CHAPTER XVII

PURCHASING RUGS

S those who have expert knowledge of the value of Oriental rugs are exceedingly few, compared with those who admire and wish to own them, the object of this chapter is to make suggestions regarding a proper selection, when purchasing, and to point out some of the pitfalls that beset the inexperienced.

In estimating the value of any rug, three distinct qualities are to be considered: rarity, artistic beauty, and utility.

Rarity may depend on the age of a rug, the locality where it was woven, or its type. In determining the age, which as a rule is greatly exaggerated, a number of facts should be considered. One is the condition resulting from wear; though at times this is misleading, since of two pieces, one may have been handled with almost religious solicitude and the other exposed to the elements and to hard usage. Moreover, an artificial appearance of natural wear is sometimes counterfeited. Another is the pattern, since, as has been shown in the case of antique carpets, the character of drawing changed with succeeding periods. Even when the patterns of old carpets are copied in modern pieces, a distinction is usually discernible to a careful observer. The colours, too, play an important part in determining age; for not only are some characteristic of different periods, as particular shades of yellow and green of Chinese rugs and the blue of Persian; but the mellowing influences of time, acting through the agencies of exposure and wear, cause effects that cannot be produced by any artificial process.

Occasionally the time when a rug is made is recorded in Arabic numbers woven above a word denoting “year.” If they are indistinct on account of the length of the nap, they may be more clearly read in reverse order at the back. In the following lines, each of them is represented below the one in our own notation, to which it corresponds.

These numbers represent not the Christian but the Mohammedan year, which, dating from the time of the Hegira, began about the middle of July 622, or a little before the actual flight from Medina. It should also be remembered that the lunar and not the solar year is considered in Moslem chronology; which, according to our reckoning, gains about one year in every thirty-three and seven tenths years. To calculate, then, the year of our time corresponding with the year expressed in the rug, from the number should be subtracted one thirty-three and seven tenths part of itself, and then should be added six hundred and twenty-two. Thus, if the year 1247 was woven in the rug, our corresponding year would be a. d. 1247 less 37 (or 1210) plus 622, or 1832. When such dates appear in old rugs, they are generally to be depended on, but in modern ones they are more likely to be antedated to give the effect of greater age.

With reference to the time when woven, rugs may conveniently be divided into three broad classes; Antique rugs or carpets made over two centuries ago; old rugs made fifty or more years ago; and modern rugs made since the introduction of aniline dyes, or within the last fifty years.

The number of antique carpets that exist is undetermined, as it is impossible to estimate how many remain in Oriental mosques and palaces. Nor has any complete catalogue been made of those that are owned in Europe and America. They consist principally of the products of Persia, Asia Minor, Armenia, and China. Many belong to the museums and the remainder to sovereigns and wealthy collectors. Like rare porcelains and old paintings, their value increases with each passing year; and the prices received for them range according to the fancy and caprice of the purchaser. They are the most valued and the most costly of all rugs.

Plate 66. Kurdish Prayer Kilim

The number of rugs over fifty years of age, but not belonging to the previous class, is very large. A few are sufficiently prized to be placed in art museums, some belong to collectors, large numbers embellish the halls and drawing rooms of people of refinement, and others are yearly brought from the Orient; but dealers and the public already realise that their numbers are limited. Even now they are searched for in the remotest corners of every rug-producing country; and in a few years the last, now cherished as family heirlooms, will have been exchanged for western gold. Almost all are well woven, though some are too much worn to be trodden longer under foot. None are treated with aniline dyes, but the colours mellowed by time are exceptionally good and frequently contain rare tones characteristic of the first class. Moreover, large numbers contain emblems of a symbolism still shrouded in mysteries that increase their fascination. On the whole, this is the choicest stock from which to choose elegant carpetings for luxurious homes. When it is considered that rugs of this class are beginning to disappear from the market, the prices at which they may be purchased are moderate compared with the prices of more modern pieces.

The great majority of existing Oriental rugs have been woven within the last fifty years. A few of them, including many of the newest, have colours that compare favourably with those of older pieces; but a large number show the effect of aniline dyes. Some woven by nomads or dwellers in remote villages, without thought of sale, have designs and workmanship such as have characterised the fabrics of these people for past generations; but others, which are the products of the work-house system, though well woven as a rule, lack the charm of spontaneous individuality. Age alone has little influence in determining the value of these modern rugs, since they have not yet become rare; yet even in them the wear of time affects their other qualities. Other things being equal, they cost less than the old and the antique rugs.

The locality where a rug was woven is also to be considered in determining its rarity and therefore its value. Of the countless carpets that once existed in Egypt, of the very early rugs of Caucasia and Turkestan, not a piece remains; but if one were to be found it would be almost priceless. The antique carpets of Syria, or of Kirman, Shiraz, and Tabriz, woven over three centuries ago, are more valuable than others of equally good workmanship, of which relatively large numbers remain. So, too, of the rugs classed as old, but falling short of the venerable age of the real antiques, those which are now difficult to be obtained on account of their scarcity, are more valuable than those which are being produced in larger numbers. Rugs such as the Joshaghan, Tiflis, and many others of sixty or more years of age, are no longer woven. Modern products from the same districts may adopt the old names, but they are not the same. Accordingly, it will be only a short time when they too will disappear from the market. Good examples of such pieces should therefore receive more careful consideration on the part of purchasers and collectors, as their value is increasing with each passing year.

Furthermore, the rarity of an old rug is often independent of its age or the locality where it was made, and is due to its peculiar type. For instance, the Ming Rugs of China with silver threads and the so-called Garden Carpets of Iran represent types rather than localities. Likewise the Hunting Carpets of Persia, the Holbein Carpets of Asia Minor, and the Dragon Carpets of Armenia, represent, as far as we know, the textile craft of no well-defined district of limited area, as is the case of modern rugs, but rather rare types. Such pieces are valuable, not alone on account of their age, but also because they represent these rare types.

A rug is also valued for its artistic beauty. The innumerable rugs which centuries ago were in daily use soon disappeared, and only those intended for palaces or temples have been preserved. It is but natural, then, that the antique carpets representing the highest art of their time should be not only rare but also beautiful. Yet even in them is often a distinction that affects their value. Fortunately, very many of the larger number of rugs of less age, but classed as old, likewise possess artistic beauty. This chiefly depends on the drawing and the colouring.

It will be noticed that almost without exception careful drawing accompanies workmanship of a high class. This is partly due to the facts that the more excellent the weave the easier it is to clearly define patterns; and that on shortness of nap, which as a rule is found in closely woven rugs, depends accuracy of delineation. The charm of rugs often depends, also, on the graceful flow of lines, the careful balance of different parts of patterns, and the proper co-ordination between border and field. Careful attention should accordingly be given to the drawing when selecting a rug.

The artistic beauty of Oriental rugs depends still more on the colouring, since, as has been elsewhere expressed, drawing, which is intellectual, finds its highest development in the Occident, and colouring, which is sensuous, finds its highest development in the Orient. It at once suggests sumptuous luxury. In all of the antique carpets that remain and in very many rugs over fifty years of age, all the colours employed in a single piece are in tones of perfect harmony, and are so placed with reference to one another that the effect is most agreeable. But in some of the modern pieces, such as are produced in parts of Asia Minor and Caucasia, are colours which, like discordant notes of music, grate harshly on the senses. The most pleasing effect is when colours of border and field are complementary, yet so in harmony as to accentuate the qualities of each.

There are also colours which, independent of their association, are in themselves good or bad. The best are found in the antique carpets woven when the art of the dyer was an honourable profession. The colours are also very good in still later pieces; but for a century now some of the finest have not been used, and even the secret of producing them has been lost. Here and there dyers and weavers cling to early traditions, so that among modern rugs are many examples of good colouring; but the most recent pieces, excepting when softened by artificial processes, often display harsh and garish colours. This distinction is in a measure due to the fact that old colours were largely produced by vegetable dyes and the modern are too often produced by aniline. Not infrequently both vegetable and aniline colours are used in the same piece, and sometimes the quantity of aniline colour is so small that it is scarcely objectionable; but as a rule it is best never to purchase a rug that is so tainted.

One objection to the use of aniline dyes is that by removing some of the natural oil of the wool they are apt to make it brittle, so that it is less able to stand wear. Another is that in time some of the dyes, which have been applied collectively to produce a single colour, will fade or even disappear, so that the final colour may be a most undesirable shade not in harmony with those that surround it. If the fibres are brittle and become harsher to the touch when wet with water, it is an indication that aniline dyes have been used. Another test is the application of weak vegetable acids, which will make the colour spread if produced by aniline dyes, but are not likely to affect it if produced by vegetable dyes. Many native weavers can distinguish by placing the wool in their mouths, when they experience a sweet or bitter taste, according as vegetable or aniline dyes have been used. It is a mistake, however, to assume that the dyes are aniline because the wool has a brighter colour at the surface of the nap than at the foundation; or because the colour spreads when wet with water; since in time even some of the vegetable colours will fade; and when fresh they will run during the first washing in water, but afterwards they are little affected either by water or weak acids.

Even when the same colours and the same kind of dyes were used, there is a marked distinction in the appearance of old and of recently woven rugs, which is due to wear as well as exposure to sun and weather. The effect of time, imperceptible at first, is shown in rich tones of remarkable softness and beauty, that add greatly to the value of a rug. It accordingly happens that artificial processes are adopted to create as far as possible the same results without the lapse of time. Some of these are as novel as were the efforts of the distinguished viceroy of King-te Chin, in the reign of Kang-hi, to produce antique porcelains.[40] Henry Savage Landor says[41] that “to manufacture ‘Antique Carpets’ is one of the most lucrative branches of modern Persian carpet making. The new carpets are spread in the bazar in the middle of the street, where it is most crowded, and trampled upon for days or weeks, according to the days required, foot passengers and their donkeys, mules and camels making a point of treading on them in order to ‘add to age’ in the manufacturer’s goods. When sufficiently worn down the carpet is removed, brushed, and ordinarily sold for double or treble the actual price, owing to its antiquity.”

COLOUR PLATE XI—CHINESE RUG

This large Chinese carpet represents some of the best workmanship of the Keen-lung period. In it are shown the graceful drawing of leaf, fruit, flower, and butterfly, and the dainty colouring of blue, yellow, brown, and apricot on a field of ivory that are so characteristic of this time. The usual balance of designs throughout the field is maintained with precision; but, as is not always the case, different motives occupy corresponding positions. Thus a cluster of leaves and fruit may be balanced with a cluster of leaves and flowers. The conventional drawing of the corners and the somewhat formal panel that surrounds the central medallion give to the pattern strength of character while they detract nothing from its beauty.

Loaned by Mr. Nathan Bentz

Whatever may be their character, the methods employed to give softened effects to the colours are known as “washing.” Most of those in vogue in the Orient, such as washing with lime water, do little real injury. In this country to artificially mellow the colours has become a regular business of firms, who guard the secret of their different methods. Some use ammonia, borax, and soap, which also do very little injury to the rug. Others use chloride of lime, boracic acid, vinegar, or oxalic acid, that remove some of the natural oils of the wool and accordingly impair its qualities for wear. In fact, pieces are occasionally injured to the extent that the wool has become brittle and may readily be plucked out. Nevertheless, it does not necessarily follow that all rugs washed with an acid solution have been seriously injured; but the colours never have the same richness as those which have been softened by natural processes operating for a long period of years. To be sure, rugs that have been washed are often more attractive than they were in their raw colours; but the older, more beautiful rugs with genuine tones mellowed by time are always to be preferred. Over ninety per cent of the Kermanshahs, Sarouks, Kashans, Tabriz, Muskabads, Mahals, and Gorevans, and a large percentage of all other modern rugs sold in this country, have been treated by some artificial process to soften their colours or give them the appearance of age. It is generally necessary, when selecting a large rug for a floor covering, to accept a washed piece; but when a smaller rug or a runner will meet the requirements, it is preferable to choose the older unwashed piece, which, as a rule, is more beautiful and costs but little more. In the case of most pieces, the tones of colour are sufficient to enable one who is experienced to distinguish between those that are artificially aged and those that are not. In the case of others, a simple test is to rub them thoroughly with a wet rag; when, if acid or chloride of lime has been used, it can generally be detected by the odour.

The artistic beauty of a rug also depends somewhat on the fineness of the nap; as the soft, floccy fibres of some wools acquire a velvety appearance, or give to the colours a sheen and a lustre compared with which other rugs look harsh and coarse. For instance, the rugs of Shiraz and Meshed, the Beluchistans, and many Bokharas are noted for the lustre of their colours; but on the other hand many of the rugs of Asia Minor and Caucasia have colours that are without lustre, and the rugs of India which are made of dead or “Chunam” wool, seem lifeless.

When selecting any rug, then, the purchaser should carefully observe if the patterns are well drawn and their different parts show a proper balance. He should observe if the colour tones are harmonious with one another, if each colour in itself is good, and if they have been softened by natural processes acting for a long time. And he should notice if the wool is coarse, dead, and lustreless, or if it has a sheen and glint in the light of day; for these are the qualities that make up the artistic beauty of a rug.

The utility to be derived from rugs that properly belong to museums and collectors receives small consideration, though even with them the more perfect their condition the more valuable they are. But in case of the great majority of rugs, which are intended for use as well as for ornament, their utility is an important consideration to the purchaser. Rugs that have warp and weft of strong yarn and a close firm texture, will wear better than others. Also, such rugs as Bijars, in which one thread of warp to each knot is doubled under the other, will be found to wear better than such rugs as Mosuls which have each thread of warp equally prominent at the back. For durability, long nap is also to be preferred to short, since it protects the foundations of the knots from wearing and becoming loose.

Before purchasing an old rug, it should be spread on the floor to see if it lies flat and if its shape is regular. It should be examined by daylight and not by electric light, which gives a false impression of colour and sheen. It should be held up with the back turned to the purchaser, and carefully examined for weak spots through which the light may pass; since, when so held, many pieces which seem in good condition when lying on the floor, resemble a sieve. The foundation threads should also be carefully inspected, as sometimes they rot and will tear with slight tension. Moreover, as the selvage or overcasting of the sides and the webs of the ends are intended primarily not for ornament but for protection, it should be noted if they are in good condition. Sometimes the webs of the ends are entirely gone, so that continual fraying of the nap is prevented with difficulty. Sometimes the selvage or overcasting of the sides is broken and some of the threads of warp are injured. Or the sides may be well protected by a stout overcasting; but on examination it will be seen that it is not the original finishing, and that some of the border has disappeared. Again, it may have been overcast too tightly, so that the sides curl and turn under, and thus expose the border to injury when trodden on. Careful examination will often reveal surprises. In many old rugs the field is full of rents, that have been sewn together; in others entire pieces have been removed, so that they are no longer of their original length; or parts of the border are gone, or even the whole of it has been replaced by the border of another rug; yet all so deftly done that the changes are scarcely noticeable.

Nevertheless, old pieces, if otherwise meritorious, are not to be discarded on account of a few imperfections, since what can be accomplished in the hands of a careful repairer is remarkable. Broken threads of warp and weft can be mended; missing knots can be replaced with others of similar yarn; crooked pieces can be straightened by loosening here and stretching there; borders that curl can be flattened by removing the yarn and overcasting again more carefully. In fact, if the nap be not so worn that the foundation of warp and weft is exposed, it is far better to choose an old rug with some rents than a new one with garish aniline dyes. Nor should a piece be slighted, because the brownish black areas of wool dyed with iron pyrites are worn low; since often the most beautiful effects are obtained by a surface of brighter colours standing out in relief, on account of the worn blackish nap that surrounds it. Now and then a bargain can be had by buying a rug which, because of some imperfection that is not serious, has been passed by; and now and then a piece reeking with dirt has proved, when properly cleansed, to be a gem.

The foundation, consisting of the warp and weft, receives but little consideration from purchasers; yet it is one of the most important indices of the quality of a rug, and its strength is one of the most necessary conditions for utility. The warp is best observed at the ends. In most Chinese and Indian rugs and in some of the Persian, it is of cotton; in others it is of wool or goat’s hair. In the Chinese rugs the diameter of the threads of warp is much smaller than the diameter of the threads of the weft, and has little strength, but in almost all other rugs it is at least as large and as stout. The weft may readily be observed at the back. In very many of the best rugs, it consists of fine spun wool; but in many modern ones, it is of coarse wool or cotton. The number of knots to the square inch does not of itself demonstrate the quality of texture, since a rug may have only a few knots of coarse diameter and be firmly woven, or it may have many knots of fine diameter and be loosely woven; but in the same class the better rug has generally more knots than a poorer one. When selecting a rug, then, the back should be most carefully observed; for here may be seen if the yarn that forms the knots is well spun, if the knots themselves are drawn tight and well pressed down, and if the threads of weft are carefully inserted and have a texture that indicates fine workmanship. Almost invariably it will be found that if the back of a rug shows good material, and has an appearance of firmness and skilful, painstaking weaving, the front will correspond with good colours and careful drawing.

The value of antique carpets, which depends to some extent on their size, and to a much greater extent on their rarity and character, is constantly increasing; for the reason that their number is limited and each year they are more highly appreciated. It is, therefore, impossible to affix even approximate prices; but the sums paid at the Yerkes sale in 1910, when some thirty pieces were sold at auction for an average of about $9,400, will serve as a guide. The following are some of the pieces sold and the prices realised:

Carpet, size 7 feet 4 inches by 5 feet 4 inches, attributed to Western Persia, at end of XVI Century, and purchased by the Metropolitan Museum of Art, N. Y. $5,600
Persian carpet, XVI Century, described on page 86 5,600
Polish silk carpet, XVI Century, size 6 feet 6 inches by 4 feet 7 inches 4,700
Polish silk carpet, XVI Century, size 6 feet 6 inches by 4 feet 7 inches 3,500
Polish silk carpet, XVI Century, size 6 feet 11 inches by 4 feet 10 inches 12,300
Silk carpet, XVI Century, size 7 feet 2½ inches by 6 feet 5 inches, stated to have belonged to the Ardebil Mosque 35,500
Moorish carpet, XVI Century, size 10 feet 11 inches by 5 feet 10 inches, stated to have belonged to the Ardebil Mosque and purchased by the Metropolitan Museum of Art, N. Y. 15,200
Hispano Moresque Mosque carpet, size 34 feet 5 inches by 16 feet 8 inches, flat stitch 8,600
Carpet attributed to Western Iran, size 16 feet 4 inches by 11 feet 2 inches, and purchased by the Metropolitan Museum of Art, N. Y. 19,600
Carpet similar to the Mosque carpet of Ardebil, XVI Century, size 23 feet 11 inches by 13 feet 5 inches 27,000

The value of rugs over fifty years of age but not sufficiently old to belong to the antique class also depends as much on the technique of weave, drawing, colouring, and rarity as on the size; yet even this must be taken into consideration. In proportion to their size the most expensive of these rugs are the Kirmans, Sehnas, and Niris from Persia; the Ghiordes and Ladiks from Asia Minor; the Daghestans and Kabistans from Caucasia; and the Royal Bokharas and Yomuds from Central Asia. Considering both utility and attractiveness the least expensive are probably the Sarabends and Mosuls from the Persian group, the Yuruks from Asia Minor, Kazaks and Tcherkess from Caucasia, and Afghans and Beluchistans from Central Asia. As is the case with antique carpets, the prices of all old rugs in good condition are steadily advancing.

There is likewise a tendency for the prices of modern rugs to increase with each year, since on account of the gradual opening of Oriental countries to the markets of the world, and the greater demand for rugs, the wages of weavers are increasing. Some of them, as the Tabriz, Gorevans, Kermanshahs, Muskabads, Mahals, Sarouks, and Kashans, are now made almost exclusively under the direction of the work-house system, and are sold at prices that fluctuate but slightly. But in a short time the prices of all of them will doubtless be higher.

When a rug of carpet size is required, the Kermanshahs are generally preferred on account of their soft colouring and refined patterns, that harmonise with the furnishings of most reception rooms. Less expensive and more showy are the Gorevans, which are suitable for halls or dining rooms. In the Afghans, which are splendid rugs for a den, are combined durability with a moderate price. Within recent years some of the Indian rugs, as the Amritsars and Lahores, have been growing in favour, as they not only have good colours, artistic patterns, and exceedingly good texture, but are reasonable in price. Of smaller rugs required both for ornament and use, the Shiraz, Feraghan, Mosul, Bergamo, Tcherkess, Bokhara, and Beluchistan are desirable.

As is the case with other works of art, so much deception can be practised in the sale of rugs that a purchaser cannot use too much circumspection. Sometimes through ignorance or with intention, a dealer will declare that the wool of a rug which has been coloured with aniline dyes has been coloured with vegetable dyes only; that a rug washed with acid has matured naturally; that a new rug which has been artificially worn almost to the knot is an antique; or that a particular rug belongs to the class desired, as where a Shirvan is offered for a Shiraz or a Bijar for a Bergamo, which ordinarily are worth much more. It is, accordingly, discreet to buy only of such firms as have a reputation which is above reproach; and if for any reason it is difficult to learn the standing of a firm, the purchaser would do well to make an effort to test its reliability by inquiring about the qualities of some class of rugs with which he is familiar before purchasing others; and if there appears to be any intention to deceive, he should at once look elsewhere. In any event, he should take a guarantee that the rug purchased is as represented. Firms that have gained an honourable reputation by honest dealing deserve the patronage of the public, and will always be found ready to make restitution if any mistake has been made.

At times, the best rugs may be bought at auctions and at the fairest prices. Auctions such as the Yerkes, where estates are being closed or where firms are dissolved, occasionally occur, when every opportunity is given the purchaser to thoroughly examine in advance pieces which are sold without reservation to the highest bidder. On such occasions, rare pieces are sometimes bought at very moderate prices. But as a rule, unless the purchaser is a good judge and has previously carefully examined a coveted piece in broad daylight, it is better not to buy at auctions. During the sale it is impossible to properly examine a rug. The glare of electric light thrown upon it gives a too favourable impression of its beauty. The competitive bids of other real or fictitious purchasers and the words of the auctioneer too often lead beyond the dictates of good judgment. At such times one would do well to remember the old words caveat emptor.

FOOTNOTES

[1] The Yerkes sale.

[2] Stewart Dix, in “Arts of Old Japan.”

[3] The influence of the physical aspects of a country on its art as expressed in architecture is nowhere more clearly shown than in Egypt, and there is little doubt that, likewise, the character of the native rugs was influenced by the spirit of the sluggish Nile and the boundless desert wastes. But as Egypt long ago ceased to be a rug-producing country, and none of its ancient rugs remain, it will only be briefly referred to in this work, though symbolic designs which had their origin there during the Caliphate or even earlier were adopted by foreign weavers and occasionally appear with modified form in modern rugs.

[4] It is said that he carried Persian weavers as captives to Asia Minor and Constantinople.

[5] This is a product of flowers of the genus Delphinum that grows in the Himalayas. It is also obtained as a powder from Afghanistan.

[6] In a few rare instances a knot is tied to four threads of warp.

[7] Most Sehna knots are right-hand knots, but the Sehna knots of a large proportion of Khorassan rugs are left-hand knots.

[8] As far as the writer is aware, no one has hitherto called attention to the many precise distinctions there are in weaving, and to the fact that each class of rugs follows a distinct type of its own. For this reason this branch of the subject is treated more fully than would otherwise be necessary.

[9] A few of the weavers about Gozene in Asia Minor make rugs with a double foundation, in which a single thread of coarse weft crosses twice between parallel rows of threads of warp. Only rarely is this method followed in other districts.

[10] Sir George Birdwood has made the statement that “A deep and complicate symbolism, originating in Babylonia and possibly in India, pervades every denomination of Oriental carpet. Thus the carpet itself prefigures space and eternity, and the general pattern or filling, as it is technically termed, the fleeting, finite universe of animated beauty. Every colour has its significance; and the design, whether mythological or natural, human, bestial, or floral, has its hidden meaning. Even the representatives of men hunting wild beasts have their special indications. So have the natural flowers of Persia their symbolism, wherever they are introduced, generally following that of their colours. The very irregularities either in drawing or colouring, to be observed in almost every Oriental carpet, and invariably in Turkoman carpets, are seldom accidental, the usual deliberate intention being to avert the evil eye and insure good luck.”

[11] The equivalent of 106 feet square.

[12] See his work, “The Holy Carpet of the Mosque at Ardebil.”

[13] Ismael reigned, 1502-1524; Tamasp reigned, 1524-1576.

[14] 1586-1628.

[15] These will be considered in a later chapter.

[16] This group includes both antique and modern rugs.

[17] Robert Kerr Porter, a well-known traveller, stated that the floor of the audience hall of the governor at Tabriz, whom he visited in 1818, “was entirely overspread with Herat carpets, those of that manufacture being the richest that can be made.”

[18] By “Type Characteristics” is meant the characteristics of such types of the class as are most frequently seen. There are exceptions to these types.

[19] As this is the case with most rugs, only the exceptions to this feature will be noticed in the type characteristics of other classes.

[20] Ibn Batutah.

[21] Of modern Ispahans.

[22] “Burlington Magazine,” December, 1909.

[23] In “Industrial Arts of India.”

[24] Sidney Churchill in the Imperial Vienna Book says that “the dyes of Sultanabad have perhaps the most extensive colour scheme in Persia.”

[25] “Industrial Arts of India.”

[26] Mrs. Elizabeth Bishop in “Journeys in Persia and Kurdistan.”

[27] Encyclopedia Britannica.

[28] It is also to be noted that within the last few years large numbers of pieces bearing resemblance to old Oriental rugs have been woven about Constantinople.

[29] One of the most interesting is at Sivas, where are the remains of a most beautiful Seljuk gateway, with architectural lines that might well have been taken for a weaver’s model. As in many prayer rugs, the engaged columns support a high arch over which a panel rests above a figured spandrel. The outlines of each of these parts suggest most forcibly the drawing of the prayer rug, and the resemblance is carried even further; for corresponding with the border stripes is the chiselled masonry that once rested above the panel and still extends to the foot of the entrance at each side of the arch.

[30] One of these is represented in Dr. Bode’s “Knupfteppiche,” where it appears as a secondary stripe.

[31] An intermediate pattern suggested by each is found in an old Asia Minor piece owned by Dr. Bode.

[32] A. Bogolubow, in his excellent work “Tapis de l’Asie Centrale,” divides the Transcaspian Turkomans into two principal groups, the Salors and Yomuds, each of which includes sub-groups. These are again divisible into many tribes, almost all of whom weave. As their rugs, though resembling one another, show different characteristics, they might properly be separated into numerous classes; but since only a few of them are known in this country, they alone will be described.

[33] “Desert of Red Sands.”

[34] A tent in which an average of five people live.

[35] In “Industrial Arts of India.”

[36] Dr. Birdwood.

[37] Sometimes Sehna knot.

[38] Sometimes one thread of warp to each knot is doubled under the other.

[39] Rarely Iran, Feraghan, Mosul, and Kurdistan.

[40] It is stated that in the short space of a few weeks he created valuable antique porcelains to present to his noble friends by placing recent copies of old specimens in a vessel containing very greasy soup, where they were duly boiled for a month, and after that placing them in the “foulest drain of the neighbourhood,” where they remained until seasoned.

[41] In “Across Coveted Lands, 1903.”

[42] All references of an unimportant character are indicated by the page number only.

INDEX[42]