View source for The discovery of three lost Salting carpets

{{Infobox Different Types of The discovery of three lost Salting carpets
|above       =
|image       = [[File:|220px]]

|caption     = 

|image2      = 
|caption2    = 

<!--General information-->

|Name             = 
|Original name     = 
|Alternative names      = 

|Origin    = {{flag|Iran}}: 
|Manufacturer and Master designers     = 
|Weaving duration      = 
|Year of Delivery      = 

|To know more      =   

<!--Technical information-->
|Common designs      = 
|Common patterns       = 
|Common motifs       = 
|Common sizes      = 
|Number of colors used     = 
|Field color    = 
|Borde color     = 
|Dyeing method     = 
|Pile material      = 	
|Pile height      = 
|Foundation material     = 
|Weaving Technique      = 
|Knot type      =  
|Linage     = 
|Knot density     = 
|Number of nodes = 
|Weight      = 

|additional info      = 
}}



In 2007  three [[Safavid Rug|Salting]]  carpets  were  discovered in  the  Palace  of  the  Dukes  of  Bragan a,  in 
Guimarães. Two are prayer rugs with central niches, while the third has a central medallion. Finely 
knotted in wool on a silk foundation, and embellished with metal threads, this is the largest collection 
of these carpets known outside the  Topkapi  Saray  (Istanbul). For a century, their origin and 
chronology have been the source of considerable debate.  
They  take their  name from  the  collector George Salting (1835-1909) who donated a significant 
example to the  Victoria &  Albert Museum  (London), which was  originally  attributed to 16th-century 
Iran. However, the vivid  colors  of  this  and  other  carpets, and existence of  similar examples  in the 
Topkapi, led later historians to argue that they were Turkish rugs from the 18th or 19th centuries, and 
possibly even forgeries of classical Persians carpets. 


==INTRODUCTION==
They derive their name from the famous Australian-born British art colector, George  Salting (1835-1909),
who gave a particularly significant example to the Victoria and Albert Museum (London) in 1909 [1].
Defining the  “Saltings”  as a group has  proven a difficult task, but, in general, the  presence of  a silk 
foundation  (warps  and  wefts), fine  wool  pile, high  knot count,  bright colours and embel ishments  in 
metal thread, along with arabesque designs, religious inscriptions and small dimensions, are regarded 
as the most significant characteristics [1]. The question of their provenance and chronology has also 
been the source of continuous debate. 
At the end of the 19th century, they  were original y ascribed by  late 19th- and early  20th-century 
historians to 16th-century  Tabriz  (Iran)  and were recognized to be part of  a large group of  wel-preserved prayer rugs, with bril iant colours, in the Topkapi  Saray1. As  some of  the  religious 
inscriptions on the prayer rugs are almost exclusively Shi‟ite in nature, it was originally proposed that 
they  had survived  in  such good  condition  in Turkey  as  they  were not suitable for use by  the Sunni 
Ottomans. However, their  presence in Istanbul  encouraged subsequent authors, such as  Tattersall 
(1931), Erdmann (1941), and Rogers (1987), among others, to claim they were 18th- or 19th-century 
Turkish rugs, and possibly  even  forgeries  of classical  Persian carpets. More recently, a new
generation  of  carpet experts  has  returned  to  the subject and argued the „Saltings‟  are  16th or 17th 
century, and that the Topkapi rugs were sent from Safavid Iran to the Ottoman court. Michael Franses 
proposed that the carpets could have been sent from Shah Tahmasp I (r. 1524-76) to Suleyman the 
Magnificent in 1556, as suggested by a letter mentioning a gift of “carpets spread on the floor of the 
mosque for the use of the congregation”, or a decade later, to Sultan Selim II (r. 1566-1574) on the 
occasion  of  his  accession  in 1567  [1]. Jon  Thompson  has  recently  suggested, on  the  basis  of  an 
inscription  on one of  the  carpets  giving  the date of  1590-91  (AH 999)2, that they  were a gift from 
Tahmasp‟s  grandson,	Shah Abbas  I (r. 1587-1629),  to the  Ottoman  Sultan Murad III (r. 1574-95), 
possibly on the occasion of the Treaty of Istanbul signed in 1590 [2]. 
Given this complex historiography, resolving the question of the origin and date of the  „Saltings‟ is 
obviously imperative, and the discovery of three carpets in Guimarães offers a unique opportunity to 
reflect upon these problems. The type of materials used to make a carpet and how they are prepared 
can reveal groups of carpets and help to establish their origin and/or date [3]. This type of data can 
also indicate whether these groups reflect a single, coherent production of short duration, or multiple 
workshops and a wide historical timeline. Only partial scientific information is currently available for the 
„Saltings‟ [1], and hence, this study takes a multidisciplinary approach to the problems of their origin 
and date. Analyses are conducted on all of the materials used to make the three Guimarães carpets: 
namely, their fibres, dyes, mordants, and metal threads, and this data is then compared with published 
results for related examples documented in the literature to determine whether they are 1) products of 
Persian or Turkish workshops, and 2) date from the classical period (16th to 17th centuries), or are 
later, from the 19th or even 20th centuries. 
For tackling questions of provenance, yel ow and red dyes are especially important, as the former
are usually locally produced [4-7], while the latter in some cases have been shown to be cultural y and 
temporally specific [3]. In previous work on Islamic carpets the red dye obtained from the lac insect,
harvested in India and South Asia and widely traded throughout the region, has been shown to be the
main colourant for the red ground of 16th- and 17th-century Persian wool carpets [8-10]. The previous 
detection of this dye in seven prayer carpets in the Topkapi collection would appear to support a Persian (as opposed to Turkish) provenance for the group [10-11]. Identification of the precise insect lac species would provide important evidence for associating the carpets with specific centers of production. For instance, several lac species are found in India (Kerria lacca), while others species are cultivated in China (Kerria chinensis or Kerria mendingonsis) [12] (Appendix I). However, no published study to date has used the HPLC-DAD (High Performance Liquid Chromatography Diode Array) technique  to differentiate the  various  species  of  lac  insects  in order to localize  precise sources  of 
production, as has occurred, for example, for cochineal or dragon‟s blood [13-14]. Owing to the high 
complexity of the Kerridae family and the lack of accurate entomological studies on lac species, it was 
not possible to obtain specimens of the 26 species referred to in literature [6,12,15].  Nevertheless, for 
the  first time, several  lac-dye  samples  from  different geographical  regions,  as  well  as  from  different
host plants, were analysed using HPLC-DAD and PCA (Principal Component Analyses). This foremost 
lac-dye comparative database can support and promote further research on lac-dye species, as wel  
as  reveal  more  detailed  information  about  the  source of  the  red  dye  and therefore assist with 
provenance and identification of historical textiles. 
 As  for the question of  the chronology  of  the  „Saltings‟, dye  analysis  is  also useful,  as  the 
identification  of  synthetic  dyes  would provide  good evidence for a late  date,  as  demonstrated by 
Halpine for the Gobelins tapestry-upholstered furniture in the National Gallery of Art (Washington) [16],
while AMS Radiocarbon Dating (C14-AMS) can confirm an earlier one. Previous C14-AMS analyses 
conducted on a smal  number of „Salting‟ carpets [9]  have indicated a 16th- to 17th-century date for
them, but for unknown reasons a number of measurements gave “too high a carbon age” (circa 30% 
of  the  C14-AMS  total  analysis  performed), and this  question needs  to be  resolved  [17]. The 
composition and  morphology  of  the  metal  threads  is  also  relevant  for establishing  a date, although 
work in this area is still in its infancy.
Final y, the Guimarães carpets reveal an intense abrash (dappled tones) which is usual y attributed 
to subtle changes  in dye recipes,  and  an indicator  of  the  presence natural  dyes. When present in 
moderation, abrash  has  been  considered  a sign  of  high artistic  quality  [18],  but, in the case of  the 
Guimarães rugs, it could also be a consequence of degradation effects, owing to their exhibition under 
excessive light conditions for many years [19]. Mordant analysis and quantification of the dye content 
is a first step for resolving this question.   

==Technique==
The  structure of a carpet consists  of  warps, wefts  and knots, and two processes  are combined, 
weaving and knotting. The  warps are the  vertical threads that run the length of the  loom and, when 
cut, form the fringes at the ends; the wefts run horizontally across the warp. Patterns are created by 
the pile-knots  which  are  wound  around  one or two warps. After each row of  knots  is  complete,  the 
wefts are inserted across the carpet, and beaten down tightly with a comb to form a strong, resistant 
fabric. As  each coloured knot is  applied in sequence, along successive rows, carpet patterns  are 
constructed using a pointil ist technique, and involve laborious pre-planning of the design [20]. The level of detail achieved in the pattern depends on the volume 
of  the  interlaced  structure created  by  the warp, weft and knot, and 
hence, when the fibre diameter is low, designs with greater detail are 
possible. In the finest carpets, silk is used for all three threads, owing to its high strength in relation to its diameter. 
While horizontal and vertical designs can be created easily using 
 
the grid of warp and weft, diagonal and curving lines require a relative high knot density (more than 
3,500 knots/dm2), and silk is often used for the warp (and weft) in wool carpets to enhance the knot
count. This is the case for the „Saltings‟, and an S-twisted silk thread is used for both the  warp and 
weft to increase the  knot count to render very  detailed  designs, including  the  tightly  curving and 
scrolling vines  which  are  typical  of  their  designs.  The  knots  are wrapped  around  two warps  in  an 
asymmetrical  format,  open  to the  left (figure  1), which also enhances  the  detail, as  these  knots  are
more compact and the pile is more evenly spaced. 
In addition to knotting,  carpet patterns  can also be enhanced  with  decoration added in tapestry 
technique. Metal ic threads are especial y popular, and in the „Saltings‟, decorative threads comprising 
a silk core wrapped with a metal strip are introduced in selected areas, such as the inscriptions. 


==Style and Decoration==
Although the precise date of the „Salting‟ carpets has been debated for over a century, their design 
characteristics  reflect Safavid art, which was  given a major impetus  under Shah Tahmasp I (1514 - 
1576), who was an important patron of the art of the book, but also commissioned magnificent carpets 
and other objects [2]. Under his patronage, artists from Herat migrated to Tabriz, resulting in a fusion 
of the two artistic styles prominent in Iran prior to the rise of the Safavids: the art of the Timurids and 
Turkmans. This  new  style  features  tchi  clouds, angels, pheasants  and multiple layers  of  winding 
scrolls  as  major decorative elements  [2]. Figures  and  animals  were also used  increasingly, and 
integrated into  compositions  with geometric  framing  devices, such as  elliptical  and star-shaped 
medallions, cartouches, quatrefoils  and poly-lobed forms, as  seen in the  borders  of  the  „Salting‟ 
carpets.  
The splendour of this style stagnated after Shah Tahmasp became disinterested in the visual arts 
around 1555, and was only revived again with the rise of Shah Abbas I (1587-1629) [2]. At this time,
the  artists  of  the  re-established court library  began  to generate  a new  decorative style of  floral  and 
foliate  decoration, which,  in accordance  with the  Shah‟s  taste and  political  programme, was 
implemented across a wide range of arts, from il umination and bookbinding to textiles, carpets, and 
architectural  tiles. In contrast to previous  periods, the design  elements  are larger in scale, bolder in 
form  and more homogeneous, and  the  same split-leaf  arabesque can  be  found in bookbinding, a 
„Salting‟ carpet, and building decoration (figures 2 to 4).  
A new colour palette was also introduced under Shah Abbas. In textiles, there was a preference for
yellows, while the  dark  blues  and  reds  typical  of  the 16th century  gave way  to a palette of  softer
shades of beige, pink, light green, and light blue [20]. An elaborate use of metal thread for decorative 
elements, as  wel   as large areas of the  background, to create a visual effect of splendor and  luxury 
were also typical of the Shah Abbas period [22].  
The „Saltings‟ are enigmatic in sharing characteristics of both styles. On the one hand, they have 
poly-lobed medallions, cartouches, and a rich colour palette consistent with the Tahmasp style, and on 
the  other, they  have large-format arabesque scrolls  and extensive use of  metal  thread  usual y 
associated with textiles  produced  under Shah Abbas. This  mixture of  styles  is  in part responsible, 
along  with the bright colours of the Topkapi carpets, for  associating the „Saltings‟  with later Turkish
production and for  questioning their authenticity. 


==The Guimarães Carpets==
Two of the three „Salting‟ carpets found at the Palace of the Dukes of Bragan a (Guimarães) are 
small prayer rugs, symmetrical along the longitudinal axis, with a niche occurring in the central field, a 
wide colour palette, religious inscriptions, and are profusely decorated with metal-wrapped thread. The 
first rug, previously in the collection of Vital Benguiat and known as the “Benguiat Prayer Rug” (figure 
5), has vine scrolls decorating the red field, with the upper corners filled with irregularly-shaped areas 
of  colour  with  short inscriptions,  and a  continuous  inscription  outlining  the  central  niche.  The  main 
border has  inscribed cartouches  in red and green  which alternate  with lobed medallions  on  a beige 
ground. This  carpet is  related stylistically  to six  rugs  preserved  in the  Topkapi  Saray  Museum 
(Istanbul), and five in private collections or of unknown provenance.3 The carpet is in a poor state of 
conservation, and exhibits intense abrash, visible mainly in the blue, green and red areas of the upper 
corners.
The second carpet, known as the “Duff Prayer Rug” (figure 6), has a field of interlocking hexagons, 
with the upper corners fil ed with vine scrolls, while inscribed cartouches alternating with medallions in 
the  upper half  of  the  border. The  inner border  and niche  also  contain  running  inscriptions. Red 
predominates as the ground colour throughout. The hexagon field of this carpet is found in only one 
other example, in the Topkapi Saray (Istanbul)4, but the format of the border with inscriptions located 
exclusively in the upper half can be seen in ten additional carpets, now in the Metropolitan Museum of 
Art (New  York), Walters  Art Gallery  (Baltimore), Topkapi  (Istanbul), and  several  private  col ections5. 
This carpet is also in a poor state of conservation and the abrash is located almost exclusively in the 
lower main border. 

The  third,  and largest,  carpet is  an  entirely  new addition  to  the  corpus  of  „Salting‟  carpets  and  a 
very important discovery. It was previously in the col ection of Messrs Perez (London) and published
by  Stanley  Reed (1967),  but not included  in the study  of  Franses  and  Mil s  (1999), and  only  finally 
recognized in the study presented here.  
It has a central medallion (as opposed to a niche) (figure 7), and an exceptional y high knot count 
(11.155 knots/dm2), considering that the pile is wool and not silk. This knot density is more than twice 
that of  the  famous  Ardabil  carpet (V&A)  and closer to rugs  made exclusively  of  silk, such as  the 
„Kashans‟ [20]. The central field has a dark red ground covered with	tchi clouds, a blue medallion in
the  centre, and  green corner  medallions.  In the  main border,  lobed  medallions  alternate  with the 
inscribed cartouches  on a  beige  ground. The  inscription appears  to be  a Persian  poem, which 
describes it as “precious royal carpet… under the feet of such King Suleyman” (see appendix I). Reed 
has  proposed that it was  dedicated to the  Persian  Shah Suleiman  (r.1667-1697)  [22]. This  poem  is 
potential extremely  important historical  evidence and  is currently  being considered for greater clarity 
for this study by Professor Wheeler Thackston (Harvard University) (forthcoming information).  
While the „Salting‟ prayer or niche rugs are well defined as a group stylistical y, the medallion group 
remains to be properly described, and few parallels can be found for the field design and inscription of 
the  Guimarãs  carpet, in contrast to the  prayer rugs. In general  terms, the  quarter  medallions 
decoration  in the field and  of the border can be related to four recorded examples in the Hermitage 
(St. Petersburg), Musée  Historique  des  Tissus  (Lyon), and  Musée  des  Arts  Décoratifs  (Paris)6. It is 
also in a poor state of conservation, and abrash is visible throughout the carpet: in the central field and 
main border, as well as al  the colours present. 

==DYES AND MORDANTS==
The  colours  used  in 16th-  and 17th-century  Persian  and Turkish carpets, referred  to here as 
“classical”,  were obtained from  natural  sources, namely  plants  or insects. It was  only  in the  19th 
century, with the  discovery  of  mauve dye by  Perkin (1856)	and several  coal-tar-based  dyes, that 
synthetic  dyes  were  adopted  in Iran  and  Turkey  to  produce  bright and vivid colours  [10].  Regional 
taste also played  a factor  in colour  choices  and in the  composition  of  the  colour  palette. The 
combination  of  contrasting  primary  colours  is  a  general  feature of  Turkish carpets,  while in  many 
classical  Persian  carpets  red  dominates  the  central  field and  blue  or green are used  for the  border,
and in the  decoration, secondary  colours  are juxtaposed  beside primary  ones, such as  orange  with 
blue or green with yel ow [20]. Published analyses of the dyes used in Islamic carpets also reveal a 
number  of  interesting  trends  in the choice of  dyestuffs  in these two regions,  which are useful  for
studying the „Salting‟ carpets [10-11]. The number of analyses is too small to make firm conclusions,
but some general observations can be made.  
For the reds, the most common dye found in 16th- and 17th-century Turkish carpets was extracted 
from the roots of the madder plant (Rubiciaea family). 
It occurs more rarely in classical Persian carpets, there it is exclusively used for the oranges [8-9]. 
The main chromophores of madder are alizarin and purpurin, both compounds based on 1,2 dihydroxy 
anthraquinone  chromophore, table 1. In addition, dyes  prepared from  coccid insects  were popular  
[23]; Porphyrophora genera have been found in Persian silk carpets, such as the “small silk Kashans”
[8,23].  Other  red dye sources  of  animal  origin were used  in Iran  and India, such as  lac-dye  (Kerria
genera) [3], not found in Turkish carpets, along with cochineal [11]. Lac dye and cochineal are easily 
distinguished by the presence of laccaic acids and carminic acid chromophores, respectively (table 1). 
Lac-dye  has  been  identified in a  large  number  of  classical  Persian  carpets  [8-9]  usual y  in  the red 
ground; its  presence in the  „Saltings‟  is  potentially  an  important  indicator  for determining  their 
provenance and for this reason is discussed extensively below [1]. 
The blues in historical carpets from both Iran and Turkey use exclusively 
indigo dye extracted mainly from the indigo shrub (Indigofera tinctoria L.) or
from dyer‟s woad (Isatis tinctoria L.) [12,18]. The main chromophore of indigo
is  indigotin  (figure 8)  and its  identification in historical  rugs  is  an  important
indicator  of  authenticity,  as  indigo  was  the  most successful  and prolific natural blue prior to the invention of synthetic dyes in the 19th century.  
The  yellows  in classical  Persian  carpets  were generally  obtained 
from local plants, such as asbarg (Delphinium semibarbatum) which is 
composed  primarily  by  quercitin, kaempferol  and isorhamnetin 
chromophores. Weld	(R. Luteola),  rich in luteolin-7-O-glycoside  was 
also found in Persian carpets (figure 9).  
The  use of  related plants  with luteolin glycoside  derivatives  [3,11] 
have  been identified in  both Persian and  Turkish rugs. The  luteolin-based compounds are very well documented and considered characteristic of an ideal dye7, and thus 
commonly  identified  in high quality  textile products  [4]. By  contrast, the  yel ows  found  in historical 
Turkish rugs do not reflect the use of asbarg, as it does not occur in Turkey; according to the literature, 
it is unlikely that this plant was ever imported into Turkey, as more than enough indigenous yel ow dye 
plants are available for use [11,18]. 
Curiously, the oranges in Persian carpets do not include lac-dye, but, instead, madder mixed with
a yel ow dye. As for Turkish rugs, the results obtained do not offer a precise conclusion; although it is 
generally agreed that some Turkish plants such as chrysanthemus (Chrysanthemus coronarium L.) or 
dyer‟s sumac (Cotinus coggygria SCOP.) are used to obtain this colour [18].  
Greens in both Turkish and Persian rugs are obtained by using a mixture of indigo with a yel ow
dye [11].  
In both Iran and Turkey, the browns and blacks  are obtained using tannins 
from local sources, for example, knobby or gallnut oak (Quercus spp.). 
Beige colours were often obtained with natural wool in both Persian [8-9] and 
Turkish Carpets [18]. However, in some Persian carpets, analyses of beige areas 
have shown the presence of the juglone chromophore probably obtained from the walnut tree (Juglans regia L.) (figure 10). This dyestuff does not appear to have been identified in Turkish carpets [11]. 
It should be mentioned that some of the dye analyses reported in the literature were performed with 
extraction  methods  based  on  strong acids  (usually  with HCl)  [7,24]  and, even  with Thin Layer 
Chromatography (TLC) [11]. TLC was used before HPLC-DAD, but does not provide complete results, 
as it is not possible to identify the precise chromophores and hence the correct dye-source [4]. 
The  HCl  method  is  an  aggressive extraction  method  which can promote  the degradation of 
chromophores  in some dyes. In yellow dyes, it can cause decomposition of  the  glycosidic  dye 
compounds  to their  parent aglycons, with consequent loss  of  information  about the  original  dye 
source. For example there  are numerous  plants  with  different luteolin glycoside  derivatives  such as 
weld	(Reseda luteola  L.),	dyer‟s  greenwood  (Genista tinctoria  L.), true chamomile (Matricaria 
chamomil a L.) or three-leafed sage (Salvia Triloba L.) [18]. When these plants are submitted to a HCl 
extraction, the  various luteolin glycosides present are destroyed during the process and usual y only 
luteolin is detected in the final dye extract. Frequently, the identification of luteolin in dyed yellow fibres 
is attributed to the presence of weld, one of the most stable yel ow dyes. Nevertheless, when the HCl 
extraction method is used, it is not possible to exclude the utilization of other related plants  such as 
the ones mentioned previously. Therefore, mild extraction procedures such as those used in this work, and sucessfully  reported in literature, are necessary  to obtain  greater    information about the dye 
source [4,7]. 


==Mordants==
In order to  attach these natural  dyes  to the  textile fibre  to
ensure they  are fixed permanently  and will not bleed, a metal ic 
ion  (mordant)  was  usually  applied  in both  Persian  and  Turkish 
carpets  (figure 11). The  mordant  acts  as  an  intermediary 
 
between the fibre and the dyestuff, and is usually obtained from a
mineral  source  (salts  in  the  form  of  crystals, efflorescence  or 
crusts, mud, metal oxides). The choice of mordant can determine 
both the wash-resistance of a colour and its intensity [12,18].  
The presence of aluminium, iron and copper as mordents has 
been observed in both classical Persian and Turkish carpets	for 
different colours [9,18]. For the yellows and reds usually an alum 
mordant  was  used,	while  in the dark  browns  the presence of  an iron  mordant  has  been  detected 
[9,18]. For some beige colours, no mordant was detected at all. Also in blue fibres dyed with indigo, no 
mordant  is  applied, as  indigo is  a vat dye which precipitates  in the  fibres  through an  oxidation-
reduction reaction [4,8]. 


==LAC DYE LIBRARY==
The  Indian  and South Asian origin of  lac-dye  would appear  to support a Persian  (as  opposed  to 
Turkish) provenance for the  „Salting‟  carpets. However, no  study  to date has  used  HPLC-DAD to 
differentiate  lac-insect species  to identify  locations  of  dye production. Indeed, of  the  lac  species  of 
Asian origin belonging to the Kerria genera (ca. 26 species) [15,25], only Kerria lacca Kerr has been 
characterized  using  HPLC-DAD [6,12,26]. In this  study, for the  first time, 82  lac-insect historical 
samples of unknown species Kerria genera and 2 samples from Paratachardina genera, with diverse 
provenance, as  wel   as  from  different host plants,  were analyzed  using HPLC-DAD. This  data  base 
methodology  is  applied here for the  first time to lac-dye represents  a first step  towards  developing 
necessary research on lac-dye species, to reveal greater information about the precise source of the 
red dye and therefore assist with provenance identification of historical textiles.   

==Lac Insects== 
Lac-dye was highly appreciated in Asia and is documented in India as a textile dye, from at least 
the 4th century BC [12]. Lac-dye was also known and used by the Romans for dyeing purposes since 
the 2nd century AD, which suggests that lac  was probably one of the first insect dyes  to be used in 
West Asia [26]. The dyestuff is obtained from female scale insects  of the Kerridae family  (figure 12) 
which contains nine genera and approximately 100 described species. Lac-dye can be obtained only
from Kerria and Paratachardina genera [15].
The  Kerria  genera comprises  26  species  [15],  which can  be found 
mainly  in  India,  China, Taiwan,  Sri  Lanka, Australia, and Pakistan, and 
the Kerria lacca Kerr species is the most important (Appendix II). There
are several  publications  about the  Kerria  species  [5,15,28]; however, 
none  of  them  have used  molecular taxonomic  studies  for confirmative
species identification. According to Penny Gul an9 it is essential that fresh
specimens are collected from Asia and molecular taxonomic studies are 
performed to correctly identify the lac insect species as has occurred, for 
example, for the Paratachardina species [29-30].  
The  Paratachardina  genera comprises  a total  of  nine species,  which 
have been  recently  revised, mainly  from	China, India, Sri  Lanka,
Philippines and Papua New Guinea [29]. However, species belonging to
Paratachardina genera do not produce lac-dye of commercial importance and mainly Kerria spp. are exploited [15]. Nevertheless, as the Paratachardina genera is the only one 
for which accurate taxonomic  studies  have been  published, it was  included in this  study  and 
specimens were analysed. 
Several  entomologists  supplied  five  Kerria spp  and two  Paratachardina	spp  correctly-identified 
insect-specimens  for HPLC-DAD analysis. As  the  taxonomy  of  the  Kerria  genera is  not well 
established (Appendix II), it was not possible to obtain reliable insect-specimens for all of the 26 Kerria
species  described  in the literature. Hence,  in this  study,  as  a first approach to studying the	Kerria
insects, 76  unknown historical  lac-dye insect sources  from  different geographical  regions  (India, 
Australia, Taiwan, Vietnam, Laos, Singapore, Pakistan, Bangladesh Sri-Lanka and others), as wel  as 
from  different host plants  (Ficus  spp., Schleichera spp., Butea  spp., Zizyphus  spp. and Shorea sp.)
(figure  13),  were  characterized  by  HPLC-DAD and  submitted to PCA  analysis.  These 76  historical 
sources of lac-dye, dating from the 19th century and belonging to the Royal Botanic Garden at Kew, 
provide a good starting point for establishing similarities between the sources and for identifying insect 
groups according to their provenance and host plants. In future research, these insect samples should 
be submitted  to  molecular  taxonomic  studies  and  compared  with  the 26	Kerria  species  correctly-
identified by entomologists. 


==Lac-dye==
The process of production of lac dye involves thousands of lac insects attach themselves to twigs 
and the fleshy young leaves of suitable host plants (figure  13), from  Moraceae, Sapindaceae  or Dipterocarpaceae  families, usually in November [31-32]. By March, each insect covers itself with a 
predator protection layer, exhibiting the appearance of a red oval-shaped, smoothly polished lifeless 
sack entirely filled with a beautiful red liquid [33].  The coated branches of the host plants are cut and 
harvested  as  sticklac, which contains both	resinous matter and the  red  organic  dye composed  of 
laccaic  acids. In  order  to obtain the  pure deep red  colouring  matter  suitable for dyeing  purposes, 
known as  lac-dye, it is  necessary  to perform  a simple water  extraction  process. In extractions with 
100% water, the principal chromophores of lac-dye, namely laccaic acid A, B, C and E, are obtained in 
the  final  water extract  (table 2). On the  other hand, ethanolic  extractions  of  the  sticklac  product wil  
remove only the resinous matter which is composed mainly of yellow compounds (table 2).
Lac comes onto the market as a raw product, occurring in many forms; among the most common 
are the already  mentioned  sticklac,	grain/seed  lac 	(material  removed  from  the  twigs), and a 
subsequent variety which is fused and cast into moulds, caked lac  or lac dye (table 2) [33]. The pale 
perforated  kind, in which no insect exists, constitutes  the  material  for shel ac  employed  for varnish-
making.   

==Experimental==
A detailed characterization of the materials present in the three Guimarães carpets was performed 
using  different methods. The  fibres  were  identified  by  optical  microscopy  in  longitudinal  section;  the 
colours  were  characterized using  colorimetric  measurements; all  dyes  were  identified  by  High-
Performance Liquid Chromatography  with Diode  Array  Detector  (HPLC-DAD)  and when  necessary 
with  Mass  Spectrometry  (LC-MS);  all  mordants  were analysed  by  Inductively  Coupled  Plasma with Atomic  Emission  Spectrometry  (ICP-AES); and  the  metal  threads  were  characterized  by  Energy 
Dispersive X-ray Fluorescence (µED-XRF) and Scanning Electron Microscopy (SEM-EDX). C14-AMS 
analysis was performed in one sample from the medallion carpet. 
Prior to HPLC-DAD identification of the red dyes, several historical lac-dye sources were analyzed 
by HPLC-DAD and submitted to a PCA analysis. Also several parameters in the samples preparation 
were tested in order  to obtain HPLC chromatograms  with good resolution.  For more details  about 
experimental proceedings see Appendix III. 

==Results and Discussion==
The two Guimarães prayer carpets, Benguiat (PD77) and Duff  (PD78), share features commonly 
identified in the „Salting‟  group, namely the presence of  a bright colour palette, silk foundation, wool 
pile, metal-wrapped silk threads, arabesques designs, religious inscriptions and small dimensions. The 
Medallion carpet, by  contrast,  is  distinguished  by  the  absence of  religious  features, in  both its 
composition which lacks  a  prayer niche  and inscriptions  which are not excerpts  from  the  Koran but 
probably a secular poem. Analyses undertaken thus far on the „Salting‟ group have focused mainly on 
the  prayer type, and hence, the  study  undertaken  here of  the  Medal ion carpet offers  new  and 
important data for appreciating this under-studied type. With the aim of tacking the questions posed at 
the start of this study, concerning the date and provenance of the carpets, the results of the diverse 
analytical techniques utilized here (described above) are integrated and presented below, according to 
the  standard method  of  preparing  the materials  for constructing a carpet: beginning with the  fibres,
followed by the colours and dyes (with special emphasis given here to lac-dye), and ending with the 
metal ic  threads, which  were usual y  purchased  from  an  external  source outside the sphere of  the 
carpet-making workshop. As C14 offers exclusively evidence for dating, it is dealt with at the end. 


==Metal Threads==
Microscopic  analyses  indicated that the  metal  threads  used  in the  three  Guimarães  carpets 
comprise a metal  strip wound (S-direction)  around  a  yellow  silk  core,  so that the  edges  of  the  tape 
abut  but do  not overlap.  This  allows  the  colour  of  the  substrate  –  in this  case yellow  silk  -to play a
greater  role  in the  final  visual  effect, enhancing  the  golden sheen  of  the thread  (figures  1 and  2, 
appendix VIII). This technique gives a smooth, as opposed to a rough, irregular surface to the thread 
[38]. Measurements for the metal strip for the Benguiat Rug (c. 0.23 mm wide and 0.015 mm thick) are 
in accordance with previous results reported for „Salting‟ carpets (ca. 0.25 mm wide) [38]. 
 XRF and SEM analyses demonstrate that the metal lamina of the Guimarães threads is composed 
mainly of silver (Ag), gold (Au) and copper (Cu) (figures 2 and 3, appendix VIII). In the prayer rugs, Ag
was major element (circa 80%), followed by Au (circa 10%), Cu (circa 2%), with vestigial amounts of 
tin (Sn) and lead (Pb). Previous results for the „Saltings‟, however, identified the presence of Pb, but 
not Sn [38].   
The  SEM analyses  also  show  considerable degradation  of  the exterior surface,  particularly  in 
places  where sulphur (S)  and chlorine  (Cl)  were identified  and  probably  the  result of  Ag  corrosion 
products (figures 5 and 6, appendix VIII). By contrast, the inside of the strip, which is in direct contact 
with the silk  core, and therefore protected  from  contact with environmental  conditions, and is  better 
preserved. Analyses here revealed the presence of Au in higher amounts, in contrast to the exterior 
where it presumably had been worn away.
It is  difficult to establish  from  these analyses  with  certainty  how the  strips  were made, but  it is 
possible to identify the presence of a silver strip with a layer of gold. This is consistent with the method 
previously  proposed by  Eiland for the  „Saltings‟  and „Polonaise‟  carpets, in which	a  silver sheet is 
covered with a thin layer of mercury (Hg) to create an amalgam for gilding, and then forged to a very 
thin gauge and hand-cut [38]. However, confirming the use of this technique requires further research. 

==Concluions== 
The  publication of  the  „Salting‟  carpets  in 1999 represented an  important step  in  presenting  the  first 
corpus of identified carpets. However, doubts remained about their precise date and provenance. The 
rediscovery of two prayer carpets in Guimarães, and a third Medallion carpet which was not included 
in the  corpus  in 1999, offered a significant opportunity  for returning  to these  questions  using  an 
interdisciplinary approach and the application of a wide range of analytical techniques.
The  dyes  identified in the  three  Guimarães  carpets  point  consistently  to an  Iranian 
provenance. Use  of  lac-dye  in  the reds  and  pinks, but  madder  (and  not lac)  in the  oranges, is 
consistent with other  results  for  carpets  associated with Iran  in Portuguese collections. Together, 
these studies  reinforce the  notion  that this  palette is  a classical  characteristic  of  Persian  (and not 
Turkish) dyeing practices, especially as the Medal ion Carpet has an inscription written in Persian.  
Through  the  development  of  the  HPLC-DAD database for lac-dye, it was  possible to 
distinguish Kerria and  Paratachadina genera, for the  first time, and  to confirm  that only  Kerria has 
sufficient red-dye  chromophores  to  be  used  for  dyeing  textiles.	It  was  possible to  discriminate  the 
samples  from  the  Kew Garden col ection,  according to their  composition  and eventually  their  origin.
The red textile samples from the Guimarães rugs appear to be more similar to insect samples from
Pakistan, in close proximity to Iran. Further development of the database is required to establish more 
rigorous  comparisons  between  historical  textiles  and lac-insect sources. This  can only  be  achieved 
with the col aboration of entomologists, and the  identification of reliable Kerria  insect samples  which
have been submitted previously to molecular taxonomic studies, and for which their precise origin and 
host plants are wel  documented.
For the  browns  and yellows  in the carpets, tannins  and  probably	weld, respectively, were 
found. As  these are usual y  obtained  from  local  sources, further research could offer additional 
information for narrowing the provenance of the carpets. In this respect, it is noteworthy that the brown 
used in the Medal ion Carpet is comprised of ellagic acid, not encountered in the Guimarães prayer
rugs or other „Saltings‟ analysed [11], but previously identified in the „small silk Kashan‟ in the Museu 
Nacional  de  Machado Castro [20]. This  latter carpet can  be dated  to the  second half  of  the  16th 
century and represents the height of Safavid carpet production. The presence of this compound draws 
attention to the  importance of  further study  of  the  browns  in Islamic  carpets, something which has 
rarely  been considered  in the  past, owing  to the  apparent insignificance of  this  colour  to the  overall 
visual effect. No significant difference, however, was observed between the three carpets in relation to 
the mordants used, which it was hoped would be a distinguishing feature of provenance. The abrash
observed reflects  differences  in dye procedures, and the  concentration of  alum  present, and not 
degradation for light exposure.  
As for the chronology of the carpets, al  the data presented here supports an early date, such 
as the exclusive use of natural dyes and the non-industrial construction techniques used to make the 
metal  threads,  which have  also  been  identified  in  other  classical  Persian carpets. Moreover, as  “a 
carpet can only be as old as its design” [10], al  of the stylistic features are also consistent with Safavid 
art. Although these few elements are significant, they are not necessarily conclusive, of course, and 
only the AMS Radiocarbon results for Medal ion Carpet confirms a more precise date between the late 
15th  and  mid-17th  century.  This  result is  extremely  important, as  it confirms  that it is  an  authentic 
Safavid rug, as expected from its extraordinarily high quality (11.155  knots/dm2), rarely seen in wool 
carpets. However, Reed‟s proposal that the carpet was intended to be placed “under the feet” of the 
Persian Shah Suleiman I (r. 1666-1694) is not directly confirmed, as the dates of this Shah are slightly 
outside the  parameters  of  the  AMS  Radiocarbon results. Instead, a date for the  carpet in the  16th 
century seems more likely, which raises the possibility that it could have been made especial y for the 
Ottoman  Sultan  Suleyman  the  Magnificent (r. 1520-1566), and  possibly  sent in 1556, as  original y 
proposed by  Mills  and Franses  [1]. This  proposal, if  it can  be  supported by  additional  scientific  or
historical  evidence, would  make this  carpet extremely  important and  only  the second example, 
besides the Ardebil (V&A), to offer a historical name and related date.   

==

==Bibliography==
 
1. Mil s, J., and M. Franses. “Salting  Carpets.”  Oriental  Carpets  and Textile Studies. Danville, 
Califórnia, (1999). 
2. Canby, S. R. Shah ´Abbas, The remaking of Iran . London: The British Museum, (2009). 
3. Walker, D. “ exhibition catalogue.” Flowers underfoot – Indian Carpets of the Mughal Era. New York: 
The Metropolitan Museum of Art., (1997). 
4.	Ferreira, E., et  al. “The  natural  constituents  of  historical  textile dyes.”  The  Royal  Society  of 
Chemistry, 33, (2004), 329-336. 
5.	Wouters, J. “High  Performance Liquid Chromatography  of  Antraquinones: Analysis  of  Plant and 
Insect Extracts and Dyed Textiles .” Studies in Conservation, vol. 30, No. 3, August (1985), 119-128. 
6. Wouters, J. and Verhecken A. “The scale insect (Homoptera: Coccoidea). Species recognition by 
HPLC and diode-array analysis of the dyestuffs.” Annls Soc. ent Fr. (N.S.), 25 (4), (1989), 393 - 410. 
7. Zhang, X., Laursen, R. A. “Development of Mild Extraction Methods for the Analysis of Natural Dyes 
in Textiles  of  Historical  Interest Using  LC-Diode Array  Detector-MS.”  Analytical  Chemistry, 77  (7) , 
(2005), 2022 - 2025. 
8. Valssassina, M., et al. “The  Colour  of  Carpets.”  In  The  Oriental  Carpet in Portugal: Carpets  and 
Paintings, 15th-18th  centuries, exhibition catalogue, from  J. Hallett  and  T. Pacheco. (eds.)  Pereira. 
Lisbon: Museu Nacional de Arte Antiga., (2007). 
9. Armindo,  E., et  al.	“A  Persian Carpet‟s  Paradise  Garden:  Discovering  Historical  and Technical 
Aspects  Through  Carpet Conservation  and  Restoration.”  ICOM-CC 15th Triennal  conference New
Delhi. (2008), 960-966. 
10. Böhmer, H. and Enez,  N. “A  Topkapi  Prayer Rug: Arguments  for a Persian  Origin.” HALI,  41, 
(1988), 11-13. 
11. Enez, N. “Dye Research on the Prayer Rugs of Topkapi Col ection.” Oriental Carpets and Textile 
Studies. California: Berkeley, (1993), 191-204. 
12. Cardon, D. Natural dyes. London: Archetype publications, (2007). 
13. Wouters, J.and Verhecken A. “Potential Taxonomic Ppplications of HPLC Analyses of Coccoidea 
Pigments (Homoptera: Sternorhyncha).” Belg. J. Zool., 121, (1991), 211 - 225. 
14. Sousa, M. M., et al. “Flavylium Chromophores as Species Markers for Dragon's Blood Resins from 
Dracaena and Daemonorops Trees.” Journal of chromatography A (volume 1209), (2008), 153-61. 
15.	Sharma, K. K., et al.  “Role of  lac  culture in biodiversity  conservation: issues  at stake and
conservation strategy.” Current Science, 91 (7), 10 de October (2006), 894-898. 
16. Halpine, S. M. “An Improve Dye and Lake Pigment Analysis Method for High-Performance Liquid 
Chromatography and Diode-Array Detector.” Studies in Conservation, vol. 41, No. 2, (1996), 76-94. 
17. Murray, Jr. Eiland. “The Saltings Reconsidered.” Oriental Carpets and Textile Studies, V, 2nd part.
Danvil e, Califórnia, (1999). 
18. Böhmer, H. Koekboya Natural Dyes and Textiles. Germany, (2002). 
19. Frade, C., P. Cruz, E.  Lopes, M. Sousa, J. Hal ett, Santos, R. and A. Aguiar-Ricardo. “Cleaning
Classical  Persian Carpets  with Silk  and Precious  Metal  Thread:  Conservation  and  Ethical 
Considerations.” Lisbon: (accepted) to ICOM-CC 16th Triennial Meeting, (2011). 
20. Hal ett, J. and Pereira, T. P. (eds.). The Oriental Carpet in Portugal: Carpets and Paintings, 15th-
18th centuries. Lisbon: Museu Nacional de Arte Antiga, (2007). 
21. Wearden,  J. Oriental  Carpets  and  Their  Structure Highlights  from  the  V&A  Col ection. London: 
V&A Publications, (2003). 
22. Stanley,  R. Oriental  Rugs  and Carpets, Pleasures  and  Treasures. London: George  Weidenfeld 
and Nicolson Ltd, (1967). 
23.	Serrano, A. “Cochineal, A  Precious  Source of  Red:  Cochineal  Dyes  Identification  by  High 
Performance Liquid Chromatography  with Diode  Array  Detection  and  Principal  Component Analysis 
Master Dissertation, (2010). 
24. Valianou, L., et al. “Comparison of Extraction Methods for the Analysis of Natural Dyes in Historical 
Textiles by High-Performance Liquid Chromatography.” Anal. Bioanal. Chem., 395 (7), (2009), 2175 - 
2189. 
25.	United  States  Department	of  Agriculture	-  Systematic  Entomology  Laboratory. 
http://www.sel.barc.usda.gov/catalogs/kerri da/KerriaAl .htm (accessed in April, 2010). 
26. Hisao, O., et al. “Separation  of  lac  dye components  by  high-speed  counter-current 
chromatography,.” Journal Chromatography A (813), (1998), 71-77. 
27. Enez,  N. and Böhmer, H. “Ottoman  Textiles:  Dye analysis, results  and  interpretation.”  Dyes  in 
History and Archeology 14. Amesterdam, (1996), 39-44. 
28.	Karapanagiotis, I. and  Chryssoulakis, Y. “Investigation of  red  natural  dyes  used  in historical 
objects by HPLC-DAD-MS.” Annali di Chimica, 95, (2005).
29. Kondo, T. and Gullan, P. “Taxonomic review of the lac insect genus Paratachardina Balachowsky 
(Hemiptera: Coccoidea: Kerri dae), with a revised  key  to genera of  Kerri dae and description  of  two 
new species.” ZOOTAXA, 17 de October (2007), 1-41. 
30. Schroer, S. et al. “The  genetic  diversity, relationships, and potential  for biological  control  of  the 
lobate  lac  scale,  Paratachardina pseudolobata  Kondo &  Gul an  (Hemiptera:  Coccoidea: Kerri dea).” 
Elsevier, Biological Control, 46, (2008), 256-266. 
31. Hofenk  de  Graaff, J. H. The  Colourful  Past	–  Origins,  Chemistry  and  Identification  of  Natural 
Dyestuffs. Switzerland: Abbeg-Stiftung, (2004). 
32. Phipps, E. Cochineal  red: The  Art History  of  a Color, New  York: Metropolitan  Museum  of  Art, 
(2010). 
33. The Natural organic colouring matters, Monographs on industrial chemistry. Edited by Sir Edward 
Thorpe. 
34. Szostek, B., J. Orska-Gawrys, I. Surowiec, e M. Trojanowicz. “Investigation  of  natural  dyes 
occurring in historical Coptic textiles by high-perfomance liquid chromatography with UV-VIs and mass 
spectrometric detection.” Journal of Chromatography A, 1012, (2003), 179-192. 
35. Valssassina, M. “Work in progress.” comunica ão pessoal. Caparica, (2007). 
36. Balfour-Paul, J. Indigo. London: Archetype Publications, (2006). 
37. Karapanagiotis, I., Minopoulou, E., Valianou, L., Daniilia, S., Chryssoulakis, Y. “Investigation of the 
colourants  used  in icons  of  the  Cretan School  of  iconography.”  Analytica Chimica Acta 647, (2009), 
231-342. 
38. Eiland III, M. L., “Analysis of Metal Threads: New Light on the Salting Problem”. Oriental Carpets 
and Textile Studies. Danvil e : International Conference on Oriental Carpets, (1999), Vol. V. 
39. Tímar-Balázsy, A. and Eastop, D. “Metal threads.”Chemical Principles of Textile Conservation, de 
A. and Eastop, D. Tímar-Balázsy, 128-138. Butterworth Heinemann, (2004). 
40.	Zhang, X., Boytner, R., Cabrera, J., Laursen,  R., “Identification  of  yel ow  dye types  in pre-
colombian Andean textiles”, Analytical Chemistry, 79 (4) (2007), 1575 - 1582. 
41. Marques, R., Sousa, M., Oliveira, M., Melo, M., “Characterization of weld (Reseda Luteola L.) and 
spurge  flax  (Daphne  gnidium  L.)  by  high performance liquid  chromatography-diode array  detection-
mass spectrometry in Arraiolos historical textiles” Journal of Chromatography A, 1216 (2009) 1395 - 
1402. 
42. Naes, T., Isaksson, T., Fearn, T., Davies, T. Multivariate Calibration and Classification. Chichester 
: NIR Publications. 
43. Preisner, O. E., Lopes, J. A., Guiomar, R., Machado, J., Menezes, J. C., “Uncertainty assessment 
in FT-IR spectroscopy based bacteria classification models”, Chemom. Intell. Lab. Syst. 94 (1) (2008) 
33 – 42. 
44. Stuiver, M. e Reimer, P.J. Extended 14C Data Base and Revised CALIB 3.0 14C Age Calibration. 
Radiocarbon 35 (1), (1993), 215-230. 


[[Category:carpet's story]]
[[fa:داستان‌های فرش]]