File PDF .it

Condividi facilmente i tuoi documenti PDF con i tuoi contatti, il Web e i Social network.

Inviare un file File manager Cassetta degli attrezzi Ricerca PDF Assistenza Contattaci

ATN53 .pdf

Nome del file originale: ATN53.pdf
Titolo: Cover ATN 53.jpg
Autore: frisch

Questo documento in formato PDF 1.6 è stato generato da PScript5.dll Version 5.2.2 / Acrobat Distiller 9.0.0 (Windows), ed è stato inviato su il 28/07/2016 alle 11:22, dall'indirizzo IP 213.144.x.x. La pagina di download del file è stata vista 1875 volte.
Dimensione del file: 6.6 MB (61 pagine).
Privacy: file pubblico

Scarica il file PDF

Anteprima del documento

ATN is a twice yearly journal
published by the Society Friends
of ATN, hosted by DNRF CTR.


Rediscovering ancient activities: textile tools
in a 3rd - 2nd century BC context from
Herakleia, Southern Basilicata, Italy
Francesco Meo


A 16th century silk fabric fragment
from northern Finland
Sanna Lipkin and Anna-Kaisa Salmi



The scientific committee:
John Peter Wild, UK
Lise Bender Jørgensen, Norway
Elisabeth Wincott Heckett, Ireland
Johanna Banck-Burgess, Germany
Ad hoc members

Textile templates for ceramic crucibles in
early Islamic Akhsiket, Uzbekistan
Rahil Alipour, Margarita Gleba and
Thilo Rehren
Textiles and pottery: Insights into Neolithic
and Cooper Age pottery manufacturing
techniques from Romania
Paula Mazăre


Cover: Marianne Bloch Hansen
Layout: Sidsel Frisch

10th century AD textiles from female burial
Ц-301 at Gnëzdovo, Russia
Olga Orfinskaya and Tamara Pushkina


Editorial Board:
Ulla Mannering
Eva B. Andersson Strand
Margarita Gleba
Susanna Harris
Carol A. Christiansen

Print: Grafisk
University of Copenhagen

Subscription information:
To become a member of the society
Friends of the Archaeological
Textiles Newsletter and receive
ATN, as well as for institutional
subscription please visit:

Textiles from the Nile Valley Conference 2011
Henriette Koefoed


ISSN 0169-7331



Recent publications, websites,
and dissertations


Textile Calendar 2012



er No. er
53No. xx

2011 has seen much activity in the field of
archaeological textiles, including many conferences
and workshops, numerous valuable publications
and exciting new discoveries. Issue 53 includes two
articles that demonstrate how textiles may inform our
understanding of other technologies in the past, in
particular pottery making technology in the Neolithic
Europe and steel forging in the Islamic Central Asia.
The issue also includes a couple of new discoveries –
the first detailed analysis in English of the important
textile finds from the Viking-Age site of Gnezdovo
in Russia and the publication of a 16th-century silk
find from Finland. Finally, the issue includes the first
analysis of early Roman textile tools from southern
We want to remind the readers that at the Annual
General Meeting of Friends of ATN held on the 11th
of May 2011 in Esslingen, Germany, it was decided
that from 2012 the newsletter will change its name
to Archaeological Textiles Review, so Issue 53 is the
last Archaeological Textiles Newsletter. We hope it
will be an important new step in the evolution of the
journal and raise its profile.
It was also decided that from 2012 the review will
only be published once a year, so please remember
that there will be no issue in June and the next Issue
54 will come out in November 2012 which however
will be in a much larger format, i.e. the usual two
issues will be combined into one. We will focus
more on the scientific articles, while most of the
newsletter items (calendar, reviews, bibliographies,

new publications and queries) will be posted on a
rolling basis on the Friends of ATN website, which
should ensure an even faster dissemination. We
regret having to change to one issue per year, which
will certainly disappoint many subscribers/members,
but the ever increasing workload has made the
publication of two issues per year no longer feasible.
The change should also keep the constantly rising
costs down.
In the beginning of 2012 the first two issues published
in Copenhagen, Issues 46 and 47 from 2008, will
be accessible in a full colour pdf format for free on
the ATN homepage. Back issues will continue to be
published online with a three year delay. Printed
back issues will still be available via the website.
Please make sure to renew your membership at
the beginning of 2012. The ATN/ATR is edited
and produced by volunteers, all subscription fees
go towards the essential costs of printing and
distribution. We would also like to encourage the
contributors to send their articles throughout the
year to spread the editing workload. The peer review
system which is necessary for the high standard of
ATN requires more time in the editorial process.
The absolute deadline for contributions for 2012 will
be the 1st of September 2012. We very much look
forward to your contributions as ultimately the aim
of the ATR is to disseminate new research in this
fascinating field.
The editors

Archaeological Textiles Newsletter No. 53


Francesco Meo

Rediscovering ancient
activities: textile tools in a
3rd-2nd century BC context
from Herakleia, Southern
Basilicata, Italy
The following contribution introduces preliminary
results of a PhD project in Ancient History started
in April 2010 at the University of Salento, Italy,
entitled Breeding and wool industry between III and I
century BC in Southern Italy through literary sources and
archaeological data; Herakleia, its territory and the Ionic
coast between Taranto and Sinni river.
This study had its starting point in the hypothesis
advanced by L. Giardino (2004), according to
which the archaeological data, as well as literary
and epigraphic evidence, demonstrate that sheep
husbandry in the surrounding territory (chora) of
Herakleia and wool weaving in its urban areas were
among the most important economic activities of the
site in the 2nd century BC. Herakleia was a second
generation Greek colony of Southern Italy (Magna
Graecia). It was founded in 433 BC by the Greek
colonies of Taranto and Turi along the Ionic arc on a
high ground between Agri and Sinni Rivers.
Material in this study has been recovered during the
excavations carried out by L. Giardino in the western
district of the Castle Hill of Herakleia (Figs. 1-2)
in the early 1970s. Specifically, the material comes
from blocks (insulae) I, II, IV and VI, and several
closed contexts (houses individualized inside such
insulae) have been analyzed (Fig. 3). Furthermore, the
material from Herakleia is compared with three sites
located in the chorai of Metaponto and Herakleia:
Masseria Durante, San Biagio and Bosco Andriace
(Fig. 1).

In this article, the first morphological and weight
typology of the over 500 loom weights found at
Herakleia and examined to date is presented,
followed by a more detailed investigation of finds
from two closed contexts. The greatest quantity of
loom weights examined to date (468 of 536 examples)
are of discoid circular type with few rare variations
of convex discoid circular (four examples) and
hemispherical weights (27 examples). Truncated
pyramidal loom weights are relatively few (19
examples) as are flat trapezoidal ones (16 examples);
the pinched weights are even rarer (just two
Discoid circular and hemispherical loom weights
All discoid circular loom weights (here designated
Type PD2) have two holes and can be separated
in two subtypes: Type PD2a with both faces flat,
except in those cases in which a decoration is slightly
projecting outward (Fig. 4); and Type PD2b in
which both faces are convex (Fig. 5). Type PD2a is
the commonest and the thickness, measurable for
almost all examples, ranges between 1.8 and 2.2
cm, while the weight, measurable only for a part
of the weights, ranges between 100 and 200 g; the
diameter ranges between 7 and 9 cm. Fewer than
13% (63) of Type PD2a loom weights are decorated
and over 29% of the total (154) have one or more
stamps, while only about 4% (18) of the total have
one or more inscriptions consisting of one or more

Archaeological Textiles Newsletter No. 53


Fig. 1. Southern Italy. Main settlements of the Ionic arc and the 3 sites in the territory analyzed together
with Herakleia.

Fig. 2. Herakleia: the Castle Hill with the western quarter in red in the northern part; the modern town of
Policoro in the eastern part.
Archaeological Textiles Newsletter No. 53



Fig. 3. Plan of the excavated blocks of the Western quarter of the Castle Hill of Herakleia: the two houses
presented in the article are colored in red.
Greek letters. Decoration, usually present just on the
face, and stamps are moulded, while inscriptions are
both moulded inside a rectangular stamp, or, more
frequently, incised before firing.All discoid circular
loom weights of Type PD2b have two holes and
differ from the Type PD2a weights in having convex
faces and a slightly larger thickness, between 2.7 and
2.8 cm. None of them has inscriptions, stamps or
decoration. The presence of a very limited number
of examples (4 found to date) does not allow further
considerations regarding this type of loom weight.
All hemispherical loom weights (Type PF2) also
have two holes and are a variation of Type PD2 (Fig.
6): the upper part is discoid or elliptical while the
bottom one is cut horizontally to form a flat base.
Also in this case, their limited number allows only a
partial analysis of the data and it is possible only to
establish a range of thickness between 2 and 2.6 cm.
The common characteristic of all Type PF2 weights
is to have at least one decorated face with moulded
and embossed motifs; in five cases the decoration is
present on both faces.
Truncated pyramidal and flat trapezoidal loom
The situation is much more complex in the case
of truncated pyramidal loom weights. Despite
the limited number of examples, three different
types are present: Eight truncated pyramidal loom
weights with one hole (Type PTP1); and 11 truncated

pyramidal loom weights with two holes (Type PTP2),
which are further subdivided into Type PTP2a
weighing up to 300 g (7 examples) and Type PTP2b
weighing between c. 500 and c. 600 g (4 examples).
Of the truncated pyramidal loom weights analyzed
to date, four have one or more stamps (two of Type
PTP1 and two of Type PTP2a); one of Type PTP2a has
a single incised letter, while no weight is decorated.
All flat trapezoidal loom weights have two holes and
are a version of the truncated pyramidal ones (Fig. 7).
Their number is limited but they fall into two types:
Type PTR2a weighing up to 300 g (two examples);
Type PTR2b weighing between c. 750 g and c. 1.7 kg
(14 examples). Of the flat pyramidal loom weights
analyzed to date, four have one or more stamps (two
of Type PTR2a and two of Type PTR2b) and one of
Type PTR2b has a single incised letter.
Pinched loom weights
The last type of loom weight at Herakleia is the
pinched weight (Fig. 8); these have an ovoid shape
with convex faces, central hollow in the middle of
both faces and pinched laterally in the median part.
Of the two recovered examples one has one hole
(Type PP1) and one has two holes (Type PP2).
House A of Block II
Following this general morphological typology of
loom weights used in Herakleia, interesting data was
obtained from two closed contexts in which loom

Archaeological Textiles Newsletter No. 53


Fig. 4. Examples of Type PD2a weights (Photo: F. Meo).

Fig. 5. Example of Type PD2b weight
(Photo: F. Meo).

Fig. 6. Example of Type PF2 weight
(Photo: F. Meo).

Archaeological Textiles Newsletter No. 53



Fig. 7. Example of Type PTR2b weight
(Photo: F. Meo).

Fig. 8. Example of Type PP) pinched weight
(Photo: F. Meo).

weights have been found. Analysis of the spatial
distribution of weights was carried out for these
contexts room by room, excluding the surface finds.
Since the excavation was not carried out using locus
(unità stratigrafica) system, more detailed spatial
distribution analysis is not possible.
The first context is House ‘A’ of Block II (Fig. 3): this
is a ‘courtyard house’ built at the beginning of the 3rd
century BC, and continuously occupied throughout
the entire 2nd century BC, with traces of occupation
in the early imperial times (Giardino 1996, 142-150,
fig. 7; 1998, 177-183; De Siena and Giardino 2001, 144145; Giardino 2004, 402-402, fig. 7). Inside the house,
a total of 158 loom weights have been discovered: the
vast majority are of Type PD2a (144 examples) while
Types PD2b (one example) and PF2 (seven examples)
are rare; the number of truncated pyramidal (three
of Type PTP2a) and flat trapezoidal (one of Type
PTR2a and two of Type PTR2b) loom weights is very
limited. It is possible to identify two main areas in
which loom weights have accumulated: Room 8 and
the courtyard (Room 3c), for which there is a notable
dispersion of material from West toward East (Fig. 9).
Morphological analysis of Type PD2a weights shows
a presence of two distinct groups of objects: the first
group of weights present in Room 8 (23 in total)
has thicknesses between 1.5 and 2 cm and weigh
Fig. 9. Distribution of loom weights from the first cut.
between 105 and 130 g; the second group, recovered

Archaeological Textiles Newsletter No. 53


Fig. 10. Herakleia, Block II, House ‘A’: thicknesses and weights of the two groups of discoid
circular loom weights.
in the courtyard (38 in total), consists of weights
with thicknesses between 1.8 and 2.2 cm and weigh
between 135 and 175 g (Fig. 10). As warp quality and
thickness of a textile depend on weight and thickness
of loom weights, these two groups are likely to
represent the remains of two looms with which two
different warps have been worked: the first one in
Room 8 would have been used to weave finer fabrics
in comparison to those produced using the second
group of weights in Courtyard, Room 3c.
From the latter area 2 Type PTR2b weights were
also recovered, weighing over 700 g and over 1.4 kg
respectively; unfortunately the fragmentary state of
both pieces does not allow establishing their exact
The Courtyard also yielded a spool (Fig. 11) with
concave body and flat extremities (Type B1 in Gleba
2008, 143-144, fig. 99), the only such object recovered
to date among all the investigated areas. It is about
4.5 cm long and is 3.9 cm wide at the two extremities
and 2.1 cm wide at the centre of the body.
Peristyle house of Block VI
The second closed context examined here is a
peristyle house of Block VI (Fig. 3), built at the
beginning of the 2nd century BC and in use between
the 2nd and the 1st century BC (Giardino 1996, 152156, fig. 14; 1998, 177-183). Inside this house, 126

Fig. 11. Spool from House ‘A’ - Block II.
(Drawing: C. Bianco).

Archaeological Textiles Newsletter No. 53


possible to hypothesize the presence of two looms
and in the case of the peristyle group it is likely that
they were used to produce a fabric different from that
produced in House ‘A’.
The Type PTR2b weights come from the same areas
where the two groups of Type PD2a weights have
been found (except the two examples from the room
seven recovered from the less reliable surface cuts):
two from Room 36 and one from Room 1, near the
peristyle. The weights of Room 36 weigh 700 g and
1.3 kg respectively; the weight from the Room 1
weighs 700 g; the weights from Room 7 are 1 and 1.3

Fig. 12. Distribution of loom weights from the
first cut.
loom weights have been found. In this case, too, the
majority of loom weights are of Type PD2a (114 in
total, with two more examples of Type PF2), while
the truncated pyramidal and flat trapezoidal loom
weights are only eight (of which three of Type
PTP1 and five of Type PTR2b), and two are pinched
weights. The spatial distribution of the weights inside
each room (Fig. 12) also indicates presence of two
separate groups: the first one in Room 36 and the
second near the peristyle (in Rooms 1, 2, 3 and 4).
The two groups of weights of Type PD2a have
meaningful characteristics (Fig. 13). The weights of
the more homogeneous group (of 33 examples) found
in Room 36 have a thickness mostly between 1.9 and
2.2 cm, while it is not possible to identify a precise
range for their weight. The weights of the group
recovered around the peristyle (55 in total) have
instead a wider range of thickness, between 1.8 and
2.3 cm, but a well defined and narrow weight range,
between 200 and 220 g. Hence, in this case, too, it is

In addition to loom weights, a pair of iron shears was
found in Room 2, near the peristyle (Fig. 14). The
shape is common to all shears datable between the
5th- 4th and the 1st century BC recovered in Italy:
two triangular blades (knives) with rectangular tank
attached to a simple U-shaped spring (Gleba 2008,
93-97). The blades are about 11 cm long while the
total length of the shears is about 20 cm. This is the
only example of shears found in Herakleia, and the
only one along the entire Ionic arc which comes from
a house, i.e. settlement context. Two other examples
published to date come from female burials: grave T
76 of Pantanello, in the chora of Metaponto, datable
hypothetically between the end of the 5th and the
beginning of the 4th century BC (Carter 1998, 266 no.
51, 817-818 H8s, Fig. 20.19); and grave 6 of Taranto,
dated to the second half of the 4th century BC (De
Juliis 1984, 407 XXXIXs no.4).
Based on the data presented above a series of
preliminary conclusions can be drawn:
 In terms of chronology the examined contexts
attest the presence of discoid circular loom
weights in levels dated to the 3rd-2nd
centuries BC and their absence of contexts
dated to the second half of the 4th century
BC in the investigated area. This is in contrast
to the generic dating of 4th- 3rd century
BC given for this type in the archaeological

Contextual and spatial analysis of the various
types of loom weights suggests that Type
PD2a weights (with Type PD2b and PF2)
replace in Herakleia the truncated pyramidal
loom weights. Furthermore, the detailed
investigations of the two houses indicate
presence of at least four looms.

Archaeological Textiles Newsletter No. 53


Fig. 13. Herakleia, Block VI, Peristyle House: thicknesses and weights of the two groups of discoid circular
loom weights.

Fig. 14. Iron shears from Peristyle House - Block VI
(Drawing: C. Bianco).


The contextualization of loom weights inside
the houses and, more specifically, inside
specific rooms allows also to clarify the
organization of wool working: the presence
of two looms for each house seems to show
that it took place inside the houses and not
in special workshops. It is further possible
to hypothesize that wool working was not
only a simple domestic activity but a real
economic activity. Such considerations need
further research.


Type PTR2b weights to date have been
usually recovered in pairs in the various
rooms; such pairs always consist of pieces
with different weights. This situation is
common in case of the Type PTR2b and Type
PTP2b weights found outside the houses,
too. The presence of paired weights and
their very limited number suggest that it
is unlikely that they formed part of warpweighted looms. They could have been part

Archaeological Textiles Newsletter No. 53


of a pit treadle loom (Broudy 1970, 109-111),
where just one or two heavy weights were
necessary for stretching the warp. But, again,
this kind of loom was used in chronologically
and geographically distant contexts and
its presence has yet to be demonstrated in
pre-Roman Italy. At present it is not yet
possible to understand their use and further
investigation of other contexts is needed.


As far as the iron shears are concerned,
their recovery in the room of a house
where a loom has been identified and the
length of the blades which are too short to
be used for shearing animals (Wild 1970,
22; Alfaro Giner 1984, 41; Gleba 2008, 93)
suggests a connection to the weaving
process rather than to fibre procurement.
However, comparison of the dimensions
of the shears recovered in southern Italy
with the dimensions of those recovered in
northern Italy (Gleba 2008, 96-97, Tab. 3a3b) may indicate either that in the southern
regions a smaller model and with relatively
shorter blades was used, or that they were
used for another purpose (e.g. as a toiletry
A note, finally, about the spool: the presence
of a single example and its weight (around
50 g), much lighter in comparison to the
range for the loom weights of the courtyard
of House ‘A’ of block II, allows me to
hypothesize that in this case it was used as a
spool rather than a small weight (MÃ¥rtensson
et al. 2007b; Gleba 2008, 140).

Preliminary results of the analysis of the materials
presented here confirm that textile production can be
considered as one of the main economic activities in
Herakleia and in its territory in the 3rd-2nd centuries
BC. These conclusions allow to pose a series of wider
questions the doctoral research aims to examine:
- was sheep husbandry in Herakleia territory
and in general along the Ionic arc transhumant,
sedentary, or both?
- whether discoid circular and all the other types
of loom weights were used during weaving for
producing different products (e.g. as demonstrated
for other types of loom weights by MÃ¥rtensson et
al. 2007a; MÃ¥rtensson et al. 2009) and were in use in
the same historical period?
- whether textile production activities were carried
out only on household level or if there is evidence

for workshop production?
- whether the presence of such production can
be connected with the signing of the uniquely
favourable treaty of alliance with Rome (prope
singulare foedus) with Rome at the beginning of the
3rd century BC (Cicero, Pro Balbo, 22, 50; Sartori
1967, 81-88; Lombardo 1996, 25; Giardino 2003,
182)? Even though we lack the text of this treaty
of alliance, it had to be on such favorable terms
for Herakleia that, when in 89 BC the new law
Lex Plautia Papiria conferred to its inhabitants the
right to become Roman citizens, they hesitated
long because of the better conditions of the original
- was wool production at Taranto connected with
Herakleia, the colony of Taranto (Morel 1975, 293300; 1978)?
Textile production is an activity which leaves few
traces in the archaeological and documentary records
of southern Italy (for a preliminary analysis of the
historical sources see Meo in press). I hope to answer
these questions through systematic studies of loom
weights and other textile tools from south Italian sites
and to demonstrate textile production as a key part of
economic activity in this area.
I would like to thank Prof. L. Giardino for her
fundamental support during all the research for this
project; Dr. A. De Siena for facilitating the study of
materials in the capacities of Director of Metaponto
Museum and Soprintendente per i Beni Archeologici
della Basilicata; Dr. M. Gleba for her courtesy
and availability afforded to me since the ‘Roman
meeting’; Prof. M.-L. Nosch for the wonderful
welcome at the CTR; Dr. E. Andersson Strand for her
important suggestions in the research development;
Dr. C. Bianco for her patient help with drawings
and images; Prof. G. Tagliamonte for his continuous
important inputs; Dr. A. Quercia for our discussions
about loom weights; Prof. M. Lombardo and Prof. F.
Frisone for their constant supervising activity in the
PhD Course during the last few centuries BC.
Alfaro Giner, C. (1984) Tejido y cestería en la Penísula
Ibérica. Bibliotheca Praeistorica Hispana XXI. Madrid.
Broudy, E. (1979) The Book of Looms. A history of the
handloom from ancient times to the present. New York.
Carter, J. C. (1998) The Chora of Metaponto. The
Necropoleis. Austin.

Archaeological Textiles Newsletter No. 53

De Juliis, E. (1984) Gli ori di Taranto in Età Ellenistica.
De Siena, A. and L. Giardino (2001) Trasformazioni
delle aree urbane e del paesaggio agrario in età
Romana nella Basilicata sudorientale. In: E. Lo Cascio
and A. Storchi Marino (eds), Modalità insediative e
strutture agrarie nell’Italia meridionale in età Romana,
129-167. Bari.
Giardino, L. (1996) Architettura domestica a Herakleia.
Considerazioni preliminari. In: F. D’Andria and K.
Mannino (eds), Ricerche sulla casa in Magna Grecia e in
Sicilia, 133-159. Galatina.
Giardino, L. (1998) Aspetti e problemi dell’urbanistica
di Herakleia. In: Siritide e Metapontino. Storie di due
territori coloniali. Atti dell’incontro di studio. Policoro,
Italy, 31 ottobre-2 novembre 1991, 171-220. NaplesPaestum.
Giardino, L. (2003) Gli insediamenti alla foce del
Sinni in rapporto alle attività portuali delle colonie di
Siris e di Herakleia. In: L. Quilici and S. Quilici Gigli
(eds), Carta archeologica della Valle del Sinni, X suppl., 1,
181-206. Roma.
Giardino, L. (2004) Herakleia e Metaponto: dalla polis
italiota all’abitato protoimperiale. In: Atti del XLIII
Convegno di Studi sulla Magna Grecia. Taranto, Italy 2428 settembre 2004, 387-432, XXXV-XLIV. Taranto.
Gleba, M. (2008) Textile production in pre-Roman Italy.
Lombardo, M. (1996) Greci, Enotri e Lucani nella
Basilicata meridionale (VIII-III sec. a.C.): problemi
storici. In: I Greci in Occidente. Catalogo della Mostra di
Policoro, 9-27. Napoli.

MÃ¥rtensson, L., E. Andersson, M.-L. Nosch and
A. Batzer (2007a) Technical Report, Experimental
Archaeology, Part 3 Loom weights, 2007. Tools and
Textiles – Texts and Contexts Research Program. The
Danish National Research Foundation’s Centre for
Textile Research, University of Copenhagen. ctr.hum.
MÃ¥rtensson, L., E. Andersson, M.-L. Nosch and
A. Batzer (2007b) Technical Report, Experimental
Archaeology, Part 4 Spools, 2007. Tools and Textiles –
Texts and Contexts Research Program. The Danish
National Research Foundation’s Centre for Textile
Research, University of Copenhagen.
MÃ¥rtensson, L., M.-L. Nosch and E. Andersson
Strand (2009) Shape of things: understanding a
loom weight. In: Oxford Journal of Archaeology 28(4),
Meo, F. (in press) Attestazioni archeologiche di
attività laniera a Herakleia tra III e II secolo a.C. In
“Siris-Herakleia” nuove ricerche, nuove scoperte. Atti della
giornata di studio, Matera, 9 luglio 2011.
Morel, J.-P. (1975) Aspects de l’artisanat dans la
Grande Grèce Romaine. In: Atti del XV Convegno di
Studi sulla Magna Grecia. Taranto, Italy 5-10 ottobre
1975, 263-324. Taranto.
Morel, J.-P. (1978) La laine de Tarente (De l’usage des
textes anciens en histoire économique). In: Ktèma 3,
93-110. Strasbourg.
Sartori, F. (1967) Eraclea di Lucania: profilo storico.
In: B. Neutsch, (ed), Archäologische Forschungen in
Lucanien, II, Herakleiastudien, 16-95. Heindelberg.
Wild, J. P. (1970) Textile Manufacture in the Northern
Roman Provinces. Cambridge.

Corresponding author:

Archaeological Textiles Newsletter No. 53


Sanna Lipkin and Anna-Kaisa Salmi

A 16th century silk fabric
fragment from northern
A small piece of silk was found in an archaeological
excavation at a farm site in Liminka, northern
Finland (Fig. 1). Only rarely textiles are encountered
in archaeological excavations on settlement sites
in northern Finland, and this is the first piece of
silk found to date. In the 16th century, Liminka
was a relatively densely populated village with
some 80 farms situated on the mouth of the River
Liminganjoki on the shore of the Liminka Bay
(Luukko 1954, 338–339). The excavation was
conducted on a 16th century farm site on the bank
of the River Liminganjoki. During the excavation,
cultural layers associated with a wood-encased
stone foundation of a hearth were investigated. The
hearth had been a part of a building measuring c. 3
x 6 m or more. The finds included the silk fragment,
animal bones and a few pieces of red earthenware;
they originate from the earth mixed with the hearth
A small piece of silk fabric gives us a glimpse of
the textiles used by the dwellers at Liminka in the
16th century. This time period is considered as the
late Middle Age in Finland. During this time the
settlements were small and rural in nature. Only in
the following century the Swedish Crown established
towns in northern Ostrobothnia. The towns, such as
Oulu (established in 1605), soon became local centres
of trade and inhabitants of the neighbouring areas
moved into those. Previously Oulu had been a part
of the parish of Liminka. Liminka is situated less
than 30 kilometers from Oulu and in the 17th century
Oulu had gained the position Liminka had in the
preceding century.

The find consists of fine silk threads going back
and forth in one system only, and at first sight there
does not seem to be any regularity. When looked at
closer, it appears that silk threads once formed one
system of a woven textile, and the other system has
disintegrated. It was probably of plant fibre (flax?),
which survives poorly in Finnish acidic soil. For
example in the church yard of the Oulu Cathedral
only one piece of linen textile has survived. The
others are of silk or wool. The silk threads have been
loosely twisted, and it is difficult to differentiate
threads from the bunch of fibres. The textile is partly
encrusted with sand, making technical analysis
difficult. Thus, it is difficult to calculate the thread
density, but the general appearance is that it has not
been very dense (perhaps around 14 to 16 threads/
cm) compared to to the fibres (c. 0.05 mm) and thread
(c. 0.2 mm) thickness.
Using linen or wool together with silk makes the
fabric less expensive (Rothstein 2003, 528). The textile
fragment has two cut edges forming a triangle. It is
quite difficult to say what kind of textile it belonged
to. It is possible that it was a refuse piece from a
fabric that was cut and sewn into a cloth. However,
textiles like this was most likely recycled whenever
possible. The textile’s largest dimensions are 6.2 x
5.4 cm, which means that it is quite small to be used
for any other purpose than decoration. Perhaps it
was meant to be a leaf or another small part of a
Some pieces of silk fabric have survived in burial
contexts in Oulu which date to the 17th and 18th

Archaeological Textiles Newsletter No. 53


Fig. 1. Silk textile fragment from Liminka, Finland, 16th century AD (Photo: Authors).

centuries. These burials have been excavated in the
church yard of the modern Oulu Cathedral (built in
1711–1797, largely renovated after the fire of 1822),
where the previous church (build in early 1610’s)
made of wooden logs was located. Some of the silk
fragments have been found in chamber burials that
belonged to the families of higher social ranks within
the town. In single coffin burials silk is often used
in caps; flowers and leaves in headdresses made of
bronze; belts placed around the waists of children;
socks; and once in a separate man’s collar. Here, silk
is definitely a marker of high social status. While
some of the burials also have remains of needles that
were used to attach the funeral dress and fragments
of wool textiles. Most of the graves have no other
remains than nails. Textiles along with other small
items that were used to attach different costume
parts, such as buttons or fasteners, play a significant
part in the display of the social identities of the
deceased in the grave yard.
Such conclusions are more difficult to make based on
the settlement materials. But the silk textile gives an

opportunity to reflect upon the relationship of social
identity and costume in the 16th century Liminka.
Silk was an expensive import material in the 16th
century Finland, and it was certainly not available for
everyone (Pylkkänen 1956, 72–86). Riitta Pylkkänen
(1956, 72–86) estimates that only small quantities of
silk fabrics were imported to Finland in the 16th and
17th centuries and that most of these fabrics were
ordered for the needs of the court and the aristocracy.
Most of the silks were imported from the European
Hansa towns to the towns of Turku and Viipuri in
southern Finland. Carelian tradesmen also imported
some silk fabrics to northern Finland from the east.
The latter silk imports may have been Persian in
origin (Pylkkänen 1956, 76). The aristocracy and
burghers used silk in many of their clothes, for
instance dresses, vests and socks (Pylkkänen 1956).
It is also known that the wealthy farmer-tradesmen
of Ostrobothnia (where Liminka is located) dressed
handsomely in the manner of the aristocracy
(Pylkkänen 1956, 358), perhaps using silk as well. Silk
was also used in folk costume, especially for scarves,

Archaeological Textiles Newsletter No. 53


bonnets, caps and various decorations (Lehtinen and
Sihvo 1984).
The piece of silk fabric found at the Liminka farm
may have been a part of a wealthy farmer’s or his
wife’s costume, but it is not necessarily linked to high
status as we know that silk was used for decorating
and accessorizing the folk costume as well. Once
out of fashion, silk fabrics might have been given
or sold to poorer people, or made into children’s
clothing (Rothstein 2003, 529). The small fragment
of silk fabric from Liminka may have been a re-used
piece of a worn-out garment, or a refuse piece used
for decorating the costume. Re-using and recycling
materials was a common practice in northern Finland
at that time, not necessarily connected to poverty or
low status but practiced by everybody (Nurmi 2011).

Lehtinen, I. and Sihvo, P. (1984) Rahwaan
puku. Näkökulmia Suomen kansallismuseon
kansanpukukokoelmiin. Helsinki.
Luukko, A. (1954) Pohjois-Pohjanmaan ja Lapin historia
II. Keskiaika ja 1500-luku. Oulu.
Nurmi, R. (2011) Development of the Urban Mind. The
Case Study of the Town of Tornio, Northern Finland.
Pylkkänen, R. (1956) Säätyläispuku Suomessa
vanhemmalla Vaasa-ajalla 1550–1620. Helsinki.
Rothstein, N. (2003) Silk in the early Modern period,
c. 1500–1780. In D. Jenkings (ed.), The Cambridge
History of Western Textiles I, 528–561. Cambridge.

Corresponding author:


Archaeological Textiles Newsletter No. 53

Rahil Alipour, Margarita Gleba and Thilo Rehren

Textile templates for
ceramic crucibles in early
Islamic Akhsiket, Uzbekistan
The FerghanaValley, located in eastern Uzbekistan
was an important area of textile production since the
beginning of the Common Era. Located on the Silk
Road, this area was likely a crossroads of Central
Asia, which absorbed eastern and western influences.
The finds of archaeological textiles in Uzbekistan
are, however, exceedingly rare. Some of the earliest
textiles have been excavated at the Bronze Age site of
Sapallitepa in southern Uzbekistan, dated to the 17th14th centuries BC (Askarov 1977, 173-174). Numerous
silk fragments were found in the Karabulak cemetery
in southern Ferghana and date to the last centuries
BC and 1st-2nd centuries AD (Litvinskiy 1972, 133136). Recently a large number of textiles from the
Munchaktepa cemetery near Pap dated to the 5th8th centuries AD have been published (Matbabaev
and Zhao 2010). The vast majority of the surviving
textiles are silks but cotton and wool textiles were
also present. On the basis of this material, Matbabaev
and Zhao (2010, 227) suggest that silk production
in the Fergana Valley was already developed at the
beginning of the Common Era under the influence of
China. Looking at the wider region of Central Asia,
most textile scholarship has focused on the often
spectacular patterned silk finds (e.g. Schorta 2006
with extensive bibliographies), while little is known
about the more mundane and utilitarian textiles
made of other materials1.
An investigation of a large number of crucibles
excavated at Akhsiket, a city in the Ferghana Valley
of eastern Uzbekistan dated to the 9th-12th century
AD, revealed numerous textile impressions. Olga
Papakhristu (1985; 1993) was the first scholar to
discover that these crucibles must have been made
by means of a textile mould. This paper follows up

on her research by looking more closely at the textile
impressions left on the crucible fabric, advancing our
understanding of a highly standardized, industrial
scale manufacture of crucible steel. It further brings
into focus textiles used for utilitarian purposes in a
region where few textiles have been found.
Textile impressions
Impressions of textiles and basketry have long been
investigated for the information they may provide
about the perishable materials that left them.
Impressions are negatives of the original fabrics,
created, for example, when the fabric comes in
contact with clay objects or surfaces before they are
fired. Some of the earliest evidence for woven fabric,
in fact, survives in the shape of such imprints, as in
the case of the Upper Paleolithic Czech site of Pavlov
I (Adovasio et al. 1996; Soffer et al. 2000). During the
Neolithic period, several ceramic traditions existed
which used cords, baskets and textiles for pottery
decoration and entire cultures have been named after
this custom, for example the Corded Ware Culture.
Recently, results of some archaeological experiments
were published, demonstrating the importance of
such impressions for our understanding of the past
fibre technologies (Grömer and Kern 2010). To our
knowledge, textile impressions on metallurgical
crucibles have never been investigated before. They
open a new field of inquiry into the use of textiles
and the economy of their production in the past in
contexts where textiles themselves no longer survive.
Methodology and materials
More than 150 crucible fragments from Akhsiket
were investigated in this study, and 30 samples
with fabric impressions were selected for further

Archaeological Textiles Newsletter No. 53


analysis. Impressions were easily identifiable with
a naked eye but could be observed more clearly
under a magnifying glass with a raking light. The
textile fabric impressions were photographed using
a Dino-Lite USB digital microscope. In an attempt to
identify the fibre and further microscopic features
of the textiles, Vinyl Polysiloxane casts of selected
impressions were made for SEM analysis. To our
knowledge, this method has not been used before
for the study of textile impressions on ceramics.
However, unlike other types of textile impressions
which in some cases preserve the microstructure
of the fibres (Good 2001, 215; Grömer and Kern
2010), the crucible surface vitrified (Fig. 1) when
they were heated to temperatures reaching 14001500 ºC (Rehren and Papakhristu 2002, 70), thereby
obliterating any more detailed information left by the
Akhsiket is located in eastern Uzbekistan, Central
Asia (Fig. 2). During the early Islamic period and
until the Mongol invasion in the early 13th century
AD, it was the political and industrial capital of the
Ferghana Valley (Papakhristu and Rehren 2002, 69;
Rehren and Papakhristu 2000, 56; 2003, 395). During
the Samanid dynasty (AD 892-998), Ferghana became
a trade market for steel weaponry and armour

(Papachristu 1985, 123). The city structure of Akhsiket
is typical for the Sassanid era, comprising three
main parts: Quhandizh or Ark, a fortification, which
is surrounded by the main city or Shahristan, and
then the suburb areas or Rabat which included the
industrial quarters. The industrial scale of crucible
steel production in Akhsiket is demonstrated by
thousands of crucible fragments excavated at the site,
dated from the 9th to 12th century AD (Papachristu
and Swertschkow 1993; Papakhristu and Rehren
2002, 69; Rehren and Papachristou 2003, 395). The
crucible fragments included in this study come from
both stratified and surface contexts. The stratified
contexts, however, do not provide a firm chronology,
therefore we are unable to discuss the chronological
variations of the crucibles.
Crucible steel and crucibles
The manufacture of crucible steel is a highly
specialised process during which iron is transformed
into steel. During this process iron is carburized
with a variety of organic materials, e.g. wood or
charcoal, all of which are placed in a crucible which
is then sealed with a lid and heated to very high
temperatures in a furnace (Srinivasan 1994, 56; Juleff
1998, 11; Rehren 2000; Craddock 2003, 242; Rehren
2003, 210). While the metal within the crucible
absorbs carbon, the melting point of the alloy

Fig. 1. SEM image showing the vitrification of the inner surface of the
crucible fragment No. 1 (Photo: Authors).

Archaeological Textiles Newsletter No. 53


Fig. 2. Satellite view of Uzbekistan and Ferghana Valley
(Wikipedia 2010).
decreases and it liquefies, allowing the slag and alloy
to separate so that a very clean steel ingot solidifies in
the crucible (Rehren 2003, 210-211).
The Akhsiket crucibles are tall (28-30 cm) and thin
(8 cm diameter) cylindrical vessels with lids (Fig.
3). Their specific characteristics include use of a
particular type of clay and the mould-made shape
which is atypical for the domestic pottery, the
presence of corrugations on the outer surface, and
most importantly, textile impressions on the inner
and sometimes the outer surface of the crucible.
Papachristu (1993) was the first to propose that the
crucibles had been made using textile moulds, based
on the presence of textile impressions on the inner
surface of the crucible walls. The suggested process
of making the crucibles started with a textile mould
which was most likely filled with sand to give it a
cylindrical shape. The crucible was then built around
the mould. Once the clay was leather-hard, the sand
inside the mould was discarded and the flexible
textile mould was retrieved for reuse, leaving the
textile impressions on the clay. Subsequently the
crucible may have been fired before being used. Thus,
the textile impressions on the inner surface of the
Akhsiket crucibles are negatives of the original textile
moulds impressed onto the clay used to make the
crucibles and as such are a result of ‘indirect textile
trace formation’ (Good 2001, 215).
The textile impressions are mostly found on the
fragments of upper parts of the crucibles, but rarely
if ever in the lower parts. This is due to the fact that
during the carburization process the liquid steel and
the slag were present in the bottom two-thirds of the
crucible, destroying these delicate impressions. Thus,
only the upper one-third section of the crucible which
was not in contact with the slag and liquid steel

Fig. 3. Reconstruction of the Ferghana
crucible from Akhsiket:
1: Fine lid
2: The upper crucible wall with
remains of textile impression on the
inner surface
3: Slag
4: Crucible steel ingot
5: Outer surface of the crucible
showing corrugations and textile

Archaeological Textiles Newsletter No. 53



Table 1. Inner Textile Impressions of the Akhsyket crucible fragments.
The thread counts are relative, since we do not know the percentage of
clay shrinkage.
retained the impressions. These impressions provide
much information about the nature of the textiles
used to make the moulds for the crucibles, and it is
this information that we present here.
Analysis and results
Inner surface impressions
Twenty-four crucible fragments with impressions
on their inner surface were available for study. All
the impressions are from relatively balanced tabby
weaves (Table 1). Since none of the impressions
preserve evidence of the edges, the thread system
perpendicular to the height dimension of the crucible
was designated System 1. The thread system parallel

to it, which also had a more distinct and regular
visual appearance and was parallel to the seam
(where present), was designated System 2. The thread
counts range between 8 and 20 threads/cm, with
approximately the same number of threads in each
system (Table 1). The average thread count for System
1 is 13.5 threads/cm (standard deviation 2.8), and for
the System 2 it is 13 threads/cm (standard deviation
2.6). The majority of the samples have thread counts
of 12-16 threads/cm (Fig. 4).The thread counts are
relative, since we do not know the percentage of clay
shrinkage. It was not possible to measure the yarn
diameters accurately.

Archaeological Textiles Newsletter No. 53

Fig. 4. Thread counts of the inner textile
impressions of the crucibles in both
systems. Each diamond corresponds to
a single textile impression. Note that
several points on the graph (diamonds
combined with squares) correspond to
two or more textiles indicating that 12
threads/cm is the commonest thread








































Table 2. Outer textile impressions of the Akhsiket
crucible fragments. The thread counts are relative, since we do not know the percentage of clay

Outer surface impressions
A closer examination of the crucible fragments
revealed the hitherto unknown presence of textile
fabric impressions also on some of the outer surfaces
of 10 fragments (some of which also had inner surface
impressions; see Table 2). In the majority of the cases
a fuel ash glaze has obliterated the impressions, but
on some of the crucibles which had light or no fuel
ash glaze, the impressions survived. Due to the faint
nature of these impressions it was only possible to
measure the thread count in System 1, which ranges
from 10 to 20 threads/cm, with an average of 15.2
threads/cm. This is slightly higher than in the case of
inner surface impressions. Outer surface impressions
appeared finer visually as well.
Distinctive weave features
Some of the textile impressions have distinctive
features such as unusual weave elements, seams or
stitches. One of the impressions is unusual due to
the coarseness of the fabric (No. 16 inner side, Fig. 5).
The system (System 2) parallel to the seam is more
pronounced and has only 8 threads/cm, with yarn
diameter measuring over 1 mm; the other system
has 10 threads/cm; the yarn diameter could not
be measured but should be just under 1 mm. The

Archaeological Textiles Newsletter No. 53



Fig. 5. Crucible-wall fragment No. 16 with the coarsest textile fabric impression (Photo: Authors).

Fig. 6. Crucible-wall fragment No. 18 with a weaving fault (Photo: Authors).

Fig. 7. Left: Photo microscopy of fragment No. 4 with different textile patterns and a
possible border. Right: Photograph of the fragment (Photo: Authors).

Archaeological Textiles Newsletter No. 53

Fig. 8. Fragment No. 2 with the seam
and textile fabric impressions on the
inner side (Photo: Authors).

Fig. 9. Fragment No. 2, the outer surface
bearing textile fabric impressions
(Photo: Authors).

fragment has textile impressions on the outer surface
as well. The thread count for the outer impressions
is 16 threads/cm, which is clearly different in quality
from the inner surface impressions.
Fragment No. 18 has the same thread count of 16
threads/cm in both systems, but one system is more
pronounced. In one area the impression has an
irregularity in the more pronounced system which
may be a weaving fault or a much thicker thread (Fig.
Fragment No. 4 with textile impression on the inner
surface has two different patterns. It is likely that
one part of the impression is the textile border with
either a denser thread system or different yarns,
while the rest of the impression is the ground weave
in simple balanced tabby with 10 threads/cm in both
systems (Fig. 7). Visually, the diameter of the yarn
of the ground weave appears thicker than the yarn
of the border. The yarn of the perpendicular system,
which could not be measured and differentiated in
either part of the imprint, appears to have the same
diameter in both parts. This fragment also has textile
impressions on the outer surface, which are finer

in terms of thread diameter than those on the inner
Fragment No. 2 has distinctive impressions both on
the inner and outer surfaces. The inner textile fabric
is tabby with a seam running along the fragment.
The thread count for the system parallel to the seam
line is 14 threads/cm, while the other system has 18
threads/cm. Thus, while the yarn diameter is the
same, one of the systems is woven more densely. The
outside surface textile impression is again finer than
the inner, with an average thread count of 20 threads/
cm (Figs 8-9).
Fragment No. 17 also has textile impressions on both
sides. The textile on the inner surface has 12 and 16
threads/cm in the two systems. The outer surface
impression appears finer and has 18 threads/cm.
Furthermore, the binding might not be a tabby, as
one system is completely indiscernible (Fig. 10).
Apart from all the crucible wall fragments with
textile impressions, one lid fragment (No. 21) was
found to have textile impressions with thread count
of 18 threads/cm in both systems. It may suggest
that textile moulds were also used to manufacture a

Archaeological Textiles Newsletter No. 53


particular type of crucible lid (fine lid) (Fig. 11).
Crucible fragment No. 15 preserved impression of
a seam with regular stitching and 3 stitches/cm. It
has two parallel ridges along the stitches creating an
area lower than the surface of the other parts of the
crucible, suggesting the presence of an open seam
(Fig. 12). Another seam, with 2 stitches/cm, is present
on fragment No. 3. The surface at both sides of the
seam line is uneven and one side of the seam appears
to be raised and more pronounced than the other one,
suggesting a raised seam (Figs 13-14). Fragment No.
1 has a seam similar to an open seam but both wings
of the excess fabric were stitched to the body of the
mould with diagonal hem stitches which are visible
in the crucible impression (Fig. 15).
Textile fabric impressions on the crucibles are
evidence of the functional and technical processes
of crucible making. Their investigation allows
us to retrieve information about technology and
organization of not only the steel industry but also
textile production in the Ferghana region during the
Early Islamic period. The use of a mould is indicative
of a highly standardized craft and is a response to a
demanding, specialized and sophisticated production
The clay for making crucibles contains a high amount
of quartz temper, which would have been abrasive
on the hands of the potter. This factor and the very
narrow tubular shape of the crucible would have

made it almost impossible to produce the vessels by
hand and on a wheel. In fact, none of the crucibles
preserve any wheel marks. A solution to make such
crucibles would have been to prepare a mould of a
sand-filled textile to manufacture large quantities of
standardised crucibles. At the same time, the clay is
quite fine, so a relatively dense and smooth textile
template was required in order for the clay to be
shaped around it without filling the empty spaces
between the threads of the textile. Analysis of the
textile impressions indicates that fabrics used for
making the moulds are well woven, without any
obvious mistakes and with very even yarns. They are
all balanced tabbies in a relatively narrow range of
qualities. The thread counts of the impressions lead
to the categorization of the textiles into three qualities
(Fig. 4), also discernible visually:

Medium (8-10 threads/cm): 5 fragments (Nos
4, 11, 12, 14, 16).
Medium fine (12-16 threads/cm): 14
fragments (Nos 1, 7, 8, 9, 10, 15, 17, 19, 20, 22,
23, 24, 27, 28)
Fine (18-20 threads/cm): 5 fragments (Nos 2,
3, 5, 6, 21)

The relatively narrow range of the thread counts
in both systems might suggest that the textiles
were made or at least selected for the purpose.
Furthermore, the outer surface impressions are
generally finer than the textile impressions of the
inner surface of the crucibles, indicating that the
textiles which left traces on the outer surfaces were

Fig. 10. The outer surface of a crucible-wall fragment No. 17 with unusual
fabric impressions (Photo: Authors).

Archaeological Textiles Newsletter No. 53

Fig. 11. Lid fragment No. 21 with
textile impressions of a tabby weave
(Photo: Authors).

different from those used for the moulds.
The textiles on the inner surface of the crucibles are
all tabbies but they have subtle visual differences.
Hammarlund (2005) has classified different tabby
textiles based on their visual characteristics. Three of
these visual classes could be matched to the textile
impressions on the inner surface of the crucibles:

Tabby-‘character’ (Fig. 16): both systems are
straight and balanced (Hammarlund 2005,
92-93). The majority of textile impressions
investigated can be classified under this



Tabby-‘movable’(Fig. 17): in this visual
category one or both systems are moving
two-dimensionally and there are noticeable
spaces between the yarns (Hammarlund
2005, 94-95).
Tabby-‘flat’(Fig. 18): in this type, the binding
is not very obvious since one system is
loosely spun (Hammarlund 2005, 100-101).

The careful and regular stitching of the seams
also indicates a careful and possibly specialised
production of moulds. The mould would have
required a high quality textile with sufficient strength

Fig. 12. Left: Crucible-wall fragment No. 15 with an open seam. Centre: Micrograph of the seam. Right:
drawing of an open seam (After Morrell n.d. 15).
Archaeological Textiles Newsletter No. 53



Fig. 13. Left: Fragment No. 3 with a raised seam. Right: drawing of raised seam (After Morrell n.d., 15).

Fig. 14. Left: Micrograph of the raised seam of the fragment No. 3. Right: SEM picture of the raised seam
(Photo: Authors).

Fig. 15. Left: The crucible wall fragment No. 1 with an open seam and hem stitches on both excess wings
(After Morrell n.d., 6).

Archaeological Textiles Newsletter No. 53


Fig. 16. Micrograph of an example of tabby - character impression, No. 13 (Photo: Authors).

Fig. 17. Micrograph of a tabby - movable
impression, No. 22 (Photo: Authors).

and density to keep the sand. It is likely, thus, that
the textiles were either produced for the purpose, or
selected from new or used textiles of specific quality.
The later possibility seems more likely. Since the
workshops involved in the manufacture of crucible
steel were probably only few (2 or 3) at any one time
or period, and the moulds used to make the crucibles
were likely to be reused and lasted for some time, it
is unlikely that the textiles used to make them were
produced for the purpose. As and when textiles for
moulds were needed, buying a suitable piece of cloth
was probably more practicable than producing a
small amount of textile especially for this purpose;
however more information is needed regarding the
organisation of production in order to make further
conclusions. The differences in the thread counts of
the textiles indicate that they do not belong to the
same production, although individual preferences of
the craftspeople and chronological differences also
must be kept in mind.
The small differences in the textile qualities are
probably due to random variation in the textiles
chosen, rather than evidence for different workshops
for making textiles for crucible manufacture. These
slight deviations have no functional effect for the
moulds and are in line with fabric variability overall.
However, the differences in seams could suggest
either a chronological deviation in practice or that
each crucible-making specialist was making/stitching
each mould personally.
As noted above, due to heavy vitrification of the
crucible surfaces, fibre identification is no longer
possible. However, some possibilities may be
suggested. Akhsiket was a centre for silk production,
and one possibility is that textiles used for steel

Fig. 18. Micrograph of a tabby - flat style impression, No. 16 (Photo: Authors).
Archaeological Textiles Newsletter No. 53


crucible production were made of silk. Silk fabric
is very strong, elastic and does not tear easily – all
qualities which would have made it perfect for the
purpose. Another possible fibre is cotton. While not
as strong as silk, it is less elastic which may have
been an advantage in this case. Cotton was used in
Uzbekistan since prehistoric times and is the most
likely material used for the crucible templates. Only
very few comparisons could be found. The cotton
fabrics from Munchaktepa, dated 5th-8th century
BC and supposedly produced in the Ferghana
Valley were tabbies with 12 threads/cm in warp and
weft (Matbabaev and Zhao 2010, 217), which are
comparable to the ones presented here.
The textile impressions on the inner and outer
surfaces of crucible fragments have not been analysed
before. This survey gives a new direction both for
textile and crucible studies. Investigating these textile
impressions is important because, even in the absence
of surviving textiles, much can be learned not only
about the textile characteristics used in a specific
industry but also about the economy, organization of
crucible production, the skills and preferences of the
craftspeople involved in crucible making.
It is noteworthy that along with metalwork, textiles
were important commercial goods for Uzbeks,
and the Ferghana Valley had very active textile
workshops during Early Islamic times, typically
located in a ‘Tim’, a bazaar, which was a centre
of a highly specialized and intense production
area of the city (Knorr and Lindahl 1975, 50). The
textile impressions on crucibles possibly provide
a glimpse of these local products. They also
document utilitarian fabrics, as opposed to the better
investigated luxury silks.
This study suggests that the fabric used in making
the moulds is relatively fine, typically a balanced
tabby weave with around 13-14 threads/cm in both
warp and weft. Systematic differences in thread
counts between inner and outer impressions suggest
a conscious selection of textile quality, matching
divergent requirements for the two. The largest
variability was seen in the stitching of the moulds,
suggesting that this was not done by a standardised
procedure or professional tailor, but probably ad hoc
by the mould makers.
Although the actual fibres did not survive, applying
the visual analysis to the impressions led to the visual
categorization and specification of textiles used in
the manufacture of moulds. The methodology used
for this study can be used in similar cases for other

The authors wish to thank Dr. Olga Papakhristu
as the first scholar to identify the textile fabric
impressions on the crucibles. We are also grateful of
Xiuzhen Janice Li for making the vinyl polysiloxane
casts of the impressions.
1. Several cotton fragments are known from the 11th
century AD settlement of Bazardara, located in the
south-eastern Pamir, in the neighbouring Tajikistan.
One simple tabby fragment has 15.5 threads/cm in
both systems; numerous others, mostly reps, vary in
thread count and patterning and represent clothing
remains (Bubnova 1985).
Adovasio, J. M., Soffer, O., and Klima, R. B. (1996)
Upper Palaeolithic fibre technology: interlaced woven
finds from Pavlov I, Czech Republic, c. 26,000 years
ago. Antiquity 70, 526-534.
Askarov, A. (1972) Sapallitepa. Tashkent.
Bayley, J. and Rehren, Th. (2007) Towards a
functional and typological classification of crucibles.
In S. La Niece, D. Hook and P. Craddock (eds),
Metals and Mines - Studies in Archaeometallurgy, 46-55.
Bosworth, C.E. (1975) The Tahirids and Saffarids. In
The Cambridge History of Iran. Vol. 4, The Period from the
Arab Invasion to the Saljuqs, 99-135. Cambridge.
Bubnova, M. A. (1985) Поселение Базардара
(Восточный Памир. XI в.) In Древности
Таджикистана [Antiquities of Tajikistan. Exhibition
Catalogue]. Dushanbe.
Craddock, P.T. (2003) Cast Iron, Fined Iron, Crucible
Steel: Liquid Iron in the Ancient World. In P.T.
Craddock and J. Lang (eds), Mining and Metal
Production through the Ages, 231-257. London.
Eastwood, G. and Walton, P. (1983) A brief guide to the
Cataloguing of Archaeological Textiles. York.
Feuerbach, A., Griffiths, D. and Merkel, J.F. (1995)
Analytical Investigation of Crucible Steel Production
at Merv, Turkmenistan. Newsletter of the Institute for
Archaeo-Metallurgical Studies 19, 12-14.

Archaeological Textiles Newsletter No. 53

Feuerbach, A.M., Griffiths, D.R. and Merkel, J.F.
(1998) An examination of crucible steel in the
manufacture of Damascus steel, including evidence
from Merv, Turkmenistan. In Th. Rehren, A.
Hauptmann and J. Muhly (eds), Metallurgica Antiqua
(=Der Anschnitt, Beiheft 8), 37-44.
Feuerbach, A.M., Griffiths, D.R., and Merkel, J.F.
(2003) Early Islamic Crucible Steel Production at
Merv, Turkmenistan. In P.T. Craddock and J. Lang
(eds), Mining and Metal Production through the Ages,
258-266. London.
Frye, R.N. (1964) The heritage of Persia. London.
Good, I. (2001) Archaeological Textiles: A Review of
Current Research. Annual Review of Anthropology 30,

Papachristou, O. and Swertschkow, L. (1993) Eisen
ausUstruschana und Tiegelstahl aus dem FerganaBecken. Der Anschnitt 45, 122-131.
Papakhristu, O. and Rehren, Th. (2002) Techniques
and technology of the manufacture of ceramic
vessels- crucibles for smelting wootz in Central
Asia. In V.Kilikoglou, A. Hein and Y. Maniatis (eds),
Modern Trends in Scientific Studies on Ancient Ceramics,
69-74. BAR International series 1011. Oxford.
Rehren, Th. and Papakhristu, O. (2000) Cutting
Edge Technology- the Ferghana Process of medieval
crucible steel smelting. Metalla 7, 55-69.

Grömer, K. and Kern, D. (2010) Technical data and
experiments on corded ware. Journal of Archaeological
Science 37, 3136-3145.

Rehren, Th. and Papachristu, O. (2003) Similar
like White and Black: a Comparison of Steelmaking Crucibles from Central Asia and the Indian
subcontinent. In Th. Stöllner et al. (eds), Man and
Mining: Mensch und Bergbau: Studies in Honour of
Gerd Weisgerber (= Der Anschnitt, Beiheft 16), 393-404.

Hammarlund, L. (2005) Handicraft Knowledge
Applied to Archaeological Textiles. The Nordic Textile
Journal, 87-119.

Ranganathan, S. and Srinivasan, Sh. (2004) India’s
Legendary Wootz steel, an advanced material of the ancient
world. Bangalore.

Henshaw, C. M. (2010) Early Islamic Ceramics and
Glazes of Akhsyket, Uzbekistan. PhD thesis. University
College London, London.

Schorta, R., ed. (2006) Central Asian Textiles and
Their Contexts in the Early Middle Ages. Riggisberger
Berichte 9. Riggisberg.

Knorr, T., and Lindahl, D. (1975) Uzbek: the textiles and
life of the nomadic and sedentary Uzbek tribes of CentralAsia. Basel.

Soffer, O., Adovasio, J. M., Illingworth, J. M.,
Amirkhanov, H. A., Praslo, N. D., and Stre, V. M.
(2000) Palaeolithic perishables made permanent.
Antiquity 74, 812-821.

Litvinskiy, B. A. (1972) Курганы и курумы Западной
Ферганы [Kurgans and kurums of Western Fergana].

Srinivasan, Sh. (1994) Wootz crucible steel: a newly
discovered production site in south India. Papers of the
Institute of Archaeology 5, 49-61.

Matbaev, B., and Zhao, F. (2010) Early Medieval
Textiles and Garments of Ferghana Valley. Shanghai.
Corresponding author:
Morrell, A. (n.d.) The ATN Guide to Structural Sewing:
Terms and Techniques.
Papakhristu, O. (1985) Black metallurgy of Northern
Fergana. Materials of archaeological investigations
at the ancient fort Akhsyket of the 9-early 13 cc. A.D.
Abstracts of the Candidate of History Dissertations,

Archaeological Textiles Newsletter No. 53


Paula Mazăre

Textiles and pottery: Insights
into Neolithic and Cooper Age
pottery manufacturing
techniques from Romania
Numerous studies have revealed the importance
of textile imprints, especially in those areas that
are lacking actual textile remains (Schlabow 1960;
Carington Smith 1977; Wild 1988; Adovasio 1977;
Barber 1991; Adovasio et al. 1996, 1997; Good 2001;
Makkay 2001; Cybulska and Maik 2007; Marian
2009). Besides the usual technical information they
offer a new perspective upon the usage of textiles in
Most imprints are located on the flat bottoms of
vessels and they usually come from mats, even
though the textile imprints were also created by
other two-dimensional textile structures (sheets of
fabric) made by linking, knotting, knotless netting,
twining, weaving etc. (Petkov 1965; Bagolini et al.
1973; Carington Smith 1977; Makkay 2001; Adovasio
and Maslowski 1988; Tringham and Stevanović
1990; Adovasio and Illingworth 2003; Ljaško et
al. 2004; Văleanu and Marian 2004; Lüning 2005;
Kaczanowska 2006; Mazăre 2008; Richter 2005, 2009;
Chmielewski 2009; Marian 2009).
In addition to imprints on the bases of the vessels,
textile imprints have also been identified on the
walls and within the walls or bases of the vessels
(Carrington-Smith 1977, 116, Pl. 90/167, 201;
Kostelníková 1985, 197, Fig. 1; Lüning 2005, 219, Fig.
374; Kaczanowska 2006, 108; Richter 2005, 143, Fig.
8, 2009, 212, Fig. 34.1; Chmilewski 2009, 228-229, Fig.
Such textile imprints located inside of three pedestal
foot vessels were recently identified on two Neolithic
and Cooper Age sites in Romania. Alongside other
similar finds, they attest the specific role and the

intentionality of textile usage in Neolithic and Cooper
Age pottery manufacturing. On this basis it was
possible to reconstruct the production stages of the
pedestal foot vessels.
The textile imprints
The three textile imprints were identified within
three fragments of pedestal foot vessels. Two of
them are twined textile imprints identified inside the
fragments of two Neolithic pedestal foot vessels (type
A) of Vinča tradition discovered at Turdaş (c. 51004700 BC, Figs 1-2). The third one is a mat imprint
situated inside the fragment of a Cooper Age painted
pedestal foot vessel (type B) found at SebeÅŸ-Valea
Janului (PetreÅŸti Culture, c. 4450-4300 BC; Fig. 3).
They have the following characteristics:
a. Diagonal S-twist weft twining (warp thread
diameter: 0.85-1.4 mm; weft thread diameter: 0.91.5 mm; warp thread count: 5-7 threads/cm; weft
thread count: 4-5 threads/cm; unspun fibres).
Negative imprint on the bottom of a pedestal foot
vessel (type A) (Fig. 1). Its positioning indicates
that the imprint was invisible when the vessel
was unbroken because it was covered by the
vessel’s foot.
b. Diagonal S-twist weft twining (warp thread
diameter: 0.6-1.4 mm; weft thread diameter: 0.71.6 mm; warp thread count: 6 threads/cm; weft
thread count: 4 threads/cm; unspun fibres). This
is the most interesting imprint. It is actually the
positive cast of a negative textile imprint located
within the base foot of a pedestal vessel (Fig. 2).
This indicates the fact that it was produced by
contact with the vessel’s pedestal, which already

Archaeological Textiles Newsletter No. 53


Fig. 1. Negative textile imprint on the base of a Vinča/Turdaş pedestal foot vessel (type A):
a. The pottery fragment. TurdaÅŸ site. National Museum of Transylvanian History, Cluj;
b. The cast of the textile imprint (© P. Mazăre).
Archaeological Textiles Newsletter No. 53



Fig. 2. Positive textile imprint inside the pedestal foot base of a Vinča/
TurdaÅŸ vessel (type A). TurdaÅŸ site. National Museum of Transylvanian
History, Cluj (© P. Mazăre).

Archaeological Textiles Newsletter No. 53



had a textile imprint on its lower part. The cast
accuracy demonstrates that the negative imprint
of the pedestal was hard enough to not deform
under the pressure exerted by attaching the base
foot to the pedestal.
Twill 2/2 plaited mat (width of the element/strand:
2.5-5.7 mm). The imprint was located inside of
a pedestal vessel (type B), in the contact area
between the base and the vessel’s pedestal (Fig.

Reconstruction of pedestal foot vessels
manufacturing stages
Studies dedicated to pottery technology and
ethnographic examples show that vessels with
complex shapes are made by combining several

components, built in sequential steps (Balfet 1984,
185; Shepard 1985, 55-56; McCurdy 2004). In some
cases, shaping of the bases or feet is a final stage,
them being added to the upper part of the vessel
after, previously being shaped and partially dried
(Balfet 1984, 181; Leeuw 1993, 245-246). Merging
different parts requires special attention, because
attaching still soft and moist clay surface to a hard
partially dried surface is problematic as the different
parts could separate during drying or firing. A
technical solution to this problem is to create
intentionally grooved or scraped surfaces for better
adhesion, allowing perfect bonding of the different
parts (Vitelli 1987, 121; McCurdy 2004, 27, Fig. 1).
Some textile imprints found in the inner part of the
ceramic fragments, more precisely in the merging

Fig. 3. Mat imprint inside the base of a PetreÅŸti pedestal foot vessel (type B). SebeÅŸ-Valea Janului site.
“Ioan Raica” Museum, Sebeş (© P. Mazăre).
Archaeological Textiles Newsletter No. 53


area of the vessels’ components indicate that this
process was known to Neolithic potters. Examples of
such pottery fragments with textile imprints are those
from Luleč, Czech Republic (Linear Pottery culture;
Kostelníková 1985, 197, Fig. 1); Hesserode, Germany
(Linear Pottery culture; Lüning 2005, 219, Fig. 374),
Kraków-Nowa Huta-Cło, site 65, Poland (Modlnica
group; Kaczanowska 2006, 108) and Zimne, Poland
(Funnel Beaker culture; Chmielewski 2009, 228-229,
Fig. 126).
Based on these examples and charcteristics in the
present imprints it was attempted to reconstruct the
type A pedestal foot vessels. Even though some of
the technical details of the actual construction remain
unknown, it can be suggested that the vessels were
built in three stages starting with the upper part (the
actual vessel itself), and ending with the pedestal foot
(Fig. 4):
Stage I - the vessel’s body was shaped on a textile
support, and so the bottom displays, not necessarily
intentionally, an impressed surface to be used in the
next stage to bond the pedestal foot;
Stage II - the pedestal was constructed and attached
after the vessel was partially dried. In its lower part,
an imprint was produced intentionally by pressing
a textile into the soft clay, so as to ensure good
connection to the base of the foot;

Stage III - the base of the foot was shaped and
attached after the pedestal had hardened.1
In a similar way the type B vessel could have been
created, except that it would have been made of only
two parts (the hollow pedestal and the actual body of
the vessel).
Depending on their location (on the bottom, on the
wall or within the vessel’s wall), textile imprints
can be interpreted differently with respect to the
method of their production, whether or not they were
intentionally created, as well as the functional role of
the structure itself that left those imprints. While the
wall and bottom imprints can be seen more or less as
an indirect consequence of the pottery manufacturing
process2 (Vitelli 1987, 119; Yiouni 1996, 61), the inner
wall imprints are considered proof of an intentional
and planned use of textiles.
Pre-Pottery Neolithic B finds from Nahal Hemar (c.
7000 BC) and Ali Kosh (7th-6th millennium BC) have
demonstrated that worked fibres and textiles were
used for making containers before the invention of
pottery (Schick 1988, Pl. XIV/1, 3; Barber 1991, 131132). It is interesting that this tradition prevailed and
was also used on pottery, as proven by inner wall
imprints on pottery fragments found in Neolithic

Fig. 4. Reconstruction of
manufacturing stages for a type
I pedestal foot vessel
(Graphics: T. Muntean).

Archaeological Textiles Newsletter No. 53

and Cooper Age sites. For instance, J. Carrington
Smith (1977, 115) demonstrated that in the case
of the imprints from Kephala, Greece cloths were
incorporated into the walls of pots as a form of
temper to strengthen the vessels.
In the case of the presented imprints, the textiles had
two possible functions. Besides serving as the support
in which case the vessels were shaped on them,
they were used deliberately to produce the grooved
adhesive surface necessary to bond the different
parts of the vessel together. As the reconstruction
of pottery manufacturing process is based on these
isolated findings, it cannot be concluded whether it
represented an exception or a widely used practice
of shaping Neolithic or Cooper Age pedestal foot
vessels. The breakage of the base foot may be an
indication that the practice of separately shaping the
pedestal and the base foot, and subsequent joining of
the two, was not very effective and could be subject
to failure. More careful studies on textile imprints
and Neolithic and Copper Age pedestal foot vessels
could provide further clues in this regard.

Adovasio, J. M., Hyland, D. C. and Soffer, O. (1997)
Textiles and cordage: a preliminary assessment.
In J. Svoboda (ed.), Pavlov I–Northwest. The Upper
Paleolithic Burial and its Settlement Context, The Dolní
Věstonice Studies 4, 403-424. Brno.

1. The assumptions regarding the stages
of producing vessels were confirmed by an
experimental study on the use of textile structures
during the manufacturing of ceramics. For more
details see Mazăre et al. in press.
2. There are two different archaeological
interpretations concerning the imprints on the vessel
bases: a) they were produced during the shaping
phase, using mats or coarse textiles as supports for
rotating and easier building of the vessels (a primitive
variant of turning devices); b) they are a passive
consequence of placing the vessels on mats or sheets
of fabrics in order to dry after they were shaped.
Another explanation regards the textile imprints
found on prehistoric vessel walls as ornamental. They
show an intentional usage of the textiles, but their
function does not strictly concern the technological
process of manufacturing pottery, but rather the
treatment of the surface of the already shaped

Balfet, H. (1984) Methods of formation and the shape
of pottery. In S. van der Leeuw and A. C. Pritchard
(eds.), The Many Dimensions of Pottery. Ceramics
in Archaeology and Anthropology, Cingula VII,
Universiteit van Amsterdam, 171-199. Amsterdam.

Adovasio, J. M. (1975-1977) The textile and basketry
impressions from Jarmo. Paleorient 3, 223-230.

Cybulska, M. and Maik, J. (2007) Archaeological
textiles – A need for new methods of analysis and
reconstruction. Fibres and textiles in Eastern Europe 15
(5-6), 185-189.

Adovasio, J. M., Klíma, B. and Soffer, O. (1996) Upper
Paleolithic fibre technology: Interlaced woven finds
from Pavlov I, Czech Republic, c. 26, 000 years ago.
Antiquity 70, 526-534.

Adovasio, J. M. and Illingworth, J. S. (2003) Basketry
and Textile Impressions (Appendix 6.1). In E.
S. Elster, C. Renfrew (eds.), Prehistoric Sitagroi.
Excavations in Northeast Greece, 1968-1970, Volume 2:
The Final Report, University of California, 252-257.
Los Angeles.
Adovasio, J. M. and Maslowski, R. F. (1988) Textile
impressions on ceramic vessels. In A. McPherron,
D. Srejovic (eds.), Divostin and the Neolithic of Central
Serbia, Ethnology Monographs 10, 345-357. Pittsburg.
Bagolini, B., Barfield, L. H. and Broglio, A. (1973)
Notizie preliminari delle richerche sull’insediamento
neolitico di Fimon-Molino Casarotto (Vicenza) (19691972). Rivista di scienze preistoriche 28 (1), 161-213.

Barber, E. J. W. (1991) Prehistoric Textiles. The
Development of Cloth in the Neolithic and Bronze Age
with Special Reference to the Aegean. Princeton.
Carington Smith, J. (1977) Cloth and mat impressions.
In J. E. Coleman (ed.) Kephala. A Late Neolithic
Settlement and Cemetery (Keos, Results of Excavations
conducted by the Unniversity of Cincinnati under the
Auspices of the American School of Classical Studies at
Athens; v. 1), American School of Classical Studies,
114-127. Princeton, N.J.
Chmilewski T. J. (2009) Po nitce do kłębka… O
przędzalnictwie i tkactwie młodszej epoki kamienia w
Europie Åšrodkowej. Warszawa.

Good, I. (2001) Archaeological textiles: A review of
current research. Annual Review of Anthropology 30,

Archaeological Textiles Newsletter No. 53


Kaczanowska, M., ed. (2006) Dziedzictwo cywilizacji
naddunajskich małopolska na przełomie epoki kamienia i
miedzi/ The Danubian Heritage: Lesser Poland at the Turn
of the Stone and Copper Ages. Kraków.
Kostelníková, M. (1985) Otisk tkaniny z mladší
doby kamenné z Lulče (okr. Vyškov)/ Ein
jungsteinzeitlicher Gewebeabdruck aus Luleč (Bez.
Vyškov). Archeologické rozhledy 37, 197-198.
Leeuw, S. van der (1993) Giving the pottery a choice.
Conceptual aspects of pottery techniques. In P.
Lemonnier (ed.), Technological Choices. Transformation
in Material Cultures since the Neolithic, 238-288.
Ljaško, S. M., Widejko, M. J and Burdo, N. B., eds.
(2004) Encyklopedija Trypil’s’koji cyvilizaciji, Tom. 2,
Lüning, J., ed. (2005) Die Bandkeramiker. Erste
Steinzeitbauern in Deutschland. Raden/Westf.
Makkay, J. (2001) Textile Impressions and Related
Finds of the Early Neolithic Körös Culture in Hungary.
Marian, C. (2009) Meşteşuguri textile în cultura
Cucuteni/ Textile Crafts in Cucuteni Culture. IaÅŸi.
Mazăre, P. (2008) Impresiuni de ţesături pe fragmente
ceramice descoperite în situl preistoric de la Limba
(jud. Alba). Apulum 45, 315–330.
Mazăre, P., Lipot, Ş. and Cădan, A. (in press)
Experimental study on the use of perishable fibre
structures in Neolithic and Eneolithic pottery. In
First Arheoinvest Congress. Interdisciplinary Research
in Archaeology, June 10th-11th 2011, IaÅŸi. British
Archaeological Reports International Series.
McCurdy, M. (2004) Building large forms. In A.
Turner (ed.), Pottery Making Techniques. A Pottery
Making Illustrated Handbook, 25-30. Westervill, The
American Ceramic Society, 26-29. Westerville, OH.
Petkov, N. (1965). Praitoricheski pletki i tukani ot
Sofijskoto pole i blizkite mu okolnosti. Arheologija 7,

Richter, É. (2009) Our thread to the past: plaited
motifs as predecessors of woven binding structures.
In E. B. Andersson Strand, M. Gleba, U. Mannering,
C. Munkholt and M. Ringgard (eds.), North European
Symposium for Archaeological Textiles X, 189-216.
Ancient Textiles Series 5. Oxford.
Schick, T. (1988) Cordage, basketry and fabrics. In
O. Bar-Yosef and D. Alon (eds.), Nahal Hemar Cave,
Atiquot, English Series 18, 1-43. Jerusalem.
Shepard, A. O. (1985). Ceramics for the Archaeologists.
Washington D.C.
Schlabow, K. (1960) Abdrücke von Textilien an
Tongefäßen der Jungsteinzeit. Jahresschrift für
mitteldeutsche Vorgeschichte 44, 51-56.
Tringham R., Stevanović M. (1990) The Nonceramic
Uses of Clay. In R. Tringham and D. Krstić (eds.),
Selevac. A Neolithic Village in Yugoslavia, 323-388. Los
Angeles, CA.
Văleanu, M.-C. and Marian, C. (2004) Amprente
umane, vegetale şi de textile pe ceramica eneolitică de
la Cucuteni–Cetăţuie. In M. Petrescu-Dâmboviţa and
C.-M. Văleanu (eds.), Cucuteni–Cetăţuie. Săpăturile
din anii 1961–1966. Monografie arheologică, Bibliotheca
Musei Antiquitatis 14, 318–327. Piatra Neamţ.
Vitelli, K. D. 1987. Greek Neolithic pottery by
experiment. In P. M. Rice (ed.), Pots and Potters.
Current Approaches in Ceramic Archaeology, Institute of
Archaeology Monograph XXIV, 113-131. Los Angeles,
Wild, J. P. (1988) Textiles in archaeology. Shire
Publications. Aylesbury, Bucks.
Yiouni, P. (1996) The Early Neolithic pottery:
Technology. In R. J. Rodden (ed.), Nea Nikomedeia I:
The Excavations of an Early Neolithic Village in Northern
Greece 1961-1964, Supplementary of the British School
at Athens 25, 55-78. Oxford and Northampton.

Corresponding author:

Richter, É. (2005) Textil- és négyzetrendszeres
fonatlenyomatok az Alföld neolitikumából/
Woven and plaited fabrics in the Neolithic of Great
Hungarian Plain. In G. LÅ‘rinczy and L. Bende (eds.),
Hétköznapok Vénuszai, 123–144. Hódmezõvásárhely.

Archaeological Textiles Newsletter No. 53

Olga Orfinskaya and Tamara Pushkina

10th century AD textiles from
female burial Ц-301 at
Gnëzdovo, Russia
The Gnëzdovo archaeological site dated to the
10th-11th century AD comprises several barrow
cemeteries and a settlement. It is situated on the bank
of the Dnieper River near the village of Gnëzdovo,
some 12 km west of the city of Smolensk, which lies
378 km southwest of Moscow. It is one of the most
important medieval sites in Russia.1 The population
of the settlement consisted mainly of Scandinavians
and Slavs. Among them there were artisans, e.g.,
smiths, jewelers, potters and bone-carvers, warriors
and merchants. Ethnic and social attribution of the
persons buried in the Gnëzdovo barrows is based on
the analysis of mortuary practices and grave goods.
In the Gnëzdovo barrows cremations prevail over
inhumations. Textile remains have been encountered
and recorded in 12 of the total of 116 excavated
inhumations. The overall majority of these finds are
minute fragments having survived owing to contact
with metal. The only exception is the discovery
of a bundle of fabrics in a burial excavated by the
Smolensk archaeological expedition of Moscow State
University in 1982. The materials of the burial were
included in a publication dealing with the three most
interesting chamber tombs of Gnëzdovo known by
the late 1980s (Avdusin and Pushkina 1989, 190-205),
while fabrics from the site were studied partially in
M.V. Fechner’s work (Fechner 1999, 8-10).
It is worthwhile to describe once more the barrow
burial Ц-301 in connection with new finds from the
cemeteries of Kiev and Pskov where chamber burials
have been found whose rites and grave goods are
similar to those of Gnëzdovo (see Mikhailov in ATN
47), and to outline the characteristics of textiles found
in it.

Barrow Ц-301 was situated on the southern boundary
of the Central Barrow Group of the Gnëzdovo site.
The barrow was a low, slightly eroded sand mound,
circular in plan, with a small hollow on the top.
The latter was taken at first for a trace of a 19thcentury excavation pit. The barrow mound covered
a rectangular pit measuring 3 Ñ… 3 Ñ… 0.8 m which was
dug after scorching the earth or burning a small fire.
Remains of a dovetailed timber structure measuring
2.3 Ñ… 2.25 m were revealed on the bottom of the pit.
An uninterrupted layer of wood decay up to 0.1
m wide was traced along the contour of the walls
to the height of 0.3 m. The structure was probably
even higher since separate areas of wood decay
were recorded in the upper layers of the pit filling.
It was apparently covered with wooden planking,
in a north–south orientation. Planks are 0.1-0.2 m
wide and up to 0.03 m thick. The better preserved
flooring of the chamber was completely uncovered. It
consisted of planks 0.15-0.25 m wide and up to 0.05 m
thick placed directly on the virgin-soil bottom of the
pit (Fig. 1). The chamber floor had been seemingly
covered with birch-bark whose remains were traced
in the northern part of the structure. Scattered beads
(50 items), a cross-shaped pendant cut of sheet
silver with poinçonné decoration, and a fragment of
a bronze equal-armed brooch in a very bad state of
preservation were lying on the floor almost at the
centre of the chamber. Southwest of the necklace
remains, near the west wall of the timber structure,
there were the handle, iron hoops and loops of a
truncated-cone-shaped wooden bucket lying on its
side, and a small wheel-thrown pot with a linear
decorative pattern. The remains of two birch-bark

Archaeological Textiles Newsletter No. 53


discs, one whole and one half, measuring 27 cm in
diameter, lay to the southeast of the jewellery. The
discs are displaced with respect to one other, and
both have small holes along their edges, as if for
attaching the cloth walls, resulting in a combinationmaterial box. The halved disc was partially covering
the fabric bundle put in the box. The upper layer
of the bundle was visually identified during the
excavation as a coarse wool fabric. Inside the bundle,
there were fragments of an oval bronze brooch
and an imprint of a small round cup, probably of
wood. Near the bundle but somewhat nearer to the
south-east corner of the burial chamber, there were
two wax candles standing vertically almost to the
height of 0.05 m over the upper planking. Judging
from a very thin ring of wood decay some 25 cm
in diameter traced around the candles, they were
originally placed in a hollowed out wooden vessel.
Nine additional candles, either intact or broken, were
found on the floor in the eastern part of the timber
structure. The majority of candles had a melted end.
The burial was identified as a female one of
Scandinavian type, with the corpse in a sitting
position, and was dated to the 970s AD (Avdusin and
Pushkina 1989, 200, 203).

- Microscopy in non-polarized light (МБС-10) with
10Ñ… to 40Ñ… lenses was used to determine the structure
of fabrics and plaited articles.
- Optic microscopy in reflected and transmitted
polarized light (ПОЛАМ-Р-212) with 100х to 400х
lenses was used to determine the nature of textile
fibres. Permanent immersion preparations in Canada
balsam were made for the purpose.
- Organic base of gilt threads was studied by
histochemical, microchemical and microscopic
methods. Investigations were carried out by O.
Lantratova, leading research fellow of the Restoration
Department of the State Historical Museum
- Textile dyes were analysed by E. Karpova at
the N.N. Vorozhtsov Novosibirsk Institute of
Organic Chemistry using high performance
liquid chromatography (HPLC), in extractions
of mixture ethanol:water:complex DTPA and
- The chemical composition of metal was defined by
means of remote X-ray-fluorescent method with the
aid of ArtTAX device (Röntgenanalysen-Technik).
The analyses were carried out by N. Eniosova at the

Fig. 1. Plan of Ц-301 burial at
Gnëzdovo – 1982:
1: candles
2: necklace with cross
3: bucket and pot
4: textile bundle
5: fragments of two birch-bark discs.

Archaeological Textiles Newsletter No. 53

Archaeology Department of the History Faculty of
Lomonosov State University (Moscow).
Investigation included the following: fragments of
Dress 1, fragments of Dress 2, fragments of silk fabric,
fragments of linen dress, fragments of sprang, and
fragments of wool fabric.
Dress 1
Dress fragments were restored by A. Elkina,
the restorer of the Russian Research Institute of
Restoration, so by the time of analysis we had to
deal with washed, smoothed out items, which were
reinforced with white thread. The drawings enabled
us to reconstruct the main outline of the dress (Fig.
2). It is a tunic-shaped waisted gown with widening
skirt and long sleeves with cuffs. A collar fragment
8 cm high has survived on its back. Its width with
sleeves is slightly over 2 m.2 The full length of the
dress cannot be determined; the maximum length of
the surviving part is 89 cm. The dress is made from a
fabric with a decorative pattern embroidered with gilt
threads. The main pattern is placed on front and back
parts above the waist while the skirt bears patterned
horizontal stripes. Similar stripes run along the
sleeves. The entire right sleeve consisting of two parts
measuring 40 and 20 cm respectively has survived.
As to the left sleeve, its surviving part measures 50
cm in length. A deep pleat is visible on the back.
Collar and cuffs were cut from undecorated parts of
fabric. A small inner fold divides the collar in two
parts in the proportion of 1:2.
While the reconstruction of the back is secure, that
of the front part is not straightforward. The major
issue is the shape of the neck. A surviving fragment
of its left piece measuring 12 cm ran along the central
axis of the article, at a distance of 12.5 cm from the
shoulder line. Fabric, slightly folded at this place,
bears traces of a seam, which is indicative of a lined
selvedge. The bottom part is fully spread and has
punctured holes along the line of fold. It enables
us to surmise that the dress had a vertical slit at
the centre going down for some 24-25 cm from
the shoulder line. It is worth noting, however, that
fragments of the garment were restored and any folds
smoothed out. It is therefore not inconceivable that
the garment could either have been worn open or
have an additional flap on the right side.
The right flap has a seam running across it at a
slight angle to the horizontal line below the sleeve
and slightly below the main decorative pattern. The
surviving part of the seam measures 2 cm, which
does not permit us to reconstruct the juncture line

with certainty.
A small fragment of the same fabric without
decorative pattern (13 Ñ… 10 cm) has a tucked selvedge
and remnants of a hidden seam, 7 cm from the edge;
it is indicative of a closed vertical edge of a lined
article. The fragment bears no decoration, so it was
situated either above or below the narrow decorative
stripe. It has a selvedge, so it cannot be a slit at the
centre of a weaving piece. Having analyzed several
alternatives, we chose the most obvious explanation:
the fragment is a part of a side vent. We hence
conclude that the gown was long, had two decorated
stripes on the skirt and side vents (Fig. 3). The
surviving parts of the hidden seam on the fragments
under consideration and on the sleeves imply that the
article was lined; the lining covered the reverse side
completely but did not survive.
The decorative pattern is distributed top to bottom
as follows: the area with the main decoration is 75
cm wide, then there is an area without decoration 33
cm wide, followed by a decorated area 18 cm wide,
after which the surviving cloth ends. The main image
is placed symmetrically along the vertical axis of the
cloth. The composition is arranged around a disc,
which can be regarded as a celestial body, either
the moon or the sun, or as a big pearl. The most
visible figure is that of a sitting griffin, a mythical
beast with the head and wings of a bird of prey
and the body of a lion. A scaled ‘stripe’ runs near
its jaws; it is probably a part of the body of another
mythical creature, a dragon. The dragon disappears
in the clouds above the griffin, and reappears on
the other side of the disc. The head of the dragon
is depicted showing its mouth, an ear, a horn and
‘hair’ behind the ears. Other parts of the head cannot
be identified with certainty. Thus, the fabric design
features the sitting griffin probably holding a paw of
the dragon flying into the clouds (Fig. 4). The griffin
has characteristic Mediterranean traits while the
dragon is typical of China. The amalgamation of the
two traditions is only natural since the adoption of
occidental ornamental motifs by Chinese silk weavers
began already in the first centuries of the Common
Era (Lubo-Lesnichenko 1994, 189-211; Liu 1996, 18).
The decorated stripe consists of five narrower bands:
two composed of figurative festoons, two narrow and
plain, and a central band bearing floral decoration
(Fig. 5).
It seems likely that the same pattern was used in
decoration of a Chinese gown of the Yuan period
(Fig. 6), where the bodice is covered with a pattern
featuring dragons down to the waist, while decorated
stripes run over the skirt and along the sleeves (Feng
and Lin 2005, 219). A stripe some 20 cm wide runs

Archaeological Textiles Newsletter No. 53



Fig. 2. Dress 1
1. Drawing of surviving fragments and reconstruction of dress cut.
2. Reconstruction of the front part of dress.
А. Fabric section with seam running across right flap.
Ð’. Fabric section along the line of central vent.
а. Tucked selvage bearing holes punctured with needle;
b. Unfolded selvage with holes running along the line continuing that of tucking of the
upper part the selvage.
(Photo: Authors).

Archaeological Textiles Newsletter No. 53


Fig. 3. Alternative reconstructions of dress: A – Two variants of front view; B – Back view.
over the shoulders of many surviving Mongolian
gowns of the 13th century AD. Usually such a
stripe bears decoration featuring a pseudo-Kufic
inscription. The original style called ‘intertwined
Kufic’ was elaborated in eastern provinces of Iran
in the 11th century AD (Folsach 1993, 45). A study
of a female gown from the Marjani collection has
shown that a cloth piece with starting and finishing
border3 was used to make the garment. The stripe
with a pseudo-Kufic inscription runs at 10 cm from
the end of the woven piece. It can be surmised that
a similar stripe on the 10th-century AD garment
under study was wider and illustrated a particular
story. In this case the wide stripe showing the dragon
and the griffin would have been placed at the end
of the weave. Therefore a horizontal seam whose
remains have survived on the left flap would have
run somewhere near the waist on the front of the
gown. With this assumption, all narrow decorative
stripes would have been spaced evenly. Then, if
there were two of them on the skirt, the gown was
approximately 1.40 m long (Fig. 7).
Structurally the cloth of the gown belongs to the
fabrics with additional stripes of gilt weft introduced
in the ground weave. Gilt threads cover the
background of the main decorative design almost
completely. The image is outlined by thin threads of
the warp-faced weave. The distribution of ground
and supplementary threads in the decorative stripe is
approximately even.

Fig. 4. Partial reconstruction of large stripe
of decorative pattern.

Archaeological Textiles Newsletter No. 53


Fig. 5. Drawing of narrow
decorated stripe.

Fig. 7. Reconstruction of
apportionment of dress cut
on fabric. Fabric width 80
cm, length 4 m.
A: front
B: back
C: front part of skirt
D: sleeves.

Fig. 6. Mongolian noblewoman
(After Silk 2005, 219).

Archaeological Textiles Newsletter No. 53

Technical analysis of the Dress 1 fabric (Fig. 8)
Warp: red silk, single, z-twisted, 0.1-0.3 mm thick.
Décochement: 2 ground warps. Thread count: 40
Weft: Ground: red silk, without twist, 0.3-0.5 mm
thick. Thickness of threads is uneven, at some places
weft threads are up to 0.8 mm thick. Supplementary:
gold thread composed of gilt strips of serous
membrane of animal intestines (thin and somewhat
translucent) Z-wrapped around a yellow silk core,
z-twisted. The diameter of the thread measures
0.3-0.5 mm, guilt strip is 0.5-0.6 mm wide. Threads
function in pairs. Proportion: 2 ground threads/a pair
of supplementary weft threads. Pick: 1 ground weft,
a pair of supplementary weft threads, 1 ground weft.
Thread count: 18 threads of ground and 9 paired
threads of supplementary weft per cm.
Weave: weft-patterned. Ground: tabby.
Supplementary: weave of even warp threads with
the supplementary weft 1/5 Z twill (gold wefts
bound in pairs). Gilt threads are not fastened on the
reverse side. Selvedges: weft threads were cut along
the edge of the piece yet some gilt threads were left
uncut (loops). The selvedge, equipped with loops
and fringe, implies that the width of the fabric on the
loom was regulated by the outer thick threads of the
warp, removed after the fabric was taken out from
the loom. A thin stripe of gilt threads runs along the

selvedge on the patterned areas. The same technique
and design can be seen on 13th-century AD brocade
fabrics, for instance, those from the Maiachnyi Bugor
II cemetery (kept in the State Historical Museum,
Moscow; see ATN 52).
Silk warp and weft threads are red. The fibre was
dyed with madder before being spun. The cloth
contained 74 % of alizarin, 24 % of purpurin and 0.8
% of anthragallol. Thus, the original colour of the
fabric was red with a warm orange tint. The core of
gilt threads was left undyed.
The provenance of the fabric from Dress 1
In the 10th century AD, silk fabrics could have
been manufactured in China, India, Central Asia,
Asia Minor, Iran and Byzantium (Liu 1996, 20-22,
124). Different gilt threads, however, were used for
brocades in different regions, and their characteristics
have been used as a guideline in attribution of
provenance to fabrics (Falke 1921, 50-57). Thus,
fabrics with gilt threads of serous membrane of
animal intestines do not appear in the West before the
11th century AD (Sobolev 1934, 53-54; Fekhner 1982,
64; Blanco 1998, 20), therefore the cloth under study
could not have been manufactured in Byzantium.
Persian brocades4 (Owen-Crocker 1986, 187) could
have reached northern Europe at that time. It is
believed, however, that it was not pure gold but gilt

Fig. 8. Brocade fabric of Dress 1
1: Microphotograph of fabric
(right side).
2: Microphotograph of silk
3 and 4: Diagram of twills;
A-A – section towards warp
5: Diagram of passée; 1 – gilt
threads; 2 – silk weft threads.
(Photo: Authors).
Archaeological Textiles Newsletter No. 53


Fig. 9. Drawing of
wraparound dress of the
Liao dynasty (After Rossi
and Rossi 2004, 17).

Fig.10. Fragment of doll dress from the Astona
cemetery (After Min 2006, 224 Fig. 160).

silver or other alloys that were used for making gilt
threads in Iran (Bolshakov 2001, 259-260; Watt 1997,
127-141; Wardwell 1992, 371), while the use of pure
gold was typical for China (Wardwell 1992, 371; Jinke
2006, 129-145; Lantratova et al. 2002, 245-249). Besides
metal composition, the core thread used to wrap
the gilt strip around is an important characteristic
of gilt threads. Thus, cotton thread could have been
used for this purpose in Iran in the 7th-8th centuries
AD (Lubo-Lesnichenko 1994, 202). However, this
criterion cannot be used as a basic guideline for lack
of information on the other centres of production.
It is agreed that the broché (brocaded) and lancé
(weft-patterned) techniques originated from China
(Jonghe 1991, 100). To sum up, based on its technical
characteristics, it can be surmised that the cloth was
manufactured in North China.
The question of the place of making of the garment
is no less intriguing. A search for analogies in the
Chinese archaeological record led us to four gowns
of the 8th-12th centuries AD (Rossi and Rossi 2004,
12-14; Zheng 2007, 98; Feng 2007, 99). Their outline
is close to that of the garment (Fig. 9), yet they are
unfastened wrap-around garments. A fragment of a
doll’s garment from Astana (Fig. 10) dated to the 9th
century AD is also of prime interest in this connection
(Min 2006, 224, Fig. 160). We have also succeeded
in finding a triangular-necked non-wrap-around
shirt-waist gown, although it is a more recent type

Archaeological Textiles Newsletter No. 53


Fig. 11. Surviving fragments of Dress 2. 1: main fabric; 2: trimming fabric; 3: supposed outline of dress cut
(Photo: Authors).

(Gold 2005, 57). All the above comparisons enable us
to surmise that the garment under study was made
in China. However, non-wrap-around tunic-shaped
‘oriental’ form of garment with a flared outline
was characteristic of contemporaneous Iran and
Byzantium as well (Norris 1999, 151, 259, 267, 274).
Therefore, the possibility of the garment having been
made from a Chinese cloth in Iran or the Byzantine
Empire cannot be ruled out. Brocade garments from
Byzantium are mentioned in Scandinavian literature
of the 10th century AD (Krag 2007, 241). The garment
is unlikely to have been made from a whole piece of
brocade fabric in northwest Europe.
Dress 2
Only a few fragments of the second red-fabric
garment have survived in good condition. The two
largest pieces can be recombined into a bodice with
a pronounced neckline (Fig. 11). There are small
fragments of side seams connecting the bodice with
sleeves, hence the conclusion that the garment had
sleeves. A V-shaped neck is marked with a cloth strip
3 cm wide with a minute design. Lining fabric has
survived on the reverse side.
The decorative pattern on the main fabric was
probably embroidered with flat gilt threads, of which
only traces of gilding have survived. Small holes of
gilt threads are indicative of a complex design. Sadly,
we have not succeeded in reconstructing it.

Technical analysis of the Dress 2 fabric (Fig. 12)
Warp: red silk, z-twist, single and paired, single
threads are 0.1-0.2 mm thick. Décochement:
unidentified. Thread count: 40 threads/cm.
Weft: Ground: red silk, without twist; 0.4-0.6 mm
thick. Supplementary: did not survive. The analysis
of metal has shown that there is an admixture of lead
in gold. Combination of wefts and the sequence of
pick cannot be identified. Thread count: 18 ground
weft threads/cm.
Weave: weft-patterned. Ground: louisine/extended
tabby. Supplementary: unidentified. Warp threads
are distributed as follows: two paired threads, two
single; two paired, two single, etc. Selvedges: 16 side
threads of the warp consist of four threads each. In
one of the selvedges two thick threads on the border
with the main cloth consist of 6 threads each. Weft
threads turn along the edge of the cloth. Gilt threads
probably turned before reaching the selvedge.
The material was dyed with madder in the hank. The
ratio of alizarin (78 %) and purpurin (19 %) is similar
to that of the Dress 1 fabric.
The decorative stripe is 3 cm wide. We succeeded in
partially reconstructing the design.
Technical analysis of the trim of Dress 2
Warp: red silk, weak z-twist, paired, the single thread
is 0.1 mm thick. Décochement: 2 paired ground warps.
Thread count: 46 threads/cm.

Archaeological Textiles Newsletter No. 53


Fig. 12. Fabrics of Dress 2
1: Microphotograph of brocade fabric of
Dress 2
2: Diagram of ground twill (threads of
supplementary weft are lost but traces of
gilt survived on remaining threads.)
3: Microphotograph of trimming fabric
4: Diagram of trimming fabric twill
(Photo: Authors).

Weft: Ground: red silk, without twist; 0.4 mm thick.
Thickness of threads is uneven, at some places weft
threads are up to 0.8 mm thick. Supplementary: gold
thread composed of gilt strips of serous membrane
of animal intestines Z-wrapped around a yellow silk
core, z-twisted. The diameter of the thread measures
0.2-0.4 mm, guilt strip is 0.6-0.7 mm wide. The metal
thread coat shows a small admixture of copper and
bromine resulting from remelting. Threads are
single. Proportion: 1 ground weft, 1 supplementary
weft. Pick: 1 ground weft, 1 supplementary
weft. Thread count: 18 threads of ground and
supplementary weft per cm.
Weave: weft-patterned. Ground: tabby.
Supplementary: unidentified. Gilt threads have no
additional fastening on the areas with minuscule
At present the fabric is brown but dye analysis
has shown that it had been red and contained less
alizarin (53 %) and more purpurin (36 %) than the
main cloth. Consequently, the trim fabric was of more
intense scarlet colour than the main fabric of Dress 2.
Technical analysis of the lining of Dress 2
Warp: light brown silk, weak z-twist, paired, single
and paired threads, the single thread is 0.1-0.2 mm
thick. Thread count: 48 threads/cm.
Weft: light brown silk, without twist; 0.2-0.3 mm

thick. Thread count: 26 weft threads/cm.
Weave: louisine/extended tabby. Combination of
warps is similar to that of the main fabric of this
garment: two paired and two single threads.
The weave of the lining cloth has much in common
with the ground weave of Dress 2. It is the similarity
of the two different fabrics that enables one to
surmise that both fabrics were manufactured at the
same centre, having its own mechanism of securing
threads on the loom. And since two fabrics of the
same provenance are combined to form a single
garment, it seems likely that the article in question
was made not far from the place of manufacture of
these fabrics.
The silk fabric ‘with discs’
The fabric survives in two fragments. A decorative
pattern in the shape of discs positioned at a distance
of 9.5 cm from each other has survived on the larger
fragment (8 x 23 cm). Flowers forming a ‘tree of life’
are depicted within the discs.
Technical analysis of the fabric ‘with discs’ (Fig. 13)
Warp: red silk, weak z-twist, single, 0.1 mm thick.
Décochement: 2 warps. Thread count: 32 threads/cm.
Weft: red silk, without twist; 0.3-0.6 mm thick.
Thickness of threads is uneven; it depends on
the presence or absence of design in stripes.

Archaeological Textiles Newsletter No. 53

Supplementary: gold thread is composed of gilt
strips of serous membrane of animal intestines (thin
and somewhat translucent) Z-wrapped around
a yellow silk core, z-twisted. The diameter of the
thread measures 0.2-0.3 mm, gilt strip is 0.5-0.6
mm wide. Threads are single. Proportion: 1 ground
weft/1 supplementary weft. Pick: 1 ground weft, 1
supplementary weft. Thread count: 18 threads/cm.
Weave: brocaded on tabby. Selvedge: 8 warp threads
are thinned out. Weft threads were cut along the
It is impossible to reconstruct the garment to which
the two above-mentioned fragments belonged. The
fragments were washed and smoothed out and there
are no traces of seams. If there was no sewing along
the edge, the cloth could have been used as a shawl,
scarf or veil.
The linen dress
Fragments of a linen cloth belong to a single garment
made from undyed and dyed blue fabrics. All
pieces were washed and partially smoothed out,
which made them less informative. In the course
of the analysis linen fragments were divided into
groups according to their shape and type of fabric.
Fragments of the first group are narrow stripes of
a fine blue cloth with torn off ends and traces of
numerous folds (Fig. 14). The overall length of the
surviving fragments measures over 2 m along the
weft.5 Judging from a fragment measuring 50 cm in
length up to the selvedge, the textile was at least 50
cm wide. Traces of folds are unevenly distributed.
The second group is formed by two fragments
belonging to two fabrics, a thick light and a fine
blue one, sewn together (Fig. 15), with the thread

passing through folded edges of the fabrics. Folds
strengthened with additional seams were made on
the fine cloth. The fragments show that the folds on
the fine cloth are distributed unevenly along the line
of joint with the undyed coarse fabric. It seems likely
that these are fragments of the waist-part of the gown
while the fragments of the first group belong to the
A fragment of an undyed thick cloth probably
belongs to the neck-part (Group 3). It has folds on one
side and several seams. Two more layers of a similar
cloth are sewn on the main fabric in the lower part
of the fragment; these were probably patches (Fig.
16). The interpretation of this fragment is, however,
debatable. The bow-shaped seam-line and the fact
that the outer layer of the cloth is folded on the
face cannot be explained. Remains of blue threads,
probably of a blue-fabric bolster which underlined
the edge of the cloth, have survived along the upper
edge of the fragment with folds. A similar treatment
of the neck is recorded for a linen gown from Pskov
(Zubkova et al. 2009, 293-300).
Group 4 includes fragments of the fine cloth not
showing blue colour. The lack of colour may be
accounted for by the fading of the dye due to the low
quality of cloth dyeing. Fragments belonging to this
group are generally torn off on all sides, yet one of
them has a clear slanting cut. Such a cut could have
been encountered on the bodice if the outline of the
gown was similar to that of Dress 1, on the skirt if it
had gores, or on tapering sleeves.
The data available are not sufficient for an
unambiguous reconstruction of the linen gown. We
suggest the following provisional reconstruction.
The gown is a shirt-waister with the junction either

Fig. 13. Fabric with discs
1: Drawing of decorative pattern
2: Microphotograph of fabric (right
side) (Photo: Authors).
Archaeological Textiles Newsletter No. 53



Fig. 14. Fragments of linen dress, Group 1 (Photo: Authors).

Fig. 15. Fragment sewn together of two fabrics,
Group 2 (Photo: Authors).

Fig. 16. Drawing of dress fragment, Group 3.

Archaeological Textiles Newsletter No. 53

at the waistline or somewhat lower. Its upper part
is light-coloured and the lower one is blue. The skirt
is made from several, at least four, pieces of cloth
made into small folds unevenly distributed along
the joint of skirt and bodice. Probably the main folds
were concentrated on the sides. The sleeves were
seemingly long and made from a fine blue fabric.
The collar was folded into pleats and sewn onto a
bolster of a fine blue cloth. Based on the dimensions
of the neck cut in a straight line (35 cm along weft
threads to one side), the gown had shoulder pieces.
The use of fabrics of different quality implies that
the upper part of the gown made from inferiorquality cloth was covered with another garment,
allowing only the blue skirt and, possibly, the
sleeves to be seen.
Description of tabby fabrics (Fig. 17)
The thick fabric.
Warp: undyed linen, z-twist, single, 0.2-1 mm thick.
Thread count: 8 threads/cm.
Weft: undyed linen, z-twist, single, 0.2-1 mm thick.
Thread count: 8 threads/cm.
Warp and weft threads show a wide scatter of
thickness and an uneven décochement of the twist.
Linen raw material is of poor quality; there are hard
remains of plant stems in the threads.

The fine fabric:
Warp: blue linen, z-twist, single, 0.2-0.5 mm thick.
Thread count: 16 threads/cm.
Weft: blue linen, z-twist, single, 0.2-0.5 mm thick.
Thread count: 12 threads/cm.
The finished fabric was dyed with indigo, a blue vat
dye. The thickness of threads and the degree of twist
are uneven, yet the quality of threads is higher than
that of the thick fabric.
No complete 10th-century AD linen gowns have
been found so far, but a comparison of the fragments
of the two 10th-century AD gowns, one from Pskov
(Zubkova et al. 2010, 293-300) and another from
Gnëzdovo is given in Table 1.
Female shirt-waisters made from two types of fabrics
and dated to the 13th century AD were found at
Toropets and Iziaslavl in Russia (Saburova 1987, 102,
Table 86).
Samite silk fabric
Samite fragments are narrow strips some 5.5 cm
wide. One side of a strip is folded, while the other
is hemmed with an edging of the same cloth. A
similar treatment of strips was recorded for the trim
of a linen article from a chamber burial in Pskov
(Zubkova et al. 2010, 161).
Samite: 1/2 twill S. The ratio of the main to binding
warp is 2:1. There are at least two weft threads, one of
which is either blue or green.

Fig. 17. Microphotographs of linen fabrics. 1: thick fabric; 2: fine fabric (Photo: Authors).
Archaeological Textiles Newsletter No. 53


Parole chiave correlate