Yale University Library

Jka67 A16
902Bcg

l ¦¦ '

?¦

feu

YALE UNIVERSITY
LIBRARY

THE AMERICAN HANDBOOKS
ON FINE AND APPLIED ARTS

ANTIQUE RUGS FROM
THE NEAR EAST BY
WILHELM BODE AND
ERNSTKOHNEL

Prayer Rug, Rsia Minor, about 1500.
Kaiser Friedrich Museum, Berlin.

ANTIQUE RUGS
FROM THE NEAR EAST

BY

WILHELMBODE

THIRD REVISED EDITION
WITH CONTRIBUTIONS BY
ERNST KOHNEL

TRANSLATED BY
R. M. RIEFSTAHL PH. D.

E. WEYHE
710 LEXINGTON AVENUE
NEW YORK
1922

Preface to the second edition.
When "Antique Rugs from the Near East" appeared in 1902, it was the first
comprehensive monograph on this subject. Since that time new handbooks
on "antique Oriental rugs" were issued almost yearly. However, these have done
little to broaden our knowledge of the classic period of this industry as they are
usually published by dealers and are generally limited to the later specimens which
are current in the trade. The scientific study of the history of rugs has been ad
vanced mainly by the publications of the great exhibitions of Mohammadan art and
by the comprehensive work of F. R. Martin, "A History of Oriental Carpets be
fore 1800".
The first edition gave new points of departure to research and materially in
creased the interest of museums and collectors in antique rugs. However, research
of late years has afforded supplementary information and in some respects the first
volume has been superseded by more recent works. Consequently in our new
edition we substituted well founded hypotheses for worn out theories. Such a
process is necessary as there are practically no irrefutable facts in this field. At the
same time, we have arranged the text in separate chapters to organize the volume
more clearly and to emphasize its character as a handbook. The illustrations were
also looked over and several of the less important examples were discarded, while
other, particulary characteristic rugs which became well known at the Munich Ex
hibition, 1910, were substituted. In addition, questions of ownership were verified as
far as possible. In the final proofs we used information regarding the rugs belonginq
to American collectors, for which we are indebted to Dr. Valentiner.
The reader will find the most important changes in the text of the first edition
in the discussion of the vase rugs and the Armenian-Asia Minor groups. Besides
other incidental additions, new material has been used in the discussion of the
Polish rugs in tapestry weave, the Konia rugs, the late Damascus patterns, the so-
called Armenian forerunners and successors of the Sefevi patterns and in the chapter
on Indo-Persian rugs. On the other hand, we have entirely dropped the Spanish
group which was discussed very thoroughly before, as it is not strictly within the
subject. The detailed description of several examples remains entirely unchanged.
Nor is there a noticeable increase in the number of the paintings which the author
was the first to use very extensively in the identification of certain patterns.
W. Bode. E. Kuhnel.

Preface to the third Edition.
The most recent developments in the study of antique rugs and the discovery
of new examples have necessitated numerous corrections and additions in the
text of the third edition. Besides these changes, we have embodied in the present
text information which has been given by Dr. R. M. Riefstahl in the English edition
of this handbook, now being published by E. Weyhe, New York. We have espe
cially taken account of Dr. Riefstahls notes regarding rugs owned by American
museums and private collectors.
The text and illustrations have been separated in order to simplify reference
to the reproductions. In addition, several particularly important specimens have been
added to the list of illustrations (Figs. 19, 25, 38, 45, 51, 69).
W. Bode. E. Kuhnel.

Preface by the Translator.
American collectors have often expressed the desire of having a translation of
"Vorderasiatische Knupfteppiche" , the handbook by Bode and Kuhnel, which
is, and remains, the leading handbook on the early Oriental rugs. The present
translation follows the German text as faithfully as possible and is the same as
that of the German third edition. In some cases the translator, living in America,
was able to give some new information which has been placed in the "Notes by
the Translator" at the end of the book and has partly been used in the third German
editipn. The translator followed a systematic terminology in the designation of the diffe
rent parts of a rug. The words "border" and "center field"; "main border", "small
borders", and "guard stripes"; "ground of the rug", "central medallion", "attached
medallion", "corner motif", are always used strictly in the same sense. The "car
touche rug" means the type of composition which is called "compartment rug" in
the catalogue of the Victoria and Albert Museum.
In several cases the translator obtained information about matters which he was
not able to decide by himself. Mr. A. F. Kendrick of the Victoria and Albert Museum
in London kindly gave information about several matters. The translator also
wishes to express his thanks to Mr. C. F. Williams in Norristown, Pa., to Mr. John
D. Mcllhenny in Philadelphia and to Mr. P. M. Sharpies, Westchester, Pa.
New York, June 1, 1922. R. M. Riefstahl Ph. D.

Preface to the second and third Editions 5
Preface by the Translator ... 1
Introduction  9
The Persian Animal Rugs  15
Hunting Rugs — Silk Animal Rugs — Woolen Animal Rugs
Persian Rugs with floral Patterns  25
Medallion Patterns — Herat Patterns — Prauer Rugs — Vase Patterns — Garden
Patterns
Persian, so-called Polish Rugs  30
Knotted Polish Rugs — Polish Rugs in Tapestry Weave
Indo-Persian Rugs  35
So-called Armenian Rugs  . . :  37
Animal Patterns — Patterns with Flower Branches — Rugs Inspired by Sefevi Patterns
Early Anatolian Rugs  '.  41
The Konia Rugs — Animal Rugs in Old Paintings
So-called Ushak Rugs from Asia Minor  44
Smyrna Rugs — The Ushak Patterns — Prayer Rugs — Rugs with a white Ground
Rugs from Asia Minor; so-called Holbein Rugs  51
Patterns with Conventionalized Vines — So-called Holbein patterns
So-called Damascus Rugs  54
Patterns with Kaleidoscopic Forms — Turkish Plant Patterns — Turkish Prayer Rugs
Concluding Remarks  58
Notes by the Translator  63

Introduction.
Many of us will still remember the impression produced by "antique Persian"
rugs when, in the seventies, they were introduced by foreign agents
The artists were the first to throw out the insipidly colored Brussels carpets and
to replace them with richly toned Oriental rugs. The art loving public soon fol
lowed their example. Everyone was eager to acquire several small prayer rugs for
use on tables and floors, or as coverings for sofas and cushions. Indeed, it was
a distinction to own valuable specimens of the famous Persian antique knotted work,
as these rugs were termed in the art trade.
The interest thus awakened in Oriental rugs and the growing delight in colorful
interior decoration, soon brought the attention of the antique trade to truly old
specimens. These appeared in different places — in the Orient itself, particularly
in Turkey and several adjacent countries, and also in Italy, Spain and Portugal
Even earlier, silk carpets were eagerly purchased by several wealthy collectors who
laid especial emphasis upon the magnificent and artistic furnishing of their houses.
Other rugs, which had usually been discarded by churches and palaces because of
their more or less damaged condition, could be secured very cheaply in Spain,
southern Germany and especially in Italy. These were gladly purchased by various
museums and by artists who used them in furnishing their studios.
Thus scholars gradually became interested in Oriental rugs and began to study
them, though only incidentally and with hesitation. It became evident that the very
numerous specimens which dealers brought to Europe from the East were almost
all modern. But on the other hand, old paintings frequently bore reproductions of
rugs which were identical with a few in the antique trade. Then the philologists
began to study the inscriptions which appeared in many rugs. Such material, as
well as existent information from historical sources, were used for the determination
of age. The comparison of the decoration of rugs with that of architecture, metal
work, pottery, textiles and other manifestations of Oriental art, gave further indi"
cations. On the other hand, modern rugs made in the different regions of the Near
East and Central Asia were used to determine the origin of the antique specimens.
This method was especially popular among the large rug dealers who did buying
and manufacturing in those regions. Several exhibitions and subsequent publications
particularly furthered the understanding and interest in Oriental rugs-above all,
the great Vienna Rug Exhibition of 1891 , followed by the important publication
devoted to this exhibition and by a supplementary volume in 1907. In 1908 Martin's
history of Oriental carpets appeared and finally in 1911 came the publication on
the Mohammadan exhibition in Munich;
Among the various methods of research, that of studying the inscriptions had
seemed to be the most useful and dependable, as it is in the field of Mohammadan
architecture, metal work etc. But this method has thus far shown itself fruitless.

10 Bode-Kuhnel, Antique Rugs from the near east. 
if not untrustworthy. Up to the present time only two rugs have been found with
an unquestionable date (1539 A. D. Cf. Fig. 24 and 1584 A. D. of page . . ). The
attempts at more exact dating of other specimens which seemed to have significant dates
or inscriptions, have always met with strong doubts and even been proved partially
wrong, although such research has been staged with a great expenditure of learning.
Copies of the 19 th century and inferior later pieces have been declared to be five
hundred years old. These errors are explained by the unusual difficulties which research
encounters. Inscriptions on rugs are rare and as a rule difficult to read, for the Orientals
love to conventionalize their letters or to hide them in plant and animal forms.
The meaning, too, is difficult to understand unless the inscriptions consist only of
unimportant Koranic verses or sentences, as is usually the case. If an exceptional
specimen appears with the statement of a fact, the inscription generally gives the
name of a man who ordered the rug, or that of the ruler of some state. These
are easily confused, as the names are frequently repeated and localities are practi
cally never mentioned. We must also remember that rugs with inscriptions have
long been copied — especially those for purposes of worship. Such copies were
made in different periods, but the modern imitations of old rugs are particularly
faithful. This is a further difficulty for research.
More practical than the philological discussion of inscriptions is the comparison
of the earlier rugs with various other Near Eastern crafts. The latter very often
contain an exact date, while in certain periods and in certain regions the same
forms of decoration are found simultaneously on rugs and on works of other crafts.
But here, also, research encounters definite limitations, which are only too easily
overlooked and occasion false conclusions. Sometimes certain ornaments are a
result of the technique. Hence they often recur with scarcely perceptible changes in
the most widely separated periods. Then, too, Oriental art is very conservative.
For instance, the decorative motives found in the earliest woven work of Meso
potamia have been maintained more or less faithfully for thousands of years up to
modern times, despite the tremendous, almost elemental revolutions which have con
tinually recurred and the agitations which unavoidably ensued. Wars and deva
stations in the train of invading hordes shook art and industry to their very found
ations and maintained a continuous ascent and descent of civilization. Such agitations
have repeatedly produced similar changes in the textile industries — a period of
deep decline and then a slow reawakening. Therefore, because of the old traditions
and models, rugs of different periods appear of the same technique and pattern.
Hence it is difficult, especially in the earlier specimens, to recognize whether a
certain rug shows the degenerate, coarse characteristics of decadence, or the rough
primitivity of an art which is again waking to life. Superficial and unintelligent
modern reproductions are frequently considered as examples of an old, primitive
art, while characteristic specimens from earlier epochs are taken for rough reproduc
tions from a later time.
Much valuable historical information had been expected from the study of
modern rug manufacture in the Near East. It was hoped to establish the traditions
of the old patterns in the designs now woven in different districts of Asia Minor,
Persia, Turkestan, Transcaucasia and other regions of the Near East, and through
studying them to discover the place where the antique rugs were made. Thus far
the results of this research are disappointing. We must not forget that it is princi
pally the dealers who immediately recognize a kindred and familiar modern pattern
in the old rugs. Without any knowledge of history they ascribe the old rugs to
the place where the similar pattern is still made. However, such conclusions are

Introduction. 1 1

only to be drawn with the greatest care, and only when a tradition of centuries
can be proved for rug knotting in a given city or region.
The nomadic life of many tribes of the Near East, the great changes and
migrations of several periods, the terrible devastations which frequently destroyed
entire tribes and their habitations, as well as the old, scarcely altered trading routes
and their lively commerce for thousands of years, made possible the interruption of
an old established industry in one district and the migration and continuation of
famous patterns in other regions which were often widely separated. We will see
later that this can actually be proved for a number of the most interesting types of
rugs. In this branch of research, too, almost no comparative study has been done
on the spot.
Thus far the study of specimens appearing on paintings has proved to be one
of the most trustworthy and fruitful sources of our knowledge of Near Eastern rugs.
In the present short essay on the evolution of the older rug industry in the Near
East the author has used his own preliminary studies in this field. Besides the
rugs themselves, the study of the rugs on paintings forms the basis of his con
clusions. This study is still particularly interesting because it shows the importance
to the Occident of both antique rugs and Oriental textiles. It is evident both in the
decorative and fine arts, but is particularly remarkable in the development of great
schools of colorist paintings — the Venetian, and to some extent, the Dutch school.
It is well known how much Venice has been open to the influences of Oriental
art because of her close connection with the East since the time of the Crusades.
A glance at the Venetian paintings of the 15 th century shows the city of lagoons
in half Oriental attire. In every representation of Venetian festivals and processions,
in the numerous pictures with views of the city or motives from the interior of
houses, Oriental rugs hang as festive decoration from windows and balconies. Rugs
are used on the floors and tables in the private house as well as in the public pa
lace; even the gondolas were canopied and covered with them. The churches were
particularly richly decorated with rugs. Upon the steps of the altars, before the seats
of the priests, and over the balustrades lay the beautiful rugs of the Orient and offered
an exquisite pleasure to the eyes of all. Inspired by the gorgeous fabrics, glass
wares, inlaid metal vessels, leather work and other examples of domestic art which
for centuries had been imported from the East or were made by "Turkish" artisans
in the Turkish quarter of Venice (later in the Fondaco de' Turchi) the applied
arts developed surprisingly in Venice itself. At the same time, the eye of the
painter was educated in this environment. Out of a prosaic, colorless school of
painting there developed in a few decades the greatest colorist school in the
history of art. Those who are acquainted with the Oriental rugs of the 15th cen
tury will feel in the pictures of the contemporary Venetians the same instinct for
the combination and choice of colors. Only the great masters of the 16th century,
Giorgone and Titian, freed Venetian painting from this dependence upon the deco
rative arts of the East and made it entirely self sufficient.
It was not only Venetian art whose coloristic development was essentially in
fluenced by the harmonies of the Near Eastern rugs. The Near East continued to
prove its creative power two centuries later when an art based on color developed
on this side of the Alps. This influence is felt in the Flemish, and still more in
the Dutch school. In the 16th century, Persian rugs were especially prominent
among the many products of the Orient, imported indirectly into the Netherlands in
the course of her relations with Spain. But in the following century the trade
with Persia, as well as the establishment of trading stations in Asia Minor, brought

12 Bode-Kfihnel, Antique Rugs from the near east. 
rugs to the Low Countries in greater numbers. A glance upon the manifold re
presentations of Dutch interiors will prove how frequently Persian rugs were used
on tables and floors in the houses of the upper middle class at that time. And
they were especially common in the houses of painters, as one can conclude from
the frequent appearance of Oriental rugs of different patterns in their pictures. It
was not only the Dutch artists who were unconsciously influenced by such colourful
objects in their surroundings. At the same period, the English, too, felt a desire
to decorate their homes with Oriental rugs. We know how much they did to re
vive the rug industry in India. And by means of this interest in rugs they kept
alive their sense for vigorous and powerful color, which in Germany had long been
dead. During the first half of the last century the rugs of the latter country had
shown an accumulation of harsh and discordant colors, which was followed by a
fashion for anemic pastel shades. Only the art of very recent years has sought to
overcome this defect. The intensive study of Oriental art has contributed a good
deal to this movement,
J. Lessing in his well known publication "Antique Oriental Rug Patterns", 1877,
was the first to state that only a few examples of antique Oriental rug knotting had
been preserved. He was also the first to approach this subject in a scientific way.
Riegl, too, held fast to this belief in his book "Altorientalische Teppiche", 1891.
This view was refuted, however, by the great exhibition of rugs in the Commercial
Museum in Vienna in 1891 and recently by the Exhibition of the Masterpieces of
Mohammadan Art in Munich, 1910. In the first of these exhibitions almost 150,
and in the latter over 200 antique rugs were placed on exhibition. In view of
the considerable number and variety of specimens there was no unanimity among
experts regarding the difficult questions of age and origin. They seemed rather to
despair of ever finding a solution. However, matters are no longer in such an un
fortunate state. Although we cannot as a rule date the rugs within the year or
the decade, nevertheless we have reliable evidence which enables us to give the
exact century for practically all specimens. For the sake of clarity one must put
aside, first of all, countless of the less valuable and only incidentally picturesque
and charming examples from the end of the 18 th and 19 th centuries, as well as
modern imitations. We must then attempt to survey the material available for
study in museums, palaces, church treasures, in private collections and in the art
trade. Perhaps the most comprehensive collection of antique Oriental rugs and the
most interesting because of its number of very early pieces belongs to the Mo
hammadan department of the Berlin Museum. This department is still lodged in
the Kaiser Friedrich Museum and is later to be taken over into the Asiatic Museum,
now in process of construction. The collections of the Austrian Museum of Art and
Industry in Vienna, the National Museum in Munich, the Manufacture des Gobelins
and the Museum of Decorative Arts in Paris, the Historical Textile Museum in Lyon,
the Metropolitan Museum in New York the Museums in Boston, Cleveland and
San Francisco and above all the collection of the Victoria and Albert Museum in
London, (which has particularly valuable pieces in the very best condition), are
of equal importance and useful because of their careful exhibition. The collection
of the Poldi Pezzoli Museum in Milan and that of the Berlin Museum of De
corative Art, which both possess a beautiful animal rug, are also rich in treasures.
The Ewkaf Museum in Constantinople, the Czartoryski Museum in Ccacow, the
Museums of Decorative Art in Cologne, Dresden, Leipzig and Budapest, the
Cluny Museum in Paris and others, have at least several good and interesting

Introduction. 13

pieces. The houses of almost all members of the Rothschild family, especially
those in Vienna and Paris, are rich in magnificent examples. A number of ex
cellent rugs are also owned by Professor Friedrich Sarre and other collectors in
Berlin, Dr. Fritz Harck in Seusslitz, W. von Seidlitz in Dresden, Baron Heinrich von
Tucher in Nuremberg and by several Austrian patrons of art, in particular Prince
Liechtenstein, Prince Schwarzenberg, Count Clam-Gallas, Dr. Figdor and others.
In the United States the collections of B. Altmann and Isaac D. Fletcher are now
in the Metropolitan Museum of Art. The rugs of the Theodore Davis Estate and
of the Joseph Lees William Memorial Collection are loaned to the Metropolitan
Museum in New York. We may further mention the collections of James F. Ballard
in St. Louis, James Deering in Miami, Florida, J. F. Mocllhenny in Philadelphia,
E. H. Myers in Washington. Mrs. H. Pratt in New York, Dr. Denman Ross of Cam
bridge, Mass., P. M. Sharpies in Westchester, Pa., P. A. Widener in Philadelphia,
Senator Clark in New York1. The painter E. Goupil owned a small, very choice
collection, including several silk rugs which are illustrated in the catalogue of the
public sale of his collection in 1888. The Yerkes Collection in New York was
one of the most important private collections ever brought together and was
sold at auction after the death of the owner. A larger number of pieces, especially
interesting from a historical point of view, are owned by the author. Besides,
there are several church treasures, especially those in Italy — the Cathedral at Padua,
S. Francesco at Brescia and marty others — which still have isolated examples of
good antique rugs, even though the hunt for such things during the last twenty
years has considerably diminished their number. The collections of various dealers
who only a few years ago had a valuable stock of antique rugs, have now been
almost completely bought out. Only Stefano Bardini in Florence, Kelekian in Paris
and New York and Beghian and Kafaroff, both in Constantinople, Bernheimer in
Munich, have still a certain number of antique pieces. Not a few of the important
rugs have found their way into the stocks of great New York dealers, such as Du-
veen Brothers and French & Co.
Out of this abundance a closer study reveals a number of none too numerous
types which repeat themselves with slight variations. The origin and development of
these types during the course of two or three centuries can be proved through
the representations of rugs in old paintings. Several of these types were so
widespread and have maintained themselves for so long that they still appear in
many dozen specimens; others appear to be unique and perhaps only originated
through particular circumstances. This book, does not embrace a complete history
of the art of rug knotting in the Near East and perhaps we will never have the
necessary foundation for such a history. However, the antique rugs which have
been preserved are almost entirely from the 16 th to the 18 th centuries, and only a
fewdate back to the 15 th. There are only two or three rugs known which could be
long to a still earlier period — to the 13 th century. Up to this same century we can
also trace them in pictures and miniatures. On the other hand, isolated rug-like
designs upon Sassanian silver vessels or Assyrian reliefs cannot be claimed with cer
titude as knotted rugs, as they were presumably executed in tapestry technique. Then,
too, the descriptions left by ancient authors of the magnificent rugs of the early middle
ages are too much adorned by the flowery speech of the Orient to evoke even a fairly
clear idea of the originals. Thus the sources of exact information for the history
of rugs in the Near East are exhausted. They only give the barest framework for
the history of the evolution of rugs. For this reason our historical survey must
be limited to the more recent period. While discussing rugs from the 16th and

14 Bode-Kflhnel, Antique Rugs from the near east.

17th centuries, only a few weak and uncertain gleams of light will be thrown upon
the darkness of the previous primitive periods.
Save for the hanging rugs in the mosques, the Orientals have used the knotted
rugs exclusively as a floor covering, both today and in ancient times. Hence we
are not concerned with all kinds of wall hangings, pillow coverings etc. The costly
knotted rugs, have always lain, and still lie today, in the middle of the room. This
makes it possible for anyone seated on the divans running about the wall to rest
his eyes on this masterpiece, the pride of the house. When walking about their
rooms the Orientals use only the small woolen runners of coarser workmanship
which run along the sides of the main rug in the center.
For the knotting technique2, as well as for similar general questions which
concern rug handicraft, the reader is referred to "Altorientalische Knfipfteppidie",
the very scholarly work of A. Riegl. For the sake of completeness let us quote
his excellent description of the technique.
"Wherever rug knotting is still done in the Orient, the technique is practically
the same. This results from the simplicity of the process. The worker grasps two
warp threads strung next to each other and lays across them a short piece of yarn,
about two and a half inches long and usually of wool8- He then pulls the two
ends of the yarn, first behind the warp threads and then between both of these
back again to the front. Both ends of the thread therefore rise above the warp
like a double tuft which is secured in place by a simple encircling of the warp
threads. Thereupon, the worker seizes the next pair of warp threads and repeats
the same process, and so on, all along the entire width of the warp,"
"When a row of tufts has been knotted in, the weave has to be consolidated.
As yet, only the two adjoining warp threads are held together by encircling them
with the same strand of woolen yarn. The warp threads have to be joined together
into a firmly constructed weave by the insertion of one, or generally two wefts,
as in plain linen weaves. The weft thus inserted is pressed closely down upon
the row of tufts and in this manner the weaver not only obtains a resistant texture,
but counteracts the loosening of the individual tufts."
" To simplify the work, the strands set aside for knotting are given a length
which makes it easy to interlace them between the warp threads. The practical, as
well as the artistic value of a knotted rug increases, if the nap is cut short as in
velvet, and not left rather long, as in plush. After the cutting, the weave must
be pressed closely together with a comb, as in a loose weave the wefts between
the rows of tufts might appear on the surface and the short knotted tufts might
work loose."

The Persian Animal Rugs.
Among antique Oriental rugs the silk rugs with animal representations deserve
particular interest because of their costly material, their beautiful coloring and
their excellent design. We have neither dated specimens nor any other documen
tary evidence to prove their date or place of origin; nevertheless, there is full
agreement about these points. For this reason the animal rugs are a particularly
fit starting point for our investigation.
The largest and most magnificent piece of this group and the most famous
antique rug at the present day is the so-called "Hunting rug", formerly in the
possession of the Austrian Imperial family and now property of the Austrian State
The "Hunting rug" (fig. 1, 2) owes its name to the peculiar composition of mounted
hunters which adorns the large center field. The salmon-colored ground is decorated
with numerous flowers attached to delicate vines. Some of these flowers are native
to Persia, others show forms of refined conventionalization. On this ground
horsemen move to and fro, pursuing lions, antelopes, ibexes, wild boars, hares,
foxes, jackals and other wild animals. The horsemen may be recognized as Persian
by their costume, likewise the horses and their trappings are characteristically Persian
and all the wild animals are of that country. The two winged figures which recur
regularly on the red ground of the broad border among delicate vines with flowers,
buds and leaves, may be characterized from their costumes as Persian "genii".
Representation, conception and conventionalization of the plants, as well as the
drawing of the animal and human figures and their costumes correspond throughout
with the decoration of miniatures, metal and leather work, and with other Persian
minor arts. The composition and design of the rug — the proportion of the center
field to the main border with its accompanying small borders on the inner and
outer side; the round, scalloped middle shield with smaller attached medallions
running the length of the rug, the wedge-shaped corner pieces conceived as quarter
divisions of the center field, even the characteristic masks (lions' and mens' heads).
inside the large flowers which occur in the outer border, all have analogies in other
branches of Persian decorative art of the middle ages.
In the large middle shield and in the four corner pieces which exactly corre
spond, we see a fight of fabulous animals represented on a ground of floral sprays..
Two dragons turned back to back, each show their teeth to a mythical giant
bird which swoops upon them. This representation is repeated four times in the
center of the rug in tasteful, symmetrical composition. Such phantastic creatures
have nothing to do with Persia. Indeed, the first glance shows that they are
borrowed from Far Eastern legend and art. The dragon in combat with the
phoenix is one of the most famous decorative motive in all kinds of Chinese works
of art. It has been interpreted as the official symbol of the Ming dynasty which
ruled China from 1368 for almost three centuries.4 In addition to this motif there

16 Bode-Kuhnel, Antique Rugs from the near east.

are other Far Eastern elements in this hunting rug. A peculiar gnarled ornament
appears in various forms and runs inconspicuously through the decoration of the
border. It is not at all Persian. It is one of the most important symbols of Chinese
mythology — the emblem of immortality — the "Tschi". It is variously formed:
either small, scalloped, compactly massed like a shell, drawn out like a ribbon, or
surrounded by conventionalized clouds and lightning. It corresponds exactly with
the sacred fungus of true Chinese conception. A glance at any collection of Chinese
art of the Ming period or later will prove this conclusively.
There is another hunting rug closely related to this magnificent piece. It has
almost the same dimensions and is also executed in silk and silver. This rug is
owned by Baron Maurice Rothschild in Paris (Fig. 3) and was found in Italy. About
1879 the Marchese Torrigiani of Florence sold it for #30 to the antiquarian Stefano
Bardini who sold it later to Baron Adolphe Rothschild for #6000. Today this rug
would be worth almost #200000. The composition is the same as in the Vienna
rug. The center field also shows mounted Persian hunters, although the figures are
less numerous. The round middle shield has similar small attached medallions above
and below. Instead of the dragon it shows phantastically formed panthers battling
with the phoenix or simurg, which in the corner pieces is represented alone. In the
border we see three nearly identical groups in which a noble Persian is served by
a retainer. The ground of the center field is of a magnificent dark green. It is
decorated with flowers, buds and leaves on slender vines, while the border shows
shrubs of more realistic form — blooming almond and magnolia trees with bright
birds in the branches under which the figures move. The Chinese "Tschi" is hidden
in the outer small border, where it is combined with flying cranes and very deli
cately conventionalized plant motives. The cranes, too, are found on other rugs of
the same group; this is one more proof that they are also borrowed from Chinese
legend and art.
In the Royal Palace in Stockholm is a third hunting rug in silk and gold, similar
to the one just mentioned. It has only a few animal motives on the red back
ground and the yellow border. A smaller one, with woolen pile, in the Maciet
collection in Paris (Fig. 4) shows the degeneration of this type.5 In the centre
field — quite as in the Indian group — balanced composition has entirely dis
appeared. The design of the animals and plants is coarser and shows much less
care; the conventionalization of the flowers and leaves has much less character.
As in the other rugs, the centre field shows Persian hunters; in the lower corners
the large phoenix appears. Dr. R. M. Riefstahl claims that these are five scenes of
the famous "Khamse" of Nizami, each scene being represented twice. In the border,
a round medallion with a crouching Persian alternates with a genie kneeling in a
large flower, and repeated between them is an antelope surprised by a panther. In
a woolen rug of the former Yerkes collection purchased at the sale by I. B. Trevor,8
the hunting motif becomes confined to the central medallion. The field shows ani
mal fights. The latter also recur in the rosaces in the deep blue border alternating
with figural medallions. A rug with human figures closely related to those mentioned
before was exhibited in the Petrograd Retrospective Exhibition of Decorative Art of
1904 as a loan of Prince Sanguschko. (Plate 18 in the publication of the Exhi
bition by Prachoff, 1907). This rug was said to have been captured in 1621
from a Turkish tent. The central motif shows the broad arabesque which
usually appears in the borders, combined with four medallions with human figures.
In the attached medallions appear musicians and genii; in the corner motives,
mounted hunters; in the field, fighting animals, and in the elaborate border, we find

The Persian animal rugs. 1?

peacocks in opposite reprensentation, standing figures, the dragon fighting with the
phoenix and other scenes with fighting animals.
The velvet-like woolen rug in the Poldi Pezzoli Museum in Milan (Fig. 5), is particu
larly valuable because of its material, its technical perfection and faultless condition down
to the very fringe. It has no hunting scene in the center field, but is otherwise
very closely related to the two famous hunting rugs. Besides its high artistic and
decorative value, it has a particular interest because of its unusual richness in
Chinese motives. There are no corner pieces in this rug and the round central
medallion, without any attached medallions, takes up comparatively little space. It
is decorated with birds among flower vines while the innermost star shows cloud
bands. Blooming trees stand symmetrically on the flowered, rose-colored ground
with birds and monkeys swinging in their branches, while tigers and dragons attack
each other at their roots.
The Chinese element is represented at either end of the rug by flying cranes
among cloud bands intermingled with a good deal of "Tschi". In the corners and
in the center we also find masses of this holy Chinese fungus of unusual form.
However, the most striking representation is found at either side above the central
medallion, taking the place and shape of the usual attached medallions. Two Persian
genii, a flower upon their heads instead of the usual crown, crouch before a vase
with fan-like, flower decorated ornaments, arched over by a phantastic canopy
decorated with flowers and birds. Despite the purely Persian form and dress of
the genii they are quite obviously the fanciful reproduction of a Chinese sacrificial
altar, while the flowers before the altar are a transformation of the Chinese globe-
pattern, the "Tschintamani".
The finial of this composition likewise proves that this interpretation is not too
daring. It is in the shape of a canopy and copies the bat symbol, although in a
distorted and phantastic form. The decoration of this canopy — two ducks between
clouds and Tschi motives — is another proof of Far Eastern influence. The border
shows large palmetto flowers connected by ornate undulated ribbons filled in with
animal motives.
The very subjects represented make it probable that such magnificent rugs were made
in silk and silver or in woolen pile for a Persian noble, perhaps indeed, for the Shah
himself. A number of highly skilled workers would need several decades for the
production of such a rug, which even at that time cost many thousands. We have
not heard of other rugs so large and costly as these, but perhaps in time one
piece or another may be discovered in the harem of sultans and other rulers and
nobles of the Near East. "Animal rugs" in the true sense of the word, have been
preserved in comparatively large numbers. They are of the same technique and with
similar decoration, save that they are almost entirely without human figures. But
there are more of them in the palaces of princes and wealthy men of the Occident
and in several museums, than in the Orient itself. They are either in silk,
some with interwoven silver threads, or in wool. The latter are usually of large size.
A silk animal rug of smaller measurements, but remarkable for its faultless con
dition, was lent by Prince Lobanow to the Vienna Rug Exhibition, where, with the
imperial "hunting rug", it was greatly admired. It now belongs to the Stieglitz
Museum in Petrograd (Figs. 6 and 7). Its discovery in one of the palaces of the
sultan is of particular interest. Composition and decoration are especially reminiscent
of the magnificent title pages and sumptuous bindings of Persian manuscripts. It
would hence seem very probable that the illuminators made the designs for the
court rugs. In other details it resembles the two large hunting rugs, save that
Bode-KQhnel. 2

18 Bode-Kuhnel, Antique Rugs from the near east.

there are no human figures, which are always rare in Near Eastern rugs. Wild
animals among delicate flower vines occupy the center field of this rug which
shows dragons, panthers, tigers, fallow deer and jackals looking at and pursuing
each other. The centermost star medallion, which is round with a wavy edge,
is entirely filled with plant decoration. The corners show birds among flower
vines-parrots, pheasants, and cranes seized by falcons. The border bears large
elongated cartouches with decorative inscriptions in Persian Talik characters, which
alternate with eight-lobed rosaces showing floral decoration. Both of these are
connected by small circles enclosing an animal. All ornaments and the space bet
ween them are decorated with scattered floral sprays which also adorn the small
border. The animals in this rug are particularly realistic and well designed, and
the arrangement and drawing of plants are very refined. As in most of these rugs,
the large inscriptions in the border contain only quotations from Persian poets.
A new feature, however, is a beautiful leaf motif, similiar in design to an out-
curved halberd which is usually designated as the "arabesque" because of its long
and frequent appearance in "Arabian" art. It dominates the pattern of the central
medallion as well as that of the round medallions in the border of the Lobanow
rug. Such arabesques sometimes occur in the hunting rug, but they are so sub
ordinated and small that they are scarcely noticed. The "cloud band" which winds
symmetrically among the flower vines in the center field and the border seems to
be a new motif but it is only the Chinese Tschi in ribbon-like development and
delicate Persian conventionalization. In just this form it became a particularly popular
and characteristic element of Persian decoration at the Sefevi period.
Several dozen silk rugs of this kind have gradually become known. According
to the character of Oriental art, they all show individual differences in conception
and decoration. The majority are in the different Rothschild houses — in those of
Baron Maurice and Baron Gustave Rothschild in Paris, in that of Baron Alfred in
Vienna, etc. The museums of decorative art in Paris, London, Berlin and Petro-
grad also own several examples. Others belonged to the art dealer Bardini in
Florence, the late Mr. Morrison, and to the painter Goupil, whose rugs were sold
at auction in Paris in 1888 with his valuable collection of antique Oriental art
One among them, now in the Musee des Arts Decoratif in Paris (Fig. 8), is similar
to the Lobanow rug, except for the more strongly conventionalized design, especially
in the center medallion. Besides the dragons which are picturesquely used to form
the outline of the corner motives, we find the mythical Chinese Kylin in its deer
like form. This Kylin motif is used alone as well as in combat with a lion-like
monster, which seems to be the Kylin taken over from Korea to China. The car
touches with inscriptions are also in Persian cursive characters, but they are not of
the monumental type found in the Petersburg example. The same pattern, identical
in all details, but knotted in wool, reappears in a rug in the former Yerkes collection
and with slight differences in another rug in the Pannwitz collection.' A related
example in the Czartoryski Museum in Cracow shows a noticeable difference only,
in the border which is enlivened by animals, and the inscriptions are confined
to the smaller, inner border. The Far Eastern influence appears plainly in a
rug of particulary beautiful coloring and design which has been given to the
Berlin Museum of Decorative Arts by the author. (Fig. 9.) The center field
shows fighting animals upon a flowered ground — lions, tigers and panthers tearing
ibexes and antilopes to pieces, and also the lion throttling the deer Kylin. In the
central medallion the motif of the dragon in battle with the phoenix is repeated
four times, elegantly grouped around fully opened flowers, while in each of the

The Persian animal rugs. 19

corner pieces we find three birds upon flower bushes. In the deep green border
each large palmetto flower is flanked by a pair of golden pheasants which pick the
berries from the vines on which they sway. In this movement they form a silhouette
which in other rugs is conventionalized into an interesting undulated ribbon pattern
(Compare Fig, 5). The inner small border shows the cloud band alternating with an
open flower. This rug is entirely of silk. The corner motives, too, the ground of.
which in this type of rug is usually interwoven with silver and gold threads, are
in this case of golden yellow silk. In refined composition, in the bold and vigorous
design of the animals and flowers, as well as in the strength and harmony of its
rich colors, this rug is one of the most beautiful which has been preserved.8
Another silk rug, (Fig. 10) in the Musee des Arts Decoratifs in Paris, originally
came from the Goupil collection. It is closely related to the Lobanow rug except
for the large oval central medaillon which is in the shape of a cross with an orna
mental center motif and a peacock in each of the four arms. This central motif is
surrounded by a composition of single branches or shrubs laden with flowers or
fruit. The cartouches of the border are filled with Talik writing. An animal rug
belonging to Mr. Baker in New York is similarly arranged, but much more elaborate9
We must also mention woolen pile rug in the collection of the Chamber of Commerce
of Lyons with some interwoven silver threads which has almost the same, although
not quite so refined a border as the Poldi Pezzoli rug in Milan. Two ducks between
"Tschi" motives form the center of the beautifully designed central medallion around
which sitting figures and serving genii are grouped. Smaller medallions of bat-like
form are attached. The same motif is repeated with slight variations in two beau
tiful pieces of smaller size belonging to the Metropolitan Museum In New York,
one of these rugs being from the Yerkes Collection, and a gift of Mr. J. Fletcher.10
Both show a small central medallion with ducks, surrounded by a company ban-
qetting. All kinds of animals — most of them taken from Chinese mythology — run
around in the center field among flowers. An analogous example is also in the
Textile Museum at Lyons.
Some of the woolen rugs show silk rug patterns executed in a less costly
material; others have developed a style of their own. Their design is more vigorous,
more angular, and sometimes coarser. This is not only due to the coarser knotting,
but also because they often belong to a somewhat earlier epoch of Sefevi art, for
which severer design, more vigorous conventionalization and more energetic color
composition are characteristic. If the silk rugs are gorgeous and alluring, the woolen
ones are imposing and dignified. Examples in this material are just as scarce as
the silk rugs, perhaps because they have been less valued or less carefully preserved.
The gift of such a rug by Shah Tahmasp I. to the funeral mosque of his dynasty
at Ardebil in 1539 proves that they were highly esteemed in their time. One of
these rugs with a border very similar to that of Lobanow rug, now belongs to the
collection of the Manufacture des Gobelins in Paris. In both of the smaller medallions
attached to the central medallion two peacocks stand in opposite representation. Two
other large rugs which are practically identical — one in the possession of Count
Boucquoy in Vienna (Fig. 11), the other in the Museums of Decorative Art in Berlin,
are similar, but somewhat severer in design. The Berlin rug has a large rose-
colored central medallion with similar design and the border is divided almost
identically into rectangular cartouches and four-lobed quadrangles, which are filled
solely with flower and leaf decoration. The white center field shows isolated, roughly
designed animals, among them the Kylin. Although there are no corner motives, the
center field is considerably reduced by the two smaller medallions attached to both

20 Bode-KuhneL Antique Rugs from the near east.

ends of the central medallion — the outer one heart-shaped, the inner one rectangular.
The vigorous pattern of this rug reveals a strong feeling for style; the coloring is
rich and harmonious. The same holds true for the above-mentioned rug belonging
to Count Boucquoy and for another rug which appeared in the Paris art trade in 1 892.
We do not know the present location of this rug, which was the most beautiful
of them all.
Another group, closely related within itself, differs even more from the silk
rugs. Examples are found in the Kaiser Friedrich Museum in Berlin, in the Cathedral
of Cracow, in the Collection of Prince Adolf Schwarzenberg, in the possession of
Mr. Williams of Norristown, Pa. and of Mr. Bardini in Florence. The Berlin rug
(Fig. 12) was formerly in an old synagogue in Genoa.11 The corner motives of
the center field contained human figures. Such representations are forbidden by the
Jewish commandment and therefore a strip of the center field, some twenty inches
wide, was cut away on both sides next to the border which was then sewed loosely
to the rug. A mate to this rug is said to have been temporarily in Windsor Castle.
We do not know whether this is the same rug wich Dr. R. M. Riefstahl has esta
blished as belonging to Clarence H. Mackay in New York, and which was formerly
in the collection of Mr. M. Perry in Providence, R. I. Considering the dimensions
of the rug, the pattern is very rich and of large proportions. The arrangement
and design have a consummate feeling for balance and fine convention-alization, while
the rich coloring (white, black, yellow, rose, cinnabar red, salmon pink, cerulean
blue, deep blue, grayish blue and bluish green) is as vigorous as it is harmonious.
The conception of the center field as a wood with trees, bushes and flowers springing
from a white ground and enlivened with all kinds of animals, is characteristic of these
rugs, and differentiates them from those previously discussed. In the hunting rug
belonging to the Austrian State, the salmon red ground is similarly decorated with
blooming plants. However, these merely produce an effect similar to the flowery
background in the rugs mentioned before, because of their appearance between the
large figures of the hunters and animals. But in the rugs we are now considering,
the majestic trees and bushes with their luxuriance of flowers, leaves and fruit are
the most important feature. The arrangement of the central medallion with its small
attached medallions is also characteristic. These are brought into strong relief both
by their coloring, and by their ornate contour and rich decoration of animals and
flowers. The same holds true for the corner motives.
The details of the design are of particular interest. In the white center field
we find all the mythological Chinese animals which we have seen here and there
in the other animal rugs. Above hovers the phoenix with the lion-kylin snarling at
it; the deer-kylin below looks fearfully at the lion. At the side, at the foot of two
high cypresses, we find the dragon. Toward the center of the rug only Near
Eastern animals appear: panthers, bulls, fallow deer, ibex, jackals, hares, and dogs,
while upon the trees are monkeys and song birds. The animals are not fighting!
but are represented as grazing or looking at each other and growling. Despite their
conventionalization, most of the trees and shrubs upon or under which the animals
move, may be recognized as cypresses, plane, almond, medlar trees, etc.
The design of the central medallion and that of the corner motives is even
more original. The former is entirely filled with walking, fluttering, and flying cranes
between phantastically knotted ribbon motives in various forms. These knottings
are easily recognizable as the Chinese fungus, the Tschi. Their ribbon-like, connected
design and their twisted form prove that — as in China — they are very obviously
conceived as the symbolical representation of clouds. They are sometimes massed

The Persian animal rugs. 21

together, sometimes drawn out very thin with looped, ribbon-like attachments similar
to older Chinese representations. A different motif repeated beneath the standing
cranes, which consists of a smaller, slightly twisted band with zigzagged pro
tuberances may be considered as the lightning motif. The Far Eastern origin
of1 all these motives is established beyond doubt by a comparison with purely
Chinese works of art. Both the crane and the Tschi, are Chinese symbols of
immortality. Similarly, deer reclining between cranes are used as a symbol of
longevity. The representation which was repeated in the four corners was also
purely Chinese in its inspiration. This is shown in the New York specimen where
it is completely preserved. We can recognize a figure in a long Chinese coat and
Chinese shoes with high soles. This figure stands before another crouching figure
who is offering a goblet. The space between the figures is also filled with peculiar
ribbon motives of more intertwined and angular form than those in the central
medallion. The knot-like Tschi motif, too, is either unrecognizable or else very
indistinct. It is possible that this motif must also be regarded as a representation
of Chinese clouds. The three globes enclosed by the cloud band beneath the
standing figure are particulary worthy of note. Their appearance as the main
decorative motif in several rugs of widely differing types proves their particular
significance (Cf. Fig. 73). In these rugs the globes form triangles like those in the
large Berlin animal rug. The use of this symbol in connection with Chinese figures
again offers most trustworthy proof that the entire composition is of Chinese inspi
ration. Lastly, this theory is supported by the parallel, slightly twisted ribbon
motives, the frequent appearance of which in rugs of widely divergent types may,
according to the above, be explained as a Far Eastern tradition. These motives
may be interpreted as clouds, waves or lightning (fire). In such an arrangement,
the globes would represent the Tschintamani, the holy symbol of the teachings of
Buddha, repeated so frequently and so diversely in Far Eastern Art.
Another, but obviously slightly'later rug, with smaller medallions and without figural
corner motives, is related to the piece under discussion. This rug, as well as the one
reproduced in Fig. 4, was formerly in the Maciet Collection before its bequest tothe Museum
of Decorative Art in Paris. A well preserved example of the same type (Fig. 13) which
belongs to Prince Adolph Schwarzenberg, was brought before the public in the Vienna
and Munich Exhibitions. As this rug is of somewhat smaller dimensions, the central
medallion and its attached medallions are considerably reduced and the corner motives
have disappeared. Cypresses, plane, and fruit trees may be identified in the thicket of
blooming trees and shrubs in the center field. Only a few animals appear among
these trees — lions, panthers and birds — whereas the phoenix holding a little
bird in its beak is used as a corner motif. The smaller inner field of the central
medallion shows four pairs of ducks upon waves (or clouds). Around it are large,
ribbon-like arabesques upon a ground of small flowers on elaborate vines. In the
small, rectangular medallion attached to either end of the central medallion, two
peacocks are represented in the same way as in several rugs previously discussed.
The border design is very unusual. An angular, undulated ribbon motif, winding
reciprocally through the entire width of the border, divides it into fields with alter
nating white and red grounds decorated with vines bearing blossoms and leaves
with bird motives in between. Cloud bands of slender, exceedingly pure Chinese
form "run through the entire border. The Tschi motif is not used elsewhere in this
rug except in the corners of the center field next to the phoenix. The full flowers
of the border show in their centers smali masks of lions and panthers — the latter
in profile. In the Berlin rug such masks are found much less frequently: — that is,

22 Bode-Kuhnel, Antique Rugs from the near est.

only in the central medallion and its smaller attached medallions. The animal rug
belonging to Mr. Williams has a border of somewhat the same type. The deco
ration of its beautifully toned green ground is similar to that in the Cracow rug.
The latter, however, shows a yellow center field with a red border, and its design
is an intermediate phase between the two which we have already described in detail.
The Victoria and Albert Museum in London owns by far the most magnificent
and stylistically perfect rug of this type (Figs. 14—17). This rug, strange to say,
has no large central medallion. In its place there is only a very small round white
medallion with fish motives; both halves of the rug show a beautiful and energetic
pattern. In the middle of this pattern we find a large, circular medallion with
slightly curved outline bearing thin plant decorations upon a black ground. Above
and below are large Chinese flower vases supported by three lion pillars and held
by two dragons. The bodies of the vases are decorated with a pair of geese in
opposite representation. Four smaller oval medallions, each with a pair of ducks
flying among flowers and cloud bands upon a black ground, are attached diagonally
to the four sides of the central medallion. The reddish brown ground of the center
field is decorated with a few trees — blooming almond and fruit laden pomegranates.
Wild animals and mythical- creatures move among them, while birds sway in the
branches. The wide border (Fig. 17) also shows the undulated ribbon motif. This
causes the side with the black ground to be a kind of matrix for the other side with
the reddish brown ground. Dragons and Fohos on the black ground , t spitting at
each other, alternate with lions seizing bullocks or antelopes on the red ground. All
these figures appear among vines. They are markedly Chinese in character. The
narrow outer border has a yellow ground with masks of various animals among
vines. The rich and vigorous color harmony with its predominating note of black
is thoroughly in keeping with the powerful and severe style of the design and the
interpretation of the plants and animals. This rug is without doubt one of the
earliest works of the Sefevi period. The rugs of this era are to later silk rugs,
what the art of the Quattrocento is to that of the Cinquecento in Italy.
Several other rare woolen rugs of large dimensions form an essentially different
group. The center field of these rugs does not show the large compositional units
found in the examples previously discussed. Instead, it is entirely divided into small
recurring fields or cartouches in which ornate animal motives alternate with floral
sprays and blossoms. A particulary valuable specimen of this type, bought by
Mr. Yerkes of New York from Adolf Thiem, is now in the collection of the Metro
politan Museum12 (Fig. 18). Dr. R. M. Riefstahl points out a mate of this rug in
the Textile Museum in Lyons. This rug shows an all over pattern of cartouches on a
cream white ground. Large round cartouches with scalloped outlines bear represen
tations of the fight between the dragon and the phoenix upon a light blue field.
Other cartouches with a black ground show lions among vines. Small, heart-shaped
cartouches show upward flying cranes among flower vines or curved vine orna
mentation with blossoms, on blue or rose-colored grounds. The beautiful main
border with its rose-colored ground has long cartouches in pale blue alternating
with small black rosaces. The latter again show the symbol of the Ming dy
nasty18 — the phoenix fighting with the dragon, while the former are charmingly
decorated with arabesques and cloud bands on flower vines. The same type of
design characterizes a figural rug in the collection of the Duke of Buccleugh (Fig. 19).
Although the decoration is slightly more conventionalized, it is incomparably richer
and in composition reminds one of the repeat designs of the contemporaneous tex
tiles. Falcon hunters, musicians and revelers are represented in oval fields; in smaller

The Persian animal rugs. 23

medallions standing and flying birds appear. The ground is filled with quadrupeds:
antilopes, lions, panthers and foxes, both in trees and among vines. In the border
the Chinese motif of the fight between the dragon and the phoenix which predo
minates in the New York piece, alternates with six-lobed interlacings of two vari
colored dragons. Between these are scattered all kinds of animals among vines.
The outer border again shows the animal mask motif.
Another rttg of very characteristic and beautiful type similar in composition,
but entirely different in details was loaned by Countess Clam-Gallas to the Vienna
and Munich Exhibition (Fig. 20). The small units or cartouches of the center field
show varied decoration. The animal motives are limited to a pair of peacocks or
pheasants in opposite representation which alternate with arabesques, cloud bands
and palmetto rosaces. In the other cartouches are brandies and shrubs bearing
blossoms and fruit or other motives in correspondingly light and delicate colors. This
breaking up of the center field into rows of regulary formed cartouches goes perhaps
back to an earlier period of rug-making. This type of composition — as well as that
with corner motives, central medallion and attached side medallions — can be compared
with the art of the book decorator. It is particularly reminiscent of the gold tooled
bindings of the 16th century, with their characteristic frame of cartouche motives.
There is another group of rugs with particularly close knotting. In spite of
minor dissimilarities, they, too, are clearly from the Sefevi period. Their center
field, which has no division into medallion or cartouches, is decorated with single
animals or groups of animals. Sometimes they move among conventionalized flower
vines; in other cases the center field is conceived as a picture showing a meadow
or wood with all kinds of varicolored animals. Professor Sarre of Berlin owns a
very fine, remarkably well preserved rug of the 16th century which belongs to the
first type (formerly in the collection of A. Thiem). A mate of this piece, formerly
in the Yerkes collection, has been secured by the Metropolitan Museum in New
York (Fig. 21). The fragment of an exceedingly beautiful rug of this same type, whose
field and border show analogies to many of the specimens discussed before, is in
the Austrian Museum (Fig. 22). Another is in the North Bohemian Museum of
Decorative Art in Reichenberg, while a third was exhibited in Munich by Mr. Bohler.
In this chapter we have classified the various types of animal rugs according
to their typical variations, without discussing exhaustively the examples preserved.
In spite of their manifold variations, many similarities make it probable that they
originated in the same period and in localities not far from one another. They all
belong to Eastern Mohammadan art. This is shown by their composition and by
the use of motives which are either familiar to the art of Eastern Islam or have
been transformed in a way characteristic of this art. Among such motives the
arabesque and the fully opened flower motif are especially popular. This latter
motif is sometimes termed palmetto, sometimes pomegranate, although it was primi
tively intended to be a lotus flower. The Persian origin of these rugs is proved
by the use of human figures, by their costumes and occupations and the kinds of
animals represented. But the most conclusive proof lies in the artistic character of
the entire decoration which is closely related to that of Persian miniature painting,
leather work, textiles, decorative tiling, and other kinds of applied art.
As these works, and the miniature paintings, in particular, can be dated much
more easily than the rugs, they are a valuable clue for dating the group now under
discussion. Rugs of this type belong to the dynasty of the Sefevi (1502—1736),
who established their power after a number of severe struggles and then brought
Persian art to a high state of development Inscriptions on rugs in the Persian

24 Bode-KQhnel, Antique Rugs from the near east.

characters of this period confirm this date, while the Far Eastern motives in the decoration
are a final corroboration. These Chinese influences are not due to any lack of
native creative power, nor is their use entirely to be explained by the practice of
borrowing and working over all kinds of foreign motives, which is a general
characteristic of Persian, as well as of Mohammadan art. It is better to interpret
them as the result of a consciously planned artistic policy of the dynasty, particularly
of Shah Abbas the Great. This dynasty highly esteemed the flourishing Chinese
art of the day, thus continuing the traditions of the previous Mongolian period. For
this reason we find the same Far Eastern motives in other branches of Persian deco
rative art from the end of the 15th until far into the 17th century. But in no
other form of art do they appear so numerously, so frequently and so purely as in
the rugs, not only in those of Persia, but also in those of Asia Minor and Armenia.
This is especially true of the "cloud band" which appears in all Near Eastern rugs,
with or without the fungus motif, the Tschi. The representation of various Mon
golian animals, especially the Kylin, the phoenix and the dragon are also quite
widespread through Persia. Furthermore, one occasionally finds the Buddhist globes
(the Tschintamani), the bat ornament and a motif representing either lightning or clouds.
These general points of view may help us to understand the hunting and animal
rugs, but as yet we cannot make a definite assignment of these rugs, to definite
provinces and manufactories, or to an earlier or later phase of their evolution. The
places which were especially favored as royal residences in this period, first Herat
Tabriz, later Ispahan, may well have been of importance.

Persian Rugs with floral Patterns.
In the animal rugs of the 16 th century the center field almost always shows the
same large units of composition — a central star medallion with one or two
smaUer. medallions attached at either end and corner motives which often take the
form of a quarter section of the central star medallion. The entire rug is covered
with vines bearing blossoms and leaves, among which animals and figures move
as if upon a blooming meadow. Leaving out the animals, a delicately interlaced
decoration of blossoms, buds and leaves on slender stems remains. This decoration
is more or less regularly set into the main pattern which is always of a broad
conception. A considerable number of closely related medallion rugs which may
be divided into two groups show very similar decoration. These are without
representations of living creatures, or show only very few and small animals in
an entirely subordinated form. The older group, which corresponds with the animal
rugs in composition and design, have vines with blossoms, buds and leaves as
their sole motif and shows them in severe conventionalization. In masterly sense
of composition, in purity of design, in tasteful and rhythmical development of the
main pattern, in the stylistically perfect form of the blossoms and leaves, and in
richness and delicacy of 'coloring, these rugs are almost superior so those with animal
representations. Their material is less costly, as they are almost always knotted in wool.
Their dimensions are usually rather large, to correspond with the large proportions
of the elaborate pattern. Pieces characterized by inscriptions donot seem to have been
found among the examples preserved. Nor do old paintings give us any evidence as to date.
Sufficient information is gained, however, by comparing these rugs with the
datable rugs, both those with animal representations and those of related types.
Several characteristics prove that they were made in a period previous to the greater
number of the animal rugs — the beginning of the 16th and the latter part of the
15 th century. These characteristics are the vigorous conventionalization of com
position and decoration, the regular appearance of numerous arabesque motives and
the predilection for colors characteristic of the earlier rugs — we refer particularly to
the white and yellow backgrounds. The striking similarity to the famous rugs of
the Sefevi period, especially in the border, offers still further assurance that these
"flower" rugs originated in Persia. The occasional and very inconspicuous use of
isolated animal motives further supports this view.
A very beautiful early rug in the Berlin Museum of Decorative Art is of this
variety. The same collection also possesses another good example with a more cur
rent pattern. A smaller, particularly well proportioned medallion rug is in the Cologne
Museum of Decorative Art. The Kaiser Friedrich Museum in Berlin exhibits two
others. Several beautiful examples of very large size which the antiquarian Bardini
secured from Italian churches are now in the collections of Baron von Tucher and
Mr. Yerkes of New York, now in the collection of I. Seligmann (Fig. 23).14 The
author knew of another such rug in the church of San Salvadore in Venice, the most
magnificent of this type. Unfortunately the border has been cut away at either end.
Another example, of which only the inner border has been preserved, is in the col
lection of the Manufacture des Gobelins in Paris (Fig. 24). Its most important feature
is the very beautifully designed large central star medallion upon a red ground.
Small birds are scattered among the vines in the corner motives, but these do not
detract from the decorative unity of the composition.15 This type is also represented

26 Bode-Kuhnel, Antique Rugs from the near east. 
in the Metropolitan Museum (Altman Collection). Dr. R. M. Riefstahl also calls our
attention to an incomplete piece in the collection of Mr. C. F. Williams, Norristown,
Pa., which on a background of vines shows two riders in sketdiy design, one of
them seemingly the captive of the other. An example owned by a dealer (Fig. 25)
is particularly important because of its obvious relation to the Ushak group, in which
the designs of these Persian prototypes are interpreted in a very original way.
(Compare page 93) This specimen is of conventionalized, angular design and of
strong coloring. It has a large round central medallion and a severe border of
broad arabesques which on both ends differs in its solution of the corner design.
In still another rug, sold in the Zander auction in Amsterdam in 1914, the transition
to the Uskak type is still more obvious, especially in the breaking up of the border
design; the corner motives also are lacking.
Another example in the Victoria and Albert Museum, the famous Ardebil rug,
is slightly different from the rugs described above, but has points of similarity with
the "cartouche" rugs belonging to the Metropolitan Museum and to the Countess Clani-
Gallas (Figs. 18 and 20). This is perhaps the largest rug which has been preserved
from the earlier period; it measures 38X17 feet (11,52, 5,34 m; Figs. 26 and 27).
The dark blue center field is entirely covered with small blossoms and conventionr
alized peony like flowers of various sizes and shapes upon delicate vines. The
middle bears a large star medallion decorated with flower vines, arabesques and
cloud bands upon a yellow ground. To the pointed projections of its ornate edge
are attached sixteen small ovals of diverse forms resembling Chinese lanterns. A
large mosque lamp is suspended from the two ovals in the longitudinal axis of the
rug. The corners are decorated to correspond with the central medallion. The ex
traordinary richness of this design, whidi is executed with remarkable precision, does
not impress us as confused and over decorated, because of the large surface of
the rug and the beauty and harmony of its colors. The main border, framed by
small borders of elegant design, shows alternating round and rectangular medallions
with flower vines and clouds bands upon a black ground. Up to the present day
this rug is unique among all Persian rugs on account of the date which is placed on
a small white shield at the end of the center field, close to the border. The rug was
woven in 1539 A.D. This specimen is also of great scientific value because, as an ex
ception to the general rule we have information regarding its origin. It was discovered
some twenty years ago in the funeral mosque of the Sefevi in Ardebil (Western Persia),
for which it was ordered by the Shah Tahmasp16. Its manufacture in Kashan is
attested by the inscription, which at the same time offers one more proof that all
classes of rugs previously discussed originated in Persia. On the other hand, it
also gives the most trustworthy support for dating all of these rugs in the 16th
century. According to Dr.- R. M. Riefstahl, an almost identical piece, but much cut
down, was bought by Duveen Brothers in the sale of the Yerkes Collection.
Another rug of large dimensions and similar design, which lay for a long time in a
mosque in Constantinople, is reproduced by F. R. Martin (Plate 2). This rug forms
in some respects a transition from the type just described to the less pretentious
rugs of this group.
The omission of all figural representation was a self understood requirement in
all rugs intended for a sanctuary. But there are also a number of palace rugs- in
which the animal motif is entirely, or practically entirely suppressed. However, in
their magnificent material, delicacy of design and excellent knotting, these rugs' are
in no way inferior to the magnificent animal rugs from the Sefevi period and were
probably made in the same manufactories.

Persian Rugs with floral Patterns. 27
This is especially true of a few silk rugs, one of which is in the collection of
the Manufacture des Gobelins in Paris (Fig. 28). A mate of this rug belongs to the
Altman collection in New York (No. 86) now at the Metropolitan Museum); a
third is owned by Baron Albert Rothschild in Vienna (Fig. 29). A fourth of
equal beauty, was bequeathed by George Salting to the Victoria and Albert
Museum. The Paris specimen is very closely related to the animal rug in the
Berlin Museum of Decorative Art (Cf. Fig. 9) but in the Paris rug the fine compo
sition and the exquisite design of the fully opened flowers in border and center
field appear more strikingly. The beautifully undulating arabesques in the central
medallion and the corner motives are a further distinction of this rug, for in almost
all other silk rugs such arabesques are entirely subordinated to the animal forms or
completely lacking. The only Far Eastern motif is the Tschi. It appears among the
floral decoration of the center field in purely Chinese form as snail-like balls, either
with projections in the shape of lightning or surrounded by the cloud band. The
Tschi is similarly used in a silk rug with a beautiful deep red ground which was
in the Bardini auction in London. The Rothschild rug which is especially distinguished
by its depth of color (cherry red predominating) and by the beautiful conventionalization
of the flowers, shows a large number of inscriptions which are very skilfully sub
ordinated and concealed. In other instances, the delight of the Orientals in convention
alized inscriptions occasionally goes so far that the entire surface of the rug is
almost covered with them. This is even true of silk rugs from the classical period.
The painter Goupil in Paris owned a very characteristic rug of this type.
As in several animal rugs, the division of the center field into a central medallion
with smaller attached medallions and corner motives shows the first signs of deca
dence in the group which we have just discussed. The corner motives are often
entirely missing and so are the attached medallions. The central medallion itself is
often very small and is not set off by the contrast of its color with that of the
center field. (Compare two otherwise remarkable specimens, in the Musee des
Arts Decoratifs.) It was now only a step to the dissolution of the entire com
position of the center field. This was then confined to vines with beautiful pal
metto-like flowers of different shapes and sizes and to small star flowers and
leaves. Only the cloud band undulates symmetrically, but inconspicuously among
the flower vines. In several rugs, most of them early examples, a few scarcely
visible birds sway in the vines. The border, too, shows the same type of deco
ration — large and small peony-like palmetto flowers, usually attached in alternate
arrangement to undulating, delicate vines with blossoms and leaves. Very rarely
these rugs show the type of border with alternating long cartouches and small
rosaces. When these do occur they have the customary decoration of graceful
flower vines. One early, particulary fine example was in the Bardini sale in London
in 1898. It was previously in the collection of Allessandro Castellani who secured
the rug in Nuremberg. The colors in this type of rug afe as arbitrary as the deco
ration. The center field usually has a deep red, the border a dark green or a dark
blue ground. Usually such rugs are of medium or large size. They are always of
wool, but knotted with great care and very closely. Several smaller pieces are of
silk interwoven with gold and silver. In the numerous existent examples of this
type one can clearly follow the development from the severe, stylistically perfect
forms still related to the animal rugs of the Sefevi with their deep, harmonious
coloring and close, finished knotting, to those with more careless design, brighter
or inharmonious coloring and rougher, looser knotting. Because of their relationship
to later examples dealers have adopted the designation "Ispahan rugs" for this entire

28 Bode Kuhnel, Antique Rugs from the near east.

class, although they frequently appear in literature as "Herat rugs", and are then
classed among the Eastern Khorassan rugs.
Their date is not alone to be determined by their relation to the animal rugs,
which are often more severe and classical, and hence usually from an earlier period.
The numerous paintings in which such rugs appear establish their date still more
reliably. Although as far as we know, Italian paintings show no such rugs, Rubens
and Van Dyck owned various examples. In the work of the former we may cite
as proof his series in the Louvre depicting the life of Maria de Medici, and in that
of the latter the portrait of the Children of Charles I in Windsor Castle and Dresden as
well as many other paintings. Such rugs are also found in the works of Spanish artists
of the 17 th century — for example, in a painting by Moy a in the Pinakothek in
Munich. In Dutch paintings these rugs appear very often. They are most frequent
in genre paintings — in works of Codde, Terbordi, Metsu, Netscher, Slingelarit,
Vermeer, P. de Hooch, Eglon van der Neer, and Frans Mieris, Troost and Quink-
hardt From the evidence of these paintings we may date the manufacture of this
group of rugs from the rend of the 16 th to the beginning of the 18th century.
Besides their close relationship to several animal rugs from the end of the 16th
century, the frequent appearance of such rugs in Portugal and Holland gives evi
dence that they originated in Persia. At this time both Portugal and Holland had
the closest trade relations with Persia. Holland gradually defeated her older rival.
Both in Lisbon and Amsterdam there were settlements of Persians — chiefly Ar
menian tradespeople, among whose imported goods rugs are expressly mentioned.
In Persia, too, Herat rugs still appear in trade, particularly those of poorer quality.
An interesting collection of this type of rugs, almost entirely from Spain and
Portugal, belongs to Baron H. von Tucher in Vienna and Nuremberg (Fig. 30).
Another famous collection is that of Senator Clark in New York (Fig. 32). Both
collections show this group from its origin up to late times. Several valuable
pieces are in private collections in Austria (Vienna Exhibition, 1891). Others are in
the possession of the Duke of Buccleugh Mr. H. von Dirksen in Berlin (Fig. 31)
and in the collections of several Paris connoisseurs and in the art trade as well. In
the museums (Vienna, Lyons, Paris, Berlin, Leipzig etc) such rugs are also quite
common, but of varying importance. A very lovely example in the Victoria and Albert
Museum (Fig. 33) shows arabesques in the field and pheasants in the border.
The few prayer rugs preserved from the prime of the Persian knotting art are
best classified with the Herat type17- Their design is naturally based on a more
or less conventionalized prayer niche which in the mosques of the faithful points
the direction to Mecca. J. Bohler exhibited a rug of this type in Munich in 1910,
the coloring of which was very similar to that of the Herat group. This rug, which
bears many religious inscriptions, is chiefly in tones of red and green. A rug of
similar design, the epigraphic parts of which are confined in an unusual way to
the upper part of the rug— the arch of the niche, the border and the edges— was in
the Yerkes collection. Two good examples, one from the Bardini, the other from the
Fletcher Collection, were secured by the Metropolitan Museum17 and the most beautiful
prayer rug known is probably that in the Musee des Tissus in Lyons. It has silver threads,
an olive field, red corner motives and a bright border with black inscriptions. Two simpler
rugs with bright coloring were lent to the Munich exhibition by L.Bernheimer18. One of
these pieces is now in the James F. Ballard collection, St. Louis. Of course, all of these
rugs are decorated with floral forms without any use of living representations.
The so-called "vase rugs" obviously come from an entirely different region of
Persia, although they also belong to the Sefevi period. In many respects they form

Persian Rugs with floral Patterns. 29
the link between the types previously discussed and the so-called Armenian group
which we will study later. These rugs most frequently show a pattern dating from
about the middle or the second half of the 16th century, which is particularly effec
tive because of its conventionalized flowers and rich colors. They do not appear
in paintings, but a few years ago they could still be found in Italian churches and
in the bazaars of the East. Delicate vines bearing large fully opened flowers of
both star and peony-like form run diagonally through the field. These vines either
enclose vases with a small boquet of flowers or large single flowers. Among the
latter there is a beautifully conventionalized lily motif, somewhat fuller, but other
wise very similar to that in the well known coat of arms of Florence. Between
these large flowers rise slender stems bearing a luxuriance of delicate small flowers
of more realistic formation, among which we recognize asters, blue bells and other
flowers. Occasionally the entire plants, even with an indication of the roots, are
used instead of the flower branches. In the earliest rugs of this type we occasionally
find the Chinese Tschi motif in voluted snail form among the flowers. This is the
case in a beautiful fragment in the Austrian Museum (Fig. 34). As a rule, such
rugs have a narrow border. Occasionally , it is reduced to the size of an enlarged
"small" border. In other cases it is formed of large arabesques, like the borders
sometimes found in the medallion and animal rugs. These arabesques frame pal
metto chalices and are filled in with floral vines and small scattered motives. Be
sides the above-mentioned rug, there are several beautiful examples of this type,
of different sizes, some of them fragmentary, in the museums in London and Berlin,
in the collections of Baron von Tucher, Professor Sarre, Mr. Lamm in Naesby
(Sweden) and elsewhere, especially in England. An example in the Ottoman Museum
in Constantinople shows an allover pattern of lozenges framed by lanceolated leaves.
This rug is probably the most magnificent specimen of this type (Fig. 35). An inte
resting example in the Leipzig Museum of Decorative Art has a few animal figures
and is also remarkable for the powerful arabesques in the field. Some of the rarer,
smaller examples particulary fine in design and color were exhibited in the Vienna
rug exhibition of 1891 and in Munich in 1910 (Fig. 36). The diagonal vines are
often so slender and so inconspicuous in their color that the divisions entirely fade
away, and the entire design has the appearance of a richly varied and skillfully
divided flower pattern. As usual, the pattern is of a severely symmetrical design,
but never has a monotomous effect, because the different motives are generally so
skillfully varied in color that they seem to take on different contours.
A pattern of great interest, and thus far comparatively rare is more distantly related to
this group. Its decoration consists of large, loose floral sprays and small rectangles
with flowers or shrubs. These have a unique frame of canals with small ponds at
the points of intersection, enlivened with fish and ducks. These so-called "garden
rugs" are a textile interpretation of a formal garden with its flower beds, little fish
basins and streams. They are perhaps a vague reminiscence of a mythical rug of
colossal dimensions — the so-called "Spring of Khosrau" which was been described
by old Arabian writers and valued at many hundred thousands. It fell into the hands of
the Arabs when in the year 637 they plundered the palace of the Sassanids in
Ktesiphon. This historic "rug" of the early middle ages must not only have been
completely different from the above-mentioned class in material and composition,
but was undoubtedly of tapestry weave and not knotted. However, the recollection
of its beauty may have lead to the repetition of many features almost a thousand
years later — a remarkable evidence of the vitality of all Oriental tradition.
Among existent examples,, that of Dr. Figdor in Vienna (Fig. 37) is executed

30 Bode-Kuhnel, Antique Rugs from the near east. 
in excellent technique, an infrequent occurrence in rugs composed in small fields
and from such an early period. This rug is knotted in wool, but in such a way
that an effect of silk is produced. Gold and silver threads are interwoven in nume
rous places, especially in the flowers and animals. These gold threads are intro
duced in the same loose technique used in the so-called "Polish rugs" which we
will discuss later. Combined with the deep, gleaming color of the knotted ground,
the metal strands produce a brilliant and picturesque effect. All kinds of vari-colored
birds sway in the flower vines, while in two of the fields an undefinable quadruped
moves beneath the trees. The very narrow border, framed only by an unadorned
band on either side, shows slender vines with alternating open and closed flowers
between lancet leaves. A more elaborate type of this pattern appears in a garden
rug coming from Constantinople and formerly in the collection of the antiquarian
Wagner in Berlin (illustrated in Martin's book). This rug is now in the United States
It is divided in four large canals which issue from a central reservoir. The remaining
space is filled by a large number of animals between trees.
The same plan may also be recognized in a later series which shows details
of less variety and a design already becoming stiff. Two analogous examples of
this type, one of bright, the other of deep tonality, are in the Kaiser Friedrich
Museum and in the collection of Mr. Lamm at Naesby, Sweden (Fig. 59). Other
examples belong to Professor Sarre in Berlin, Mr. Sidney Colvin in London, Theodore
Davis in Newport now loaned to the Metropolitan Museum, to the Frankfurt
Museum of Decorative Art etc.19. This revival of the garden pattern probably ori
ginated from 1700 to 1750 in northwestern Persia, where at the date mentioned
other kinds of classical rugs were also copied, as we will see in our discussion
of the Armenian rugs. The stiff conventionalization of these copies gives the
appearance of a primitive style, but it is in reality only a symptom of decadence.
A rug which obviously came from a Shiite sanctuary in Persia and which was ac
quired by the Mohammadan department of the Berlin Museum in 1920, may be cited
as a transitional specimen in the development of the garden rugs from the classical
to the decadent style. It was probably made in Armenia, as early as the 17 th century.
Although of considerable size, this specimen is incomplete; the prevailing colors are
red, blue and green (Fig. 38).
There are a few woolen rugs, probably from southern Persia, which show some
similarities to the garden type. These have varied trees with elaborate blossom or
leaf decoration in alternate arrangement upon the dark center field. The space be
tween is filled with palmetto flowers on vines, upon which appear the contrasting
figures of small birds. The border is small and shows a simple, angular vine
pattern with an alternating open and closed blossom. Such a rug of lesser dimen
sions and of rich harmonious coloring, probably from the end of the 16th century,
was secured by the Kaiser Friedrich Museum in Berlin (Fig. 39) from the Bardini
collection. The Victoria and Albert Museum possesses a rug which is similar save
for a cruder type of design and a different border pattern. A more imposing type
of tree rug has an undivided field filled with symmetrically arranged large cypresses,
fruit and flower trees, with palmetto flowers filling in the space between (Fig. 40).
The severe border of wide arabesques reveals that this rug originated in the early.
part of the 16th century, while details of the design show similaritties to a group
of the animal rugs as well as to several with vase patterns.

Persian, so-called Polish Rugs.
Another class of rugs is included among the most precious products of the
manufactories of the Sefevi Shahs. These are the so-called "Polish" rugs20.
They usually show ornamental motives in bright colors, and have a silk pile upon a
ground of silver or gold threads. Despite their costliness, a considerable number
of them have been preserved. The designation "Polish rugs" originated at the
Paris World's Exposition of 1878 where varied examples of this type were
exhibited by Prince Czartoryski. These are now in the Museum in Cracow.
Among them was a rug bearing the Czartoryski coat, of arms (Cf. also Fig. 41,
after a rug in the National Museum in Munich). Experts even thought that they
could decipher the letter "M" in the border, which they interpreted as the mark of
the Mazarski manufactory in Slucz, where the well known Polish brocade sashes
were woven in the 18th century. The manufacture of these sashes was started
with workmen imported from the Orient, while their pattern and technique followed
the example of the Persian silk weaves. But we have no proof that silk rugs, too,
were made in this factory. These rugs must have been made during the prime of
the rug knotting industry, as is shown by the comparatively large number of ex
istent examples — numbering at least several hundred — , by their costly materia.
and their technical perfection. The conditions necessary for such a high develop
ment of the rug knotting industry did not exist in Poland at this period. We can
furthermore prove that they date from the first half of the 17 th century or earlier,
which is a hundred years before Mazarski and his enterprise.
The Oriental origin of these rugs is fully assured by the entirely Persian
character of the decoration. The center field often has the well known central star
medallion and occasionally the corresponding corner motives. All the floral motives
in the center field appear in conventionalized form. These include the well known
arabesque, similar to an outcurved halberd, the fully opened peony palmetto, the
smaller peach blossom and other flowers on elaborate vines. Frequently the cloud
band in various forms winds among the floral motives. On the other hand, the
pattern is less sharply pronounced than in the animal rugs, and in the later spe
cimens it becomes somewhat vague and confused. The effect desired is merely that
of picturesque color, and if these rugs are still in good condition, their beauty is
extraordinary. The loose knotting, the high pile, the wide surfaces of tapestry
woven silver or gold threads which lie deeper than the pile: — all produce a rich
girandole of flashing light combined with delicate, light hues of color. Compared
with other Persian silk rugs the workmanship of these rugs is superficial and careless
and hence shortens their life. Therefore, among scores of examples, compara
tively few are in a perfect state of preservation. The border either shows alter
nating flowers, with or without arabesques, or an angular ornament which we may
call the "reciprocal lily" pattern. This pattern, which was found even in mediaeval

32 Bode-Kuhnel, Antique Rugs from the near east.

textiles, still appears in Oriental rugs of the 15 th and 16th centuries. The design
and conventionalization of the flowers and other ornaments prove that these rugs
were made in the 17 th century. Several with borders of especially fine and vigorous
design, such as the rug formerly in the possession of V. Robinson (Fig. 42) and
another in the collection of Prince Johann von Liechtenstein (Fig. 44) were probably
made about 1600. Others with a very much more decadent form of decoration
may have originated as late as the beginning of the 18 th century.
. The similarity of this type of rug to the authenticated examples of the Sefevi
period, is sufficient proof for their origin in one of the Persian court manufactories,
even though the majority of examples preserved have been found in Poland, Austria,
Hungary, Russia, Turkey and in other countries adjacent to Turkey. In these coun
tries they may frequently be traced in the hands of the same family up to the
18th and even into the 17th century. In Persia, on the other hand, such rugs do
not seem to appear. At all events, they have not been brought from Persia to
Europe in the channels of regular trade. At the Vienna Exhibition dozens of "Polish"
rugs were shown, a number of them of great beauty. Among these we may
mention those belonging to the Emperor of Austria, the King of Saxony, the Austrian
Museum, Prince Johann von Liechtenstein, Baron Nathaniel and Baron Adolphe
Rothschild, Prince Sdiwarzenberg, Count Schonborn and others. At this exhibition
the above mentioned facts led to the assumption that the manufactory of the „ Polish"
rugs must be sought in Constantinople or near by, where a very well developed
silk and faience industry was in its prime at the same period. This hypothesis has
now been definitely given up, and at the time of the Munich Exhibition in 1910,
their Persian origin was no longer questioned. This exhibition received its first
impetus from the discovery of an entire collection of Polish rugs in the Royal Palace
in Munich by Prince Rupredit. The absence of these rugs in the country of their
manufacture is explained by the fact that they were given as presents by Persian
rulers to European princes. For the same reason, the appearance of coats of arms
and occasional concessions to European taste are no longer surprising.
In the Palazzo Barberini there were formerly a number of such rugs which had
been presented to Pope Urban VII. Others, of similar origin, were in the possession
of the Colonna, Corsini and other Roman families. A particularly beautiful example
of magnificent color harmony and very long pile is owned by the Correr Museum
in Venice. It is probably the rug which an ambassador of the Shah presented to
the Doge in 1604. In 1639 a half dozen Polish rugs were also presented by a
Persian embassy to the Duke of Holstein Gottorp. These are still in Rosenborg
Castle in Copenhagen and were reproduced and described by F. R. Martin. One
of them shows a central shield with two beautifully conventionalized eagles. In the
Provincial Museum in Hanover hangs a particularly beautiful example with a green
field and salmon colored border (Fig. 45). This was also quite obviously a present
to the court of Hanover. In addition to these rugs there are several examples
which belong to the Imperial Armory in Moscow, the Prussian and Austrian Court
and to several Swedish castles. A rug in the collection of the Duke of Buccleugh
is analogous to the specimen reproduced in Fig. 43, save that the design of the
field is more delicate and the border has full arabesques with palmettoes. Several
dozen Polish rugs may be pointed out in private and public collections in America,
partly from the former Yerkes collection21. (An important series in the collection of
Senator Clark in New York, several examples in the New York and Boston museums).
A few excellent specimens, very well preserved, are in the National Museum in
Munich. Another, with a coat of arms is in the Victoria and Albert Museum98. Ac-

Persian, so-called Polish Rugs. 33

cording to Dr. R. M. Riefstahl there is a specimen with a eoath ot arms, which
cannot be identified, in the Mobilier National of the French Republic. Still another rug
from the Royal Palace in Munich (Fig. 46) has a center field composed of two longitudinal
strips with an unusual representation of fighting animals between trees. These groups
are not in opposite symmetrical representation, as would be expected in a rug, but
are identical, as if they were two juxtaposed strips of a patterned brocade. This
rug probably dates from the late 17 th century. We cannot authenticate the „Polish"
rugs in paintings, as such articles of luxury, when they appeared at all in trade,
were too exorbitant for the purse of an. artist. In a portrait by G. Pencz in the Berlin
Gallery, a „Polish" rug seems to be spread over the table against which the man
represented is leaning. Both the yellowish green coloring and the ornamental design
give clear evidence that this is a Polish rug. The date of the portrait (1534) gives
us a welcome chronological hint for the beginning of this class of rugs.
We wish to state, on the other hand, that rugs were knotted in Poland. Some
of these may have been made soon after 1700. However, these rugs have all the
characteristics of Occidental peasant art. They are extremely crude in design and
technique; and their color, especially, has no similarity to the almost too refined
examples which we have just discussed. A very interesting rug which Dr. Kuhnel
saw in the Hospicio of Cordova, may well belong to this class. It had a yellow
ground, with a colorful pattern resembling the "vase,, type. A coat of arms in the
centre bears the insignia of the order of Merced, this being the monastery from
which the rug came. Below is the inscription: "Se hyzo ano de 1773 siendo co-
mendador el R.P.M.F. Antonio Escribano".
The term "Polish rugs" is incorrect, as we have proved above. But if its use
is to be continued for the designation of rugs made for Europe at the Sefevi period,
it must also be applied to the tapestry woven rugs, which were made for the same
destination as the pile rugs, but which were of course intended as wall hangings28.
A certain number of this type of rug were shown for the first time at the Munich
Exhibition (1910). The Bavarian Court contributed several of the most interesting
examples, among them a rug with the Polish eagle, that had come by inheritance
to the house of Wittelsbach. These "Kilims", which are of silk, frequently inter
woven with silver threads, are distinguished by their rich and vigorous color har
mony. Their decoration often takes individual turns. Animal representations are
particularly frequent and even hunting scenes, genii and other figural motives are
no rarity. (Compare one of the above mentioned rugs in the Royal Castle in Munich
and another rug in the Louvre.) Another belonging to Dr. Figdor in Vienna is one
of the most beautiful examples. It is covered with lozenge-shaped cartouches,
showing animals either fighting or standing alone, on a yellow ground. The blue
border is decorated with bright arabesques surrounding bird cartouches. (Fig. 47).
Another, in an unusually good state of preservation, is in the Mohammadan Department
of the Berlin Museum (Fig. 48). The central medallion shows the dragon with the phoenix;
the field, birds among floral motives; the corners spitting Kylins, and the exquisite
border contains lions and tigers between animal masks in four-lobed rosaces. All
this is in a free naturalistic style, reminding one of the classical silk rugs, while
the coloring is very vigorous: carmine, salmon red, Imperial yellow, lemon yellow,
green, light blue, slate blue, dark blue, fawn color, purplish pink, black and white.
In four places the word "Padishah" can be read, which proves that the rug was
made in a court manufactory. The same pattern is repeated in a fragment of the
Copenhagen National Museum. We must also mention a very well preserved specimen
.in the Karlsruhe Museum, a rug with very similar pattern formerly in the possession
Bode-KMinel. 3

34 Bode-Kuhnel, Antique Rugs from the near east.

of the Royal House of Saxony, and another one belonging to the Countess of Beam.
The latter is almost entirely decorated with varying forms of the Tschi. (Fragments
are in the museums of Prague, Dresden etc.) The National Museum in Munich also
owns an embroidered rug, the coloring and design of which place it among the so-
called "Polish" rugs.
In this connection we must mention the unattractive, socalled "Portuguese" rugs.
These were ostensibly knotted for Portugal or for Goa and other European settle
ments in the East Indies. Their Persian origin is as certain as their comparatively
late period of manufacture. They generally have a lozenge-shaped central medallion
which is always filled with floral motives. The outlines of this lozenge are paralleled
by numerous frames of zigzag lines. In the corners are small boats with Europeans
or stiff animal figures. The coarse border shows predominantly arabesque-like forms
and palmettoes. Examples of this type of rug are in the Berlin Museum of Deco
rative Art, the Musee des Tissus in Lyons24, the Lamm collection and elsewhere.

Indo-Persian Rugs.

We cannot enter into a discussion of the Turkoman rugs. Although their home
country is perhaps the cradle of the art of rug knotting in general, the spe
cimens existing can scarcely be traced back as far as the 18th century. This tech
nique was doubtless at its height at the period of Timur and his successors, in the
region of Samarcand and Bokhara. Driven by the Sefevi from their seats in Western
Turkestan, the Timurids founded the empire of the Great Moguls in India and
brought with them the knotting technique as well as the arts of miniature painting,
silk weaving, etc
Until very recently, and even in the first edition of this book, a group of rugs
which are obviously of the 16th and 17th centuries, had been classified among the
Persian rugs of the classic period. Their decoration has a certain . relation to the
style of the classic Persian rugs, but some particularities of the design and the color
place them in a class by themselves. We may now definitely state that they ori
ginated in Northern India. Indeed, this new group is best termed "Indo Persian"
to show that it is a variation of Persian art on Indian soil.
The earliest example of this group (very probably about 1500), is a phantastic
animal rug. Two fragments of this rug, belonging to Dr. Roden in Frankfort and
to the Jeuniette collection in Paris, (Fig. 49), were exhibited in Munich25. Two
entirely analogous fragments, which may not come from the same rug are in the
collection of Kouchakji freres in New York and of Mr. Henry Lowenfeld in Paris.
All these fragments show a peculiar decoration of various grotesque animals which spring
from each other's mouths in wild confusion. A piece in the Musee des Arts decoratifs, show
ing animal heads on vines, represents a later form of this motif. The Mogul manu
factories attain their most complete development in several rugs of very realistic
design, decorated with both plant and animal motives. The Austrian Museum in
Vienna owns two by far the most beautiful examples of this type. The first (Fig. 50)
shows all kinds of birds which are realistically conceived and designed with great
freedom. These include peacocks, cranes, hens, turtle doves, hoopoes, partridges, etc
moving among trees and bushes with luxuriant foliage. The field has entirely lost
its textile-like appearance and is purely pictorial without any symmetry, repeated
motives, or restraints of architectonic composition. The border alone is reminiscent
of the more severe style of the rugs of the Sefevi period. It shows large palmetto
flowers with lion masks upon slender curving stems. This design is interrupted at
intervals by small panthers and bulls spitting at each other. The dashing and
picturesque effect of the design is especially reminiscent of certain Persian and In
dian lacquer bindings of the 16th — 17th century (among others, those in the Hamburg
Museum for Art and Industry) which have obviously served as models. The same
analogy suggests itself in a magnificent piece in the Boston Museum, showing a
fierce hunting scene with all kinds of Far Eastern and Indian animals (Fig. 51).

36 Bode-Kfihnel, Antique Rugs from the near east.
Besides the wild animals, this rug has a representation of a hunting car drawn
by an ox and a glimpse of the diversions in a garden pavilion. The border shows
palmetto masks of grotesque design. There are several animal rugs similar to the
two examples mentioned above, except that the decoration is coarser and less ela
borate. These are in the Widener, closely related to the Boston specimen, Goupil,
Yerkes, Karthaus (Potsdam) collections, two in the Metropolitan Museum, gift of
J. P. Morgan etc. Others are in the London and Paris museums of decorative
art. A fragment from the Sarre collection also deserves mention. It shows two
elephants of fine color and design in opposite representation.
The second of the magnificent rugs of the Austrian Museum (Fig. 52) has the
characteristic niche of the prayer rug type. Naturally its religious purpose prohibited
this use of figural motives. Hence the field is entirely covered with closely growing
flowers of delicate hues which all spring luxuriantly from a single shrub. Another
Mihrab pattern with a simple, loosely designed thistle plant, is represented by rugs
in he collections of Mr. Yerkes26, Mr. Engel-Gros and Mr. Sassoon London. Almost
the identical pattern is found in Persian brocades from the Beghian and Kelekian
collections. As a rule, the most reliable identification for the Indian rugs is their wine red
ground, the tonality of which is essentially different from the typically Persian shades.
The complete disregard of symmetry is, also, as we have seen, a typical characte
ristic of Mogul rugs. On the other hand, if one finds examples with a repeat, as
in textile patterns, it is quite safe to conclude that they are of later date and that
they were influenced by the velvet industry of southern Persia and northern India,
which was highly developed in the 17th century. For example, this influence is
unmistakable in a silk rug with flower bushes and birds in the Musee des Arts
decoratifs (Fig. 53) and still more plainly in a third specimen in the Austrian Museum
(Fig. 54). The latter has a monotonous field the entire decoration of which consists
of two inconspicuous floral motives set in unending rows. Even the somewhat
severer border cannot give to this decoration the character of a rug.

So-called Armenian Rugs.
The rugs of the Sefevi period show the knotting art in its highest develop
ment. When studying them one always feels that both technique and design
must have gone through a number of preliminary stages before reaching such per
fection. Therefore it would be of the greatest importance to find earlier examples
showing these previous stages of development. But difficulties arise when we take
stock of the existent examples of the early knotted rugs. To be sure, there are a
number of specimens from Asia Minor and Spain which are unmistakably to be
attributed to the 15 th century and a few less numerous examples from an earlier
epoch. But only one group is related to the Persian rugs and can be defined with
some exactitude. These are the so-called Armenian animal rugs.
Examples of this type are for from numerous and it is also difficult to identify
them in paintings. Indeed, little was known about Armenian rugs before the first
edition of this book. In the Munich Exhibition they were represented by half a
dozen examples.
Instead of the large homogenous patterns of the Sefevi period, these rugs show
a combination of a number of dissimilar compartments.
The elaborate flowered ground which in other types of rugs serves as a back
ground for the animal representations, is either entirely lacking or very much reduced.
The animals appear most frequently upon a plain colored, unpatterned ground,
showing, at the most, single angular flowers or other small ornaments used to fill
in the space. Instead of the usual wide and elaborate borders, they have only a
narrow border with slender, severely conventionalized vines, leaf ornamentation and
similar decoration. As a rule, the usual small borders on the outside and inside
of the main border are missing. The different compartments are separated by strips
of poorly designed geometrical figures or else by thin and severely conventionalized
floral scrolls. The colors are vigorous and primary. The ground is usually yellow,
rarely white, red, purple or black. The wool is rough, with a high pile and good
knotting. All these elements make these rugs austere and primitive, sometimes even bar
baric, yet at the same time, they are of an extremely logical and rhythmical con
ception. Therefore, we may at once exclude the assumption that the Armenian rugs
are a decadent form of the 16th century animal rugs. On the other hand, one
could overestimate the chronological distance between these rugs and those of the
classical Sefevi period and place a yawning gap between both. F. R. Martin in his
"Oriental Carpets" seems to take this view by placing these rugs as early as the
middle of the thirteenth century. However, they are certainly not that old, for
despite dissimilarities in style, several decorative motives show a very close relation
ship to certain Sefevi patterns, especially those in the vase rugs. In the latter a narrow
border and rather severely conventionalized decoration are equally frequent. A well-

38 Bode-Kuhnel, Antique Rugs from near east.

known example — the large vase rug in the Ottoman Museum (Fig. 35) — shows an
entirely analogous division of the field. So even if we regard this group as direct
predecessors to the magnificent Persian rugs, they must have originated in a part
of Persian where the conditions for this craft were essentially different. Furthermore,
we may ask: Are these rugs an article of luxury which were made in some cul
tural center in a manufactory conducted by artists, as was doubtless true of the
majority of the rugs previously discussed, or do they represent examples of a highly
developed folk art, intended for the palaces of the rich but made by peasants or nomads
in whose tribes rug knotting was an ancient tradition? Without definitely answering
these questions, we can establish this much: the primitive character of the early types
may be readily explained as a particular style created by the requirements of a
particular technique. In this style all decorative motives are deprived of their real
ism and are expressed according to the technique of knotting. The other rugs of
the 16 th and 17 th centuries, are an outgrowth of the general refinement of
artistic taste, and attain such technical perfection, that their spirited design almost
equals the art of painting. But the Armenian group had already been influenced
by foreign civilization, as several decorative motives prove, particularly the Far Eastern
fabulous animals, which are found on practically all the examples preserved. The
use of these animals may be readily explained by the admiration of the Persians
for Chinese art. In this instance, these animals are obviously characterized as Mon
golian heraldic animals, despite their barbaric design. The dragon standing upright,
and the dragon fighting with the phoenix are rough reproductions of the Ming
symbols27. Besides these motives the Foho with the Kylin also appear. It might
be claimed that these rugs are of Mongolian or Chinese origin. But this is im
possible, for different reasons: the sacred animals of Chinese mythology are repro
duced without intelligence, and, in some cases they are intermingled with animals
of diverse types; some of which have no relation to China whatever and are of
purely Near Eastern type.
We must therefore seek elsewhere for the origin of these rugs. Perhaps the
hypothesis of their Armenian origin which has been gradually adopted is right One
must remember that Armenia was repeatedly in close connection with Persia, not
only culturally, but politically, and that knotting is still practised there in many
places. Indeed, we may perhaps discover the last influences of the old patterns in
the technique, color and design of many 17 th and 18 th century rugs which
doubtless originated in Armenia. In rugs of this type the animal representations
have become transformed into ornamental motives. Several examples of this type
are in the Austrian Museum. A piece mentioned by F. R. Martin shows an Ar
menian inscription and the date 1684. The use of Far Eastern motives in a rug
made in the northwestern border territory of Persia is not at all astonishing. On
the one hand, the first Seljuks brought all kinds of inspiration from the East On
he other hand, after the invasion of Hulagu — i.e. from the second half of the 13th
century — an entire flood of Far Eastern motives poured in upon all the decorative
arts of the Near East.
As we have said, there are only a few existent examples of these rugs. Several
very large pieces show dragons standing upright alternating with other animals.
One, found in the Orient, and formerly in the collection of Theodore Graf in
Vienna, has been secured by the Kaiser Friedrich Museum in Berlin (Fig. 55). Two
other examples, once in an Italian church, are now in the Bardini collection28.
Others are in the collections of Mr. Williams, Norristown, Pa., Mr. Sharpies, West
chester, Pa., Mr. J. Mcllhenny, Philadelphia; D. C. Jackling, San Francisco, Professor

So-called Armenian Rugs. 39

Sim Khovitch, New York, James F.Ballard, St. Louis, George Hewitt Myers, Washington
and the Boston Museum. Another formerly in the collection of the author, is now
in the Kaiser Friedrich Museum. Others are in the collections of Mr. Lamm in Nasby,
Sweden, and Mr. van Stolk in the Hague and in the Victoria and Albert Museum.
All these rugs are very similar. One specimen, found in a church near Venice, has
rows of lozenges with a dragon motif alternating with others whose design seems
to represent a pair of running stags near a tree. In the first of the above-mentioned
rugs, running bulls, panthers, hares and other animals which always appear in pairs,
are used with the dragon motif. In others, the dragon and Kylin are represented
fighting with the Foho. They are doubtless of different periods, the later ones with an
obvious tendency towards simplification and distortion of the pattern; the one repro
duced in Fig. 56 may be regarded as a transitional specimen. The color scale is
usually very rich. In the Berlin rug, for instance, we find the following tones: —
white, yellow, red, salmon red, blue, sky blue, bluish green, vernal green, light gray, light
brown, dark brown, black. In the Arabian Museum in Cairo there is a later example
with stiffly designed human figures. In this rug, the alternating of the animal fields
with those bearing flowers or trees, and the peculiar borders of the lozenges, are
noteworthy. Formerly there have been attempts to find traces of Arabian writing
and all kinds of cryptogrammatic enigmas in the conventionalized vines of the
border. The Viennese Orientalist Karabacek even thougth that he could recognize
an inscription in the rug belonging to Mr. Graf. According to his reading this
inscription proved that the rug originated in the 13th century in northern Syria.
This is all pure imagination; the rugs have no inscriptions whatever.
Several other rugs belonging to the same class have no representations of
mythological or real animals. However, they show many analogies in technique. In fact,
the mark of a comparatively early date actually appears in only one other group,
in which large floral sprays are the characteristic form of decoration. Although this
group has many motives peculiar to itself, the general conception seems to be the
same. An example of this type is partly in the Kaiser Friedrich Museum partly in
the collection of Mr. C. F. Williams, Norristown. (Fig. 57). 29 A second is in the
collection of Mr. Lamm in Nasby. A third belongs to Mr. M. K. W. Vanderbilt
in New York. These rugs have comparatively large longitudinal cartouches divi
ded into larger and smaller halves of a peculiar polygonal form. They appear in
rows of four, alternately pointing upward and downward. The two outer ones
are cut by the border and hence only partially visible. The larger half of
these units shows a central decoration of one large, beautifully conventionalized
flower, the smaller one a more inconspicuous palmetto of similar design. Both are
enclosed by a vine with small flowers. The sections between these fields are always
filled with a compact floral branch which has a heavy stem. In the one row these
sprays bear small blossoms like the almond; in the other they have fruit motives,
similar to the pomegranate. The more naturalistic form of these motives affords a
fine contrast to the conventionalized flowers in the cartouches. The border design
of the Berlin rug establishes the relationship between this type and the Armenian
animal rugs. The original and powerful style of its decoration manifests itself in
the characteristic division of the field into cartouches, but in the present group the
small flowers are in complete geometrical conventionalization,
The color composition of this rug is as original and powerful as the design.
The colors are gorgeous, pure and vivid and are blended in a very piquant manner.
This group of rugs, too, is fundamentally different from those of the Sefevi period,
although they have been influenced by the latter in the design of the fruit and

40

Bode-Kfihnel, Antique Rugs from the near east

flower bushes and by the occasional breaking up of the Medallion composition
(Cf. Fig. 20).
Armenia, conceived as the northwestern border territory of Persia, must be
considered as the home of an entire group of knotted rugs, which are best defined
as the successors of the Sefevi pattern. They seem to have come into favor at the
end of the 17 th century after the classical period of the Persian manufactories had
passed, and were then adopted in that region. Most frequently they show a field
with animal combat scenes loosely distributed between cypress trees and other plant
motives and framed by a simple border. All these elements of decoration are
borrowed from the magnificent rugs of the classic period, but everything is reduced
to a very modest scale and is rendered in an awkward and stiff style of design,
often without an understanding of the motif. A close comparison with the classical
patterns of the Sefevi period and with those of their above-mentioned predecessors,
makes it obvious that these rugs are not primitive, but decadent specimens. The
simplification of the decorative problems is evident in the confusion of the design,
the lack of color harmonization , the use of arbitrary border motives and in other
signs of decadence. In spite of their later origin, such animal rugs are rather rare.
We may mention specimens in the possession of Count von Leuchtenberg in Petro-
grad, Koudiakji Freres in Paris, the Cologne Museum of Decorative Art and the
Kaiser Friedrich Museum in Berlin (Fig. 58). Dr. Martin (P. 12 of the text volume),
reproduces a rug of this class, but he ascribes it to the 13th century. On the
other hand, he is probably correct in attributing a very conventionalized hunting
rug of the same class (P. 55 of the same book) to the "Southern Caucasus, about
1680". The later garden rugs, which we discussed before in detail, also belong to
this category (Fig. 59).

Early Anatolian Rugs.

The Nomad tribes of Turkestan who still use the oldest type of purely mathe
matical linear design, have always practised rug knotting.
It is therefore not remarkable that the earliest evidences of the knotting technique
have been found in this territory and in Chinese Turkestan, especially. They belong
to the finds of the Turfan Expeditions of the Berlin Museum and to the discoveries
of Mr. Aurel Stein. These specimens may perhaps be from the 5th-6th century
A. D. R. Berliner and F. Sarre have very recently called our attention to these first
feeble attempts at a type of art which was destined to such an important development.
In the Middle Ages, therefore, the Turkish population which settled in Asia Minor
have contributed most to the art of rug knotting. This art seems to have flourished
even at the time of the Sultanat of Konia, for Marco Polo, who travelled in the
Seljuk empire at the end of the 13th century, mentions that the best and most
beautiful rugs in the world were made there. Several existent examples may be cited
as proof of this statement. Three of these rugs, which must have been made about
the 13th century, laid in the columned hall of the mosque Ala-ed-din in Konia until
their transfer to the Ewkaf Museum in Constantinople. A fourth specimen appears
to have served young Giotto as a model for a curtain in one of the frescoes of the
upper Church of Assisi. The decoration (Figs. 60—62), consists of simple geome
trical motives which in three of the rugs extend symmetrically over the entire field.
But in the fourth rug, the field is composed of completely irregular octagons. In
three examples the border still shows vigorous conventionalized Cufic writing, while
in the fourth (Fig. 62) this writing has become a mere ornament. The primitive
character of these rugs is beyond question. They are without the slightest trace o
floral or figural motives , and even the ornamentation is extremely simple and primitive.
The coloring, predominantly red, yellow and blue, also lacks all subtlety of shading.
Several Spanish Moorish rugs of the 14th — 15th centuries, discovered some time
ago, are probably most closely related to this group, and in some respects seem
to be their successors. However, we are not concerned with these rugs.
In the 13 th or 14 th centuries animal forms may also have entered the con
ceptions of the rug knotters of Asia Minor, but naturally, such motives were first
expressed in very severe linear style. We have positive proof for this in an in
complete rug which was found in a church in central Italy. This rug was later
brought to Rome, and is now in the Kaiser Friedrich Museum (Fig. 63). The two
fields still preserved show the same pattern — the dragon in combat with the
phoenix. The design is very barbaric ; in fact, it is almost a caricature. It is probable
that this rug originally had three, and perhaps even six fields, since there is only
a border on one of the smaller sides. The framings of the octagonal fields and
the narrow border show simple patterns of rigid geometrical conventionalization.
The colors are as simple and crude as the pattern. The ground is a vigorous yellow,

42 Bode-Kuhnel, Antique Rugs from near east.

the animals are blue and red, while the border shows red ornamentation upon a
black ground. A publication on antique Norwegian rug designs shows a similar
piece. Certain details reveal that this Norwegian rug is not of Oriental origin.
Yet it is so similar to the Near Eastern group just described and so strikingly Oriental
even in the smallest, details of composition and decoration, that it must be an imi
tation of this Oriental type. The plundering expeditions of Norwegian sea robbers
as far as the East coast of the Mediterranean might very well have brought Sara
cenic wares to Scandinavia. Besides, we are well aware of the lively trade relations
between the North and the Orient. The Berlin rug is the only original still existing,
but a fortunate circumstance has preserved the reproduction of an identical rug on
an old painting. This will allow us to determine its exact date. This painting is
a fresco in the Spedale (Hospital) della Scala in Siena by Domenico di Bartolo.
It represents the "Marriage of the Foundlings" and was painted between 1440 and
1444 (Fig. 64). The pattern is almost identical to that of the Berlin rug and is
conventionalized in the same rough way, while the corners of the octagons show
very similar degenerated ornaments. However, the borders of the individual com
partments are wider. It is worthy of note that the border of this rug shows a
motif which resembles a battlement and which in our discussion of the so-called
"Polish" rugs was designated the "reciprocal lily" pattern. The same rug is also
represented in a tournament scene upon a practically contemporaneous cassone
picture. Fortunately our judgment of mediaeval knotting in Asia Minor is not dependent
upon this one existent example. Similar rugs are also found in a number of pic
tures from the beginning of the 15 th and from the 14 th centuries, besides a few
isolated examples from the end of the 13 th century. All these confirm our early
dating of the rugs themselves.
A rug with a very primitive pattern appears in a large Madonna painting by
Fra Angelico with the figures of Saint Cosmas and Saint Damianus. This picture
was painted in 1438, and is in. the Academy of Florence. In each of the fields
animals are grouped in twos, one above the other; none of them can be definitely
identified. Small stars, zigzag lines and similar simple ornaments enclose the com
partments while the narrow border is decorated with rosaces. A rug in the "Betro
thal of Mary" by an unknown Sienese, shows similar primitive representations
with an animal in each field and ornamental Cufic writing in the border. This
picture, which dates from the 15th century, is in the National Gallery in London
(Nr. 1317). Another similar rug is found upon an almost contemporaneous portrait
by Giovanni di Paolo wrongly ascribed to Pisanello, which is in the Galerie Doria
in Rome. The pictures of the Trecento also show very simple rugs with severely con
ventionalized animal motives. However, their designs are not as crude as those
from the first half of the Quattrocento. Birds are the main motives in the fabrics
of this period, and in the rugs they seem to be the only existing patterns, as far
as we can recognize them with certitude. They seem to occur in the frescoes,
particularly in those by Giotto in Assisi, but unfortunately they are restored beyond
recognition. The different bird patterns are closely related. A single bird, or two
birds in opposite representation are shown in octagonal or quadrangular compart
ments of vigorous coloring. In the corresponding compartments the animals are
usually in opposite representation according to the character of the rug, and the
colors are also reversed. The small framings and even the main borders are as a
rule without adornment save at their intersection where a small, angular ornament

Early Anatolian Rugs. 43

is regularly used. Small ornaments, similar to these, decorate the bodies of the
animals. The colors are mainly yellow and blue or yellow and red in combination
with white and black. One of the pictures showing such rugs is "The Betrothal of
Mary" by Niccolo die Buonaccorso of Siena. This painting is in the National
Gallery in London (No. 1199; painted about 1380, Fig. 65). It shows very conven
tionalized yellow and red birds which are seemingly eagles, standing in opposite
representation upon a red and yellow ground respectively. There is a similar rug
in one of the Madonna pictures of the Berlin Gallery ascribed to Lippo Memmi
(Nr. 1072. Fig. 66). This picture was painted about 1350. The fields show two
long legged eagles in heraldic representation on either side of a conventionalized
tree. The framings and the borders show only thin ornamentation. The large pain
ting of St Louis of Toulouse by Simone Martini inS. Lorenzo of Naples shows
a rug with a double eagle repeated in all the fields. This painting is twenty or
thirty years earlier. A rug in Giotto's famous triptych in the sacristy of St.
Peter in Rome, dating from the end of the 13th century, has the same decoration.
Another animal rug of similar type is reproduced in a somewhat older religious
painting of the Sanctissima Annunciata in Florence. An exact description of this
rug is impossible, as the painting is an object of worship, only visible once a
year at a great distance. The numerous copies of this painting are very super
ficial. Most of the reproductions of rugs in early pictures show the patterns in a
very much reduced scale. When we try to imagine these patterns in their actual
size, they seem strikingly bare and unimaginative. Although the Konia patterns
would lead us to expect such simple designs, we must not think that these pain
tings reproduce the original rugs with exactitude. Every building, every piece of
furniture, all textile patterns and all the objects of art in the pictures of Giotto and
his followers, when compared with the originals, appear as a kind of abbreviation,
and without doubt this is also true of the rugs of this early period. On the other
hand, we may accept without question the representations of rugs by artists of the
Quattrocento, who loved to devote themselves to the smallest details; who cons
cientiously reproduced every brocade pattern in the robes of their saints.
We may only make a general assumption as to the exact origin of the rugs
thus rendered by the artists. These pieces were probably brought from the regions
of the Near East then accessible to Italian trade — Anatolia and Armenia, and
in rare instances from Syria also. It is likewise probable that they came from dis
tricts near the sea. Unfortunately the authors of the 13 th and 14 th centuries
allude to rugs only incidentally and casually. However, they mention Anatolia as
the home of a flourishing rug industry and speak of the exportation of rugs from
Asia Minor to Italy. This leaves us little doubt as to the origin of this early class.
In it we may easily recognize the early stages of the so-called Armenian animal
rugs, which then served as a transition to the masterpieces of the Sefevi period.

So-called Ushak Rugs from Asia Minor.
We have already mentioned that the Asiatic origin of one of the types of rugs
previously described has been questioned. We refer to the "Polish" rugs,
as they were erroneously designated. But there are still several other types about
which similar doubts have been expressed, and indeed, with much greater justi
fication. If the well known Smyrna rugs are included' in this class, we must admit
that they were made for Asia. Indeed, they are almost the sole antique specimens
for which we can prove an exact region of manufacture. But they were probably
made in response to orders and commissions from Europe, and perhaps even under
partial European supervision. Therefore, they cannot be placed among the purely
Oriental rugs.
The manufacture of rugs in Smyrna and the environs is one of the most recent
industries of this kind in Asia Minor. But nevertheless, it adhered most carefully
to the old design and technique. The coarse wool, loose knotting, high pile and
subdued color harmony characteristic of this type distinguish them from all old
Oriental rugs.
Another unusual feature of these rugs is that they are approximately square,
instead of long and narrow. Many Dutch and English houses, and particularly
those of the 18th century, are furnished with Smyrna rugs which exactly fit the
rooms where they lie. There can be no doubt that such rugs were made to order.
Furthermore, the patterns of these rugs are not individual ones, but copies of diffe
rent old designs. The most common type are those with patterns reproduced in
Figs. 67, 68, 74. These designs, in combination with a simple scattered flower
pattern, have been repeated since the 17th century with little change. So we may
assume that this industry was started either in or near Smyrna by the European
colony which flourished there and whose members were mostly Dutdi. European
orders were executed by Oriental workmen. This explains the frequent appearance
of such rugs in Europe, especially in Holland and England. It also explains their
long popularity and their adherence to the old traditions. On the other hand, it
likewise accounts for the imitation of Smyrna rugs when the modern rug industry
was founded in Europe.
The old prototypes which interest us particularly have also been attributed to
Smyrna, because of the similarity of their patterns. This attribution might be correct,
if one interprets it as the surrounding province for which Smyrna was the export
center, and not the city itself. But since there is still a center of production in this
district whose recent products are similar to some of the old types, we are justified
in transferring the term "Ushak" rugs from the later to the earlier group.
Among the old specimens of this family there are various types, which differ
in design but are related in color and technique. As a rule they show some form
of the characteristic medallion composition and in this respect, as well as in details

So-called Ushak Rugs from Asia Minor. 45
of the decoration, remind one of certain Sefevi patterns. Their date also corresponds
approximately with that of the Sefevi rugs, so that Persian influence may well be
taken for granted. This is most apparent in a pattern which agrees approximately
with that of most of the animal rugs and some of the rugs with inanimate design
(cf. fig. 23 and especially fig. 25). The field shows a round central shield of
zigzag outline with delicate attached motives at either end, and the corner motives
consist of more or less exact sectors of the central shield with very small heart-
shaped motives attached toward the inside. Sometimes, especially in particularly
large rugs, a section of the central shield is repeated on the long sides of the rug,
generally that part, which is cut off by the border. The design of these fields
shows varicolored and elaborately interlaced vines with small, conventionalized
flowers and leaves. A somewhat similar design fills the ground of the center field.
It has the appearance of a scattered flower pattern, for its slender vines are com
pletely subordinated. The border, framed by two narrow strips with a ribbon-like
pattern, usually shows a vine design with alternating flowers and leaves of very
slender, conventionalized form. The cloud band with scattered flowers between is
only used occasionally. Interlaced Cufic characters are still rarer. In the central
medallion and in the corners we find large and beautifully formed arabesques. The
colors of the ground are usually deep and strong. As a rule the center field is red
or reddish brown, the border dark blue. The design is in bright yellow, bright
blue, green, with yellow contours etc
Rugs of this kind, which are generally of considerable or even very large size,
are not rare in the Orient. They were particularly common in Italian churches until
rather recently. They are found fairly often in those of Spain, Portugal, Tyrol and
southern Germany. From the churches they found their way through the art trade
into most of the large museums of decorative art and into many private homes,
where unfortunately, they are soon destroyed by daily use. One of the most beau
tiful examples, both in coloring and design, belonged to Count Gregor Stroganoff
in Rome (Fig. 67). He bought it in 1883 from the sale of the Castellani collection
which offered a rich choice of interesting old rugs, almost all of which original ly
came from Italian Churches. Proffessor Gustav Guterbock in Berlin owned formerly
an excellently preserved rug of this type almost 29 feet (9 m.) long, with high pile.
Other examples are in St. Michael's Court Church in Munich, in the Chamber
of Trade and Commerce in Bozen, in the possession of Baron Haniel, Berlin etc
The date of these rugs may be exactly determined because of their compara
tively frequent appearance in old paintings of widely different schools. We will
only mention a few of the most important of these. The earliest date is gained
from a remarkable representation of the family of Henry VIII of England which
was painted by order of Queen Elizabeth in about 1570. This picture was shown
in the Tudor exhibition in London in 1888, No. 158. Next comes the rug in a
Spanish picture — the large Zurbaran in the Louvre, which was painted about 1622.
A rug of the same pattern is in the portrait of Princess Margarita Theresa by
Velasquez. This picture, which is in the Court Museum of Vienna, was painted
about 1656. Various Dutch paintings which show the same type of rugs belong to
the same period — the large Jan Vermeer in the Dresden Gallery (dated 1656), the
"Concert" by G. Terborch in the National Gallery in London and others. This
pattern does not appear upon later paintings. According to the pictorial represen
tations, the change which it shows in the course of almost a century is a very slight
one. The later period can only be distinguished from the earlier one by greater
simplicity, clearer arrangement and by the angular design of the leaves and flowers,

46 Bode-Kfihnel, Antique Rugs from the near east.
especially those in the small borders. On the other hand, numerous rugs which
still appear in the mosques of Constantinople, Smyrna and the neighborhood, down
to the rugs now made in the Ushak region, show the gradual degeneration of this
type. In the last examples we may only recognize a weak remnant of the vigorous
style of the older rugs.
A medium-sized rug in the Poldi Pezzoli Museum in Milan is of particular
interest for the age of this large star medallion pattern and its evolution. The black
background forms a strong contrast with the large dark red medallion star of simple
outline, which fills practically the entire field with its small medallions longitudinally
attached and with the small sectors repeated in the four corners. Because of the
thin, ribbon-like design, the Saracenic flowers and arabesques can only be deci
phered with difficulty. But in the border, the flowers, at any rate, have a fuller,
more naturalistic form. On the other hand, the angular vines encircling these
flowers scarcely allow one to guess the original floral forms and the composition
of elongated fields with flowers and round star medallions. If one has not seen old
rugs in large numbers, the hardening of all forms and the rough knotting causes
one to conclude at first glance that this rug is a late decadent imitation of the
original pattern. But a careful examination reveals clearly that it is an early spe
cimen which perhaps goes back to about 1500. This conclusion is supported by
the early style of the design of the small borders which is very similar to that of
rugs upon late Quattrocento pictures. It is confirmed by the primitive, but not at
all careless technique, and above all, by the deep, powerful coloring. The fine
preservation of the black ground is especially significant, as in 16 th century rugs
and particularly in those from later times, this color was entirely eaten away by
the dyes. A very fine early rug of this class, illustrated in an excellent colored repro
duction, is attributed by Prisse d'Avennes, (l'Art Arabelll, Plate 153), to the 18th cen
tury. At the same time he describes another rug (Plate 150), also in colored repro
duction, as a work of the 14th century, although the similarity of this rug to those
with the design we have just described, makes such a difference in dating very
improbable. However, the frequent appearance of this second type of rug in pain
tings proves that it was made at almost the same time as the first example cited.
Its center field shows an alternation of large, eight-pointed stars of delicate shape
with smaller four pointed ones of similar configuration. On both sides, these stars
are flanked by similar smaller and larger stars respectively, which are cut in half
by the border. As these rugs are always of medium or large size, the total length
of the pattern is formed by three or five central stars with one half star on both
small sides of the rug. The inside of these stars is decorated with delicate and
angular vines with flowers and arabesques of slender design, which we may com
pare to those of the round middle shields of the previous group. The ground of
the center field is similarly decorated with a few flowers on slender vines of angular
design. The border of this type of rug usually shows either the cloud band com
bined with a few flowers or slender interfacings of vines, framed on either side by
a small border with very slender ribbon-like flower vines in conventionalized form
(Fig. 68). In this class of rugs the colors are still more stereotyped than in those
above. The ground of the center field is usually a beautiful carmine with scattered
flowers in brownish or green tones. The star medallions have a deep blue ground
with a yellow design and the ground of the border is generally sky blue.
Early in 1914 three such rugs from the collection of the Earl of Dalkeith, with
the old type of cartouche border and the insignia of Sir Edward Montagu were

So-called Ushak Rugs from Asia Minor. 47
exhibited in the South Kensington Museum. One of these rugs showed large and
small stars ranged in gradually increasing numbers along the width (Fig. 69). These
pieces are of inestimable importance, as two of them are dated 1584 and the third
1585. This is a very important and unassailable date for the history of rug knot
ting, in Anatolia, as important as the Ardebil rug for Persia. A. F. Kendrick has
recently been inclined to the theory that these rugs were woven in England, although
they are in the original Ushak style. We continue to believe that they are Oriental
rugs, woven to English order. Upon the evidence of paintings from all the nations
which at that time imported rugs from the Orient, this group may be dated from
the middle of the 16th to the middle of the 17th century. Among those showing
especially characteristic examples, the famous "Ring", by Paris Bordone, deserves
particular mention. This picture, which is in the Academy of Venice, was painted
about the middle of the 16th century. In this picture the throne of the Doge
stands upon a magnificent Ushak rug of this type.
A similar rug is shown in a large portrait dating from 1614, painted by Marcus
Gerard who was then working in England. A painting by Zurbaran in the Raczynski
Gallery in Berlin, dating some twenty or thirty years later, shows another such rug
lying on the floor. Rugs of this class appeared rather frequently in Italian trade.
One example of particularly beautiful color and design, is owned by the Church of
St. Ann in Augsburg. A similar, although incomplete specimen, belongs to the
Museum of Decorative Art in Dresden; an early, small specimen of excellent quality
to Dr. Fritz Harck in Seusslitz; a finely preserved large rug to Count Donhoff-
Friedrichstein. Other examples are owned by the Austrian Museum, the Kaiser
Friedrich Museum in Berlin, Mr. Williams in Norristown and others.
Another rug from the Kaiser Friedrich Museum shows this type in incomplete
development (Fig. 70). The stars in the center field have not yet taken a definite,
entirely regular form, while their design shows full and beautiful flowers of angular
design instead of the later thin and rigid interpretation. But in the border this con
ventionalization has advanced further. The coloring is very deep and magnificent
The grounds of the center field and of the border are dark cherry red; the medal
lions are dark blue. The design and conventionalization prove that this isolated
type belongs to the beginning of the 16 th century.
A pattern appearing even more seldom (Fig. 71), is related to all three schemes
we have mentioned: — to the first in composition, to the second in design and to
the third in certain details. As the technique and colors, too, are identical, it is
safe to place this rug as early as the 16 th century. The deep blue ground is de
corated with the usual vines, which are generally yellow. As in the above men
tioned type, the center field shows either three or five medallions with a red ground.
These are of elongated hexagonal shape and of equal size. On both sides of the
rug, between these central medallions, are stars cut in half by the border. They
have a blue ground and small attached medallions. Analogies can easily be found
for the floral and arabesque decoration and their conventionalization, as well as for
the coloring and the design of the border. Besides the example from the Austrian
Museum above mentioned, a very good specimen of this type belongs to Baron
H. von Tucher in Nuremberg (Fig. 72). This rug is simpler and clearer in design
and stronger in its color contrasts. Related to this second group is another, pattern,
which is met quite frequently and which we illustrate by a magnificent specimen
of archaic type with beautiful full flowers and a beautifel border of extremely pure
design. (Fig. 73).
A very different motif, of which to our knowledge, there is only one example

48 Bode Kuhnel, Antique Rugs from the near east. 
(Fig. 74), belongs to this class. This motif is of peculiar importance for the prece
ding as well as for the later period and shows a more compact arrangement of the
second pattern which we discussed above. Angular vines with thin and scarcely
recognizable leaves and blossoms of thin form, twine in both directions from one
star medallion to another. In this wag the large floral stars are enclosed in a
rhomboidal frame. The border shows the cloud band with scattered flowers between,
exactly as in the other pieces. (Fig. 68). The rich, brilliant coloring also corre
sponds with that in the other rugs. The ground of the center field is a deep blue,
which is rare in rugs of this type, and that of the border and of the large star
flowers is vermilion. The decoration is in bright yellow, various shades of green,
bright blue, salmon, red, and white. These characteristics make it most probable that this
rug originated in the same region as the other varieties. The date of its origin is
also approximately the same. This variant of the Ushak pattern has been a favo
rite design in the "Smyrna rugs" manufactured for Europe and has been copied
since the 17th century.

In connection with the Ushak group above described several types of prayer
rugs also originated, which in accordance with their purpose are of small size.
These small dimensions necessitated a selection from the variety of motives deve
loped in the patterns of a larger scale. One of the commoner designs, in which
for the sake of symmetry the niche is used on both ends, appears in paintings for
nearly a century80. Formerly originals of this type appeared rather frequently in
the Italian art trade (Fig 75). The center field shows a single colored ground
bearing a small four or six sided star with a detached conventionalized hanging
lamp in each niche. The corner pieces have motives which may be recognized as
degenerated cloud bands. The border shows either the same motif combined with
scattered flowers or slender interlacings of flowers and leaves. Occasionally the
arabesque is also used, very much as in the large patterns discussed above. In
addition, we generally have the two typical small borders, sometimes even a larger
number. The colors are very deep and gleaming. The center field is a vigorous
red which contrasts with the deep blue ground of the border. The small star in
the central medallions is also deep blue; the corner motives are usually deep green.
The design of the medallions as well as of the border is in varied colors, with red
predominating. Rugs of this kind are still quite common in private collections
(Baron von Tucher, Dr, Bode and others). They also appear in paintings dating
from the middle of the 16 th century. Among these are the double portrait by
Francesco Veneziano, dating from 1561 (belonging to Mr. Holford, Westonbirt);
many pictures by Tintoretto (for example, those in the Brera in Milan); a small
painting by Rottenhammer in the Berlin Gallery (from the beginning of the 17 th cen
tury); a painting by Matteo Rosselli in the Florence Academy, dated about 1620
and many others. However, they are not found much later than this date.
Another type, of the same small dimensions and very similar composition,
showing a much more elaborate decoration in the field of the niche, and a much
weaker border design (Fig. 76), is in the Hungarian Museum in Budapest, and
another in Skokloster, Sweden. There is no star in the center field. The corners
are decorated with slender arabesques, and the center shows slender vines with
angularly designed blossoms and leaves growing from two vase-like hanging lamps

So-called Ushak Rugs from Asia Minor. 49
in the middle. In the border, elongated cartouches alternate with small, eight pointed
stars. The former show delicate vines with small flowers, arabesque motives., and
beautifully conventionalized fully opened flowers. In a somewhat later stage even
this simple border pattern is further reduced to cartouche motives, and the field and
corner designs are also simplified as much as possible (Fig. 77). This is the most
current type of the so-called Transylvanian or „Seven Mountain" rugs. This desig
nation is incorrect, but it is still in favor with dealers and collectors, probably
because the rugs were found so frequently in Hungary and Transylvania. Other varieties
showing less and less connection with the Ushak group, have only a single central
motif. The border is somewhat wider, with cloud bands or similar motives. The
colors of all of these rugs are usually pure and vigorous. The ground of the center
field is generally red, more seldom blue. The period of their manufacture may be
established from pictures of the Dutch school, in which such rugs frequently appear
We may mention paintings by Thomas de Keyser, C. de Vos, P. Codde, C. Netscher,
Jan Verkolje, A. de Snaphaan, Cornelis van Plan and others. They also appear rarely in
the^works of the later Italians, such as B. Castiglione. The last paintings showing rugs bf
this type are the large portrait of the Empress Maria Theresia in the Belvedere in
Vienna and a painting by Battoni in the Ermitage in Petrograd. Therefore, they
were made some time between the beginning of the 17 th and the middle of the
18th centuries. In earlier times originals of this type were frequently found in
Transylvanian and Hungarian churches and occasionally in those of Poland, Italy
and Southern Germany, where the dealers secured them.
According to the above evidence, one might be tempted to ascribe all prayer
rugs of Asia Minor to the seventeenth century. Fortunately, there are still a few
examples which prove that the Mihrab pattern led to characteristic forms even
earlier than this date, although such examples are very rare. In fact, there is only
one specimen of this entire class to serve as proof, but it is a most conclusive piece
of evidence. This is the rug formerly belonging to Mr. von Angeli in Vienna and
now in the Kaiser Friedrich Museum in Berlin. It is a beautifully harmonized rug
with a large cloud band of grandiose design. This fills the entire lower field and
encloses an exquisitely conventionalized full Saracenic flower. (See color plate.)
The main motif, the smaller ornaments and particularly the severe arabesques in the
corners, are of archaic conception. Not only the design, but also the coloring, which
is skillfully harmonized with the predominating note of dark blue, and the very
small double borders prove the early origin of this rug. The knotting is rather
coarse, but it cannot weaken the striking effect of this really perfect work of art.
This piece is without doubt older than all the Ushak types. However, a close rela
tionship with several of them is not to be denied, even though the different com
position almost conceals it.

There is another type of rug related to the Ushak family in many respects.
This group originated in a different region of Asia Minor and is difficult to classify.
It has a white, or infrequently a light yellow ground and a center field with an
all over pattern. Such rugs, which formerly appeared rather frequently in Italian
trade and occasionally in that of Southern Germany and Turkey, show a characte
ristic conventionalization of archaic floral forms in geometrical rigidity of design.
The pattern in this type consists of rhomboidal forms, repeated in regular disposition
Bode-Ktthnel. 4

50 Bode-KOhnel, Antique Rugs from the near east,  ^^^^^^
between cattered sflowers. This rhomboidal design, which has been interpreted as
a pattern of birds pointing in two directions, was more probably derived from an
early leaf motif. The period when this pattern originated (Fig. 78), is proved by
its appearance upon a picture by Varotari in the Ermitage in Petrograd, painted
about 1625 and upon a fresco by Peter Candid in the Munich Royal Palace. Accor
ding to this evidence these rugs were manufactured from about the end of the 16 th
to the middle of the 17 th century. The border is of characteristic Ushak type. It
either shows the cloud band combined with slender vines bearing small flowers or
a characteristic reciprocal triangle design in which completely conventionalized pal-
mettoes may be recognized between distorted arabesques. This same border motif
is occasionally used elsewhere, quite frequently in combination with a rarer center
pattern which shows three superimposed globes above two narrow bands. (Fig 79.)
This motif reminds one of the Tschintamani symbol, which we found occasionally
in rugs of the Sefevi period. (Fig. 12.) Like the rugs just described, this type, too,
usually has a white ground; the design is of rich and vigorous coloring. The
"lightning and globe", or "cloud and triple moon" motives are found in entirely
analogous design upon Turkish brocades about 1600, so that the dating adopted
above seems to fit both groups. In addition to the specimens in the collection of
the author, the following collectors own rugs of this type, most of them of large
size: Prince Schwarzenberg in Vienna, privy Councillor Zander in Berlin, Mr. Wil
liams in Norristown, Dr, Kennedy in Boston, the Antiquarian Simonetti in Rome and
many others, besides specimens in several museums of decorative art

Rugs from Asia Minor; So-called Holbein Rugs.
One of the characteristics of the early Asia Minor animal rugs was the trans
position of motives into a severe, even mathematical style. In other instances
the most varied plant motives were transformed into a geometrical pattern and in
such cases it is just as difficult to recognize the blossoms and leaves as the animal
forms. The farther we go toward the beginning of the 16th century, the stronger
this tendency appears, and in the 15th century it seems to have been fairly general.
Yet even in this period the relation to plant forms may always be established with
careful study.
This is especially true of a type which was imported to Europe in sudi great
numbers that even today hundreds of originals remain. Rugs with this pattern also
appear in old paintings far more frequently than any other type (Fig. 80). From
the evidence of pictures we may at once deduce an unusually long period of manu
facture with but slightly changed design. We may follow the pattern for nearly
two centuries. The basis of the design consists of angularly conventionalized vines
bearing the fully opened "pomegranate" or "palmetto" flowers, combined with the
arabesque. We first mentioned this type of composition in our description of the
Persian rugs, and found it again and again in the most varied groups. The design
covers the entire center field like a network. In the earliest period the border usually
shows ornamentally transformed Cufic characters which gradually become an angular
ribbon pattern. Later a border appears with alternating fields of severe plant or
nament Occasionally the cloud band or thin flower vines are found instead. During
the 17 th century the design of the center field as well as that of the border be
came increasingly obscure and coarse and the knotting looser and of poorer quality.
The choice of colors is also true to type — a rather bright red ground in the
center field, showing a design in yellow with some blue and occasionally some
green touches. The border has a somewhat richer design upon a blue or green
field. Occasionally one finds more lively color contrasts and an approach to the
Ushak types (Fig. 81). The small prayer rug size is the most frequent, but larger
pieces are also no rarity.
The oldest paintings showing this type of design are works of the Venetian
school. Let us mention paintings by Lorenzo Lotto, Girolamo de' Libri (in San
Giorgio in Braida, about 1520 and in the Museo Civico in Verona, dated 1530); Luca
Longhi (1544); a portrait of Cesare Borghia, attributed to Bronzino (1535), which
in 1819 was secured by Baron Alphonse Rothschild from the Borghese Gallery and
many others. This type of rugs is also found very frequently in the works of the
later painters of the Netherlands — from Flanders as well as from Holland. Jan
Brueghel, Frans Francken, Cornells de Vos, Simon de Vos, Hendrik van Steenwyck,
and other painters from Antwerp must obviously have possessed such rugs them
selves. They are equally frequent in the work of Gabriel Metsu, Gerard Terborch,

52 Bode-Kuhnel, Antique Rugs from the near east.
Jan Steen, Bernart Fabritius, Jan van Bronckhorst, Nicolas Maes, Pieter de Hooch,
Godfried Schaldten and many other Dutch painters. Toward the end of the 17 th
century this class of rugs, so common but a short time before, suddenly disappears
from paintings. However, well-preserved examples are not rare in museums of deco
rative art. A beautiful early example is owned by the Frankfort Museum of Deco
rative Art.
A similar design of less compact composition and probably coming from a
different region of Anatolia, is the basic principle of another class of rugs which
also appears frequently in old paintings. The center field of this type is usually
divided into complicated star and' cross forms, sometimes into other geometrical
figures (Fig. 82, 83). In these rugs we find interlaced ribbons with angular turns,
and similar linear designs combined with other motives whose original plant forms
may still be guessed. The arabesque, in particular, may still be discovered. Small
stars of kaleidoscopic design and of purely mathematical form are used to fill in
the space between the larger ones. The main border pattern consists of decora
tively interlaced ribbon motives in white upon a colored ground. At first these were
an imitation of Cufic writing, but later they became a purely geometrical shaft or
nament. Occasionally small polygons or rosaces alternating with stars appear as a
border pattern, and in earlier examples we find a thin vine motif with leaves. The
accompanying border occasionally shows the zigzag pattern which we have described
in another connection.
This group of rugs is characterized by close knotting and vigorous coloring,
generally upon a red ground. They show certain points of similarity with several
examples of Turkoman work which have probably gone through a similar evolution.
We may trace these rugs from about the middle of the 15 th to the middle of
the 16 th century, chiefly in Italian, although sometimes in German paintings and
particularly in the work of Hans Holbein, one of the greatest German masters, after
whom they have been called. The following list comprises a number of the most
important paintings which show the designs particulary clearly : Mantegna's famous Ma
donna painting in S. Zeno in Verona (1459); Carpaccio's series with the story of St Ursula
(1495) in the Academy of Venice; Pinturicchios's frescoes in the Libreria of Sienna dating
from 1505; a few Madonna paintings by Domenico Ghirlandajo in the Academy and in the
Uffizi of Florence (about 1480); a painting by Sebastiano Mainardi in the Gallery
of Naples; Rafaellino's early Madonna in the Berlin Callery and the Family of
Licinio Pordenone in Hampton Court (1524); Bassano's Madonna in the Pinakothek
in Munich, Hans Holbein's Portrait of Gisze in the Berlin Gallery (1532) and his
"Two Ambassadors" in the National Gallery of London (about 1533); the fresco
by Piero della Francesca in San Francesco at Rimini (1451); the Annunciation of
Baldovinetti in San Miniato near Florence (about 1480); the early masterpiece of
Lorenzo di Credi in the cathedral of Pistoja (about 1480); Ercole Roberti's throned
Madonna in the Brera in Milan (about 1480); and paintings in the same museum
by Francesco Bonsignori and Gaudenzio Ferrari; Mantegna's frescoes in the castle
at Mantua (soon after 1460); Badile's Madonna in the Verona Museum (1546); a
man's portrait by Parmegianino in the Naples Museum; another by Dosso Dossi'in'
the Corsini Gallery in Rome and the portrait of a cardinal by Pontormo in the Gal
leria Borghese. A rug of the older type of this group may be identified in a large
portrait group of the Guildmasters of the City of London. This group, which was painted
in 1604 by one of the Dutch painters working in England, is in the National Por
trait Gallery. Such a late representation is plausibly explained by the fact that the
artist showed the worthy burghers of London in their Guild Chamber in which the

Rugs from Asia Minor; So-called Holbein Rugs. 53
rug may have served as a table covering for several decades. Another interesting indi
cation for the chronology of these "Holbein" rugs is given by a large woolen embroidery
dated 1533, which Dr. Kiihnel saw in the National Swiss Museum in Zurich. This
embroidery was intended as a wall hanging and is an exact copy of a Holbein rug.
For a full century the design of this type remains almost unchanged except
for small variations. Until some years ago original specimens frequently appeared
in Italy and in Southern Germany, but as far as we know, never in Spain, which at
that time still had an important rug industry of its own. Examples are found in
the collections of many European museums of decorative art and they still appear
occasionally in the Italian art trade, for example, in the collection of the antiquarian
Simonetti in Rome.
Another series of rugs may be considered as a variety of the present type. They
show the same character of draughtsmanship and seem to borrow from the same
stock of motives. However, in this type we find that the pattern is of larger pro
portions, since only a few stars fill the field. In the other type the stars are in
close rows and of considerably smaller size, and therefore not in very strong con
trast with the ground. Original rugs of this second type are not any too frequent
(Fig 84 — 86). Their vivid colors differentiate them sharply from the other more
softly toned types. Their successors may be found in several types of the later,
so-called Bergama group. Sometimes they have been rather misleadingly designated
"Holbein rugs" because this German master repeatedly used such rugs in his pic
tures, as well as the smaller patterned ones discussed before. Rugs of this type also
appear in other paintings — especially in those of the Flemish and Italian schools
They are always in severe geometrical conventionalization and are most frequent in
the 15th century. The pictures of Hans Memling are rich in such patterns. Other
rugs of this type are found in the paintings of Baldovinetti, Ghirlandajo, Raffaellino
del Garbo, Crivelli, Carpaccio, Foppa and numerous contemporary painters.As far as we
know, a portrait by Sofonisbe Anguissola in the Raczynski Gallery in Berlin, dating
from 1560, offers the latest example. In this instance the painter perhaps used a
rug that was several decades old. In the Venetian paintings, particularly in the
Carpaccios, these patterns are reproduced in an impressionistic style, so as to render
their picturesque effect. In the others they appear in a strikingly hard and severe
interpretation, although the specimens preserved show a rather free design. Conse
quently it is very probable that the artists more or less altered their different models
by making small changes or striving for simplicity. This is especially true of Jan
van Eyck and his pupil Petrus Christus, who liked to introduce Gothic motives. It
applies also to various paintings of the Quattrocento, among others, to a panel by
Fra Angelico in the Academy of Florence, to Piero della Francesca's altar paintings
and his Annunciation in the Uffizi, and a fresco by Foppa in the Brera dating
from 1485. On the other hand, Holbein is entirely trustworthy, since his accuracy
of reproduction also extended to Oriental rugs. Memling, who seldom omits the
rug at Mary's feet in his pictures of the throned Madonna, it also reliable to some
extent The colors of these rugs are fresh and vigorous. The ground tone is usually
yellow or red; the borders are small and of decorative, conventionalized design,
either with inscriptions or plant forms. The size is small and particularly fit to be
used by the artists.

The So-called Damascus Rugs.
Among the Near Eastern rugs, one category . is 1 the subject of especial
contention. This group is completely isolated and is therefore ascribed to
the most varied regions of the Mohammadan world. The color scheme is very
striking (Fig. 87). A more or less deep cherry red, generally used as the back
ground, is combined with bright green and a few touches of pure blue, producing
a very harmonious, but not particularly colorful effect. The material is usuaUy the
wool of the Angora sheep which has a peculiar lustre, and to make it more effec
tive, it is knotted into a silk warp. Occasionally these rugs are entirely of
silk and then they are exceldingly beautiful because of the soft shimmer
of the light which is broken into thousands of hues by the design. The
latter consists of a very pleasant harmonization of manifold geometrical forms,
kaleidoscopically thrown together, with small, diversely conventionalized flower
motives. The center is generally formed of a large, many sided star medallion; in
very large rugs several of these stars lie next to each other. In the border there
is frequently an alternation of circular and elongated fields, as in certain of the Sefevi
patterns. However, in the type under discussion, neither the design or coloring
sharply differentiate the border from the field.
Rugs of this kind are seldom found in the earlier paintings, perhaps because
of their less striking decoration. They may only be definitely authenticated upon
several pictures of the Venetian school. "The Debate of St. Stephen" by Car-
paccio in the Brera shows the pattern very plainly, but in accordance with the
artist's technique, the reproduction is rather superficial. The design is shown much
more exactly in a portrait by G. B. Moroni, dating from about 1550, which in
1893 belonged to Charles Fairfax Murray in London. On the other hand, originals
of this type are not rare. Until a short time ago they often appeared in Italian
churches and occasionally in those of southern Germany, from which they were
secured by dealers and museums. In exceptional instances the bazaar in Con
stantinople also offered such rugs for sale. Most of the rugs of this type are in
bad condition, because the fineness and perishable nature of the Angora wool
makes them unfit to withstand usage as floor rugs in the churches. J. Lessing
sets the date of their origin as "about the middle of the 16th century". However,
the manufactory must have been in its prime a half century earlier, as is proved
by the above mentioned reproduction by Carpaccio from the end of the 15th cen
tury. Among the well-preserved examples we may mention a specimen in the
Kaiser Friedrich Museum (Fig. 87), a similar rug, which was lent by Count Baillet
Latour to the Vienna Exhibition, and a large specimen of elongated shape
belonging to the dealer Sangiorgi in Rome, formerly in the Simonetti collection.81
Most beautiful of all is an incomparable silk rug formerly in the possession of
the Austrian Imperial family and now property of the Austrian State, which is

The So-called Damascus Rugs. 55

without question a marvel of Oriental knotting technique (Fig. 88). Beautiful fragments
are among others in the Berlin Museum of Decorative Art and in the Kennedy collection
in Boston. A rug in the Kaiser Friedrich Museum in Berlin (Fig. 89) which is dupli
cated in a few other collections, belongs unmistakably to the same group, as is
proved by the typical coloring and other characteristics. But at the same time its
pattern shows a slight relation to that of several groups of Asia Minor rugs, which
we have already discussed. Each of the stars in the center is completely sur
rounded by severely conventionalized small cypresses, rosaces or flowers. As an
exceptional instance, this rug has a border which is entirely different from the field,
except in coloring. It shows two palmetto flowers between severe arabesques in reci
procal arrangement. Similar patterns are found in paintings from the first half of
the 16 th century. We may cite as an example the altar painting by Lorenzo Lotto
in S. Giovanni e Paolo in Venice (about 1530) and another almost contemporaneous
one by Torbido in the Academy of Vienna. Previously this class of rugs was ascribed
to Morocco on the basis of the similarity of their geometrical pattern to the mural
decorations of many Moorish buildings and to many other products of the arts and
crafts of Western Islam. This relationship is in reality rather superficial and does
not give sufficient grounds for attributing the rugs to Morocco, which never had
an art of its own. Furthermore, in the 16th century this country was not in a
condition for an art of such high technical and decorative development. In fact,
Morocco was always dependent upon Spain, where rugs were knotted in the 14th
and 16th centuries. Such rugs were also geometrical in character, but entirely
different in all other respects and there are still fewer analogies in the wretched
modern examples of Moorish knotting. A second theory which has enjoyed a certain
popularity among dealers would without any logical reason establish a parallel between
these rugs and certain Eastern Nomad rugs and would defend the theory of their
origin in the region of Samarkand with arguments which are not even worth dis
cussing. The frequent appearance of Damascus rugs in Italy and their absence in Spain
points to the fact that they originated in the Levant. As the Persian origin of these
rugs is obviously impossible, Asia -Minor, Egypt and Syria come into considera
tion. In favor of the latter country we may bring forward the fact that these rugs
are apparently identical with the highly valued "tappeti damaschini", which were
frequently noted in the inventories of Venetian families in the 16 th century and
were especially mentioned as table covers. For this reason the designation "Damas
cus rugs" has gradually come into favor. This hypothesis has at any rate far more
probability than the two others mentioned above, although we know nothing about
a rug industry in Syria. But Damascus might have been a trade centre for rugs
of this type. F. Sarro has recently proposed a new solutfon to the problem. He
thinks an Egyptian origin can be established from the similarity in the composition
of these rugs with the Cairo book bindings and from the fact, mentioned by Turkish
historians, that Egyptian carpet weavers were transferred to Constantinople. This
theory is also supported by the records of European travelers who sometimes mention
rug knotting establishments in Cairo. However, we are not quite convinced of it.
But this theory has also the advantage that it takes account of another group of
rugs, which by their material, technique and coloring appear clearly to be a later
phase in the evolution of the Damascus rugs. And on the other hand, the elements
of their decoration have to be assigned to the Turkish Empire. Indeed, this group
shows exactly the same, partly naturalistic, partly elaborately conventionalized floral
decoration which is characteristic of the so-called Damascus and Rhodos faience

56 Bode-Kuhnel, Antique Rugs from the near east.
of the 16 th— 17 th centuries. These were all probably made in Nicaea and
not partly in Damascus, as was originally supposed. Therefore the rugs related
to them may have originated in Asia Minor. They show an unmistakably Persian
influence, as does all Turkish decorative art of the period. However, the Persian
influence only gave the impetus toward the development of very characteristic
motives of their own, which were limited to strictly floral elements. The leaf and
flower motives of this new style are particularly luxuriant in several prayer rugs
belonging to the Austrian Court, to Prince Liechtenstein, and to the Berlin and
London museums of decorative art (Fig. 90). This type is also represented by several
so-called table rugs of square shape. There are specimens of this type in the Vic
toria and Albert Museum; one is represented in a painting by Boniface Pitati in the
Academy of Venice; another is in the possession of Mr. von Dirksen (Fig. 91). In
the older Damascus pattern. This is undeniable proof for the development of one
type from the other. (See the example in the Kaiser Friedrich Museum). As a rule
the latter ones the central medallion often has a geometrical form, in the style of
in this second type a yellowish color is added to the others, which retain exactly
their old shadings. A change in the color scheme only appears when the decora
tion becomes broader, less coherent and less unified. The use of the cloud motif
in the shape of lightning contributes materially to this effect. This motif, already
familiar in other types of rugs, is now used in more vigorous form than before. At
his point a rug of comparatively subdued design and coloring in the Kaiser Friedrich
Museum (Fig. 92), forms a transitional type to the rugs with exceedingly vivid
designs. Among the latter, particularly beautiful examples may be mentioned in the
London, Paris, Frankfort and Leipzig Museums of Decorative Art (Fig. 93). The
collection in Cologne, on the other hand, includes an example in which there is a
recurrence of the type of composition with central medallion (in this instance
showing arabesque palmettoes) and four corner motives. The rest of the field is
covered with small clouds in close arrangement, while the border bears beautiful,
fully opened blossoms and vines (Fig. 93).
The skilled design and technical finish of both types of these rugs find parallels
only in the Sefevi period of Persia. We still think that they represent examples of
a court manufactory of the Turkish sultan. This manufactory probably lasted from the
end of the 15th until the middle of the 17th century. In all likelihood it was in Anatolia,
not far from Constantinople. As the knotting industry was never really established
in Syria proper, the designation "tappeti Damaschini" may have been one of those
inexactitudes of regular occurrence in dealings between the Orient and the Occi
dent. The fact that this was probably a court manufactory offers a most feasible
explanation for the gross contrast between these rugs and the other rugs of Asia
Minor which belong decidedly to peasant art. It also explains the sudden transition
from the so-called geometrical kaleidoscopic style to naturalistic flower motives. Un
less a final confirmation is found for Mr. Sarre's theory which assumes an Egyptian
phase, succeeded by a later evolution in Turkey.
Several prayer rugs which are dated about 1700 show the influence of the last exam
ples of the so-called Damascus pattern. Theses rug are chiefly characterized by the promi
nence of the niche, which is usually red. Their spirited design has nothing in common
with the severe contours of the Mihrab rugs of the Ushak group, and the border
decoration, in particular, is reminiscent of the "Damascus" type. Despite these
characteristics this class begins to show signs of decadence. Examples of this group
are in the Berlin Museum of Decorative Art (Fig. 95), the Sarre Collection, etc. The
greater number of the prayer rugs made in Asia Minor since the middle of the

The So-called Damascus-Rugs. 57

18 th century have arisen from this last stage in the development of the classical
rug style. These include the Giordes, Kula, Ladik groups etc. each of which con
tain many varieties. However, their decorative motives as a whole show only
degenerations of earlier patterns, arbitrarily composed and mutilated.
In Asiatic Turkey, about 1600, a hitherto unknown form of composition was
adopted. This took the form of a manifold, so-called family prayer rug, produced
by means of a row of niches covering the field, one of the few earlier examples
is in the National Museum in Munich. The details of the design are as unin
spired as this new pattern itself. They are dragged together from all possible Per
sian, Indian and Turkish models and then arranged with doubtful taste. This proves
that the weakening of creative imagination and the appearance of eclectic taste
produced the decadence of the industry.

Concluding remarks.

The different groups of rugs which have been classified in the preceding chapters,
far from exhaust the stock of antique examples preserved. Even among the
known originals and upon the reproductions in paintings there are a number of
rugs which cannot be definitely placed in any of the categories which we have
established. These rugs are not discussed, since they have very few typical charac
teristics. Several, even among the earlier examples, show the decadence of a certain
pattern; others are occasional departures from an established type. A certain
wavering of the design reveals- that still others are timid compromises between
different patterns. A severe adherence to the aesthetic principles of style and tech
nique are general characteristics of Oriental rug manufacture until its complete deca
dence in the middle of the 18 th century. But even in the comparatively short
period which we are fairly well able to survey there was transformation after trans
formation. Their beginning may be considered as a sympton of decadence.
So far we have only the barest evidence for a history of the evolution of the
different patterns. Today we still lack the most important requirement for this
history: — a more exact knowledge of the districts in which the different groups
of rugs were manufactured. Nor can this knowledge be extensively advanced in
the future, since there are practically no specimens preserved in the place of their
manufacture, and everything which might yield information is destroyed hastily and
pitilessly in the complete demoralization of Near Eastern affairs. Whatever speci
mens are occasionally saved for the European market go through the hands of
speculators who are often more concerned with obscuring than with clarifying the
sources from which they draw their stock. The modern rugs, too, furnish practi
cally no evidence, although their exact localization is frequently possible. And even
these modern specimens cannot always be placed, despite the amazing certitude
with which every dealer unhesitatingly classifies every rug. For, the decline of the
rug industry was not so much technical, as aesthetic, and produced such a deterio
ration and mixture of patterns that rugs now recognized as characteristic products
of this or that province are seldom entirely an outgrowth of old traditions. Indeed,
frequently such patterns are not a creation of the people at all, but consist of mo
tives which they have borrowed and then unintelligently developed into a more or
less profitable article of peasant craft.
F. R. Martin, in his valuable history of rugs, attempted to establish more de
finitely the localization of the older rugs, and especially of those in the Persian
group. He did this by consulting all accessible sources. But although many of his
hypotheses may be clear and striking, yet as a rule, they seldom lead to valid
conclusions, and in many cases they have already' been proved erroneous. For this
reason we must be content for the moment to differentiate only the larger Near

Bode-Kiihnel, Antique Rugs from the near east. 59
Eastern groups. If we wish to subdivide, we must use the different motives of the
decoration as a basis. A stylistic and historical connection then results automatically
where it clearly exists: — for instance, the relation between the medallion rugs
with animals and those without them, or the connection between the vase rugs and
the so-called Armenian patterns. But with our present knowledge we could not
honestly attempt a distribution of the different types of animal rugs among the
other classes of Persian rugs.
It is evident that our survey of the material on hand was necessarily very
general and limited. As a foundation for it we had to use trustworthy criteria and
ihe reliable dates which we secured. The Sefevi period was, of course, the basis
of our investigation, because it is only period which we know fairly well from the
number of examples preserved and from the evidence at hand for ascertaining their
origin. The Persian rugs of this period are imposing and beautiful in their com
position, and are usually characterized by their decoration with floral motives. In
their design and composition a thoroughly realistic perception vies with a fine instinct
for style. The animal forms and also the occasional human figures which blend
with the decoration usually appear as a subordinate feature, or if we may use the
expression, as a kind of insert in the floral decoration of the field. In the beginning
of this period the design and its execution still showed an almost severe strength
and occasionally even a certain thinness of form. But about the middle of the
16th century these same patterns attain a soft abundance, perfect taste in com
position, rhythm of movement, imaginative conception and magnificent coloring. The
intermingling of Chinese motives is the dominant feature of this prime of Oriental
rug manufacture. These Chinese motives, in addition to other elements of decoration,
also found their way to India, where at the same period the technique reached a
similar climax. With the decline of the Sefevi rule in the 17 th century the Persian
designs began to be simpler. The animal figures disappear from the plant decoration
which becomes still more luxuriant, although harsher in color and coarser. On
Turkish soil near Constantinople, the classical Persian decoration seems to have
gone through its last vigorous and individual transformations. This was in the
17 th century. A little later it was more and more misinterpreted in the prayer
rugs, and finally, diluted with senseless degenerations, it floods the entire territory
of the Near East.
In the 15th century, in the northwestern border territory of Armenia and the
Caucasus, the Persian manufacture, properly speaking, was preceded by an industry
obviously connected with the former. The patterns of this manufactory show a
very similar form of composition, and they also have floral decoration as the pre
dominating feature, but the design is stiffer and tends to harden into geometrical
formations. If we trace this development still further back, the forms are created
in an entirely different spirit. This new type dates from about the 14 th century
and is found most frequently in reproductions on paintings. The representation of
animals comes to the fore; in fact, they form the dominant motif in the decoration.
They are used not merely to enliven an elaborate floral decoration, as in the silk
rugs of the Sefevi. Large and isolated, they become the leading motif of the com
position. They stand out, either singly or in pairs, against the strong, plain back
ground of the meagerly framed compartments and are repeated in regular rows.
The angular conventionalized design of the figures has an effect similar to that of
heraldic animals and it is evident that they originally had a symbolic or religious
meaning. Such rugs can be traced from the end of the 13 th until the beginning of
the 15 th century and we can state with comparative certainty that they originated

60 Concluding remarks.

in Asia Minor, where the animal designs seem to have followed a geometrical
linear style which we may justly regard as the oldest decorative principle in the
entire knotting technique. The earliest authenticated use of this type of decoration
in combination with Cufic writing is in the 13th century. It then influenced the
floral decoration of Asia Minor. Here it appears most prominently in the so-called
Holbein rugs of the 15th— 17th centuries, in which the plant forms, especially, can no
longer be recognized. It is shown most characteristically in the kaleidoscopic design
of the so-called Damascus rugs, and least in the Ushak group. The latter were sufficiently
guarded against the threatening conventionalization by their connection with Persia.
In recent years early rug designs have been copied in the Orient, some as a
whole, some in part. Yet both types of these reproductions are usually superficial.
The early originals may be detected by their general appearance and by the finer
knotting and more beautiful and magnificent coloring, but they are chiefly distinguished
by a number of apparently unimportant details, which in the modern rugs are either
entirely lacking or unintelligently rendered. When one has steeped himself ,for a
time in the classical rug designs it is comparatively easy to recognize the actual
imitations by all kinds of slight mistakes and inconsistencies. As in other branches of
art, the modern rugs never copy a known original faithfully, but borrow their motives
from several related models, and in so doing commit errors against the old laws of style.
In the following summary the characteristics of the designs are again briefly stated.
The Persian rugs from the Sefevi period are the most easily recognizable
type. For instance the peculiar Far Eastern motives of which they show a larger
or smaller number, are characteristic of them alone and are completely lacking in
late rugs. These motives include not only the mythical Chinese creatures — the
phoenix, the Kylin and the dragon, combined with such symbolical animals as deer,
cranes, etc. The symbol of immortality, the Tschi, is also very frequent, whereas
the emblem of Buddhism, the triple globes or Tsdiintamani, is more rarely used.
There can be no doubt concerning the meaning and origin of the Tschi, which in
the form of the "cloud band" plays such an important part in most of the early
rugs. In several specimens this motif is quite as purely conventionalized and diversely
formed as in contemporaneous Chinese art. During the entire 16th, and even in the
first half of the 17th century, not only rugs of the most diverse types, but miniatures,
faience, bronzes and textiles show this motif in the most varied ornamental trans
formations. The prayer rug which is reproduced in color shows the cloud band in
particularly beautiful form and yet with a trace of early austerity. The large beautiful
clouds take up almost the entire field. We find similar forms in early monuments
with Far Eastern and Islamic influences, and they are especially frequent in Italian
textiles, where they are generally surrounded by rays. In Italy such motives may
be traced back to the 14th century. The characteristic scroll work (consisting of
small rolled up ribbons) which appears so frequently in different types of rugs from
the early 16 th century, is also perhaps due to the stimulus of the Chinese decora
tive arts. The globe motives or the Tsdiintamani, with or without so so-called
lightning is far rarer. Like the Tschi, this motif even reached Asia Minor
where it was particularly popular in the later groups of the so-called Damascus
rugs and sometimes appeared in very decoratively conventionalized form. It is
obviously identical with a motif which is frequently the main design in various
Oriental textiles, especially those of the 14th century. This consists of three half
moons set one against the other, with small spheres in the openings, the whole
repeated between flame motives. This design is very similar to the coat of arms
which the Strozzi family of Florence has born since the 14th century. We have

Bode-KQhnel, Antique Rugs from the near east. 61
already mentioned the magnificent silk rug of the Poldi Pezzoli Museum in Milan
as an example of the use of a Chinese altar with a canopy-like superstructure in
the shape of a bat Thus far, this is the only authenticated instance of its use.
In this period the purely Persian motives are as individual and as beautifully
formed as the Chinese ones. Such rugs show a careful balance in the composition
of the field and in the relation between the various elements of the pattern, and
every motif reveals a fine sense of style and is developed in the most manifold
ways. When discussing several of the most magnificent specimens, we pointed out
the great variety of animals, quadrupeds, as well as birds which all belong to the
Persian fauna and are so skillfully reproduced that they reveal an exact study of
nature. Combined with these animals, we find masks of lions and panthers and
sometimes of jackals. They are a favorite motif of the early 16th century.
On the other hand, the artists seem to have hesitated to use human figures.
Indeed, it is only in the hunting rugs that they play a larger role, although occasionally
they also appear in other ' silk rugs as angels or genii [in true Persian form. But
in the latter case they are completely subordinated to the main composition from
the standpoint of decoration.
The striking feature of all these rugs is their floral decoration in which two
main forms again predominate. These are the fully opened palmetto flower and
the arabesque leaf. Both are often designed with architectonic severity; both are
characteristic ancient symbols from the Saracenic stock of ornamentation. The
palmetto, which in textiles is frequently called the pomegranate, seems to have
originated from the water lily or lotus flower. Since the beginnings of art this
flower was native to the Nile and Tigris as well as to the great rivers of the Far
East and for thousands of years it has been a vital and productive motif in Asiatic
design. It may be alternately in full bloom or closed like a bud, large or small,
in side view or seen from above. In all these forms this flower is the intrinsic
basis of design during the prime of Persian art. Small flowers of different kinds,
usually in the shape of stars, are combined with the palmettoes and used to fill in
the ground. In his "Stilfragen" Riegl has thoroughly explained the origin of the
arabesque as a creation of Mohammadan art from the undulated vines of late Roman
art. The arabesque gave the rug knotter a fine opportunity to show his mastery
of design in the most elegant linear composition, despite all technical difficulties.
The arabesque is generally used in the border as a frame on either side of the fully
opened flowers. Both of these motives are frequently found in the center field in
similar arrangement.
The botanical determination of the flowers or shrubs which are found in the
earlier rugs in combination with the palmettoes, is difficult or entirely impossible
because of their conventionalization. The smaller blossoms are similar to those of
the apple or almond tree. The more vigorously conventionalized flowers, especially
the tulips, carnations and hyacinths, which are particulary common in Turkish prayer
rugs of the 17th century, are strikingly similar to the decoration of the so-called
Rhodos faience and to that of many brocades and velvets of the same period.
Therefore, at that time they seem to have been a general motif in the decorative
art of Asia Minor. We may recall that the almost complete geometrical conven
tionalization of all plant motives is a characteristic for most of the types of rugs
from this territory. This goes so far that the plants are scarcely recognizable. We
have only one example of such an evolution in Persia, and even this is not entirely
authenticated. This is the pointed, so-called reciprocal lily pattern which was used
as a border motif, especially in the "Polish" rugs.

62 Concluding remarks.

Only a few patterns of the Sefevi period show actual inscriptions. These
almost always take the form of the flowing Persian Talik style. The vertical,
rounded Tsuluts characters appear more rarely, while the archaic, angular Cufic is
practically never used.
The modern products, if they have any originality, no longer show any trace
of Chinese motives, while the purely Persian or general Mohammadan motives, if
used at all, are only a misinterpretation of the past. That is all that remains of
the varied elements which appeared so numerously and in rhythmic balance in
the rugs of the 16 th and 17 th centuries. For this reason, it is not difficult to
distinguish rugs from the prime of the Persian knotting art with animal and plant
decoration, from the so-called Armenian reproductions from the period around
the year 1700, which show these motives in the first stages of decadence, and yet
in rather successful unity of composition. With the progress of time the signs of
decadence naturally became more evident
This is not true of the older types of rugs the designs of which are based on
tnterlaced ribbons and other geometrical features. This type of decoration is most suited
io knotting technique. It therefore appears in all periods and its simple motives
afford no grip for critical methods. Furthermore, these patterns have been imitated
more or less faithfully in modern times, not only in places where European merchants
ordered copies of old models but also in regions where ancient traditions had
been actually preserved in their purity and had not come in contact with other
spheres of artistic production. Such modern rugs therefore show very characteristic
signs of the earlier periods; — interlaced and rolled motives, hook-like ornaments,
hexagonal or octagonal fields with staff-like motives, battlement-like decoration in
the borders, letter forms, etc. Despite their separation in time, the ornaments from
both the recent and early rugs are quite frequently identical. Even the knotting is
the same, as the older examples show a rather loose, careless technique, similar to
that in the better Nomad rugs. The arrangement of the single motives, the relation
of the center field to the border or other criteria of style, are also of no avail, as
in this respect, the early period too is arbitrary. This is shown by the fact that
the borders are sometims very small, while in other cases they are very wide or
appear in double or triple number. In such cases one may best avoid error by
keeping the colors of the classic rugs as a standard of judgment The old spe
cimens appear in such pure and harmonious shadings that their contrast with all
later rugs (to say nothing of those with aniline dyes), is sufficiently striking evi
dence for anyone who has once felt the coloristic beauty which characterizes,
above all else, the primitive rugs of the East.

Notes by the Translator,

Note 1 (Page 13): At the present day the most important collections of rugs in the United
States are those of the Metropolitan Museum of Art in New York, including the former Alt-
man and Fletcher Collections, the Museum of Fine Arts in Boston and the Museum of Art in
Cleveland. Among private collections we may mention those of James F. Ballard of St. Louis,
the late Henry C. Frick of Pittsburg, James Deering of Miami, Fla., John D. Mcllhenny of
Philadelphia, George Hewitt Myers of Washington, D. C, Mrs. Herbert L. Pratt of Brooklyn,
Dr, Denman Ross of Cambridge, P. M. Sharpies of Westchester, Pa., P. A. B. Widener of
Philadelphia and Senator Clark of New York. The important collection of the late Mrs. Phoebe
A, Hearst is now in the San Francisco Museum, while that of the Theodore M. Davis Estate
is in the Metropolitan Museum of Art in New York.
Note 2 (Page 14): For descriptions of the knotting technique American readers may con
sult the handbook of Walter A. Hawley, "Oriental Rugs, Antique and Modern", and that of
G. Griffin Lewis, "A Practical Book of Oriental Rugs".
Note 3 (Page 14): As far as our knowledge goes, the woolen thread which is to form
the pile, is never cut in short pieces which are inserted between the warp threads. The
knotted thread is only cut from the hank after the knotting operation is finished.
Note 4 (Page 15): To our knowledge, no coats of arms in the European sense are used
by the Chinese. Dragons as well as phoenixes are extremely frequent symbolic motives in
Chinese art, but to our knowledge no representation of the dragon fighting with the phoenix
exists in Chinese art. This composition seems to be an absolutely Near Eastern motif, al-
hough its components are shaped after Chinese models. Considering this Near Eastern origin
of the composition, we might perhaps better call it the fight of the simurg and the dragon.
Note 5 (Page 16); This rug is now in the Musee des Arts Decoratifs. The representations
are well known scenes from >the "Khamse" of Nizami. The scene in the row next to the
border represents Khosrau seeing Shirin bathing. The next row shows hunters on horseback,
which may or may not be an illustration of one of the numerous hunting scenes of
the Persian poet. The third row illustrates two scenes of Medjnun and Leila: Leila
traveling on a dromedary to visit Medjnun in the desert, and the death of Medjnun
in the arms of Leila. The fourth row again shows hunters on horseback. The rug
is cut off at the fifth row. This row represents the fight of the phoenix and the dragon. The
entire phoenix is preserved, but we see only the head of the dragon; the remainder is missing.
It is possible that there is less than one half of the rug preserved. The rug may have con
tained ten rows of representations, the missing five identical to the preserved ones, but facing
in opposite directions. This rug is not identical with another rug in the Maciet Collection,
now also in the Musee des Arts Decoratifs in Paris, of which the other half is in the Cathe
dral of Cracow. (The latter reproduced in Martin, "Oriental Carpets".)
Note 6 (Page 16): This rug, which was No. 221 of the Yerkes Collection, was bought
by J. B: Trevor at the sale.
Note 7 (Page 18): This rug was No. 218 of the Yerkes Collection. We were not able to
ascertain its present owner.
Note 8 (Page 19): The Altman Collection in the Metropolitan Museum of Art in New
York contains a silk rug, the border of which is identical with the Berlin rug. However, the
center field is identical with that of the silk rug formerly in the Aynard Collection and now
belonging to D. K. Kelekian, New York. (See R. M. Riefstahl "Three Silk Rugs in the Alt-
man Collection", Art in America, Vol. IV, pages 147—161).
Note 9 (Page 19): This rug now belongs to Mr. Kelekian, New York.

64 Bode-KQhnel, Antique Rugs from the near east.
Note 10 (Page 19): The rug formerly in the Yerkes CoUection has been bequested to the
Metropolitan Museum of Art by Isaac D. Fletcher (reproduced in Martin , Fig. 127 and in the
Yerkes Catalogue No 206; Metropolitan photo No 366S8). The mate of this rug passed from
the collection of Stefano Bardini into the hands of Alexander Smith Cochran who in 1908
gave it to the Metropolitan Museum of Art. (Reproduced in "Early Oriental Rugs", Exhibition
of the Metropolitan Museum of Art, 1910—11, No 32; Metropolitan photo No 3504.) The rug
in Lyons seems to be identical with the two Metropolitan rugs. This is the first case known
to the translator in which an important animal rug has been preserved in three identical
examples. Note 11 (Page 20): There are two rugs of this type, one in the Kaiser Friedrich Museum in
Berlin, of which the corner pieces are cut off, the other now in the collection of Mr. Clarence
H. Mackay, New York, and formerly in the collection of Mr. Marsden J. Perry in Providence,
R. I. There is no rug of this type in Windsor Castle, but the present rug was used in the
coronation ceremonies of Edward VII in Westminster Abbey, having been borrowed from a
former British owner by his Majesty's Office of Works. Mr. A. F. Kendrick of the Victoria and
Albert Museum in London was kind enough to give this information and to confirm to the
translator that there is no rug of this type in Windsor Castle, and that it is the custom to
obtain tapestries, carpets and draperies of all kinds from all possible sources for coronation
ceremonies in Westminster. The rug belonging to Mr. Mackay is complete and will soon be
the subject of a special publication by the translator. This rug shows the motif under dis
cussion in complete form. There is one angel standing, the other flying up and holding a
bottle with a tall neck. Certain details of the design show Chinese influence, but on the whole
these figures are not more Chinese than the numerous representations of angels in black and
white Persian or Turkestan brush drawings, which have been preserved in considerable number.
(See specimens in F. R.Martin, "Miniature Painting" and Ph. W. Schultz, "Miniaturmalerei".)
Note 12 (Page 22): A mate of this rug is in the Musee des Tissus in Lyons.
Note 13 (Page 22): See Note 4.
Note 14 (Page 26): The rug formerly in the Yerkes Collection is now in that of J. Selig-
mann. The Altman Collection in the Metropolitan Museum of Art in New York and the
Collection of Mr. C. F. Williams in Norristown, Pa., include fine specimens of this group.
Note 15 (Page 26): Another very fine rug of this type came with the Altman Collection
into the Metropolitan^Museum of Art. Particularly interesting is the half of a rug of the same
type in the collection of Mr. C. F. Williams of Norristown, Pa.i which may help in placing this
group of rugs. The miniature figures of two men on horseback, the one seeming to be the,
prisoner of the other, are inserted in the middle of the fine spiral and floral design. The abbre
viated design of these figures reminds one of the human and animal figures on later Cau
casian rugs.
Note 16 (Page 26): There were two identical rugs in the mosque of Ardebil. Both were
somewhat damaged. The better preserved one of the two, was repaired with pieces from the
second rug, and is now in the Victoria and Albert Museum in London. The second rug was
a good deal reduced in size through repair and damage and was provided with a border
taken from another and slightly later Persian rug. As it is, this second rug is still of remarkable
beauty. It passed from the Yerkes Collection into that of Captain De Lamar, after whose
death it was purchased by Duveen Brothers.
Note 17 (Page 28): Two very fine prayer rugs of this type are now in the Metropolitan
Museum in New York. One of them formerly belonged to Bardini in Florence. (See Martin
"Oriental Carpets", Fig. 203.) The other passed from the Kelekian Collection into that of
Isaac D. Fletcher who gave it as bequest to the Metropolitan Museum.
Note 18 (Page 29): One of these rugs is now in the James F. Ballard Collection in
St. Louis, Mo.
Note 19 (Page 30): The rugs of the Theodore M. Davis Estate are now on exhibition at
the Metropolitan Museum of Art in New York, among them the garden rug.
Note 20 (Page 31): The translator calls these rugs "Polish rugs", not "Polonaise rugs".
"Polonaise" is an American mispronunciation of the French word "Polonais" in "tapis polonais",
which means in plain English "Polish rugs".

Notes by the Translator. 65

Note 21 (Page 32): The most important collection of Polish rugs is now that of Senator
Clark of New York. Fine specimens are in the Metropolitan Museum in New York and in the
Museum of Fine Arts in Boston.
Note 22 (Page 32): Mr. A. F. Kendrick is kind enough to inform the translator that there
is no Polish rug with a coat of arms in the Victoria and Albert Museum. The silk rug
formerly in the Salting Collection (Catalogue No 678, Vienna Oriental Carpets, Plate 96), con
tains two central motives which look like coats of arms, but which are not. The Mobilier
National of the French Republic possesses a very large silk rug with a coat of arms in the
center. The translator was not able to identify this coat of arms, even with competent
heraldic assistance.
Note 23 (Page 33): There is no Persian miniature painting known to the translator in
which tapestry weaves are represented as wall hangings. On the other hand, woolen tapestries
(Kilims) are used all over the Near East as floor coverings, particularly in summer. There are
huge Kilim weaves preserved of rectangular size, which correspond to typical sizes of Oriental
floor rugs, but which would not be suitable for wall decorations on account of their oblong
shape. All kinds of fabrics, brocades and embroideries are, and have been used as floor cover
ings in the Near East, as the miniature paintings prove.
Note 24 (Page 34): The Metropolitan Museum has no rug of this type. There is an
Armenian rug with similar lozenge composition, but without the characteristic corner com
positions, and of entirely different weave.
Note 25 (Page 35): Another fragment of the same rug is in the Kouchakji Collection in
New York. .
Note 26 (Page 36): This rug was No 202 of the Yerkes Sale. We are not able to as
certain its present owner.
Note 27 (Page 38): See Note 4.
Note 28 (Page 38): The translator does not know what became of the Bardini rugs. He
knows four rugs of this type in the United States. One belongs to Mr. James F. Ballard in
St. Louis, Mo., the second to Mr. C. F. Williams, who acquired the rug from a personal friend;
the third to Mr. John D. Mc Ilhenny of Philadelphia, Pa., and the fourth to Mr. P. M. Sharpies
of Westchester, Pa. The two last mentioned rugs were acquired in Constantinople, as
Mr. Williams was kind enough to inform the tranlator.
Note 29 (Page 39): There are two fragments of the primitive rug, one in the Kaiser
Friedrich Museum in Berlin, the other in the Joseph Lees Williams Memorial Collection in the
Metropolitan Museum in New York. Mr. C. F. Williams informs the translator that the rug of
Mr. M. K. W. Vanderbilt is of a later date and so is the rug belonging to Mr. Lamm of
Naesby, Sweden, which was exhibited in 1910 in Munich.
Note 30 (Page 48): Such rugs with the niche design on both ends are called "hearth rugs"
or "odjalik" in order to distinguish them from the "namazlik" or prayer rugs.
Note 31 (Page 54): This rug passed into the hand of the antiquarian Sangiorgi in Rome.

w

*3*- f;;\#*4 --v*-^ :-i%v ¦ v^"-'.'^'^4'~. »;^^4.
4 ^l£* >V % Wi&h^it' 4 -^^^4#fe^^4 Wc
^-*s .-v-' 3JT<*- - wtv — -<to-^ .peTI'" ^ Vflpt -»4^i, .¦ / -'¦¦l
^"^^•.jwD; fe + '- ~^yMf:, \ y\

; ftS" *a"r7S)r , 
^¦?$4^~ '*'

r&ni

\&m

¦•¦"'

9rKiif i " W ' >"; ' : ' '' *V ^ W a- ' ' ' ' ' " 4 < ' ' '

"m.C"Y~' ' ";' ?!

Fig. 1.

Silk Hunting Rug formerly in the Possession of the Austrian Court. 5
Bode-Kuhnel, Vorderasiatisdie KnOpfteppidie.

Fig. 2.
Border of No 1.

Fig. 3.
Silk Hunting Rug from the Collection of Baron Maurice de Rothschild, Paris.

Fig. 4.
Hunting Rug in the Mus6e des Arts decoratifs, Paris.

Fig. 5.
Animal Rug in the Poldi-Pezzoli Museum, Milan.

Fig. 6.
Animal Rug in the Stieglitz Museum, Petrograd.

Fig. 7.
Detail of No 6.

Fig. 8.
Animal Rug in the Musee des Arts decoratifs, Paris.

Fig. 9.
Silk Animal Rug in the Museum of Decorative Art, Berlin.

Fig. 10.
Silk Rug in the Musee des Arts decoratifs, Paris.

Fig. 11.
Woollen Animal Rug in the Collection of Count Boucqoi, Vienna.

Fig. 12.
Persian Animal Rug in Wool in the Kaiser Friedrich Museum, Berlin.

Fig. 13.
Woollen Animal Rug in the Collection of Prince Sdiwarzenberg, Vienna.

Fig. 14.
Woollen Animal Rug in the Victoria and Albert Museum, London.

00

§ Ti

2 Ol
O •

' tW .--^A

5 :': •^?4lJr1*'fe "^YYYY"^.
1^

ST -Si

... rn^PW%
^^44;4444

sc^ss

ItHH

YYY. I i

Fig. 16.
Detail from No 14.

Fig. 17.
Detail from No

14.

Bode-Kiihnel, Vorderasiatisdie Kniipfteppidie.

Fig. 18.
Woollen Rug with Cartouche Pattern in the Metropolian Museum, New York.

Fig. 19.
Wollen Rug with Figures and Cartouch Pattern in the Collection of the Duke of Buccleuch.

Fig. 20.
Woollen Rug with Cartouche Pattern in the Collection of Countess Clam-Gallas, Vienna.

Fig. 21.
Woollen Animal Rug in the Metropolitan Museum, New York.

Fig. 22.
Persian Animal Rug in the Austrian Museum for Art and Industry, Vienna.

4 :

'' '&£%<¦"

>4?f

: -j

iWfc

-v;
LH-
" t-b4j v
H)y Ji^i v ~'st .ft
• V,* Vi'- .'¦i'Y&«il'
Fig. 23.
Persian Woollen Rug formerly in the Collection of Charles T. Yerkes, New York.
Fig. 24.
Persian Woollen Rug in the Collection of the Manufacture des Gobelins, Paris.

Fig. 25.
Persian Woollen Rug with Central Medallion.

#4*4*liifH

gw^v*

4C P? tz^HV-jM^^Skff'

Y^'fy^u--M^X % 4AWK|l
ftrtwwtw

a

- •¦- -

: 44v 4 y< V l&w*^"^^:??^^ . * 4 ' 79 * '. $£ .

4<£'

•-;-*•

. ¦¦¦v,\V ¦>>. Hr^ffiwfr iwtiBtiffMr

Fig. 26.
Persian Woollen Rug from Ardebil, dated 1539, in the Victoria and Albert Museum, London.

Fig. 27.
Detail of No 26.

Fig. 28.
Persian Silk Rug in the Collection of the Manufacture des Gobelins, Paris.

i¥mmmM

V'&(Yw<&

:*

]hYf^^

4 Lar! ^^yj^^Tu

Fig. 29.
Persian Silk Rug in the Collection of Baron Albert de Rothschild, Vienna.

Fig. 30.
Persian Woollen Rug in the Collection of Baron H. von Tudier, Munidi.

Fig. 31.
Persian Wollen Rug formerly in the Collection of Dr. von Dirksen, Berlin.

:

Fig. 32.
Persian Woollen Rug in the Collection of Senator Clark, New York.

•#• ix ¦ '& ,- iU

Fig. 33.
So-called Herat Rug in the Victoria ar.d Albert Museum, London.

Bode-Kiihnel, Vorderasiatisdie Kniipfteppidie.

Fig. 34.
So-called Vase Rug in the Austrian Museum for Art and Industrie, Vienna.

Fig. 35.
So-called Vase Rug in the Imperial Ottoman Museum, Constantinople.

Fig. 36.
So-called Vase Rug in the Collection of Mr. W. von Ginzkey, Vienna.

Fig. 37.
Persian Woollen Rug with Garden Pattern in the Collection of Dr. Albeit Figdor, Vienna.

Fig. 38.
So-called Garden Rug in the Kaiser Friedridi Museum, Berlin.

Fig. 39.
Persian Woollen Rug with Tree Pattern in the Kaiser Friedridi Museum, Berlin.

Fig. 40.
Persian Woollen Rug with Tree Pattern from the Collection of Mr. C. F. Williams, Norristown, Pa.

.W4>.

%$$ *0$*™

Fig. 41.
So-called Polish Rug with Coat of Arms in the National Museum, Munich.

'ii '] II iii' •>'*'; .ft1.,' I :' ' , .'. (Sl/ii'l . tV I 'J' if'' i
Mm iiiiiiipil

Hii .',>:,". w. | s;li

i 's.

CM 4.

' M ilii ,

ll

'? rr '-,! . "

Fig. 42.
So-called Polish Rug in Silk formerly in the Robinson Collection, London.

¦B^W^-'aSw!

Fig. 43.

So-called Polish Rug in Silk, from the Collection of Prince Johann Liechtenstein, Vienna.

Fig. 44.
So-called Polish Rug in Silk, from the Collection of Prince Johann Lieditenstein, Vienna.

Fig. 45.
So-called Polish Rug in Silk in the Provincial Museum, Hannover.

Fig. 46.
So-called Polish Rug in Silk in the Residence Palace, Munich.

Fig. 47.
So-called Polish Tapestry Carpet in the Collection of Dr. Figdor, Vienna.

Fig. 48.
So-called Polish Tapestry Carpet in the Kaiser Friedridi Museum, Berlin.

Fig. 49.
Indian Animal Rug formerly in the Collection of M. Jeuniette, Paris.

Bode-Kuhnel, Vorderasiatisdie Knupfteppidie.

M

*-^"J^"''w^P^*',''-'';/-''''''' ¦¦ :¦"¦•¦'¦¦ ', :-"' - -"-- ¦¦-,¦'- m&f&w "

Fig. 50.
Indian Woollen Rug in the Austrian Museum for Art and Industry, Vienna.

Fig. 51.
Indian Wollen Rug in the Museum of Fine Arts, Boston.

Fig. 52.
Indian Prayer Rug in the Austrian Museum for Art and Industry, Vienna.

Fig. 53.
Indian Silk Rug in the Musee des Arts decoratifs, Paris.

Fig. 54.
Indian Woollen Rug with Floral Decoration in the Austrian Museum for Art and Industry, Vienna

Fig. 55.
So-called Armenian Animal Rug in the Kaiser Friedrich Museum, Berlin.

Fig. 56.
So-called Armenian Animal Rug in the Collection of Mr. P. M. Sharpies, Westchester, Pa.

Fig. 57.
So-called Armenian Rug in Wool in the Collection of Mr. Williams, Norristown, Pa.

S'i)

. Fig. 58.
Late Animal Rug in the Kaiser Friedridi Museum, Berlin.

Fig. 59.
Late Garden Rug in the Collection of Mr. Lamm, Naesby, Sweden.

MhSS

Fig. 60.

Fig. 61.

Fig. 62.
Seljuk Rugs from the Mosque Ala ed- din,' Konia, Asia Minor.

Fig. 63.
Early Woollen Rug with Phoenix and Dragon Pattern in the Kaiser Friedridi Museum, Berlin.

Fig. 64.
Rug in the „Wedding of the Foundlings" by Domenico di Bartolo.

Fig. 65.
Rug in the „Betrothal of Mary" by Nic. di Buonaccorso in the National Gallery, London.

Fig. 66.
Rug from a „Madonna" by Lippo Memmi in the Berlin Gallery.

Fig. 67.
So-called Ushak Rug formerly in the Collection of Count Gregor Stroganoff.

Fig. 68.
So-called Ushak Rug with Star Pattern in the Collection of Dr. Bode, Berlin.

Fig. 69.
So-called Ushak Rug, dated 1584, in the Collection of the Lord of Buccleuch.

Bode-Kiihnel, Vorderasiatisdie Kniipfteppidie.

Fig. 70.
So-called Ushak Rug in the Kaiser Friedrich Museum, Berlin.

Fig. 71.
So-called Ushak Rug in the Austrian Museum for Art and Industry, Vienna.

Fig. 72.
So-called Ushak Rug in the Collection of Baron H. von Tudier, Munich.

Fig. 73.
So-called Ushak Rug in the Kaiser Friedrich Museum, Berlin.

Fig. 74.
Asia Minor Woollen Rug in the Kaiser Friedridi Museum, Berlin.

Fig. 76.
Asia Minor Prayer Rug in the Hungarian Museum of ^Decorative Art, Budapest.

JZ--
»'#^i
*V*.#
km*.
-*,*-
- J
5»^pil >>i*%!L >M>^b»l KwRfB
f>:x|^f^#Ji^
^f'^'^^y *V ****** ^ ****
Fig. 78.
Asia Minor Woollen Rug (with white ground) in the Collection of Dr. Bode, Berlin.
Fig. 79.
Asia Minor Woollen Rug (with white ground)

in the Collection of Dr. Bode, Berlin.

¦-<©.

Fig. 82.
Asia Minor Woollen Rug with so-called Holbein Pattern in the Kaiser Friedrich Museum, Berlin.

Fig. 83.
Asia Minor Woollen Rug with so-called Holbein Pattern in the Kaiser Friedrich Museum, Berlin.

Fig. 84.
Asia Minor Woollen Rug with so-called Holbein Pattern in the Collection of Dr. Bode.

ej^JXSS0**

Fig. 85.
Asia Minor Woollen Rug with so-called Holbein Pattern in the Collection of Dr. Bode, Berlin.

Fig. 86.
Asia Minor Woollen Rug with Geometrical Pattern in the Kaiser Friedrich Museum, Berlin.

Fig. 87.
So-called Damascus Rug in the Kaiser Friedrich Museum, Berlin.

Fig. 88.
So-called Damascus Rug in Silk formely in the Possession of the Austrian Court.
Bode-Kiihnel, Vorderasiatisdie Kniipfteppidie.

Fig. 89.
So-called Damascus Rug in the Kaiser Friedrich Museum, Berlin.

Fig. 90.
Turkish Prayer Rug in the Museum of Decorative Art, Berlin.

jSlJlg^K^JSJTi-' t-J^JJ' .-'6 fo™s*in?

Fig. 91.
Turkish Rug from the Collection of Dr. W. von Dirksen, Berlin.

Fig. 92.
Turkish Rug in the Kaiser Friedridi Museum, Berlin.

Fig. Ij.
Turkish Rug in the Museui ' of Decorative Art, Leipzig.

Fig. 94.
Turkish Rug in the Museum of Decorative Art, Cologne.

Fig. 95

Turkish Prayer Rug in the Museum of Decorative Art, Berlin.

00483 9271

/y

s'k

nH

p^"fC