AUSTRALASIAN lllSTORlCAL ARCI-IAEOLOGY, 19.2001

Every Procurable Object: A functional analysis of the Ross FactoryArchaeological Collection

ELEANOR CONLIN CASELLA

ESlablished ill 1848 for accommodalion o/tral1sported/emole convicts, the Ross Female Fae/DlY operated1111/;/ early 1855, when the cesso/ion ofBrWsh penal fransporlaNonlo Jlcm Diemen sLand caused the closurea/this rural place a/confinement. From 1995 (0 1999, the Ross Factol)' Archaeology Project comparativelyexamined material culture .Ij·ol1J three wards of the prisoll site. Sfraligraphic aI/cl architectural dOlodemonstrated that a total q/32jeatures and deposits could be related to the female-coJ/vict period a/sireoccupatiol/. This paper presents results ala/uHc/iollal analysis conducled 011 arlf;{acts recoveredfrom theseFactory-period deposits. Based 01/ st/fdies of nilleteenth-celltlllJ' sites in tlte American West, this .~peqlic

metlIodology was adapted to examine tile nature and relalive frequency of.(tmclional types particlllar~v

relevant to an archaeology oIcOJ?/inemeJ1f.

INTRODUCTION

There must be a con like me in every prison in America.Ilm the guy who can get it for ya. Cigarcttcs. A bag ofreefer. if that's your thing. A bottle of brandy tocelebrate your kld's High School graduation. Damn nearanything, within reason. Yes sir, 1'111 a regular Sears andRoebuck. So when Andy Dufrcne camc to me in 1949and asked me to sllluggle Rita Haywol1h into thc prisonfor him, I told him <no problem'. (Tile Shaws}ulIlkRedemption, 1994; Castle Rock Entel1ainmcnt)

From 1995 to 1999, the Ross Factory Archaeology Projcelexamined material aspects of gender identity and powerrelations within the Rass female Factory, a mid-ninelecnth­century, female-convict prison site in rural Tasn1ania. Thispaper will prcscnt results from post-excavation work onrecovercd material assemblages. Following completion of theRoss Factory Archaeology Project, the Rass collection wasdonated to the Queen Victoria Museum and Art Gallery ofLaunceston, Tasmania, for cmation, display and pell11anentstorage. Access to both the Ross collection and the detailedanefact catalogue is available upon request through the QueenVictoria Muscum and Art Gallery, and the Culnlral HeritageProgram of the Dep<u1ment of Primary Industries, Water andEnvironment, Tasmania.

The Rass Female Factory Historic Site

From 1803 101854, Van Diemen's Land received over 74 000convicts transported from the British Isles (Eldershaw 1968:130). Approximately 12 000 of these felons were women,primarily convicted ofpcny domestic theft (Oxley 1996). Upontheir arrival in the colony, most spent time incarceratcd withinthe Female Factory System, a network of women's prisonssc'Hered across the island (Brand 1990; Damollsi 1997).

Named 'factory' as a cOlltraction of the word'manufactory'. these penal institutions were modelled afterthe British Workhouse System, a nincteenth-ccnnlry publicsocial welfare system which required standardised rates oflabour from all inmates to hasten their social and moralsalvation from delinqucncy, idleness and poverty (Driver1993). Upon tbeir entry to a Factory, female convicts wereassigned to the 'Crimc Class' and incarcerated for a minimumof six months. While serving this probationary sentence, thewomen were intended to 'reform' through Christian prayerand daily labour a\ thc acceptably feminine industries ofsewing, textilc production and laundry work. lnsubordinationwas punished through lenglhy periods of confincment inSolitary Cells, accompanied by severe rcduction of food

rations. Once the probationalY period had been successfullyserved, inmates were reclassificd into the 'Hiring Class' andawaited assignmcnt to local properties, completing theirconvict sentences as domestic servants for free colonists(Ryan 1995; OxIcy 1996; Damollsi 1997: Daniels 1998)_

A/though detailed discussion of the history andarchaeology of the Ross Factory can be found in alternativepublicalions (Casei la 1997; Casei la 200Ia), the majoroccupation periods must be slll1unarised to contextualise theRoss collection. Locmed on the southern edge of the RossTownship in the rural midlands of the island, the first historic­period occupation occuned in 1833, when a thatched-roofbrick hut was constructed for the acconullodalion of maleconvicts during construction of the ornately carved sandstoneRass Bridge. Aner completion of the Bridge in 1836, the silewas abandoned. In 1841, it was reused as a Road Gang Stationfor 172 male convicts working on the 'Midlands Road'between Hobart Town and LaUJlCeSlOn. Historic plans indicatea dramatic cxpansion of the station, with the original brickbuilding incorporatcd inlo a quadrangle of sandstone barracksarranged around two adjoining musteryards (AOT PWD266/1693). The original brick hut was modified into a 'Bakeand Cook Housc' and three brick ovens were installed.

During 1847. the station underwent significant physicalaheraliol1s as it was transformed into a female-convict prison.Many of thesc conversions proved archaeologicallysignificant. 'When stone noors within the convict dormitorystructures were removed, they were replaccd by stone joistsupports and suspended timber noars, thereby creating a newunderfloor spacc below the Crime Class and Hiring ClassDomlitory 0001' boards (Terry 1998: 32)_ The i11lemal layolllof the original brick Bridge Gang hut was modified, with theovens removed and a new internal doorway added (AOTPWD 266/1695). Historic plans indicate that the renovatedstrucUlre houscd the' Assistant Superimcndent's Quaners'.

The Ross Female factory operated from 1848 to 1854,when Britain ceased convict transportation lO the VanDiemen's Land colony (Figure I). Al the height of operationin 185 I, the Ross Female Factory incarceratcd 124 femaleconvicts and 44 of their children (Rayner 1980: 31)_Disciplinary problcms in 1850 directly led to the constructionof Cl new block of 12 sandstone Solitaly Cells locatedapproximately 30 m south of the Main Compound of theFactory (AOT PWD 26611 696)_ Opened in 1851, the newSolitary Cells were intended to improve the ptlllitive isolationand deprivation of recalcitrant inmates (Scripps & Clark1991). After closure in 1854, the site was transferred tocivilian managemcnt and experienced a series of municipal

25

Poltugal S\reel

Figure J.. Site plan I\'ilh inset (!( Van Dielllen's Land (Tasmania).

and domestic occupations, It was gazetted as a historic reservein 1980 and is now administered through the Cultural HeritageBranch of the Tasmanian Department of Primary Industries,Water and Environment.

The Ross Factory Archaeology Project

Begun in 1995, the Ross Factory Archaeology Project wasdeveloped to examinc archaeological remains from the tlueedifferent classification stages experienced by female inmates.Excavation trenches were therefore located to sample from theCrime Class Dormitory (Area A). Hiring Class Dormitory(Area B) and Solitary Cells (Area C). Trenches within Area Balso sampled from the Assistant Superintendent's Quarters(Casella 1997). Following completion of the 1997 excavationseason, all recovered artcfacts and soil samples underwent 14months of post-excavation work at the Tasmanian Parks andWildlife Service laboratory in Hobart.

LABORATORY METHODS

Three stages of laboratory processing were dcveloped forrecovered artefactualmaterials. During the excavation seasons,retrieved materials underwent field management procedures.A fter being transported la the Hobart laboratory facil ities of theTasmanian Parks and Wildlife Service, artefacts from the RassFactory collection were cleaned, identified and catalogued byfabric. Finally, a specialised functional analysis was conductedon materials collected from stratigraphic units related to thcfemale-convict occupation periods. This section will briefly

26

summarise the methods and procedures developed for eachstage of artefact management and laboratolY analysis. Detaileddiscussions of the analysis and interpretation of assemblage~

from the Ross collection are available through otherpublications (Casella 1999,2000, 200Ia).

Stage One: Field Management

During the 1995 and 1997 excavation seasons, field cre\\'iconducted prim31Y cleaning, identification and stabilisation ofartefacts through an on-site Field Laboratory. Culturalmaterials and soil samples were identified by excavation area,trench number and stratigraphic context number. The va~

majority of recovered finds was collected by lot provenancewith artefacts spatially provenanced as a group and linked IQ

thcir slratigraphic cOlltext number. However, followingstandard field practices, three-dimensional point provenan~~

data was also taken for 'special finds' or those objectidetemlincd to be diagnostic, photogenic, valuable, pal1icularJ!fragile or otherwise ullusual (Davies & Buckley 1987: 16i,Di'on 1994: 36-37; Praetzellis & Praetzellis 1990).

Stage Two: Identification and Cataloguing of Finds

The second stage of management involved laboralOl}processing of recovered non-indigenous artefacts. A numberof Aboriginal lithic al1efacts were recovered from disturbedsecondary contexts during both seasons of excavation. MrCharlie Beasley, Tasmanian Aboriginal Heritage Officer fOl

the Ross Factory Archaeology Project, inventoried these lilh~

•

s

sSita

fifiSIc

Adornment

Table 1: Factory Related Contexts, RFAP.

PRESENTATION OF RESULTS

Specific classification categories used for this functionalanalysis consisted of Adornment. Clolhing, Domestic.Ecofact, Fuel, Indulgence, Literacy, Monetary. Miscellaneous,Social COl1lrol and Structural. Tables 2 to 5 present all datadiscussed in the following section. For some of the artefactasscmblages, estimates of the MNI could be calculated. Wherethese estimates could be made. the MNI count appears in thetablc directly following the total weight in grams, separated bya semi·colon.

Crime Class Dormitory underfloor: 1022, 1033,1039, 1020, 1036, 1035

Dormitory underfloorlfirst demolition: 1008, 1015Dormitory exterior: 1048, 1028, 1054

Asst. Superintendent's Quarters east room: 2033,2020

Quarters west room: 2034Quarters chimney feature: 2035Hiring Class Dormitory underfloor: 2023Dormitory exterior: 2028, 2026

Solitary Cells interior first floor (Fl): 3028. 3040,3037,3024

Central cell interior burnt feature: 3036Cells interior second floor (F2): 3017, 3035. 3018West cell interior stash pit: 3022Cells northern exterior: 3034. 3021, 3020Cells southern exterior: 3025, 3014. 3013

Area A

Area B

Area C

Three types of artefacts were classified under this category.Defined as objects relating to personal ornamentation, thematerials found within factory-related deposits consisted ofonc composite copper and glass necklace or earring pendant(Area B, Table 3). seveo glass beads (Area A, Table 2), andtwo clear-glass rings (Area A). Although this last artefact typewas classified as adomment, it is also possible these objectswere part of a closure device for medicine bottles (M. Carney,pers. comm. 1998). No adornment-related artefacts wererecovered from Area C, the SolitalY Cells (Table 5).

Seven glass beads were recovered from deposits within theCrime Class Dormitory (Table 2). Using the standardarchaeological classification system for glass trade beads(SpeelOr 1976; Karklins 1982; Sprague 1985) all sevenartefacts were identified within the Canadian Kidd typology as'Type IVa' (Kidd & Kidd 1970). This bead type coosists ofdrawn, hot.nlll1bled, undecorated, polychrome glass beads.

Factory period. These depositional events contained materialsmost representative of the physical layout and daily use of theRoss Factory from 1847 to 1855. Although detailed discussionof the strati graphic and historical evidence used foridentification of Factory-related deposits is available throughother published sources (Casella 1997, 200 Ia), Table Isummarises the specific archaeological contexts subjected tothe tertiary stages of post-excavation analysis.

The functional classification system used for this researchwas inspired by the work ofAdrLan and MalY Praetzellis at theCultural Resources Facility of Sonoma State University. Aseries of fUllctional categories were generated from thePraetzellises' report on a nineteenth~century urban domesticsite in Sacramenta, California (Practzellis and Praetzcllis1990). The specific functional categories utilised in this studywere modified to address institutional and Australian socialcontexts. This paper will now turn to consider results from thisfinal stage of analysis.

artefacts for identification purposes (Casella 200 Ia: Appcndix7). All Aboriginal artcfacts recovered through the RossFactory Archaeology Project were repatriated to theTasmanian Aboriginal Community. through the AboriginalHeritage Unit of the Parks and Wildlife Service and theTasmanian Aboriginal Land Council. at the end of eachexcavation season. Thus. 14 months of curation.identification. cataloguing and analysis were conducted on theRass Factory assemblages of non-indigenous artefacts.

To encourage future inter-site comparative studies ofTasmanian convict collections, identification andclassification methods for the Ross materials were dcvelopedfrom the Fabric Key produced in the early 19805 by the PortArthur Conservation and Development Project (Davies &Buckley 1987: 184-189). Artefacts were first categorised intoone of four fabric types: ceramic, glass, metal or 'other'.Ceramics underwent two further stages of classification:separated by ceramic Iypes. the earthenwares. stonewares,porcelain and terracotta artefacts were then divided byfunction into domestic versus architectural and industrialceramics. Owing to the highly fragmentary nature of theceramic assemblage, vessel type could not be determined withaccuracy. Rather. minimum number (MNI) counts wcreinterpreted from differences in fabric and decorativecharacteristics. F1lI1her detailed study of the Ross Factoryceramic assemblage would certainly offer additionalinfonnation on the nature of food ways within female convictinstitutions.

Glass artefacts were first classified as window, bottle orother. and then divided by specific glass colour. Metals wereseparated into aluminium. copper-alloy. ferrous, lead. silver,composite or other. Nails were specifically identified withinthe collection because of their particular value aschronological indicators. The general category of 'other'contained all fabrics except ceramic, glass or metal. Thisdesignation included bone artefacts and ecofacts. cement,coal. concrete, leather. mortar and plaster, plastic, rubber,shell. seeds, stone and wood (including charcoal).

After an artefact lot was separated into fabric categories,mdh-idual objects were then catalogued within theassemblage. A unique catalogue number was issued to everyindividual specimen. This catalogue number consisted of fourparts: Excavation Area, Trench Number, Context Number andArtefact Number. Thus. an object catalogued as.~.~.1015.104' would be the 104th artefact from cootext 1015III trench 2 of Area A, the Crime Class region of the site.Counts and weights in grams were taken for all catalogueentries; MNI counts were estimated and recorded for theceramic. glass, copper-alloy sewing pin and ferrous nailassemblages. When possible, measurements and descriptiveidentification data were recorded. Funding limits preventedthe faunal and floral assemblages from undergoing specialistanalysis. Thus. while individual elements were catalogued,bone, seed and shell artefacts could not be identified byspecies. Furure analysis of the fauna I and floral assemblageswill pro\ ide significant perspectives on both the colonialenvironmental landscape and the presence of food remainswithin the Ross Factory.

Stage Three: Functional Analysis

Since the Ross Factory Archaeology Project was specificallyinvestigating the Female Factory period of site occupation,archaeological contexts most directly linked with this horizonwere subjected to tertiary stages of analysis. Cultural materialsfrom these deposits underwent a functional analysis as thefinal stage of post-excavation work. Through a combination ofstratigraphic and historical analysis, a total of 32 depositionalcontexts were identified as mosl strongly related to the Female

27

Table 2: Functional analysis results, Crime Class Dormitory, Area A.

Assemblage underfloor

Adornment glass bead 0.539; 6glass ring 0.19; 1Total 0.639 (0.04%)

Agricultural ferrous horse shoeTotal

Clothing buttons 3.179; 1copper eyelets 0.569; 2copper pins 4.219; 19copper thimble 3.559; 1ferrous buckleTotal 11.499 (0.7%)

Domestic ceramic 83.979; 17clear glass tumblerclear bottle glass 52.279; 5brown bottle glassferrous containerTotal 136.249 (8.5%)

Ecofact bone 174.979oyster shell 2.259Total 177.229 (11.1%)

Fuel charcoal 18.119coal 40.369organic slagwood 1.299Total 59.769 (3.7%)

Indulgence tobacco pipes 4.119; 2olive bottle glass 80.909; 4Total 85.019 (5.3%)

literacy slate pencil 7.259; 3slate tabletTotal 7.259; 3 (0.4%)

Monetary coin 8.49; 1tokenTotal 8.49 (0.5%)

Social Control ferrous chain link 11.359Total 11.359 (0.7%)

Structural brick 96.799mortar/plaster 22.859ferrous nail 351.879ferrous other 570.839copper otherleadsandstone cobbleslate 3.209clear lamp glassclear window glass 60.29Total 1105.749 (69%)

Total (9) 1603.099 (99.9%)

underfloor & demolition

0.119; 10.319; 10.429 «0.01%)

421.529; 1421.529 (2.2%)

7.529; 62.259; 62.389; 182.629; 10.869; 115.639 (0.08%)

332.909; 47191.99; 1348.379; 6

7.59; 1880.679 (4.7%)

762.2398.39770.539 (4.0%)

310.54950.039552.896. 549919.919 (4.9%)

19.319; 41447.859; 111467.169 (7.8%)

20.469; 79.43; 229.899 (0.2%)

524.209524.209 (2.8%)

3878.569167.4491112.118026.7290. 17912.9951.79

68.19; 1500.73913818.439 (73.3%)

18,848.369 (100%)

exterior

5.079; 2

5.079 (1.2%)

11.14; 2

21.639; 20.59;1

33.279 (8%)

92.61 9

92.619 (22.2%)

43.169

7.82950.989 (12.2%)

1.159; 245.959; 347.109 (11.3%)

115.70931.7890. 39925.929

14.389188.179 (45.1%)

417.209 (100%)

The variety recovered from Ross Factory Crime Classunderfloor deposits were all transparent red over an opaquewhite interior (Figure 2). The shape of all seven beads roughlycorresponded with the 'doughnut' shape category defined inJanet Spector's 1976 classification article (Spector 1976: 25).Two sizes were recovered: one was 2.5 n1l11 in diameter(weighing O.02g), and six varied from 4 mm to 5 mm indiameter (cach weighing a.llg).

In a 1997 report on trade beads recovered from anineteenth-century Native American site in California,archaeologist Lester Ross observed that Kidd Type IVaconstituted the second most comll1on bead typc recoveredfrom archaeological sites in the western United States (Ross1997: 185). The red-on-white variety was often termed'comaline d'Aleppo' or 'Hudson's Bay Company' beads, and

28

was especially common in sites dated from l800 to t860(Rass 1997: 185).

To promote international studies of trade beads, AustraliaJlhistorical archaeology has predominantly utilised pre-existingCanadian bead classification systems (Iacono 1996: 46-4i:Birmingham 1992). Although these artefact reports havetended to identify beads under the functional category of'decorative clement', other functions have been suggested.Australian literature has reported on the use of glass beads inCatholic rosaries and decorative weights for sewing bobbil15(Jacono J996; Lydon 1993). However, in thc specific contextof thc Ross Female Factory, it seems unlikely that theseartefacts related to lacemaking or performances of reiigiollidevotion. Numerous historic studies of the Australian FemaleFactories have demonstrated that conflicts over conviCl

Table 3: Functional analysis results, Assistant Superintendent's Quarters, Area B.

Assemblage east room west room chimney

Adornment composite pendant 1.079; 1Total 1.079 (0.08%)

Clothing butlons 12.229; 15 3.939; 4 0.349; 1composite (shoe frag) 0.229; 1copper fastener 1.879; 1copper pins 0.929; 5 0.699; 4 0.259; 2ferrous fastener 0.729; 2Total 15.959 (0.6%) 4.629 (0.4%) 0.599 (0.6%)

Domestic ceramic 279.419; 30 66.09; 13clear glass tumbler 13.78g; 1clear bottle glass 19.78g; 5 3.439; 2green botlle glass 2.609; 1blue bottle glass 2.579; 2 0.259; 1stoneware bottle 38.769; 3stone marble 10.35g; 2 15.779; 3Total 367.259 (14.6%) 85.459 (6.5%)

Ecofact bone 273.169 91.699 27. 199peach seed 2.399egg shell 0.379oyster shell 0.379 0. 159 0.21 9Total 273.539 (10.9%) 94.69 (7.1 %) 27.49 (27.9%)

Fuel charcoal 12.939 2.89 8.479coal 3.489Total 16.419 (0.6%) 2.89 (0.2%) 8.479 (8.6%)

Indulgence tobacco pipes 5.629; 2 1.59; 1olive bottle 91ass 39.819; 2 254.389; 1Total 45.439 (1.8%) 255.889 (19.4%)

Literacy slate pencil 6.429; 3Total 6.429 (0.3%)

Miscellaneous quartz crystal 77.839; 2 1.039; 1 22.599; 1Total 77.839 (3.1%) 1.039 (0.08%) 22.599 (23%)

Monetary coin 9.429; 1token 15.459; 1Total 9.429 (0.4%) 15.459 (1.2%)

Social Control ferrous chain link 37.139ferrous padlock 156.679Total 193.809 (7.7%)

Structural brick 193.239mortar/plaster 33.909ferrous nail 624.279 187.329 39.029ferrous other 645.79 653.619clear window glass 11.68g 20.31 9Total 1508.859 (60%) 861.249 (65%) 39.029 (39.8%)

TOla1(9) 2514.899 (100%) 1322.149 (99.9%) 98.079 (99.9%)

unifonns permeated the disciplinary structures of these penalln51i,u,;o"S (Casella 1999; Damousi 1997; Daniels 1998). Thepresence of glass beads wilhin the underfloor deposits of theCrime Class Dormitory may relate to this perpetual conflict,and thus might represent inmale circumvention of Factoryregulations over dress code and personal possessions (CaseI la2000).

Agricultural

Only one object related to agriculture or similar activities wasrecovered from Factory.related deposits: a ferrous horseshoefound within Area A (Table 2). Since this altefact came fTommixed underfloor and demolition deposits (contexls 1015 and1008) this objecl may have entered the site during demolitionactivities after closure of the Female factory in 1855.

Clothing

This category included materials related to clothing use andproduction. Artefact types included within this category were:buttons of various fabric, a composite leather and copper shoefragment. copper eyelets. copper pins, copper thimbles,copper fasteners. ferrous fasteners, and fen-Clls buckles. Noclothing-related materials were recovered from the HiringClass DOlll1itory interior of Area B (Table 4). In contrast, theCrime Class DOflnitolY interior of Area A yielded both thehighest concentration and greatest diversity of clothing relatedanefacts, including seven buttons, eight copper eyelets, 37copper pins, and two copper thimbles (Tablc 2)_ A smallferrous buckle was also included within the clothing category,although it might have been a fastener on a leather horse

29

Table 4: Functional analysis results, Hiring Class Dormitory, Area B.

Assemblage interior exterior path exterior drain

Clothing buttons 11.409; 15' 0.369; 1copper eyeletscopper pins 0. 139;1ferrous buckle 1.249; 1Total 11.539 (15.4%) 1.69 (0.08%)

Domestic ceramic 1.289; 1 0.569; 1 31.909; 11clear glass tumbler 18.029; 1clear bottle glass 4.919; 2 168.659; 8blue bottle glass 0.079; 1 158.699; 2stoneware bottle 148.379; 4Total 1.35g (0.4%) 5.47g (7.3%) 525.63g (27.4%)

Ecofact bone 57.429 13.149 124.859Total 57.42g (16.3%) 13.14g (17.6%) 124.85g (6.5%)

Fuel charcoal 6.089 0.779 31.639coal 1.139 0.649 16.039Total 7.219 (2.0%) 1.419 (1.9%) 47.669 (2.5%)

Indulgence tobacco pipesolive bottle glass 2.139; 1 1.149; 1 457.029; 6Total 2.139 (0.6%) 1.149 (1.5%) 457.02g (23.8%)

Literacy slate pencil 1.819; 1 1.419; 1slate tabletTotal 1.819 (2.4%) 1.419 (0.07%)

Miscellaneous quartz crystal 4.229; 1Total 4.229 (0.2%)

Monetary cointoken 7.879; 1Total 7.879 (0.4%)

Social Control ferrous chain linkferrous powder flask 297.779Total 297.77g (15.5%)

Structural brick 53.409 54.059mortar/plasterferrous nail 197.919 35.799 77.859ferrous other 32.039 2. 129 293.279copper otherlead 17.149, slateclear window glass 0.339 2.259 9.069Total 283.679 (80.6%) 40.169 (53.8%) 451.379 (23.5%)

Total (9) 351.789 (99.9%) 74.669 (99.9'10) 1919.409 (99.9%)

·special find 575, a cache of 4-hole bone buttons held between sandstone flags of context 2018.

harness. Artefacts specifically related to clothing production(sewing pins and thimbles) were also in greatest conccntrationwithin the Crime Class Dormitory. Materials related toclothing production generally appeared to be concentratedwithin interior spaces of the Main Compound. The CrlrneClass Dormitory yielded the highest frequency of coppersewing pins (MNV = 37), followed by the AssistantSuperintendent's Quarters (MNV = 11). No clothingproduction related materials were recovered from the SolitaryCells of Area C (Table 5).

The highest concentration of buttons (MNV = 36) wererecovered from thc exterior yards and AssistantSuperintendent's Qual1ers of Area B (Tables 3 and 4). Fifteenof these buttons f0l111ed special find number 575, a cache ofidentical four-hole, sew-through bone buttons held betweentwo large sandstone flags of the pathway immediately north ofthe Assistant Superintendent's Quarters. The Solitary Cellsheld the second greatest concentration of buttons, with 22recovered from Area C (Table 5). Nine buttons were recoveredfrom Area A (Table 2). A detailed analysis and interpretation

30

of the Ross button assemblage can be found in other publishedsources (CaseJla 2000; CaseJla 200Ib).

Domestic

The domestic category was used for artefacts related to foodstorage, preparation and consumption. Domestic itemsincluded tableware (ceramics and clear glass tumblers).bottles (clear, blue and brown glass, and stoneware), andferrous containers. The greatest concentration of domesticartefacts was recovered from the Crime Class Dormitol)'interior, with a minimum of 77 individual vessels (I 016.91 g:sce Table 2). Comparatively, the Hiring Class Dormitory ofArca B yielded two vessels (1.35 g; see rable 4). The exteriooof these structures demonstrated contrasting results, with 29vessels (531.10 g) recovered from the Hiring Class sandstonepathway and drain system. and only five vessels (33.27 g)recovered from the exterior and drains of the Crime Class. InbOlh regions, ceramic vessels dominated the domesticassemblages.

The Assistant Superintendent's Quarters held a total of

Table 5: Functional analysis results, Solitary Cells, Area C.

Assemblage F1 burnt layer F2 stash pit northern ext. southern ext.

Clothing buttons 12.329; 9 17.699; 7 5.559; 4 0.959; 2ferrous buckle 14.09; 1Total 12.32g (0.4%) 31.69g (2.3%) 5.55g (0.5%) 0.959 (0.04%)

Domestic ceramic 44.899; 8 6.889; 5 8.779; 2clear glass bottle 59.429; 4 6.759; 1 3.489; 3 42.379; 1blue glass bottle 0.259; 1ferrous container 375.509; 1 2.159; 1 28.229; 1Total 104.56g (3%) 6.75g (0.5%) 375.5g (43.5%) 12.51g (1.2%) 79.36g (3.1%)

Ecofact bone 1318.479 40.479 265349 26.499 667.699 292.99oyster sheH 0.699snail shell 0.09gTotal 1319.259 (39%) 40.47g (24.9%) 265.34g (20%) 26.49g (3.1%) 667.699 (65%) 292.9g (11.4%)

Fuel charcoal 192.45g 95.30g 311.639 6.199 63.969 35.019coal 0.449wood 0. 139 6.569 1.71 9 1.049Total 193.02g (5.7%) 95.30g (58.5%) 318.199 (24%) 6.19g (0.7%) 65.679 (6.4%) 36.05g (1.4%)

Indulgence tobacco pipes 68.539; 7 1.01 9;1 1.579; 1 3.19; 1 14.249; 3 6.319; 3olive glass 221.149; 6 2.779; 1 27.249; 2 60.929; 1 90.279; 3 608.72; 5Total 289.69g (8.5%) 3.78g (2.3%) 28.81g (2.2%) 64.02g (7.4%) 104.51 9 (10.1 %) 615.03g (24%)

Monetary coin 4.689; 1Total 4.68g (0.1%)

Social Control ferrous chain link 67. 399 40.409Total 67.39g (2%) 40.40g (3.9%)

Structural brick 805.059 18.779 604.01 9 11.959 839.219mortar/plaster 10.51 9 1.51 9ferrous nail 318.829 4. 149 23.379 32.81 9 46.979 258.649ferrous other 259.059 33.529 358. 109 67.499 77.91 9copper other 2.019lead 83.559timber 54.779clear window glass 26.899 0. 309 0.519 6.89g 219.849Total 1411.82g (41%) 23.21g (14.3%) 671.92g (51 %) 390.91g (45.3%) 133.3g (12.9%) 1535.43g (60%)

Total (g) 3402.739 (99.7%) 162.76g (100%) 1322.70g (100%) 863.11 9 (100%) 1029.63g (100%) 2559.72g (99.9%)

If't 2

t t " eo 10_ x2.

I""", 2- I .3 ,

Rf',o,p 97~

A'Z'101S'15e>

D;z....,~

lZfi'f' ~7 ~

A·"3· ~I':> '133

Figure 2: Top.' kaolin cia)' pipe bowl and ,I'few (Are(/ C) and l:ed-ol1-lI'ltile glass head (Area A). BO{/Qm: slate pencil m,,1 tahJel.fraglllellls (Area A).Drawings by t?lIssell W(ll'l1Ian.

452.70 grams of domcstics, minimally representing 63 vessels(Table 3). The greatest diversity of domestic materials camefrom this structurc, Stone marbles were included in thedomestic category because of their possible use in bottleclosures. FUl1hcr analysis of the domestic assemblage will benecessary to adequately date and source these materials,

Inside the Solitary Cells, 15 vesscls wcre recovered,weighing a total of 486,81 grams (Table 5), The two highestconccntrations of domestic artefacts occurred in the first-noorhorizon and northern exterior of the cellblock, Although agreater minimal vessel count of ceramics was prescnt, oneferrous container dominated the assemblage by weight

Ecofact

Although this category of cultural materials primarilyconsisted of bone, other finds included oyster shell, snail shell.egg shell. and peach seeds, Due to financial constraints, theRoss Factory ecofact assemblage did not undergo specialistanalysis, During the secondary stage of post-excavation work,basic counts and weights (in grams) were recorded as pan ofthe cataloguing process, Species identification andcalculations of the MNI present were not undertaken,

A comparative examination of bone weights recoveredfrom the three excavation areas produced limited results, Byweight, Area A yielded over five times more faunal matcrialthan the Hiring Class region of Area Br A total of I 029,81grams of bone was rccovered from the interior and exteriorregions of the Crime Class Dormitory, as opposed to 195A Igrams from interior and exterior spaces of the Hiring ClassDormitory (Tables 2 and 4), The largest amount of bone waspresent in the Solitary Cells, with 2 611:36 grams recoveredfrom Area C (Table 5). This faunal material concentrated inthe first-floor layer of the three excavaled cells, with over 50perccnt of the bone from Area C recovered from thisslratigraphic horizon. Other areas of concentration includedthe southern exterior (11%) and the second noor layer (10%),

These preliminary results are significantly biased by both

32

variations in the weight of different skeletal elements from thesame species, and by differcnces in the weight of skeletaelements among different species. Funher specialist studies ofthe fauna I remains from the Ross collection are essential fafuture research on this site, Identification of species preseo:and M I estimations would yield significant new perspectiVe!on diet and foodways within the Ross Female FactOT)':

Fuel

The fuel category incorporated both flammable materials,~by-products of connagration, Four types of materials weft

classified as fucL Charcoal and coal were the most COIl1IlKl

types within thc collection, found within all three excavati(!areas. Wood was recorded separately from charcoaldifferentiate firc-relatcd activities within the site, SOIIX

overlap existed between the Fuel and Structural functiona:categories, as both wood and charcoal could be classifiunder either heading.

Charcoal appeared in highest frequency on the interiorsthe Solitary Cells, with 704.54 grams recovered from Area ((Table 5), Concentrated in the second-nom horizon. charcalso constituted over 58 percent of the cultural materia:recovered from context 3036, a hearth-like deposit of blll'l1jmaterials located between the two tloor horizons in the souther.lhalf of the Central Cell (see Casclla 2001a for discussion"stratigraphic evidence), This data has been interpreted ~

representing an intentional aCl of arson by Factory inmalOundergoing 'separate treatment' (Casella 200Ic).

The category of 'organic slag' consisled of wood Illodifiecby carbonisation, with rapid combustion producing a puff!charcoal resembling slag (Schweingruber 1978: 204), Thtype of material was only recovered in Area A, from mixlXunderfloor and first-demolition period deposits on the intenof the Crime Class Dormitory (Table 2). When combined Wl~

the high frequency of charcoaL the presence of 552:8 grams~

organic slag within this depositional layer might indicate lhl.the dormitory structure was partially damaged by firt

following closure of the Factory in 1855. Coal was alsorecovered in greatest quantity from Area A, specifically fromcontexts on the interior of the Crime Class DormitOly.

Indulgence

The Indulgence classification category was adopted directlyfrom the functional analysis model developed by Adrian andMary Praetzellis (1990). Consisting of kaolin tobacco pipesand olive-glass alcohol bottles, this category represented thepresence of activities forbiddcn to female convict inmates ofthe Ross Factoly. Detailed analyses and interpretations of thisfunctional categOlY are available in recent published andunpublisbed sources (Casella 1999; 2001b). Withio tbisrepon, some patterns in the distribution of this functionalcategory will be briefly summarised.

Dating the kaolin pipes recovered from FacLOry-relateddeposits proved diITicult as no artefact from this highlyfragmented assemblage displayed identifiable makers' marks(Oswald 1975). The most intact artcfact from this assemblagewas an unmarked bowl-and-stem fragment (Figure 2).Displaying a moulded rib design on the bowl segment, thisnon-diagnostic pipe fragment was common throughout thenineteenth-century (Ayto 1979). The fragment was similar inbOlh manufaculre technology and decorative style to clay pipebowls typical to Australian historic sites (Gojak and Stual11999:]9 and Figure 4). At Ross, tobacco pipes appeared to beconcentrated within the Solitary Cells, as both the highestminimum number of vessels (16) and grcatest amount by\\cight (94.76 g) were recovered from Area C (Table 5). AreaAcontained a total of24.57 grams of tobacco pipe fragments,representing a minimum of eight pipes (Table 2). As notobacco-pipe fragments were recovered from the Hiring ClassDOllnitory ofArea B, these results might indicate a di fferentialuse of illicit tobacco within the Crime Class region of RossFactory. At a minimum estimate, three pipes were recoveredITom the Assistant Superintendent's Quarters of Area B,weighing 7. I2 grams (Table 3).

The lowest frequency of olive-glass alcohol bottles wererecovered from Area B. A minimum total of 11 bOllles(754.48 g) were collected from this excavation area. Themajority of this material was recovered from the drain systemlocated On the exterior of the I-liring Class Dormitory (Table4). Only 294.19 grams (MNV ~ 3) of olive bottle glass wasfound within the Assistant Superinlendent"s Quarters,rcpresenting the smallest amount recovered from thec;.:ca\'ated Factory stmctures (Table 3).

In contrast, the majority of olive glass bottles recoveredfrom Area A were collected from the interior of the CrimeClass Dormitory. While a total of 1 574.70 grams (MNV ~ 18)was recovered from Area A. approximately I 528.75 grams(MNV :=; 15) of this glass was found inside the Dormitory"TUClure (Table 2).

Comparing the olive glass distribution between Areas Aand Cproduced some noteworthy results. Although both areascontained a minimum of 18 bottles, a greater weight of oliveglass was recovered from the Crime Class region, withI 57.+.70 grams recovered from Area A, versus 1011.06 gramsrecovered from Area C. When interior deposits of thesestructures are compared, the same differential pattemingexists; I 528.75 grams (MNV = 15) of olive glass wascollected from inside the Crime Class Donnitory, and only312.0 grams (MNV = 10) of olive glass was found inside lheSolitary Cells. However, while a greater minimum numberand weight of olive glass bottles were recovered from Area A,an area three times greater had been excavated in this regionthan in Area C. Furthcnnore, olive glass constituted only 55percent of the total glass assemblage from the Crime ClassDormitory underfloors, compared with over 77 percent of the

total glass assemblage from the Solitary Cell interiors. Thus,olive glass bottles occurred. much less frequently inside theCrime Class DormitolY than within the Solitary Cells.

These results could have been affected by such factors asoccupation density, differential preservation of the record, anddepositional processes. The greater presence of these artefactswithin the earthen floors of the Solitary Cells might also havereflected the limited options for disposal of incriminatingevidence. In her study of the glass assemblage from the BoottMills of Lowell, Massachusetts, Kathleen Bond noted thesignificantly high quantity (by weight) of undiagnosticsmashed glass artefacts in the courtyards that surrounded theworkers' boardinghouses (Bond 1989). She suggested that thisoccurrence might indicate a 'smash and scatter' strategyemployed by the mill girls 10 safely disperse all incriminatingevidence of their illegal alcohol consumption. The lowerfrequency of tobacco pipes and alcohol bOllles within theCrime Class DormitOlY underfloor deposits might indicate agreater variety of options available to those inmates fordisposal of their indulgences.

Literacy

The functional category of Literacy consisted of objectsrelated to reading, writing and numeracy. Within the Rosscollection, two types of artefacts constituted thjs category:slate pencils and slate tablets (Figure 2). Both weights andestimates of minimum numbers present were taken for thiscategory of artefacts.

Artefacts related to literacy were concentrated in Area A,specifically on the interior of the Crime Class DOlmitory. Aminimum often slate pencils (27.71 g) were recovered fromunderfloor deposits within Area A (Table 2). Approximately9.4 grams of slate tablet were also collected from theseslratigraphic contexts. Minimally representing two tablets,these fragments constituted the only presence of slate tabletswithin Factory·related deposits. Area B contained 9.64 grams(MNl ~ 5) of slate pencil fragments (Tables 3 and 4). Theseartefacts were found in highest frequency in the east room ofthe Assistant Superintendent's Quarters, although two pencilswere recovered from deposits on the exterior of the HiringClass Dormitory. No literacy-related artefacts were recoveredfrom Area C, perhaps reflecting the absence of this activitywithin thc Solitary Cells (Table 5).

While embarked upon transport ships and incarceratedwithin the Female Factories, convict women were encouragedto develop their literacy skills through regular training inreading, \.vriting and numeracy. Drawing upon fashionablephilosophies of penal reform, the Convict Departmentconsidered the cultivation of literacy necessary forencouraging obedience and mental discipline, and forenhancing the convicts' value as skilled domestic workers.Female convicts in Australia tended to be more literate thantheir free, working-class counterparts in either England orIreland (Oxley 1996: 161). Since convict women assigned tothe Crime Class were encouraged to attend literacy instructionwithin their dormitories each evening (Seripps & Clark 199 I),the high concentration of literacy-related artefacts within theinterior oflhe Crime Class Dormitory may reflect this activity.

Miscellaneous

One particularly ambiguous type of artefact was recoveredwithin Factory-related deposits. During excavations, fiveqUaltz crystals were recovered from the earthen floors of theAssistant Superintendent's Quarters and the drain system onthe exterior of the Hiring Class Dormitory (Tables 3 and 4).Weighing a total of 105.67 grams. these crystals were foundonly in deposits from Area B. The functional purpose of theseartefacts remains unknown.

33

Monct.uy

The Monetary category accounted for two types of artefacts:legal tender, and merchant tokens from colonial businesses.Three British coins were recovered from excavation trenchesat the Ross Female Factory (Figure 3). Onc copper-alloypenny was found within the eastern room of thc AssistantSuperintendent's Quarters of Area B (Table 3). Depicting afcmale bust, this coin was identified as a Victorian Era issue.On the reverse, the date 1866 could be discerncd. Thepresence ofa post·Factory period coin within thc earthen floorof the Assistant Superintendent's Quarters suggested thisstnJcture had probably experienced some occupation or useafter closure of the Female Factory. Thcrcforc artefactualmatcrials recovered from these earthcn floors were lessdirectly associated with the Ross Factory inhabitants.

Context 1020, an underfloor deposit within the CrimeClass Dornlilory of Area A, contained a highly oxidizedcopper-alloy coin (Table 2). Inscribed with a laurcate bust. andthe word' ...ORGIUS'. no date could bc read on the reverse ofthis coin (Figure 3). Comparative stylistic analysis of coinsissued during the rcign of George 111 identified this artefact asa British copper pcnny, p3l1 ofIhe fourth issue n·om Matthewl3oulton's Soho mint in Binningham (Seaby 1961: 32; Cooper1983: 21). This pcnny best matched thc diamctermeasurcment of those issued during 1806. Since it wasrecovered from context 1020, an Area A underfloor deposit

stratigraphicaily linked to the female-occupation period, itspresence could reOect the illicit possession of legal tender b~

a female convict accommodated within the Crime ClassDomlitory.

Within Area C, a copper-alloy 1823 George IV Britislfarthing was recovered from the first-floor horizon of thecentral solitary cell (Table 5). This artefact was recovered mgood condition. with only a slight degree of oxidisatic.obscuring its surfaces (Figure 3). Stylistic analysis suggest'this coin had been printed during the Soho mint ofBirmingham's first issue for George IV, which ran from 1821to 1826 (Seaby 1966: 212). The location or this coin suggest"it had been deposited during female·convict occupation ofdxsite.

The colonial economy thrived on a second layer 01material exchange. A general dearth of legal currency withicVan Diemcn's Land encouraged lhe development ofcconomknetworks ofsemi-legal barter throughout the colony. Althougtthe majority of Van Dicmen's Land token originated itHobart. copper-alloy merchant tokens were issued by •different businesses locatcd in six different towns around tiltisland (MeNeicc 1969: 6 ). Van Diemen's Land token!typically camc in dcnomin<1lions of either onc penny or one·halfpenny. They wcre widely circulated from the early 18501to 1875, when all non-governmental currency was rccalledb)the Brifish J\1onies Act of 1875. Two of these merchant tokcrn

1<-. Z. ~

.::::-. :';~, .~.~.:

'zf~;/j\;q l<r""'7 [10201 M".1. 1020'1

~ '17 \"""2>1\ iiE::,G" 30372>7

Figl/ll! 3: Top: Brirish coillsfrom ,he ROS.f Collection80110m: merchallllOkells (Area D). Dra"'jllgs by RIJlhrmall.

34

r

were recovered within the Assistant Superintendent's Quartersof Area B (Figure 3). As discussed above, the presence of an1866 Victorian-issue British coin within the earthen floor ofthe eastern rOom indicated a degree of post-Convict Eraoccupation of this structure. However, although the twomerchant tokens could not be directly associated with inmatesor staff of the Female Factory. they do suggest that someFactory or post-Factory inhabitants acknowledged. and likelypanicipated in. the semi-legal economy of mid-nineteenth­century Van Diemen's Land.

The first token was recovered from a deposit of darkbrown silt that had accumulated within the underbarrel drainof Area 8. Approximately 28 mm in diameter, this merchanttoken had been issued by olr H. Hedeberg. a grocer on ArgyleStreet in central Hobart Town. Named the 'Swedish House.'Hedeberg's business catered to a particular ethnic minorityamongst the free settlers.

The second token was recovered from the earthen floor ofthe western room inside the Assistant Superintendent'sQuaners. This copper-alloy token had been issuecl hy R.Andrew Mather, a leading draper and clothier of Hobart Town.The Mather token was 33 mm in diameter. It appeared to bethicker in cross-section. and stamped with less detail than theHedberg token. Although these differences might haverenected variations in the manufacture process, a similarMather token recovered from Port Arthur~a male-convictsite on the Tasman Peninsula of Van Diemen's Land-did notshow similar signs of substandard manufacture (McGowan1985: 33.116). Recovered rrom a post-I 848 domestic depositinside 'Lithcnd,' a staff cottage within the Port Arthur penalcompound, this Li\hend Malhcr token waS slmilar in size,thickness and clarity of design details to the Hedeberg tokenfrom the Ross Factory collection. Tasmanian numismatistRoger McNcice observed thal the Mather token wasfrequently forged. Although these illegal copies were usuallyslamrerl in lead, other melal fabrics were also used. Thus, it ispossible Ihal the Mather token recovered from the RossFactoI)' Assistant Supcrintendent's Qum1ers was a forgely,although post·depositional impacts cannot be fully dismissed.

Social Control

The category of Social Control consisted of objects related tome phy~ical enforcement of domination. Three types of suchanefacts were found within Factory related deposits. Ferrouschain links were recovered from all three excavation areaS.The greatest concentration (524.20 g) of these artefacts wasfound within underfloor and demolition deposits on theinterior of the Crime Class Dormitory (Table 2). This evidencecould suggest that these chain links entered the site afterclosure of the Factory in 1855. Rather than relating to thee:<ercise of social conlroL they might have served somestructural or demolition related function. Within Area C, 67.39grams of fcrrous chain links were recovered from the first­floor horizon, and 40.40 grams from the northern exterior oflhe cellblock (Table 5). 0 ferrous chain links were recovered\\ithin the Hiring Class Domlitory and only 37.13 gramsfound within the eastern room of the AssistantSUp'nnlendent's Quarters (Table 3).

Two other objects related to social control were recoveredfrom Arca B. One ferrous padlock was recovered from theeaSlcm room of the Assistant Superintendent's Quarters.He,wily oxidised, it was intact, heart-shaped and weighed156.67 grams.

Finally, a ferrous gun-powder Illusket flask was recoveredfrom sill deposits within the underbarrel of the sandstone drain~ystem on the exterior of the Hiring Class Dormitory of AreaB (Table 4). Capped with a copper-alloy self-measuringrelease valve SpOUI, this pear-shaped artefact was

manufactured from 1750 to the late nineteenth-century (Heldand Jcnkins 1959: 134). Heavy oxidisation of the felTous bodyof this artefact obscured identification details, including anydistinguishing military insignia or decorative marks.

This ferrous artefact was recovered from context 2026, abrown alkaline silt deposited within the base of the Area Bdrain underbarrel. It was located approximately 70 cm into thecourse of the box-drain. Context 2026 was an alluvial silt, andhad acculllulated through the noollal functioning of the drainsystem in Area B. These drains were installed in 1841. duringexpansion of the penal site to provide accoJlllllodation formale Road Gang convicts. After closure of the Factory in1854. the drains were no longer maintained regularly. Sinceseasonal floods of the Macquarie River frequently inundatelhe site, the drain llnderbarrel would have rapidly cloggedwith alluvial sediments. Thus. the musket flask probablyrelated to one of the convict periods of site occupation.Depositing this flask within the drain underbarrel would haverequired raising the heavy sandstone drain bowl. Thus. themusket nask appeared to be intentionally hidden deep withinthe drain underbarrel~a rather inconspicuous andinappropriate location for a cOlllainer of gunpowder within apenal institution.

Structural

Structural materials consisted of artefacts relatcd to thephysical fabric of Factory buildings, architectural features,and furnishings. The functional category included: tcrracottabrick, mOl1arlplaster, ferrous nails, fermus other (screws,bolts, llutS, wire, flat fragments, and unidentified fragmcnts),copper othcr (wire, flat fragments. and unidentifiedfragments), lead, sandstone cobblestones, slare. timber(fragments of joists, beams or roof shingles). clear lamp glassand clear window glass. By weight. rhe 'Structural' categoryformed a dominant component of the recovered artefactualassemblage in al\ three excavation areas. A \though morcresearch must be completed to comparatively analysematerials from this category, some preliminary trends can bediscussed.

The greatest concentration of structural materials wasrecovered from mixed undernoor-demolition contexts withinthe Crime Class DonllilOry of Area A (Table 2). While thisresult was in itself unremarkable, the panicularly highfrequency of clear window glass present in this stratigraphiclayer (500.73 g) corroborates with documentary evidence forthe installation of paned windows in the convict d01111itories in1847, in preparation for the Female Factory occupationperiod. Within Area C. a relati\'ely high amount of clearwindow-glass (219.84 g) was also recovered from thesouthern exterior of the Solitary Cells (Table 5). Extrapolatingfrom contemporary penal design. the southern windows weremost likely small and rcctangular, installed just below ceilinglevel to provide necessary air circulation. while limiting visualstimulation and preventing escape (Evans 1982).

Although a far greater amount of structural materials wasrecovered from the interior, versus exterior. of the Crime ClassDormitory, the oppositc trend appeared in the Hiring Classregion of Area B (Table 4). The exterior of the Crime ClassDormitory yielded 188.17 grams of structural artefacts, whileexcavations of the exterior pathway and drain system of rheHiring Class produced 491.53 grams of these materials, orapproximately 2.6 limes morc structure-related artefacts. Thisdiscrepancy could represent a combination of causes,including differences in the process of dcmolition employedon the two dormitories during later periods of site occupation,or increascd robbing and recycling of structural materials fromthe Hiring Class Dormitory building. Further research isrequircd to adequately evaluate these results.

35

DISCUSSION

Results of this fUllctional analysis present some interestinginter-regional material patterns. Comparison of all threeexcavation areas demonstrates that both the greatest amountand greatest functional diversity of artefacts were recoveredfrom Ihe Main Compound of the Factory (Tables 2 to 4). asopposed to the SolitalY Cells (Table 5). Although this is arather predictable distribution pattcm, it is nonetheless worthnoting the significant amount and diversity of culturalmaterials recovered from the Solitary Cells-that region orthesile dedicated to the enforcement of austere disciplinarydeprivation.

Within the Main Compound. the AssistantSuperintendent's Quarters (Table 3) cOllwined the greatestdiversity of cultural materials. Unfortunately, this region of thesite was also related to the Factory-occupation period with theleaSl degree of cel1ainty. The brick Assistant Superintendent'sQuarters was originally constructed for the accommodation ofmale convicts during the Bridge Gang occupation period of theearly 1830s. Stratigraphic evidence suggests that unlike theCrime Class and Hiring Class Dormitories, the earthen-floorcdQuarters had never been installed with wooden fioorboardsduring preparations for the establishment of the FemaleFactory in 1847. Thus, the earthen floors oflhe excavated eastand west rooms most likely held a mixture of pre-1848 andFactory-relatcd artefacts. Finally, as presented earlier, therecovery of an 1866 Victorian penny within the eastern roomdeposits also suggested a significant degree of post-Factoryimpact on this structure.

The two regions of the site predominantly occupied byfemale convicts were the Crimc Class Dormitory (Table 2)and the Solitary Cells (Table 5). Comparison of rcsults fromthese two areas reveals that materials from the two functionalcategories related to personal appearance, Adoll1l11ent andClothing, appeared to be concentrated within the Crime ClassDormitory. With the significant exception of buttons,Clothing-related artefacts were recovered in greater number,weighl and diversify from the D0l111itory. All seven beadsrecovered at the Ross Factory were found in the Dormitoryunderfloor deposits; no Adornment-related artefacts occulTedwithin the Solitary Cells.

Finally, matcrials from the Indulgence category wererecovered in greatest concentration from the Solitary Cells. Aspresented before this pattern might reflect a lack of disposalalternatives within the Cells. Although detailed evidcnee forthe acquisition and Iranspol1 of these illicit artefacts can befound in other sources (Casella 2000; Casella 2001 b), it iswOl1h noting lhat the archaeological presence of olive-glassalcohol bottles and kaolin tobacco pipes might matcriallysignify internal networks of cavelt blaekmarket exchange.These' illicit objects' could therefore renect subversions of theboundaries, fencelines and architectural isolation of the RossFactory. Ultimately, the high presence of forbidden'indulgences' in Area C suggests that female convictssentenced to periods of 'separate treatment' cnthusiasticallymaintained their access to diverting luxuries.

FUTURE DIRECTIONS

The Rass FaclOry Archaeology Project was designed toexamine reciprocal power relations as reflected in both thebuilt environment and the disl'I-ibution of portable artefacts.Although results of the functional analysis have illuminatedsignificant pallerns of site use during the Facwry-occupationperiod, fUflher research could reveal additional aspects of lifewithin this unique convict site, and would provide newdirections for the archaeology of con fi nemenl. Two avenuesseem pal1icularly rich for future study.

36

The Fabrication of Discipline

Excavations of the Solitary Cells (Area C) only examinedapproximately 17 percent of this important structural feature.Givcn this limited sample, one particularly valuable futuredirection would be the continued analysis of this region of theRoss Factory site. Opening fUlther trenches to the north andsouth of the Solitary Cells would provide comparative data fortesting interpretations of the internal architecture of thestructure. Interweaving slratigraphie and architectural data. Ihave argued that these cells contained twO layers of earthenfloor, separated from the entrance level by a drop of at leasl30 cm (Casella 1997; Casella 200 Ic). However, if these cellshad been original1y floored with suspended woodenfloorboards, a common building feature during the mid­nineteenth-century, il is possible that the stratigraphic contextsidentified as the upper 'F2' floor actually represented theconstruction surface. In this case, anehlcts recovered from thishorizon would have been deposited by the male convictlabourers as they built the sandstone cellblock. Following thisalternative model, the stratigraphic contexts defined as thelower 'FI' noor would pro-date the Solitary Cells. Artefactsrecovered from thal earlier horizon would represclll thegeneric scaner of nineteenth-century materials washingdownhill from the Main Compound of the Ross Station duringthe 1830s and 1840s.

However. results generated by a functional analysis of theartefacts recovercd from excavalions might contradict thisalternate architectural interpretation. In Table 5, resultsdemonstrate that the First Floor (F I) contained I 411.82 gramsof structural materials. while the Second Floor (F2) yieldedonly 671.92 grams. Thus, although structural materialsaccounted for a relatively grcater component of the totalartefact assemblage in the Second Floor (55% of F2 versus41 % of F I), the First Floor contained twice as many structure·related artefacts by weight. With no building construction ordemolilion activities historically documented in this southernregion of the site before erection of the Solitary Cells in 1851(Scripps & C1ark 1991), this high rrequency or strueturnlmaterials in the first-floor horizon could identify that featureas a combination construction-and-first-occupation floor.rather than a generic scattcr of cultural debris related 10

occupation of the Main Compound before 1851.

By extending excavation trenches lO the immediate northand south of the cellblock, wc could evaluate the alternmh'earchaeological interpretations or Area C. To sustain Ill}"

interpretation of two superpositioncd floor features wilhin thecells, the density of artefactual materials would need toconcentrate on the interior of the cells, and diminish to thenorthern and southern exteriors. If no change in artefacldensity were present, the lower first-noQl' horizon is morelikely to represent the "background noise' of seconda~

deposition, rather than an architectural floor fcature. Thesencw data would then enable a re-evaluation of physicalevidencc for the presence of cultural deposits and socialactivities within Area C.

Food and the black market economy

Preliminary identification and inventories of ferrous, faunaland ceramic assemblages suggests a significant discrepanC}betv,/een the physical evidcnce for food within convictdormitories and cells, and the official scale of food rationscalculated by the colonial Convict Department (Scripps &Clark 1991: 20). Furthermore, in their 1842 testimony to tllrCommittee of Inquiry into Female Convict Prison Discipline.convict informers explained how their fellow inmates easil)supplemented their officially issued rations through illegaltrade in rood (AOT CSO 22/50). A detailed study or raun.materials and ethnobotanical rcmains from the Ross collcctiOll

would directly investigate this issue of illicit food exchangewithin the prison. A detailed specialist analysis of the ferrouscontainers and ceramic assemblage recovered froIn the CrimeClass DonnilOry and Solitary Cells would also provide newdata on the procurement and distribution of food within theFactory. Ultimately, such research would provide furthermaterial il1'5igh\ into anothcr means by which female convictstransgressed the disciplinary regulations and spatial barriersthat structured their institutionaliscd lives.

The Ross Factory Archaeology Project added ncwmall:rial perspectives to current research debates on Australianfemale convicts. Using functional analysis methods developedin American historical archaeology. this research hasexamined the cultural remains related to tbe Female Factoryoccupation period. Results of this analysis provide us withnew physical evidence for both the institutional structures ofconfinement and the covert aspects of life within the RossFactory.

ACKNOWLEDG EM ENTS

Funding for the Ross FaclOlY Archaeology Project wasprovided by the Wenncr-Grcn Foundation. lhe University ofCalifornia Pacific Rim Research Program. and thc StahlEndowment of the Archaeological Research Facility,Lniversity of California at Berkeley. Final stages of analysiswere supported through grants from The \Voodrow Wilsol1Foundation and the American Association of UniversityWomen. The Parks and Wildlife Service of Tasmania and thcQueen Victoria Museum and Art Gallery (Laullceston,Tasmania), the Tasmanian Wool Centre of Ross, and theRiggall family of the Sommercoates propclty all generouslyprovided essential support throughout the years. Mary Reitcnand Matthew Chester offered skilled assistance during post­excavation stages of laboralory work. Particular thanks toGinni Woof for directing the field laboratory during bothc'\cavation seasons. Finally. I would like to remember MartinDavies who helped initiate the Ross FactolY ArchaeologyProject

BIBLIOGRAPHY

ArchIVes of Tasmania (AOT) CSO 22/50. ColonialSecretary's Office.

AOT PWD 266/1693. Plan of road station barracks at Ross.c.1842.

~OT PWD 26611695. Plan of the alterations of tbe RossProbation Station 10 the Ross Female House of Correction.184 .

~OT PWD 266!1696. Plan of the alterations of the RossFemale Faclory, c.1851.

.~~lO. E. G. 1979. Clay Tobacco Pipes, Shire Album No. 37,Shire Publications Lld, Aylesbury. United Kingdom.

Birmingham, 1. 1992. Wybalellna: The Archaeology ofCultural Accommodation ill Nineteenth CenturyTasmania, The Australian Society for HistoricalArchaeology Incorporated, Sydney.

Bond. K. 1989. 'The Medicine, Alcoltol, and Soda Vesselsfrom the Boot! Mills Boardinghouses', in M. Beaudry &S. Mrozowski (eds), IlItel'disciplincuy Investigations a/theBool/ Mills, LoweJl. Massachusells. Volume 3, CulturalResources Management Study No. 21, National ParkService, NOlth Atlantic Regional Office, Boston, USA.

Brand, J. 1990. The COl/vict Pl'Obarioll System, Blubber HeadPress. Hobart.

Casella, E. 1997. ' "A large and efficient establishment":preliminary report on fieldwork at the Ross FcmaleFactory', Australasian Historical Archaeologv 15: 79-89.

Casella, E. 1999. Dangerous Girls and Gentle Ladies:archaeology and nineteenth centllry Australian fcmalecOllvicts. unpublished PhD dissertation. Department ofAnthropology, University of California, Berkeley, USA.

Casella, E. 2000. ' "Doing Trade": a sexual economy ofnineteenth-century Australietll female convict prisons',World Archaeology 32 (2): 209-221.

Casella. E. 200 la. (in press) Archaeology o/The Ross FemaleFactory: Female Incarceration in Van Diemell S Land,Australia, Records of the Queen Victoria Museum and ArtGallery. Volume 106. QVMAG Publications. Laul1ccston,Tasmania. Australia.

Casella. E. 200 I b. . Landscapes of Punishment andResistance: A Female Convict Settlement in Tasmania.Australia·. in B. Bender & M. WineI' (eds), ContestedLandscapes: Landscapes of MOl'emel11 and Exile. BergPress. Ox ford. pp. 103-118.

Casella, E. 200\c. 'To Watch or Restrain: Female ConvictPrisons in 19th Century Tasmania. Australia',Imematiollal Journal of Historical Archaeology 5 (I):45-72.

Cooper, D. R. 1983. Coins alld Minting, Shire PublicationsLtd .. No. 106, Aylesbury, United Kingdom.

Damousi, J. 1997. Depraved alld Disorderly, CambridgeUniversity Press, Cambridge.

Daniels, K. 1998. COllvict Women, Alien & Unwin, Sydney.

Davics, M. and K. Buckley 1987. Port Arl!1l/r COllservatiollal/d Development Project: Archaeo/'Dgical ProceduresMalltial, Parks & Wildlife Scrvlcc (DPlWE), OccasionalPaper Number 13, Hobalt.

Dixon, P. W. 1994. Crickley Hill: The I-lill/OI"I De/ences,Crickley Hill Trust and the Department of Archaeology,University of Nottingham, United Kingdom.

Driver, F. 1993. Power and Pauperism: The WorkhouseSystem, 1834-1884, Cambridge University Press,Cambridge.

Eldershaw. P. R. 1968. 'The Convict Department'. TasmanianHistorical Research Association. 15(3): 130-149.

Evans, R. 1982. The Fabrication of Virtue, CambridgeUniversity Press, Cambridge.

Gojak, D. and I. Stuart 1999. 'The Potential for tbeArchaeological Study of Clay Tobacco Pipes fromAustralian Sites'. Australasian llistorical Archaeolog)' 17:38-49.

Held, R. and . Jenkins 1959. The Age ofFirearms. CasseIl &Cornpany Ltd, London.

lacono. N. 1996. Cumberland/Glollcester Streets SiteArchaeological Investigations 1994, Volume 4: ArtefactReport, Miscellancous. unpublished report prepared forthe Sydney Cove Authority and the Heritage Council ofNew South Wales, Godden Mackay Heritage Consultants.Sydney, Australia.

Karklins, K. 1982. ·Guide to the Description andClassification of Glass Beads', Parks Canada: Hist01Yand Archaeology 59: 83-117.

Kidd, K. and M. Kidd 1970. 'A Classification System forGlass Beads for the Use of Field Archaeologists',Canadial/ His/oric Sites: Occasional Papers illArchaeologr Gnd His/my, Parks Canada, OI13\.\/a, I; 45-89.

Lydoll, J. 1993. 'Task differentiation in historical archaeology:Sewing as material culture', in H. du Cros & L. Smith(eds). Women ill Archaeology: A Feminist Critique,Occasional Papers in Prehistory No. 23, Department ofPrehistory, Research School of Pacific Studies, theAustralian National University. Canberra, pp. 129-133.

37

McGowan, A, J985. Excavalions at Lithend. PorI Art/HII'Historic Site, Parks & Wildlife Service (DPIWE),Occasional Paper No. 11. Hobmt.

McNeice, R. V 1969. Coins and Tokens of Tasmania,18()3~1910, Platypus Publications, Hobart.

Oswald, A. 1975. Clay Pipes for the Archaeologist, BritishArchaeological Reports. British Series 14.

Oxley, D. 1996. COl/viet Maids, Cambridge University Press,Cambridge.

Praetzellis, M. and A. Praelzellis J990. For A Good Boy:Victorians 011 SaCl'GlllenlO :" J Street, Cultural ResourcesFacility, Anthropological Studies (enler. S0110ma StaleUniversity, Rohncrt Park, USA.

Rayner, T. 1980. l-liSlorical Survey of the Ross Female FactorySite. Tasmania, unpublished repon prepared for theNational Parks & Wildlife Service, Hobart, Australia.

Ross, L. 1997. 'Glass and Ceramic Trade Beads from theNative Alaskan Ncighborhood', in K. G. Lightfoot, A. M.Schiff & T. A. Wake (cds), The Archaeology alldEl/lI1ohistOJY of Fort Ross. Cal{fornia, Vo/. 2: The NativeAlaskan Neighborhood, Contributions of the University ofCalifornia Archaeological Research Facility, No. 55,Berkeley. USA.

Ryall. L. 1995. 'From Stridency to Silence: the Policing ofConvic' Women 1803~1853', in D. Kirkby (cd.), Sex,Power and Juslt'!;e, Oxford University Press, Melboume.

Schwcingrubcr, F. H. 1978. Microscopic Wood Anatomy(Trans. K. Baudais-Lundstrom), Swiss Federal Institute ofForestry Research, Zurich.

Scripps, L. and 1. Clark 1991. The Ross Female Factory,unpublished report for the Tasmanian Parks & WildlifeService, Depactmcnl of Parks, Wildlife and Heritage,L-lobal1, Australia.

Seaby, H. A. 1961. British Copper Coins and their Values.Parr 1: Regal Coins, Seaby's Numismatic Publjcations,London.

Scaby, 1.-1. A. 1.966. Standard Catalogue o.lBritish Coins. B. A.Scaby Publications, London.

Spector, J. 1976. 'The Interpretive Potential of Glass TradeBeads in Historic Archaeology'. Historical Archae%KL'10: 17-27.

Sprague. R. 1985. 'Glass Trade Beads: A Progress Report',Historical Archaeology 19 (2): 87~105.

Terry, I. /998. Ross Female Convict Station Hisroric Site:Conservation Plan, unpublished report for the Parks &Wildlife Service, Department of Environment and LandManagcmenL, Hobart, Australia.

38