Hst1998 Report

8/2/2019 Hst1998 Report

1/87

Hofstair 1998Framvinduskrslur/Preliminary Reports

Ritstjrar: Adolf Fririksson & Orri Vsteinsson

Fornleifastofnun slandsFS062-91016

Reykjavk1998

8/2/2019 Hst1998 Report

2/87

2

Fornleifastofnun slandsBrugtu 3101 [email protected]://www.mmedia.is/fsi

Reykjavk 1998

8/2/2019 Hst1998 Report

3/87

3

Efni / Contents:

ADOLF FRIRIKSSON & ORRI VSTEINSSON.......................................

1.0 HOFSTAIR 1998. SAMANTEKT OG YFIRLIT ...............................5

Inngangur.....................................................................................................................5Markmi ......................................................................................................................5Svi A .......................................................................................................................6

Mannvirki fr 18./19. ld. .........................................................................................6Mannvirki fr 9.-11. ld............................................................................................6A2 ............................................................................................................................7A3 ............................................................................................................................7Suurgafl skla (AB) ................................................................................................7

Innbyris afstaa mannvirkja svi A, D og G. .....................................................7Svi D .......................................................................................................................8Svi E........................................................................................................................9Svi G .....................................................................................................................10Niurlag .....................................................................................................................11

GAVIN M LUCAS 2.0AREA A EXCAVATION REPORT................ 13

Introduction................................................................................................................13Excavation Results.....................................................................................................13

Phase IV (20th century) ...........................................................................................13Phase III (18th/19th century) ....................................................................................14

Phase II (12th

-18th

century) .....................................................................................14Concluding Discussion...............................................................................................18

RAGNAR EDVARDSSON 3. 0 AREA D EXCAVATION REPORT......23

Introduction................................................................................................................23Excavation Results.....................................................................................................23

Phase III (20th century) ...........................................................................................23Phase II (c. 1100-1908)...........................................................................................24Phase I (9th 11th century) ......................................................................................25

Conclusions................................................................................................................29

HOWELL ROBERTS 4.0 AREA E EXCAVATION REPORT................. 37Introduction................................................................................................................37Results .......................................................................................................................37

Structure E1............................................................................................................37External Deposits ...................................................................................................44Structure E2............................................................................................................46Skli A/B................................................................................................................48Context 1004..........................................................................................................49

Discussion / Conclusions............................................................................................49Structure E1............................................................................................................49Structure E2............................................................................................................51Heat Sources and Context 1004 ..............................................................................52Table 4-1. List of contexts in Area E.......................................................................56

8/2/2019 Hst1998 Report

4/87

4

TOM MCGOVERN 5.0 AREA G EXCAVATION REPORT................ 58

Feature G 1998 Season Objectives..............................................................................58Investigations Carried Out in 1998 .............................................................................58Recovery Methods......................................................................................................60

Conditions of Preservation..........................................................................................60Finds ..........................................................................................................................62Bone and Shell ...........................................................................................................62Botanical Remains......................................................................................................62Stratigraphic Observations and Interpretation .............................................................62Acknowledgements ....................................................................................................64

KAREN MILEK, IAN SIMPSON & SARAH BEVERIDGE

6.0 GEOARCHAEOLOGICAL SAMPLING REPORT............................. 65

Introduction................................................................................................................65

Soil and Sediment Sampling At Hofstair: Rationale and Procedure...........................65Deposits Within Structures .....................................................................................65Midden Deposits: Area G .......................................................................................71Soil Profiles within the Infield ................................................................................76Possible Frost-Heave Features ................................................................................76

Floor Deposits Collected from ver ..........................................................................77Methods for Processing and Analysis of Samples .......................................................79Conclusion and Proposal for Future Work ..................................................................80Acknowledgements ....................................................................................................80

APPENDIX 1: LIST OF FINDS .................................................................81

HEIMILDIR/REFERENCES......................................................................84

8/2/2019 Hst1998 Report

5/87

5

Adolf Fririksson & Orri

Vsteinsson

1.0 Hofstair 1998. Samantekt

og yfirlitFornleifastofnun slands, Brugtu 3, 101

Reykjavk.

Inngangurri 1991 hf Fornleifastofnun slands

rannsknir a Hofstum Mvatnssveit.r minni httar athuganir sem gerar vorurin 1991, 1992 og 1995 leiddu ljs a Hofstum voru kjrastur til rannskna

upphafi bsetu slandi. ri 1996 varv hafist handa vi yfirgripsmiki rann-sknarverkefni og er fyrsta fanga ess(1996-1998) n loki.

ur hefur veri ger grein fyrir rangriverksins rin 1996 og 1997 (sj AdolfFririksson & Orri Vsteinsson 1996,1997a-b, 1998) en hr verur rakinframvinda og rangur uppgraftarins 1998.Rannsknin skiptist nokkra tti og hafaveri skrifaar stakar skrslur um hvern

eirra (sj kafla 2.0-6.0). viauka aftast essu skrslusafni er a finna skr yfirfundi.essi skrsla hefur ekki a geyma

endanlega rvinnslu uppgraftargagna,heldur er hn fyrsta samantekt og drg atlkun eim mannvistarleifum semrannsakaar voru sumari 1998. msumathugunum eim efnivi sem safna varsumari 1998 er enn ekki loki, auk esssem margar af eim byggingaleifum sem ljs komu eru ekki fullrannsakaar og erv endanlegra niurstana ekki a vnta br.

Fornleifarannsknir Mvatnssveit hafaveri studdar af mrgum ailum.Rannsknarr slands greiddi kostna vifornleifauppgrft, Sktustaahreppur forn-leifaskrningu og jminjasafn slands,NABO, Hunter College, National Geo-graphic, og hsklarnir Sheffield, Stirling

og Cambridge fornvistfrirannsknir.Samlia rannsknunum var haldi nm-

skei verklegri fornleifafri. Nmisttu 15 erlendir nemendur og styrktirkissjur sklahaldi. Verkefninu stjrnaAdolf Fririksson og Orri Vsteinsson.

Gavin Lucas hafi heildarumsjn meframkvmd uppgraftarins og stjrnai jafnframt uppgrefti svi A. RagnarEdvardsson stjrnai uppgrefti svi D,Howell M. Roberts svi E og ThomasMcGovern svi G. Garar Gu-mundsson hafi umsjn me fornvist-frirannsknum llu svinu, Ian A.Simpson stjrnai jarvegsrannsknum.Karen Milek jarvegsfringur annaistrannsknir glflgum. Vi rannsknirnar

unnu auk eirra fornleifafringarnirHildur Gestsdttir og Mjll Snsdttir.Auk fjrstunings sja og stofnana hefurHofstaaleiangur 1998 noti astoarmargra Mvetninga, ekki sst smundar ogGumundar Jnssona Hofstum. Hafieir bestu akkir fyrir. essu yfirliti vera rifju upp heildar-marmki verkefnisins og lst sturannskna hverju svi.

MarkmiMarkmi rannskna Hofstum er arekja sgu bskapar og bsetu fyrstukynsla slendinga. Lg er hersla anota aferir fornleifafri ogfornvistfri til a afla nrra heimilda umetta efni. Rannsknunum er srstaklegatla a varpa ljsi nokkur samtengdatrii:? hsakost landnema og run

byggingakosts fyrstu ratugina.? ntingu aulinda umhverfi Hofstaa,

einkum sambandi vi eldivi,byggingarefni og efni til sma.

? innflutning efni og matfngum.? herslur skipulagi bskapar, veia og

hsdrahalds, m.t.t. mataris oglfsafkomu

? hrif nttru og umhverfis (veurlag oggrurfar) bskaparhttifrumblinganna og hrif eirra umverfisitt (beit, rktun og efnisflun).

? efnahag, flagsger, mannfjlda,

8/2/2019 Hst1998 Report

6/87

6

heimilisstr og mtun samflagsgerar,einkum m.t.t. runar valds ogmismunandi agangs a aulindum.

Vi rvinnslu rannsknargagna er tefltsaman rangri uppgraftarins Hofstum

og heildarskrningu fornleifa Mvatnssveit.

Vi framkvmd verkefnisins var essumheildarmarkmium skipt upp fanga. 1998lauk rum fanga Hofstarannskna semstai hafi san 1996. Undirmarkmiina essu sinni voru a:a) ljka uppgrefti svi D,b) ljka uppgrefti svi E,c) ljka uppgrefti fyllingu gryfjunni

svi G ogd) rannsaka suurgafl skla A/B og

samhengi hans og gryfju.essum markmium var a mestu n sumar og verur rangurinn rakinn hr stuttu mli.

Svi A

ri 1998 var ntt svi opna vi

suurenda sklans. Ljst var eftir arannsknir hfust jarhsi og fyllingum G sunnan sklans 1995 og afhsi visuvesturhorn hans 1996-7 a samhengiskla, jarhss og afhsa yri ekki fyllilegaskili n ess a rannsaka etta svi. Aukess hafi ekki veri gengi r skugga umhvort mannvirki vri a finna visuausturhorn sklans ri 1908. uppdrtti Bruuns fr 1896 eru snd afhsbi vi suvestur- og suausturhorn

sklans en sarnefnda byggingin var ekkiathugu frekar og kemur ekki fram sariuppdrttum. rannsknarskrslum fr 1908er fullyrt, a ekki hafi veri torfveggur fyrirsuurgafli sklans. Markmi rannskna svi A sumari 1998 var rtt:a) a kanna afstu skla, afhss D og

jarhss G.b) a leita ur ekktra mannvistarleifavi suausturhorn skla ogc) a kanna leifar suvesturgafls skla.

rangur uppgraftarins er rakinn skrslu Gavin Lucas, en hr verur gefi

stutt yfirlit um helstu atrii.Opna var strt svi vi suurenda

sklans, 17 m langt (A-V) og fr 3,4 til 9,4m breitt (N-S). Svi var breiara visuurenda sklans og var ar grafi mta

langt suurfyrir sklagaflinn eins og inn hsi, en austan vi ni uppgraftarsviekki jafnlangt til suurs. 1908 hafi aeinsveri grafi innan r sklanum essusvi nema einum sta ar sem mjrskurur hafi veri grafinn gegnumsuurgaflinn. uppgraftarsvinu komu ljs byggingaleifar fr fjrumtmaskeium.

Mannvirki fr 18./19. ld.

Utan vi sklatftina, suausturhorniuppgraftarsvisins undir brekkunni ofantftanna, fundust nlegar mannvirkjaleifar.Aldurskvrun eirra byggir gripumsem ljs komu og afstu til gjskulaga.Til essa hafa eingngu fundist minjar fr9.-11. ld vi uppgrft Hofstum.Hofstaabndur minnast ekki a essumsta hafi stai mannvirki og ekki er essheldur geti lsingum fornfringa fr19. og 20. ld. Ekki var anna eftir afessum byggingum en regluleg grjthrgasem sneri A-V og virist vera norurhli byggingu. essar veggjaleifar skeratorfvegg sem einnig virist vera norurhli mannvirki, sennilega skylt hinu semgrjtveggurinn tilheyri. essarveggjaleifar voru byggar ofan gryfju semni lengra til norurs og hafi veri grafin gegnum mun eldri byggingaleifar. Umr verur fjalla sar. Gryfjan sker

gjskulagi fr 1717 og er v yngri ena, lklega fr 18.-19. ld. Hlutverkessara mannvirkja er ekki ljst entluverar heyleifar fundust sem benda tila gryfjan gti hafa veri heytft.

Mannvirki fr 9.-11. ld.Fyrir utan au svi ar sem jarvegur

var fjarlgur vi uppgrftinn 1908 erallsstaar jarlagastabbi me gjkulgumfr 1104/58 til 1717 yfir minjunum, en

LNS undir eim.Uppgrefti var ekki loki neinum af eim

8/2/2019 Hst1998 Report

7/87

7

minjum sem komu fram sumari 1998, enhins vegar fkkst mynd af svinu, semgefur ga von um rangursrkaframhaldsrannskn:

A2Austan vi suurenda sklans kom ljstorfveggur sem virist vera suurgafl hsssem hefur veri sambyggt sklanum.Veggurinn er hlainn r grgrnum strengog snr A-V. Hann sst 4.8 m kafla vinorursni uppgraftarsvisins. yfirborim sj mta fyrir hsinu, enda teiknuuBruun (1896) og Brynjlfur (1900) afhsvi sklann essum sta. Bruun hefur hinsvegar ekki grafi a 1908. yfirbori

minnir essi bygging D1, .e. afhsi visuvesturhorn sklans, a ru leyti en va A2 virist hafa veri byggt upp visklavegginn, en nokku bil er millisklans og D1. A2 ltur t fyrir a vera um10 m langt og um 5 breitt. A2 er klrlegayngra en sklaveggurinn essum sta ena ru leyti verur ekki sagt um gerhssins ea afstu ess gagnvart rummannvistarleifum a svo komnu mli.

A3Um 3 m austan vi suausturhorn sklans,og 2 m sunnan vi suausturhorn A2 komu ljs leifar af stakri byggingu. Lti vareftir af henni ar sem hn hafi a strumhluta veri eyilg egar gryfja var grafinofan hana 19. ld (sbr. a ofan). sstnorurgafl og hluti af vesturvegg og viristhsi hafa veri 2.2 m breitt og um 5 mlangt. Snilegar leifar af veggnum eru allt

a 0.6 m breiar og er hann hlainn rstreng. vesturveggnum voru dyr semhfu veri fylltar upp. Hsi er suausturhorni uppgraftarsvisins og ersuur- og austurhluti ess enn grafinn.

Suurgafl skla (AB)Hreinsa var ofan af suurenda sklans,

um 4 til norurs fr suurgafli. ljs komutorfhleslur og tvfld steinar sem amestu leyti voru gu samrmi vi

teikningar Bruuns. Mannvistarleifar innansklans voru ekki kannaar a essu sinni,

heldur var lg hersla a kannasuurgafl hans og leita vsbendinga um

jarlagaskipan og samhengi vi gryfjuna(G) og minjar vi suvesturhorn (D). ljs kom a s tlkun Bruuns og Finns

Jnssonar um a enginn torfveggur hafiveri fyrir suurgafli sklans er rng. areru augljsar leifar af torfvegg rklmbruhnaus. Bruun hefur grafi aessum vegg, og talsvert inn hann en afeinhverjum stum ekki tali hannuppfylla r krfur sem hann geri tiltorfveggja. skrslum talar hann um apallur r torfi hafi veri essum sta og hann ar kannski vi essar veggjaleifar.a er illskiljanlegt hversvegna Bruun s

ekki vegginn v hann er eins og arirveggir sklans sem grafnir hafa veri fram.Helst gti muna a hr er hann vi lgrien annarsstaar. 1908 hafi veri grafialllangt inn innri hli sklaveggjanna ogvi vesturvegginn sst a torfhleslurnarhafa upphaflega n a innri steinarinniog jafnvel upp hana. Allt a 0,8 m afveggjunum voru fjarlg vi uppgrftinn1908 og suvesturhorni sklans var grafivert gegnum klmbruhlesluna ogsuurendi austurveggjar hafi einnig verifjarlgur gersamlega. essu svivirist sklaveggurinn allsstaar veratvskiptur, me um 1 m ykkriklmbruhleslu a innan en reglulegristrengjahleslu a utan, sem virist hafaveri heldur ykkari a vestan en austan.Frekari athuganir ytri brn austurveggjargeta tt eftir a leia anna ljs.Suurgaflinn er um 9 m a lengd a

utanmli. Engar dyr eru suurgafli.

Innbyris afstaa mannvirkja svi A, Dog G.Hsin sem stu austan vi sklann (A2

og A3) voru reist ofan uppmokstri ogtorfhrunslgum. Rannsknin ni ekki tilneri laga a essu sinni en mtti sj aundir essum lgum var grtt ruslalag(0106) sem svipar mjg til lags ea

lagasyrpu sem fundist hefur va Hofstum (0004, 1004). etta lag fannst

8/2/2019 Hst1998 Report

8/87

8

yfir efstu ruslalgunum sem fylla jarhsiG, undir veggjum D1, D2, E1 og E2, envirist ekki vera undir sklaveggjum svum E og A. Sambrilegt lag ssteinnig inni sklanum. Ekki er ljst hvort

um sama lagi er a ra llum essumstum, en ljst er a hvort svo er, er eittmikilvgasta rlausnarefni Hofstaa-rannskna nstu rin. A sunnan liggur0004 a sklaveggnum en virist ekki faraundir hann ekki s a tiloka a svokomnu mli. Annahvort hefur lagimyndast eftir a sklaveggir voru reistir,ea a var fjarlgt sklastinu egarsklinn var reistur. ar sem afstaa skla og0004 er ljs, er fyrri tlkun (Adolf

Fririksson & Orri Vsteinsson 1997b) afstu skla og jarhss uppnmi. Virannskn 1995 var tali a ruslalagi semliggur ofan fyllingu jarhssins gengiundir suvesturhorn sklans. N er ljst aa sem tali var suvesturhorn sklansreyndist vera suurveggur tengibygginguD2 vi suvesturhorn sklans. Verur v nsta rannsknarfanga leitast vi a finnanjar vsbendingar um innri afstu sklaog jarhss. Hafa komi fram hugaverarvsbendingar sem vert er a skoa nnaregar rannsknarsvi verur stkka: milli LNS og gra ruslalagsins 0004 hafava komi ljs ltilshttar ummerki ummannaferir. milli er unnt lag afljsleitri fokmold me dreifum kola-flekkjum. suursvinu hefur komi ljs anna lag milli 0004 og LNS. a erljsleitur jarvegur me flekkjum af Heklu3 og er ljslega uppmokstur r

forsgulegum lgum. Hefur a myndastegar teki var fyrir jarhsi, skla og jafnvel rum byggingum. Verur avifangsefni nsta rannsknarfanga aathuga hvort finna megi samhengirkastslaga fr jarhsi gagnvartsklagrefti. Einnig er nausynlegt a skoadreifingu gra ruslalagsins nnar. a erfremur unnt sunnan vi sklann en erykkara bi austan og vestanvi. a ereinnig mjg ykkt og sumstaar tvfalt

undir D1, ynnist t milli D og E og veruraftur ykkara og sumstaar tvfalt svi

E. Bendir etta til a lagi gti hafadreifst fr fleiri en einum upprunasta.

Svi Dri 1998 var uppgrefti svi D haldi

fram en hann hfst 1996. Rtt er a rifjaupp aalatrii fyrri rannskna. Bruun grfofan ferhyrnda tft vi suvesturhornsklans en gaf nr enga lsingu henniutan a e.k. bygging hafi veri vi hlisklans essum sta, n dyra, og nsnilegra umerkja annarra en akolaflekkur var nlgt miju. ri 1996var fyllingin fr uppgreftinum 1908hreinsu upp og 1997 voru gerar nnarirannsknir hsinu, sem var fari akalla D1. Kom m.a. ljs a dyr hafa veri sunnarlega vesturvegg, sem hfu verifylltar upp. Einnig komu ljs dyr mijum austurvegg. milli skla og D1reyndust jafnframt vera byggingaleifar, D2.ri 1998 var rannskn D1 loki amestu og reyndist etta svi mun flknara tlkun en tlit var fyrir fyrstu, enda erum nokkur byggingarstig sama blettinuma ra. r minjar sem sust yfirbori

eru leifar af yngsta mannvirkinu svinu.Er a tft ea hs r torfi sem snr einsog sklinn, .e. N-S, 11.20 m langt og 6 mbreitt a utanmli, en 8 m langt og 3,20 mbreittt a innan. Veggir vestan, noran ogaustan eru hlanir r rauleitu strengjatorfi.Veggirnir eru annig gerir a innan ogutan er 0,4-0,5 m breitt lag af streng en milli er um 0,30 m moldarfylling.Suurgaflinn er ruvsi og virist gerurr hrun- og ruslalgum sem gtu veri

eldri en torfveggirnir. hsi D1 eru rjrdyr, tvr austurvegg og ein vestuveggsem san hafi veri hlai fyrir. Ljst era hsinu hefur veri breytt, jafnvelnokkrum sinnum. Er mgulegt a dyrnar vesturvegg su elstar. Hsi hefur staistakt og ekki gengt r v skla. Sanhefur veri fyllt r dyr og gerar arar austurvegg og gng (D2) yfir sklann. hefur austurveggurinn einnig veri rofinnnoran vi gngin og ar eru riju dyrnar,

annahvort til tgngu ea inn annaherbergi (D3) milli skla og D1. Inni D1

8/2/2019 Hst1998 Report

9/87

9

var lag sem virist hafa aki botn hssinsa mestu, en a hafi va veri fjarlgt rannskninni 1908. etta lag er ekki lktglflagi, heldur er a bleik-brnt a lit, amestu r plntuleifum. Lagi er n

greiningu, en a virist einfaldlega verahey og er sennilegt a lokaskeii hafi D1veri e.k. hlutft. Undir essu lagi voruglfleifar, sem virast eldra stig torfbyggingunni, en undir v vorustoarholur og leifar af timburstokk, semtilheyrt hafa timburhsi er sneri N-S og varum 2.7 m breitt a innanmli og a.m.k. 5.4m langt. etta hs sneri eins og torfhsi enhefur veri heldur mjrra. Stoarholurnarliggja me reglulegu bili mefram austur,

norur og vesturvegg, en ekki er ljst hvortr hafi horfi niurgrefti vi suurenda,ea hvort timburhsi hafi veri styttra ensari byggingar sama sta. Einnig ermgulegt a torfveggirnir hafi veribyggir utan um timburhsi, en ekki erfullt samrmi milli stoaraanna ogtorfveggjanna. Talsvert meira bil er t.a.m. milli hola og veggjar vestan megin enaustan. botni hssins eru rjr holur: s minnsta

mijum norurhluta, nnur str mijuhsi og s rija skammt suvestan hennar.Hinar fyrrnefndu voru undir heylaginu ogtilheyra v eldra byggingarstigi. Ekki erljst hvaa hlutverki r gegndu, en einneldsprunginn steinn fannst strri holunni.rija holan var full af kolum og sku og henni voru eldsprungnir steinar. Djparskorur voru vi brnir holunar, lkt ogsteinhellur hafi stai ar upp rnd.

Holan sker eina stoarholuna og gefur atil kynna a hn s eldsti fr saribyggingarstigum D1. suurenda kom ljs steinlg sttt sem

er lnu vi dyrnar austur ogvesturveggjum. Gti hn hafa veri lg erD1 var tengt vi skla um gngin D2, en er engin vissa fyrir v. Einnig er mgulegta hn s eldri og hafi veri framanvisuurgafl timburhssins.

Allar essar mannvistarleifar eru fr

tmabilinu 9. - 11. aldar. LNS sst stkusta svi D. Eru elstu

mannvistarleifarnar rtt ofan vi LNS. Einsog ur er nefnt hefur grtt ruslalag (0004,1004) sem er aeins yngra en LNS fundist llum rannsknarstum Hofstum oger a me elstu ummerkjum svinu.

etta lag er einnig snilegt nokkrumstum undir veggjum D1 og D2. a.m.k.tveimur stum svi D sstmannvistarlag, torfefni, undir 0004, en ekkier ljst hvort a er vsbending um a svi D hafi stai e.k. bygging ur engra ruslalagi myndaist ar. Samskonarummerki hafa komi fram svi E.Vi rannsknir jarlagaskipan nest

D1 voru tekin sni r nestu lgunum og erniurstana r eim athugunum a vnta

innan skamms. nsta rannsknarfangavera gngin D2 og svi D3 knnunnar.

Svi Eri 1998 var loki rannsknum

byggingaleifum svi E. ar hfukomi ljs veggjaleifar vinorvesturhorn skla (E1) og skammtvestan eirra voru leifar annars hss (E2).

Leifar E1 voru mjg skemmdar af skuri Esem Bruun grf langsum gegnum svi1908. Vi sari rannsknir ar hefur verigert r fyrir a E1 vri leifar af ferhyrndriog aflangri vibyggingu, en erfitt er afullyra a svo s. Aeins sust leifarlangveggjar ("norurveggur") sem nr frsklavegg, um 5 m vestur. Enginntorfhlainn vesturgafl var snilegur ogeinungis var a sj mja rnd af meintumsuurlangvegg essarar byggingar. essi

veggjarstfur ni ekki alveg avesturveggs skla, heldur var um 0.70 mbil ea dyraop ar milli. Talsvert rusl,einkum drabein, var dyraopinu og via. Hafi vesturgafl veri E1, er ekkitiloka a hann hafi veri r timbri, v kverkinni yst vi norurvegg voru rjrholur r sem gtu veri leifar eftirtrverk. milli eirra var mjg tt,dkkleitt og troi lag sem gtu verirskuldur, en ar sem langskururinn fr

1908 var grafinn vert gegn um ettasvi verur ekki fullyrt um etta. Bruun

8/2/2019 Hst1998 Report

10/87

10

taldi engar dyr hafa veri sklavegg svi E en a er rangt, v sj mtti a opvar ar sem E1 liggur a honum. Bruunhafi einungis fundi inngang hsiaustanmegin, sem snr beint upp

brekkuna. Er sennilegt a E1 hafi veriaalinngangur sklann einhverju stigimlsins. ar sem ljst er a einnig hefurveri gengt inn sklann svi D minnirsvipur hans meir sklann Skallakoti,en ar eru tveir inngangar annari langhli.rtt fyrir a essi hluti

rannsknarsvisins s mjg skemmdureftir rannsknina 1908 var ess freista areyna a ra hlutverk E1 me nnariathugunum nestu jarlgunum innan

ess. Tekin voru sni r eim, . m. bleik-brnu, lfrnu lagi sem minnir mjg meint heylag hsi D1.

Hs E2 er 3.40 m vestan vi E1 og 8.30 mvestan vi sklann. Jarlg milli E1 og E2gefa til kynna a E1 s yngra en E2. E2 varallt a 5.80 m langt N-S og 3.80 m A-V autanmli, en 4.40 m og 2.10 m ainnanmli. Veggir ess voru hlanir rstreng, fr 0.60 til 1.05 m breiir og er hveggjaleifanna n um 0.20 m.Vesturveggurinn skar sig r v a innan erhann hlainn me stru grjti. eim veggmijum eru mjar dyr me voldugumsteinum til beggja handa. Innan vi dyrnarer renna sem liggur mefram nr llumvesturveggnum inni hsinu. Innan vihana voru fjrar strar og djpar holur. Yfirhsinu l lag af hrundu torfefni. egar avar fjarlgt kom ljs grunnt, bleikt ogleirkennt lag blanda brnleitri mold og

stku torfusneplum. Minnir etta meintarheyleifar sem fundist hafa D1. Undiressu lagi var san grleita, vfrlaruslalagi 1004 sem er undir llu hsi E2,undir veggjum og utanvi . Er E2 vyngra en gra lagi. ar sem 1004 nraeins a sklaveggnum virist sem hanns eldri en a og ar me eldri en E2, enekki er hgt a tiloka essu stigimlsins a 1004 hafi veri fjarlgt vibyggingu sklaveggjarins.

rennunni vi vesturvegg E2 var gulbrnt,leirkennt lag me miklu af fiskbeinum og

lfrnum leifum. Voru essi lg veri tekin sni heild sinni til frekari rannskna.Eiga essar minjar sr ekkert fordmi slenskri fornleifafri og er a hulinrgta til hvers etta hs hefur veri

nota.

Svi GVi rannsknir 1908 og 1965 mtti sj a

gryfjan sunnan sklans var full af sku,beinum og rum rgangi. ri 1995 kom ljs a undir rgangslgunum voru leifaraf glfi jarhsi. Eru essi rgangslgtalin vera fr 10. ld. Jarhsi var ekkikanna frekar, en er nr rannsknarfangihfst 1996 var kvei a leggja herslu a grafa fyllinguna sem nkvmlegast tila greina drabeinin og arar leifar. rin1996-1997 var fyllingin grafin upp a hlutaog ri 1998 var marki sett a ljkauppgrefti fyllingunni. v markmii varn og hafa ll ruslalg veri fjarlg uppr jarhsinu og blasir ar vi torfhrunofan holunni sem vntanlega er hruniak- og veggjaefni yfir jarhsinu. Virannsknirnar 1998 komu ljs sem fyrr

miki magn af vel varveittumdrabeinum; kilingar, klfar, unglmb oggrsir, sem og bein r kindum, svnum,nautgripum og geitum. ar fundust einnigfuglabein, eggjaskurn, skeljar og bein rsilungi og orski ea su.Sni r rum rgangslgum, sku,

koluum viarleifum og eldsmerktirsteinar, voru tekin til frekari rannskna ogverur reynt a greina skuna til uppruna.Til samanburar var safna snishornum af

ekktu eldsneyti ngrenni Hofstaa, .e.torf, mr, sauata, kamykja, birki, lyngog hrs, en rekaviur og ang var stt strndina vi Skjlfanda. Verur essumefnum brennt og reynt a einangra ausreinkenni skunnar sem geta hjlpa tila bera kennsl samskonar sku fr 10.ld. ruslalgunum fundust sem fyrr nokkrirforngripir, . m.: ltil stika ea skaft rbeini me krossi endanum, mannbroddur,

jrnnaglar, brni og perlur og perlubrot ogjrngjall.

8/2/2019 Hst1998 Report

11/87

11

nsta rannsknarfanga verur rist uppgrft jarhsinu sjlfu.

NiurlagN er loki fyrsta fanga rannskna

fornblinu a Hofstum Mvatnssveit.Er fyrri mynd af minjastanum ngerbreytt. ur var tali a ar hafi staistr, aflangur skli, me suurgafli rtimbri, ruslahola ea sohola sunnan hans,en norarlega vi vesturhli moldarbingurog tftarbrot gn sunnar. ljs hefur komia sklinn er ekki eina hsi stanum,heldur er ar raun yrping hsa af msumgerum fr tmabilinu 880-1050. Elstahsi er lklega jarhsi sunnan skla semur var tali vera sohola, en tla m ahafi veri fyrsti bstaur frumblinga.Fljtlegt hefur veri a reisa hbli af essutagi til a veita mnnum og jafnvelskepnum skjl ar til rrm gafst tilfrekari hsbygginga. Vi suvesturhornsklans var torfhs sem skagar upp avera af smu str og hinir minnstueldasklar sem fundist hafa slandi. arur virist hafa stai eldra hs r timbri.

Ekki er ljst hvort a hs er eldra ensklinn. milli essara hsa og sklansvoru einhverju stigi mlsins bygg gngog jafnvel btt vi enn einu herbergi millisklans og essa afhss. Vinorvesturhorn kom ljs a a sem urvar tali moldarbingur reyndist vera leifaraf aflangri byggingu sem fst var visklavegg. Er sennilegt a hn hafi veriinngangur inn sklann, e.k. forskli.Skammt vestan hans var lti, nnast

ferhyrnt hs me undarlega strum steinum vesturvegg og rennu og fjrum strumholum. Mannvirki af eirri ger hefur ekkifundist ur slandi. ur var tali aekki hafi veri suurgafl r torfi sklanum, en a hefur n veri hraki. Visuausturhorn fundust leifar enn fleiri hsa:eitt aflangt fast vi sklavegg og annastaksttt skammt austar, en milli eirra erdld sem gti bent til a ar s einnig

jarhs.

Ekki hefur enn veri skori r umbygggingasgu Hofstaa a llu leyti, en

ljst er a jarhsi tilheyrir elsta skeii,nstelstar eru lklega eldri byggingar visuvesturhorn sklans, stri sklinn ensar yngri byggingar kring um hann bia austan og vestan.

Frumathuganir beinaleifum benda til aHofstaabar hafi va afla matfanga, eirhafi ntt sr til fullnustu veiiskap Mvatni en einnig lagt herslu skepnuhald me svipuu snii og ekkistfr seinni ldum. Frlegt verur a rna skuleifar og komast a v hvaaeldiviur var notaur. Einnig er spennandia ba eftir niurstum jarvegs-rannskna glfleifum r Hofstaahsum,en r gtu varpa nju ljsi hlutverk

bygginganna.Fyrirhuga er hefja njan, riggja rafanga rannskna 1999 v augnamii aafla enn frekari vitneskju um framvindulandnms slandi. eim fanga verurlg hersla a n fram heildarmynd afforsgulegum byggingaleifum Hofstum. Jarhsi sunnan sklansverur grafi fram, byggingar viausturhorn hans, afhsi vi norurendasklans og allur sklinn. Vegna einstakravarveisluskilyra Hofstum hefur veriunnt a greina sundur mismunandi jarlg llu uppgraftarsvinu, sem auveldarmjg hi vandasama verk a rekja upprunaog run bygginganna. Samhlia uppgreftimunu rannsknir rgangi, sku ogglfleifum halda fram. A auki erfornleifaskrning Mvatnssveit langt veg komin. Loki er svisskrningu llum Sktustaahreppi (1250 minjastair)

og hafa rmlega 750 stair veri skrir vettvangi.Landnmsrannsknir mvetnskum

fornleifum eru umsvifamestu fornleifa-rannsknir slandi til essa. ljsi eirram lesa brot r forsgu slands, jafnt um bnaarhtti sem ekktust fram eftirldum, sem er liu undir lok fyrirupphaf ritaldar.Hofstaaminjar sna a ekki svo lngueftir a frumblingurinn ar ruddi sr lnd

hafa afkomendur hans geta ntt srfjlbreytta landkosti Laxrdal til bsetu

8/2/2019 Hst1998 Report

12/87

12

og viurvris og skapa sr mikla ausldsem egar sr merki 10. ld. Er agrandi verkefni nsta fanga a leitavsbendinga um hvaa forsendur lgu abaki essum fyrsta sigri slensks

landbnaar hrjstugri fsturjr.

8/2/2019 Hst1998 Report

13/87

13

Gavin M Lucas2.0Area A Excavation Report

Introduction

1998 saw the opening of a new area ofinvestigation at Hofstair, a large trench c.17m long and 9.4m wide at the west,stepping in twice to a width of 3.4m at theeastern side. It was laid over the southernend of the Skli which linked up with AreaD to the west but stopped short of Area Gleaving a standing baulk c. 2m wide. Theaim of the investigation was primarily toestablish the nature of the southern wall of

the Skli as Bruun had not recognized anystructural feature substantial enough tostand as a back wall:

Sondre gavl af rummet havde, som allerede be-rrt, ikke nogen egenlig jord-eller grstrvsvg,men kun en 20 cm. hj, foroven flad,fodskammel som de, der ofte ses ved foden afbrddavgge.

[As already mentioned, the southern gable of theroom did not have a proper earthen or turf wall,

but only a 20cm high flat-topped footing similarto those often seen at the base of walls made ofwooden panels.]

- Bruun 1909, 270 - trans. M. Snsdttir

A secondary aim of the trench plan was toidentify any further activity or structuresaround this part of the site; for instance, it isknown that Bruun on an earlier visit to thesite noted a feature to the southeast of theSkli while in 1900 Brynjlfur Jnssonidentified a field dyke/enclosure wallrunning parallel and east of the Skli(Bruun 1897, 174-5; Jnsson 1901, 12-13& fig.V). Bruun in his 1908 excavation didnot investigate any area to the southeast.Beyond what previous investigators ob-served however, it was considered im-portant to open up as large an area aspossible precisely because a thoroughunderstanding of the spatial layout of a sitecannot be gained through trenching

strategies focused solely on surface visiblestructures.

The excavation strategy followed that usedpreviously on other areas, removal of allsoil including turf and topsoil by hand andemploying single context recording.

Selected contexts considered suitable weresampled or sieved. The report given herepresents the results of these investigationsin phased sequence starting with the mostrecent remains and ending with the oldestfeatures so far uncovered. Table 2.1 gives afull list of contexts uncovered while Figure2.2 shows the stratigraphic matrix for thesite; only in a part of the area was thebeginning of the occupation sequencereached while subsequent fieldwork will be

needed to complete this narrative. Thephase groups are preliminary only andsolely applicable to Area A.

Excavation Results

Four main phases of activity have beenidentified within Area A (Fig. 2.1) andthese are discussed in reverse order, i.e. the

latest first:Phase IV (20th century)The latest activity identifiable in the areacomprised previous archaeologicalinvestigations: running through thesouthwest corner of the open area was the1995 FS excavation trench([0102]/[0103]) while coming out of thenorthern limit of excavation, DanielBruuns 1908 trench through the Skli

([0002]/[0104]). Both merit a few words interms of their relationship to the widerarchaeology uncovered. The 1995 trenchwas originally cut to establish stratigraphiclinks between the sunken-floored buildingin Area G and the Skli (A/B), but it wassuspected after opening Area D and nowcorroborated, that this trench by-passes theSkli wall and merely clips the collapse.This has ramifications for the interpretationof the relationship between the Skli A/B

and Structure G which will be discussedbelow.

8/2/2019 Hst1998 Report

14/87

8/2/2019 Hst1998 Report

15/87

15

8/2/2019 Hst1998 Report

16/87

16

occupation phases and is characterised bytwo sub-phases of windblown soilaccumulation across the area. For most ofthis period, there is a fairly rapid and sterile

aeolian deposition of fine pale brown silts(loess), which include two identified tephralayers (V1717 and a V1477) andprobably others not seen which are muchfiner and less continuous (such as V1410,H1300, K1262 and possibly H1104/1158;see Sigurgeirsson 1998). This layer [0016]occurs continuously across the site; beneathit lay another, less extensive layer [0109]which covered the Skli and the eastern partof the area but feathered out to the west and

south. This had a similar loess matrix butincorporated some charcoal and decayedorganic elements giving it a slightly sandiertexture and darker colour. Its boundary with[0016] was not sharp and what it representsis a period after the abandonment of the sitein Phase I when there was still culturalactivity in close proximity but whosecessation is marked by the gradualtransition to the sterile loess above.

Phase I(9th-11th century)The Hekla 1104/1158 tephra was neverunambigously identified in Area A thisseason but on the basis of its positionobserved elsewhere on the site, any depositssealed beneath [0016] in this area haveprovisionally been dated to the pre-12thcentury. Excavation of previouslyundisturbed contexts from this earliestphase of the site did not progress very far

this season, nevertheless a substantialpicture of what is present has been gainedwith good expectations for a subsequentseason. Broadly two main sub-phases ofstructural activity have been identified, thelatest including two small structures (A2and A3) and the earliest belonging to theconstruction of the Skli (A/B). A possiblefourth structure is also being consideredalthough the only indication of its presenceso far remains a depression in the overlying

deposits, a depression which was first notedat the 1477 tephra level where it was very

marked. The depression may have beencaused merely by the sloping depositscoming from Structures A2 and A3 butanother possibility could be a sunken-

floored building and although confirmationmust await further excavation, this muchwe can say, that if there is a structurebeneath these deposits, it belongs to anearly phase of the site, contemporary withor even earlier perhaps than the Skli andthe sunken-floored building in Area G.

Of the structures we have so far exposed,Structures A2 and A3 are the latest,although their relationship to each other

remains to be resolved. They are situated tothe east of the Skli, Structure A2appearing to be connected to/abutting itwhile A3 lies around 3m to the southeast.All over the eastern side beyond the Skliare various deposits of turf collapse, whichlower down inter-digitate with other upcastand midden-type ashy deposits. Of theupper layers, three discrete contexts([0118], [0119], [0120]) appear torepresent different episodes of wall decayand collapse from Structure A2 which isdefined at present by a single east-westlength of turf strengur wall [0125] comingout of the Skli for 4.8m before returningbeyond the northern limit of excavation.On the surface there is some indication ofits extent northward suggesting a structurecomparable to D1, i.e. about 10m long and5m wide. Its comparability to D1 is furthermade significant by the fact that it has been

untouched by Bruun although otheractivities may have done some damage.This is the case with Structure A3 whichalthough had a little more of its planuncovered, suffered from the later post-Medieval insertion of a silo pit through itsinterior.

Two sides of Structure A3 suggest abuilding 2.2m wide and c.5m long,although only half of its length occurs

within the trench, the limit of excavationrunning through an apparent blocked

8/2/2019 Hst1998 Report

17/87

17

doorway opening to the west. The wall[0126] survives to a width of 0.6m alongthe northern side and is constructed solelyfrom turf strengur. In size it is slightlysmaller but comparable to Structure E2,

especially in the position of the doorwaybut given that all internal features appear tohave been cut away by the silo, there is noway of knowing how comparable it mayhave been in terms of function. Initialobservations also suggest this may be arebuild of an earlier structure but this mustawait confirmation. Both Structures A2 andA3 were constructed over upcast dumps([0124], [0111]) and other turf collapse([0114]) which spreads over the eastern part

of the area. This was the level at whichexcavation ceased and these depositsdisplay fairly complex inter-digitationsuggestive of simultaneous deposition,whether gradually or rapidly. Whatever thesequence, one thing does seem probable,and that is that a very early charcoal-richlayer [0106] continues beneath thesedeposits but stops at the Skli wallindicating that both Structures A2 and A3post-date the Skli by some length of time.

This charcoal-rich layer [0106] is partlyequivalent to [0004] in Area G and has beena key marker horizon for much of theinterpretation on the site since 1995. Itsstatus is undeniably significant but somecare must be taken with its identificationacross the site as a single deposit. The layerdesignated [0106] in Area A is a fairlycharcoal-rich, dark grey brown sandy silt

which is very thin (

8/2/2019 Hst1998 Report

18/87

18

would have been 0.8m, but it is uncertain asyet whether the wall extended over to theinner line of stones or whether thesedemarcate a benched area inside. However,on the exterior of the herringbone wall we

found evidence of a strengur course ofsimilar width which had been ramped upagainst it and this may have given thenecessary support and thickness to thestructure.

The relationship of this wall to otherdeposits in the area was not fullyinvestigated this season, although as wehave mentioned, the charcoal layer [0106](partially equivalent to [0004]/C4) did not

continue beneath it. We did howeveruncover additional activity beneath thislayer in the form of a thin aeolian deposit[0123], relatively sterile but with veryoccasional charcoal flecks/fragmentssuggestive of human activity. This is theloess typically found directly over theLandnam tephra, although it is notcompletely continuous as it tended tofeather out by the edges of the Skli walland became thicker southward. Sealedbeneath this but still over the Landnamtephra was a discontinuous deposit ofupcast [0122] with heavy Hekla 3 tephrainclusions which had been dumped intohollows to the southwest of the Skli in theuneven ground immediately after thedeposition on the Landnam tephra. Thisappears to represent a part of the veryearliest activity on the site and is probablythe result of construction upcast from either

the Skli or the sunken-floored building inArea G. It continues toward Area G and itsrelationship to the wall collapse of G willbe critical; it also continues alongside thesouthwestern side of the Skli and itsrelationship to contexts here also needfurther investigation.

Concluding Discussion

The aims of the 1998 season have been

more than adequately answered - we foundthe missing southern wall and three,

possibly four new structures besides.However, a whole new set of questionshave been opened up by this investigation,not least that the original intepretation ofthe building sequence proposed in 1995 has

to be revised. We need to ask again what isthe relationship of the Skli to the sunken-floored building in Area G and what is therelation of these to the new structuresrevealed? How coherent is layer [0004] asa single context, especially as we movebetween Areas G and D? In concluding, Iwould like to address two main themes -one, looking back to Bruun and to try andunderstand why he excavated andinterpreted the southern end as he did, the

other looking forward to objectives for afollow-up season in terms of the questionsraised here which need answering.

Bruuns claim for the insubstantial natureof the southern wall is primarily a result ofhim having removed most of it beforeactually seeing it; even so, its survivingdimensions are comparable to thoseobserved elsewhere in Areas D and E in thecurrent investigations which makes it hardto square his observation of this part of thestructure compared to others. Hisinterpretation in this context may be more aresult of his prejudices prior to excavationwhere from the surface it appears as if thesouthern end has no wall compared to theeastern and western walls, an observationalso made by Brynjlfur Jnsson in 1900.The irony is, the more substantial longwalls of the Skli as they appear on the

surface are in fact not really the walls at allbut cumulative collapse and build up on theoutside from other structures. The highestpart of the eastern wall as seen on thesurface for example sits over Structure A2- the actual Skli east wall lies muchfurther west. The fact that Bruun couldrecognize a wall when he saw it makes ithard to understand why he should have cutso much into them; at least twopossibilities raise themselves, either or both

of which may be partly true. In the firstinstance, by starting from the presumed top

8/2/2019 Hst1998 Report

19/87

19

and working down as he appears to havedone, he started too far out and inconstantsupervision of the workmen would haveeasily resulted in removal of the walls. It isalso possible however that the walls had

severely decayed and collapsed and theywere unrecognizable until cut back someway.

Whatever Bruuns perception of the site, weare also re-interpreting our own excavationsas the example of the 1995 trench shows.This will, and should, be an ongoingprocess and says something about thecomplexity of Hofstair and the richness ofits archaeology. We have questioned the

reliability of the C4 marker horizon as asite-wide tool, but regardless of that, it doesnot appear to be useful anymore inassessing the relationship between the SkliA/B and Structure G since both pre-date it.However, we have found another potentialdeposit to link these two structures, theupcast [0122], and one aim for the nextseason when excavating the collapse fromStructure G would be to link up Areas Aand G to follow this deposit. The newstructures to the east of the Skli revealedthis year also clearly need furtherinvestigation and this will involveexpanding Area A to the south and north aswell as a little to the east to uncover theirfull plans. Beyond this, work wouldcontinue on excavating all deposits down tothe beginning of the sequence within theopen area in the expectation of completingthe picture already emerging from this

season.

AcknowledgementsMany thanks to the core team in Area A whoworked so hard and accomplished so much in threeweeks: Sarah Beveridge, Andrew Leykem, Ralph

OConnor, Bengta Ryste and Jonas Wesley; and toothers who helped out at various times: Lisa Alter,Colin Amundsen, Hildur Gestsdttir, Abby Hunt,Linda Livolsi, Sandra Meadows, Meredith Vastaand Orri Vsteinsson.

8/2/2019 Hst1998 Report

20/87

20

Context Description

0100 Stone footing of Structure A10101 Redeposited blocks of turf (not structural), cut from elsewhere and dumped in a hollow in

the northeastern corner of the area

0102 1995 excavation trench backfill0103 1995 excavation trench ("R")0104 Bruun's 1908 excavation trench0105 Turf wall which has been truncated, (Structure A1)0106 Very thin layer, thickens to east and west; either deliberate or naturally formed, heavy

incorporation (e.g. midden outwash) of cultural material into a silt accumulation. Part ofinfamous "C4" layer.

0107 Abuts wall [0105] - part of robbed structure [0100]? Includes some finds from silo backfill[0115].

0108 silo pit cut with beam/stone slots at the base for internal partition/containment.0109 Mixture of windblown loess and fine organics - long-term accumulation of natural aeolian

soil and either weathered midden material or deliberate incorporation of midden

(fertilizer). Possibly less rapid deposition to the cleaner loess [0016].0110 Foundation cut for wall [0100]0111 Mixed upcast dump - limits uncertain as it appears to lens under and over turf debris

[0114]. Incompletely excavated in 1998.0112 Turf collapse/decay from southern end of the skli (Structure A/B) wall - eastern extent

uncertain and unexcavated in 1998.0113 Arbitrary cleaning unit over the northeast part of the area - probably the interface between

[0109] and various turf debris below0114 Turf debris/collapse - limits not fully defined in 1998 or excavated. Interdigitates with

[0111] and [0124]0115 Backfill of silo pit0116 Main wall of skli, southern end (Structure A/B). Severely truncated by Bruun along all

edges leaving only a thin spine of the original herringbone wall upstanding.

0117 Basal fill of pit, not continous but filling voids and slots in the base of the silo pit [0108].0118 collapsed turf, probably from wall [0125]0119 collapsed turf - very distinct iron-rich oxidised turf fragments, unlike any of the walls

visible in 1998 but its position and orientation suggests collapse from a structure to thenorth, i.e. [0125]

0120 mixed turf collapse and midden-type material, probably representing initial decay ofstructure/wall [0125]

0121 Very mixed material, settling in top of hollow over Structure A1, sealing all other layersassociated with it.

0122 Upcast filling hollows/natural depressions over the Landnam layer in the southeast part ofthe area. Hollow appeared very regular and linear when exposed in plan but uponexcavation proved to have avery irregular and undulating base and sides.

0123 Loess - windblown silt lying over the LNL which also levels out much of the uneveness of

the previous surface. Inlcudes some cultural material in the form of charcoal suggestive ofhuman activity in the vicinity

0124 Mixed upcast and midden-type material0125 Turf wall of Structure A20126 Turf wall of Structure A3

Table 2.1. List of contexts from Area A

8/2/2019 Hst1998 Report

21/87

21

Figure 2.2 Area A Matrix

8/2/2019 Hst1998 Report

22/87

22

8/2/2019 Hst1998 Report

23/87

23

Ragnar Edvardsson

3. 0 Area D Excavation Report

Fornleifastofnun slands, Brugtu 3, 101Reykjavk, [email protected]

IntroductionIn 1998 excavations continued in the areadesignated D to the southwest of the mainhall (A/B) at Hofstair. Area D wasoriginally investigated by Daniel Bruun in1908 but he did not come to a firmconclusion regarding the extent or functionof the remains (Bruun & Jnsson 1909).

A new campaign of excavations in area Dbegan in 1996. In this first season the mainobjective was to clear Bruuns trenches toenable a better understanding of his work in1908. On the basis of findings in 1996excavations were continued in 1997 withthe aim to investigate structures called D1and D2. D1 is an independent structurewest of the skli A/B and D2 is the area andstructural remains connecting the skli andD1. A new code, D3, will be introduced for

the area between A/B and D1, north of D2.In 1997 the emphasis was on excavatingfill, floor-layers and doorways within D1which had not been previously disturbed byBruun and this work was continued in 1998focusing investigations in equal measure onthe area immediately outside D1. Thetrench was extended to the north to exposeall structural remains associated with D1,and to the east to expose a part of thewestern wall of A/B. This was done to

facilitate better understanding ofrelationships between D1 and A/B andother structural remains in the area.

The same basic methodology was appliedas before, single context planning continuedfrom the plans and context records madeduring the 1997 excavation. Table 3.1 givesa full list of contexts uncovered whileFigure 3.4 shows the stratigraphic matrixfor the site. Samples were taken fromselected contexts for a variety of analysiswhich will contribute to the understanding

of the function and development of theexcavated structures. The 1998 excavationsin D added considerably to findings madein earlier seasons and revealed previouslyuntouched archaeological deposits. This

report presents the results of the 1998excavations in area D. Four main phaseshave been identified in area D and they willbe described in sequence, beginning withthe most recent remains.

Excavation Results

Phase III (20th century)Phase IV has been described in reports ofthe investigations in 1996 and 1997 (AdolfFririksson & Orri Vsteinsson 1996,1997a) and very limited new excavation ofthese deposits was undertaken in area D in1998. Therefore only a brief description ofthis phase is given here for betterunderstanding of the 1998 work and of theoverall relationship between deposits in thearea. This phase consists of 20th centuryactivity and to a large extent the remains ofDaniel Bruuns excavation in 1908.

A layer of topsoil (001) has accumulatedsince the excavation of 1908 with traces ofmidden material deposited on the field asfertiliser within living memory. This layeris nowhere more than 10 cm thick.After excavating the skli (A/B) in 1908Bruun turned his attention to the area on thesouthwestern side of the skli (A/B). Bruunexpected to find a room or an adjoininghouse but was not satisfied with the resultsand only gives the briefest of mentions of

this part of the building complex in hisreports. On his published drawing of thesite he shows a rectangular trench withlimited archaeological remains, but on hisfield sketch he drew a line indicating thearea of D1 which he excavated.The extent of area D trench in 1998included Bruuns trenching (059, 060) ofthe skli (A/B) on the eastern side of thewestern wall of that structure. In theexposed area it is apparent that Bruun

aimed to put his trenches on the top of theturf-walls (015) and then followed the turf

8/2/2019 Hst1998 Report

24/87

24

on the inside, truncating the walls to anunknown degree. In an effort to locate adoorway on the western side of the skli(A/B) Bruun cut a trench through the skliwall and into two parallel walls forming a

passage connecting the skli and D1. Heseems to have managed to follow thesouthern wall (537) of the passage withoutcutting into it much but he does not seem tohave understood the northern wall (536)which is truncated by Bruuns irregulartrenching and badly damaged.Bruun excavated a rectangular area insidestructure D1, cutting into the western wall(24) but leaving a strip of turf-debris on theinside of the southern wall-like deposits

(see below) as well as two rectangularchunks of turf-debris in the northern end ofthe structure. The edges of his trench werecut right down into the floor layers severingthe relationship between the walls and thefloor but leaving a part of the floorlayers inthe middle. Bruun did not remove fills inthe two doorways in the southern part of thestructure but excavated a doorway on themiddle of the eastern wall of D1. On theeastern side of this wall Bruun encounteredthick turf-deposits which he was not able tounderstand, his confusion apparent from theirregular trenching (D3).

Phase II (c. 1100-1908)Preceeding Bruuns excavation was a longperiod of inactivity in area D. Layer 016represents aeolian and humic accumulationwith bands of tephra from the abandonmentof the structures prior to 1104/1158 to their

excavation in 1908. This context is clearlydivided in two by the a tephra from AD1477 which provided a useful markerhorizon.In the lowest part of layer 016 there is athin and often indistinct band of tephrafrom either or both Hekla eruptions in 1104and 1158. This tephra occurs just above thecultural layers and was frequently observedin plan just before the turf-walls of D1 andA/B appeared indicating that hardly any

soil accumulation had occurred between thecollapse of the walls and the deposition of

the tephra. Outside the western wall of D1a layer was identified underneath 016. Thislayer (070) was similar to 016 in texture butwas lighter brown and included charcoalbits and patches of turf debris. This layer is

interpreted as accumulation of earth againstthe walls of D1 after the structure had beenabandoned. Traces of layer 70 could beidentified on top of the western wall. Tracesof a similar layer (061) were found in D3,abutting the eastern wall of D1. Layer 70and 61 could not be connected.Structure D1 was abandoned before1104/1158. Thick layers of turf collapsewere identified, both inside D1 and outsideit, particularly in the area between D1 and

A/B, now called D3. In 1997 twoundisturbed contexts (021,043) representingturf collapse within D1 were recorded in thesouthern and northern end of the structure.Similar deposits were revealed in 1998 inD3. This turf collapse was right under layer016. The 1104/1158 tephra layer could beidentified right above the turf collapse. InD3 layers representing collpase of thewestern wall of the skli and the easternwall of D1 were identified (062, 063, 064).Layer 062 and 063 are turf debris, mixtureof greenish and reddish turf. The reddishturf is clearer along the eastern wall of D1but the greenish alongside the western wallof A/B. Layer 064 is very similar to 062and 063 but it was harder and morecompact. There is a possibility that 064 is awall. Therefore it was recorded as adifferent context. Hardly any traces of turfcollapse were identified on the outside of

the northern and western sides of D1.The difference between the collapse fromthe walls of the skli (A/B) and D1 wasvery clear. The skli wall was built withgreen-grey klmbruhnaus turf (015) but D1with red-orange strengur turf (024). Fromthe layers in D3 a number of animal boneswere recovered, many of which werefragments of sheep and cattle skulls (seeMcGovern, this volume).

8/2/2019 Hst1998 Report

25/87

25

Phase I (9th 11th century)Phase Ib: The structural remains mostclearly visible in area D, those which wereabandoned at the same time as the skliA/B, were all in use at the same time and

are all ascribed to the same phase althoughsome development of construction andlayout can be discerned. Furtherinvestigations are needed to clarify theorder in which the apparent structuralchanges were made but a tentativehypothesis can be put forward, dividing thephase into three sub-phases. First is theconstruction of a rectangular structure (D1)with a doorway on the southern side of itswestern wall, second is the filling of this

doorway and the construction of aconnecting passage between D1 and theskli, and third is the partial dereliction ofD1 and its usage for the storage of hay.

The turf walls of D1 are aligned north andsouth as the skli measuring 8 x 3,20m onthe inside but 11,20 x 6 m on the outside. On the west, east and north side arestanding turf walls (024), measuring 120cm wide and 80-100cm high. On thesouthern side a turf construction is lackingbut in its stead there is a pile of mixeddeposits, mainly turf depris but also ash andmidden deposits, rising to a similar heightas the other walls and which may at leasthave served a protective function (seebelow). The turf walls are made of red-orange strengur turf. The turf was laiddown horizontally in two stacks on theinner and outer face of the wall and the

space between was filled with loose earth(541). Layer 541 is 30 cm wide and 10 cmdeep were it could be seen. The length ofeach strengur could not be identified as thestrengur in the walls had fused together.The walls in D1 are different from the Skliwall exposed along the east part of thetrench in 1998, as well as in areas E and A.The Skli wall is constructed withklmbruhnaus made of green-grey turf.The house had an entrance on the southwest

side, excavated in 1997. A small wall (538)extends to the west on the northern side of

the doorway, abutting the western wall ofD1 and measuring 120 x 100 cm. A testtrench was dug at this position in 1996cutting this wall in half. This wall was builtwith similar turf as the other walls in D1

and seems to have been built as a weathershield at the entrance to the house.

On the inside of the turf walls the floor ofthe building had been cut into the subsoil(003). The depth of the cut could not benoted in most places as the 1908 excavationhad disturbed the layers inside D1.However in the northern part of D1 thebuilding had been cut from 15 20 cm intothe subsoil. The 1997 excavation had

already shown that undisturbed floorlayersinside D1 in its northern end were higheralong the east and west walls than along themiddle axis of the houses. This was taken toindicate that there might have been raisedbenches along the walls. Correspondingfeatures could not be found elsewhereinside D1 because of truncation by the 1908excavation.

Along the west and east walls of D1postholes were identified, eight on eachside (contexts 081, 088, 091, 097, 094,099, 501, 503, 505, 507, 533, 525, 526,527, 528, 529, 539, 531). Three more werefound, two in the north end and one in thesouth. The depth of the postholes variedfrom 20 40 cm. The smallest postholemesured 18 x 18 cm and the largest 25 x 24cm. All these postholes were cut into thenatural, and some had a stone which may

have served as packing. The holes wereregularly spaced, with an average intervalof 25 to 60 cm. The postholes on both sidesin the southern two thirds of D1 had beendamaged by Bruuns trench removinglayers to the extent that it was not possibleto see from which surface the holes hadoriginally been cut. The postholes in thenorthern third of the house, on the otherhand, had not been damaged by earlierexcavations. These postholes did not

become visible until layer 051 had beenremoved and some of the postholes cut

8/2/2019 Hst1998 Report

26/87

26

8/2/2019 Hst1998 Report

27/87

27

layer 052. Layers 051 and 052 had beenrecorded in 1997 and interpreded as haylayers. The fill in all the holes was similar,lightbrown earth with inclusions of H3

tephra (contexts 082, 089, 098, 093, 096,500, 502, 504, 506, 508, 534, 510, 511,513, 514, 516, 540, 530). Some of thepostholes were associated with a smallerhole (090,092,095,512,515), located 4 - 10cm in front of the main hole. These holesmeasured about 6 x 6 cm. The function ofthese smaller holes is not clear but they areclearly connected with the postholes insome way.

Inside D1 three fairly large depressions hadbeen cut into the subsoil. The smaller andmore irregular depression (075) was in thenorthern part of the structure, in the centralaisle and this was sealed by the pinkishstraw layer. This hole mesured 60 x 60 cmand about 10 - 15 cm deep. The largerdepression (542) is in the middle of D1,measures 160 x 160 cm and is 10 15cmdeep. Both these depressions were sealedby a pinkish straw layer, no other infillcould be identified. The larger one can wellhave been a fireplace when the structurewas in use. Both of these holes hadprobably been cleaned out before thestructure was used as a hay storage. Onefireburnt stone was identified in the largerone which might indicate its usage. Thethird depression (085) was identifiedtowards the western wall and seems to havetruncated one of the postholes. It was

recorded in 1908 as a patch of charcoal andwas revealed again 1996 but not excavateduntil 1998. Inside this depression there areoblong deep grooves along three sides (S,W and N), suggesting that vertical slabshave been removed from the sides. Betweenthe grooves the bottom of the depression isuneven. It was filled with peat or turf-ash,mixed with charcoal and soil and included anumber of fire-cracked rocks. There washowever no scorching of the soil beneath

the fill, indicating that fire had not burnedon the ground in this location. This

depression truncates one of the post holesand is therefore more recent than the timberstructure, suggesting that the postholesrepresent an earlier phase of building in this

location than the turf walls and floor layers.The function of the depressions is unclearbut the larger one may be the remains ofsome sort of raised fire place where the fireburnt on a horizontal slab set upon threevertical ones.

A pavement (535) was discovered in thesouth end of D1 aligned with theentranceways on the east and west walls.The stones in this pavement are lava rocks,

measuring from 60x60 cm to 20x20 cm. Itis possible that Bruuns excavationremoved stones nearer the westernentrance. It was noted in 1997 that thewestern entrance had been sealed onpurpose. It can be suggested that this wasdone at the same time as the passage D2was constructed, but no stratigraphicrelationship is available to prove this.

The working theory in 1997 was thatBruun had removed all floor layers in D1except for the northern third of the house.There the remains of a floor layer (052)were discovered in 1997. The 1998excavation showed that Bruun had cut thefloor layers away alongside the walls butleft them intact along the middle. The 1908excavation had also cut floor layers away inthe narrower trench which had been cut intothe northern end of the house in its middle,

seperating floor layers on each side of thehouse. Right along the middle of thestructure dark brown compact layers084/086 were discovered. Where these darkbrown layers could be traced, they wentstraight up to the south wall and did not gounder them. They could not be identifiedalong the east and west walls becauseBruuns excavations had removed them.Layers 084 and 086 are similar in texture tolayer 052 which was excavated in 1997 and

was interpreted as a floor layer. Layers 084and 086 had been cut in the northern part of

8/2/2019 Hst1998 Report

28/87

28

D1 by Bruuns excavation and could not belinked with layers 052 but their apparanceand texture suggest that they all belong tothe same layer. These layers are organic anddark reddish brown. They are up to 9 cm

thick. Thickest near the stone pavement(535 ) in the south end, thinning towardsthe north, east and the west until theydisappear. Their extent is about 100cm x120cm. In the northern part of D1 along theeastern wall a dark brown layer (509) whichis probably as same as layer 052 wasrecorded. Embedded in this layer weretimber remains which were set in a verticalslot between two of the postholes. Avertical beam in this place suggests that the

house was a timber structure, with woodpanelling filling the gaps between the posts.A comparable beam slot was however notfound on the western side where layers hadbeen left intact by previous excavators.Depending on the nature of layer 509 thismay indicate that the panel-house wasdismantled at some stage and replaced by amore conventional turf-house construction.

The construction of a passageway (D2)between D1 and the skli (A/B) seems torepresent a separate and later stage in thedevelopment of this building. The passageis aligned east west with turf walls on thenorth and south sides. The northern wall(536) is badly damaged by the 1908excavation but the southern one (537) isnearly intact. Both passage walls are builtin the same way as D1, with strengur turf.The walls in the passage are built of reddish

turf while the skli walls are of greenishklmbruhnaus turf.

The southern wall of the passage abuts thelayers representing the southern gable of D1suggesting that it is later than both skli andD1, although this has yet to be confirmed.The skli wall seems to have been cut andthe remains of a repair, which may beassociated with this action, are still visibleon the northern face of the passage where

the skli wall has been truncated. The repairis a 20 cm thick slice of reddish-orange

strengur turf like that in D1, stacked againsthe grey-greenish klmbruhnaus turf in theskli wall proper. Immediately west of thisthe northern wall of the passage has beentruncated by the 1908 excavation, and

another truncation is at the juncture of thepassage wall and the eastern wall of D1,making the analysis of the relationshipbetween the walls difficult. This will bepossible with the removal of layers on theoutside of the walls (in D3) at a future date.The relationship between the southernpassage wall and the skli wall is not yetclear either. Bruuns trench also cuts awayall the layers inside the passage straightdown to the underlying sheet midden 004.

The 1908 trench stops short of the westernwall in D1. There Bruun obviously thoughtthat he had found the western wall of thestructure in D. The 1997 excavationdemonstrated that where Bruun thought thathe had found a wall there was actually aninfilled doorway, 1,40 x 120 cm. The fill(066) in this doorway was of the same sortas the turf debris 021 and 043 which hadfilled D1 suggesting that it was filled in atthe same time as the rest of the building.Underneath the fill was a pinkish strawlayer (077), similar to layer 051. Thisdoorway seems to have been in use after thewestern entrance had been blockedalthough it cannot be precluded that theywere both in use at the same time before thewestern one was filled in. The thirddoorway in D1 is on the middle of easternwall and leads to the area (D3) between theskli and D1. This doorway was excavated

in 1908, removing all evidence ofstratigraphic relationships. It is quitepossible that the doorway leads to anotherroom in area D3 but D3 has not been fullyexcavated at this point.

The last stage of occupation in area D waswhen the structure was used as a storageprobably for hay. In 1997 a pinkish strawlayer (051) was found in the northern partof D1. This layer was sealed by turf debris

layer 043 and was on top of a morefloorlike layer, 052. 051 was spread all over

8/2/2019 Hst1998 Report

29/87

29

the northern part of the structure and couldnot be found anywhere outside it. It wassampled and the examination showed ahigh proportion of decomposed grass(Simpson 1997).

During the excavation in 1998 similarstraw layers were found elsewhere in D1.The last remains of layer 051 were removedand then it became evident that it sealed thepostholes in the north of D1 and the singleposthole in the southern end. Similar layers(051,074,079) sealed the depression (075)in the north end and the large depression(542) in the middle. The same layers alsosealed the posthole in the south end and

were on top of the pavement (074,079). Itwas not possible to see whether these strawlayers had continued up against the easternand western walls in the southern two thirdsof the building because there the excavatorsin 1908 had cut right through to the subsoil.No straw layers were found in the passageD2 except beneath the fill in the easterndoorway. It is possible that the 1908excavation had removed them. The fact thatlayer 051 sealed the postholes and otherelements within D1 suggests that all thetimber structures had been removed beforeits use for hay storage.A section was cut in the middle of D1

along the north/south axis. This was donefor micromorphological sampling carriedout by Karen Milek. In this section twopinkish layers, 074 and 079, very similar to051 were described and it was noted thatthey were separated by a thin layer of turf

debris (078). Below these layers layers 084and 086 were revealed, interpreted as floorlayers.

Phase Ia: Beneath the turf constructions inD the sheet midden 004 (C4 in previousreports) can be seen in various places. Nofurther indications of the nature and extentwere revealed in 1998. The layer has beenobserved in several discontinuous locations;underneath the southern wall of the passage

D2, under the western wall of D1 and insection in the northern side of the northern

entrance on the eastern wall in D1. In thatlocation turf-debris layers have beenobserved underneath this marker horizon. Asimilar layer has been observed in Area E,also underlying the principal structures

there, E1 and E2, but its relationship withA/B is not yet clear, although evidence ismounting that the layer is more recent thanat least parts of the skli. This sheet middenand the deposits sealed by it represent aseparate phase of occupation in area D,earlier and quite different from thestructures D1 and D2. It is one of the tasksof the upcoming seasons of investigationsin Hofstair to explain the nature of thisremarkable deposit.

Conclusions

In 1998 the excavation of D1 on the insidewas completed, apart from possible layerswhich may be capped by the pavement inthe southern end of the structure, which wasnot removed. Surface layers associated withthe passage D2 were also examined but the

relationship between the structural elementsof D1, D2 and A/B awaits clarificationthrough examination on the outside,particuarly in D3 and on the junction of D2and A. The 1998 season also saw the lastremains of Daniel Bruuns excavationexplained and removed.

Bruun had realised that in the area he calledD there was some sort of a structure but hefailed to make anything of it. It can now be

seen that his failure was in part due to thefact that he, inadvertently, removed anumber of structural elements. This isclearly seen in the passage between theskli and D1 and also inside D1, as well ason the inside of the western wall of the skliitself.

It is striking that Bruun saw neither thefloor layers nor any other structuralelements inside D1. Similar features have

been observed both in areas E and A. Thisraises the question to what extent Bruuns

8/2/2019 Hst1998 Report

30/87

30

conclusions are based on observationsduring his excavation or whether they aremore a result of his observations andpreconceived ideas before he actuallyexcavated. Area D was obviously excavated

by Bruun from what he saw on the surfacein 1908. This underlines the importance ofreopening the skli to revaluate the 1908excavation results.

The relationship between the skli A/B, D1and D2 is now becoming clear. First astructure was constructed aligned with theskli. This may originally have been atimber house with turf walls added later. Ata later stage a passage was built connecting

the skli and D1. The excavations in area Dhave not yet been able to show that the skliwall is necessarily earlier than all thestructural remains in D1.

If indications which have appeared in 1998that 004 is later than the skli A/B as wellas the midden fills in G, it suggests that atleast the turf structure D1 must be later thanthe skli. The complexity and variablecomposition of the deposit which has beencalled 004 and 1004 (in E) does howeverwarn against making assumptions about itshomogeneity and usefulness as a singlechronological marker.

Excavation of D3 was not completed in1998 but this part of the area may containthe key to the understanding of therelationship between the structural elementsin D. The evidence indicates that D3 was an

open area between the skli and D1 but thedoorway in the middle of the eastern wallof D1, and layer 064 which may be remainsof a wall in D3. This indicates thepossibility of another room in that area.

8/2/2019 Hst1998 Report

31/87

31

8/2/2019 Hst1998 Report

32/87

32

8/2/2019 Hst1998 Report

33/87

33

Context Description

0016 Undistrubed over cultural layers.0051 Straw layer in north end.0055 Turf collapse in south gable wall.0059 Bruuns cut in A/B wall.0060 Fill in cut in A/B wall.0061 Isolated layer in D3 (same as 16).0062 Turf collapse in D3.0063 Turf collapse in D3.0064 Turf collapse in D3 but little differnt in texture from 62/63.0065 Turf collapse in D1 (small patch).0066 Fill in passage dorway.0067 Turf collapse in bottom of Bruuns trench in D1.0068 Pinkish layer (Piet) outside north wall in D1.0069 Turf collapse in D1, bottom of DB trench.0070 Accumulation against western wall (under 16).0071 Collapse from A/B wall (green turf).0072 Turf collapse in D1.0073 Single block of stratified material.0074 Organic/pinkish layer. Straw layer. (same as 51)0075 Cut. A hole in north end of D1.0076 Fill in charcoal pit in D1.0077 Straw layer in eastern doorway.0078 Turf layer between hay layers in D1.0079 Straw layer pinkish (under 78).0080 Layer (over postholes).0081 Posthole cut in north end of D1.0082 Posthole fill .0083 Turf collapse between longhouse and D3.0084 Dark redish/brown layer. Floor in D1.0085 Cut in charcoal pit in D1.0086 Dark brown layer (under 84) Floor in D1.0087 Cut for fill 82. Posthole.0088 Cut for posthole in D1. North end.0089 Posthole fill.0090 Cut. Support pole for a post.0091 Cut. Posthole.0092 Small hole beside the posthole.0093 Fill in posthole.0094 Posthole cut.0095 Small hole beside posthole.0096 Fill in posthole.0097 Posthole cut.0098 Posthole fill.0099 Posthole cut.0500 Posthole fill.0501 Posthole cut.0502 Posthole fill.0503 Posthole cut.0504 Posthole fill.0505 Posthole cut.0506 Posthole fill.0507 Posthole cut.0508 Posthole fill.

8/2/2019 Hst1998 Report

34/87

34

0509 Dark brown layer. Wood paneling.0510 Posthole fill.0511 Posthole fill.0512 Fill in a small hole.0513 Posthole fill.0514 Posthole fill.

0515 Fill in a small hole.0516 Posthole fill.0517 Fill.0518 Fill.0519 Fill.0520 Fill.0521 Fill.0522 Fill.0523 Fill.0524 Fill.0525 Posthole cut (510).

0526 Posthole cut (511).0527 Posthole cut (513).0528 Posthole cut (514).0529 Posthole cut (516).0530 Posthole fill.0531 Posthole cut.0532 Posthole.0533 Posthole cut.0534 Posthole fill.0535 Pavement in D1.0536 North wall in passage.0537 South wall in passage.

0538 Wall in western entrance (bslag).0539 Posthole cut (south D1).0540 Posthole fill.0541 Fill in turf wall (24).0542 Cut in large hole (fireplace)

Table 3-1. List of contexts from Area D

8/2/2019 Hst1998 Report

35/87

35

Figure 3-4. Area D Matrix.

8/2/2019 Hst1998 Report

36/87

36

8/2/2019 Hst1998 Report

37/87

37

Howell Roberts4.0 Area E Excavation Report

IntroductionInvestigations in Area E focused on theremains of two small structures (E1 and E2)partially uncovered during excavation in1997. Continuation of that work hasclarified the nature and extent of StructuresE1 and E2, and also the relationshipsbetween Structures E1, E2 and Skli A/B.Excavation facilitated the recovery ofnumerous environmental samples and a

large quantity of faunal remains, and thesewill contribute towards a broaderunderstanding of the role of Area E withinthe Hofstair complex.

Stratigraphic excavation demonstrated thatStructure E1 must have been built later thanboth the Skli A/B and Structure E2.Structures E1 and E2 can both been shownto be later than a widespread refuse horizonidentified at many locations across the site.

Within Area E, this refuse horizon appearsto respect the upstanding remains of SkliA/B and is apparently later than theconstruction of Skli A/B.

The Skli A/B is the earliest constructionwithin Area E, and Structure E1 is thelatest. All structural remains within Area Ecan be shown to belong to a phase ofoccupation at Hofstair that can be dated tothe 9 - 11th centuries by the relationship ofthese remains to knowntephrochronological horizons. Of theremaining layers, context 1016 represents aperiod of aeolian deposition from 1104 ADup to 1477 AD, and context 1001 representsthe continued deposition of soils from 1477AD up to the modern period. Contexts 1002and 1027 represent Daniel Bruunsexcavation in 1908.

In 1998 Area E was slightly enlarged bothto the south and east of the previous limits

of excavation in order to addressoutstanding questions about the nature ofStructure E1, and its relationship to theremains of Skli A/B. Excavation inStructure E2 proceeded in opposing

quadrants in order to facilitate the recoveryof micromorphological samples from theentire extent of this structure. Singlecontext recording was continued from theplans and context records made during the1997 excavation. Further investigation ofthose contexts unexcavated in 1997 hasallowed the development and re-evaluationof evidence gathered in 1997.Table 4-1 lists all contexts recorded in AreaE in 1997 and 1998, and a stratigraphic

matrix (Figure 4-4) illustrates therelationships between those contexts.Excavation in Area E was conducted bystudents of The Field school in IcelandicArchaeology, closely supervised andinstructed by the author. Excavation byquadrant within Structure E2 was carriedout by the author and Karen Milek(University of Cambridge) who wasresponsible for micromorphologicalsampling within Area E.

Results

Structure E1Removal of all backfill, trample andmodern material from Trench E, opened byDaniel Bruun in 1908, and re-opened in1992, revealed some small remnants ofsurviving archaeology at the base of thetrench, within Structure E1 (fig. 4-2).

Careful study of the sections of Trench Eduring 1997 revealed the presence of asequence of fine internal deposits, beneathseveral layers of mixed turf debris.Although heavily truncated by Trench E,these internal deposits could be seen toform similar sequences of deposition to thenorth and south of Trench E. The mostcomplex of these sequences was to be seento the north of Trench E, where truncationhad been least severe.

8/2/2019 Hst1998 Report

38/87

38

8/2/2019 Hst1998 Report

39/87

39

Excavation in 1998 commenced with theremoval of a block of turf (context 1048)which could be shown to lie over theuppermost of the fine internal deposits(context 1030). Context 1048 was

composed of yellow/grey/green turfstrengur, and is interpreted as an episode ofstructural collapse. At the base of Trench Ethe truncated remains of a possible floorlayer were located (contexts 1079 and1080). Context 1079 was a mixed pinkishbrown silt located at the centre of StructureE1, and exhibited signs of compaction. Thelower limit of context 1079 was formed bya lens of paler pink material with afibrous/organic component, sitting over

natural deposits. Context 1080 was locatedto the west of context 1079 and appearedsomewhat darker in colour. The compactionof context 1080 was highly variable, andthis layer was found to contain occasionalflecks of white H-3 tephra along withcharcoal, peat and turf fragments. Theremoval of context 1080 revealed a smallirregular linear depression (context [1081])believed to be consistent with the removalof a post or post-pad.

Context 1030, a mid grey brown sandy silt,had been largely exposed by excavation in1997. Upon excavation, it was seen toinclude occasional small rounded stones,occasional charcoal, and fragments of bone.Removal of context 1030 revealed adiscrete deposit of soot and soil (context1096) located towards the centre ofStructure E1. Context 1096 was largely

black in colour and measured c. 0.80meast/west, 0.35m north/south and up to65mm in depth. The latter context wassemi-circular in shape, as it had beentruncated by Trench E. Context 1096included several fine lenses of yellowbrown silt, and therefore is believed to haveaccumulated over some period of time, andis unlikely to represent a single episode ofdeposition. Soil samples were recoveredfrom this context for further analysis.

Context 1093, a small patch of turf debristowards the west of the structure, was found

to be beneath both contexts 1030 and 1088.Context 1088 was a similar small patch ofturf debris, situated over context 1093, butseparated from it by elements of context1030. The removal of these latter contexts

exposed a soft pale pinkish brown deposit(context 1097) with a fibrous component.Context 1097 was situated in the centralnorthern portion of Structure E1, andmeasured 2.09m east/west x 0.71mnorth/south with a maximum depth of15mm, being thickest at its truncation byTrench E. Context 1097 overlay above asimilar pinkish brown deposit (context1100) extending further to the west.Context 1100 measured 1.78m east/west x

0.67m north/south with a maximum depthof 55mm, and was seen to be somewhatdarker and more mixed than context 1097,containing charcoal, occasional small turffragments, and fine lenses of green grey siltat its upper horizon. Beneath context 1100was another thin deposit (context 1106) of avery similar nature. Context 1106 extendedover a somewhat larger area withinStructure E1 filling most of its northernportion and measuring 4.42m east/west x0.95m north/south with a maximum depthof 30mm. Context 1106 was distinguishedfrom context 1100 by the presence of finediscontinuous lenses of green grey silt atthe upper horizon of context 1106. Sealedby context 1106 was a thin mixed depositof dark greyish brown sandy silt (context1109), found to include occasional smallfragments of turf, some charcoal, bone andoccasional small lenses of material very

similar to the overlying deposits 1093,1097, 1100, and 1106. Context 1109measured 4.92m east/west x 0.91mnorth/south with a depth typically of c.25mm, but thickening to a depth of up to80mm at its western limit. Removal of thislayer revealed a variety of underlyingdeposits. At the west of Structure E1 was apatch of mixed reddish turf debris (context1132) overlying a yellowish brown siltlayer (context 1133). Both latter deposits

lay up against the internal face of thenorthern wall of Structure E1. In the central

8/2/2019 Hst1998 Report

40/87

40

8/2/2019 Hst1998 Report

41/87

41

northern portion context 1109 sealed a verypale pinkish brown thin layer (context1113) contained within a shallow

depression at the centre of Structure E1.Context 1113 contained occasional smallpatches of yellow green turf debris and afew tiny fragments of calcined bone. East ofcontext 1113 was a small patch of a similarbut separate deposit (context 1115).Context 1115 extended towards the easternlimit of Structure E1. At the junction of thelatter structure and Skli A/B context 1109lay over a small truncated patch of greysandy silt (context 1180). Context 1180

may extend into the internal area of SkliA/B, and has been left unexcavated at thistime. The removal of layers 1113 and 1115exposed the remains of a number of smallinternal post holes that may represent anearlier period of use within Structure E1.

Located at the internal edge of the northernwall of Structure E1 were two very smallpost holes (features [1169] and [1171]),along with a single larger post hole (feature

[1162]). Feature 1169 measured 75mm indiameter and 115mm in depth and wasfilled by a soft mixed brown silt (context1168) with traces of reddish organicmaterial. Feature [1171] measured 65-80mm in diameter and 100mm in depth,and was filled by a yellow brown silt(context 1170). The latter features wereseparated by a distance ofc. 0.20m. Feature1162 was situated c. 0.11m to the west offeature 1169, and c. 1.90m from the westernlimit of Structure E1. Feature 1162measured 195mm in diameter with a depthof up to 0.25m, and was filled by a mixedgrey brown sandy silt (context 1161). Boththe cut and fill of feature 1162 exhibitedindications of a post pipe. The centralportion of context 1161 was found to beloose and soft, and upon excavation thebase of feature 1162 was stepped, being80mm deeper at the centre.

A further small post hole (feature [1159])

was located towards the centre line ofStructure E1, c. 1.10m from the westernlimit of that structure. Feature 1159

measured 90-100mm in diameter by 0.22min depth, and was filled by a yellow brownsandy silt (context 1158). A shallow sub-square post hole (feature [1118]) waslocated 1.65m from the eastern limit ofStructure E1, 0.45m south of the northernwall. Feature 1118 measured up to 0.22meast/west and 80mm in depth, but had beentruncated at its southern edge by Trench E.Feature 1118 was filled by a mixed greybrown silt (context 1117), including

occasional small lenses of pinkish brownsilt.

To the south of Trench E, a similarsequence of deposits and features wereobserved. Context 1075, a mid grey brownsandy silt (very similar to context 1030) hadbeen exposed by excavation in 1997,following the removal of layers of turfdebris (contexts 1049 and 1051). Furtherinvestigation at the eastern end of the

southern wall of Structure E1 showedcontext 1051 to be filling and extendingbeyond an opening between that wall andthe western wall of Skli A/B. To the southof Structure E1, context 1051 was overlainby a series of deposits of grey green turfdebris (contexts 1045, 1050 and 1054excavated in 1997, and contexts 1108, 1110and 1114 excavated in 1998).

Context 1075 measured 2.52m east/west by0.58m north/south, with a maximum depthof 40mm, and was found to containfragments of turf. Beneath context 1075was a pale pinkish brown silt with anorganic content (context 1094), very similarto those seen north of Trench E. Context1094 measured 0.92m east/west, 0.34mnorth/south and between 2-10mm inthickness, and was sampled for furtheranalysis