You are on page 1of 65

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/296331790

Tapestry Conservation (Part I & II)

Experiment Findings · September 1997


DOI: 10.13140/RG.2.1.4775.9122

CITATIONS READS
0 2,679

1 author:

Foekje Boersma
The J. Paul Getty Trust
20 PUBLICATIONS 44 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Panel Paintings Initiative View project

Managing Collection Environments Intitiative View project

All content following this page was uploaded by Foekje Boersma on 29 February 2016.

The user has requested enhancement of the downloaded file.


General introduction

In August 1996, a research project into the conservation of tapestries was started at the
Central Research Laboratory (since January 1997 merged into the Netherlands Institute for
Cultural Heritage) in Amsterdam. The aim of the study was to investigate the issue of
supporting tapestries, using a practical and scientific approach. The working group for
tapestry conservation of the Dutch Textile Conservation Group (TRON) and the ‘Werkplaats
tot Herstel van Antiek Textiel’ (HAT) in Haarlem initiated this research. The project was given
an international character due to the involvement of the Textile Conservation Centre (TCC)
based at Hampton Court Palace UK.

As tapestry conservation is a very wide and complex field, it was necessary to narrow down
the investigation to a manageable portion. The first research period of 6 months was therefore
dedicated to one of the main points of discussion among tapestry conservators, i.e. the
material(s) used for supporting tapestries. Several cotton and linen fabrics were selected for a
comparison study, as these materials are most commonly used as tapestry supports. Their
ageing characters and behaviour to changes in relative humidity were investigated.

It should be noted here that the subject of tapestry support is a very difficult and complex
matter on its own: immediately related to it is the subject of full or partial supports, stitching
patterns and materials, hanging mechanisms, etc.. This first research period should therefore
be seen as an initial step in tackling this complex field.

The following report is divided into three parts. The first part provides general background
information about tapestries. It gives an introduction into the technique of tapestry weaving,
the materials used and the history of tapestry weaving. It also relates the history from tapestry
repair to tapestry conservation, including a literature survey of tapestry restoration and
conservation.

The second part represents a literature study of the chemistry and physics of fibres, yarns and
fabrics. This information is necessary to understand the degradation processes of textile
materials and their behaviour in climatic changes.

The third part is bound separately and describes the scientific research program as carried out
by the Central Research Laboratory. It includes the outcome of a questionnaire, the artificial
ageing program followed by the physical mechanical tests (tensile strength) and the research
into the behaviour of fabrics when exposed to changes in relative humidity.

Acknowledgement

First of all I would like to thank the Central Research Laboratory for dedicating the funds,
knowledge and equipment for starting this research project. It is an important step in
achieving a better understanding in the conservation of tapestries. I am thankful to the
working group for tapestry conservation of the Dutch Textile Conservation Group (TRON) and
the ‘Werkplaats tot Herstel van Antiek Textiel’ (HAT) in Haarlem for initiating this research. I
am also very grateful to the Textile Conservation Centre for supporting this project by allowing
me the necessary time off to carry out the research in both England and Amsterdam.

The committee, established to discuss the outcome of the research project and its possible
continuation, has been and hopefully will be a critical body between the researchers and the
field of conservation. Special thanks to its members: Judith Hofenk de Graaf, Wilma Roelofs,
Bert de Graaf, Karin van de Wateren, Mieke Alberts, René Lugtigheid and Jenny Barnett.
Without the help of conservators from the field of tapestry conservation, this project would not
have been so successful. I would therefore like to thank all the people who completed the
questionnaire, by which they contributed tremendously to making an inventory of existing
conservation and/or restoration methods and materials used. I would like to express my
special thanks to Susanne Cussell, tapestry conservator at Chevalier Conservation, who
translated the questionnaire in French.

The following conservation workshops I would like to thank for showing me around and
discussing tapestry conservation with me:

In The Netherlands:
- Andelos Textielrestauratie Jenny Barnett
- Stichting Werkplaats tot Herstel van Antiek Textiel René Lugtigheid
- Rijksmuseum Amsterdam Karin van der Wateren
- De Tiendschuur Marianne Maar
Inge Specht
- Textielrestauratie Mieke Vogel
Loutje den Tex

In England:
- Textile Conservation Danielle Bosworth
- Textile Conservation Sheila Landi
- Textile Conservation Studio (National Trust) Ksynia Marko
- Textile Conservation Studio (Royal Palaces) Lyndsay Shephard
- Textile Conservation Centre Alex Clarke

In Belgium:
- Koninklijke Manufactuur Gaspard de Wit Yvan Maes

There are finally a few people I would like to thank personally for their help and assistance:
David Howell, my loyal E-mailer, for the interesting discussions and exchange of ideas and
information, Karin van Nes for her ideas and support and Afke Bruinsma for her help in
preparing the materials for testing.
TABLE OF CONTENTS

Part I

Tapestries: General Background Information

1 Tapestry weaving 5
1.1 The technique of tapestry weaving 5
1.2 Materials 10

2 Brief overview of the history of tapestry 10

3 Types of damage found in tapestries 11


3.1 Damage related to the materials and techniques used in tapestry weaving 11
3.2 Damage caused by previous repairs 11
3.3 Damage caused by the environment 12

4 History of tapestry repair 13


4.1 The restoration of tapestries before WWII 13
4.2 The restoration of tapestries after WWII; reweaving 14
4.2.1 Side-track: the use of synthetic adhesives 16
4.3 The development from restoration to conservation 16
4.3.1 Great Britain 16
4.3.2 The Netherlands 17
4.3 Conservation and restoration alongside and even combined 18
4.4 Conservation of tapestries 20
4.5 Techniques and materials used in conservation/restoration 25
4.5.1 Strap support of tapestries 25
4.5.2 The choice of material for supporting tapestries 26

Part II

Chemistry and Physics of Flax (linen) and Cotton

1 The textile fibres 28


1.1 Flax 28
1.1.1 Production 28
1.1.2 Structure and properties 29
1.2 Cotton 31
1.2.1 Production 31
1.2.2 Structure and properties 32
1.3 Comparing flax (linen) with cotton 34
1.3.1 Chemistry of cellulose 34
1.3.2 Properties of flax and cotton 35
1.3.3 Degradation of cellulose 36

2 Physical properties 39
2.1 Tensile properties 39
2.2 Time effects 41
2.3 Effect of humidity/temperature on the tensile properties 45

3 Moisture in textiles 49
3.1 Fibre shrinkage 51
3.2 Yarn shrinkage 52
3.3 Fabric shrinkage 53
3.3.1 Studies in fabric shrinkage 55

Bibliography 57
Part I

Tapestries: General Background Information

1 Tapestry weaving

In order to explain the different types of damage found frequently in almost all surviving
tapestries, it is crucial to understand the basic structure of tapestries, i.e. the way they were
woven (and not as many laymen still believe embroidered!) and the materials used. The role
tapestries fulfilled in both past and present and the history of their care and 'preservation' are
also important issues when discussing tapestry conservation.

1.1 The technique of tapestry weaving

'Tapestry' is a type of a plain (tabby) weave structure, i.e. a fabric constructed of one set of yarns
under tension (the warps) interlaced at right angles by another (the wefts). Tapestry weave
however has a special feature; the two sets of interlacing yarns are unequal in character and
weight. The warps are coarser, usually undyed and relatively widely spaced, whereas the wefts
are finer and often dyed. The warps are completely covered by the tightly packed wefts. The
presence of the warps is perceived only indirectly in the parallel ridges or ribs which are so
characteristic for tapestry weaving.

Tapestries can be woven on two types of looms, the high-warp loom (Fr. haute lisse) and the
low-warp loom (Fr. basse lisse). The two looms produce essentially the same kind of fabric.
The weavers work on both looms at the reverse of the tapestry. The difference lies mainly in the
operation of the loom, i.e. the means of separating the alternate warps to facilitate the passage
of the wefts. The manner in which the design is translated from the painted cartoon to the woven
fabric is also different.

In a high-warp loom (see fig.1), the warp is stretched vertically between two rollers and is held
under tension. The division between the odd and even warp yarns is created by a shed rod.
One set of warps is stationary, the other set of alternating warps is drawn towards the weaver by
means of pulling small loops (also called 'heddles'), which are attached to each warp of this set
(see fig.2). This means that the weaver always has to use one hand for pulling the heddles,
which leaves only one hand for inserting the weft. The cartoon with the design of the tapestry
hangs behind the weaver on the wall. He has to memorise the design, as he is not able to see
the design while he is weaving. To guide his memory, he draws the outlines of the design on the
warps. In order to check the work, the weaver could either walk to the front or peer through the
warps to see a portion of the completed fabric reflected in a mirror placed in front of the tapestry.
Fig.1 High-warp loom after Diderot

Fig 2 The system of heddles of the high-warp loom after Diderot


The low-warp loom is thought to be developed from the ordinary shuttle loom (see fig.3). The
warp is stretched horizontally between two rollers. The division between the odd and even warps
is created by a time-saving device of pedal-controlled heddles which freed both hands of the
weaver (see fig. 4). Every warp is attached to a heddle (instead of every alternating one in a
high-warp loom). The warps are raised in groups by pressing the pedals beneath the weavers
feet. Several weavers usually sit next to each other, each one having their own set of heddles
which control the warps in their own section. The cartoon, a full scale design for the weaver to
follow, is cut in vertical strips and placed underneath the warps. The weaver could reproduce the
most intricate details of the design quite easily by looking down on the warps and the underlying
cartoon.

Fig 3 Low-warp loom after Diderot


Fig 4 Detail low-warp loom after Diderot

The disadvantage of this type of loom however is that the weaver is unable to check the progress
of the work nor see a large section of the tapestry at a time as the front of the fabric is invisible.
He could only see small sections by inserting a mirror underneath. The low-warp loom was
adapted in the 2nd half of the 18th Century to overcome this problem; the engineer Vaucanson
invented an improved version of this loom for the 'Manufacture de Gobelins', which could be
tipped up so that the weaver could follow his work with almost as much freedom as the high-
warp loom worker.
Another disadvantage was the fact that cartoons intended for the low-warp weaving had to be
made in mirror-image, because the face of the tapestry was in direct contact with the front of the
cartoon.

In tapestry weaving the design and the fabric are constructed simultaneously and interdepen-
dently. The design is created by the dyed wefts alone, and is visible on both the obverse and
reverse of the tapestry. The weaver works with a number of bobbins, each loaded with yarns of
different colour and/or material. The bobbin is worked back and forth across the section of the
warp where the design required that specific colour or material. This results in another very
important feature of tapestry weaving: the weft does not extend across the full width of the fabric
(from selvedge to selvedge), but only as far as a colour is required in the design. One could say
that tapestries are constructed as a 'weft-mosaic'. This phenomenon has very important
consequences for the structure of the tapestry.

Where two different colours meet along a straight line, parallel to the warp, a slit occurs in the
fabric when the two adjacent warps are used as selvedges for both colours. There is no
common weft to bind the two together. In this way a straight and clear line of colour division is
made, involving the creation of a gap or slit. Most slits are closed by stitching once the tapestry
has been finished. Small slits are sometimes made deliberately to enhance certain features in
the design. These are not closed by stitching.
Fig 5 Stitching of the slits after Diderot

To avoid slits, which are undermining the structure of a tapestry, weavers sometimes 'interlock'
the wefts from different coloured areas. The different coloured wefts are looped around each
other where they meet, before returning in opposite directions in the next row of weaving. Thus
the line of colour division is less clear, but the weave structure much stronger. Another way of
avoiding slits is 'dove-tailing'; alternating wefts from two adjacent coloured areas are passed
around a single warp lying in between the two areas of colour.

Tapestries are woven side-ways on both types of loom, which implies that the finished fabrics are
hung with the warp running horizontally. An explanation is perhaps to be found in the way
tapestries were used. Tapestries were mainly intended to cover walls. The lengths of these
walls could vary considerable and was usually longer than the height. On a loom, the warp
direction is in theory endless, whereas the width is a fixed measurement. By weaving the design
side-ways, one could therefore create tapestries as long as possible. Another explanation is
related to the technique of weaving: most important elements in the design of a finished tapestry,
e.g. architectural features, trees and outlines of figures, are vertically orientated. These lines are
easier to weave on the loom when they run horizontally. It is also noted that a tapestry can
shrink up to 5 cm per meter in the warp direction, once it is finished and removed from the loom.
This is because the woollen warp, which is extremely elastic, recovers from being stretched on
the loom. It is thought to be less disturbing when elements of design (e.g. figures, trees) shrink
in their width than in their length. Finally, some people argue that the light reflection on a tapestry
with horizontally orientated ribs creates a clearer image. Light falling on vertically orientated ribs
would create lines of shadow which would obscure the design. Anyhow, the fact that tapestries
often hang on their wefts is seen as a great disadvantage, because all the weight of a tapestry
2
(on average approximately 1 kg/m ) has to be carried by the tapestry's weakest element, i.e. the
discontinuous wefts.
1.2 Materials

In tapestries woven in Europe, undyed woollen yarns are usually selected for the warps.
Occasionally linen yarns are used as warps, particularly in tapestries woven in the Germanic
countries. In modern tapestries cotton is often used for the warp. Wefts are mainly of dyed
woollen yarns. Wefts of silk for highlights and lighter colours are used sparsely at first (medieval
period), but later on with increasing frequency and concentration. For very luxurious and rich
tapestries, different types of metal threads in gold or silver colour were used.

2 Brief overview of the history of tapestry

The earliest form of tapestry weaving is the so-called 'kelim', a fabric in which the slits between
different coloured areas are made intentionally as part of the geometrical design. These slits are
not closed by stitching afterwards and therefore remain open. This technique is thought to be
developed around 7500 A.D. in Asia-Minor. It is still unknown how the technique entered Europe
but one theory is that it was introduced by the Crusaders on their return from the Holy Land. The
first fragments of tapestry weave known to us today are the 11th Century 'St Gereon' fragments,
which originally hung in the church of St Gereon in Cologne, but which are now scattered around
in several European museums. It is thought that these and other early tapestries, of which only
very few examples survived, were woven on a small scale in monasteries.

In the 14th Century the art of weaving became more widespread over Europe. Large studios
were established, mainly in France (Paris and Arras), the Elzas and Flanders. In the next
hundred years the art-form transformed into a very important and financially powerful industry.
Tapestries were woven not only for churches and monasteries but also for royalty and wealthy
civilians. They were part of the movable furniture and were so indispensable to the owner, that
they accompanied him on journeys from estate to estate, sometimes even went to war with
them. To own a tapestry was a sign of wealth, grandeur and power.

In the 15th Century the Flemish cities of Doornik and Brussels became important centers for
tapestry weaving, with the 16th and 17th Century being referred to as the 'belle epoque' of the
Brussels industry. From a point of view of quality, tapestry weaving was at its peak in the late
14th, 15th and 16th Century, when the designs reflected and exploited the techniques of weaving
to the full. In 1545, almost 15,000 persons were in some way involved in the Brussels' tapestry
industry, which is a considerable amount taking into account that the city had a population of
60,000. In the 18th Century the tapestry industry reclined and in 1794 the last Brussels'
workshop, belonging to Jacob van der Borcht, closed. (Delmarcel 1994, p.13)

In the 17th and 18th Century France became more important and surpassed Flanders,
especially with the establishment of the 'Manufacture Royale des Meubles de la Couronne' near
Paris in 1667 and of another 'Manufacture Royale' at Beauvais in 1664. In the 18th Century the
latter one became an active production centre together with Aubusson. Tapestries were
imitating paintings more and more, as a wider range of colours were introduced. The typical
techniques in weaving employed to create optical effects were gradually lost. At the end of the
18th Century the tapestry industry died out as there was no demand anymore for this expensive
art-form.

William Morris (1834-1896) revived the technique of tapestry weaving as it was applied in
Medieval tapestries. This resulted in a short-lived revival of the technique in Europe in the late
19th Century. The French artist Jean Lurçat (1892-1966) designed several tapestries using a
scheme of approximately 20 to 30 colours. He also used Medieval tapestry-techniques for
achieving colour changes. In the 1960s tapestries became extremely monumental and often
three-dimensional.
3 Types of damage found in tapestries

3.1 Damage related to the materials and techniques used in tapestry weaving

One of the factors which first of all establishes the live-expectancy of a tapestry is the quality of
the materials used and the quality of work, as Karin Finch and Ksynia Marko pointed out:

..."The quality of tapestries varies enormously both in technique of weaving and design. While
the weaving of some may be rather loose and coarse, those which were made by later weavers
in some of the very famous European workshops are in very fine, close weaving. While many of
the early tapestries which have survived were woven almost entirely of wool, weavers working
later used much more silk, both for highlights in the actual designs and also for quite large areas
of sky."... (Finch 1977, p.56). ..."Discrepancies of warp tension often occur, manifesting
themselves as cockling or producing 'seer sucker' effect."... (Marko 1995)

The materials used for weaving were sometimes manufactured in such a way that they became
almost self-destroying. For example the wool. ..."The undyed wool used for tapestry warps has
usually lasted well. In contrast, the wool used for the weft is in a very mixed condition. Broadly
speaking, for Gothic and Renaissance tapestries a small range of colours was used, mostly dyed
with straightforward vegetable dyes which did not harm the wool. Notable exceptions are black
and dark brown which were produced by oxidation of iron which resulted in degradation of the
fibres. [...] From the Baroque period, as detailed imitations of oil paintings became fashionable,
the number of dyes in use doubled. The new dyes were often unstable, sometimes harmful to
the fibres. Consequently, later tapestries are often much more faded than earlier examples."...
(Pow 1970)

The same accounts for the use of silk. ..."Silk was at first used only for highlights or for small fine
panels. Its use increased with time and by the end of the 17th century, expanses like skies
would be entirely of silk weft. [...] The strain was often too great for these, especially the 'skies'
which carried the weight of the entire tapestry below. The silk weft of most 17th- and 18th
century tapestries is now in a very delicate state. Where it has not actually broken or split, it is
very brittle and can easily brush off in a fine powder or break up in fluffy tufts under the pressure
of a needle. Although an area of silk tapestry may look quite solid, under the slightest strain it
may disintegrate."... (Pow 1970)

Secondly, one of the major damaging factors is incorporated in the type of weave used when
creating tapestries, i.e. the presence of slits where two colours in the design meet. When the
tapestry weaving was finished, most of these slits were sewn up, generally with silk or linen
threads. These stitching materials perished relatively soon. The weight of the tapestry then
caused the slits to gape open which resulted in enormous strain on both ends of the gaping slit.
This could result in severe structural damage if left untreated.

3.2 Damage caused by previous repairs

Almost all tapestries have been repaired at least once in their life-time. These repairs were not
always carried out by professionals. Structural damage has often been caused by tight stitching
and darning, cuts, etc.. The article about the conservation of the "Roman de la Rose" tapestry
(woven in Tournai around 1460) by the Textile Conservation Centre at Hampton Court Palace
(1985-86) gives an example of an extensively repaired tapestry.
..."The tapestry had been extensively repaired in the past, at several different times, resulting in
many areas of weakness. One of these was the dark brown wool weft already referred to.
Some of these dark brown areas had been repaired by simply cutting away both warp and weft
from the weak areas, making clean, sharply cut holes in which brown wool tapestry-woven
patches were inserted. These patches were visually quite acceptable but were themselves now
degraded and very poorly secured, giving inadequate support. The insertion of these repairs had
involved the loss of some of the original weave; more of the weave was now being lost as
tension at the joins of these patches pulled weft yarns off the warp. There were also several
large holes where both warp and weft were missing, and around the edges of these the fibres
were brittle and yellowing. These areas, and some of the weak areas at the edges of the
tapestry, had been supported on patches made from a great variety of materials, including loop
pile carpet, double cloth, triple cloth, fulled wool, plain weave fabric, and patterned weaves in
wool and linen. [...] Some, but not all, of these repairs were visually quite effective. However,
they were causing physical stress to the tapestry's structure in various ways: some patches gave
support to larger holes, but were too thick and were attached in such a way that they only added
to the stress on the fibres and in some places were causing pulling and puckering. An unsightly
thick brown thread had also been woven into large areas of the piece. The colour was
inappropriate; it looked dead and flat. Looking at the reverse of the tapestry, it was obvious that
this thread had never matched the original. The yarn was too thick and was not always woven at
the correct tension or sometimes even in the correct technique. This too was therefore causing
much distortion and stress in the weave structure."... (Borg Clyde 1992, p.153)

3.3 Damage caused by the environment

Those factors which are damaging to textiles in general, also account for the accelerating
degradation of tapestries, i.e. light, fluctuating relative humidity (temperature), insects, etc.. In
part 2 of this report more detailed information will be given about the physical behaviour of
textiles. For the moment only one extreme example of damage caused by unstable
environmental conditions is presented here.

..."The castle of Châteaudun is not provided with a heating system. All of the tapestries are
hanging on humid walls, attached by nails on wooden beams. Under these circumstances,
variations in temperature and humidity can provoke dramatic problems, such as the one that
developed during the very cold winter of 1986-1987. A sudden change in temperature led to an
exceptional concentration of humidity on the walls. The tapestries soaked up the water, which
then converted to ice, causing considerable damage by its weight."... (Maes 1987)
4 History of tapestry repair

4.1 The restoration of tapestries before WWII

The repair of tapestries is an old profession. In the early days tapestries were moved from one
residence to another. Wear and tear in these circumstances was considerable, and they had to
be repaired at high costs. Records dating back to the 14th Century describe the restoration of
tapestries, which included cleaning and repair of holes and tears, carried out by professional
tapestry weavers.

..."Jules Guiffrey, citing the Lille archives published by Jules Houdoy, shows the Burgundian
dukes sending their tapestries for restoration to Brussels workshops in the fifteenth century. The
Battle of Liege, woven at Arras for John the Fearless (died 1419), was restored in 1499 and
again in 1541-43, and Pieter de Pannemaker, head of the leading tapestry workshop of his days
at Brussels, undertook the repair of the Gideon tapestries of the Order of the Golden Fleece in
1529, accounting eight hundred and seventy-one days work on them."... (Wingfield Digby 1971,
p.57).

The situation at the English court was as follows:


..."The English Crown, like most monarchies, had its own staff for the restoration of "Arras"
tapestries. They were under the control of the Master of the Great Wardrobe and were directed
by the Yeoman Arrasworker. The latter was often chosen from among the leading owners of
tapestry workshops. His own workshop might provide new tapestries for the Crown, which
figured in the Wardrobe accounts, but his chief function was to supervise the cleaning and repair
of tapestries by six or more men employed at daily rates solely for this purpose."... (Hefford 1976,
p.67)

It can be ascertained from surviving records that tapestries in constant use were cleaned and
repaired approximately every 20 to 30 years (Hutchison, 1990, p.10). Repair work involved
restitching of slits (essential to the main structure of the tapestry), reweaving damaged areas,
adapting borders, etc. It was also practice to paint or chalk certain areas to enhance the design
or to retouch mistakes in weaving.

At the time when tapestries became more and more an integral part of the interiors in wealthy
houses and palaces, they were often cut in order to fit in certain spaces. Tapestries are likely to
have suffered severe damage while being on permanent display, where they were subject to
dust, soot from open fire places, insects and sometimes direct sunlight. The repair work of these
damaged tapestries was entrusted to official tapestry weavers, but in the 19th and early 20th
Century also to art-dealers and amateur repairers. Damaged areas were usually cut out and
sometimes rewoven or replaced by patches of other tapestries, painted linen cloth, etc.. Smaller
holes were usually darned, often creating tension and distortions in the surrounding areas.
Materials used for the repair work were often of bad quality and frequently dyed with unstable
dyes.

The repair of tapestries could sometimes involve extreme intervention:


..."There was once a quite serious restoration attempt to recapture the original appearance of a
faded tapestry by turning it back to front and taking all the loose ends back through the tapestry
to the other side, but, apart from the considerable difficulties encountered in that process and the
peculiar effect made by all the figures on the tapestry appearing as left-handed, it was then
realised that the colours would inevitably fade, and the attempt was not repeated."... (Finch 1977,
p.62)
4.2 The restoration of tapestries after WWII; reweaving

After the second World War, the restoration of tapestries became more professional, for
example at the 'Werkplaats tot Herstel van Antiek Textiel' in Haarlem, the Netherlands
(henceforward referred to as HAT).

Most tapestries treated by the HAT in the early days of its existence underwent total restoration,
i.e. the replacement of weak and damaged material and of all silk. The decision was based on
the following considerations:

..."Many of you will think this reweaving of all worn materials rather radical and would have
thought the filling in of the missing parts enough. At first we also took this point of view, but soon
after starting the work it became evident that the condition of the original silk of these tapestries
was so bad that the silk was only a mass without structure, sticking together and showing the
original connection and the original design but disturbed at the slightest touch. It appeared that
even during restoration or soon after parts we had thought intact and had spared also fell out, or
were torn away by the new materials put against them; this applied particularly to the dark brown
wool. So we judged it advisable to reweave at one time all the silk and that part of the wool
which was in the same condition."... (Diehl 1964, p.105)

Pow describes the special technique for reweaving which had been developed at the HAT:
..."The Haarlem workshops' method of reweaving:
The Haarlem Workshops have devised several ingenious refinements on the usual method of
repairing warps in tapestries. Normally, where the warps are broken or missing, new lengths of
thread of the same gauge are simply inserted under the weft alongside the originals and drawn
across to fill the gap. The thickness of the thread used in this procedure inevitably results in
some inequality of density around the repair, and there is a risk that, as it is inserted between the
weft and drawn through to be secured at the back, the coarse new thread will break up the brittle
old fabric.
At Haarlem two finer threads are used to replace each old warp, one a darker shade of cream.
Any loose weft is picked away from the damaged area until about 3/16 inch of the end of each
old warp is exposed. The new warps are cut very long and tied to the old with weavers' knots at
varying distances from the remaining weft on either side of the break. the loose end of each new
warp is now forced along the appropriate old warp opposite, under the weft. Finally, the long
ends are secured by large stitches through the tapestry.
The new weft is woven in the usual manner, except that new warps are picked up singly by the
sewn-in ends until the knots are concealed behind, and then they are picked up in pairs. This
gives a smooth continuous transition from old to new weavings, while the extent of the repair is
clearly indicated on the back of the tapestry by the lines of knots.
When the reweaving is finished the long ends of new warps are trimmed off where they were
secured by stitches through the tapestry. They are firmly held by the new weft and the weavers'
knots. The original warps of early tapestries were of unbleached wool, but wool of similar
strength and density is no longer available. At Haarlem, as elsewhere, the desired effect can
only be obtained by re-warping with strong cotton of suitable gauge."... (Pow 1971, p.82)

One of a series of tapestries thus treated was the 'Devonshire hunting' tapestry series. In the
late 19th Century these tapestries were hanging at Hardwick Hall (UK). They had previously
been cut into strips and sections and were hanging two to three layers deep on the walls in some
places. The pieces were send to the 'Victoria & Albert Museum' to be joint together into four,
almost complete hangings. Prior to this restoration procedure, they were first cleaned by
'Messrs. Pullar & Son' and then sent on for restoration to the 'Decorative Needlework Society' at
17 Sloane Street in London in the period 1900-1910. Damaged areas were rewoven, but the
weight and quality of the woollen threads used for reweaving was ill matched to the original. In
1957 the tapestries came permanently into the collection of the museum. They had become
very dirty again and nearly all restoration threads had changed colour. One of the tapestries was
in the meantime restored in Paris by 'M&G Petit' as it went on loan to exhibitions in France.
The three others were sent to the HAT in Haarlem (1955-1966) to be repaired using specially
dyed yarns, which were carefully selected for their light-fastness (6-7). The joints were newly
rewoven. (Wingfield Digby 1971)

As mentioned above, the restoration material used for reweaving, had to have very good light
fastness. This would prevent the rapid fading of newly woven parts, which would make the
repairs very obvious and obtrusive. Pow refers to two workshops dealing with the issue of dye
fading in different ways:

..."The two main considerations in choosing wool for new weft are texture and resistance to
fading. [...] As most dyes fade rapidly at first and then more slowly, the tapestries will have
reached a degree of fastness seldom achieved with a new dye. Haarlem deals with this problem
by dyeing their own materials, the Tabard workshop at Aubusson, France, by exposing the new
wool to daylight for about a decade before using. [...] A solution to the problem of finding silk in
colours fast enough for reweaving does not appear to have been found, except the negative one
of repeating the treatment every twenty years."... (Pow 1970, p.141)

Not only at the HAT reweaving was carried out. It was a technique which remained (and in some
cases still is) in practice along side the development of new conservation methods. For example
in the tapestry workroom at Hampton Court Palace:

..."Up until the late 1970s the tapestry workroom had continued the tradition of reweaving
restoration techniques that had been set up by the firm of William Morris & Co. under the
direction of H.C. Marillier in 1912. [...] Trained weavers from Merton Abbey Tapestry Workshops
(established in 1881) were thus based at Hampton Court to oversee the work, to wash tapestries
prior to their restoration and to train new workers. The restoration techniques carried out over
the half century changed little. All materials - wool, silk and occasionally metal threads - had
been continuously and exclusively supplied from the Merton Abbey workrooms. By the 1970s,
Merton Abbey having closed down in the 1940s, the remaining supplies of yarn at Hampton
Court were seriously depleted and all subsequent restoration work was therefore carried out
using an ever decreasing range of yarns in suitable colours, weights and condition. [...] In June
1974 the restoration/reweaving work was begun on the tapestry entitled 'The battle with King
Porus of India', and in 1976 the second tapestry 'Battle of Granicus' was begun. The work
carried out on both tapestries was typical restoration: all weak, degraded weft fibres were
removed from the weave and replaced with similar threads using 'needle-weaving' techniques.
[...] the reweaving invariably resulted in a definite and irreversible loss of fine detail, quality and
subtlety, as large amounts of original weft threads were removed, often over very large areas."...
(Shephard 1996, p.721-722)

More recent at the Canadian Conservation Institute, reweaving was chosen as the means to
preserve a Brussels 16th century tapestry (Jakobiec 1993).
..."Compatible yarns of wool, silk or cotton were used to reweave all areas where warp threads
were still intact, and new linen warps were inserted where required. In areas where sufficient
evidence of the original design existed, it was recreated. Where there was not sufficient
evidence for recreating the original design with confidence, the area was rewoven in an
aesthetically pleasing manner that respects the integrity of the original. This traditional approach
is more time consuming than other repair techniques; however, it is more harmonious, both
physically and aesthetically, with the original technique."... (Jakobiec 1993, p.3)
4.2.2 Side-track: the use of synthetic adhesives

In the sixties synthetic thermoplastic adhesives were introduced in the field of restoration and
conservation. Some experiments were carried out in both the Netherlands and England to
investigate the possibilities of using adhesives in the conservation of severely degraded
tapestries.
The adhesive could be ..."applied in two ways, as impregnants for consolidating fibres, and
secondly as adhesives for attaching textiles to supporting backings. The resins so far suggested
for impregnation are polyvinylbutyral and soluble nylon. The Haarlem workshops have used
polyvinylbutyral for some years to consolidate tapestries which are to be stitched down. Used at
4% in ethylalcohol, it attaches silk fibres to each other and to the warps and prevents the silk
from drifting off in powdery clouds as the tapestry is handled.
Unfortunately, a noticeable change is wrought in the colours. There is an overall darkening
which may give a superficially richer effect. Soluble nylon appears to consolidate in the same
way, with less colour change, but with greater stiffening of the textile. Neither of these
impregnants will support a damaged tapestry alone. They are merely a means of making it solid
enough to handle. Backing with adhesive coated net is a fuller treatment - in some cases the
sole repair method employed. A strong Terylene net is coated with polyvinylacetate and ironed
to the back of the textile to give it support and cohesion. It is regrettable that one is sticking the
net not to the warps, which are covered in the tapestry weaving technique, but to the back of the
wefts. Since the weft is often worn or broken on the front of the tapestry along the ridge of each
warp, there is a danger that, when the tapestry is hung and creeping at a different rate from the
net, where the net sticks to the weft it will give on the front and come away from the rest of the
tapestry. [...] Further difficulties arise in sticking the net to the tapestry. The most obvious
difficulty is presented by the loose ends of the weaving which are left hanging on the back. [...]
More fundamental is the fact that PVA does not adhere so readily to wool as to other fibres."...
(Pow 1970, p.147-148)

The use of synthetic resins in tapestry conservation has never been widely spread, because the
results were not satisfactory.

4.3 The development from restoration to conservation

In the late sixties, the concept of tapestry conservation, i.e. supporting the weak and damaged
areas of a tapestry onto a support fabric, is further developed.

4.3.1 Great Britain

Karin Finch was one of the initiators of the development of conservation techniques. She studied
weaving and design at the art school in Copenhagen. In 1946 she married and moved to
England, where she was able to get a job at the 'Royal School of Needlework' in Hampton Court
Palace. New embroideries were worked there and old ones restored. Also tapestries were
sometime brought in for repair work. In 1954 she was employed by the Conservation
Department of the 'Victoria & Albert Museum' in London to work on the preservation of
tapestries. In those days, repair or reweaving was still carried out by stitching into the
surrounding sound parts of the tapestry - with unavoidable visual damage to their design. Karin
Finch was allowed to conserve one tapestry by taking out darns and pieces, which were cut from
other tapestries and inserted. She supported the weak parts onto linen patches with no rewoven
guesswork of what might have been.
..."I felt that using linen patches on the back of the tapestries would enable us to preserve what
was left of both color and design and still make the tapestry safe to hang. I also wanted to
restrict the use of foreign material on the front of the tapestry so that inevitably fading and colour
changes would be less noticeable and cause minimal interference with any part of a design. [...] I
used neutral colour for the stitching."... (Finch 1990, p.71)

In 1959 she left the 'Victoria & Albert Museum' to return to weaving and design and to further
study the textile techniques as practised at the 'Historical Archaeological Research Centre' at
Lejre in Denmark. She set up a private textile conservation workshop, which moved to Ealing in
1968. In this period she taught the basics of textile conservation to approximately eighty people
from nineteen nationalities . She also started teaching textiles to the 'History of Dress'-students
at the 'Courtauld Institute of Art'.

Danielle Bosworth, a tapestry restorer from France, joint her on a daily basis.
..."When Danielle Bosworth joined me in 1968, she took over the development and led the work
on the Esther series then undertaken for the V&A. She introduced colour indications into the
couching and began to work on the problem of holes and their eye-catching effects on the
viewer. [...] To restore a smooth surface to tapestries marred by holes, we eventually settled on
stitching new warp yarns, dyed to match the old, onto the supporting patches and carrying
background design features across the new warps. The intention is to lead the eye across the
gap by using the most unobtrusive methods possible."... (Finch 1990, p.72)

By 1970 the amount of work began to overtake the studio. Karin Finch became gradually
convinced that a National Institute of Textile Conservation had to be the answer to all problems.
In early 1971 a memorandum was circulated proposing such an institute (the Textile
Conservation Centre). In October 1973 the long awaited course in textile conservation became
reality and it was granted a Grace and Favour Apartment at Hampton Court Palace in 1975. In
1986 she retired. (Finch 1985, 1989)

4.3.2 The Netherlands

In the Netherlands, the conservation of tapestries, based on the ideas of Karin Finch and
Danielle Bosworth, was slowly introduced and it ran alongside the practice of restoration. The
HAT describes the considerations when choosing between restoration or conservation.

..."A particular procedure to cover every case cannot be given. Total restoration is extremely
expensive. The alternative is conservation, which is considerably cheaper but yields less
satisfactory results. Generally speaking, the destruction of the fabric will be delayed for a short
time only, while all the damaged spots, missing parts and suchlike remain completely visible. It
should be realised that bad spots or missing parts, repaired from behind by a piece of material of
a totally different nature may be eye-catchers, which can have a great influence on the aesthetic
impression created by the whole. The method chosen will depend on the relation between the
cost and the value of the tapestry, while art-historical and historical value, and the personal value
to the owner play a great part. It might be that total restoration is being considered but that the
cost is an objection. A temporary conservation may be a substitute but it is advisable to have
such a tapestry photographed, in detail, with a view to later restoration. For tapestries where a
linen warp was used, total restoration is out of the question because of breaking of the warp
threads. Conservation is the only way to keep such a tapestry in existence for some time"...
(Diehl 1972, p.162-163)

This point of view was more or less generally accepted, as other articles from those days
indicate. Pow (1970) also refers to stitching weak and damaged areas onto a support fabric,
usually linen patches.
..."There is always the danger that when hanging, due to their different rates of creep, a patch
which is too large will be at variance with the tapestry and will either pull it up or hang in useless
blisters behind. This is particularly apparent where a tapestry has been repaired by stitching the
weak areas to an overall lining. This usually means that, when it is suspended freely from the
upper edge, the tapestry drops further than the lining and is caught up at all its most vulnerable
points."... (Pow 1970, p.145-146)
At the Rijksmuseum, tapestry restoration was carried out under guidance of Mr Bloedhouwer,
who was head of the department from 1940 to 1980. In general, reweaving of holes and missing
areas was carried out, although the old and weakened silk was not replaced. The disadvantages
of this technique were the frequent manipulation of the tapestries during treatment, which caused
weak areas to disintegrate while working. The dyes of the restoration materials were not always
of high light fastness quality, resulting in colour differences between original and repair. Mr
Stephen Cousens, the successor of Mr Bloedhouwer, was a tapestry weaver and trainee of
Karin Finch and he introduced the conservation techniques, which were by then practised at the
Textile Conservation Centre. (Specht-den Boer 1991)

4.3.3 Conservation and restoration alongside and even combined

Throughout the seventies and eighties, both restoration and conservation treatments were
carried out, quite often in combination with each other.

Kajitani (1976) describes the reweaving of damaged areas in the restoration of Medieval
tapestries at the Metropolitan Museum, which had been an accepted practice because tapestries
were seen as pictorial art objects. In her opinion experience was necessary to determine the
extent to which the pictorial elements were recreated and degraded yarns substituted. A
tapestry-weave or a non-textured monotone dyed fabric could be used for the infilling of large
holes.

Lemberg (1977) also mentions the same mix of restoration and conservation when describing
the treatment of a Caesar tapestry (woven in Tournai 1465-1470) which was treated in the period
1965-1974. The tapestry was previously restored in the late 19th century, when badly worn
woollen wefts were uniformly replaced with black wool. The repair wool was however too thick
and had discoloured to an olive green. The black wool was therefore removed and replaced by
new wool, which was dyed in different shades of dark colours based on dye-analysis of the
original material. Weak areas were given a patched support of linen fabric.

In Belgium at IRPA ('Institut Royal du Patrimoine Artistique'), restoration and conservation were
both accepted treatments. Masschelein-Kleiner (1984) and Boeck (1986) refer to the restoration
of tapestries, i.e. rewarping of holed areas and reweaving of missing wefts, as being the
preferred technique. In their articles they claim that rewarping and reweaving could reinforce the
tapestry and return the original visual aspect. Old repairs were often removed, unless they were
not disturbing the structure nor the design. Rewarping was then carried out with yarns that
resemble the original, but were slightly thinner. The missing area were rewoven with specially
dyed silk and wool.

The IRPA was also critical towards this technique as it viewed the high costs as a disadvantage
plus the fact that the original design of the missing areas was often unknown, involving
guesswork in recreating it. The conservation of tapestries, involving the support of weak and
damaged areas onto different coloured linen patches, became therefore more and more
standard practice. The bare warps were then secured using laid and couch stitching in the weft
direction. The tapestries were fully lined afterwards with linen fabric, attached with vertically
orientated, staggered lines of stitching.
Dolcini (1990) describes the treatment of the tapestries of the 'Sala dei Duecento' in the 'Palazzo
Vecchio' in Florence. The conservation of the set of twenty tapestries (dating 1545-1553),
started in 1985. Only three tapestries out of the set had ever been treated before! The following
steps were taken in the conservation treatment:
- slits and less damaged areas were repaired in such a way that the tapestries would not require
a total supporting structure. Cotton thread was used for restitching slits,
- areas of silk weft loss near slits were rewoven, using silk thread,
- small damaged areas located next to well-preserved areas were rewoven, only if there was no
doubt about the original design,
- more severely damaged sections were given a patched support, using three different weights of
linen fabric.

Hutchison (1990) also gives two different approaches in conservation in the treatment of
Gluttony and Avarice, two 16th Century Brussels tapestries.
..."Reweaving well-defined losses of silk or dark brown wool has proven appropriate in many
cases. For wool repairs, commercially dyed wool yarns are usually chosen. For replacing silk,
DMC cotton embroidery thread is preferred because of its longer life expectancy and because it
has less sheen than modern silk thread. It is preferable to incorporate the repair by stitching into
the construction of the tapestry if possible, rather than attaching many little patches to the back.
This repair is best achieved using either wool or cotton thread in an open tabby mend, spaced
about 0.4 cm apart and extending well into the undamaged surrounding area. This type of repair
adapts more sympathetically to environmental changes than those repairs that have yet a third
component of a fabric patch. On the other hand, large areas of loss - 30.5 cm square or larger,
where warps are well exposed, are best treated by adding a support patch. A suitable fabric for
patching is washed twill cotton, which may be commercially dyed to match the repair area. If
warps are missing or broken, they can be superimposed on the fabric. An open tabby mend
sewn with two strands of DMC cotton embroidery thread attaches the patch support to the
tapestry. The edges of the patch are secured with a zigzag stitch into the undamaged
surrounding area of the tapestry."... (Hutchison 1990, p.91-92)

Three quarters of the tapestry Avarice were in need of support and therefore a full support of
cotton twill fabric was chosen. The support was attached by a grid of vertical lines of running
stitches in a brick pattern. Warp couching was carried out through the support using cotton
DMC. A light weight cotton print cloth was used for lining.

Ewer (1989) and Ward (1988) describe the tapestry conservation project at Biltmore House in
Ashville, North Carolina (USA). The ..."important tapestries are a prominent part of the Biltmore
House collection and include two sets of 16th century tapestries, a 17th century tapestry, and a
late 15th/early 16th century tapestry fragment. [...] The tapestry conservation project was given
top priority because of the tapestries' importance to the collection. [...] The current treatment
approach began early 1987. [...] Repair stitching of the tapestries has three main functions. The
most important function is to support the weak areas so that the tapestry is stabilised and
protected against future damage. Secondly, conservation stitching is used to repair harmful or
unsightly repairs which were made in the past. And third, repairs may enhance and preserve
aesthetic details. The basic techniques used to perform these functions involve repairing slits,
repairing broken warps, support stitching and some reweaving. [...] The most commonly used
replacement thread is made not of silk but of linen, which is more durable. Also, many of the
woven areas of the tapestries, which were originally executed in silk, have become friable with
time. These weak areas are given support stitches which are spaced 1/8 - 1/4 inch apart in a
tabby weave pattern through the remaining silk. Areas in which the silk no longer exists are
rewoven in a color similar to the original, using DMC stranded cotton. Broken warps are
replaced with several strands of fine wool, while the wefts are replaced with only one strand."...
(Ward 1988, p.381-387)
Bauer (1991) describes the history of tapestry preservation in the 'Kunsthistorischen Museum' in
Vienna. Until the 1980's, the tapestries were fully restored, i.e. weak and damaged areas were
rewoven, unfortunately using materials dyed with unstable dyes. The technique changed slightly
in the eighties. Bauer mentions in the article that reweaving still took place, using only natural
materials (silk and wool). The difference was that the newly woven parts were much loser, in
order to enable visual distinction from the original and to facilitate the reversibility when
necessary. Holed areas were rewarped using cotton yarn. Cotton was selected because it was
thought to be smoother than woollen yarn and therefore easier to insert into the original. The
cotton yarns were pre-shrunk and dyed to match the original warps. Rewarping and reweaving
of holed areas was preferred to applying a patched support; it was felt that the introduction of
new warps restored the tension within the original warps. When using a patched support, this
would be much more difficult and distortions were likely to occur. Tapestries were always lined
after conservation with a specially woven, dense cotton sateen fabric.

4.4 Conservation of tapestries

From a literature survey it seems that tapestry conservation is more widely accepted and carried
out more frequently during the eighties up to the present day. The difference between
restoration and conservation is clearly outlined in the following citation:
..."Conservation seeks to preserve a piece in its present state, providing favourable conditions for
its long-term survival. The aim is to use a minimum of new materials and to disturb the piece as
little as possible. Restoration, on the other hand, seeks to return an object as far as possible to
its original state. In the case of tapestries this can be very disruptive to the structure and to the
original material, and sometimes very confusing historically as well. When conserving a tapestry,
therefore, it is sometimes appropriate to give it support but not to try to reconstruct any of the
design. Generally, however, in order for the tapestry to continue to fulfil its function as a pictorial
piece, the flow of design should appear uninterrupted to the eye of the viewer. How this is
achieved depends on the particular tapestry, its age and design, its strength and the wishes of
the client."... (Borg Clyde 1992, p.156)

A selection of some articles describing conservation treatments and ethics is given here in
chronological order.

Drysdale (1983): the conservation of a 16th Century Flemish Tapestry in the Poldi Pezzoli
Museum in Milan. The tapestry was in fair condition; the silk was fragile, the brown wool brittle
and the metal thread tarnished. It was given a full support of linen scrim. Bare warps were
couched with a neutral coloured wool, including the silk areas. In darker areas of colour, a dyed
wool was used. ..."At first it seemed alien to use wool in silk areas, and to have so limited a
colour range. However despite its fraying, the softness of the wool causes the least possible
disruption to the dessicated silk surface, and its width helps to hold down chipping threads. The
matt quality of the wool goes well into the silk which has largely lost its sheen. Polyester thread
would be too harsh, too thin and too shiny, though more enduring than wool, silk too shiny and
too short-lived in light, mercerized cotton again too shiny."... (Drysdale 1983, p.38). Slits were
restitched using Gutermann's Mara polyester thread.

Simon (1987): the conservation of a Rubens tapestry (1975-1986). The tapestry was in very
poor condition and there was an extensive loss of silk. The tapestry was therefore given a full
support of a cotton rep fabric (pre-shrunk at 60° C), using vertical support lines of 20cm long and
10cm apart together with laid and couch stitching for securing weak and damaged areas.
Marko (1990): the conservation of a seventeenth-century Antwerp tapestry and an eighteenth-
century English Soho tapestry. Linen patches were used to support weak areas and dyed cotton
ribbed fabric for the infilling of missing areas.
..."How to support the resulting vulnerable silk area. Generally speaking, our conservation of
tapestries involves the use of suitably coloured threads worked across single warps to indicate
missing portions of design or to strengthen weak areas. Every other warp is caught down in the
first row of stitches, the alternate warps are secured in the second row, and so on. Regular
intervals separate these rows of stitching; however, the distance between rows may vary
according to the fineness of the original weaving, the extent of the damaged area, and
sometimes, the color that is being worked. Our method aims not only at providing support, but
also at maintaining the quality of the original design as much as possible without resorting to
restoration. Following from this, it may be necessary, for example, to work supporting stitches
closer together in a coarsely woven tapestry to avoid a 'thin' appearance in a given area.
Likewise, when working with dark-colored wools, care is taken to avoid creating a 'spotty'
appearance that can attract the eye of the viewer and become a dominant feature of the
tapestry."... (Marko 1990, p.98)

Biddulph (1988): the conservation of the Sheldon tapestries at Hatfield House, woven at
Barcheston Manor in Weston (Gloucestershire, UK) around 1611. In the 19th and early 20th
Century this series was repaired by the household staff, who stitched whole groups of bare
warps together. One tapestry was restored at the Morris workshops in the 19th Century; dark
brown outlines were rewoven with a fashionable gingery brown colour. In the late 1980's, the
series was conserved by a group of volunteers under guidance of Joan Kendall (a trainee of
Ksynia Marko). The tapestries were given a full support of pre-shrunk linen scrim. ..."A
generous amount of scrim is used, allowing the tapestry room to move and thus reducing the
strain on the threads. [...] To conserve the tapestries, the seamstresses stitch across the bare
warps in parallel lines one eight of an inch apart. Bare warps are stitched down to the scrim
using either one thread of Appleton's Crewel wool for wool areas, or two toning threads of special
polyester fibre spun with the same spring as silk (the Victoria & Albert Museum has
demonstrated that this is more long-lasting than silk). Where there are holes, small patches of
100 per cent cotton rep are stitched to the back."... (Biddulph 1988, p.114/115). Dyed threads
were used for stitching in areas where the original colours could be identified. In areas where the
original was unknown, warp-coloured threads were used.

Neugebauer (1991): tapestry conservation of the Adalbero-Gobelin of Stift Lambach. Small


holes and areas of damage were supported with woollen patches, using silk thread for stitching.
Lager holes were given an infill of neutral coloured woollen rep fabric, painted with the missing
design, using aquarel/gouache paints. The painted section was 'varnished' with methyl cellulose.

Blyth (1992): the conservation of the Unicorn tapestry, a Flemish 15th Century piece in the
collection of the Victoria & Albert Museum. It was given a full support of linen scrim, onto which
weak areas were couched using woollen yarns. Missing warps were replaced with specially
dyed woollen yarns and slits were strengthened with polyester thread. Many old repairs of holed
areas, which were previously rewarped with cotton and rewoven, were removed. New warps of
wool yarns, dyed to match the original warp, were inserted and couched down using woollen
yarn, dyed to match the colour of the original background. The couching stitching was set more
closely together than usual, to minimise the speckled look caused by the dark replacement wool
on the lighter coloured warps.
The author puts some questions to the chosen conservation technique: ..."As the tapestry has
already received extensive reweaving, should it have been further subjected to extensive
conservation repairs?"... (Blyth 1992, p.10/11). However, from a curatorial point of view, the time
spent on the conservation seemed to be warranted. The old repairs were ugly and badly
executed. The removal of several repairs significantly improved the appearance. The only
aspect of the treatment that the author may have chosen differently was using a patch of blue
rep woven fabric as a support patch underneath areas where the reweaving was removed,
instead of rewarping and couching.
Borg Clyde (1992): the conservation of the Roman de la Rose tapestry (woven in Tournai around
1460) at the Textile Conservation Centre in Hampton Court Palace (1985-86).
The tapestry was given a full support of linen scrim. The support had an allowance in the width
to be taken up by the conservation stitching. This allowance was not added in the height of the
tapestry because the support had to carry the weight of the tapestry. Large holes in intricate
patterned areas were supported by a supplementary dyed linen scrim patch between tapestry
and support and rewarped using dyed woollen yarn to match the original. The new warps were
secured by warp couching using wool and cotton. Large holes in areas of a solid colour were
masked by inserting a suitably dyed woollen patch. The conserved tapestry was fully lined with
cotton twill fabric.

Brutillot (1992): the conservation of the Anbetung der Konige und Darstellung im Tempel, an
early 15th century tapestry with a linen warp and wool and linen weft. The brown wool was
severely degraded due to the iron-mordant used for dying. The tapestry was given a patched
support of dyed linen patches. Damaged areas were secured by laid couching using dyed silk
thread. After conservation the tapestry was fully lined with cotton fabric. The tapestry was given
extra support during display by putting it on a sloping board.

Maes (1993): the conservation of the tapestries of the 'Patrimonio Nacional'. A project
concerning twenty tapestries, of which sixteen were dating from the 16th century. Maes gives an
overview of three different techniques. The first, i.e. restoration by rewarping and reweaving,
was only considered for small holes, where the design was not open to interpretation and the
surrounding parts were in wool, strong enough to anchor the new warps. Maes criticises the
technique by stating:
..."It is evident today that all artistic creations reflect, even to the smallest details the era in which
they were created. Each attempt to recreate a work of the past bears the mark of the author and
his time."... (Maes 1993, p.117)
Reweaving was also seen as a severe technique when applied to a fragile piece. The rewarping
in fragile silk areas was felt to cause destruction of the weaker threads.
The second technique was conservation, aimed to halt the degradation of weaker parts of
weaving and to conserve the remaining parts, while ensuring that any intervention would remain
reversible. A patched or full support of bleached linen fabric, which had been dyed in a suitable
colour, was used. The surrounding area of a damaged part was first consolidated with sewing
lines in a check pattern. Then the area of damage was treated; all warps were fixed to the
support with a neutral coloured silk thread.
The third technique used is referred to as 'integration-conservation', a technique which held the
decorative aspect of a tapestry in mind. It had two aims, to fulfil the objectives of consolidation
and reversibility and to re-integrate damaged parts. Again a dyed support fabric (full or patches)
was used. The surrounding area was first consolidated as above. In small holes the warps were
stitched with silk threads in the colour of the original weft. In larger holes, 'remote' reweaving
was carried out. The surrounding fragile areas were left untouched and newly inserted wefts
were fastened to the consolidation linen at regular intervals. Not tightly rewoven, but clearly
separated.

The technique of 'remote' reweaving is in detail described by Maes (1990):


..."This technique works as follows: each new woven line begins with a fixation stitch in an
undamaged part of the tapestry at the top of the gap and ends at the bottom in another
undamaged part with another fixation stitch. The eight-filament silk threads are woven up and
down through the warp and secured every 10 cm by stitches through the consolidation textile.
Then, returning from the bottom to the top, the threads are secured again at the same intervals,
with stitches overlapping each other every 10 cm. Reweaving at this distance created, in this
case, a density of 3-4 lines/cm; the resulting structure of the textile inhibits the woven vertical
lines from moving to the left or right and ensures that the warp is securely anchored to the
consolidation textile. The new woven structure created in this manner is essentially lighter than
the original or than the weft structure created by reweaving with normal density. A lighter
structure is better for the conservation of the surrounding parts and allows conservation of more
original areas; the heavier the new weft, the sturdier must be the original part in which it is
anchored."... (Maes 1990, p.75)
After consolidation the tapestries were lined with pre-shrunk linen fabric.
Maran (1993): the conservation of a 16th Century Flemish tapestry fragment, with a woollen
warp and wool and silk wefts. The brown woollen weft was degraded, and the silk weft had
become weak and fragile. The original slit stitching carried out with linen had weakened.
There were also many old repairs; degraded areas were once cut out and rewarped with cotton
threads and rewoven. There were also stitched repairs on patches and infills of tapestry weave
patches. These patches were removed prior to wet cleaning. The fragment was then given a full
support of linen scrim. Areas of damage were couched with Appleton wool for wool areas and
stranded cotton for silk areas. Areas of loss were rewarped and couched. Gutermanns
polyester was used to restitched slits. The tapestry was lined with pre-shrunk cotton sateen.

Masschelein-Kleiner (1993, 1994) describes the developments in tapestry conservation at IRPA


(see also 4.3.3). At that time reweaving was not carried out anymore. The conservation
technique was now preferred because of respect for the original, a minimum of intervention and
the possibility of total reversibility.
..."Gradually during the last decade reweaving has been eliminated because of a number of
negative findings: (1) the new wool is much softer and more hairy than the original and does not
mesh perfectly with the old parts; (2) anchorage of new warps creates tension and loss of
original threads; (3) uneven fading of original and new materials may occur; (4) the new weaving
would be difficult to remove should it become unattractive as it ages; (5) reweaving of lost details
often implies alteration of the original design. [...] Consolidation begins with the sewing of open
slits with blanket stitches from the back of the tapestry. Fragile areas are reinforced by relining
them with linen patches. [...] using laid and couching technique. [...] We found, however that this
procedure did not blend well with the texture of the tapestry and also was not strong enough to
sustain the weight of the tapestry when hung. After several mechanical tests, we decided to
replace the laid and couching technique by alternating rows of stitching that follows the weft. Silk
threads are chosen for sewing despite their sensitivity to UV light and stretching. Other fibres
exhibit even more disadvantages: (1) wool lacks the desired luster and has a long extension at
break; (2) mercerised cotton is much stronger than the original old fibres and has a tendency to
shrink; (3) polyester is even stronger and, in contrast to the old natural fibres, is not at all
hygroscopic; and (4) nylon is very damaging because of its scissoring effect. [...]
Patches and other linings are made of thin linen fabric weighing about 200 grams per square
meter. Flax is the natural fibre with the highest breaking strength and the lowest extension at
break, even after aging. These qualities are necessary to ensure that in time the lining will not be
hanging from the tapestry instead of the reverse. [...] We thoroughly boil the fabric before use to
eliminate any finishing products and to relax the internal tensions that are responsible for
shrinkage. We dye the patches to match the colors of the surrounding area of the tapestry."...
(Masschelein-Kleiner 1993, p.74-75)
A full lining was applied after consolidation. This lining had to provide support and should protect
the reverse of the tapestry from dust. Similar fine white linen (pre-shrunk) as used for the
patches was selected. When the condition of the consolidated tapestry was considered to be
sound, the lining was attached while the tapestry was hanging, with a conservator on either side
of the tapestry. By passing the needle from back to front, they stitched the lining to the tapestry.
Staggered vertical lines of running stitches, over one warp on the front and approximately 2cm at
the back, were made. The vertical lines were 45 cm long and 15 cm apart. Along the top more
lines were made to assist in carrying the weight of the tapestry. The bottom of the lining was
kept loose.
Shephard (1996) describes in an interesting case-study the conservation of two partially restored
18th Century Brussels tapestries. (see also 4.2.) ..."In october 1993 both tapestries were taken
out of storage for a condition assessment and to devise a method whereby the restoration
reweaving could skillfully integrated into contemporary conservation techniques. The restoration
had stopped abruptly in a straight line running from top to bottom on both tapestries. [...] The
unrestored sections of each tapestry were in a seriously degraded condition and were failing fast.
Additionally, previous repairs had been carried out at different times, probably during the 19th
Century. They consisted of patches of medium-weight linen applied randomly to the back of the
tapestry and stitched intensively using thick linen threads. [...]
The philosophy of the TCS is to carry out work that preserves both the visual integrity of the
tapestry and the structure of the weave using contemporary, acceptable conservation techniques
and materials. The stitching techniques employed allow the tapestry's imagery to be preserved
and enhanced from a viewing distance whilst, on close examination, any conservation work can
be clearly identified and, if necessary, removed. [...]
It was decided that the 19th Century repairs would all need to be removed prior to conservation.
[...] Some of the smaller, isolated areas of reweaving were also removed where the colour,
texture and technique were unsuitable, providing that evidence of the original weft existed and
that the removal itself did not result in further damage to the surrounding structure. [...] It was
decided to continue with some reweaving in the areas where the warps were completely
exposed and devoid of any original wefts. This reweaving would continue for a short extent at
the same closeness as the earlier reweaving but would gradually be opened out between each
weft passing to gradually resembling the technique of conservation couching. [...] After the
completion of the reweaving stage a fine linen holland cloth (15 ends per cm), pre-shrunk and
scoured to remove any dressing, was applied to the back of the tapestry. [...] The conservation
work was carried out using accepted, contemporary methods, whereby the linen cloth, applied in
sections over the entire back of the tapestry, distributes the weight and provides overall support
to all areas, whilst responding to any movement of the tapestry. A calculated amount of excess
of support fabric was allowed in the direction of both the warp and weft of the tapestry. This
excess is taken up by the conservation stitching and by the varying thickness of the weft ends on
the back of the tapestry. The linen support was attached in 15cm wide sections and the
conservation stitching was carried out within these sections, the closeness of the lines of
stitching varying according to the specific areas, the weaker weft requiring more intensive stitch
work than stronger wool weft areas. Lines of stitching were between 2 and 20 mm apart. [...] In
areas where little or no work was required, lines of widely spaced, regular, straight or zigzag
stitching were applied. [...] Broken or missing warp threads were replaced or strengthened with
wool yarn matched in thickness, ply and twist to the original. Replacement warp yarns were also
dyed to match the original, although in areas where there were large areas of missing warps, i.e.
in the outer selvedge galloon (guard), they were colour matched to that of the weft, thus
eliminating the necessity of close couching and providing a more visually acceptable
appearance."... (Shephard 1996, p.722-725)
4.5 Techniques and materials used in conservation/restoration

In the previous chapters, a great variety in different techniques and materials used in
conservation and restoration has been mentioned. In this chapter two issues will be discussed in
greater detail, as they appear in the literature to be subjects of discussion amongst tapestry
conservators and restorers. The first issue is the use of straps as a means of supporting
tapestries and secondly the choice of material used for the support (patched or full) of tapestries.

4.5.1 Strap support of tapestries

In some articles the use of straps as a method of providing overall support to tapestries is
discussed.

Hutchison describes in the case-study of the treatment of Glutony and Avarice (1990) the
application of straps. About the decision to use this type of support system, Hutchisons says:
..."Strap supports are designed to secure the tapestry without binding it; they allow for a certain
amount of fluctuation in the environment. Considering the age and condition of most tapestries,
a strap support is usually warranted. In the case of Gluttony, nineteen 7.5 m tapes made of
washed cotton in a herringbone twill were positioned approximately 30.5 cm apart across the
back of the tapestry. These tapes were allowed to hang freely from the back. While two people
applied a slight amount of opposing tension across the face of the tapestry, the tapes were
pinned in place, starting at the top and working out from the center tape and continuing down the
back of the tapestry in that order. The tapes were then attached with heavy-duty cotton thread in
a continuous running stitch down each side of the tape. The stitching passed over one warp of
the tapestry approximately every 5 cm. [...] Over the last fifteen years I have observed tapestries
with straps, some of which have been on continuous display. I have seen no evidence of the so-
called swag effect that is sometimes raised as an objection to the use of straps. Another
objection is that they form bands of shadow on the face of the tapestry. This same case can be
made against patches, namely, that they create shade distortions, especially by trapping
environmental dust. The use of a dust cover in conjunction with support straps has overcome
this argument against their application. The straps also offer the advantage of easy access to
the back of the tapestry. A further advantage is that, in event of moisture, a pocket of air
between the dust cover and the tapestry protects the object from direct contact with
dampness."... (Hutchison 1990, p.94)

Kajitani (1976) also favoured the application of straps in order to strengthen the tapestries after
conservation. She described the use of non-stretching straps which were secured by zigzag and
back stitches to the back of the tapestry, after which a loose lining of a light weight fabric was
attached along the top and sides of the tapestry.

Marko (1995) comments on the use of straps:


..."This method consists of spaced vertical straps of varying widths of closely woven cotton or
linen, attached to the reverse by stitching through to the front face. The straps are used as a
method of support, and are designed to take the weight of the hanging tapestry without adding to
it the weight of a full support. They do not carry repairs and so can be easily removed for future
treatment and cleaning. However, strapping is really only viable if used on a reasonably sound
object and if applied with an understanding of its natural movements, but so often these spaced
straps cause uneven tension resulting in a 'festooned' effect. A second problem occurs in the
filtration of airborne dust [...] creating a striped appearance which is exaggerated if the tapestry is
unlined."... (Marko 1995, p.ii)
4.5.2 The choice of material for supporting tapestries

The argumentation of which material to select for supporting a tapestry is given only in a few
articles . Following are some of the remarks made in the literature.

In America the use of cotton has been more widely accepted than for example in Europe. Two
American textile conservators give the following arguments on their choice:

..."Two issues raised by this discussion deserve additional remarks. The first regards the use of
cotton as opposed to linen for the support backing of textiles. Obviously, it would seem desirable
to use like fabrics as backing for textiles - i.e. silk for silk, linen for linen, etc. - because they
would respond in kind to environmental changes; however, some fabrics do not provide the
needed supportive qualities. For a historic tapestry of wool or silk, certainly neither of these
fabrics would be practical as material for support backings. For large historic textiles, we often
use cotton instead of linen support backings because the differences in their relative
hygroscopic properties are virtually inconsequential in the United States, where most art objects
are destined for environments that are at least moderately controlled. Even in poorly controlled
environments, cotton does not react as quickly as other fabrics to seasonal fluctuations, which
can be significant in the United States. Good-quality cotton is readily available in the United
States and is therefore the practical choice...." (Hutchison 1990, p.93-94)

..."When my predecessors noted that the tapestries imported from the Old World had been
strapped and lined with linen cloth, they assumed that linen was preferred to cotton, and they
conscientiously followed this European tradition, importing the linen goods from abroad.
Actually, linen was used in Europe not by choice but by necessity because it was the only cloth
available domestically. As the industrialisation of textile manufacturing progressed and labor
costs soared, linen industries began to economise by using inexpensive, harsh chemicals to
hasten the process of extracting fibres from the flax stalks, instead of following the time-
consuming and labor-intensive natural retting process.
The linens used for conservation work in the 1950's were probably made of such industrially
processed cloth. This explains why these linens have, after only thirty years, gone to shreds on
those tapestries that have been hanging in the galleries continuously from this period. [...] By
contrast, the linen linings and webbing tapes used during the eighteenth and early nineteenth
century are still in good condition. Jute webbing tapes and silk sewing threads likewise break
down, usually within a twenty-year period. [...]
In 1966 we began to use cotton cloth - prewashed for dimensional adjustment and sizing
removal - for strappings and linings; cotton or synthetic fiber for webbings of hangings; and
cotton, linen and synthetic fibre threads for sewing."... (Kajitani 1990, p.59-60)

The choice of material is in Europe only very rarely explained, like in Dolcini's article on the
conservation of the tapestries of the 'Sala dei Duecenti' in the 'Palazzo Vecchio' in Florence:
..."We also considered the possibility of using cotton, but the results of the application tests
carried out by the conservators were disappointing. The fabrics were tested for weight, density,
resistance to traction, and stretching. On the basis of these we chose the best-quality linen
available."... (Dolcini 1990, p. 86)
During the UKIC conference Lining and backing: the support of paintings, paper and textiles the
necessity of more scientific research into the materials used in the conservation of textiles is
made predominant in the article of Brooks et all.:
..."In textile conservation the aim of support is to enhance the long-term preservation of historic
textiles by 'strengthening' damaged or deteriorated areas and stabilising fragile and vulnerable
parts. The objectives are to enable them to be handled, displayed and/or stored safely.
Imparting dimensional and structural stability are also stated objectives. [...]
It is important that a support fabric has sufficient strength to carry the weight of the textile. de
Graaf discusses the factors to be considered, stressing that it is critical that the weight of the
textile is distributed evenly. [...] Compatibility, in terms of fibre type, construction, colour and
transparency/opacity, is also important. There is a strong body of opinion arguing in favour of
'like with like', i.e. the use of support fabrics which are made from the same fibre type as those of
the historic textile. [...] The choice between natural or synthetic fibres has also long been a
contentious issue. [...] synthetic fibres do not correspond to the behaviour of historic textiles, and
as a result could lead to stress and damage to the textile and failure of the support. In contrast,
there is concern amongst some conservators over the long term stability of certain natural
materials. [...] Conservators may elect to use synthetic fibres and fabrics in such situations (very
variable RH) because they are considered more durable. [...] A review of textile conservation
literature confirms that although much attention has been given to the choice of adhesive
techniques and the choice of support mechanisms in general, some important aspects in
determining the choice of a support system would benefit from a more informed debate. For
example, a wide range of reasons are give for the choice of materials, but few of these are
underpinned by objective testing of their relative benefits in terms of fibre, construction and
behaviour. [...] Although this experience is valid, there is an increasing awareness within the
textile conservation profession of the need for accurate and reproducible scientific data to enable
conservators to make better informed choices about support techniques."... (Brooks 1995, p.7-
11)
Part II

Chemistry and Physics of Flax (linen) and Cotton

1 The textile fibres

This chapter discusses the chemical and physical properties of flax (linen) and cotton. As
mentioned before in the general introduction, the first research period was narrowed down to a
comparitive study in these two materials. Flax and cotton were selected for this study, as the use
of these materials as support fabrics for the conservation of tapestries has been a point of
discussion for many years amongst textile and tapestry conservators. The following information
is given as background information to the scientific research project discussed in part III.

1.1 Flax

1.1.1 Production

Flax is a bast fibre, coming from the inner bark of the stem of the annual plant Linum usitatissi-
mum, which grows in many temperate and sub-tropical regions. The flax plant can reach a
height of 90 to 120 cm. It has a single slender stem which lacks side branches other than those
which bear the flowers. When the plants have flowered and the seeds are beginning to ripen,
the crop is pulled up by the roots (by hand or by mechanically pullers). About one-quarter of the
stem consists of fibre strands, which are composed of long, slender, thick-walled cells. The flax
fibres are held together in the stems by woody matter and cellular tissue. The fibres are freed
from this matter by a fermentation process called 'retting'. Retting can be carried out by different
methods, e.g.:
- Dew-retting: the fermentation of the crop by moulds is encouraged by wetting it with dew or rain
for several weeks. Dew-retting tends to yield a dark-coloured fibre.
- Tank-retting: the seed bolls are stripped from the stems and the stems are tied in bundles,
which are then packed into concrete tanks. The tanks are filled with water and artificially heated
to about 30°C. This retting process is completed in about 3 days and produces some of the best
and most uniform fibres.
- Chemical retting: the flax straw is treated with chemical solutions, like caustic soda, sodium
carbonate, soaps and dilute mineral acids.

The next stage in the processing of flax is 'breaking'; i.e. the stems are passed between fluted
rollers in a breaking machine, so that the woody core is broken without damaging the fibers. The
woody matter is then separated from the fibres by beating the stems with a blunt wooden or
metal blade on a so-called 'scutching' machine. After scutching the fibers are usually combed or
'hackled' by drawing them through sets of pins, each successive set being finer than the previous
one. The aligned fibres are collected into a loosely-held rope of fibre, called a 'rove'.

The spinning of the rove can be carried out dry for coarser yarns and wet for finer yarns. In wet
spinning, the rove is led through a bath of hot water, which softens the gummy matter holding the
fibre strands together, enabling them to be drawn out and aligned more perfectly as the rove is
elongated and twisted.
1.1.2 Structure and properties

The fibre cells of flax vary in length from 10-60 mm with a diameter of approximately 0.02 mm
(Cook 1993, p.9-10). The cells are long, transparent cylindrical tubes with swellings or 'nodes' at
many points and they show characteristic cross-markings.

Fig.1
Microscopic image of a
flax fibre (after Cook 1993, p.12)
Compound % of weight Source

Cellulose 65 Beurteilungs. (1986)


71.2 Hearle (1963)
71.3 Cook (1993)
64.1 Sommer (1960)
Hemi-cellulose 16.0 Beurteilungs. (1986)
18.5 Hearle (1963)
18.5 Cook (1993)
16.7 Sommer (1960)
Pectines 3.0 Beurteilungs. (1986)
2.0 Cook (1993)
1.8 Sommer (1960)
Protein 3.0 Beurteilungs. (1986)
Lignin 2.5 Beurteilungs. (1986)
2.2 Hearle (1963)
2.2 Cook (1993)
2.0 Sommer (1960)
Wax and fats 1.5 Beurteilungs. (1986)
1.7 Cook (1993)
1.5 Sommer (1960)
Moisture 8.0 Beurteilungs. (1986)
10.0 Sommer (1960)
Ashes (Mineral) 1.0 Beurteilungs. (1986)
other - Beurteilungs. (1986)
8.1 Hearle (1963)
4.4 (water soluble) Cook (1993)
3.9 (water soluble) Sommer (1960)

Table 1 Main components of flax


Flax fibres have a higher crystallinity than cotton, which is crystalline for 70%.

Flax is a strong fibre with an average tenacity of 5.8 g/dtex (= 6.4 g/denier). (Cook 1993, p.10)
The strength of flax increases when wet as reported by Hollen (1988, p.12); the tenacity when
dry is 3.5-5.0 g/denier and 6.5 g/denier when wet.

Flax is a particularly inextensible fibre. It stretches only slightly as tension increases. The
elongation at break is approximately 1.8-2.0 % dry, and 2.2 % wet (Cook 1993, p.10, Hollen
1988, p. 13). Within its small degree of stretch, flax is an elastic fibre. It will tend to return to its
original length when tension is relaxed; it has a 65% recovery from a 3% stretch (hollen, p.15). It
has a high degree of rigidity and resistance to bending. (Cook 1993, p.10)

Flax has a moisture regain figure of 7-12% (expressed as a percentage of the moisture-free
weight at 20°C and 65% RH). It has the ability to absorb water rapidly.

It is highly heat resistant to decomposition up to about 120°C, when the fibre begins to discolour.
Flax will withstand dilute, weak acids but is attacked by hot dilute acids or cold concentrated
acids. Flax has a good resistance to alkaline solutions.
When boiled and bleached, flax is virtually pure cellulose, having a high resistance to rotting.
Under severe conditions of warmth, damp and contamination, however, mildews may attack the
cellulose of flax, but resistance is generally high. (Cook 1993, p.11)

1.2 Cotton

1.2.1 Production

Cotton grows inside the seed pods of a wide variety of plant species in the Gossypium family.
The cotton fibres are attached to the seeds of these plants, probably serving to accumulate
moisture for the germination of the seed. Nowadays, the worlds cotton is grown by raising
annual crops (sown in spring), the plants reaching a height of between 1.2 and 1.8m. The plant
develops many cream-white flowers, which turn pink towards the end of the first day. On the
third day the flower dies to leave a small green seed pod or boll. When it reaches maturity, the
boll bursts and the cotton appears as a soft wad of fine fibres, attached to the seeds inside the
boll.

Cotton is usually picked in the autumn by hand or by machine. After picking, the cotton fibres are
mechanically separated from the seeds (by a 'gin' machine). The 'ginned' cotton, often referred
to as 'lint' is pressed and packed in bales. Before the actual spinning of the cotton, the fibres are
subjected to several preliminary processes. These include opening up of the tightly packed
cotton bale and cleaning of impurities such as dust, leafs and twigs.
The next process is called 'carding', carried out by a carding engine, which separates the cotton
fibres, taking out most of the remaining impurities and removing short and immature fibres.
From the carding machine, the cotton is delivered as a filmy web, which is collected to form a
loose rope, the 'sliver'. Slivers are stretched or drawn in stages by passing them through a
drawing frame. If the cotton is to be made into fine or high-quality yarns, it is combed to align the
fibres more accurately and to remove more of the shorter fibres. The stretching and drawing is
continued on other machines such as the 'slubber' and the 'roving' frame. (Cook 1993, p.49-51)
Cotton is normally spun into yarns and threads without undergoing any treatment other than the
mechanical processes mentioned above. A small amount of mineral oil may be added during
spinning, and yarns may be treated with a sizing. The non-cellulosic materials will remain in
cotton and if the yarns, threads or fabrics are furthermore untreated, they are known as 'grey
goods'. Before these grey goods can be dyed and finished, they are cleaned and purified. They
can undergo an alkali treatment to remove the wax from the surface of the cotton fibres. The
cotton is boiled in a pressure vessel with a solution of caustic soda at about 118°C. Cotton is
sometimes bleached. In recent times, hydrogen peroxide has been used increasingly as cotton
bleach instead of the traditional chlorine compounds. (Cook 1993, p.54-55)

1.2.2 Structure and properties

The individual cotton fibre is a seed-hair consisting of a single cell. Each cotton seed may
produce as many as 20,000 fibres on its surface, a single boll will contain 150,000 fibres or more.
The fibre length varies, depending on the variety of plant, the conditions under which it has been
grown and the state of maturity at the time of picking. Lengths vary from 6 to 65 mm. In general
the fibre is fairly uniform in width with an average fibre diameter of 11-22 mm (Gohl 1993, p.41).
The primary wall is a tough, protective layer formed during the early days of the fibre growth. It
contains wax, protein and pectinaceous substances, as well as cellulose. The secondary wall
forms the bulk of the fibre. This is cellulose that is laid down during the second stage of growth,
after the fibre has attained its full length, when consolidation of the cell wall takes place. Layers
of cellulose are added one after another to the thin cellulose membrane from inside the cell.
Each growth-ring corresponds to a day of growth and cellulose-deposition. The cellulose is laid
down in the form of spiral fibrils, some 1,000 or more to each ring. (Cook 1993, p.62)

Fig.2 Morphological diagram of cotton after Gohl (1993, p.43)


The main components of cotton are:

Compound % of weight Source

Cellulose 87.0 Beurteilungs. (1986)


91.8 Hearle (1963)
95.3 Sommer (1960)
Hemi-cellulose - Beurteilungs. (1986)
6.4 Hearle (1963)
Pectines 1.2 Beurteilungs. (1986)
0.99 Sommer (1960)
Protein 1.3 Beurteilungs. (1986)
Lignin - Beurteilungs. (1986)
- Hearle (1963)
Wax and fats 0.6 Beurteilungs. (1986)
0.75 Sommer (1960)
Moisture 7.0 Beurteilungs. (1986)
7-11 Hollen (1988)
8.5 Cook (1993)
7-8 Morton (1993)
Ashes (Mineral) 1.2 Beurteilungs. (1986)
0.86 Sommer (1960)
other 1.7 Beurteilungs. (1986)
1.8 Hearle (1963)
1.10 (sugar, org. acid) Sommer (1960)

Table 2 Main components of cotton

Cotton has a crystallinity of about 70% (Hearle 1963, p.392).

The cotton fibre undergoes a metamorphic change when the seed boll opens. Until this
happens, the fibres maintain a tube-like appearance. But as the boll opens, the moisture
evaporates from the fibres. As the liquid disappears, it leaves an almost empty channel, the
lumen. As a result, the cell wall collapses, forming a ribbon-like structure. The fibres not only
collapse but also twist lengthwise forming convolutions which are very characteristic of the cotton
fibre. These twists take place in both directions in the fibre. The number of convolutions varies
greatly, on average a fibre will have some 50 twists per cm.

Cotton is a moderately strong fibre. Its strength increases when wet as reported by Hollen
(1988, p.12); the tenacity or tensile strength when dry is 3.5-4.0 g/denier and 4.5-5.0 g/denier
when wet. It has an elongation at break of 3-7% (at 65% RH and 70F), which increases
considerably when the fibre is wet to 9.5%. The fibre is fairly elastic; at 3% extension it has an
elastic recovery of 75% (Hollen 1988, p.12,-15).
Cotton has a moisture regain figure of about 7-11 (expressed as a percentage of the moisture-
free weight at 20°C and 65% RH). It has the ability to absorb water rapidly.

Cotton has an excellent resistance to degradation by heat. It begins to turn yellow after several
hours at 120°C and decomposes at 150°C as a result of oxidation. It is attacked by hot diluted
acids or cold concentrated acids, in which it disintegrates. It is not affected by cold weak acids.
It has an excellent resistance to alkalis. It swells in caustic alkali (mercerisation) but is not
damaged. Cotton is attacked by fungi and bacteria in hot and moist conditions. Mildews, for
example, will feed on cotton fabric, rotting and weakening the material. Mildews are particularly
troublesome on cotton that has been treated with starchy finishes, and much of the damage can
be avoided by thorough scouring. (Cook 1993, p.69)

1.3 Comparing flax (linen) with cotton

Both flax and cotton are composed almost completely of cellulose.

1.3.1 Chemistry of cellulose

Cellulose is a linear chain of ß-glucose units, linked together by hemi-acetal linkages in the 1:4
position. The repeating unit in the cellulose polymer is cellobiose which consists of two glucose
units. Flax consists of about 18000 cellobiose units and cotton about 5000 (Gohl 1983, p.210-
211). This is referred to as the degree of polymerisation (Dp).

Fig.3 Chemical structure of cellulose (after Gohl 1993, p.45)

The most important chemical groups on the cellulose polymer are the hydroxyl groups (-OH) and
the methylol groups (-CH2OH). These polar groups give rise to hydrogen bonds between the -
OH groups of adjacent polymer chains. These hydrogen bonds are formed between covalently
bonded hydrogen and the strongly electronegative oxygen atoms on adjacent polymer chains,
which are less than 0.5 nm apart. Van der Waals forces are defined as very weak electrostatic
forces which attract neutral molecules to each other. Compared to the hydrogen bonds these
forces are of little significance.
Both flax and cotton are crystalline fibres. Each polymer tends to form part of several amorp-
hous and crystalline regions. This assists in the cohesion of the polymer system. When several
polymers are aligned or orientated in more or less parallel order they form a crystalline region.
Wherever the polymers are aligned at random or are not orientated, amorphous regions are
formed. These crystalline and amorphous regions do occur at random throughout the polymer
system of the fibre.

Fig.4 Fibre orientation (after Gohl 1993, p.29)

The estimated amorphous regions of cotton are approximately 30%, the crystalline regions 70%.
Flax is more crystalline than cotton because of its longer polymers (Gohl 1983, p.210-211).

1.3.2 Properties of flax and cotton

Flax and cotton have very similar properties as they consists both of cellulose. The major
differences are the degree of polymerisation and crystallinity, resulting in some differences in
their properties.

Both cotton and flax are strong fibres, attributed to the good alignment of their long polymers and
the formation of numerous hydrogen bonds between adjacent polymer chains. Flax is stronger
than cotton, as its more crystalline system allows the extremely long polymers to form more
hydrogen bonds than cotton polymers. The strength of both cotton and flax increases when the
fibres are wet. It is believed that this is a result of a temporary improvement in the polymer
alignment in the amorphous regions, causing the formation of more hydrogen bonds.

Both linen and cotton goods are crisp and stiff to handle. This is a result of their crystalline
nature. Linen goods will wrinkle more readily than cotton goods, as flax is more crystalline.
Cotton and flax are very hygroscopic, as a result of the numerous polar -OH groups in the
polymer chain which attract water. Water molecules can however only enter the polymer system
in the amorphous regions, as the interpolymer spaces in the crystalline regions are too small for
the water molecules. The swelling of the fibres is caused by the separation of polymer chains by
water molecules in the amorphous regions only. Cotton appears to respond more rapidly to
changes in the relative humidity. This is probably due to its shorter fibres and more amorphous
structure, which makes it easier for water molecules to enter the structure.

Both cotton and flax have the ability to conduct heat energy. Flax is the best heat resistance and
conductive of all commonly used textile fibres. There appears not to be a satisfactory
explanation for this.

1.3.3 Degradation of cellulose

Numerous studies have been carried out in the degradation processes of cellulosic material.
Following is a brief overview of some of these researches and their findings.

Both Scicolone (1996) and Feller (1986) describe the degradation of cellulose to take place in
two to three steps. There are several types of degradation of cellulose, e.g. thermal, hydrolic,
photolytic, photochemical and enzymatic degradation. They basically all result in scission of the
cellulosic chains. The degradative scission of cellulosic chains frequently appear to follow zero-
order chemical kinetics, i.e. the situation in which an equal number of bonds are broken in a
given length of time. Therefore, the degradative scission can be followed by measuring the
change of the so-called 'intrinsic fluidity', expressed as the inverse of the average degree of
polymerisation.

Scicolone looked at the kinetics of cellulose fiber degradation caused by light, heat and acid-
catalysed hydrolysis (heterogenous hydrolysis). Feller described the heterogeneous degradation
of cellulose by acid hydrolysis, by photochemical and by enzymatic attack. Both mention the fact
that the degradation seems to display two distinct stages: a fast initial stage followed by a
significantly slower one. ..."One may suggest that there are not two but at least three major
stages in the overall hydrolytic degradation of cellulosic fibers, extending from the initial attack to
the final point when all of the cellulosic material has been converted to soluble fragments."...
(Feller 1986, p.334). The rapid initial stage is considered to represent the breaking of a limited
quantity of 1,4-ß-glucosidic bonds which are known to be particularly sensitive.

The second stage proceeds at a slower rate and essentially represents the attack on the normal
1,4-ß-glucosidic bonds located in the amorphous regions of the fibers. It has been observed that
the degree of polymerisation of cellulose fell rapidly to a low value, designated as the 'levelling off
degree of polymerisation' (LODP), beyond which the last stage of degradation proceeds at an
extremely slow rate (LODP cotton DPn = 200-235). It represents the breaking of the cellulose
chains bound into the crystallite bundles or zones that usually make up a significant fraction of
the fibers. The levelling off point has been characterised at the physical level as the limit of
acceptable strength.

Scicolone compares the degradation rates of cotton and linen. The first step of degradation is
faster for raw cotton than for raw flax, while the second slower stage is characterised by a similar
rate for both.
Sook (1990) looked at the chemical reactions occurring when cellulosic material was exposed to
light. The specific chemistry of these reactions was found to be dependent upon the energy of
the light absorbed, the purity of the material and the presence of moisture and oxygen.
..."'Photosensitized degradation' occurs when impurities, eg, dyestuffs, metal ions, degradation
products, etc., in the material absorb near UV light at wavelengths between 340 and 400 nm and
transfer some of this absorbed energy to a cellulose molecule. Light of this energy cannot cause
chemical bonds to break, but can raise impurities in the cellulose to an 'excited state'. These
excited species can then induce the cellulose molecules to react with oxygen or other reactive
species in the environment."... (Sook 1990, p.302).
The mechanisms of photo-oxidation are not well defined, but it is thought that the formation of
free radicals is the first step. Carbonyl (-C=O) and carboxyl (-COOH) groups are created and
chain scission occurs.

Berry (1977) examined three techniques for artificially aging cotton fabrics: acid hydrolysis,
heating, and irradiation with high-voltage electrons. He found that, when comparing the strength,
extent of oxidation, crystallinity, molecular weight, and fracture mechanics of the degraded
fabrics with those of several pre-Columbian cottons, none of the artificially aged fabrics
duplicated the ancient samples in all properties. The artificial aging experiments revealed
several different degradation reactions. Two types of degradation can be caused by radiation:
photolysis produced by electromagnetic radiation (ultraviolet light is an example of electromag-
netic radiation) and radiolysis produced by ionizing radiation.
Cellulose can be hydrolysed by acids and by enzymes derived from bacteria and other micro-
organisms. Enzymatic hydrolysis breaks the 1,4-ß-glucosidic link, resulting in a decrease of the
molecular weight, involving a rapid loss in strength. The crystallinity is initially increased during
hydrolysis as severed chains in the disordered regions separate and reform into more crystalline
patterns. In general, acid hydrolysis produces similar degradation. Since acids usually involve a
much smaller molecule as the attacking agent, acids penetrate the amorphous area of cellulose
more easily than enzymes and produce a faster drop in the degree of polymerisation.
Thermal degradation follows different paths depending upon whether moisture and oxygen are
present. Above 140°C, however, moisture does not influence degradation.

Needles (1989) aged flax fibers by heating them at 180°C in air for periods of up to ten hours.
The heat-treated linen became progressively darker and more brown in colour and showed
progressively losses in tensile and abrasion properties. Wide angle X-ray scattering suggested
that the heat-aged linen was less crystalline than untreated linen. Scanning electron microscopy
showed that heat aging caused long crevices, longitudinal to the fiber axis. As a result the
accessibility and surface area of the fibers increased. It was however noted that heat-aged linen
dyed to lighter and slightly different shades as it had fewer dyeing sites available for direct dyes
than untreated linen. Therefore it can be concluded that flax oxidation during heating apparently
led to some breakdown of crystalline regions in the cellulose, but did not provide additional
dyeing sites. The loss in dyeing sites is thought to be due to heat-induced crosslinking of the
amorphous regions in the flax.
Needles concludes that heat ageing of cellulosic causes accelerated discolouration, loss in
tensile properties, increased fiber breakage, reduction in the degree of polymerisation, reduction
in the degree of swelling, in some instances increases in crystallinity, increases in the carbonyl
and carboxyl group content, decreases in moisture content, and reduced dye uptake of the fibre.
These property changes are similar to changes in natural-aged linen.

Hackney and Hedley (1993) examined the natural aging of linen canvas. The samples, both
unimpregnated linen canvas and wax/resin impregnated linen canvas (washed, 39 warp and 42
weft threads/inch), were prepared in 1956 and kept for 24 years at 'gallery' conditions in the Tate
Gallery in Central London. The samples were arranged so that some had been exposed to light
and some protected from it.
After aging, the tensile strength at the moment of breaking of individual yarns was measured.
The unimpregnated canvas samples exposed to light and to air circulation were the weakest of
all samples tested. Compared with the strength of the reference sample, the strength had fallen
from 2.98 to 1.21 and 1.12 kg, a loss in the order of 30-40% of their original strength in only
twenty-four years of 'gallery type' exposure.
Unimpregnated samples aged when protected from light, where stronger. The ratios of the
strength of a sample aged in the light to its strength when aged in the dark were 0.61 and 0.68.
When measuring the pH of the samples, there appeared to be some correlation between the pH
of the extracts and the break load of the yarns; however, other trends were also discernible.
Exposure to light weakened the canvas without a substantial increase in acidity, whereas
exposure to freely circulating air showed an increase in acidity greater than the corresponding
weakening effect. They concluded that apparently the pH depends on the mechanism of
deterioration, and may only be used as a guide to the condition of the yarns.

Conclusion:
When analysing al these studies into the degradation of cellulosic materials, one might conclude
that there are several differences as well as comparisons in the degradation behaviour of linen
and cotton. Scicolone and Feller showed that both materials first show a rapid decrease in the
degree of polymerisation until a levelling off point has been reached, after which the degradation
proceeds more slowly. They noticed however that the first stage of degradation was faster for
cotton than for linen. This might be caused by the fact that linen is more crystalline than cotton
and has a higher degree of polymerisation.
2 Physical properties

The behaviour of yarns and fabrics will depend on several factors such as the fibrous material it
is made of, the construction of the yarn (i.e. the amount of twist, is it a single or plied yarn, etc.)
and the interlacing of the yarns in the fabric (i.e. type of weave, knitting, braiding, etc.). The
behaviour is furthermore directly related to the temperature and relative humidity of the
atmosphere. It may be obvious that the issue of the physical properties of fabrics is a very
complex one.

In order to reduce the number of factors, the focus will be on plain weave fabrics only, made of
cotton or linen, as these materials are relevant to the topic of this research project. There have
been a few researches into the physical characteristics of these types of fabrics, especially in the
field of painting conservation in order to understand the behaviour of paintings on canvas. The
terminology used to express different mechanical properties will be explained, followed by an
account of the specific behaviour of fibres, yarns and fabrics when responding to changes in the
atmospheric conditions.

Please note that what will be said in the following about fibres is also applicable to yarns and
fabrics. However, an extra complicating factor is that in yarns and fabrics movements relative to
each other may occur (i.e. between fibres in the yarn and between yarns in the fabric), much like
the sliding of chain-molecules in the fibre. This may cause extra permanent deformation in the
yarns and in the fabrics.

2.1 Tensile properties

When looking at the behaviour of fibres put under a certain load, one can observe that most
fibers will behave in a similar way. In order to compare different fibres with each other, the load
applied can be expressed as gramforce or Newton per tex (gf/tex or N/tex) or per denier (gf/den
of N/den). This quantity is called the specific strength or stress and is also sometimes referred to
as tenacity and makes it possible to compare the strengths of different kinds of fibres. (Tex and
denier are measures of the thickness of a fibre; Tex is the weight in grams of 1,000 meters of
fibre or yarn, denier is the weight in grams of 9,000 meters of fibre or yarn).

In a stress-strain or load-elongation chart the specific stress or strength is plotted vertically


against the tensile strain, which is represented by the percentage of extension of the fibre (the
elongation) when compared to its initial length.
Fig.5 Hypothetical load-elongation curve (after Graaf 1980, p. 55)

In the beginning, the reaction of the fibre to a weight-load is represented by a straight slope; the
load applied is directly related to the elongation of the fibre. The fibre is still in its elastic region,
meaning that the fibre will return to its original length when the load is removed. The transition
between parts a and b of the curve is called the yield point, and beyond this point certain
structural changes take place in the fibre, causing changes in its physical properties. At this
point, any further elongation of the fibre will cause permanent deformation. This is because
fibres are not ideally elastic, but have so-called visco-elastic properties. If the initial load does not
exceed the yield point, the fibre essentially regains its original length (elongation 0%) when the
load has returned to zero. When the fibre does not regain its original length, a permanent
elongation remains. This permanent deformation is greater the higher the initial load has been,
respectively the higher the initial elongation has been. The reason for this is the fact that a
structural change has taken place in the arrangement of the chain-molecules in the fibre.
Beyond the yield point the cohesive forces between the chain molecules of the textile material
are not strong enough to prevent these to slide along each other. This slipping of chain-
molecules causes permanent elongation of the fibre. (Graaf 1980, p.57)
When the load applied becomes to great, the fibre will eventually break, represented by the end
of the curve.
Most fibres show the type of curve described above, when subjected to a load. Yarns, threads
and fabrics have a similar behaviour.

Fig.6 Stress-strain curves of different fibre types (after Graaf 1980, p.56)

Looking at the graph, one can conclude that flax is the strongest fibre, and also the most rigid.
Wool behaves totally different and only needs a very small load to extend to almost 1.5 times its
original length. (See also Table 3 and 4)

2.2 Time effects

Creep, recovery and relaxation are all phenomena characteristic to the visco-elastic behaviour,
caused by the sliding along each other of chain molecules in the material.

Creep is the time-induced phenomenon of the extension of a fibre under an applied load: the
complementary effect is called relaxation - the reduction of stress with time under a given
extension. In other words, when a fibre is held stretched, the stress in it gradually decays. It
may drop to a limiting value or may disappear completely.
The following figures depicts what happens to a fibre when stretched by a constant load for a
certain length of time.

Fig.7 Creep and recovery (after Graaf 1980, p.57)

At the first moment the load causes an elongation corresponding to the load-elongation curve.
The extension occurs at first relatively fast, later on slowing down. When the load is high with
respect to the breaking strength of the fibre (e.g. 3/4 of the breaking strength) then the fibre may
even break after some time, although the load applied was smaller than the breaking strength.
When the load is removed, the elongation of the fibre will partly decrease immediately and will
then shorten further, at first relatively fast and later on slowing down. The ultimate length of the
fibre may however be greater than the original length. The fibre has undergone a permanent
deformation. The total extension may be divided into three parts: the immediate elastic
deformation which is instantaneous and recoverable; the primary creep which is recoverable in
time; and the secondary creep which is non-recoverable.
Fig.8 Relaxation (after Graaf 1980, p.58)

When a fibre is kept under constant elongation during some time, the initially load to cause this
elongation will decrease, again in the beginning rather fast, but later slowing down. This
phenomenon is depicted in figure 8 and is called relaxation. Note that after relaxation the fibre is
still under a certain load and that this load is in its turn able to cause creep in the fibre.

Textile fibres possess the ability to recover their size and shape after deformation, a phenome-
non called elasticity. A single stress-strain curve taken from zero to rupture provides no
information about the recovery abilities of that fibre. Therefore, cycles of loading and unloading
are carried out. Depending on the weight of the load and the time applied, the fibre may either
totally recover from this deformation, although it will follow a different curve (hysteresis).

If however secondary creep has occurred, which is non-recoverable, the fibre will remain longer
than its original length. As a measure of tensile elasticity at any given stress or strain, the ratio of
the extension recovered in a reasonable time to the total extension at any given stress may be
taken as a representation of elastic recovery.
The extension of a fibre can be divided into three components as explained above; the
immediate recovery on the release of the stress, a delayed recovery called primary creep that is
recoverable after sufficient time and last the non-recoverable secondary creep. Elastic recovery
can therefore be defined as the ratio of the sum of the first two components to the sum of all
three components. It has been found that cellulosic fibres have an inferior recovery properties
compared to the animal fibres wool and silk. (Meredith 1956, p.75,76)

Fig.9 Cyclic loading curves at 65% RH and 20°C (after Meredith 1956, p.75)
2.3 Effect of humidity/temperature on the tensile properties

Several factors can greatly influence the tensile properties of materials. First of all aging, which
causes the breaking point to move back down the curve. Secondly the moisture content of the
fibre, which is related to the relative humidity of the surrounding atmosphere. Different types of
fibres have different responses to changes in the relative humidity, as is depicted in the curves
below. The tensile strength and the elastic recovery are influenced considerably.

Fig.10 Stress-strain curves at various relative humidities (after Morton 1993,


fig.13.26)

The following two tables (3 and 4) show that linen and flax become stronger when wet. This is
believed to be a result of a temporary improvement in the polymer alignment in the amorphous
regions, causing the formation of more hydrogen bonds. Cook explains it as follows: ..."The dry
fibres, constructed from its fibrils of cellulose, is a fairly stiff, rigid entity. The cellulose molecules
are held tightly together inside the fibrils, bound by bonds established between molecules lying
close alongside one another. Water is able to penetrate into the cellulose network making its
way into the capillaries and spaces between the fibrils and into less tightly bound areas of the
fibrils themselves and attaches itself also by chemical links to groups in the cellulose molecules.
Thus, water molecules tend to force the molecules of cellulose apart, lessening the forces that
hold the cellulose molecules together and destroying some of the rigidity of the entire structure.
Water acts as a 'plasticiser'. It permits the cellulose molecules to move more freely relative to
one another. The mass of cellulose is softened, and can change shape more easily under the
effect of an applied force."... (Cook 1993, p.66 )
Wool and silk become weaker when wet. When wool absorbs moisture, the water molecules
gradually force sufficient polymers apart to cause a significant number of hydrogen bonds to
break. Silk, which is as crystalline as the cellulosic fibres, also looses strength when wet. This is
due to water molecules hydrolysing a significant number of hydrogen bonds, which will weaken
the fibres.

Fibre Tenacity Source


dry wet

cotton 3.5-4.0 g/den 4.5-5.0 g/den Hollen 1988


cotton 3.0-5.0 g/den 20% stronger Cook 1993
cotton 0.45 N/tex 0.50 N/tex Morton 1993
flax 3.5-5.0 g/den 6.5 g/den Hollen 1988
flax 5.8 g/dtex 20% stronger Cook 1993
flax 0.54 N/tex Morton 1993
wool 1.5 g/den 1.0 g/den Hollen 1988
wool 1.0-1.7 g/den 0.8-1.6 g/den Cook 1993
wool 0.14 N/tex 0.10 N/tex Morton 1993
silk 4.5 g/den 2.8-4.0 g/den Hollen 1988
silk 3.5-5.0 g/den 75-85% of dry Cook 1993
silk 0.38 N/tex 0.35 N/tex Morton 1993

Table 3 The effect of humidity on the tenacity of fibres. (dry = 20°C 65% RH)
As the relative humidity increases, all the fibres become more extensible and the extension at
break becomes larger. This is due to the absorbed water reducing the cohesion of the chain
molecules in the amorphous regions of the fibre.

Fibre % Elongation at break Source


dry wet

cotton 3-7 9.5 Hollen 1988


cotton 5-10 Cook 1993
cotton 6.8 7.9 Morton 1993
flax 2.0 2.2 Hollen 1988
flax 1.8 2.2 Cook 1993
flax 3.0 Morton 1993
wool 25 35 Hollen 1988
wool 25-35 25-50 Cook 1993
wool 42.9 57.1 Morton 1993
silk 20 30 Hollen 1988
silk 20-25 33 Cook 1993
silk 23.4 38.1 Morton 1993

Table 4 Effect of humidity on the elongation of fibres. (dry = 20°C 65% RH)

The following table shows the effect of the moisture content and the temperature on the tensile
strength and breaking extension.

Ratio of values:
wet at 95°C /wet at 20°C
Fibres Tenacity Breakings
extension

cotton 1.00 1.00


flax
wool 0.55 1.37
silk 0.71 0.96

Table 5 Effect of humidity and temperature on the tenacity (Morton 1993, p.327)

If the relative humidity stays constant, over a range of 16°C to 32°C, the values of the moisture
regain are not affected significantly. However, if the absolute humidity remains constant while
the temperature increases, the equilibrium moisture regain decreases. (Ballard 1995, p. 17)
Different fibres exhibit different levels of recovery. Under low stresses - producing 1%
extensions - the level of elastic recovery is reduced at higher relative humidity. At higher
stresses - ones that produce 10% extensions - the level of elastic recovery is reduced still more,
to less than half of the recovery levels at 1% extension. Textile technologists note that with large
(10%) extensions, the reduced elastic recovery is slightly improved at 90% RH. (Ballard 1996,
p.668)

Fibres 1% extension 5% extension 10% extension


RH 60% 90% 60% 90% 60% 90%

cotton 91 83 52 59 - -
flax 90
wool 99 94 69 82 51 56
silk 84 78 52 58 34 45

Table 6 Effect of humidity on the elastic modulus


(after Morton 1993, p.327 and Gohl 1993, p.210-211)
3 Moisture in textiles

The ability of a fibre to absorb moisture depends on its molecular composition, i.e. the number of
hydrophilic groups, and the geometric arrangement of the molecules. Water molecules can only
enter the fibre in its amorphous regions. The molecules here are linked together only at the few
points where they are close to one another. Consequently, most of the active groups in the
polymer chains are available to the water molecules. The first water molecules must be
absorbed directly onto the hydrophilic groups present but, for those absorbed after this, there is a
choice. They may be attracted to other hydrophilic groups, or they may form further layers on top
of the water molecules already absorbed. In the amorphous regions the structure is
comparatively open and is therefore easily deformed (or swollen) by allowing water molecules to
enter.

The amount of water in the fibre is generally expressed as a percentage of the weight of the dry
fibre, when it is called 'moisture regain', though occasionally it is referred to the combined weight
of fibre and water, when it is called 'moisture content'.

Fibre moisture regain difference in de- Source


(20°C, 65%) sorpt. and absorpt.
regain (20°C,65%)

cotton 7-11 Hollen 1988


cotton 8.5 Cook 1993
cotton 7-8 0.9 Morton 1993
cotton 7 Beurteilung 1986
flax 12 Hollen 1988
flax 12 Cook 1993
flax 7 Morton 1993
flax 8 Beurteilung 1986
flax 10 Sommer 1960
wool 13-18 Hollen 1988
wool 14-18 1.2 Morton 1993
silk 10 2.0 Morton 1993
silk 11 Cook 1993
silk 11 Hollen 1988

Table 7 Moisture regain of fibres

The relation between moisture regain and the relative humidity of the surrounding air at constant
temperature is represented in a sorption isotherm. The isotherms of the natural fibres have a
characteristic sigmoid shape. This implies that relative large changes in relative humidity in the
region of 30-70 % have far less effect on the moisture content of a fibre than changes at
extremely high or low relative humidity. Howell (1996) gives an example of wool; ..."the
percentage change in moisture content of wool when taken from 30 to 60% RH is exactly the
same as taking wool from 82 to 92% RH."... (Howell 1996, p.692)
The amount of water held by a fibre at a given relative humidity is greater for a fibre which has
previously been in equilibrium with a higher relative humidity than for the same fibre having
previously been in equilibrium with a lower relative humidity. The desorption isotherm therefore
does not correspond to the absorption isotherm, but lies above it. This is called hysteresis. At
constant relative humidity the moisture regain of cellulose fibres varies only slightly with
temperature. (Meredith 1956, p. 35, 36)

Fig.11 Moisture regain isotherm (Collins 1939 , p.46)


One of the most important phenomena when examining moisture regain of fibres is the time that
textiles require to respond to changes in relative humidity. The approach to equilibrium is
exponential, i.e. there is a rapid initial reaction followed by increasingly slower changes. Also, the
smaller the change in relative humidity, the longer it takes to come into equilibrium. Howell
discovered the same in his experiments investigating the 'moisture response time' of new and
degraded silk. The moisture content of all samples took an extremely long time to come into
equilibrium, there still being measurable changes after three weeks. It took a day for 38% of the
change to take place, after which the rate of change became increasingly slow. (Howard 1996).

Ballard (1995) describes: ..."The most astonishing aspect of hysteresis is the time dependency
associated with it. That wool fiber can reach equilibrium at 65% from a dry state in 1.5 hours, but
only after 103 days from a wetted state. Similarly, cotton takes 2.5 hours or 99 days; silk, 5
hours and 39 days."... (Ballard 1995, p.16)

3.1 Fibre shrinkage

In order to understand the behaviour of fabrics when responding to changes in relative humidity
of the surrounding air, it is logical to start with the behaviour of individual fibres. When a fibre
absorbs moisture, it will swell longitudinally along its length and transversely across its width.
The length changes of the fibres are so small as to be negligible; from ordinary air-dry condition
(45% RH) to wet conditions, the fibres show an extension of less than 1%. On the other hand,
the increase in diameter is comparatively large. As the relative humidity of the air increases, the
fibre diameter grows progressively till a maximum is reached in liquid water. The swelling is
reversible; the fibre diameter decreases when the water is dried out. These diameter changes of
the fibres play one of the fundamental parts in almost all shrinkage effects of textile materials.
Sommer (1960, p.282) gives the values of cross-sectional swelling of cotton fibres.

% of dimensional change in
a cotton fibre
Relative humidity (%) Length Cross-section

100 (0.09) 48.6


90 0.46 29.7
80 0.49 22.2
60 0.30 15.7
30 0.15 9.3
0 0 0
30 0.23 7.1
60 0.36 12.2
80 0.60 17.2
90 0.76 25.0
100 (in water) 1.124 43.9

Table 7 Dimensional changes of cotton fibre at different relative humidities.


As can be read from this data, the cross-sectional swelling of the fibre follows a hysteresis loop.
Morehead (1952) and Hedley (1993, p.113) mentioned the remarkable phenomenon of the
extreme increase in swelling between 97% and completely wetting out. Hedley gives as data for
the swelling of cotton at 90% relative humidity 11.5%, at 97% RH a swelling of 19.1% and when
wetted out a swelling of 33.5%.

Different types of fibre display different swelling behaviours as can be seen in the following table.
The amount of swelling depends on the amount of amorphous material, the size of the
crystallites and the presence of polar groups.

Fibre Swelling in water Longitudinal Source


% increase in diameter swelling (%)
(transverse swelling)

cotton 44-49 Lewin 1983


cotton 20, 23, 7 Morton 1993
cotton 20-26 1.1 Morehead 1947
flax 47 (area swelling) Morton 1993
flax tow 47 Morehead 1947
wool 32-38 Lewin 1983
wool 22-26 1.2 Morehead 1947
silk 19-20 1.3-1.7 Morehead 1947

Table 8 Swelling of different fibre types

3.2 Yarn shrinkage

Several studies have been made into the behaviour of yarns and woven structures to changes in
relative humidity. On old study by Collins (1939) explains the shrinkage of yarns as follows:
..."The important features of the yarn structure are the twist factor and its compactness. [...] The
length of a fibre round the yarn is determined by the distance the fibre goes round the yarn and
the distance it goes along the yarn. When the yarn diameter increases, the distance round the
yarn increases and in order for the fibre to continue to span both as far round and as far along
the yarn as it did before, it would have to stretch. This it is reluctant to do if any easier course is
available and a little consideration will show that there are two possible ways whereby the fibre
can avoid stretching. The first is by untwisting, whereby the distance of the fibre round the yarn
is reduced. [...] In most cases however, the yarn is not allowed to untwist. [...] the fibre has to
span as far round the yarn as before and the only means of shortening its path is for it to span a
shorter distance along the yarn, i.e. the yarn must shrink." ... (Collins 1939, p.48)

When a yarn takes up moisture from the air it will swell, because the fibres from which it is spun
swell, and the more compact a spun yarn, the closer its swelling will match that of the fibres. The
yarn shrinkage itself is directly related to the yarn twist. In a yarn of zero twist there would be no
need for the fibres to stretch and thus no shrinkage forces would occur. Greater twist will
therefore cause greater yarn shrinkage as well as an increasing yarn swelling.
3.3 Fabric shrinkage

It might be obvious that the shrinkage of both fibres and yarns can only show at most a few
percent. It is therefore that the extreme shrinkage of some cotton and linen goods cannot be
solely explained by the phenomena described above. It is evident that the large shrinkages must
be due to some features special to the woven fabrics themselves.

Shrinkage of fabric can basically be allocated to two causes:


- relaxation of stresses induced during weaving
- fibre and yarn swelling resulting in crimp

Relaxation shrinkage is caused by the tensions introduced during spinning, weaving and
finishing. As fibres are being spun into yarns, yarns woven into fabrics and fabrics subjected to
finishing treatments, the fibres are stretched. So long as the fabric is kept dry, the fibres will be
unable to recover their normal sizes and shapes. But once the fabric becomes wet, the fibres
are softened and lubricated and are able to relax into more natural positions. When so-called
grey state goods of sized, unbleached cellulosics, are wetted out, they tend to shrink 10-15%.
However this is often an early stage in fabric processing, prior for example to bleaching, dyeing,
and other wet finishing. Thus, finished textiles, like garments, are generally not subject to more
than 1 to 2% shrinkage. (Ballard 1996, p.667)

The second cause is much more complex. The following figures show what happens when a
fabric structure takes up moisture.

Fig 12 Chan- ge in fabric structure


from yarn swelling (after Hedley 1993,
p.113)
The fibres and yarns take up moisture and swell. As the yarns swell, the path which they have to
follow in the weave becomes greater in length. There are now two possibilities; either the yarn
must elongate to account for the greater distance, or the interlacing yarns must be pulled
together so that the distance remains the same. Since cotton and linen yarns are rather
inextensible stiff structures they are less likely to extend themselves and, if the fabric is not
restrained, they will instead tend to pull the interlacing yarns closer together. In practice such
forces will work in both the warp and weft directions of the fabric and an interchange of crimp will
occur, with the final configuration being a balanced equilibrium of the forces developed.
However, in general, the greater the initial crimp of a yarn the greater will be the shrinkage forces
and, since warp yarns frequently display the greater crimp, it is not uncommon for such fabrics to
show most shrinkage in the warp direction. According to Collins, crimp interchange of this kind
can account for shrinkages of 10% of more in cotton fabrics. (Hedley 1993, p.113)

Bilson (1996) describes that the shrinkage of a woven cellulosic fabric is preceded by a short
term loss of crimp and tension as a consequence of longitudinal yarn swelling. Sudden wetting
out may cause canvas to expand briefly before it shrinks. He describes a two-stage response of
tension fall and regain. The crimp effect of transverse swelling are delayed because it can only
take place after the free spaces within the weave have begun to be filled. Instances of free
space include gaps between individual fibres within yarns and spaces separating adjacent yarns,
as well as intermittent contact between intersecting yarns.

The degree to which a woven structure may shrink due to transversal swelling of the yarns,
depends very much on the weave structure, i.e. the type of weave and the weave density. The
translation of transverse yarn swelling into shrinkage can only occur at points in a woven
structure where warp and weft intersect. Therefore, a weave like satin or damask, which is
characterised by a high proportion of floating yarns and few interlacings, will be less likely to
shrink in wet and humid conditions than a plain weave of the same yarn count. A plain weave
with a very high density, i.e. very closely packed with few free spaces, will not shrink as much as
a plain weave with a lesser density. This is because it is not possible to pull the interlacing
threads any closer together and the weave structure is 'jammed'. Very open weave structures
will also not shrink that much, because of the abundance of free spaces present.

There is also the phenomenon of successive shrinkage. If a piece of cotton or linen fabric is
washed more than once it is usually found that it continues to shrink in the succeeding washes.
The amount of such successive shrinkage becomes progressively less with each wash and the
fabric approaches certain limiting dimensions. Pre-washing a fabric will cause shrinkage and,
though a single washing is unlikely to release all shrinkage potential, it will reduce the degree of
shrinkage on further wetting. Achieving the final shrinkage state could take six or more machine
washes. Collins writes about this: ..."There is both practical and theoretical interest in this
successive shrinkage effect. From the practical point of view, the most important aspect is that
the shrinkage reaches a limit but almost equally important is the knowledge of the proportionate
amount of successive shrinkage that is likely to occur. It is found that this depends on the
washing treatment. Fig.5 shows the result obtained when a bleached 5-shaft satin was washed
successively by 2 methods, immersion in still water and rotation in a washing machine. It is clear
that though the fabric will eventually shrink the same amount by either method, the use of the
machine ensures that the shrinkage more rapidly approaches the limit."... (Collins 1939, p.57)
2.2.1 Studies in fabric shrinkage

There have been a few studies into the behaviour of woven fabrics, especially in the field of
painting conservation. Linen and cotton canvasses have been used by artists as a carrier for
their paintings. The shrinkage behaviour of these canvasses is, of course, of great importance
for the stability of the paint layers. Some of these studies will be briefly summarised here.

In the study by Abbott (1964), 'The mechanism of fabric shrinkage: the role of fibre swelling', the
theory of fabric shrinkage described above is confirmed.
The fabric (viscose) specimens (strips of 1 inch wide and 6 - 8 inch long (parallel to the warp))
were immersed in distilled water for some time prior to testing in order to minimize relaxation
shrinkage. While still wet, they were mounted in an Instron tester. A sample was then extended
5% at a rate of 0.2 inch/min (arbitrary values). After the tension in the sample had become
reasonably constant, the specimen was allowed to dry to an equilibrium with the laboratory
atmosphere (20°C, 65% RH). When the dried specimen was sprayed again with water, an
immediate drop in tension could be seen, caused by the longitudinal swelling of the fibres and
the accompanying decrease in fibre modulus. When the swelling progressed to a state of close
packing, the effects of lateral swelling of the fibres became evident, causing the rise in tension to
a steady wet-state condition.
When the drying air was turned on, no effect was noticed except for a slight decay of the wet-
state tension with time until all liquid water had disappeared and the fibres begin to de-swell. The
gradual drop in tension then observed resulted from lateral de-swelling of the fibres. Eventually a
state was reached in which the structure became sufficiently opened up so that further de-
swelling would not produce any further decrease in crimp. At this point, the effects of longitudinal
de-swelling and increase in fibre modulus became important again and the tension gradually
rose until it reached the value characteristic of equilibrium with the atmosphere of 65% RH.

Measurements were also taken of the tension changes which occurred when the relative
humidity of the atmosphere surrounding the fabric was changed from 65% to 98% and back
again to 65% without actually any liquid water touching the specimen. When the fabric was
exposed to the high-humidity atmosphere, the equilibrium tension was reached only very slowly.
It was only after about fourteen hours that the tension arrived at a relatively constant value.
Though the rate of change is much slower, the general shape of the curve obtained is essentially
similar to that obtained with liquid water. The initial relatively rapid change is attributable to
longitudinal swelling and modulus decrease of the fibres, and the slow rise in tension can be
attributed to the effects of the cross-sectional swelling which occurs as the fibre moisture content
reached equilibrium with the 98% relative humidity atmosphere.

Abbott also looked at the behaviour of a fabric when kept under constant tension.
..."These changes in length during a cycle of wetting and drying, bear an inverse relationship to
the direction of the tension changes observed at approximately corresponding stages of wetting
and drying when similar specimens were held at constant length. The rapid increase followed by
an immediate decrease in length, which takes place on wetting, corresponds to the rapid drop
and rise in tension which was observed previously. Therefore, the length increase may be
attributed to longitudinal swelling of the fibres and the length decrease to the effects of lateral
swelling of the fibres."... (Abbott 1964, p.T124)

Bilson is currently carrying out a research into canvas shrinkage.


..."This preliminary study compares the response of a new, unprepared linen canvas to a glue-
sized and oil-primed mid 19th century loose lining canvas which was known to shrink in the
presence of moisture. Both were plain-woven from a single Z-spun yarns: the approximate
counts being 15 warps and 16 wefts per centimetre for the new sample. [...] A qualitative
assessment of cross-sectional area based on the major diameter showed the wefts in each case
to be generally 20 to 30% thicker than the warps. A qualitative assessment of twist showed the
warps to be medium spun, whilst the wefts were loosely spun. [...]
The samples were allowed to shrink restrained only by friction between the underside of the
canvas and a supporting surface. To accomplish this, 6 x 10 cm pieces of the two canvases
were placed on a stretched membrane of Reemay 2006 - a spun-bonded polyester chosen for its
porosity, smoothness and relative insensitivity to moisture. The whole assembly was then placed
in a humidity chamber with space allowed for air to circulate above and below the samples.
Humidity levels were maintained with a portable ultrasonic humidifier in combination with a fan,
and recorded with a KM004 combined temperature and humidity probe (accuracy + 2% RH, +
1°C). Temperature within the chamber was maintained at ambient levels of around 18 to 22°C.
Measurements of yarn swelling and shrinkage within the samples were taken using a Sony
three-chip CCD camera linked to a Power Macintosh computer running Optilab image analysis
software. [...]

Images from the two samples at 10x magnification showing areas of canvas approximately 21 x
16 mm were taken at RH levels of 30, 47, 65, 72, 86, 94 and 100% (saturation achieved by
spraying) - the samples being maintained at each level for three hours. [...] Three separate sets
of measurements were taken from each image: those of changes in free space between each
individual pair of intersecting and adjacent yarns, changes in yarn diameter and spacing along
the length of each warp between intersections, and dimensional changes in an area bounded by
four marked yarns."... (Bilson 1996, p.245-250)

It is clear that transverse yarn swelling can be detected at even low RH, and that its progression
as the precursor of shrinkage can be followed, despite the opposite reaction of the size and the
balancing action of longitudinal swelling. There was a difference in transverse swelling between
the weft (loose spun, large diameter) and of the warp (medium spun/small diameter). The
predominance of transverse swelling in the weft is related to its diameter being larger than that of
the warp, a condition which would induce greater dimensional shrinkage in the warp direction.

Another recent study is by Young (1996), 'Biaxial properties of sized cotton-duck'.


She looked into the behaviour of modern sized cotton-duck fabric, used as a carrier for paintings.
She found that both the weft and warp of the fabric exhibit creep on loading and hysteresis on
onloading, the weft having a higher degree of non-linear stiffness. In all the tests she carried out,
the weft creeped more than the warp when stretching the raw fabric. Fabric samples, both raw
and sized, were tested in a specially designed biaxial tensile tester. Just after wetting, the stress
in the fabric was reduced to 0.5 Mpa (25N). This initial decrease in tension was associated with
a release in fibre strains locked into the fabric during manufacturing. The water reduced the
friction at intersection points allowing yarn slippage and stress relaxation. After two minutes the
biaxial tension started to rise reaching a maximum, where the weft stress was 3.3 MPa and 3.0
MPa in the warp. This is the point where the yarn fibres have swelled because of water
absorption and crimp increases. The weft and warp yarns now have longer geometrical paths
resulting from increased yarn diameter. (Young 1996, p. 327-328)
Bibliography

Abbott, N. J. et all. The mechanism of fabric shrinkage: the role of fibre swelling. In:
Journal of the Textile Institute, 1964(55): T111-T127.

ASTM designation D 5429-93: Standard practice for pretreatment of backing fabrics used in
textile conservation research.

Bacci, G. Il consolidamento del tessuto: aspetti tecnici. In: Gli arazzi della Sala
dei duecento., Studi per il restauro. Florence: Ministero per i Beni Culturali, 1985:
192-200.

Ballard, Mary W. Hanging out: strength, elongation, and relative humidity: some physical
properties of textile fibres. In: ICOM-CC 11th Triennial Meeting Edinbrugh, 1996.
London: James & James Ltd, 1996: 665-669.

Ballard, M. Hanging out with tapestries, carpets and quilts: the effect of textile
properties on appearance. In: The Textile Conservation Group Newsletter,1996 Mar,
18(3): 1-4.

Ballard, M. W. How backings work: the effect of textile properties on appearance. In:
Lining and backing: the support of paintings, paper and textiles., Papers delivered at The
UKIC Conference 7-8 November 1995. The United Kingdom Institute for Conservation,
1995: 34-39.

Ballard, M. Mechanicle properties: preview and review. In: Textile Conservation


Newslettter, 1995, spring: 14-28.

Bauer, R. Zur Geschichte der Tapisserienpflege am Kunsthistorischen Museum in


Wien (1). In: Restauratorenblatter, 1991, Band 12: 57-66.

Bennett, A. Garry Five centuries of tapestry from The Fine Arts Museums of San
Francisco. San Francisco: The Fine Arts Museums.

Berry, G.M. et all Reinforcing degraded textiles., part I: properties of naturally aged cotton
textiles. In: Preservation of paper and textiles of historic and artistic value. Washington
D.C.: American Chemical Society, 1977: 228-248.

Biddulph, F. Point Counterpoit., Behind the scenes at Hatfield House, Hertfordshire.,


skilled volunteers conserve the finest early 17th century English tapestries. In:
Traditional Interior Decoration. 1988: 108-116.

Bilson, Thomas Canvas shrinkage: a preliminary investigation into the response of a


woven structure. In: ICOM-CC 11th Triennial Meeting Edinbrugh, 1996. London: James
& James Ltd, 1996; 245-252.

Blyth, V. Restoration of the unicorn tapestry. In: SSCR Journal, 1992 May, 3(2):
10-11.

Boeck, J. de La restauration et la conservation des tapisseries. In: Tecniche di


conservazione degli arazzi, 1981, Florence. Florence: Leo S. Olschki, 1986: 29-38.
Boeck, J. de, et all The treatment of two sixteenth-century tapestries at the Institut Royal du
Patrimoine Artistique. In: The conservation of tapestries and embroideries, 1987,
Brussels. The Getty Conservation Institute, 1990: 113-116.

Borg Clyde, N. The "Roman de la Rose" tapestry. In: The art of the conservator.
(Oddy, A.(editor )). London: British Museum Press, 1992: 151-162.

Bosworth, D. The problems of ethics and aesthetics. In: Seminar International la


restauration et la conservation des tapisseries, 1984, Paris. Paris: IFROA: 102-103.

Brooks, M. et all Supporting fragile textiles: the evolution of choice. In: Lining and backing: the
support of paintings, paper and textiles. Papers delivered at The UKIC Conference 7-8
November 1995. London: The United Kingdom Institute for Conservation, 1995: 5-13.

Brutillot, A. Conservation d'une tapisserie au Bayerisches Nationalmuseum. In:


Seminar International la restauration et la conservation des tapisseries, 1984, Paris.
Paris: IFROA: 133-137.

Bruttilot, A. Conservation of a fifteenth-century tapestry from Franconia. In: The


conservation of tapestries and embroideries, 1987, Brussels. The Getty Conservation
Institute, 1990: 75-79.

Brutillot, A. Konservierung der Tapisserie "Anbetung der Konige und Darstellung im


Tempel". In: Zeitschrift fur Kunsttechnologie und Konservierung, 1992, 6(2): 367-372.

Cavallo, A. S. Tapestries of Europe and of Colonial Peru in the Museum of Fine Arts,
Boston. Boston: Museum of Fine Arts.

Clarke, A. and F. Hartog The cost of tapestry conservation. Paper presented at the
Reconstruction and Camouflage Techniques in Tapestry Conservation, 1994,
Amsterdam. Unpublished.

Collins, G. E. Fundamental principles that govern the shrinkage of cotton goods by


washing. In: Journal of the Textile Institute. 1939(30): 46-61.

Cook, J. Gordon. Handbook of textile fibres., I. Natural fibres. Durham UK: Merrow, 1993.

Cousens, S. The effects of old repairs on the choice of conservation treatment of


tapestries. In: Seminar International la restauration et la conservation des tapisseries,
1984, Paris. Paris: IFROA: 138-142.

Diehl, J. M. Some remarks concerning the restorationwork on the Zealand


tapestries. Haarlem: Foundation workshop for the restoration of ancient textiles;
unpublished.

Diehl, J. M. and F. Visser. Tapestries. In: Textile Conservation (Leene, J. E. editor).


London: IIC, Butterworths, 1972: 153-163.

Diehl, J. M. The workshop for the restoration of ancient textiles, Haarlem. in: Delft
conference on the conservation of textiles, 1964, Delft. IIC: 105-108.
Dolcini, L. Arazzi, tapestries. I Restauri dell'Opificio Firenze, 1990: 61-66.

Dolcini, L. Il consolidamento del tessuto: aspetti metodologici. In: Gli arazzi della
Sala dei Duecento., Studi per il restauro. Florence: Ministero per i Beni Culturali, 1985:
201-209.

Dolcini, L. Integrazione delle lacune negli arazzi: etica e unita metodologica. In:
Conservazione e restauro dei tessili., Convegno internazionale, 1980, Como. Milan:
CISST: 107- 113.

Dolcini, L. The tapestries of the Sala dei Duecento in the Palazzo Vecchio. In: The
conservation of tapestries and embroideries,1987, Brussels. The Getty Conservation
Institute, 1990: 81-87.

Drysdale, L. Conservation of a 16th century Flemish tapestry in the Poldi Pezzoli


Museum, Milan (part 1). In: V&A Conservation Newsletter, 1983, Spring(17): 34-39.

Erlande-Brandenburg, A. Deontologie de la restauration des ouvres d'art. In: Seminar


International la restauration et la conservation des tapisseries, 1984 , Paris. Paris:
IFROA: 104-106.

Ewer, P. Tapestry conservation project - Biltmore House. In: Textile conservation


newsletter. 1989; Spring: 11-14.

Feller, R.L. et all The kinetics of cellulose deterioration. In: Historic textiles and paper
materials. Washington D.C.: American Chemical Society, 1986:329-345

Finch, K. Problems of tapestry conservation. In: V&A Conservation Newsletter,


1982, Winter(16): 40-43.

Finch, K. Problems of tapestry preservation. In: Tecniche di conservazione degli


arazzi,1981, Florence. Florence: Leo S. Olschki, 1986: 39-45.

Finch, K. Evolution of tapestry repairs: a personal experience. In: Seminar


International la restauration et la conservation des tapisseries, 1984 , Paris. Paris:
IFROA: 125-132.

Finch, K. Special problems. In: Seminar International la restauration et la


conservation des tapisseries, 1984 , Paris. Paris: IFROA: 144-146.

Finch, K. Tapestries: conservation and original design. In: The conservation of


tapestries and embroideries, 1987, Brussels. The Getty Conservation Institute, 1990:
67-74.

Finch, K. The Textile Conservation Centre, tapestry seminar - Paris 1984. In:
V&A Conservation Newsletter, 1985, 24(summer): 19- 24.

Finch, K. Textiles as documents of history and theory and those who care for
them. Paper submitted at the Joseph V. Columbus Tapestry Symposium), 1989,
Washington D.C.

Finch, K. and G. Putnam Caring for textiles. Barrie & Jenkins, 1977.
Flury-Lemberg, M. Textile conservation and research. Bern: Abegg-Stiftung, 1988.

Gohl, E.P.G. Textile science; an explanation of fibre properties. Melbourne:


Longman Cheshire, 1993.

Graaf, A. J. de Tensile properties and flexibility of textiles. In: Conservazione e


restauro dei tessili., Convegno internazionale, 1980, Como. Milan: CISST: 54-61.

Hackney, Stephen and Gerry Hedley Linen canvas artificially aged. In: Measured opinions.,
collected papers on the conservation of paintings. London: UKIC, 1993: 70-75.

Hackney, Stephen and Gerry Hedley Measurements of the ageing of linen canvas. In:
Measured opinions., collected papers on the conservation of paintings. London: UKIC,
1993: 57-65.

Harrison, P. W. Textile degradation. Manchester: The Textile Institute, 1991.

Hartog, F. Tapestry linings. In: Conservation News, 1993 July (51): 48-49.

Hearle, J. W. S. and R. H. Peters Moisture in textiles. Manchester: The Textile Institute,


London: Butterworths Scientific Publications, 1960.

Hearle, J. W. S. and R. H. Peters Fibre structure. Manchester: The Textile Institute,


London: Butterworths, 1963.

Hedley, Gerry Measured opinions., collected papers on the conservation of paintings.


London: UKIC, 1993.

Hedley, Gerry Relative humidity and the stress/strain response of canvas paintings:
uniaxial measurements of naturally aged samples. In: Studies in Conservation,
1988,(33): 133-148.

Hedley, Gerry et all Artists' canvases: their history and future. In: Measured opinions.,
collected papers on the conservation of paintings. London: UKIC, 1993: 50-56.

Hefford, W. “Bread, brushes and brooms”: aspects of tapestry restoration in


England 1660 1760. In: Acts of the tapestry symposium, 1976, San Francisco. San
Francisco: The Fine Arts Museum San Francisco: 67-73

Hefford, W. The restoration of the tapestries. In: The Devonshire hunting tapestries
(Wingfield Digby, G. editor). London: V&A, c1971: 57-82.

Herald, J. Notes on English tapestries. Unpublished: Textile Conservation Centre,


1996.

Herrmann, H. Die Konservierungs- und Restaurierungsmethoden an Bildwebereien.


In: Zeitschrift fur Kunsttechnologie und Konservierung, 1987, 1(2): 157-163.

Hollen, N. Textiles. New York: MacMillan, 1988.


Howell, David Some mechanical effects of inappropriate humidity on textiles. In:
ICOM-CC 11th Triennial Meeting Edinbrugh, 1996. London: James & James Ltd, 1996:
692-698.

Hutchison, B. From restoration to conservation: parallels between the traditions of


tapestry conservation and carpet conservation. In: The Textile Museum Journal, 1990.

Hutchison, R. B. Gluttony and Avarice: two different approaches. In: The conservation of
tapestries and embroideries, 1987, Brussels. The Getty Conservation Institute, 1990:
89-94.

Jakobiec, W. Treatment of a 16th-century tapestry. In: CCI Newsletter, 1993, (12):


1-3.

Janssen, E. La conservation et la restauration de tapisseries. In: Les fresques


mobiles du Nord., Tapisseries de nos regions, XVI-XX siecle. Anvers: Gaspard De Wit,
1994: 27-37.

Kajitani, N. Conservation maintenance of tapestries at the Metropolitan Museum of


Art, 1987. In: The conservation of tapestries and embroideries, 1987, Brussels. The
Getty Conservation Institute, 1990: 53-66.

Kajitani, N. Conservation of "Courtiers in a rose garden", a fifteenth-century tapestry


series. In: Conservation Research., studies of fifteenth- to nineteenth century tapestry (Joseph
V. Columbus Tapestry Symposium), 1989, Washington D.C. . Washington D.C.:
National Gallery of Art, 1993: 79-102.

Kajitani, N. The preservation of medieval tapestries. In: Acts of the tapestry


symposium, 1976, San Francisco. San Francisco: The Fine Arts Museum San
Francisco: 45-63.

Keyserlingk, E. Backing research project. In: Textile Conservation Newsletter, 1984,


fall: 9-11.

Kjellberg, P. Tapestry craftsmen. In: Connaissance des arts.

Landi, S. The textile conservator's manual. Oxford: Butterworths, 1992.

Lehrgang fur die berufliche Bildung., Beurteilungsmerkmale textiler Faserstoffe. Berlin:


Bundesinstitut fur Berufsbildung, 1986.

Lemberg, M. The problem of brown wool in mediaeval tapestries: the restoration of


the fourth Caesar tapestry. In: Studies in textile history., in memory of Harold B.
Burnham. Ontario: Royal Ontario Museum, 1977: 178-183.

Lewin, Menachem Handbook of fiber science and technology. New York: M. Dekker,
1983/84.

Lugtigheid, R. "Een perspectyff tapeet". In: HAT Bulletin, 1995, 2: 1-4.

Maes, Y. The conservation/restoration of the sixtheenth- century tapestry "The


Gathering of the Manna". In: The conservation of tapestries and embroideries, 1987,
Brussels. The Getty Conservation Institute, 1990: 103-111.
Maes, Y. The conservation treatment of the tapestries of the patrimonio nacional.
In: Golden weavings., Flemish tapestries of the Spanish Crown. Malines: The Gaspard
de Wit Foundation, 1993: 114-123.

Mally, A. Die Kombinationsfahigkeit von Textilmaterialien bei der Nahtechnischen


Restaurierung eines Gobelins. In: Arbeitsblatter, 1995, 2: 211-223.

Maran, M. Conservation of a 16th century Flemish tapestry fragment. In: SSCR


Journal, 1993, 4(3): 15-17.

Marko, K. All or nothing - or something., a flexible approach to tapestry


conservation. In: Postprints of UKIC meeting "Compromising situations: principles in
everyday practice", 1993. London: The United Kingdom Institute for Conservation, 1993:
33-34.

Marko, K. Tapestry conservation - a confusion of ideas? In: Lining and backing:


the support of paintings, paper and textiles. Papers delivered at The UKIC Conference
1995. London: The United Kingdom Institute for Conservation, 1995: i- iv.

Marko, K. Two case histories: a seventeenth-century Antwerp tapestry and an


eighteenth-century English Soho tapestry. In: The conservation of tapestries and
embroideries, 1987,Brussels. The Getty Conservation Institute, 1990: 95-101.

Masschelein-Kleiner, L. Study and treatment of tapestries at the Institut Royal du Patrimoine


Artistique. In: Conservation Research., studies of fifteenth- to nineteenth century
tapestry (Joseph V. Columbus Tapestry Symposium), 1989, Washington D.C..
Washington D.C.: National Gallery of Art, 1993: 71-75.

Masschelein-Kleiner, L. Summary of the discussions. In: The conservation of tapestries and


embroideries, 1987, Brussels. The Getty Conservation Institute, 1990: 1-3.

Masschelein-Kleiner, L. and J. de Boeck Contribution to the study of the conservation of


monumental tapestries. In: Preprints of the 7th ICOM triennial meeting, Copenhagen,
1984: 84.9.33-84.9.37.

Masschelein-Kleiner, L. and V. Vereecken De conservaring van wandtapijten. In: S.O.S.


Wandtapijten., redding van 24 belangrijke kunstwerken. Brussel: Koninklijke Musea voor
Kunst en Geschiedenis, 1994: 21-38.

Melville Smith, L. The exception to the rule., conservation of a tapestry fragment. In:
Textile History, 1984, 15(2): 209-218.

Mendonca, M. J. de Un probleme de restauration d'une tapisserie Bruxelloise du XVIe


siecle. In: Delft conference, 1964. Delft: IIC, 119-120.

Meredith, R. The mechanical properties of textile fibres. Amsterdam: North-Holland,


1956.

Michalski, Stefan Relative humidity: a discussion of correct/incorrect values. In: ICOM-CC


th
10 Triennial Meeting Washington DC, 1993. Washington DC, 1993, 2: 624-629.
Michalski, Stefan and Debra Daly Hartin CCI Lining project: preliminary testing of lined
model paintings. In: ICOM-CC 11th Triennial Meeting Edinbrugh, 1996. London: James
& James Ltd, 1996, 1: 288-296.

Morassutti, A. M. Struttura, funzione e funzionamento della attrezzature per il restauro.


In: Gli arazzi della Sala dei Duecento., Studi per il restauro. Florence: Ministero per i
Beni Culturali, 1985: 172-180.

Morehead, Frederick F. A method for studying the effect of humidity on the cross-sectional
swelling of some common fibres. In: Textile Research Journal, 1952, August(22):
535-539.

Morehead, Frederick F. Some comparative data on the cross- sectional swelling of textile fibers.
In: Textile Research Journal, 1947(17): 96.

Morton, W. E. and J. W. S. Hearle Physical properties of textile fibres. Bath: The Textile
Institute, 1993.

Needles, H. Heat induced aging of linen. In: Historic textile and paper materials II.
Washington D.C.: American Chemical Society, 1989.

Neugebauer, H. Gobelinrestaurierung: de Aldalbero-Gobelin aus Stift Lambach. In:


Restauratorenblatter, 1991, 12: 67-72.

Nouak, E. Restaurierung von sechs Gobelinsessel aus Schloss Schonbrunn. In:


Restauratorenblatter, 1991, 12: 73-76.

Ordonez, M. T. and A. A. Ordonez Evaluation of mounting techniques used on vertically


hung textiles. In: Preprints of the 7th triennial meeting, 1984, Copenhagen. ICOM,
1984: 84.9.38-84.9.40.

Palmer Simpson, L. Abrasiveness of certain backing fabrics for supporting historic textiles.
In: JAIC, 1991, 30: 179-185.

Pow, V. The Haarlem workshops’method of reweaving. In: The Devonshire hunting


tapestries (Wingfield Digby, G. editor). London: V&A, c1971: 82.

Pow, C. V. The conservation of tapestries for museum display. In: Studies in


Conservation, 1970, 15: 134-153.

Putnam, G. The conservation of tapestries., research report. Hampton Court: The


Textile Conservation Centre, 1983.

Rice, J. W. How humidity may affect rug, tapestry, and other textile collections. In:
Principles of textile conservation science, IX: 53-56.

Rusznak, Istvan Interaction of aqueous systems with fibers and fabrics. In: Handbook of
fiber science and technology: Volume I., Chemical Processing of fibers and fabrics.
Fundamentals and preparation, part A. (Lewin, Menachem and Stephen B. Sello). New
York: Marcel Dekker, 1983: 51-92.

Russell, W. H. and G. A. Berger The behaviour of canvas as a structural support for


painting: preliminary report. In: Science and technology in the service of conservation,
1982, Washington DC.. London: IIC: 139-145.

Scicolone, Giovanna C. et all Kinetics of cellulose fiber degradation and correlation with some
tensile properties to plan consolidation or lining interventions. In: ICOM-CC 11th
Triennial Meeting Edinbrugh,1996. London: James & James Ltd, 1996, 1: 310-315.
Shephard, Lynsay K. The conservation treatment of two partially restored 18th Century
Brussels tapestries. In: ICOM-CC 11th Triennial Meeting Edinbrugh, 1996. London:
James & James Ltd, 1996, 2: 721-725.

Simon, A. Die Restaurierung der Rubensteppiche des Kolner Doms. In: Zeitschrift
fur Kunsttechnologie und Konservierung, 1987, 1(2): 150-156.

Sommer, H. Handbuch der Werkstoffprufung: band V. Die Prufung der Textilien.


Berlin: Springer, 1960.

Sook, K. and I. Block Photodegradation of cellulosics, part 1: Effects of temperature and


humidity on the tear strength retention. In: Preprints of the 9th triennial meeting, 1990,
Dresden. ICOM, 1990: 302-306.

S.O.S. Wandtapijten., redding van 24 belangrijke kunstwerken. Brussel: Koninklijke Musea voor
Kunst en Geschiedenis, 1994.

Specht-den Boer, Inge De conservering van wandtapijten., Verslag van een stageperiode bij de
Stichting tot Herstel van Antiek Textiel. Unpublished: Opleiding Restauratoren, 1991.

Varoli-Piazza, Rosalia and Rosanna Rosicarello. The use of photogrammetry in determining the
correct method of displaying a textile artefact: the cowl of St Francis of Assisi. In:
ICOM-CC 11th Triennial Meeting Edinbrugh, 1996. London: James & James Ltd, 1996,
2: 726-731.

Ward, S. and P. Ewer. Tapestry conservation at Biltmore House. In: The International Journal
of Museum Management and Curatorship, 1988, 7: 381-388.

Wingfield Digby, G. (ed) The Devonshire hunting tapestries. London: V&A, c1971.

Yardley-Jones, A. Preliminary treatment of a 16th century tapestry. In: Textile


Conservation Newsletter, 1988, fall: 23-24.

Young, Christina R.T. Biaxial properties of sized cotton- duck. In: ICOM-CC 11th Triennial
Meeting Edinbrugh, 1996. London: James & James Ltd, 1996, 1: 322-331.

View publication stats

You might also like