Archaeological excavations at Lungi Tepa, south Uzbekistan. Report for Season 2021

145STUDIA HERCYNIA XXVII/2, 145–175 A chaeological exca a ions a Lungi Tepa, sou h Uzbekis an. Repo o Season 2021 Ladisla Damašek– Daniel Pilař– Samuel Ke és– Shapula Shaydullae ABSTRACT This epo summa ises he p elimina y esul s o he second season o exca a ions a he si e o Lungi Tepa in he Kugi ang Piedmon s, sou h Uzbekis an. The esea ch was conduc ed by aCzech -Uzbek eam in au umn 2021. The esul s sugges ha he si e was se led in he pos -Kushan pe iod, he Ea ly and High Medie al pe iod wi h aces o se lemen om he La e Medie al pe iod. La e i was used as abu ial g ound o he local popula ion. T aces o me al p oduc ion we e ound a he si e. KEYWORDS Cen al Asia; Tokha is an; Lungi Tepa; High Medie al Pe iod; Ea ly Medie al Pe iod; a chaeology; exca a ion; se lemen ; s a ig aphy; me al p oduc ion. INTRODUCTION The High Medie al pe iod in he Kugi ang Piedmon s is no e y well co e ed by a chae- ology and his o y. Only a ew si es om his pe iod ha e been exca a ed so a in he a ea (Bobokhojae – Annae – Rakhmano 1990; Mok obo odo 2007; Šejko 2011). La ely, an ex ensi e ield su ey was ca ied ou by he Czech -Uzbek a chaeological expedi ion in he Kugi ang and Hissa piedmon s con i ming many o he p e iously ound si es as well as disco e ing new ones (Damašek 2019, 48). Based on ou wo ks (co obo a ing he opinions o o he schola s) he e was asigni ican g ow h in popula ion in he High Medie al pe iod in his egion (S ide 2004; Augus ino áe al. 2016; S ančo 2019). This can pa ly be a ib- u ed o he exploi a ion o me als, aces o which we e ound in he a ea o he Kugi ang and Hissa piedmon s (S ančoe al. 2019, 145, 162). Bu aclose look a bo h he a chaeological and his o ical e idence is needed o ge abe e unde s anding o he p ocesses which led o he apid se lemen g ow h in he High Medie al pe iod. The Czech -Uzbek expedi ion commenced i s ield wo k on he si e o Lungi Tepa in 2019 and in 2021 a e aone -yea b eak came back o inish he wo k. In 2019, ou main phases o he se lemen we e es ablished based on he s a ig aphy bu nei he he lowe le els o he se lemen no he subsoil we e eached (Damašeke al. 2020). The main phases con ained ase lemen o he High Medie al ( om he A ab phase o he Mongol in asion, 8 h–13 h cen u y AD) and he Ea ly Medie al pe iods (5 h–8 h cen u y AD), as well as a16 h–18 h cen u y AD bu ial g ound. The eam wo king on he si e in 2021 consis ed o a chaeologis s om Cha les Uni e si y, namely Ladisla Damašek, Daniel Pilař, and Samuel Ke és. The wo k was done unde he auspices o he Czech -Uzbek a chaeological expedi ion led by Shapula Shaydullae (Te mez S a e Uni e si y) and Ladisla S ančo (Cha les Uni e si y). Kach aman Toshalie (Te mez S a e Uni e si y) p o ided logis ic and adminis a i e sup- po o he expedi ion. Fou local wo ke s om he illage o Khojaunkan pa icipa ed in he exca a ions as wo ke s. 146 STUDIA HERCYNIA XXVII/2 LOCATION Lungi Tepa is loca ed on he eas e n edge o he mode n -day illage/mic o -oasis o Khojaun- kan (Fig. 1) in he Kugi ang oo hills (She obod Dis ic , Su xonda yo P o ince). The illage o Khojaunkan (37°53’48.549”N, 66°46’33.414”E) is si ua ed app oxima ely 30 km as he c ow lies om She obod own. The si e Lungi Tepa consis s o acen al mound and apla o m su - ounding i . The cen al mound ( ell se lemen , in local e minology epa) has an o al shape wi h dimensions o oughly 60 × 40 m a he base (Fig. 2). The a e age heigh o he cen al mound is app oxima ely 6.5 m. The pla o m has aclea bo de on he eas e n and pa ly on he no he n side, bu on he sou he n and wes e n side i is dis up ed by con empo a y houses and o he mino cons uc ions such as ences and ou buildings. Thus, i is di icul o de ine i s p ecise o iginal shape and a ea. Fig. 1: Loca ion o Lungi Tepa inside he illage/mic o -oasis o Khojaunkan (L. Damašek). PREVIOUS ACTIVITIES Lungi Tepa was known by local a chaeologis s om he 1980swhen he i s a chaeological esea ch was ca ied ou a he si e (Bobokhojae – Annae – Rakhmano 1990). The Czech -Uzbek expedi ion i s ook an in e es in he si e in 2017. Asu ace su ey was ca ied ou and da ing o he si e o he Medie al pe iod was con i med (S ančoe al. 2017). In 2019, a eam consis ing o an h opologis s led by Rebecca Kinas on (Uni e si y o O ago, NZ) and a chaeologis s led by Ladisla Damašek (Cha les Uni e si y, CZ) selec ed he si e o a ial 147LADISLAV DAMAŠEK– DANIEL PILAŘ– SAMUEL KERTÉS– SHAPULAT SHAYDULLAEV exca a ion, whose p elimina y esul s ha e been published (Damašeke al. 2020). Howe e , he exca a ion o he si e was no inished in 2019. This was mainly because he main ell mound was used as aceme e y in la e his o ical pe iods (which had no been an icipa ed) and he exca a ion o he g a es ook up much o he ield wo k ime. In au umn 2021 a e aone -yea b eak caused by co id es ic ions, he membe s o he Czech -Uzbek expedi ion e u ned o Lungi Tepa o inish he exca a ions s a ed in 2019. SURVEYED AREAS In 2021, he main goal o he Lungi Tepa eam was o inish he s a ig aphic sec ion o he cen al mound ( ench S1), which was s a ed in 2019. In addi ion o S1, ou new a eas we e explo ed in 2021 a ound he cen al mound o Lungi Tepa, namely S5, SB3, SB4, and SB5 (Fig.2). The las h ee men ioned enches a e spo s whe e po e y was collec ed om he su ace o he pla o m. No sys ema ic app oach was used o collec ing he po e y sha ds. SB3 and SB5 a e spo s whe e he pla o m was dis up ed by ecen human ac i i y. SB3 is aplace o clay ex ac ion loca ed in aga den o he wes o he cen al mound o Lungi Tepa. I is in he i- cini y o he es pi S2 om 2019 and i was dug deep in o he pla o m. Po e y was collec ed om he walls o his clay ex ac ion si e. SB5 is ashallow hollow way on he sou heas edge o he pla o m. In his place adi oad c osses he edge o he pla o m causing e osion. Po e y sha ds we e collec ed in and a ound he hollow way. SB4 is aplace indica ed by local inhab- i an s as he ind spo o acomple e ce amic essel (Fig. 3). S5 is loca ed on he eas e n edge o he pla o m, i s main objec i e was o explo e/examine he s uc u e and s a ig aphy o he pla o m. The pla o m has as eep slope on he eas e n side which was dis up ed by clay digging o anea by mud ence wall (Fig. 4). We ook ad an age o his dis up ion and he exis ing p o ile was simply s aigh ened o see i he basic s a ig aphy can be de e mined. I s inal dep h as well as wid h was app oxima ely 2 m (Fig. 4). Po e y sha ds we e p esen only in he uppe 20 cm o soil. The assump ion ha he basic s a ig aphy o he pla o m Fig. 2: Map o he Lungi Tepa si e (L. Damašek). 148 STUDIA HERCYNIA XXVII/2 would be e ealed h ough he cleaning o an exis ing e ain dis up ion was no con i med. In o de o ge mo e in o ma ion abou he na u e o he pla o m, as a ig aphic sec ion o b oade exca a ions a e needed. Such an unde aking was no possible gi en he ime limi- a ions o he Lungi Tepa eam. Fig. 3: A essel ound by local inhabi an s in he ind -spo SB4 (pho o by S. Ke és). Fig. 4: T ench S5 be o e and a e exca a ion (pho o by L. Damašek). TRENCH S1 In 2019, he bo om o ench 1 was co e ed wi h aplas ic shee and hen back illed wi h soil a he end o he exca a ion. In 2021, he soil was emo ed o he le el o he plas ic shee and hen he exca a ion wo k con inued deepe in o he Lungi Tepa main mound. On i s sou he n side he ench S1 was ex ended ( om 9 m in 2019) owa ds he base o he cen al mound up o app oxima ely 16.4 m. I s wid h o 2 m and o ien a ion o he ca dinal di ec ion was main- ained. The inal dimensions o S1 we e hus 16.4 m in leng h, 2 m in wid h, and almos 4 m 149LADISLAV DAMAŠEK– DANIEL PILAŘ– SAMUEL KERTÉS– SHAPULAT SHAYDULLAEV in dep h (in he no he n pa ). Two baulks labelled K1 and K2 we e le in he no h pa o S1 and a es pi labelled R1 was exca a ed in he sou he n pa o S1. All he digging was done manually using hand ools. Ame al de ec o was used o check bo h he exca a ed soil and he exca a ed le els inside S1. G a e illing and selec ed con ex s we e sie ed using a2 mm sie e. S1 as well as o he places subjec o exca a ion we e measu ed (loca ion, al i ude) by o al s a ion. Topog aphical e e ence poin s c ea ed in 2019 we e used o se up he o al s a- ion. The e o e, he 2021 measu emen should ha e he same accu acy/e o as ha o 2019. Pho og aphic documen a ion was made along wi h d awings o selec ed si ua ions. Selec ed si ua ions we e also documen ed o 3D models. The po e y om s a ig aphic uni PI was coun ed and weighed. Selec ed sha ds om PI ( ims, deco a ion, e c.) we e labelled, pho og aphed, and d awn. All he po e y om he es o he s a ig aphic uni s was labelled, pho og aphed, and d awn wi hou excep ion. All o he inds (bones, glass, s one, me al, and o he s) we e coun ed pho og aphed and in some cases d awn. EXCAVATED UNITS The newly unco e ed sou he n pa o S1 yielded con ex s om all he ou ecognised phases (PI, PII, PIII, PIV) al eady dis inguished in 2019 (Damašeke al. 2020). In he o iginal pa o S1, wo k con inued om he le el o PIII/PIV. In 2021, anew phase labelled PV was ecognised. Subsoil was eached in he sou he n pa o S1 by he base o he epa. G a es (= PII) we e expec ed in he newly exca a ed (sou he n) pa as hey we e p esen unde nea h su ace laye s (PI) in 2019 and g ea e a en ion was paid o hei iden i ica ion. PI The ench was dug om he su ace by spi s un il any dis inguishable ea u es we e ound. All hese uni s belong o he phase PI, he dep h o which a ies depending on he s a ig aphy p ese ed unde nea h. Logically, i is hicke a he base o he ell mound due o he e osion. Anumbe o small me allic inds was de ec ed by me al de ec o in he con ex s o his phase (Tab.1; Fig. 18:52, 18:56). Te aco a igu ines o abi d and o ano he unde e mined animal we e ound among he ma e ial o his phase (Fig. 2:4–5). In 2019, aSubphase IA (PIA) was es ablished (Damašeke al. 2020, 166). A e e isi ing he si ua ion, his subphase was can- celled and he pi (labelled OBJ12), which was i s only componen , was eassigned o phase III. Fig. 5: G a es (d awing by L. Damašek). 150 STUDIA HERCYNIA XXVII/2 Fig. 6: G a es H9, H11, and H12 (pho o by L. Damašek). PII The exca a ion s a ed by emo ing he su ace laye s in he newly (2021) unco e ed sou he n pa . Benea h hese laye s, ou g a es we e ound: H9, H10, H11, and H12 (Fig. 5–6). Th ee o hem we e child en’sbu ials (Fig. 6) and one was p obably an adul . The child en’sg a es (H9, H11 and H12) we e exca a ed bu no he adul one (H10). I s dep h and posi ion allowed he skele on o be le un ouched wi hou in e e ing wi h wo k on he s a ig aphic sec ion. Jus he uppe pa o he H10 g a e -pi and illing was emo ed, while he skele on was le in ac . Thus, he assump ion ha i is an adul g a e is based en i ely on he dimensions o he g a e -pi . All he g a es ha e an iden ical sou h -no h o ien a ion wi h he head acing wes (which could be only assumed in he case o H10). The e we e no g a e goods and he g a e pi s we e ela i ely na ow. H9 was co e ed by la s ones and wo pieces o wha we e p obably iles (Fig. 7). H10 had s ones inside he g a e back ill. H11 and H10 had aniche dug in hei no he n pa (Fig. 5). H11 and H12 we e in mu ual supe posi ion. H11 dis up ed he sou he nmos pa o H12 (Fig. 5). All he g a es we e dis u bed by bio u ba ion namely by mice. Mouse holes oge he wi h an abundance o mouse bones we e ound in all he g a e illings (Fig. 8). The h ee child en’sg a es (H9, H11, and H12) we e in each case incomple e, pa s o he skele ons we e missing. Fig. 7: G a e H9 wi h and wi hou s one co e (pho o by L. Damašek). 151LADISLAV DAMAŠEK– DANIEL PILAŘ– SAMUEL KERTÉS– SHAPULAT SHAYDULLAEV The e we e no ac o s, which would sugges ha hese g a es we e no associa ed wi h hose (H1–H8) p e iously ound in 2019. The e o e, all he g a es (H9–H12) om 2021 we e assigned o he phase PII. This phase is ep esen ed by abu ial g ound which co e ed a leas he cen al mound o Lungi Tepa. Two g a es H4 and H5 o PII om 2019 we e da ed using he adioca - bon me hod. In bo h cases a ib bone was used o he da ing. The sample om H4 was da ed be ween 1510–1660 AD and he sample om H6 be ween 1460–1640 AD (Fig. 9:1–2). Thus, he bu ial g ound can be oughly da ed o he ime span om he 15 h–17 h cen u y co esponding o he pe iod o he Uzbek and Bukha a khana e. I is wo h no ing ha om 12 g a es de ec ed in 2019 and 2021 se en we e child en’sg a es and only i e belonged o adul s. Fig. 9: Radioca bon da ing o H4, H5 and wo pieces o cow bones om laye 13. Fig. 8: Mice bones inside H11 (pho o by L. Damašek). 152 STUDIA HERCYNIA XXVII/2 PIII The g a es o e lie he ho izon PIII es ablished in 2019 and consis o all ea u es om he Islamic pe iod (High Medie al pe iod; 8 h–13 h cen u y AD). The emains o apakhsa ( ammed clay) wall (OB8) emained in he no he n pa o S1 (Fig. 11–12). The emains o he wall we e disman led in 2021, and anew ea u e was ecognized which was no iden i ied in 2019. I was p obably a ound pi wi h a la bo om (labelled 24) which was cu in o he pakhsa wall – OB8 (Fig. 11). I s eas e n pa emained unexca a ed, because i was beyond he eas e n p o ile o S1 and i s sou he n pa was des oyed by he g a e -pi o g a e H8. I s ela ion wi h he uppe laye s emains unce ain, bu being cu by ag a e -pi (phase PII) and con aining po e y simila o po e y collec ions om PIII, he pi ( ea u e 24) was assigned o PIII. The base o OB8 was ound, i o med a la su ace which was easily ecognisable because he clay blocks b eak o om i (Fig. 10). The base o OB8 was he uppe pa o he ounda ion o OB8 (labelled 30) which was ha dly ecognisable du ing he exca a ion bu was e y well dis inguished on he p o iles especially on he wes e n one (Fig. 12). The abo e -men ioned ea u es we e he emains o PIII in he pa o S1 exca a ed in 2019. O he ea u es om PIII we e exca a ed in he pa o S1 unco e ed in 2021. The mos signi ican o hese we e he e- mains o h ee walls (labelled 25, 28, and 34), as o age pi (labelled 39) and se e al ho izon al laye s (labelled 32, 501, and 502). Walls 25 and 28 we e in supe posi ion and hey ollowed he con ou s o he main ell mound (Fig. 11–12). I looked like hey o med ings a ound he main ell mound, bu i is ha d o d aw any conclusion om he 2 m wide ench. Wall 25 was made o s one (Fig. 13) and i cu h ough wall 28, whe eas wall 28 was eco ded only as nega i e in he o m o adi ch. On he eas e n side, bo h walls we e dis u bed by wo child en’sg a es H11 and H12. Unde nea h walls 25 and 28, laye 32 ollowed which con ained ag ea many po e y sha ds. Laye 32 con inued sou h o he base o he ell mound whe e i lay on laye 37 and subsoil (Fig. 11). In he sou h pa o S1 a es pi labelled R1 was placed o check he dep h o he subsoil (Fig. 12, 14). The illing o he es pi labelled 400 was exca a ed in one go. Only a e ha , laye s 32 and 37 we e dis inguished in he p o ile inside R1. The emains o awall (34) we e loca ed unde nea h laye 32 a he base o he ell mound. I was apakhsa wall bu only asmall piece o i was p ese ed. I is bes isible on he wes e n p o ile (Fig. 15). Wall 34 did no go h ough he en i e wid h o S1 bu i ended in he middle o S1 whe e i was cu by s o age pi 39. In he eas e n pa o S1 in he place o 34 he e is al eady subsoil (Fig. 11). Small inds om he con ex s o PIII wi h he excep ion o s o age pi 39 include small glass agmen s, an a owhead, and apiece o nac e (Tab.1; Fig. 18:57; Pl. 4/1:6–10). Fig. 10: Clay wall OB8 (pho o by L. Damašek). 153LADISLAV DAMAŠEK– DANIEL PILAŘ– SAMUEL KERTÉS– SHAPULAT SHAYDULLAEV Fig. 11: Eas e n p o ile o S1 (d awing by L. Damašek). 160 STUDIA HERCYNIA XXVII/2 Fig. 22: Po e y phase III– s o age pi (d awing by L. Damašek). 161LADISLAV DAMAŠEK– DANIEL PILAŘ– SAMUEL KERTÉS– SHAPULAT SHAYDULLAEV Fig. 23: Po e y phase III– s o age pi (d awing by L. Damašek). 162 STUDIA HERCYNIA XXVII/2 Fig. 24: Te aco a and s uc u al elemen s (d awing by L. Damašek). (Pl.4/2:10–18; 3:9–14), geome ic deco a ion in black on awhi e backg ound composed o do s and lines– apseudo ege al mo i (Pl. 4/3:1–4), and monoch ome g een glaze (Pl. 4/2–3). Whi e opaque glaze wi h g een splashes was p esen on wo sha ds (Pl. 4/2:3, 4/3:17). Wi h only a ew excep ions (Pl. 4/2–3) almos all he glazed agmen s which a e big enough o allow de e mina ion a e om bowls. All he main deco a ion g oups ecognized a Lungi Tepa a e well known in Cen al Asia and we e de ec ed in ma e ial om old Te mez. The splashed sg a - ia o wa e was widesp ead in Cen al Asia in he 9 h –11 h cen u y AD (Wilkinson 1973, g oup 2; Wa son 2004, F splashed wa es 199–203) and was cap u ed in Te mez (Fe e ase al. 2020, 261; Houal 2021, pl. 256, 84). I was he mos nume ous (26%, 25 agmen s) deco a i e mo i among he glazed wa e. The geome ic deco a ion in black on awhi e backg ound is da ed o he 10 h–11 h cen u y (Houal 2021, P3-1 251, pl. 79). The monoch ome g een glaze belongs o wo di e en g oups, he agmen om s o age pi 39 (Pl. 4/2:2) seems like an ea ly a ian o he g een monoch ome glaze which is dimme . The es o he agmen s (3 pieces) a e likely he la e a ian wi h ab igh su ace (Pl. 4/3:16). The ea ly a ian is usually da ed o he 8 h–10 h cen u y and he la e a ian is da ed o he 11 h–12 h cen u y (Houal 2021, M1 245, pl. 75, M3 247, pl. 75). Wi hin he unglazed (Fig. 25–27) wa e, one shape s ood ou , apla e o alid (un o una ely only he ims a e p esen , no he whole shapes) wi h an e e ed im which is deco a ed wi h ca ed o namen s (Fig. 25). These we e no iced du ing he i s exca a ion o Lungi Tepa (Bobokhojae – Annae – Rakhmano 1990, ig. 5:5–8). The closes analogy ound a e lids om Qashqada yo wi h aca ed deco a ion (Isamiddino – Khasano 2000, ig. 30:II, 141). O he unglazed shapes (excluding po e y om s o age pi 39) include po s, basins, pi che s, and as o age ja (Fig. 26, 27). 163LADISLAV DAMAŠEK– DANIEL PILAŘ– SAMUEL KERTÉS– SHAPULAT SHAYDULLAEV Fig. 25: Po e y phase III– lids/pla es (d awing by L. Damašek). 164 STUDIA HERCYNIA XXVII/2 One e aco a igu ine was ound in he con ex s om PIII apa om hose om he s o age pi (Fig. 24:2). The same ypes o glazed wa e we e de ec ed in he assemblage in 2021 as in 2019 namely monoch ome and polych ome wa e. Fig. 26: Po e y phase III (d awing by L. Damašek). 165LADISLAV DAMAŠEK– DANIEL PILAŘ– SAMUEL KERTÉS– SHAPULAT SHAYDULLAEV Fig. 27: Po e y phase III (d awing by L. Damašek). 166 STUDIA HERCYNIA XXVII/2 PIV PIII was ollowed by PIV. This phase was also es ablished a e exca a ions in 2019 and include ea u es om he p e -Islamic pe iod (Ea ly Medie al pe iod, 5 h–8 h cen u y AD). Two adi- oca bon da es we e ob ained om ma e ial (cow bones) ga he ed om laye 13 in 2019. Bo h da es i in he same span o 650–780 AD (Fig. 9). Fea u es om his phase a e concen a ed in he no he n pa o S1 close o he uppe pa o he main mound and na u ally deepe in he body o he main mound. In 2021, he same cha ac e is ics as in 2019 came o ligh . The e we e no many dis inguishable ea u es in phase IV (Fig. 11, 12), laye s (labelled 33, 41, and 16) we e hicke and con ained an abundance o animal bone and ce amics. Laye s om phase IV we e o en exca a ed by spi s due o hei hickness and abundance o ma e ial. Se e al i on objec s we e ound including akni e, wo a owheads, and asocke (Tab.1; Fig. 18:54, 58, 63, 64). Po e y om phase IV was mainly handmade wi h acoa se ce amic ab ic. Two main g oups can be dis inguished. Fine wa e and coa se wa e. The ine wa e is wheel h own whe eas he coa se wa e is handmade wi h alo o inclusions. Among he coa se wa e a e ep esen ed mos ly big jugs and po s (Fig. 28, 29, 30) which o en had aces o use on an open i e. Two almos comple e jugs and one pa ly comple e jug we e ound in laye 33 (Fig. 28). Ho seshoe -shaped handles, ho izon al handles, a ched handles, and a ched handles wi h ings in he middle we e ound in 2019 as well as in 2021 (Fig. 29). Rims o s o age ja s wi h ab own - ed su ace we e p esen in he assemblage (Fig. 29:1–3). The po e y has simila cha ac e is ics as o he Ea ly Medie al assemblages om Su xonda yo such as po e y om Dabilku gan (Solo jo 2013, ig. 8), Shu obku gan (Bolelo 2004, 53–61), and o he si es Fig. 28: Po e y phase IV– jugs (d awing by L. Damašek). 167LADISLAV DAMAŠEK– DANIEL PILAŘ– SAMUEL KERTÉS– SHAPULAT SHAYDULLAEV Fig. 29: Po e y phase IV– po s (d awing by L. Damašek). 168 STUDIA HERCYNIA XXVII/2 Fig. 30: Po e y phase IV– po s (d awing by L. Damašek). 169LADISLAV DAMAŠEK– DANIEL PILAŘ– SAMUEL KERTÉS– SHAPULAT SHAYDULLAEV (Annae 1988, ab.I–XVII). An in e es ing agmen o po e y which is mos p obably Ea ly Medie al was ound in laye 24 (phase III). I has well p ocessed clay and i is deco a ed wi h as amp in he o m o as ylised human ace (Pl. 4/1:14). Analogies o his ype o s amped deco a ion can be ound in po e y om he 5 h–7 h cen u y AD (Ma shak 2012, 356). PV In 2021, anew phase Vwas ecognised. This phase was dis inguished based on he s a ig a- phy and change in cha ac e o laye s and composi ion o ce amics. Fea u es om his phase V(laye s labelled 36 and 44) we e concen a ed in he no he n pa o S1 and hey ep esen ed he deepes unco e ed a chaeological si ua ions (Fig. 11, 12). The subsoil was only eached in he sou he n pa o S1 nea he base o he ell mound, so phase V emains he las eco ded uni . I is no likely, bu he possibili y o ano he phase o phases ollowing PV canno be excluded, since he subsoil was no eached in he no he n pa o S1 below PV. In e ms o absolu e da ing, phase Vcan be da ed o he pos -Kushan pe iod/Kushan -Sasanian pe iod (3 d–5 h cen u y AD). This da ing is indica i e and needs o be con i med by exac da ing me h- ods. The laye s and o he ea u es o his phase con ain ewe a e ac s compa ed o he es o he s a ig aphy. They almos lack any ma e ial excep po e y which was spa se compa ed o he laye s belonging o phases III and IV. Laye 44 did no con ain any a e ac s. An impo an cha ac e is ic o phase Vis ha i con ains big agmen s o s o age ja s, bo h in clus e s and isola ed (Fig. 31). This is he case o laye 36 which mus ha e been accumula ed in asho pe iod o ime because some o he bigge agmen s o s o age ja s (app oxima ely 30–40 cm) we e in a e ical posi ion. I is clea ly isible on he no he n wall o baulk K1 (Fig.32). All he laye s o PV we e also compac ed and ha d o dig using ine ools. In some cases, apickaxe had o be used. The only no iceable ea u e in PV was asmall o en co e ed om he sides by agmen s o s o age ja s (Fig. 33). I is ha d o de e mine he pu pose o his Fig. 31: F agmen s o s o age ja s (pho o by L. Damašek). 198 STUDIA HERCYNIA XXVII/2 Pl. 4/1: Glass and po e y (d awing by L. Damašek; pho o by S. Ke és). 199PLATES Pl. 4/2: Glazed po e y (d awing by L. Damašek; pho o by S. Ke és). 200 STUDIA HERCYNIA XXVII/2 Pl. 4/3: Glazed po e y (d awing by L. Damašek; pho o by S. Ke és).