Housing as a Source for Industrial History: A Case Study of Blaenafon, A Welsh Ironworks Settlement, from 1788 to c1845 Author(s): Jeremy Lowe Source: IA. The Journal of the Society for Industrial Archeology, Vol. 8, No. 1 (1982), pp. 13- 36 Published by: Society for Industrial Archeology Stable URL: http://www.jstor.org/stable/40968025 . Accessed: 10/04/2013 21:02
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].
.
Society for Industrial Archeology is collaborating with JSTOR to digitize, preserve and extend access to IA. The Journal of the Society for Industrial Archeology.
http://www.jstor.org
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History: A Case Studyof Blaenafon, A WelshIronworks Settlement, from1788 to cl845
JeremyLowe
In Fieldworkin Local History,W.G. Hoskins wrote that stone-walledand slate-roofedbuilding, with a total floor "housesat all sociallevels and in all periodsspeak their own area of about 400 to 450 squarefeet (Figures 3 and 12). "l testimonyof the past-if only we can construeit. Mostlythese houses were "double-fronted," with a central Workers'housing has beenaccepted as partof theprovince doorwaybetween windows on eitherside. They werebuilt of industrialarcheology, but itsrelevance is generallyseen in rows often facingacross a common access path or in termsof socialhistory, the study of laborconditions and "bailey" onto smallfront gardens. Usually the back wall domesticlife. was blank,with neither door norwindow openings, except fora smalllarder ventilator. The adoptionof thistype of The themeof thisarticle is thatthe archeology of housing house as a standardfor industrialworkers was a gradual has a widerreference, that houses can be an importantpart process.Its originswere rural;in the 18thcentury it was of those "unintentionalsources" which, as the editorof used forsmall farmhouses in East Wales.Early in the next thisjournal has remindedus, industrialarcheology is well centurysome rowsof double-frontedfour-roomed houses fittedto provide.2The ThirdInternational Conference on werebuilt by industrialistsin the EnglishMidlands (Figure the Conservationof IndustrialMonuments study group of 13), whilein SouthWales the typebecame more common; 1978 recordedthat "housingusually outlives the industry after1825 it was used forentire settlements.4 But before " withwhich it is associated, and thusmay still survive when then,in the earlyyears of the industrialrevolution, many othermore obviouslyrelevant records and evidencehave other types of workers'houses were triedout. Some of 3 beenlost. Suchhousing can be treatedin itsown right as a themcontinued or adaptedthe traditions of ruralbuildings, classof document,producing data whichis relevantto the whichwere still strongin Britain.Others were specially studyof industrialprocesses and businessdecisions. Just developed to meet the needs of the new urban and what this contributioncan be will depend on circum- industrialcommunities. stances. The contextchosen here is a case studyof an importantbut poorly documented coke- fired ironworks in At Blaenafonin SouthWales, where this case studyis set, South Wales,for whichthe housing evidence is particularly thereexisted just such a wide varietyof housing.5The richand varied. ironworksaround which the towngrew up was foundedin 1788 by threeEnglish businessmen who werenewcomers to Wales.6 The remainsof the worksthey began are very impressive,and some partsare now preservedby theState. In the middleof the 19thcentury, the most common type Documentaryevidence is relativelyscarce, though, especi- of house builtin South Walesfor industrialworkers had ally for the period of the foundingpartnership, which fourrooms, arranged two down and twoup in a substantial lastedup to 1836. The mainsources are mapsand property
13
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology deeds,with a few early lettersand wills. There is little back before1900. Wheneverpossible the actualprocess of informationabout the technicaloperation of the works demolitionwas watched, so that hidden featuresof untilafter it wassold in 1836. constructioncould be recorded.
Blaenafonis in hillycountry on the northernedge of the Mapsare stillthe mostimportant documentary source, for South Walescoalfield at the easternend of a beltof coal they providea seriesof time horizonsfrom which the and iron ore outcropsabout 30 mileslong. In 1788 the growthof the housingstock at the workscan be followed. districtwas thinly populated; most of it was poormoorland Withthem the developmentof the workscan be relatedto grazing,intersected by woodedriver valleys which deepen itshousing in fourphases: the south.The mostaccessible of ironore toward patches Phase 1 The EarlyColony: 1788-1792.The extentof had beenworked on a smallscale since the 16th to century, the worksis shownapproximately by Dadford'smap supplycharcoal blast furnaces,but the coal whichalso of the MonmouthshireCanal and its feedertram- was not of much value appearedat the outcrops practical roads,dated 1792.10 untilthe of were into the techniques coke-firing brought Phase 2 Expansion: 1792-1813/14.The numberof the nexttwo half-a-dozen area in 1757.7 During decades, furnaceswas increasedfrom two to five,and their coke furnaceswere built ironmasterswho were by English outputmuch augmented. The two-inchmap of 1813 theiractivities to Wales. These furnaceswere extending and a local plan of 1814 make the horizonfor this sitedin the because neededsources of water valleys, they period.11 fortheir The economic of power blowingengines. working Phase 3 Consolidation: 1814-1833/36.Output was theore fieldson the moorlandbecame higher possibleonly stabilizedand an organizedprogram of improvements withthe of in development steam-poweredblowing engines carriedthrough. In 1833 an attemptwas madeto sell the 1780s.The Blaenafonfurnaces were the firstin South the business,but no sale took place until1836. The Walesto be blown directsteam successfully by power.8 one-inch map of 1832 (Figure 1) is the best available.12 Blaenafonbetween 1788 In all, fivefurnaces were built at Phase 4 The Blaenavon Iron and Coal Company: 1810. werein use at leastuntil but from and They 1880, 1836-1841.In itsearly years the new company began 1860 on became to a new ironworksbuilt they subsidiary an extensivebuilding program brought to a sudden ownerson a site.That works by the same nearby produced end by a financialcrisis in 1841. This phase is to but it has now been iron, and then steel, up 1938, summarizedby the large-scaleTithe maps of completelydemolished. The earlierfurnaces survived, as did 1843-1 846.13 some of the around them,though they were buildings Afterthe crisis of which the robbedof muchstonework. In 1969 were 1841, nearlybankrupted they recognized all further investmentwas concen- as one of the classicironworks sites in South Wales,and company, majorcapital trated on the "New" site, where the furnaceswere laterthey were taken into State guardianship. They are now eventuallybrought into use in 1861. beinginvestigated and conserved,and will be a splendid showplace.9
Unfortunately,only a smallarea aroundthe furnaceswas Phase1: Housingin theEarly Colony: 1788-1792 givenState protection.Almost all the workshousing was outsidethis area, scatteredamong the old drifts,quarries, Contextand Data and spoilheaps (Figure 2). Some of thesehouses were still occupied in 1969, others were derelict,but all were scheduledfor demolitionas unfitfor habitation. Though Thoughthere is no large-scalemap of theBlaenafon works the full historicalvalue of such buildingswas scarcely in its earliestdays, its layoutcan be reconstructedfrom a appreciatedthen, some field records were made duringthe viewof the furnaces(Figure 5) drawnin 1799, and from long processof clearance,which continued up to 1978. the small-scale(1:25,000) map, dated 1792, of the tram- Informationabout the houseswhich had been destroyed road fromthe worksto the MonmouthshireCanal.14 The beforefield workbegan was gatheredfrom maps, public originalworks consisted of two stone-builtfurnaces with healthrecords, a fewphotographs, and some recollections the steamblowing engine sited between them. Number 2 of elderlyresidents. In one case thisoral evidencewent Furnacesurvives, probably much rebuilt. The furnaceswere
14
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History
cl829. One-InchOrdnance Survey (Sheet Figurel.Blaenafon, " Enlargedpart of 42 "Brecon, published1832); parishnames and boundariesadded.
15
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology builton one side of a naturalhollow, against a rockface createdby stone quarrying.The 1792 map shows that aroundthem there was a scatterof half-a-dozenbuildings. Fiveof thesecan be pickedout on laterlarge-scale maps as blocksof workers'housing (Blocks A to E on Figure2) but thesixth remains unidentified. % iitir-fyy~t "'zx&i;=f n Therewere three types of housein all. BlockA consistedof four-roomhouses of the commontype. It stillexists and ;■•i %¥ ' ré^^=^Ñ^m>. has been studiedin fulldetail. Blocks Bl, C, and D wereall back-to-backhouses. FortunatelyBl surviveduntil 1972 whenit was investigatedarcheologically during demolition. ~ C and D weredestroyed long ago but mapsare a guideto T2 Bl theiroriginal form. Block E was stillstanding as recentlyas Key 1962, but it had been so muchaltered by then that its originaldesign is uncertain. Blocks Bl and B2 (BunkersHill or BunkersRow) were bothdemolished in 1972. Bl is markedon the 1792 map. Until 1970 it was thoughtthat both Bl and B2 were reconstructionsof the mid-19thcentury, but thenit was noticedthat withinand below the visiblestructure large partsof earlierbuildings survived. Archeological investiga- tion during1971-1972 recoveredalmost everyessential detailof theseearlier buildings.17
The easternend of Bl was the firstto be built.Although the block stands perpendicularto the cart-roadwhich linkedthe ore mineswith the furnaces,the /is*,.'' **.- s*T .."^' s£-*3T early ground belowits floorsand foundationswas clean and undisturbed (Figure 14). In contrast,the groundbelow block B2 was Block A (Stack Square and EngineRow) is close to the heavilycontaminated with small pieces of coal and ironore furnacesand its U-shapedplan is easilyrecognized on the whichhad spreadacross from the road. EvidentlyBl was 1792 map. The two parallel rangeseach contain four built beforethe transportof materialsto the furnaceshad substantialfour-roomed houses of the commontwo-story, begun.It contained20 one-room-and-loftdwellings, mere double-frontedtype (Figures 3 and 12).15 The linking cabins measuringabout 10 by 11 feet internally,built cross-rangehas anotherof thesehouses at its upperend, back-to-back,separated by four and one-halfinch brick then a buildingwhich, in 1819 if not earlier,was the walls, witha smallfireplace set acrossone cornerof the companyshop, and at its lower end, a largethree-storied onlyroom, a door and a littlewindow in the outsidewall, buildingwhich might have been the works manager's house, and a sleepingloft contrivedalmost entirely in the roof or the worksoffice.16 Although they are of thesame type, space, but witha smallventilation opening at floorlevel the houses in the two parallelranges vary in size and (Figures3 and 6). As withBlock A, theoutside walls were quality.In theupper range, the two end-houses have a floor built in rubblestonework, but at one angle therewas a area of 615 squarefeet and fireplacesin bothground-floor brickquoin whichshowed that originally the buildershad rooms,while the center pair have an areaof 525 squarefeet intendedto use brick.18The constructionshowed technical and one fireplaceon each floor.In the lowerrange, all the competence,but slapdash execution, as ifthe builders knew houses measure 475 square feet and have only one well what theywere doingbut cared not at all about its fireplace.The whole block,including the crossrange, has accuracy.Everything was as cheapas possible.There was no rubblestone walls about 22 inchesthick, steeply cambered carpentryjointing; plain lengthsof low-costtimbers like fourand one-halfinch brickarches over the ground-floor poplaror willowwere used, except possibly for structural windowsand door openings,jointed oak beam and joist purlinsof elm. The rafterswere unsawn poles, and theloft floors,and oak trussedand bracedroofs. floors,which had four-inchsquare joists about threefeet
16
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History
I i I Q> «O rsT *^ I 1 Q$ _^ <2> oo ^Q
"^ <^ .s| •s¡ Ì8
s § itti
2l e.11 -a
17
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology
' [ 1 I I . ! A First Floor Second Floor Section
-4^---4- fíJS c B1 èi si [1 InlnlI ?
First Floor Second Floor Section |
Cellar First Floor Second Floor Section I ■
First Floor Second Floor Section
First Floor Second Floor Section
^ I - - ilii iV
First Floor Second Floor Section Figure3. Blaenafon: house types. For details see text.
18
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History apart,were too lightlyconstructed to bear any bedroom furnacewith a water-poweredblowing engine, necessarily furniture(even if it could have been taken up the sited in one of the valleysof the coalfield.The steam- ladder-typestairs). powered cylinderblowing engines developed by Isaac Wilkinsonand James Watt in the 1770s freedthe iron industryfrom this dependence on waterpower. Steam was rapidlyadopted by Englishindustrialists for severalnew ironworksin the Black Countryof Staffordshire,to the westof Birmingham,where coal and ore wereplentiful but waterresources insignificant, and withina decade the new sourceof powerwas broughtto Wales.21 *% CJ Dù All threeof the partnerswho set up the Blaenafonworks came fromthe Birminghamregion and Thomas Hill, the seniorpartner, had extensiveinterests in Staffordshire.22 Map evidencesuggests that Blocks C and D weresimilar to The new workswas obviouslyan applicationof Stafford- Bl, thoughperhaps different in the actual dimensionsof shire technologyto South Wales, but what was the each house.Block C (CoaltarRow) was sitednear some of connectionbetween the two districts?A late 19th-century theearly coal levels.19It was demolishedbefore 1910, but memoirsuggested that the partnerswere so dependenton the 1844 Tithe map (1:2,376 scale) showsthat it was a MidlandEnglish skills for the constructionof the furnaces back-to-backrow too, similarin widthto Bl. The western that the vitalhearth-stones were carriedall the way from side was used forworkshops in 1844, but the easternside the BlackCountry by bargeand cart.23A specialroad was was still used for housing.Block D, whichstood in the built,it is said,for some four miles across the hills to make furnaceyard, was demolishedbefore 1880. It too was a such a movementpossible. This traditioncannot be tested back-to-backrow withhouses on one side and, by 1844, by lookingat theindustrial remains for nothing survives of officeson theother. the 18th-centuryfurnaces in Staffordshirewhich would have been the prototypefor Blaenafon.Even if a com- Block E (Stable Row) was sited on the edge of a small parisoncould be made it is debatablewhether an industrial paddockcalled Cae-White(Mr. White'sfield) whichhad structurelike a furnacewould revealidentifiable regional beenleased by the ironworkspartners.20 When demolished features. about 1966, it consistedof housesof severaltypes, but the originalnucleus was probablya rowof six dwellings,about On the otherhand, the Blaenafonhousing of 1788 does half-wayin size betweenBlocks A and Bl . possessregional features which can be seen and recognized today.The archeologyof BlocksA and Bl definitelyshows Discussion of Phase 1 that non-Welsh-probablyMidland English-builders were at Blaenafonin the 1788-1792.In BlockA The documentarysources, and the ruinsof the furnaces- working period it is the which has a non-Welshcharacter. The the"normal" data of industrialhistory- show well enough masonry solid of the floorsand roofscan be matchedin thatthe Blaenafonworks in its firstphase was a sizeable carpentry every featurewith local Welsh but the industrialdevelopment using a new technology(steam house-building, chunky is not in the sametradition. In blowingpower) brought to SouthWales by outsiders.But masonry particular, the cambered, brickarches are a telltale was it an independent"colony" unrelatedto thelocal iron steeply single-ring characteristicof Midland 12 and and industry?Who actually built the furnaces,and who England(Figures 13), evidencethat the masons on BlockA came operatedthem? What was the scaleof operation,and did it positive working fromthe northor westof in requirenew managerialmethods? The data fromhousing region Birmingham.Possibly, fact probably,these masons built the furnaces, a few areinformative on all thesepoints. only yardsaway.
Buildingthe Works Block Bl was even more alien to South Wales in its Untilthe late 18thcentury the ironindustry was generally character.In 1788, the back-to-backhouse was onlyjust dependenton waterpower. In South Wales the typical emergingas a developingbuilding type. The firstknown ironworks,whether charcoal or coke fired,was a single examples(destroyed many years ago, but identifiedfrom
19
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology propertydeeds) had beenbuilt in Birminghamin 1774, and The firstBritish data forhouse occupancyare given by the othersin Nottinghamin 1775.24The typehad beenevolved 1841 Census-ratherlate, but the best available. This census to suitthe congested conditions found in suchlarge towns, showsthat in smallhouses, like Bl , therewas an averageof so it was not muchused in SouthWales, even later in the 2.2 workersper house,and in largerhouses an averageof 19thcentury. Everything about BlockBl indicatedthat it 2.7 to 3 workers.Block A wouldhave housed about 25 to was a cheap and quickjob carriedout by a house-builder 30 workersand Block Bl about 45. Thesewould be actual absolutelyfamiliar with the back-to-backtype and with workers,not includingany non-employeddependents.27 brickconstruction. Such a mancould onlycome from one The maps of Blocks C and D show that theycontained of thelarge towns of England,most probably Birmingham, between20 and 36 smallhouses, depending on whether thoughthe actual place mustremain uncertain because no theywere built as theyappeared in 1844, or originally(as comparablebuilding has survivedin England.25The differ- seemsmore likely) with houses on bothsides. They would ences betweenBl and B2 confirmthis deduction. If B2 is havehoused another 45 to 80 workers.Block E wouldhave the workof local builders,modeling their techniques and takenabout 15 workersand another1 5 or 20 could have dimensionson Bl, thenBl itselfmust have been the work lodged in local farms.The limitsof the workforcethus ofan outsider. rangefrom a minimumof 140 to a maximumof 190. At Beaufort,a single-furnaceworks set up in 1778 fivemiles westof Blaenafon,the workforcewas 70 to 80 menso the Blaenafonestablishment marks a bigincrease in thescale of operations.28 TheLabor Force The number,size, and type of the houses built by the The sitingof the individualblocks seems to be relatedto foundersof the worksis an obviousguide to the size and thevarious workplaces, and givessome idea of thenumbers structureof the laborforce they intended to employ.The engagedin each part of the works.The skilledfurnace earlierwater-powered single-furnace works were sited in the craftsmenfor whom Block A musthave been intended were valleyswhere farming settlement was also concentrated, relativelyfew. The provisionof a gradedseries of houses, and so theywere dependent on the samelabor resources. withtwo (or four)of each size,shows that it was expected The charcoal furnaceshad a seasonal workingcycle, thatthere would be a definiteorder of importanceamong operatingonly through the fall and winterwhen the call for them,and thateach of thetwo furnaces would have its own farmlabor was at its lowest,and thecoke-fired works were workingteam. House size at that time was much more also liable to be slowed by shortagesof water in the closelyrelated to statusthan to the expectednumber of summer.26The ironmastersemployed their key furnace occupants:"Twelve feet is a widthsufficient for a dwelling workersall the yearround, and builthousing for them (of thatis to be deemeda cottage;if it be widerit approaches whichnothing remains). They do not seemto have made too near to what I would call a house for a superior any special provisionfor theirlaborers, who were still tradesman."29The housesof BlockA wereall 15 feetwide. treatedas casual workersand lodgedin the farmhouses,or Of course,there was no questionof eachcraftsman having a in primitivecottages built on wasteland. house forhis own exclusiveuse. It was normal,as late as 1841, to find lower-statuslodgers in craftsmen'shouse- Blaenafonwas too isolated,and theworks too largein scale holds.The laborers'houses were also relatedto workplaces. to fitinto thispattern. The highcapital cost and working Block D, in the furnaceyard, would have housed at most expenseof the steamblast engine could only be met by 30 laborersto attendthe furnaces.Block Bl stood by the continuousworking, which required the employmentof a road leadingto the higheriron ore mines,and BlockC was large permanentworkforce. It would have been quite close to some of the coal workings.30Together these two impossiblefor all theselaborers to findlodgings in the 15 rows could have housed 90 to 100 minersand colliers. or 20 houses scatteredaround Blaenafon.The housing Block E was sited next to the paddock,Cae-White. Later evidence,the number of buildings,the way they are placed there were stables next to the houses. Was this row aroundthe works,the use of the non-localback-to-back intendedfor the workerswho looked afterthe partners' form,the relationshipbetween the pairs of houses in horsesand mules? Block A and the buildingof two furnacesall indicatethat Blaenafonwas a completeindustrial colony, planned from The size differencesbetween the houses may indicate more thestart to housethe workers needed. than the statusof variousworkers. In the 19th century
20
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History therewere still craftsmenof MidlandEnglish origin at up much of the nationalsurplus in taxationand govern- Blaenafonand at Ebbw Vale in 1796 key workershad ment borrowing-and in part to the sheerextent of the Englishnames.31 The large,multi-roomed houses of Block partners'mineral concession, which seems to have proved A wouldhave beena necessaryinducement if the partners richerthan they had expected. Even before the war, in 1791, intendedto attractworkers from the comparativelyurban- Hill had soughtfunds from the Bristolbanking firm of ized settlementsof the EnglishMidlands to theisolation of Harfordand Co., to help himbuild a secondironworks at the Welshhills, where even the everydaylanguage was Nantygloon the westernedge of the concession.That unfamiliar.32On the otherhand the dwellings provided for ventureturned out badly because a disputebetween the the furnacelaborers, miners, and collierswere so much Blaenafonpartners and Harfordskept the Nantygloworks inferiorin qualitythat the recruitmentof theseworkers out of use and tied up the fundsHill had spentthere. An was obviouslynot a problem.Some wouldhave been the attemptto startoperations in 1802 failed,with a further local minerswho up to 1786 had workedthe ore veins on a accumulationof useless debts.36The partnersmay also seasonal basis for the charcoalfurnace at Llanellyfour have lost moneyin the failureof the VartegForge, which milesto the north.Others would have beenattracted from was two milesdown the valleyfrom Blaenafon. It worked the undevelopedrural areas of South and West Wales, only from1804 to 1811, but the reasonfor its closureis hopingto findregular work. They would have been inured not known.37Most of the housesbuilt in thisperiod were alreadyto the poor housingconditions which were the demolishedbefore 1970. Onlyblocks B2 and K have been commonlot of farmlaborers in Walesat thattime, and recordedmore recently. Yet even thoughwe have to rely would,no doubt,have foundthe housesof BunkersRow heavilyon mapsfor information about housing of thistime, quitebearable.33 theyshow that there was a widevariety of housetypes and sizes,and a blockof plans.
Phase2: HousingDuring the Block B2 (BunkersRow) stood on the partners'leasehold Periodof Expansion:1792-1813/14 land on theother side of themine cart-road from Block Bl. Probablybuilt soon after1792, and demolishedin 1972,it Contextand Data contained14 back-to-backhouses, built just as roughlyas Two maps dated 1813 and 1814 providethe next stock- Block Bl but usingstonework throughout. The thicker takingpoint. The earlierone showsthe whole district at the internalwalls reduced the usable area of each house relativelysmall scale of 1:31,680.The othercovers only the slightly.The roof was moresteeply pitched and covered area roundthe furnaces,at a scale of 1:2,500. Fromthe withslates (Bl had been roofedwith stone slabs). Further documentarysources relating to thisperiod we mayinfer extensionswere added after 1805 to both blocks; six thatthe steady growth of theironworks was interruptedfor houseswere built onto Bl and eightonto B2. The datingof a whileby insecurity of tenureand a shortageof capital. these extensionshas been establishedby dendrochrono- logicalstudy of the rooftimbers.38 Later still, but before The thirdfurnace (on the left,Figure 5) was added before 1812, rangesof detachedlarders, with brick vaults, were 1794.a4 Then,although the output was probablyincreased, builton bothsides of thetwo blocks. therewas a lull in the buildingprogram. The wholeof the firstdevelopment of 1788-1792had been undertakenon land whichhad been leased for only 21 years,with an optionto extendfor seven more. This gaveno securityfor furtherbuilding either of industrialplant or of houses.Not until1806 was a long-termlease obtained, for 63 years,at a muchhigher rent.35 Two morefurnaces were begun soon afterwards,and broughtinto productionabout 1809 and 1810. Outputrose from4,318 longtons in 1796 to some 13,880long tons in 1813/14.It was notjust theshort lease whichcaused a problem.From an earlydate therewas a shortageof capital,which can be attributedin partto the early deathsof Hill's originalco-partners, in partto the generaleffects of the NapoleonicWar- which was soaking
21
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology
Block G (Stable Yard) was sitedon freeholdland acquired the houses to the boundaryof Leigh's land, as close as in 1801, and was demolishedabout 1900. It contained19 possibleto the Blaenafonfurnaces. The basicform (Figure houses in three ranges,each consistingof a different 3) of the ground-floorplan- a frontkitchen, with a small house-type.The stables formedthe fourthside. The bedroomleading off it at the back,and a staircasein the westernrange consisted of single-frontedhouses similar in adjacentcorner- is foundin manysingle-fronted houses of plan proportionsto the standardhouses built later on, the time,but featureslike the cellar,the big upper-floor duringPhase 3 of thedevelopment of theworks. sleepingchamber open to the roof, and the unusually substantialconstruction gave the houses a quiteindividual BlockH (RiverRow) stoodon theedge of thesame field. It character.The roofs,supported by massivebolted A-frame hadjust beenstarted in 1814, buteventually extended to a trussesof oak, wereparticularly interesting and impressive. lengthof 18 smallhouses, averaging about 290 squarefeet total floorarea. Nothingelse is knownof thisbuilding, whichwas abandonedsoon after1880 and buriedby waste tips. Discussion of Phase 2 By 1799 there was a serious shortageof housingat Blaenafon-"a want of habitationsfor the increasing numberof families"wrote Archdeacon Coxe afterhis visit. Since 1789 the workforcehad increasedto 350 men.40All the housingbuilt by thepartners up to thatdate- that is to say BlocksA, B1/B2(without their extensions), C, D, and E- could have accommodatedat most some 215 workers. Even of the S •' W ^V.-ifcS2 o/^^A/xn y by the low standards the day, workerswould fin havebeen badlyovercrowded. The firstbuilding of houses on freeholdland did not take placeuntil after 1801, and it is unlikelythat on the leaseholdrestarted before Block Jl (NorthStreet) stood on the leaseholdland, and building 1806-for even the extensionsof B1/B2 wereheld back was demolishedbefore 1962. It was an irregularrow of untilthat year. Coxe tells us whatwas beingdone. Apart housesof varioussizes, mostly of thecommon four-roomed fromsome dwellings formed under the arches of a tramroad double-frontedtype. Block J2 continuedthe line of Jl on bridge,an "ingeniouscontrivance" which he illustratedin the other side of the Drum and Monkeypublic house his book (but whichhas neverbeen located on theground), (whichafter 1844 was takenover as the store).It company he saw houses all aroundthe works area: "the was demolishedabout 1966. It was more in form springingup regular hollowsof the rocksand sidesof thehills are strewedwith than Jl, and consistedof houseslike those in Block G, numeroushabitations." The Tithe Surveysof 1843-1846 similarin planarea to thelater standard house. show that these houses were built singlyand in small groups;they varied greatly in size and type.Since clearly theywere not company-built,they must have been erected by independentminers and colliers,most of whomwould have workedas sub-contractorsto the partnersrather than as theiremployees. To protecttheir mineral rights, and to increase their property,the partnershad to retainan interestin these buildings.One arrangementmight have been for the partnersto sub-leasea sitefor a shortterm, suchas 21 years,on conditionthat a house be built.Since the partners'own tenurewas so short,such an agreement Block K (Shepherd'sSquare) was differentfrom all the would be by wordonly, possibly confirmed by a note in otherrows in thatit was builtby an adjoininglandowner the partners'day-books (now lost). Whenthe termof the and leasedto the partnersfor £50 per year.39The builder lease expiredthe siteand the house wouldbe surrendered was Capel Hanbury Leigh, the ironmasterof nearby to the partners,and the householderbecame a normal Pontypool.The buildingswere recorded, but onlyin part, rent-payingtenant.41 Some houses in the centerof the duringtheir demolition in 1970. The blocklayout brought worksmay have been builtin thisway. Block Jl, which
22
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History straggledalong the line of the tramroadto Pontypool,had riverat the VartegForge. The Forge Row there,almost theplan irregularity which would be theresult of individual certainlybuilt in 1804, is now in the care of the Torfaen uncoordinatedconstruction. MuseumTrust (Figures 3 and 8). It is a maturedesign using the four-roomedsingle-fronted plan which was probably For a fewyears after Coxe's visitthe workforcemay have imported,like BunkersRow (Bl), directlyfrom the towns stabilized,at least untilthe lease had been renewed.Then of the EnglishMidlands.44 It is a buildingwith some theremust have been a bigincrease as the fourthand fifth aestheticcharacter. At thatdate it was quiteexceptional to furnaceswere constructedand broughtinto use about findin Welshworkers' housing such well cut and laid stone 1809/10.Though some houseswere built at thistime, for work,with neat archesover all ground-flooropenings and The examplethe additions to BunkersRow (B1/B2),the Stable with a definite small plinth and stone cornice. Yard (G) and BlockJ2, they were too fewto accommodate constructionsystem is stable and efficient,using plain all the new workers.It is clearthat the shortage in 1814-18 purlinedroofs withouttrusses (thus quite differentfrom was as bad as it had been in 1799, or maybeeven worse. ShepherdsSquare) and framedtimber separating walls Thereare several of urban Eventuallythe partnersdid mountan extendedprogram of betweenalternate houses. signs Each househas frontand back house-building(this comes in our Phase 3), but theirfirst influencein the plan. doors, the to the back door as a larder.Such expedientwas to seek helpfrom C. H. Leighand it washe with passage serving who financedthe constructionof ShepherdsSquare (K). access fromboth sidesof therow was reallyonly necessary are rarein districtsof The £50 per year rent, at five percent(the normal in towns,and back doors country therewere maximumof the time),sets the maximumcost of the 17 SouthWales. On the otherhand, in thecountry housesand land at £1,000or £55 perhouse. This was good no retailmarkets or shops,so a good larderwas important. eventhe valuefor such solid construction; possibly some materials- Forge Row is somewhatdefective in thisrespect; The stone,bricks for the internalwalls-were supplied by the poor housesof BunkersRow weregiven better larders. Row are anotherurban feature. Blaenafonpartners. £60 is a commonlyquoted pricefor large windowsof Forge of a housesin Britishindustrial villages at thistime.42 Glasswas expensiveand heavilytaxed; in 1806 thecost windowwas about threetimes the costof thesame area of narrow Furtherevidence of the pressureon the partnerscomes stone wall.45 Large windowswere desirablein fromthe decisionto extendBunkers Row (B1/B2) after streetsand courts,especially where industrial outwork in 1806. The design of these houses was by then quite the home was usual, as in textileworking areas, but they out-of-date.The extensionsdid have improvedheadroom, werenot so necessaryon ironworkssites. especiallyon the upperfloor, and the floorstructure was morerobust, allowing the roomsto be furnished,but their of the larders floorarea was not increased.The addition Phase3: HousingDuring the soon afterthe extensionsshows that the inadequacyof all Periodof Consolidation:1813/14-1836 thesehouses was before1812. recognized Contextand Data BlocksG and J2 show,on the maps,the irregularityand The mapsavailable for this period divide it intotwo parts. varietyof planwhich seems to be a generalfeature of Phase For the firstthere is a large-scaleplan (1:3,050) dated 2, thoughfrom their plan area and shapethe houses appear 1819. Thoughroughly drawn, it showsmuch of theworks to be of betterquality than B1/B2.They wereprobably area, and marksindividual rows of houses.46The only builtquite late in the periodand once againthey seem to suitablemap for the rest of the period is the One-Inch be relatedto placesof work.Stable Yard (G) was sitednear OrdnanceSurvey (1:63,660) which was surveyedduring grazinglands in the valleybottom. Its constructionshows 1828-1829and publishedin 1832. It is a beautifulwork, how necessaryto the workswas the care of its transport drawnwith astonishing clarity (Figure 1), butat thatsmall animals.43 scale the detailof individualrows of housesis not always reliable. Despite all the pressureon resources,the tendencyfor standardsto rise is discerniblein much of the house- Thereare stillonly a few documentarysources, of which buildingof thistime. Shepherds Square (K) was solidand the most usefulare the sale particularsof 1833, and the quite roomy.The same is true of anotherhigh-quality recordof iron shipmentson the MonmouthshireCanal.47 housingdevelopment which is sitedtwo milesdown the (The use of the latteris discussedbelow.) In 1813/14the
23
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology
Blaenafonworks were prospering.The partnershad sold housesis builtdirectly over another.50 The housesin both theirinterests at Nantygloin 1811, and by sublettingpart rows were the commonfour-room double-fronted type. of theirmineral rights had reduced theirown rent by The lowerones wereentered from North Street. Their back £2,075.Five furnaceswere in production.With the market walls were builtagainst a heap of wastetwo storieshigh, foriron still buoyant, the business must have been showing the top surfaceof whichwas level with the "ground" floors a good workingsurplus. In 1815 SamuelHopkins died. His of the upperhouses, which had theirentrance doors in the sharepassed to hispartner's son, Thomas Hill Jr., who took blankback-wall at thislevel. theleading part in the management.48Under his direction, theoutput of ironwas not increased-in facttotal tonnage Block N (Quick Buildings),on the freeholdland, was built declinedslightly- but investmentwas directedtowards before 1819 and demolishedabout 1910. Its designwas more efficientand economicalworking. In 1816/17 a probablyone of the steps towardsthe evolutionof the water-poweredforge was builtat Garnddyrys,on a hillside standardhouse. There were 33 houses in all in the row. well to the northof the furnaces(Figure 1). This forge, They were about 15 percentsmaller on plan than the which had been needed for some time,was capable of standardhouses; people who rememberedthem said they producingabout 100 tonsof valuablebar-iron each week.49 were "like the New Ranks (Blocks Q and R) but a bit Hillcompletely reorganized the tramroad routes around the lower." works,constructing a tunnel that linked the furnaces to the northernworkings at Pwll-duand the Garnddyrysforge. Nearerstill to thestandard house type was the Lower Rank The tramroadwent on fromthe forgenorthwards to the at Pwll-du(not shownon Figure2). Thisrow too was built canal system.Several rows of houseswere built at Pwll-du before1819 and it was demolishedin 1963. It consistedof and Garnddyrys.At the same time new coal and iron threeterraces, each of ten houses,joined end-to-endto workingswere opened up at Blaenafon,to the westof the follow the line of the limestonequarry tramroad. The furnaces. appearanceof this row has been workedout fromthe familysnapshots and memoriesof two formerresidents House-buildingin the earlypart of this periodcontinued (Figure7). Like Shepherd'sSquare (K) thesehouses had a the generaltrends of the previousphase, with a steady frontkitchen, a ground-floorback-bedroom, and a big growthin the amountof building.Different designs were singleroom upstairsopen to the undersideof the roof. stillbeing tried out, but the numberof housesin each row There wereno stairs,just a ladderto the upperfloor, and was greater,and the layoutmore regular, than in Phase 2. no cellar, the larder being where the stairs were in Eventuallya standarddesign was evolved,and from1820 ShepherdsSquare. onwardsthis was theonly typeused; standard houses were built fiveat a timein regularrows, usually of 10 or 20 houses.
Block L (NorthStreet and StaffordshireRow) was one of theunsuccessful designs. Remarkable to look at, it had no It was before1819 • •" effecton subsequenthousing. begun . n V. ^ wJ ¡ M 'yJJ iJaL-L-V0 ¡SS [r^-fj and was demolishedin 1963. Unlikeany other row at Blaenafonit used the "dual" principlein whichone rowof
24
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History
7870 7820 7830 7840
Last in this period is the standardhouse itself.Block Q Figure4. BlaenafonIron Works:iron shipmentsand housing New BlockR New andBlock stock, 1806-1841. Iron shipments:Average tonnage,heavy line (Upper Rank), (Lower Rank), includes correctionfor extra value of wroughtiron. Housing T (Ten Houses)all followedthis design (Figure 3).51 They stock: Cumulated to the year shown; separate indicationof werethree-room houses on the linesof ShepherdsSquare houses of two, three,and four or morerooms; close shading shows additions since wide shows (K), witha frontkitchen 12 feetsquare, a smallbedroom previouscalculation, shading stock thenexisting; "standard houses'' are cross-shaded. behind,the larderbeside it, and a bigsleeping space over thewhole house, open to the undersideof the roof.There wereno cellarsand no stairs,just a simpleladder behind the buildingof the Garnddyrysforge increased the value,and frontdoor. The designincluded a littlearchitectural quirk, thelabor content, of a substantialpart of theoutput. In the whichoccurred in all knownstandard houses. The position diagram (Figure 4) each ton of iron shipped from of the upper-floorwindow alternates: in one house it is Garnddyrysis shownas equivalentto two tons of pig-iron abovethe ground-floor window, and in thenext it is above fromBlaenavon.52 Because the shipmentsfluctuate con- the door (Figure 9). This patterncontinues for the full siderably,output is calculatedas a five-yearaverage, plotted lengthof each row and gives a kind of architectural in the fourthyear. The diagramshows that with the signatureto the standardhouses. It occurredin Lower allowancefor wrought-ironproduction, output continued Rank,Pwll-du, but not in ShepherdsSquare or ForgeRow. to rise until about 1825. The labor force must have increasedtoo.
Discussion of Phase 3 Output doubled between the turn of the centuryand At the beginningof Phase 3 the demandfor housingat 1813/14, and went up another 35 percent by 1825. Blaenafonwas stillnear its peak. Thoughthe amountof Correspondingly,the labor forceof 350 workersin 1799 house-buildingwas increasedby Thomas Hill Jr., the would have risento about 750 in 1813/14and thento a workforcewas still gettinglarger. There was a slight maximumof around 1,000 in 1825. The diagramalso decreasein the actual tonnageof ironshipments, but the shows changesin the partners'housing stock. They had
25
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology
Figure5. Blaenafon Worksin 1 799. From W. Coxe, An Pwll-du, 63 at Blaenafon, and ten at the outlying Garn- HistoricalTour in Monmouthshire 1 (Brecon, 904). yr-Erw coal pit. Houses of this kind would have cost between £60 and £90 each, so the whole programrequired an averageyearly expenditure of about £1,000 to £l,500.54 This was not a large outlay, seeing that it was said in 1833 owned or controlled about 200 workers' houses in that the works made a profit of £8,000 per year "on 1813/14. During Phase 3 they added another 235 to this pig-ironalone."55 Hill does not seem to have triedto satisfy total. When the works was put up for sale in 1833, there the demand for houses by a crash program of were about 430 workers' houses, one for every 2.3 building. Outside agencies still played a part in providing workers.53 them. In 1826 Thomas Deakin, the works manager,built Block P (Deakins Row)- it also had the "window signa- The ad hoc characterof house-buildingin Phase 2 is still ture"- and by the edge of the open land at Garnddyrys quite evident in the early part of Phase 3. A new designfor there still exist the ruins of about 25 houses which were River Row (H) was followed by another buildinginnova- built by the adjoining landownersto meet the demand for tion in North Street (L) and a third for Quick Buildings homes near the forge. (N). With the buildings of Lower Rank, Pwll-du, about 1818, the choice of designs settles down. The window- The standard house was an efficientdesign, much better patternsignature of the standardhouse appears, as well as than the dismal two-roomboxes built at Nantygloafter the its general form and plan. From this time onwards end of the Napoleonic Wars,or the inefficientthree-room house-buildingactivities take on the characterof a coherent "catslide" houses preferredat MerthyrTydfil.56 It was program.Within the next ten years 158 standard houses robust and simple to construct. Some of the rows were were built,35 at Garnddyrys,50 (includingLower Rank) at occupied for 150 years and, though without modern
26
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History
Figure6. BunkersRow, Blaenafon in 1970. (Block Bl) south side: whiteoutlines show facade of originalbuilding of 1 788-1789. All photographsby author unlessotherwise stated.
Figure7. Lower Rank: Pwll-du,c 1960. Drawingreconstructed from snapshots lent I L by W.H. Wedlock.
27
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology
Figure8. Forge Row, VartegForge, cl971. wide varietyof householdsliving in these rows. Some Rear view back door showingpaired houses were occupied simplenuclear families of the openings(nearest one blocked). by wage-earner,his wife,and theirchildren. Others had large mixed populations.In Lower New Rank (R) fivehouses heldrespectively 11, 11, 12, 13, and 13 people,and a house in UpperNew Rank(Q) was occupiedby a marriedcouple drainageand services,were still warm and dry.The standard withseven lodgers. house did not need any elaboratemasonry or carpentry. Manyof its materials-stone, bricks, lime, and sand-were The evolutionof the standardhouse has beensketched out the normalraw materials,or products,of the ironworks. onlybriefly here. Obviously it drawson local traditionsas Thus the fireplaceopenings were bridgedby standard did ShepherdSquare; the Gellidegrow at MerthyrTydfil, ^section cast-irontramplates, often showing considerable built before 1814, has the same plan-formand wide wheel wean It would have been quite possiblefor the spreadingroof.57 It also shows outside influence,both ironworkscraftsmen to supplyall the skillsneeded. The throughthe precedentof ForgeRow and in theway that it designwas also suitablyflexible in use. It wasnot lavish as uses rationalalternatives such as the tramplatelintels, to some"philanthropic" works housing was; theladder to the traditionalcraft solutions. From the pointof viewof this upperfloor was a verybasic feature. But the arrangement articleone of its main featuresis the way in whichit of the sleepingspaces was a realisticanswer to the needs illustratesthe new-stylemanagerial control exercised by and attitudesof the day. A marriedcouple could havethe ThomasHill overthe old craftsof building.It is a genuine privacyof thedownstairs back room, but all othermembers industrialproduct, rationally developed to meet a well- of the householdslept upstairs. The 1841 Censusrecords a definedneed.
28
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History
Phase4: Figure9. UpperNew Rank, Blaenafon 1971: "standardhouse" The BlaenavonIron and Coal Company:1836-1844 (Block Q), typical facade withalternating window pattern. Contextand Data The fourthphase ends formally with the TitheSurveys of Harry Scrivenor,the historianof iron, took over. All 1843-46. Practicallyit ended in 1841 with the near- buildingwas stopped,and no morehouses were built until bankruptcyof the BlaenavonIron and Coal Company. about 1860. Thereis virtuallyno datafor the last three years of thefirst partnership.The new companybought the worksin 1836 AlthoughAshwell repaired many houses he builtonly two and the minutebook of its generalmeetings was begunin newrows. Both have been fully recorded. March 1837. The manager,Mr. Ashwell,was an engineer who foundmuch in the company'sproperty that needed alteration.As the mainlease had only33 yearsleft to run, he beganto builda new works,with three furnaces, on the freeholdland. He also spent freelyon the leasehold propertywhich he found"in a mostdilapidated state."58 He used architectureopenly to expressthe new company image.New workshopson theold sitewere given a Roman pedimentand the waterbalance lift he builtnext to the furnacesstill survivesas a splendidpiece of masonry. Almostinevitably he overextendedthe company'scapital Block F (VictoriaRow) was builtin 1837/38and was not resourcesand hisrule ended in a financialcrisis in 1840/41. demolisheduntil 1978. It stood on the freeholdland and
29
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology consistedof 24 housesarranged in a "Palladian"composi- prioritythan adequate internal planning. Ashwell showed a tion.There was a centralpedimented projection, containing more practicalside to his philanthropyby includingthe two houses, and end pavilionsmade by projectingone chapelin his buildingscheme, and (probably more appreci- house slightly(Figure 10). The row continuedthe plan ated) by raisingwages by ten percentwhen he took over formand generalconstruction of the standardhouse, with and anotherten percentafter his firstyear in office. stairsand a back door in place of the larder.Ashwell Certainlyhe succeededin keepinghis workforcerelatively obviouslywanted an impressivebuilding, but essentially contentedat a difficulttime. When the otheriron-workers VictoriaRow was a standardhouse rowwith a raisedfront of Monmouthshirerose in rebellionin 1839,and joined the walland an unequalroof pitch which was so badlyset out Chartistmarch on Newport,the Blaenafonworkers stayed thatthe back of theupper room would have been unusable at home.61 (Figure 3). It seemsthat the "tradition"of the standard house was so establishedthat the workscraftsmen firmly Conclusion were unable to modifyit satisfactorily.Obviously no drawinghad been made to workout thenew design before The basichypothesis of thisarticle is thathousing can be a it was built. The "do-it-yourself"nature of the row was sourceof evidencefor matters that are not directlyrelated foundin its fittings:a works-madecast-iron "kitchener" to itsfunction as domesticshelter. I havetried to showthat rangesurvived in Number21 untildemolition (Figure 11). in suitablecircumstances (and I admithere that Blaenafon does offeran unusuallyfull record of earlyhousing types) evidencedrawn from housing can significantlybroaden a threadof historicalcontinuity. It can evenbridge gaps in the documentaryrecord. Housing can be regardedas superiorevidence: it is amongthe sourceswhich the editor of this journal has classifiedas "unintentional"and thereforefree of interpretationalbias.62 But to use houses '" in thisway as documents,we mustlearn to read them.In W) l'~*Thc l^y;!!!!!.1.1.1.11'1''!.'!!)-?,^; ;:;:r^ presentingthis study I have been inhibitedto some degree by an assumptionthat I haveto explainnot just theextent Block W was also on freeholdland and used (Chapel Row) of the evidence,but also its nature,and at the same time the same Palladian motifs on a scale.59 Its grander attemptto analyzeits significance. centerpiecewas a chapeldated 1839 whichwas demolished about 1955. The houses surviveduntil 1971.60 They The ordinarinessof workers'housing, especially of thesort formedtwo each with centerand end terraces, pedimented mostlikely to meetthe eye of the industrialarcheologist, The of the houses was like Victoria pavilions. plan just has tendedto obscureits actualvariety of detail.In Britain but thecross section was The floors Row, improved. upper the historianswho have consideredearly industrialand weredivided into two bothwith headroom rooms, proper urbanhousing comprise two overlapping groups- those who and windows.Both these rooms werestill open to theroof feelthat it is notreally usable as data,and thosewho think as in the standard but otherwisethis was a space, house, thatall houses whichmight have been usefulas data have four-room good design. been destroyed.Historians of the firstsort reject actual housesas data becausethey see any particularexample as Discussion of Phase 4 eithermeaninglessly commonplace (an attitudewhich I can The evidenceof housingin this phase supplementsthe see creepinginto my introductionto this paper) or else pictureobtained from the documentarysources. Ashwell's unrepresentativeand biased. E. Gauldie writesin Cruel managementis shownto be proneto grandgestures, but Habitationsthat "what might seem the most concrete of all lackingthe controlof detail which was so strikingin evidence,houses actually survivingfrom the period,is Thomas Hill's time. The inept work on Victoria Row necessarilymisleading, because what now existsis a sample indicateshow muchhouse-building was stilla craft,carried selectedby time,almost certainly better than the average on by practicedhabits, rather than a processgoverned by forits periodor it wouldnot have survived."63 The second drawingsand specifications.It also offersa commentaryon groupof historianswould like to use housesas data,but the attitude,which Ashwell seems to have broughtwith believesthat the relevantexamples have been destroyed. himto Blaenafon,that an impressiveexterior had a higher Thus W.G. Hoskinswrote in 1967 (fiveyears before the
30
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History
Figure 10. VictoriaRow, Blaenafon (Block V) cl975.
Figure11. VictoriaRow, Blaenafon(Block V) cl975: works-made cast-iron"kitchener" (now preservedby TorfaenMuseum Trust).
31
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology demolitionof BunkersRow): "As forhouses at the other value of a corpus of inventoriesand distributionmaps. end of the socialscale, it is probablytrue to say thatwith Industrialhousing lacks thissolid background. Hence many thealmost total clearance of slumsin ourtowns during the readersof thisarticle may weigh up as tenuousat bestmy past generationwe may have lost informationabout assertionthat the shape of an arch over a windowis house-typesand housingconditions which can neverbe evidenceof the place of originof the craftsmanwho made recorded.It is oftenthe commonplacewhich is most in it.66In fact,it is moreconvincing to arguefirst that the need of recording."64Both thesequotations are soundly constructionand formof BunkersRow (Bl) is evidenceof based, yet both have proved unduly pessimistic.The non-localbuilders on the site,and secondthat the window corollaryof Hoskins'slast sentenceis thatone does not archesof Stack Square (A) tell us wherethose builders knowthe commonplaceexists as a historicalphenomenon mighthave come from.The archshape, I maintain,would until one startsto study it. Certainlythat is true of identifya Midlandereven if therewas no documentary Blaenafon.It wasonly the survival and identificationof the evidenceat all about the originsof themen who set up the furnaceswhich attracted any attention(belated and at too furnacesin 1788. low a levelof thoroughness)to thehousing, and thusled to the discoveryof buildingswhich both Hoskins and Gauldie Muchof the data presentedhere is muchless tangible than wouldrecognize as exceptionsto theirgeneralizations. the survivingwindow arches. Roughly drawn maps, snap- shots,and recollectionsof thingspeople saw or heard70 In theirtheme-paper to the 1981 ICCIH conference,Roger yearsearlier have all beenessential to thereconstruction of Chartierand Yves Leguin suggestedthat there should a pictureof Blaenafonhousing. Oral evidence prompted the be a corpus of industrialand working-classhousing.65 archeologicalinvestigation of the interiorof BunkersRow Studentsof ruralvernacular houses recognized long ago the (Bl). An elderlyresident told me how herfather used to
Figure12. Number 1-4 Stack Square, Blaenafon (Block A) in 1969.
32
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History recall in his youth that "they rose up the roofsof the Housinghas an analogousvalue as a continuoushistorical houses,"thus signaling to mea morethorough investigation phenomenon.It existedin some formwherever there was of the structurebelow the surfaceof thewall-plaster. Maps humanactivity. While its absence todayfrom a particular too haveto be interpretedcorrectly. The Tithemaps are a place may be due to destruction,it may meanthat one is vitalsource for early industrial housing in Britain,but they looking in the wrongplace, or for the wrongkind of arenot reliable in all respects.They were made in the 1840s housing.It is not entirelyby chancethat Blaenafon is the to show the use of agriculturalland, not as a censusof place wherethe earliestworkers' housing in South Wales houses. They can be comparedwith later surveysfor has been found.The absenceof housing(except of houses topographicalaccuracy, but thequality of thedetail shown for skilled workers)at other,earlier ironworks sites is mustbe weighedup in termsof internalconsistency. Thus evidencefor the continuationof a quasi-ruralsociety when the Llanoverparish map of 1844 marksbuildings reason- theywere firstestablished, and the alien characterof the ablyaccurately as to theform and theplan shape- whether housingfound at Blaenafonsupports this analysis. In the it is back-to-back,and so on-but notas to thesize or even same way the absenceof identifiablehouse-building by the the exact numberof individualhouses in a row. Maps, Blaenafonpartners at the turnof thecentury, and theway howeverrough, do have the advantageof beinga con- in which house-buildingwas re-startedwith the outside tinuousrecord of a location;they treat the area surveyed financingof ShepherdsSquare, reveal the partners'capital (not alwaysthe sameas thearea represented) as a whole.67 shortagewith as much or more certaintythan do the One can be reasonablysure that if a featureappears in one documentarysources. Documents tell us of the lossesthat partof a map it did existthere, and thatwhere it does not the partnerssuffered, but we wouldhave to inferthat this appear,it didnot exist. mighthave affected house-building. The absenceof house-
Figure13. NailersRow, Coalbrookdale,early 19th century.A Midland Englishrow of four-roomdouble-fronted houses.
33
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology
Figure14. BunkersRow, Blaenafon,in 1971, the old cartroad passingbetween Block B2 (left) and Block Bl (right). building,at a time whenthe labor forcewas rising,is a the mainenterprise. They saw thatit could be worthwhile positiveindicator of theirdifficulties. to providehousing even when there was no greatmanagerial need forit, and to siteand designthese houses so thatthey In the early industrialcolony, housing was an essential broughtin the maximumreturn, instead of just to suitthe componentof businessorganization. From a studyof the needs of a processor plant. This phase was reachedat housingof cotton factorycolonies, L.D.W. Smith has Blaenafononly afterabout 1825. Withthe formationof proposeda basic distinctionbetween the managerialand the BlaenavonIron and Coal Companyin 1836 thereis a the entrepreneurialobjectives of house-buildingby early perfectexample of the reversionto managerialobjectives industrialists.68He suggeststhat at first(and also earlyin which Smith suggestedwas characteristicof the new the life of each new management),house-building was management. usuallymanagerial in intent.Management's aim was to secure,at minimumcost to the business,the regular In stressingthe non-domesticfeatures of the evidence attendanceof a labor forceadequate for the disciplined obtainablefrom housing, I havetended to pass overmany needsof the Lookingat suchhousing, features enterprise. social questionson whichthe Blaenafonhousing provides whichtended to increasethe (and moral)welfare physical information.How far,for example, can we takethe changes of the workersare as well as obviously"managerial" in house size as an indicatorof the social relationships and so are those which (sometimes)"philanthropic," withinindustry? Does the changefrom a factorof threein economizedin directcosts to the impressedon company, the rangeof late 18th-century house areas to an absenceof the workersthe watchfulnessof the managers,and gavea anyrange at all in thestandard houses of the 1820s indicate senseof common also had an important purpose.Housing a reductionin the statusof craftsmen-a growthof the managerialrole in the controlof cash-flow.The company proletariatout of the earlysociety of independentsubcon- whichrecouped all rentsmuch reduced its need forcash, tractingmasters and theirgangs? Conclusions of thatkind particularywhen its house-buildingcould be used, as mightbe expectedto emergefrom the study of housingevi- Thomas Hill's standardhouse programprobably was, to dence.Perhaps more striking is the way it illuminatesour even-outthe employmentof the buildingcraftsmen who generalimpression of theseveral Blaenavon managers. wereessential for the operationof an ironworks.Later in the lifeof businesseswhich owned housingmany propri- etors came to recognize,according to Smith,that the Ashwell'srecord was known already from the documentary supplyof livingspace could be a profitableentrepreneurial sources,but his housingshows that he had realproblems in activityin its own right,either alone or as a subsidiaryto gettinghis ideas carried out. Criticsof hisspending program
34
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions Housingas a Source forIndustrial History could well have used VictoriaRow as an exampleof his 1970), p. 95. Rhymni(Mid-Glamorgan): Plantation Street, ineffectualuse of the resources. 1837-41. See JeremyB. Lowe, WelshIndustrial Workers company's Hopkins's 1775-1875 Museum as a was also established Housing (Cardiff,National of Wales, reputation good employer already 1977), p. 56. Pentrebach(Mid-Glamorgan): Triangle, 1839-44. matters.69The by his interestin religiousand educational See JeremyB. Lowe and J. Gross,"Triangle, Pentrebach, evidenceof housingis thatstandards of accommodationdid MerthyrTydfil," IndustrialArcheology [G.B.] 15, 2 rise his of direction, he and his n 980V146-1 56. during period though ' partnerwere unable to allocate sufficientresources to 5. The modernspelling; the Welsh "f pronouncedas English"v." The olderform "Blaenavon" was used in the names. house-building.The man whose style is most clearly company The moderncounty name is Gwent;before 1973 it was revealedby the housingevidence is ThomasHill Jr. He can Monmouthshire. be seenas the forceful,well-organized controller of a busi- 6. J. Lloyd,the younger, of Brecon,The Early History of theOld nessthat can neverhave been easy to runbecause of itsgeo- South Wales Iron-works,1760-1840 (London, 1906), graphicalsituation. Such an impressioncould not be deduced pp. 160-161.There is no authoritativehistory of theBlaenafon ironworks.The most recent account is J.K. fromthe scanty material relating directly to theironworks. by Knight, "BlaenavonIronworks: 1789-1976: A PreliminarySurvey," Journalof South-East WalesIndustrial Archaeology Society 2, Finally,what is the relevanceof thisarticle to a mainly 3(1977):26-38. NorthAmerican readership? Writing in Britain,about one 7. By John Mayberyof Worcestershireat the Hirwaun(Mid- Britishlocation, I hope thatit hassuggested lines of inquiry Glamorgan)furnace. D.M. Rees, Mines,Mills and Furnaces and of whichwill fruitfulwhen (London:HMSO, 1969), p. 62. principles analysis prove 8. The Dowlaisfurnace erected on thehigher land in 1759 useda appliedin othercontexts. It is notpossible for me to tryto cylinderblowing engine powered by waterwheel. The reservoir predictwhat these contextsmight be. Even withinthe was keptfilled by a steampump. The systemwas not reliable: limitedperiod of the IndustrialRevolution in Britain a steamblowing engine was installedin 1802. Rees,Mines 66. housingprovision can rangefrom the completecolony, (note7), p. 9. Ibid.,p. 62. suchas Blaenafon,to at all.70Each situation nothing along 10. Map of Newport-Pontnewynydd Canal and associatedTram- thisrange has its own implications,both for the relation- roads. Copy at Newport(Gwent) Museumand ArtGallery, shipof industryto the pre-existingeconomy, and forthe dated1792, by Thomas Dadford Junior. organizationand managementof theindustry itself. 11. Two-inchOrdnance Survey Map, sheet 194 (unpublished), dated 1813, London,British Library. Plan of BlaenavonIron WorksMonmouthshire, dated 1814. Copy in collectionof TorfaenMuseum Trust, Pontypool, Gwent. Acknowledgments 12. One-inchOrdnance Survey Map, sheet 42 "Brecon," Old Series,published 1832. The author wishesto thankMr. J.K. 13. Tithe Surveys:Trevethin (1843), Llanfoist(1844), Llanover particularly Knight, and of AncientMonuments Mr.D.N. (1844), Llanwenarth(1846) parishes.Aberystwyth, Inspector (WelshOffice); NationalLibrary of Wales. Anderson,especially for his help withcensus material; and 14. W. Coxe, An HistoricalTour in Monmouthshire(1801). 2nd also Mr. A. Babbidgeand Mr. MartinLawler of Torfaen ed. by E. Davies(Brecon, 1904), unnumbered illustration. MuseumTrust for the references supplied by them. 15. Plan and photographin Lowe, WelshWorkers Housing (note 4), p. 9. 16. Mapsin collectionof TorfaenMuseum Trust. At theNantyglo works(1792-1795), the livedin thecenter of Notes managingpartner a long row of workers'dwellings. Lowe, WelshWorkers Housing(note 4), pp. 16-17. 1. W.G. Hoskins,Fieldwork in Local History(London: Faber, 17. Ibid., p. 15; and JeremyB. Lowe and D.N. Anderson, 1967),p. 23. "BunkersRow, Blaenavon,Monmouthshire," Iron Industry 2. Dianne Newell,"IA, DiscoveringMore About Human Be- HousingPapers 6 (Cardiff:UWIST, 1973). haviourin the Past,"in "Symposium,Industrial Archeology 18. If brickhad been used, the outsidewalls would havebeen and the Human Sciences,"SIA OccasionalPublications 3 thinner,and theinternal depth of therooms would have been (1978):2. 12 feet-a commonstandard dimension for setting out houses. 3. MichaelBussell, "Working Group L: Minutes,"in The Indus- Brickswere taxed, but stonewas both cheapand tax-freeat trial Heritage,Trans TICCIM 3, ed. M. Nisser(Stockholm: Blaenafon. NordiskaMuseet, 1977), p. 409. iy.
35
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions IndustrialArcheology
and Stourbridge.His partnersBenjamin Pratt and Thomas 47. H. Scrivenor,A ComprehensiveHistory of the Iron Trade Hopkinsare moreshadowy figures who wereboth dead by (London,1841), pp. 126-129,294-295. 1792, leavingHill to continuethe businesswith Samuel 48. Knight,"Blaenavon Ironworks" (note 6), p. 31. Therewere Hopkins,who was his nephew. otherpartners, for example Mr. Robert Wheeley in 1830-1836. 23. T.D. Steel,"Blaenavon Works" (1893), printedin Pontypool 49. Sale particulars(note 35). The 100 tons per week was only andDistrict Review 11 (October1972): 12. sustainedfor the first year or so. Thelong-term output was 67 24. C.W. Chalklin, The Provincial Towns of Georgian England tonsper week. (London:Arnold, 1974), pp. 198,203. 50. The dual typeappeared at Nantygloin 1793. Lowe, Welsh 25. J. Burnett,A Social History of Housing 1815-1970 (Newton WorkersHousing (note 4), pp. 16-17. StaffordshireRow is Abbot: David and Charles, 1978), p. 71, shows a surveyed illustratedin /. Indust.Archaeol. 1 (1964-5). drawingof a back-to-backterrace of 1843. No surveyof any 51. Lowe,Welsh Workers Housing (note 4), pp. 30-31. earlierhouses has been published. bl. blaenavoniron and Coal CompanyMinute Book, uwent K.U. 26. "Observationson the makingof iron . . at Llanelly" by John D.751.356. Pig-ironin 1836-1837fetched £5-8s-0d per ton; Hanbury. In H.R. Schubert,History of the BritishIron and finished(wrought) iron £lO-lls-6d per ton. Meetingof 15 Steel Industry from 450 B.C. to A.D. 1775 (London: Sept.1837. Routledgeand KeganPaul, 1957), Appendix XVI, p. 427. 53. Sale particulars(note 35). Theseshow the works housing stock 27. In 1841, the averagehousehold in BunkersRow (B1/B2) to be 483 houses,50 of whichwere for officials or on the contained2.2 workersand 2.7 dependents,4.9 personsin all. farmswhich the partners owned. BlockA had beenaltered, but in theslightly smaller houses in 54. Powell,An EconomicHistory (note 42), pp. 11-12, 49-50. theNew Ranks (Q and R) theaverage was 2.7 workersand 3.6 55. Sale particulars(note 35), in a noteby P. Gabb,the auctioneer. dependents,6.3 personsin all. 56. Lowe,Welsh Workers Housing (note 4), pp. 28, 37. 28. Gray-Jones,Ebbw Vale(note 4), p. 38. 57. Ibid.,p. 8. 29. J. Wood, A Series of Plans for Cottagesor Habitations of the 58. MinuteBook: GwentR.O. D.751.365. 6th meeting22 May Labourer,etc. (London,1781). 1806 edition,p. 6, footnotek. 1840. 30. Coxe,An HistoricalTour (note 14), p. 200. 59. Lowe,Welsh Workers Housing (note 4), p. 42. 31. Knight,"Blaenavon Ironworks" (note 6), p. 31; Gray-Jones, 60. Some of them,at thewest end of therow, were destroyed by Ebbw Vale(note 4), p. 35. slag-tippingabout 1910. 32. R. Warner,A Walk ThroughWales in 1797 (Bath, 1807), 61. Knight,"Blaenavon Ironworks" (note 6), p. 31. pp. 27-28. 62. Newell,"IA, DiscoveringMore" (note 2), p. 2. 33. P. Smith,Houses of the WelshCountryside (London: HMSO, 63. E. Gauldie, Cruel Habitations. A History of Working-class 1975),pp. 310-311. Housing(London: Allen & Unwin,1974), p. 96. 34. J. Fox, General View of the Agricultureof the County of 64. Hoskins,Fieldwork (note 1), p. 67. Monmouth(Brentford: for Board of Agriculture,1794). 65. RogerChartier and Yves Lequin, "The LivingQuarters of Reprintedin Monmouthshire Medley 1 (1976): 132-1 33. IndustrialPopulations," Trans ICCIH 81 (InternationalCon- 35. "Particularsof the . . . BlaenavonIron and Coal Mines, ferenceon the Conservationof the IndustrialHeritage), November1833" MS notesfrom sale catalogue. Gwent Record forthcoming. Office. D.7J94. 66. A documentarysource, such as an advertisementby Thomas 36. Lloyd,Old SouthWales Iron-works (note 6), pp. 166-170. Hill in a Birminghamnewspaper of 1788 forbricklayers or ó/, j.b. Lowe, ine iirsttorge at Biaenaion, journal oj me masons,would probablybe foundmore acceptable, even if it South East Wales Industrial Archaeology Society 3, 1 containedno hint of the place where they were to be (1978):3M0. employed.The advantageof physicalevidence is that its 38. Loweand Anderson, "Bunkers Row" (note17), p. 51. locationhas to be explained. 39. Sale particulars(note 35). 67. Some of the Tithesurveyors did not recordfeatures of any 40. Coxe,An HistoricalTour (note 14), p. 200. kindon areaslike highways that were not subject to tithe(the 41. Lowe,Welsh Workers Housing (note 4), p. 63. Churchtax on crops).Houses that encroached on theverges of 42. CG. Powell, An Economic History of the British Building highways,very common in some ruralareas, were omitted Industry1815-1979 (London: ArchitecturalPress, 1980), fromthese maps. pp. 11-12; Chalklin, The Provincial Towns (note 24), p. 224. 68. L.D.W. Smith,"Textile FactorySettlements in the Early 43. The partnersmaintained a specialbreed of mulesduring the IndustrialRevolution." Ph.D. dissertation,Aston University, earlydays of theworks. Knight, "Blaenavon Ironworks" (note 1976. 6), p. 36. 69. Knight,"Blaenavon Ironworks" (note 6), pp. 30-31. 44. Chalklin,The Provincial Towns (note 24), pp. 198-200. 70. RichardArkwright chose Cromford in 1770 as thelocation for 45. Wood,A Seriesof Plans (note 29), pp. 9-10and 14-15.Rubble his firstspinning mill because there was alreadya community stoneworkcost 6V2old penceper square foot, 18 inchesthick. of lead-miners there; he was able to recruittheir wives and A windowcost 18% penceper square foot, not includingthe childrenas his firstlabor force.No housingwas builtuntil laborof formingthe opening.Glass was taxedfrom 1767 to 1777. See S.D. Chapman,"Workers Housing in the Cotton 1845. FactoryColonies, 1770-1850," Textile History7 (1976): 46. Planof theTramroad from Blaenafon Furnaces to Pwll-duand 112-139,118. Garnddyrys.29 Sept.1819. GwentRecord Office.
36
This content downloaded from 132.206.27.24 on Wed, 10 Apr 2013 21:02:10 PM All use subject to JSTOR Terms and Conditions