plan denotes these as havingcast-iron cores, and photographs takenin the1970s when the Woolwich Rotunda was restoredsubstantiate this (Fig. 5).7 Cast-iron archedbraces with open-work spandrels link these posts circumferentially, andsimilar but shorter braces spanbetween inner and outer posts. The plan of the ground floor clearly indicates members connecting thebases of the inner posts, as wellas radialand diagonal members linking them to the outer wall or foundations.Corresponding to those above, these elements were presumably sited beneath the floorboards.However, unlike those braces at higher level, it is unclearwhether they were made of timber

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions Cones, not Domes : JohnNash and Regencystructural innovation oriron, or meant to be actingin tension or compression. This drawing also shows an innermost ring of twelveposts set at the centre of the building, each of lesser diameter than those already described, but stillwith an iron core. According to a descriptionofthe fête in The Times , this circular array 'supported a garlandof artificial flowers in the shape of a temple... usedas anorchestra for two bands'.8 The Times alsogives us someidea of how impressive this structure must have looked on the day of the ball, with decorativegroins formed between the inner and outer posts 'from which arose an elegant umbrella roof ... decoratedwith large gilt cords ... paintedto imitate white muslin, which produced a very light effect'.9 ThePolygon Room was a hugesuccess. Nash had succeeded in creating an illusionisticset-piece, onethat from the outside looked familiar, like a hugetent, but whose sumptuously decorated interior, withoutthe expected central pole, must have occasioned feelings of surprise and awe. Thatmere cloth couldseemingly sustain itself overhead in a funnel-likecontour must have bamboozled even the more soberguests, and appealed to theirnotions of the sublime. Certainly itfound favour with the Prince Regent,who retained itand the three adjoining apartments until 1818, when negotiations were set in train forits removal to Woolwich. Almost all the other temporary rooms had been dismantled in1815. Before consideringthe structural design of the Rotunda in more detail, it is worthrecounting itsdismantling at CarltonHouse and subsequentre-erection at Woolwich,not only because the records present an unusuallydetailed insight into this early instance of préfabrication andportability, but also because it tellsus muchabout how Nash thought about the structure.

Fig.5. Photographtaken during the early 1970s restoration showing the cast-iron arched braces and timberdiagonal bracing at theroof perimeter. (NMR: PSA Rotunda G 189815.)

Dismantlingthe Polygon Room, 1818

Despitebeing retained by the Prince Regent at CarltonHouse, where it saw occasionaluse as a ballroomduring summer fêtes, it was clear by the summer of 1818 that the Polygon Room had outlived itsoriginal purpose. The celebrations marking Allied Victory were long over, Carlton House itself was indecline, and the expense of maintaining theremaining temporary buildings was a drainon the finances ofthe Office of Works. On 7 August1818, the Surveyor-General, Colonel Benjamin Charles Stephenson

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions JonathanClarke wasinformed that it was 'thePrince Regent's Pleasure that the great circular Room in the Gardens of CarltonHouse should be removed from thence ... [and]transferred toWoolwich, there to be appropriated to theconservation of thetrophies obtained in thelast war, the artillery models, and other military curiositiesusually preserved in theRepository of theRoyal Artillery'.10 Nash, stationed in Brighton, clearlyhad other hopes for his building, for later that month he wroteto Stephenson,mentioning a 'propositionI had to make related to the application ofthe building as a Church#...#when I come to town I willshow you a designand calculation on thatsubject'.11 In fact,Nash considered the building 'applicableto many purposes ... indeedI thinkit would be a usefulbuilding in any of the dockyards and othergovernment establishments'.12 ThePrince Regent's wishes prevailed, almost certainly influenced bySir William Congreve (1772-1828), Comptroller ofthe Royal Laboratory atWoolwich and inventor of thewell-known Congreve rocket. Congreve was a longstandingacquaintance of thePrince, and neededmore room to display the museum pieces that had arrived from Paris. Even so, his proposal was notapproved without considerable wrangling, his applicationinitially being turned down by the Treasury,which, fearful of the costs of dismantling, transporting andre-erecting the building, thought thatit should be soldto a privatebuilder.

Becauseit was such a largeand intricate building, Stephenson wrote to Nash for advice as tohow to dismantlethe Polygon Room. Nash gave these instructions:

Thebest course to takein pulling down the polygonal buildings in CarltonHouse gardens will be thus.After the boarded casing - thetimbers filling in between the posts and the boarded floor - shall be removedand the outside covering taken off, a scaffoldabout 8 or9 feetsquare should be erected inthe centre - the cupola taken down to the main ribs and the hoops taken off which bind the upright timbersof the roof together and form the cylinder - then to take out the purlins on each side of any oneof the ribs leaving the other purlins to steady the other ribs - whenthe rib is so disengagedraise thefoot of it up from the plates by a jackthe rib being suspended by the blocks of a triangle- and the upperend of the rib being also suspended by the blocks of another triangle on the scaffold - a guide ropeat thefoot of the rib will pull it overthe plate outwards and at thesame time disengage the uprighttimbers atthe upper end of the rib from the rest of the uprights which together form the centre cylinder- and the rib being thus entirely separated and disengaged may be easilylowered down by theblocks of the triangles ateach end - theother ribbs one by one may be takendown in the same way- afterwhich the joists of the floors may be takenup andthe posts and the plates forming the wallstaken down.

As thebuilding has been constructed so as totake to pieces and be refixed- I advisethat the posts andplates from the walls be numberedand marked a drawing being previously made of them with thelike marks of the drawing - so thatthe timbers may readily come into their places should the buildingever be put up again, there or elsewhere - and in like manner the floor the boarding of which hasfor that reason been framed in compartmentsas well as thejoists - theribs forming the roof shouldalso be drawnand marked, and numbered and the ribs taken down whole and the purlins markedand numbered.13

Detailedthough these written instructions were, Stephenson foresaw the need for onsite assistance. In hisreply to Nash, he stated I havegiven Mr [William]Wyatt who acts as labourerin Trustat CarltonHouse your detailed directionsfor taking down the Polygonal Building, and have desired he will wait upon your for such furtherinstructions as will be necessaryin the progress of so difficultanundertaking; forI conceive thatthe great size andweight of the Ribs, the height of the cupola, and the very nice mechanical constructionofthe whole Building will require more than ordinary care and abilities, to remove the samewith safety.14 47

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions Cones, not Domes : JohnNash and Regencystructural innovation

Nash wroteback to Stephensonon 30 August,explaining the more troubling aspects of the disassembly,including the tricky part of disconnecting the ribs, each of which has been calculated to weighfour tons.15 Nash reassured Stephenson that he had 'consultedNixon who built it undermy directionsand he is ofthe same opinion with me that with the centre scaffold I proposed and the jacks trianglesand blocks, the ribs may be taken down with great ease - andthat the scaffold being erected and a competentnumber of men they may be taken down in a week'.16In a furtherletter, dated 4 September, Nashoffered further reassurance, saying that he would make Nixon available on-site so thatthe carpenter could'explain the construction ofthe building and advise on thetaking of it down'.17 However, Nash addedthat 'he [Nixon]cannot stay longer than Monday or Tuesdayas theworks at Brightonare extremelyurgent and so difficultofexecution that I cannottrust him a momentto be away'.18 TheRotunda was duly dismantled inthe autumn of 1818.Only five men were employed to do this, andjust two, a carpenternamed Jeffry Wyatt (who had been entrusted with its original erection), and a labourercalled Francis Swinney, did the hard toil. William Wyatt, Labourer in Trust at Carlton House, superintendedthework, and also sorted and numbered the materials; William Nixon was on hand for the firsttwo days in an advisorycapacity.19 The totalcost for its dismantling was £21010s. 41/2d., an amountrepaid by The Ordnance Department tothe Office of Works. The accounts show that £44 was savedby the reuse of 'old lead', a materialwhich had presumably been used to cover the joins between thesheets of canvas covering the roof.20

Re-erectionat Woolwich,1819-20

Ifthe dismantling ofthe Polygon Room was somewhat delayed, then its re-erection as The Rotunda waslong-drawn-out. When it reached Woolwich is notclear, but it was in early December that Congreve requestedthat orders be givento theCommanding Engineer at Woolwich'to makethe necessary arrangementsforits erection on the brow of the Hill at the Eastern boundary ofthe Repository Grounds, thatspot being the most convenient as wellas themost picturesque situation for it'.21 Bureaucracy delayedprogress for a furthersix months, and it was not until 14 June 1819 that Congreve could order 'thatthe erection of the Rotunda be proceededupon'.22 During this period it was decided to substitute brickwalls for the original timber enclosure, owing to the 'exposed situation' of the new site. The Board ofOrdnance had nearly one million bricks in storage, and it made sense to use some of these.23 During thisperiod - or perhaps a little later - thedecision was taken to introduce the central column that survives. In anycase, on 21 OctoberLt-Colonel John J. Jones, the Commanding Engineer in chargeof the re- erection,informed Congreve that 'The Buildingis nowin suchan advancedstate, that besides the interiorfurnishing, little remains to be executed, but painting and nailing on about one third of the canvas coveringto the weather boarding of the roof, raising about half of the masonry column to supportthe dome,and repairing the fixing in the window and door frames'.24 However, a particularly severe winter seemsto havestymied completion, and it was notuntil May 1820that Jones was ableto declarethe buildingas 'beingin a stateto be occupied'(Fig. 6).25 In overallform and structural behaviour the re-erected building that stands at Woolwich is muchthe sameas thebuilding erected in Carlton House Gardens. Some of the original materials were found to havehad decayed beyond re-use during the course of rebuilding, but they were replaced by duplicate items.These included a largenumber of joists, rafters, floor and roof boards, as wellas spikes,nails, bolts,etc. and a newinner and outer canvas lining for the roof, replacing that which had been supplied initiallyby James Baber's Patent Floor Cloth Manufactory.26 Besides the addition of brick walls, which assistedthe wall posts in supporting the wall plates, the principal modification was the introduction ofa giantstone Doric column (Fig. 7). It has longbeen presumed that its purpose was to helpbear the additionalweight of thelead roofing, thought to havebeen added in c.1819when the Rotunda was rebuilt.27However, the lead covering was notadded until the mid nineteenth century, probably in C.1861,when General Lefroy undertook a major refurbishment ofthe museum.28 Until then, the roof

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions JonathanClarke coveringcomprised solely of half-inch deal boarding under canvas sheeting.29 So whywas the column introduced?The reason doubtless stems from the fact that the structure's status was now permanent. It mayhave been intended as a failsafe,in the event that the iron hoops binding the ribs together broke, or to ensurethat a particularlystrong wind-generated deflection of the timbersdid not end catastrophically.30Thenagain, its imposition may have been as muchsymbolic as structural,dignifying themuseum interior with a featureassociated with permanence and authority. Certainly there is a commemorativeaspect to it, for the names of former artillery officers adorn the shaft, although at what pointthe painting of these began is unknown.But we can assume that its structural role became all the morecrucial from c.1861, because the weight of the new lead roof would have imposed considerable loadson the timber roof structure otherwise, causing it to spread outwards and thrust excessively against theenclosing walls.

Fig.6. A viewof the Woolwich Rotunda published by R.W. Lucas in June 1820 ,just weeks after its completion.(© RoyalCollection Picture Library, RCIN 701653.)

' Fig.7. Sectionalviews of the Woolwich Rotunda as itsurvives today. The central 'freestone column wasone of the last components tobe addedwhen re-erected in1819-20, after the roof structure. (Drawingby Andy Donald .)

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions Cones,not Domes : JohnNash and Regencystructural innovation

THE STRUCTURALDESIGN OF THE WOOLWICHROTUNDA Thereare three inter-related aspects to thestructural design of theRotunda that because of their specialinterest, deserve comment: the unusual geometry ofthe structure, theform and behaviour of the trussedribs, and the use of laminated timber and specialised iron fasteners orconnectors (Fig. 8). Taking thefirst of these, Nash was faced with the difficulties associated with the peculiar tented form he wanted toachieve. Framing a three-dimensional hemispherical structure over a circularspace was complicated enoughin itself,especially if the span was large, since it brought with it the problems of intersecting structuralmembers, ofproviding the basic frame for the outer dome shape using straight timbers, and, wherean inner dome that rose above the wall plate was required, finding ways around the use of simple tiebeams to counteract outward spread of the ribs. Framing a three-dimensional timber structure that lookedconvincingly like a tentboth inside and outside, with a concaverather than convex curve, and whichcould not therefore rely on the mechanics of arch action, was perhaps more problematic. The fact thatthe structure ought not to exert any significant outward pressure on the (original) timber enclosing wallscompounded the complexity ofNash's mission. In essence, Nash was tasked with creating a rigid catenoidalenclosure, one that covered an unobstructedarea of some 10,600 sq.ft, and whose 116 foot- spanactually exceeded the clear diameters of manyof whatwere then the world's largest domes, includingSt Sophia,Istanbul (115 ft) and St Paul'sCathedral, London (112 ft).

Fig.8. Perspectiveviews of the Woolwich Rotunda as itsurvives today ; showing the tented form of the roofstructure. Not until c.1861 did weighty lead sheeting replace the canvas covering - some 40 years afterthe column was introduced. (Drawing by Andy Donald.)

TheRotunda's elegant curvature, both inside and out, was based on thecatenary, the curve formed bya chainor rope of uniform density hanging freely from two fixed points not in the same vertical line. Tentprofiles were archetypal examples of catenaries, as were the hanging chains or ropes of suspension

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions JonathanClarke bridges.But although the catenary curve had been observed for centuries, its correct mathematical equationwas notfinally deduced until the early 1690s, appreciation of its inherentstructural and geometricaladvantages following. In 1697the Scottish mathematician David Gregory published his theoreticaldeterminations onthe subject. These showed that if the catenary curve was the most efficient profilefor transmitting tensile forces, then it followedthat the inverted catenary curve was themost efficientprofile for the arch and the dome, acting in compression.31 Wren and Hooke used this logic of convertingpure geometry into mechanics in theirfinal design for a self-supportinginverted-catenary inner-brickdome at St.Paul's Cathedral.32 Later, in 1744,Euler showed that the catenoid, or shape formedby the revolution ofa catenaryabout its axis, had the minimum surface area for a givenbounding circle.33In architecturalterms this might have translated into the eighteenth-century realisation that, for a givencircumference andheight, catenoids required less cladding than hemispherical domes. Whether Nashwas aware of these developments, and whether he thoughtto exploitthem in the Rotunda seems doubtful.Nash was not mathematically minded, but he did possess an intuitive and discerning 'feel' for structure.Atthe very least, he was concerned with arriving atthe correct catenary curve, as a lightpencil sketchon the 1814 section shows (Fig. 9).

Fig.9. Sketchfor a catenarycurve on 1814sectional drawing. Nash seems to have been considering a 96-ft-dimaterstructure rather than the final 120-ft diameter roof. (TNA: PRO WORK43/575.)

Fig.10. Divided tie beam truss arrangements as illustrated inAsher Benjamin's American Builder's Companion(1811) and in the Polygon Room (1814).

Takinga secondaspect of thestructural design, the form and behaviour of the trussed ribs, it is evidentthat the overall structure was conceived as a three-dimensionaldivided tie-beam truss, with the uprighttimbers of each opposing pair of half ribs forming the king post, the upper chords acting as the principalrafters, and the lower chords functioning as the tie beam (Fig. 10). As DrDavid Yeomans has shown,early nineteenth-century American carpenters seemingly devised and refinedan ingenious trussedarrangement toaccommodate raised ceilings, one that differed markedly from the usual British solutionof raising the tie beam above the wall plate.34 The American architect Asher Benjamin, for

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions Cones, not Domes : JohnNash and Regencystructural innovation example,illustrated a divided tie beam arrangement in the second edition of hisAmerican Builder's Companion, published in 1811. Bydividing the tie beam into two lengths and slanting each half upward tothe king post, a higherceiling could be accommodated without the problems associated with the raised tie-beamarrangement, themost acute of which was ensuring sufficiently strong connections totransmit thesubstantial tensile force between the tie and the principal rafters. But the major reason why Nash usedthe divided tie-beam truss for his Polygon roof was because, made up of two identical halves with a sharedkingpost, it lent itself to a radialarrangement that eliminated the problem of intersecting tie beams.To haveradially united 12 trussesof raised tie-beam configuration would have been almost impossible,and would not anyway have provided the requisite interior volume or curvilinear geometry. Viewedin this light, the roof structure ofthe Rotunda is madeup of 24 radialtrussed half-ribs, bound togetherby their upright ends which form a giant shared king post. Notwithstanding thesteeply concave upperand lower chords, the arrangement ofeach truss bears a noticeablesimilarity to that shown by Asher,which itself is looselybased around the king-post theme, with pairs of secondary posts and struts usedto give support to the tie and strengthen theprincipals against bending. In bothcases, the lower chordor tie beam provides restraint against the outward thrust of the principals, which in the case of the Rotundais quiteexcessive, since, rather than being straight orcurving inwards, itcurves outwards. The weightof theseasoned timber roof structure alone, without canvas or boardedroof covering, is calculatedto be some58 tons,and the outward thrust would have naturally tended to straighten out the ties.Thus the strength ofthe joints between the tie beams and the upright timbers forming the cylindrical kingpost, and the hoops binding these together, was absolutely critical to the success of the structure. Totalreliance was placedon theability of wroughtiron to transmitand resist tensile forces at these points.Because of the inaccessibility ofthis part of the roof structure, itis unclearexactly how the tie beamsare connected to the uprights, but a detailof Nash's original drawing suggests that heavy bridle strapsand long coach bolts were essential (Fig. 11). Withthe half-ribs given stiffness intheir own plane bycircumferential purlins and diagonal braces, and restrained from spreading outwards by the divided tiebeam, the weight of the roof was transferredvertically onto the double ring of wallplates, with, ostensibly,little need for these to containthrust. Again, these, like much of theroof structure, are inaccessibleto inspection,and it is possiblethat the individual plates forming them are strapped end-to- end,producing two continuous polygonal 'tension rings' that may resist any remaining outward thrust.35 However,the original drawings do notindicate any ironwork connecting the plates (Fig. 12).

Fig.11. Detail of 1814 sectional drawing, Fig.12. Detail of 1814 roof plan, showing showingironwork used to connect the tie triangularlybraced double wall plate. beamsto the upright timbers forming the cylindricalkingpost.

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions JonathanClarke

Thethird noteworthy aspect of the structural design is theuse oflaminated timber and specialised ironfasteners or connectors.The upper and lower chords are made up fromfour layers of vertically laminatedplanks, strapped together so thatthe inner two break joint with the outer ones. The posts and raftersare also madefrom laminated timber, although in just three and two thicknesses respectively. Curiously,the 1814 drawings make no indication oflaminated timber, although we must presume itwas originalto the design since there is no indicationin theaccounts of the re-erection that the ribs were modifiedor replaced. The 1814roof plan also showsa differentarrangement ofrafters, and it would seemthat there must have been a slightmodification tothe design after May 1814 when the plans were drawn,due perhaps to thepractical advice of one of the carpenters involved at CarltonHouse, if not Nixonpossibly John William Hiort (1772-1861), who Nash employed as Clerkof Works and Measurer.36 Theuse oflaminated timber was a logicalchoice for the upper and lower chords as itgave them the necessarypolygonal curvature without compromising structural continuity. Forces were transmitted fromone plank to the other by a combinationofdirect timber-to-timber frictional bearing and the bearing on thecoach bolts that passed through the connecting straps. Some of these connectors seem to have beenspecially designed for the purpose, including forged shaped fishplates (Fig. 13), and Y-shaped two- waystraps, used to connect the lowest brace with the foot of the lower chord (Figs. 3, 5 & 7). Forthe mostpart however the planks were simply clamped and bolted together atthe points where the posts and strutsare joined to the chords, with the more usual bolted iron stirrups passing over the heads and feet ofthe posts, and over the back of the upper chord around the struts (Fig. 14). Althoughthe structural necessityofpassing the straps over the back of the upper chord is questionable,itmade sense practically, sincethe flat base of the inverted stirrup ensured vertical alignment ofthe roughly-hewn planks, creating a flushsurface for the boarded casing. Passing the u-shaped pieces of iron over the foot of the posts was howeverstructurally imperative, since they slung up the lower chords. Here, the use of laminated timber cameinto its own, for the outer pair of planks forming the posts were lapped over the innermost plank ofthe lower chord and the feet of the struts, enabling forces to be transmittedbyboth bolts and direct timberto timber bearing. Somewhat inelegant of appearance, itnonetheless avoided complex carpentry joints,and was thereforeideally suited to use in structuresintended to be demountable,and where constructionwas in anycase designedly hidden from view. A similarjointing technique with doubled postsof this type was used by William Wilkins in the roof over the hall at Downing College, Cambridge, erectedsome time between 1807 and 20.37

Fig.13. Forged-iron fishplate, used here Fig.14. Inverted u-shaped iron stirrup , here tosplice the laminated timber upper chord fasteningthe laminated upper chord to two (JonathanClarke). struts(Pat Payne , AA048109).

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions Cones, not Domes: JohnNash and Regencystructural innovation

Fig.15. Marquees , including a giant bell tent , erected for an eventfor the Kentish military volunteers heldat Mote Park in 1800, in the presence of the Royal Family. Engraving by Alexander W. Medium; (© BritishLibrary , Maps K. Top.17. 48.)

STRUCTURALINFLUENCES AND RESPONSIBILITY Illustratedforerunners Thereis no knowndirect precedent for the Polygon Room at CarltonHouse. It was a highly innovativestructure, a classic example of structural form following architectural appearance. That it took itscatenoidal shape from military and festive tents there can be littledoubt, as thiswas a formfrequently ondisplay in late Georgian Britain (Fig 15). Specifically,Nash seems to have emulated the form of the BritishArmy Bell Tent, a typeintroduced inc. 1794 to replace the inferior ridge-pole infantry tents and onethat figured prominently inWellington's Peninsular campaigns.38 But appearance was one thing, actuallyframing a large conical enclosure was another. Besides apprenticeship, thechief medium for the transmissionand diffusion of structuralcarpentry techniques in theeighteenth and early nineteenth centurieswere illustrated carpenter's manuals. As Yeomanshas shown, besides providing instruction in theframing of roofsand floors, manuals of this period also dealt with elementary geometry and how it couldbe appliedto the practical problems of setting out work (Yeomans, 1986). FrancisPrice, in the secondedition of hisBritish Carpenter, orA Treatiseon Carpentry, published in 1735,illustrated a numberof 'curvilinealRoofs of great Extent'. One of these, which he thoughtwould suit as a round temple,gave a sweepingtent-like exterior, although the raised tie beam framing arrangement only allowedfor a hemisphericalinterior (Fig. 16). Itis difficulttosee how this trussed structure might have beenrealised three dimensionally given the problem of intersecting timbers, besides, it did not really resolvethe problem of outward thrust, which Price thought was 'thechief difficulty tostruggle with'.39 His solutionwas to reinforcethe circular plate or curbby boltinga polygrammaticarrangement of timbersto it in thehorizontal plane, an inelegantcountermeasure that would have diminished the opennessand circularity ofthe interior. A similardesign appeared later in the century, within the 1769 editionof William Pain's The Builder's Companion and Workman 'sDirector (Fig. 17). Painwas perhaps themost prolific British author of carpentrymanuals in thelate eighteenth century, and one of the hallmarksof hiswork was clarityof constructionaldetailing. Like Price's design, on whichit was certainlybased, this was essentially a raised tie-beam truss, embellished with additional principals, both curvedand straight. But this design came closer to producinga conical interior that more closely mirroredthe exterior, and it showed the widespread use of iron strapping rather than the pegged joints andbolted connections ofPrice's drawing. 54

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions JonathanClarke

Fig.16 and17. Designs for framing circular temples, that on the left illustrated inFrancis Price's BritishCarpenter (1735); that on the right in William Pain's The Builder's Companion and Workman's Director(1769). Bothare elaborations ofthe raised tie beam truss.

Itwas, however, the thinking ofthe period's most accomplished British carpenter that most directly informedthe structural design of the Polygon Room. Peter Nicholson (1765-1844), the prodigious and influentialcarpenter-mathematician, produced a design that seems, with the benefit of hindsight, tohave brokennew ground. First illustrated inhis Carpenter's & Joiner's Assistant of 1797,his design for a 'circularbuilding' simultaneously solved the problems of outward thrust, intersecting members and of preservinginterior volume and shape (Fig. 18). Whathe proposed was a radialarray of half trusses, their tiebeams connected to the foot of a centralshared king post via a multi-armediron strap. In thiscase eighttie beams could be united,but, as he putit, the iron strap could consist of 'as manybranches as thereare tie-beams to be united'.With the strap firmly bolted to the king post, and with the tie beams firmlybolted to the strap, the whole structure would, according to Nicholson, be 'render[ed] ... secure andpermanent'.40 Itwas an adroit divided tie-beam arrangement that, thanks to the iron strap, connected thetwo halves of each opposing tie as wellas theties to the king post. However,itdid not explain how theprincipals were united around the head of the king post, and it can only be presumedthat a similar ironconnector was to be employedthere also. Forstability italso required that the lowest struts were extendedbelow the tie beams to brace against the polygonal walls, circumscribing theopen space below thewall plate. Nevertheless, theinnovative design continued tobe publishedin subsequent editions of 41 theCarpenters & Joiner's Assistant and in other works by Nicholson and his son,42 as wellas being 43 appropriatedbyothers as lateas 1843. Thedesigners of the Polygon Room would thus have had ample opportunityto see Nicholson'sdesign before May 1814,when the unsigned drawings of the Polygon Roomwere produced.

Fig.18. Peter Nicholson's design for framing a circular building. As wellas beingthe first recorded exampleof a divided-tiebeam truss , it incorporated an ingenious iron device for uniting the 8 tie beamswith the king post. Carpenters& Joiner's Assistant (1797), plate 77.

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions Cones, not Domes : JohnNash and Regencystructural innovation

Whodesigned the roof: Nash or Nixon? Responsibilityforthe structural design of the Polygon Room is unclear.This is partlybecause the distinctionbetween architect and carpenter was still blurred in the early nineteenth century, and partly becauseJohn Nash (1752-1 835) was consistently atthe forefront ofarchitectural technology. He started outas a carpenter-architect,designing for example a newdouble roof for St Peter's Church, Carmarthen, in 1785(with a youngarchitect called Samuel S. Saxon),and by the beginning ofthe nineteenth century whenhe had reached gentleman-architect status, he waswell practised in both structural carpentry and ironconstruction.44 Nash was experienced intimber-dome construction, giving, for instance, Caledon House,County Tyrone, two domed extensions inc. 18 12. Whenone of these domes began 'swagging', he gaveexplicit remedial instructions to the owner, demonstrating both his appreciationof timber shrinkageand the value of iron straps:

theconstruction bywhich the dome is supportedis capable of sustaining treble the weight... I am of opinionit will be found that the timbers have shrunk which all timbers will do - & thatscrewing them upinto their place again will bring the whole right again - itis alsopossible that the carpenter who '45 fixedthe framing may have neglected to wedge it up tight or may have omitted the Iron straps

Nearthe end of his career Nash maintained that 'No founderever furnished me with a designfor any castingI ever used'.46 Summerson thought that 'Nash was the last English architect toconsider himself notonly an architect but an engineer'.47 Nash'sproficiency instructural design is onlyhalf the story, for T. F. Huntwrote in 1830 that the roof of theRotunda 'was designedor inventedby, and executed under the direction of, the late William 48 Nixon,a modestand retiring man, of rareworth and talent' Unfortunately,Hunt is notentirely reliable,and may have conflated the London carpenter William Nixon who died in 1826or 1827 with a Carlisle-basedbuilder and surveyor of the same name, who died in 182449 A lateredition of Hunt's 50 book,published in 1836, saysthat he came from Carlisle and designed the county gaol there, shortly beforehis death; there is noevidence that Nixon the carpenter worked in the north. In factvery little is knownof Nash's William Nixon at all. Tradedirectories show that two carpenters, William Nixon and WilliamNixon Jnr. (presumably father and son) were working from Cockfosters, Enfield Chase, in 1808.5 1 By 18 1 1 JamesWyatt, Surveyor General, had made Nixon a superintendentofworks at Carlton House,and it was there that Nixon came into Nash's orbit. Nash recounted the circumstances ina letter ofMarch 1815:

Inthe course of the work I foundit impracticable tocommunicate my directions through the Clerk of Worksor Labourer in trust, their other avocations allowing them very seldom to be there.I therefore availedmyself of a veryintelligent man (Wm Nixon) whom I foundat Carlton House placed there bythe late Mr Wyatt, and through whom I directedthe works which have been executed.52

ThroughNash's patronage and his employment with the Office of Works, Nixon became clerk of worksat Brighton and general superintendent atBuckingham Palace. In his will, written from St James's Palaceand proved on 30 April 1827, he bequeathed his architectural books to his two sons, William and CharlesNixon.53 Nixon had clearly developed an excellent working relationship with Nash, one of his executors,over the last fifteen years of his life. Nashpaid tribute to Nixon for his work at Brighton in a letterof c. 1821, declaring him 'the most diligent attentive and the most honest Clerk of the Works that I haveever met with'.54 His son, William Nixon Jnr. (c. 18 10- 1848), who eventually headed the Office of Worksat Edinburgh,enjoyed an evencloser working relationship with Nash, and was probably employedas Nash'sagent and clerk of works on the Isle of Wight.55 Thestructural design of the Polygon Room must have been intensely collaborative atall stages.As we haveseen, in 1818Nash stated that Nixon built the Rotunda 'under my directions'.56 This leaves

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions JonathanClarke scopefor Nixon to haveresolved detailed aspects of thestructural carpentry, such as theingenious methodof connecting24 halftrusses using beveled kingposts, hoops and wedges (Fig. 19). This practicalmodification ofNicholson's contrivance was probably derived from experience or knowledge of cooperingor timberwinding-drum manufacture. Nixon probably had moreneed, and more opportunity,tokeep himself abreast of developments instructural carpentry. He wasno doubt familiar withNicholson's work if not the man.57 Nixon may even have introduced the idea of divided tie-beam trusses,as theonly practical means of structuringtheinterior catenoidal form that Nash wanted. It should,finally, be rememberedthat on top of other commitments Nash had to design 15-20 temporary buildingsinless than three months. He cannothave devoted much time to any.

Fig.19. TheWoolwich Rotunda's 24 individualbevelled king posts, strapped together with hinged iron hoopstand fortified against timber shrinkage by wedges and packing pieces. Barrelmaking might haveinspired this arrangement. (Jonathan Clarke).

CONES AT BRIGHTON In 1815,the year after the Carlton House fête, Nash began remodeling the Brighton Pavilion for the PrinceRegent. His preliminary design for the east front was a somewhatuninspired affair, with a large oniondome in the centre, flanked by smaller, squatter domes either side.58 The design that was erected in1817-20 is thecelebrated medley of domes, cones and minarets that survives today, with two sweeping concavetent-roofs balancing the central convex dome. Nashstated that he hadplaced both designs beforethe Prince Regent in 1815,and that the Prince chose the conical version, even though itwas more expensive.Some historians, however, maintain that the final solution was onlyarrived at duringthe courseof building work in 1817. 59 Whatever the truth of the matter, the tented roofs over the music and banquetingrooms were direct progeny of the Polygon Room, smaller, but in someways more refined. Brighton'stwo soaring conical timber frameworks span 40ft between wall plates, each being made up of 20 radiallaminated timber half-ribs strapped together by their upright inner parts (Fig. 20). Forthese lower,flatter ceilings it was possible to tie the lower chords with collars, accomplished by an 20-armed bolted-timberarmature that avoided problems of intersection. The feet of the ribs are seated on cast-iron brackets,which project from a timberwall plate that in places is strengthenedbyiron beams, and which is variouslycarried on timbersquinches, flitched beams and brick walls. Thiscomplex 'flying wall

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions Cones, not Domes: JohnNash and Regencystructural innovation plate'construction seems to have been designed to counteract, orsafeguard against, the spreading ofthe timbers.But the most surprising difference between these structures and their prototype is the absence ofmembers connecting the lower and upper chords. Each half-rib has just two posts, the upper, which is accessibleto view, being of doubled form, like in the rotunda, with timber bridging pieces between (Fig.21). This paucity of web members is possible because the laminated upper and lower chords were considerablystronger, made up for the most part of five thicknesses ofclosely fitting planks. Only in thehigher part of theupper chords did they reduce in thicknessto threeplanks. Consequently, the laminatedchords are better equipped to resistbending, with less need of supportalong their length. Unlikein theRotunda, the planks were cut to therequired curve, and, it is worthnoting, the overall qualityof carpentry is greatly superior.

Fig.20. Sectionthrough Banqueting Room roof, Brighton Pavilion, drawn up by William Nixon, 1827. (Bycourtesy of the Royal Pavilion, Brighton).

Fig.21.Upwards view towards the apex of the laminated-timber roofover the Music Room, Brighton Pavilion.The inclined members are the array of upper struts, joining the upper and lower chords (Derek Kendall).

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions JonathanClarke

PerhapsNash placed greater reliance on the inherent strength oflaminated timber chords, rather than trussing,because he had seen William Porden's remarkable Brighton stable block (Fig. 22). Erectedin 1803-8,this 80ft-diameter dome was framedusing twenty-four vertically laminated timber ribs each measuringjust 12 inches by 9 inchesat the bottom, diminishing to9 inchessquare at the top. Porden, a pupilof James Wyatt and a friendof Nash, had based his design on the celebrated dome of the Halle au Blé(Corn-market), inParis.60 That structure, erected in 1782-3by master cabinet-maker A. J. Roubo to designsby J. G. Legrandand J. Molinos, was itself the first major exemplar of the technique inover two centuries,reviving and improvingmethods pioneered by Philibertde L'Orme(c.1510-70).61 The acclaimedFrench dome even inspired Thomas Jefferson to give his home,Monticello, a laminated timberdome in c.1800, thereby introducing thetechnique to America.62 Nash himself never got to see theHalle au Bléduring his only recorded visit to Paris in 1814,as ithad already burned down. He did, however,see andadmire its replacement, a virtual facsimile in cast iron, and it is temptingtothink this mighthave inspired the elaborate iron structure framing the Pavilion's centerpiece, the great dome over theSaloon.63 Curiously the drawings for this, like so manyof the others held at Brighton, are not signed byNash, but carry the words 'Drawn by William Nixon, 1827' in the corner. Could this be theNixon, inhis final months, or else Nixon junior, putting the record straight?

Fig.22. Viewinside the central Rotunda , part of William Porden s RidingSchool and Stables, Brighton Pavilion, erected in 1803-8. This dome was probably the first large-scale use of laminated timber in Britain(Brayley : E.W., Illustrations ofHer Majesty's Palace at Brighton,London, 1838).

CONCLUSION:REGENCY STRUCTURAL INNOVATION

Theearly nineteenth century was a highlyexperimental period in the history of building technology, notjust in terms of iron construction, which was taking bold strides, but also in timber engineering. The WoolwichRotunda, or PolygonRoom, embodies a numberof progressivestrands in lateGeorgian

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions Cones, not Domes: JohnNash and Regencystructural innovation structuralcarpentry. Techniques of timber lamination, ofspecialised trussing, and of three-dimensional framingwere all harnessed to createa newtype of roof structure, thefreestanding catenoidal enclosure. To thisend, Nash and Nixonseem to havedrawn on thesomçwhat abstract example devised by Nicholson,but in taking its guiding principles into the real world, they contrived their own technique of bindingthe whole structure. The similarity inthe trussing arrangement oftheir half ribs to that depicted by BenjaminAsher in 1811 is palpable. However,independent invention spurred by similar circumstancesis more likely than any westerly transatlantic transmission of ideas. Accordingto Yeomans,New England carpenters adopted the divided tie-beam truss largely because their churches and meetinghouseswere timber walled, and hence unable to withstand the outward thrust of low-pitch roofs. Similarconditions prevailed briefly in the grounds of Carlton House, where a suiteof semi-permanent panel-walledstructures was erected. Only one of these still survives, but, besides Brighton, Nash and Nixon'sinnovative tent room may have produced further offspring, such as theTent Room of Hertford Villa(later St Dunstan's), in Regent's Park (1827) (Fig 23).04 Designed by the young Decimus Burton (1800-81),a protégé of Nash, this freestanding convex enclosure was used by Lord Hertford, "the Caliph ofRegent's Park", to stage 'his extravagant and costly entertainments ona scale that only crowned heads couldafford'.65 Unfortunately, Burton's surviving drawings do notshow the structure, but it probably reliedon divided tie-beam trusses, ifnot laminated timber.66 Other comparable structures were probably builtin thisdecade, but thereafter Victorian sobriety most likely saw the demise of the quirky, semi- permanenttent room, and with it a particularstrand of timber engineering.

Fig.23.The south front of St. Dunstan's, Regent's Park. Decimus Burton's self-supporting quadrilateralTent Room (1827), which probably borrowed carpentry techniques pioneered at Carlton Housegardens, marked the end of the line of Regency tented structures. Source : DecimusBurton, Plansof the Villa erected in the Regent's Park... ( London, c.1827), (©Architectural Association Library.)

ACKNOWLEDGEMENTS

I shouldlike to expressmy gratitude to PeterGuillery for reading a draftversion of mytext, to AndrewBarlow and colleagues at theRoyal Pavilion, Brighton, and to MalcolmTucker and David Yeomansfor sharing insights.

Correspondence:Jonathan Clarke, English Heritage, 1 WaterhouseSquare, 138-142 Holborn, London,EC1N 2ST

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions JonathanClarke

References: 1. T. F. Hunt,Exemplars of Tudor Architecture, adapted to modern habitations (London 1830), p. 97. 2. J.M. Crook,and M. H. Port,The History of the King's Works , Vol. VI: 1782-1851(General Editor H. M. Colvin)(London HMSO, 1973), p. 317. 3. M. Berry,Extracts of the Journals and Correspondence ofMiss Berry from the Year 1783 to 1852 (ed.Lady J. Lewis) Volume 2, (London.1865), p. 481. 4. Queen'sGallery, Carlton House: The Past Glories of George IV's Palace (London:The, Queen's Gallery,Buckingham Palace, 1991), p. 31. 5. (Cole2006) 6. J.M. Crookand M. H. Port,op. cit. 7. NationalMonuments Record, PSA G (CN) 21156 11. 8. 'Fêteat Carlton House', The Times , 23 July1814, p. 3. Thisdescription was reprinted inThe Annual Register, (1814), p. 64; seealso TNA: PROLC11/17. 9. ibid. 10. TheNational Archives (TNA): Public Record Office (PRO) WORK 1/9, p. 43. 11. TNA:PRO WORK 19/11/5, f.39. 12. TNA:PRO WORK 1/9, p. 61; WORK19/11/5, f.38. 13. TNA:PRO WORK 1/9, pp. 59-61. 14. TNA:PRO WORK 1/9, pp. 61-2. 15. Basedon a measuredsurvey of the timber structure, and assuming a timber density of 6 kN/cu.m. 16. TNA:PRO WORK 19/11/5, f.39. 17. TNA:PRO WORK 19/11/5, f.41. 18. ibid. 19. TNA:PRO WORK 4/23, f. 269 and f. 307; WORK 1/9, p. Ill; WO 44/642,f. 356. 20. TNA:PRO WORK4/23, f. 269 andf. 307; WORK 1/9,p. Ill; WO 44/642,f. 356. Lead was frequentlyused to hold down canvas: e.g. 'And they spread wide the floating canvas roof, And made itfast, and fixed it down with lead...', 'Victor Hugo's Legend of the Ages', Eclectic Magazine , April 1860,p. 511. 21. TNA:WO 44/642,f. 358. 22. TNA:WO 44/642,f. 368. 23. TNA:WO 44/642,f. 362 24. TNA:WO 44/642,f. 382-3. 25. TNA:WO 44/642,f. 385. 26. In a letterdated 2 May 1815,Nash informed Stephenson that 'Mr Baber's oil clothwhich covers theroof was never properly finished, and Mr Baber should be calledupon to repaint it, or at least suchparts of it as admitthe wet.' TNA: PRO WORK 19/11/5. 27. Forexample, Kaestlin noted 'the central pillar was in fact erected during the winter of 1819,after thewinter of 1819,after the structure was reassembledand boarded over, to takethe additional weightof a leadcovering'. Major J. P. Kaestlin,'Tale of a Tent',Firepower Archives: MD 939/17 (2),p. i. 28. PRO WORK43/577; Major R. St. G. G Bartelot,'A ConciseHistory of theRotunda Museum (1778-1978)',The Journal of the Royal Artillery, vol. 105 no. 2 (1978),pp. 108-9. 29. Earlyand mid 19th-centuryillustrations of the Rotunda depict a horizontalboarded cladding, visuallydistinct from the vertical roll-joints shown in theearliest photographs of the later 19th century.Furthermore, Statements of Landsand Buildingsat theRoyal Military Repository producedfor the years 1821, 1830 and 1851 describe only a boardedroof covered by canvas, (that for1840 failed to mention the Rotunda); TNA: PROWO/55/2498; WO/55/2703; WO/55/2864 and WO/55/3034. ' 30. Thedocumented decision of April 1819 that, for the perfect security of the building in the exposed

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions Cones, not Domes: JohnNash and Regencystructural innovation

situation,''...the walls should be ofgreater strength than in thesheltered spot where it originally stood'may have been accompanied by a similar(but unrecorded) one to erect a centralcolumn, on similargrounds. ThePenny Magazine (23 March1839) in itsbrief description ofthe building, offerssome corroboration forthis: 'At first [i.e. at Carlton House] wholly unsupported inthe centre, but not beingconsidered perfectly secure, a pillarwas subsequentlyerected as a central support'[emphasis added]. 31. EncyclopœdiaBritannica , 2005. 'buildingconstruction', Encyclopaedia Britannica Online, http://search.eb.com/eb/article-60125 . 32. See JamesW. P. Cambelland Robert Bowles, 'The Construction ofthe New Cathedral' in St Paul's: TheCathedral Church of London 604-2004 eds. Derek Keene, Arthur Burns and Andrew Saint (LondonYale University Press, 2004), pp. 215-17; George L. Hersey,Architecture and geometry in theage ofthe Baroque (London 2000), pp. 69-70; 134-37. 33. e.g..E. Weisstein,2005. http://mathworld.wolfram.com/Catenary.html; www.2dcurves .com/exponential/exponentialhc .html 34. D. T. Yeomans,The Trussed Roof: its history and development (Aldershot 1992), pp. 131-35. 35. Plausibly,the boarded cladding, nailed onto the upper chords, and the canvas covering, may have alsohelped restrict outward spread of the roof structure through tensile membrane action. 36. TNA: PROWORK 6/27, f. 239. 37. D. T. Yeomans,The Trussed Roof p. 171. 38. see KeithRaynor, 'British Army Bell Tentsof theNapoleonic Period', Napoleonic Notes and Queries, No. 5. (1992),pp. 20-28. 39. FrancisPrice, The British Carpenter, orA Treatiseon Carpentry, (London 2nd ed. 1735),p. 32. 40. PeterNicholson, The carpenter and joiner's assistant ...To whichare added, examples of various roofs,etc. (London1797), p. 66. 41. Editionsof 1805,1810, 1815 and 1826. 42. MichaelAngelo Nicholson, The Carpenter and Joiner's Companion inthe geometrical construction ofworking drawings ... Improvedfrom the original principles of... R Nicholson(London, 1826). 43. PennyCyclopaedia of the Society for the Diffusion ofUseful Knowledge (London 1843), p. 147. 44. JohnSummerson, The life and work of John Nash architect (London 1980), p. 10. 45. Nash,J, 1815. Autographed letter to theEarl of Caledon, Co. Tyrone,dated 25 November1815. http://manuscripts.co.uk/stock/20442.HTM 46. Appendixto thesecond report from the Select Committee on WindsorCastle and Buckingham Palace:14 October1831, as quotedin R.J.M. Sutherland, 'The Age of Cast Iron 1780-1850: Who Sizedthe Beams?', The Iron Revolution: Architects, Engineers and Structural Innovation 1780- 1880.Essays to Accompany anExhibition atthe RIBA Heinz Gallery, ed. Robert Thorne (June/July 1990),p. 8. 47. JohnSummerson, 'Introduction' inMichael Mansbridge, John Nash: A CompleteCatalogue (New York1991), p. 15. 48. T. F. Hunt,Exemplars of Tudor Architecture, adapted to modern habitations (London 1830), p. 97. 49. Fora biographicalsketch, see H. M. Colvin,A biographicaldictionary ofBritish architects 1600- 1840(3rd ed., London 1997), p. 707. Ofthis Nixon (1760-1824), The Carlisle Journal of 6 March 1824 noted 'At Chalk Lodge on the 2nd instant Mr Wm. Nixon, in the 64th year of his age; his mild andunassuming deportment through life, and his excellence as anarchitect, rendered him one of the brightestornaments ofsociety, and an honour to Cumberland'. 50. ThomasF. Hunt,Exemplars of Tudor architecture, adapted to modern habitations (London 1836), p. 97. Therewas no mentionof Nixon in the final edition of the book (Henry G. Bohn:London, 1841). 51. W.Holden, Holden 's Annual Directory for London (London 1808). 52. TNA:PRO WORK 19/11/5 f.2 (4 March1815).

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions JonathanClarke

53. TNA:PCC PROB1 1/1725; 'the probate of the late Mr William Nixon's will' was recorded as being depositedwith the Office of Works on 7 May1827. TNA: WO 4/28p. 283. 54. JohnSummerson, The life and work of John Nash architect (London 1980), p. 106. 55. Colvin,op. cit ; Crookand Port, op. cit.; Malcolm Pinhorn, The diaries of John Nash architect: 1832 and1835, Part Two, introduced by John Summerson; edited and annotated by Malcolm Pinhorn , (Calbourne2000), pp. 122-3. 56. TNA:PRO WORK 19/11/5, f.39. 57. TreveRosoman, 'Nicholson, Peter (1765-1844)', Oxford Dictionary ofNational Biography , Oxford UniversityPress, 2004. http://www.oxforddnb.com/view/article/20147 andpers comm.) 58. see PatrickConner, 'Unexecuted designs for the Royal Pavilion at Brighton', Apollo vol. 107,no. 193(March 1978), pp. 192-199. 59. e.g.Gervase Jackson-Stops, John Nash, Views of the Royal Pavilion (London 1991), p. 28. 60. EdwardWedlake Brayley, Illustrations ofHer Majesty's Palace at Brighton(London 1838), p. 16. 61. DoraWiebenson, 'The Two Domes of the Halle au Blé inParis', The Art Bulletin , vol. 54 (1973), pp.262-79. 62. W.L. Beiswanger,'Jefferson and the Art of Roofing', The Chronicle , vol. 58 no. 1 (2005),pp. 18- 25. 63. Foran appraisalof Nash's structural iron and timberwork at Brighton Pavilion, see JohnDinkel, 'Royalrestoration', Architects' Journal vol. 186, no. 27 (8 July1987), pp. 24-27;Edward Diestelkamp,'Architects and the Use of Iron',The Iron Revolution: Architects, Engineers and StructuralInnovation 1780-1880. Essays to Accompany an Exhibition atthe RIBA Heinz Gallery, ed. RobertThorne (June/July 1990), pp. 19-20;Edward Diestelkamp, 'Building technology & architecture1790-1830', in Roger White and Caroline Lightburn, eds., Late Georgian classicism : papersgiven at theGeorgian Group Symposium, 1987 (London : GeorgianGroup, 1988), pp. 73- 91. 64. I amgrateful toDr StevenBrindle for bringing this (demolished) building to my attention. 65. DonaldMallett, The greatest collector: Lord Hertford and the founding of the Wallace Collection (London1979), p. 25. 66. CertainlyBurton used laminated timber for the dome of the Colosseum in Regent's Park (1824-27, demolished1875). Although no mention of the technique is made in Ralph Hyde, The Regent's Park Colosseum(London 1982), see 'A Chapteron Curved Beams', The Building News , 22 March1878, p. 286.

This content downloaded from 132.206.27.25 on Sat, 6 Apr 2013 16:45:28 PM All use subject to JSTOR Terms and Conditions