Mongolian designs are rarely seen in any of these pieces, which probably represent more closely than any other Persian carpets native art unaffected by foreign influences. Almost all of them are now owned in Europe.

Of the early rugs, those woven in Armenia are far less known than those from Persia. Nevertheless, it may reasonably be a.s.sumed that the high culture that was manifested in Bagdad and Ctesiphon during the sway of the Caliphs was felt among the mountainous districts to the north; and that the Seljukian rulers, who left such artistic monuments in the old Armenian capitals, appreciated and encouraged the manufacture of fine woollen fabrics. In fact, Marco Polo, who travelled through that region during the latter part of the XIII Century, referred to them as being remarkably handsome.

Probably the oldest remaining pieces are the so-called Dragon carpets, which, it is believed, were produced from the XIV to the XVII Century and possibly even earlier. Not infrequently the length is at least twice the breadth; the very narrow border occasionally consists of only a single stripe; and the field is occupied by a trellis-like pattern of narrow, conventionalised leaves, within which are designs containing archaic flowers and dragons. The ground colour of the field is generally some shade of red, that of the border white, and the leaves are yellow, blue, or green. In the borders of many of them appear an S motive from which undoubtedly was derived the design so frequently seen in panels of more recent Asia Minor prayer rugs.

In the Metropolitan Museum of Art of New York is a XV Century carpet (Plate 19, opp. Page 86), which, though widely differing from these pieces in general pattern, so closely resembles them in the essential characteristics of weave and colour that it is unquestionably of the same cla.s.s. The field is occupied by concentric diamonds with stepped sides. The encircling bands, that are mostly red, yellow, and violet, and the corners, that are white, contain numerous archaic forms, including palmettes, trees, birds, and animals. There are also numerous small designs of the tri-cleft leaf so common to the Circa.s.sian and Soumak rugs; and the ray-like edges of the central lozenge, as well as the four palmettes that rest upon it, suggest the origin of the effulgent stars of old Daghestans and Kabistans. An effort has been made to balance similar designs in corresponding parts of the field, though its centre is at one side of the geometric centre of the diamonds. The palmettes show distinctly a strong Persian influence and the animal forms likewise show that it was not woven by a sectarian Sunnite of Western Asia Minor.

Part of a very unusual carpet (Plate 20, opp. Page 88), from a district in Eastern Asia Minor, is in the Kaiser Friedrich Museum at Berlin. Its princ.i.p.al interest lies in the fact that it is very old and that its approximate age has been determined. In the hospital at Siena, Italy, a similar rug is represented in a fresco called the "Wedding of the Foundling," painted by Domenico di Bartolo about the year 1440, so that it is reasonable to conclude that this particular piece was woven not much later. In fact, its character would indicate that it or some other from which it has been copied was much older. Each of the nearly square compartments contain octagons, within which on a yellow field are represented the mythical fight of the dragon and phnix that was adopted as the Ming coat of arms. It is interesting to note that the chain pattern of the brownish-black main border stripe is not unlike what is seen in modern pieces, but the running latch hooks of the corners and the small S designs are unusually stiff. This disposition to formal drawing, which is conspicuous in all parts of the rug, shows an archaic style noticeable only in the very earliest carpets.

In the celebrated painting of Georg Gyze (Plate 21, opp. Page 92) which hangs in the Berlin Gallery, is represented a rug of a cla.s.s so frequently seen in the paintings of Hans Holbein that they are known as "Holbein rugs." Their marked dissimilarity to those previously described indicates that they were woven under different circ.u.mstances if not in different regions. Neither in the fields nor borders is any trace of Mongolian or Persian influences; and the absence of all floral, leaf, and animal forms so usual in most antique carpets is noticeable. Indeed, the fact that animal forms rarely appear in the art of the Sunni Mohammedans aids in determining the place of their origin. They came from Asia Minor or Western Armenia.

It has generally been a.s.sumed that they were woven in Western Asia Minor, because they were purchased there in former centuries and taken thence to Europe; but they possess many features that indicate they may have been woven farther to the east, whence many could easily have been transported westward in caravans. Their borders contain the well-known pattern derived from Cufic letters which, more conventionalised, appears in later years only in such rugs as the Kabistans and Daghestans of Eastern Caucasia. Most of them also contain the small octagonal discs and larger octagonal figures with Greek crosses at the centre that suggest forcibly the designs of Southeastern Caucasia. The narrow stripes of ribbon and chain pattern found in many of them also are very common in Caucasian rugs; so that it seems not improbable that these Holbein rugs were made within the boundaries of that greater Armenia which, embracing the upper Mesopotamian valley, extended over the eastern part of Asia Minor and the southern part of modern Caucasia.

[Ill.u.s.tration: PLATE 21. PORTRAIT OF GEORG GYZE BY HANS HOLBEIN, SHOWING A HOLBEIN RUG WITH CUFIC BORDER]

These rugs claim the attention not only because they have borders of such interesting origin, but by the fact that the age when they were woven is ascertainable. As Holbein lived between the years 1497 and 1543, and some other rugs of this type appear in the works of early Flemish and Italian painters, it may reasonably be a.s.sumed that some of them were made before the end of the XV Century.

A very excellent example of this cla.s.s, owned by Mr. C. F. Williams, is now in the Metropolitan Museum of Art. It has a length of about five feet with a breadth of three and a half. The ground colour of the field is an olive green and that of the main stripe of the border is red. The prevailing colours of the designs, which are entirely geometric, are blue, green, and ivory. All of these rugs are small or of moderate size, and are slightly oblong. Some of them have a ground colour of green; and yellow is frequently found in the pattern. The weaving is rather loose; and compared with Persian rugs they have fewer knots to the square inch.

Another carpet from Asia Minor that also belongs to Mr. C. F. Williams appears in Plate 22, opp. Page 94. It is the only entire rug with this pattern that is known, though a piece of a similar rug is in the Victoria and Albert Museum at London. On fields of blue and red are outlined three large four-pointed stars separated by smaller diamonds.

Within these figures and in the surrounding field is a network of tracery supporting conventionalised leaf and floral forms. Between the field and the main stripe of the narrow border is a close co-ordination of pattern, but the simple ribbon of the inner guard seems alien. It appears without modification in many later Asia Minor and Caucasian rugs.

An important feature are the double knots at the corners of the stars, since they are identical with designs found in a ma.n.u.script made for one of the Shahs in 1435, and thus a.s.sist to determine the age of the rug.

For this reason and on account of its general character, it seems not unreasonable to place it as early as the middle of the XV Century.

Similar carpets were woven during a long period, and it is probable that in the latter half of the following century they were largely influenced by the weavers that Solyman the Magnificent, after capturing Tabriz in 1534, transported to his own country. The same general features still remained, but the detail was more elaborate and ornate. Arabesques, palmettes, and floral forms, both of field and border, resembled more nearly the Iranian character. But at a later period, after the beginning of the general decadence to which every industry and art were subject, the patterns became much simpler, and the colours were reduced almost exclusively to red and blue with a little green. At length, both pattern and colours a.s.sumed the type of modern Oushaks, that by a slow process of devolution originated from these antique pieces.

In Armenia and Asia Minor it is probable that weaving existed before the Christian era, and that the earliest carpets which remain, though affected by more eastern influences, are largely the product of an indigenous art. But in India it was otherwise. It is true that Sir George Birdwood is authority for the statement that the Saracens introduced carpet-weaving there; but it is most probable that at the time of the invasion of the armies of Tamerlane and during the lives of many of his successors, whatever carpets were woven were very crude.

Even when the Moguls began to build and embellish palaces, they obtained their carpets from Persia. But at length Shah Akbar established manufacturies at Lah.o.r.e about the year 1580, and invited Persian weavers to settle there. From them the native workmen acquired much of their knowledge of patterns and technique.

It was during the reign of Shah Jahan (1628-1658), builder of the famous peac.o.c.k throne and Taj Mahal, that most of the choicest pieces that now remain were woven. In delicacy of texture they rival those of any other country, and it is not unusual to find pieces with nearly eight hundred knots to the square inch; moreover, all their designs are depicted with remarkable clearness of definition. One of the most noted of these carpets is the woollen piece, about eight yards long by two and a half wide, that was made at the royal factory at Lah.o.r.e and presented to the Girdlers Company of London in 1634. The mingling of leaf and floral forms, as well as the Herati designs of rosette and crumpled leaf, on a field of red, shows unmistakably its relation to Persian carpets. At the same period were woven large numbers of others with fields covered with an imposing display of superbly drawn flowers, of which every part from root to leaf tips was represented with astonishing realism.

Another cla.s.s included the animal or hunting carpets, which unlike their Persian prototypes seem intended not so much to portray symbolically some historic event or abstract idea, as to convey a correct impression of an actual event.

[Ill.u.s.tration: PLATE 22. OUSHAK CARPET

Loaned by C. F. Williams, Esq., to the Metropolitan Museum of Art, New York]

One of these, a woollen piece with a length of eight and a quarter feet and a breadth of five and a quarter, is in the Boston Museum of Fine Arts. The inspiration was from some old Persian piece, but the rendering is peculiarly Indian. In this representation of an Oriental jungle is a strange mingling of the real and unreal. The struggle of a monster bird with a winged beast, half lion, half elephant, and the demoniac faces of the border suggest the inspiration of early pagan mythology; but the movements of the running gazelles and the stealthily creeping tiger, the att.i.tude of the driver of the cart and his attendant, are most natural.

The drawing as a whole is exceedingly delicate. The ground colour of the field is the red of most Ispahans and Herats of this period, but the border is a cream colour, a combination not in accord with Persian tradition. The other colours are fawn, blue, pink, grey and brown. It is probably the only Indian hunting carpet of its kind.

Few strictly antique carpets from other countries of the Orient are known. Of the innumerable pieces that were surely woven in Caucasia and Western Turkestan before the end of the XVII Century, scarcely a vestige can be found. Nor are there many from the looms of Syria, though in the days of the Caliphs every mosque was adorned with magnificent carpets.

It is true a few sterling pieces of Saracenic character, that have been ascribed to the region about Damascus, still exist. There are also a few rare and beautiful pieces that have come to light in China.[15] But of the countless thousands that in almost every country of the Orient once covered floors of palaces and mosques, representing one of the most refined arts, now nearly lost, only an insignificant fraction remains.

[Ill.u.s.tration: CHART INDICATING PERIODS WHEN ANTIQUE CARPETS WERE MADE]

CHAPTER VIII

CLa.s.sIFICATION OF MODERN RUGS

RUGS contribute to the comfort of the nomad more than any other fabric. With them he closes the entrance to his tent or covers the floor and couches on which he sits and sleeps. Thrown over other objects they form the table, made into saddle bags they take the place of trunks. The followers of Islam when at prayer kneel on a rug, and in token of affection spread one over the grave of a friend. To dwellers in cities, also, rugs contribute largely to the comfort and luxury of the home. Indeed, without them the splendour of Oriental life would seem incomplete, since they are the princ.i.p.al furnishings of every house, where stout woven pieces with long pile are spread as floor coverings, and lighter ones are hung as portieres and tapestries. Yet it is in the a.s.sembly or dining hall that the finest rugs are used, though here the most valued are exposed only on great occasions.

In the East a rug receives a particular name according to which of these special purposes it is adapted. The large, almost square piece that is used to cover the centre of the a.s.sembly hall is known as the "Khali;"

and the narrow strips or "runners" that are placed at its sides and ends are known as the "Kenares." It is on the Kenares that the servants are required to walk and the less honoured guests to stand, for they are rarely of such fine quality as the former. Before the divan, that generally surrounds three walls and is covered with fine cloth and velvet, are seats on which are placed carpets called "Sedjadeh." They are nearly twice as long as broad, and since they are of moderate size and excellent quality they are frequently used for many other purposes.

The hearth rug, termed "Odjalik," can generally be distinguished from others, as each end of the field is of triangular shape with the apex at the extremity. However much any of these may be valued, the one that to every worshipper of Allah has the most sacred a.s.sociation is the "Namazlik," or prayer rug, at one end of which is an arch in token of the mosque. At call for prayer the faithful Moslem spreads his rug with arch directed towards Mecca, and kneeling with the palms of his hands at each side of the centre he bows his head till it touches the rug. As the Mohammedans of Persia are unwilling that a Namazlik be trampled by the foot of an infidel, few from there can be bought; but the Mohammedans of other countries are less scrupulous, so that many of the prayer rugs sold in America have been made solely for trade and have never been used in worship. They may be beautiful, but special interest attaches to old pieces of which the well-worn nap shows where the knees of both father and son for over half a century have often pressed. In addition to these are other rugs with technical names, but a cla.s.sification of much greater importance is that which depends on the country or district where they are woven.

When the Oriental rugs first appeared in the market of the United States, they were spoken of as "Turkish," for the reason that importers purchased them from Turkish merchants of Constantinople. But when it became known that they had been taken there by caravans from countries farther to the east, and that large numbers of them came from Persia, the name "Persian," that to the mind of many conveys ideas of splendour, was at once applied; even to-day all cla.s.ses of Oriental rugs are often spoken of as Persian. As objects of ornament or utility, their value is independent of their place of origin; yet it is known that the wool of the nap and the dyes used in some districts are superior to those in others, and that in consequence the beauty of some rugs will improve with age far more than that of others. It is also known that because in certain districts the material of warp and weft, as well as the workmanship, is of a superior quality, the rugs made there will wear better than others. The knowledge, then, of where a rug is made is important in determining the quality and value, which otherwise only a critical examination, that few people are able to make, would show.

[Ill.u.s.tration: PLATE 23. SEHNA RUG]

Furthermore, the knowledge of where a rug is made, suggesting the cla.s.s of people who wove it, adds immeasurably to our interest. When, for instance, we look at an old piece of Kurdish weave with its nomadic designs and s.h.a.ggy nap, on which a Moslem savage as an Apache often rested fully half a century ago, there is called up a picture of the dark-visaged tribesman, fearless and untamed as were his ancestors who contested the march of Xenophon over two thousand years ago. We see him wandering with his flocks over the hills while he watches for a chance to fall upon an unsuspecting stranger. We picture to ourselves the hut of brush upon the mountain side where a slender barbaric girl bends to tie, with wonderful patience, the knots one by one. So if we would enjoy our Oriental rugs, we should know what people made them, and whence and how they journeyed, before they reached our fireside.

At the request of a purchaser the vendor is ever ready to cla.s.sify a rug, but his statements are not always reliable. This is partly due to the fact that even the great importing houses are often deceived.

Throughout Asia Minor, Persia, Turkestan, even farther east, great fairs are regularly held. Here gather the representatives of tribes from far distant quarters to enjoy for a few days or weeks the gay life and abandon of the East while bartering the products of their different crafts. Here come the purchasing agents looking for rugs; and the pieces that may be brought from afar are bought and shipped by camel and rail to such great marts as Tabriz, Tiflis, and Constantinople, where the bales are unpacked and the rugs a.s.sorted, cla.s.sified and labelled, before they are resold to the importing houses of Europe and America.

Thus both in the buying from the itinerant agent of rugs a.s.sembled from different quarters and in the rea.s.sortment at the exporting cities there is frequent opportunity for errors of cla.s.sification.

The characteristics of the different groups and cla.s.ses of rugs are given in later chapters, but it should not be presumed that these are infallible guides to the locality where they were made. Often a ruler, by fostering art, has drawn to his capital artists and artisans from other districts. Thus designs and quality of workmanship characteristic of one district would be adopted in another. So, too, the great caravans that pa.s.s along regular routes eastward and westward, and the annual pilgrimages to Meshed and Mecca, have been most potent influences for the dissemination of designs. Yet taking into consideration the general pattern and smaller designs; the material of warp, weft, and pile; the knot; the dyes; the finish of sides and ends, and the peculiarities of the weave, it is possible with a reasonable amount of certainty to determine in what districts almost all Oriental rugs are woven.

It should be borne in mind, however, that the names by which some of the rugs are known in America are not the same as those by which they are known in Asia. For instance, the rugs made by some of the tribes of the Tekke Khanate are known in the Orient as "Tekkes;" but as the great depot for Turkestan carpets was formerly the city of Bokhara, they are generally known in this country as "Bokharas." On the other hand, there are local distinctions in the eastern countries not known in the western. The accompanying cla.s.sification, therefore, is slightly arbitrary, but should be convenient for reference; since the cla.s.ses represent the cities or districts where are woven the several different kinds, excepting the Chinese, which are divided chronologically. The names of the groups are not in each instance entirely satisfactory, but are probably the best that can be chosen. The fourth group, for example, has frequently been called the "Turkoman;" but as it includes some of the rugs of Afghanistan, and also those of Beluchistan, which is remote from Turkestan, that name is not sufficiently comprehensive. The district where these rugs are made is, strictly speaking, the western and southwestern part of Central Asia; but the term here employed has the authority of some German writers of note. So, too, the rugs of Herat, though it is now a city of Afghanistan, are included with the Persian group; but it should be remembered that Herat, as well as the districts of Mosul and Kurdistan, was once part of the old Persian Empire.

GROUP I. PERSIAN.

(_a_) Khora.s.san district: Herat, Khora.s.san, Meshed.

(_b_) Shiraz district: Ispahan, Kirman, Yezd, Shiraz, Niris.

(_c_) Feraghan district: Feraghan, Hamadan, Kara-Geuz, Bibikabad, Iran, Sarouk, Kashan, Sarabend, Burujird, Sultanabad, Muskabad, Mahal, Jos.h.a.ghan, Gulistan, Teheran.

(_d_) Sehna district, or Adelan province: Sehna, Bijar, Kermanshah, Persian Kurdistan, Karaje.

(_e_) Tabriz district: Tabriz, Gorevan, Bakshis, Serapi, Herez, Suj-Bulak, Karadagh, Afshar.

(_f_) Kurdistan district: Western Kurdistan, Mosul, Gozene.

GROUP II. ASIA MINOR OR TURKISH.

(_a_) West Asia Minor district: Bergamo, Ghiordes, Kulah, Oushak, Ak-Hissar, Demirdji, Kutayah, Smyrna, Melez, Isbarta, Rhodian, Broussa, Hereke.

(_b_) Central Asia Minor district: Konieh, Ladik, Kir-Shehr, Anatolian, Karaman, Sivas, Mudjar, Nigde, Tuzla, Kaisariyeh, Zile, Yuruk.

GROUP III. CAUCASIAN.

(_a_) North Caucasian: Daghestan, Kabistan, Kuba, Derbend, Lesghian, Chichi, Tcherkess.