Chapter II: Part 2

Now, as primitive peoples advance from savagery to barbarism there comes a time in the history of all kinds of textile products at which the natural technical progress of decorative elaboration is interfered with by forces from without the art. This occurs when ideas, symbolic or otherwise, come to be associated with the purely geometric figures, tending to arrest or modify their development, or, again, it occurs when the artist seeks to substitute mythologic subjects for the geometric units. This period cannot be always well defined, as the first steps in this direction are so thoroughly subordinated to the textile forces. Between what may be regarded as purely technical, geometric ornament and ornament recognizably delineative, we find in each group of advanced textile products a series of forms of mixed or uncertain pedigree. These must receive slight attention here.

Fig. 325 represents a large and handsome basket obtained from the Apache. It will be seen that the outline of the figures comprising the principal zone of ornament departs somewhat from the four ruling directions of the textile combination. This was accomplished by increasing the width of the steps in the outline as the dark rays progressed, resulting in curved outlines of eccentric character. This eccentricity, coupled with the very unusual character of the details at the outer extremities of the figures, leads to the surmise that each part of the design is a conventional representation of some life form, a bird, an insect, or perhaps a man.

By the free introduction of such elements textile ornament loses its pristine geometric purity and becomes in a measure degraded. In the more advanced stages of Pueblo art the ornament of nearly all the textiles is pervaded by ideographic characters, generally rude suggestions of life forms, borrowed, perhaps, from mythologic art. This is true of much of the coiled basketry of the Moki Indians. True, many examples occur in which the ancient or indigenous geometric style is preserved, but the majority appear to be more or less modified. In many cases nothing can be learned from a study of the designs themselves, as the particular style of construction is not adapted to realistic expression, and, at best, resemblances to natural forms are very remote. Two examples are given in Figs. 326 and 327. I shall expect, however, when the art of these peoples is better known, to learn to what particular mythic concept these mixed or impure geometric devices refer.

The same is true of other varieties of Pueblo basketry, notably the common decorated wickerware, two specimens of which are given in Figs. 328 and 329. This ware is of the interlaced style, with radially arranged web filaments. Its geometric characters are easily distinguished from those of the coiled ware. Many examples exhibit purely conventional elaboration, the figures being arranged in rays, zones, checkers, and the like. It is to be expected, however, that the normal ornament of this class of products should be greatly interfered with through attempts to introduce extraneous elements, for the peoples have advanced to a stage of culture at which it is usual to attempt the introduction of mythologic representations into all art. Further consideration of this subject will be necessary in the next section of this paper.

The processes of pure geometric elaboration with which this section is mainly concerned can be studied to best advantage in more primitive forms of art.

_Non-essential constructive features._--Now, all the varied effects of color and design described in the preceding paragraphs are obtained without seriously modifying the simple necessary construction, without resorting to the multiple extraordinary devices within easy reach. The development and utilization of the latter class of resources must now receive attention. In the preceding examples, when it was desired to begin a figure in color the normal ground filament was dropped out and a colored one set into its place and made to fill its office while it remained; but we find that in many classes of work the colored elements were added to the essential parts, not substituted for them, although they are usually of use in perfecting the fabric by adding to serviceability as well as to beauty. This is illustrated, for example, by the doubling of one series or of both warp and woof, by the introduction of pile, by wrapping filaments with strands of other colors, or by twisting in feathers. Savage nations in all parts of the world are acquainted with devices of this class and employ them with great freedom. The effects produced often correspond closely to needlework, and the materials employed are often identical in both varieties of execution.

The following examples will serve to illustrate my meaning. The effect seen in Fig. 330 is observed in a small hand wallet obtained in Mexico. The fillets employed appear to be wide, flattened straws of varied colors. In order to avoid the monotony of a plain checker certain of the light fillets are wrapped with thin fillets of dark tint in such a way that when woven the dark color appears in small squares placed diagonally with the fundamental checkers. Additional effects are produced by covering certain portions of the filaments with straws of distinct color, all being woven in with the fabric. By other devices certain parts of the fillets are made to stand out from the surface in sharp points and in ridges, forming geometric figures, either normal or added elements being employed. Another device is shown in Fig. 331. Here a pattern is secured by carrying dark fillets back and forth over the light colored fabric, catching them down at regular intervals during the process of weaving. Again, feathers and other embellishing media are woven in with the woof. Two interesting baskets procured from the Indians of the northwest coast are shown in Figs. 332 and 333. Feathers of brilliant hues are fixed to and woven in with certain of the woof strands, which are treated, in the execution of patterns, just as are ordinary colored threads, care being taken not to destroy the beauty of the feathers in the process. The richly colored feathers lying smoothly in one direction are made to represent various figures necessarily geometric. This simple work is much surpassed, however, by the marvelous feather ornamentation of the Mexicans and Peruvians, of which glowing accounts are given by historians and of which a few meager traces are found in tombs. Much of the feather work of all nations is of the nature of embroidery and will receive attention further on. A very clever device practiced by the northwest coast tribes consists in the use of two woof strands of contrasting colors, one or the other being made to appear on the surface, as the pattern demands.

An example from a higher grade of art will be of value in this connection. The ancient Peruvians resorted to many clever devices for purposes of enrichment. An illustration of the use of extra-constructional means to secure desired ends are given in Figs. 334 and 335. Threads constituting a supplemental warp and woof are carried across the under side of a common piece of fabric, that they may be brought up and woven in here and there to produce figures of contrasting color upon the right side. Fig. 334 shows the right side of the cloth, with the secondary series appearing in the border and central figure only. Fig. 335 illustrates the opposite side and shows the loose hanging, unused portions of the auxiliary series. In such work, when the figures are numerous and occupy a large part of the surface, the fabric is really a double one, having a dual warp and woof. Examples could be multiplied indefinitely, but it will readily be seen from what has been presented that the results of these extraordinary means cannot differ greatly from those legitimately produced by the fundamental filaments alone.

_Superconstructive features._--In reviewing the superconstructive decorative features in the preceding section I classified them somewhat closely by method of execution or application to the fabric, as stitched, inserted, drawn, cut, applied, and appended. It will be seen that, although these devices are to a great extent of the nature of needlework, all cannot be classed under this head.

Before needles came into use the decorative features were inserted and attached in a variety of ways. In open work nothing was needed but the end of the fillet or part inserted; again, in close work, perforations were made as in leather work, and the threads were inserted as are the waxed ends of the shoemaker.

The importance of this class of decorative devices to primitive peoples will be apparent if we but call to mind the work of our own Indian tribes. What a vast deal of attention is paid to those classes of embroideries in which beads, feathers, quills, shells, seeds, teeth, &c., are employed, and to the multitude of novel applications of tassels, fringes, and tinkling pendants. The taste for these things is universal and their relation to the development of esthetic ideas is doubtless very intimate.

Needlework arose in the earliest stages of art and at first was employed in joining parts, such as leaves, skins, and tissues, for various useful purposes, and afterwards in attaching ornaments. In time the attaching media, as exposed in stitches, loops, knots, and the like, being of bright colors, were themselves utilized as embellishment, and margins and apertures were beautified by various bindings and borders, and finally patterns were worked in contrasting colors upon the surfaces of the cloths and other materials of like nature or use.

No other art so constantly and decidedly suggested embellishment and called for the exercise of taste. It was the natural habitat for decoration. It was the field in which technique and taste were most frequently called upon to work hand in hand.

With the growth of culture the art was expanded and perfected, its wonderful capacity for expression leading from mere bindings to pretentious borders, to patterns, to the introduction of ideographs, to the representation of symbols and mythologic subjects, and from these to the delineation of nature, the presentation of historical and purely pictorial scenes.

And now a few words in regard to the character of the work and its bearing upon the geometric system of decoration. As purely constructive ornamentation has already been presented, I will first take up that class of superconstructive work most nearly related to it. In some varieties of basketry certain bindings of the warp and woof are actually left imperfect, with the idea of completing the construction by subsequent processes, the intersections being gone over stitch by stitch and lashed together, the embroidery threads passing in regular order through the openings of the mesh. This process is extremely convenient to the decorator, as changes from one color to another are made without interfering with construction, and the result is of a closely similar character to that reached by working the colors in with warp and woof. In a very close fabric this method cannot be employed, but like results are reached by passing the added filaments beneath the protruding parts of the bindings and, stitch by stitch, covering up the plain fabric, working bright patterns. Fig. 336 is intended to show how this is done. The foundation is of twined work and the decorating fillets are passed under by lifting, with or without a needle. This process is extensively practiced by our west coast tribes, and the results are extremely pleasing. The materials most used are quills and bright colored straws, the foundation fabric being of bark or of rushes. The results in such work are generally geometric, in a way corresponding more or less closely with the ground work combination.

A large class of embroideries are applied by like processes, but without reference to the construction of the foundation fabric, as they are also applied to felt and leather. Again, artificially prepared perforations are used, through which the fillets are passed. The results are much less uniformly geometric than where the fabric is followed; yet the mere adding of the figures, stitch by stitch or part by part, is sufficient to impart a large share of geometricity, as may be seen in the buckskin bead work and in the dentalium and quill work of the Indians.

Feather embroidery was carried to a high degree of perfection by our ancient aborigines, and the results were perhaps the most brilliant of all these wonderful decorations. I have already shown how feathers are woven in with the warp and woof, and may now give a single illustration of the application of feather work to the surfaces of fabrics. Among the beautiful articles recovered from the tombs of Ancon, Peru, are some much decayed specimens of feather work. In our example delicate feathers of red, blue, and yellow hues are applied to the surface of a coarse cotton fabric by first carefully tying them together in rows at regular distances and afterwards stitching them down, as shown in Fig. 337.

The same method is practiced by modern peoples in many parts of the world. Other decorative materials are applied in similar ways by attachment to cords or fillets which are afterwards stitched down. In all this work the geometricity is entirely or nearly uniform with that of the foundation fabrics. Other classes of decoration, drawn work, applique, and the like, are not of great importance in aboriginal art and need no additional attention here, as they have but slight bearing upon the development of design.

Attached or appended ornaments constitute a most important part of decorative resource. They are less subject to the laws of geometricity, being fixed to surfaces and margins without close reference to the web and woof. They include fringes, tassels, and the multitude of appendable objects, natural and artificial, with which primitive races bedeck their garments and utensils. A somewhat detailed study of this class of ornament is given at the end of the preceding section.

_Adventitious features._--Ornament is applied to the surfaces of fabrics by painting and by stamping. These methods of decoration were employed in very early times and probably originated in other branches of art. If the surface features of the textile upon which a design is painted are strongly pronounced, the figures produced with the brush or pencil will tend to follow them, giving a decidedly geometric result. If the surface is smooth the hand is free to follow its natural tendencies, and the results will be analogous in character to designs painted upon pottery, rocks, or skins. In primitive times both the texture of the textiles and the habits of the decorator, acquired in textile work, tended towards the geometric style of delineation, and we find that in work in which the fabric lines are not followed at all the designs are still geometric, and geometric in the same way as are similar designs woven in with the fabric. Illustrations of this are given in the next section.

* * * * *

I have dwelt at sufficient length upon the character and the tendencies of the peculiar system of embellishment that arises within textile art as the necessary outgrowth of technique, and now proceed to explain the relations of this system to associated art.

In the strong forward tendency of the textile system of decoration it has made two conquests of especial importance. In the first place it has subdued and assimilated all those elements of ornament that have happened to enter its realm from without, and in the second place it has imposed its habits and customs upon the decorative systems of all arts with which the textile art has come in contact.

GEOMETRICITY IMPOSED UPON ADOPTED ELEMENTS OF DESIGN.

At a very early stage of culture most peoples manifest decided artistic tendencies, which are revealed in attempts to depict various devices, life forms, and fancies upon the skin and upon the surfaces of utensils, garments, and other articles and objects. The figures are very often decorative in effect and may be of a trivial nature, but very generally such art is serious and pertains to events or superstitions. The devices employed may be purely conventional or geometric, containing no graphic element whatever; but life forms afford the most natural and satisfactory means of recording, conveying, and symbolizing ideas, and hence preponderate largely. Such forms, on account of their intimate relations with the philosophy of the people, are freely embodied in every art suitable to their employment. As already seen, the peculiar character of textile construction places great difficulties in the way of introducing unsymmetric and complex figures like those of natural objects into fabrics. The idea of so employing them may originally have been suggested by the application of designs in color to the woven surfaces or by resemblances between the simpler conventional life form derivatives and the geometric figures indigenous to the art.

At any rate, the idea of introducing life forms into the texture was suggested, and in the course of time a great deal of skill was shown in their delineation, the bolder workmen venturing to employ a wide range of graphic subjects.

Now, if we examine these woven forms with reference to the modifications brought about by the textile surveillance, we find that the figures, as introduced in the cloth, do not at all correspond with those executed by ordinary graphic methods, either in degree of elaboration or in truthfulness of expression. They have a style of their own. Each delineative element upon entering the textile realm is forced into those peculiar conventional outlines imposed by the geometric construction, the character of which has already been dwelt upon at considerable length. We find, however, that the degree of convention is not uniform throughout all fabrics, but that it varies with the refinement of the threads or filaments, the compactness of the mesh, the character of the combination, the graphic skill of the artist, and the tendencies of his mind; yet we observe that through all there is still exhibited a distinct and peculiar geometricity.

So pronounced is this technical bias that delineations of a particular creature--as, for example, a bird--executed by distant and unrelated peoples, are reduced in corresponding styles of fabric to almost identical shapes. This conventionalizing force is further illustrated by the tendency in textile representation to blot out differences of time and culture, so that when a civilized artisan, capable of realistic pictorial delineation of a high order, introduces a figure into a certain form of coarse fabric he arrives at a result almost identical with that reached by the savage using the same, who has no graphic language beyond the rudest outline.

A number of examples may be given illustrating this remarkable power of textile combination over ornament. I select three in which the human figure is presented. One is chosen from Iroquoian art, one from Digger Indian art, and one from the art of the Incas--peoples unequal in grade of culture, isolated geographically, and racially distinct. I have selected specimens in which the parts employed give features of corresponding size, so that comparisons are easily instituted. The example shown in Fig. 338 illustrates a construction peculiar to the wampum belts of the Iroquois and their neighbors, and quite unlike ordinary weaving. It is taken from the middle portion of what is known as the Penn wampum belt. The horizontal series of strands consists of narrow strips of buckskin, through which the opposing series of threads are sewed, holding in place the rows of cylindrical shell beads. Purple beads are employed to develop the figures in a ground of white beads. If the maker of this belt had been required to execute in chalk a drawing depicting brotherly love the results would have been very different.

My second illustration (Fig. 339) is drawn from a superb example of the basketry of the Yokut Indians of California. The two figures form part of a spirally radiating band of ornament, which is shown to good advantage in the small cut. Fig. 340. It is of the coiled style of construction. The design is worked in four colors and the effect is quiet and rich.

Turning southward from California and passing through many strange lands we find ourselves in Peru, and among a class of remains that bespeak a high grade of culture. The inhabitants of Ancon were wonderfully skilled in the textile art, and thousands of handsome examples have been obtained from their ancient tombs. Among these relics are many neat little workbaskets woven from rushes. One of these, now in the National Museum, is encircled by a decorated belt in which are represented seven human figures woven in black filaments upon a brown ground.

The base and rim of the basket are woven in the intertwined combination, but in the decorated belt the style is changed to the plain right angled interlacing, for the reason, no doubt, that this combination was better suited to the development of the intended design. Besides the fundamental series of fillets the weaver resorted to unusual devices in order to secure certain desired results. In the first place the black horizontal series of filaments does not alternate in the simplest way with the brown series, but, where a wide space of the dark color is called for, several of the brown strands are passed over at one step, as in the head and body, and in the wider interspaces the dark strands pass under two or more of the opposing strands. In this way broad areas of color are obtained. It will be observed, however, that the construction is weakened by this modification, and that to remedy the defect two additional extra constructive series of fillets are added. These are of much lighter weight than the main series, that they may not obscure the pattern. Over the dark series they run vertically and over the light obliquely.

It will be seen that the result, notwithstanding all this modification of procedure, is still remarkably like that of the preceding examples, the figures corresponding closely in kind and degree of geometricity.

The fact is that in this coarse work refinement of drawing is absolutely unattainable. It appears that the sharply pronounced steps exhibited in the outlines are due to the great width of the fillets used. With the finer threads employed by most nations of moderate culture the stepped effect need not obtrude itself, for smooth outlines and graceful curves are easily attainable; yet, as a rule, even the finer fabrics continue to exhibit in their decorations the pronounced geometric character seen in ruder forms. I present a striking example of this in Fig. 342, a superb piece of Incarian gobelins, in which a gaily costumed personage is worked upon a dark red ground dotted with symbols and strange devices. The work is executed in brilliant colors and in great detail. But with all the facility afforded for the expression of minutely modulated form the straight lines and sharp angles are still present. The traditions of the art were favorable to great geometricity, and the tendencies of the warp and woof and the shape of the spaces to be filled were decidedly in that direction.

In order that the full force of my remarks may be appreciable to the eye of the reader, I give an additional illustration (Fig. 343). The two figures here shown, although I am not able to say positively that the work is pre-Columbian, were executed by a native artist of about the same stage of culture as was the work of the textile design. These figures are executed in color upon the smooth surface of an earthen vase and illustrate perfectly the peculiar characters of free hand, graphic delineation. Place this and the last figure side by side and we see how vastly different is the work of two artists of equal capacity when executed in the two methods. This figure should also be compared with the embroidered figures shown in Fig. 348.

The tendencies to uniformity in textile ornament here illustrated may be observed the world over. Every element entering the art must undergo a similar metamorphosis; hence the remarkable power of this almost universally practiced art upon the whole body of decorative design.

That the range of results produced by varying styles of weaving and of woven objects may be appreciated, I present some additional examples. Coiled wares, for instance, present decorative phenomena strikingly at variance with those in which there is a rectangular disposition of parts. Instead of the two or more interlacing series of parallel fillets exhibited in the latter style, we have one radiate and one concentric series. The effect of this arrangement upon the introduced human figure is very striking, as will be seen by reference to Fig. 344, which represents a large tray obtained from the Moki Indians. The figure probably represents one of the mythologic personages of the Moki pantheon or some otherwise important priestly functionary, wearing the characteristic headdress of the ceremony in which the plaque was to be used. The work is executed in wicker, stained in such bright tints as were considered appropriate to the various features of the costume. Referring in detail to the shape and arrangement of the parts of the figure, it is apparent that many of the remarkable features are due to constructive peculiarities. The round face, for example, does not refer to the sun or the moon, but results from the concentric weaving. The oblique eyes have no reference to a Mongolian origin, as they only follow the direction of the ray upon which they are woven, and the headdress does not refer to the rainbow or the aurora because it is arched, but is arched because the construction forced it into this shape. The proportion of the figure is not so very bad because the Moki artist did not know better, but because the surface of the tray did not afford room to project the body and limbs.

Now, it may be further observed that had the figure been placed at one side of the center, extending only from the border to the middle of the tray, an entirely different result would have been reached; but this is better illustrated in a series of bird delineations presented in the following figures. With many tribes the bird is an object of superstitious interest and is introduced freely into all art products suitable for its delineation. It is drawn upon walls, skins, pottery, and various utensils and weapons, especially those directly connected with ceremonies in which the mythical bird is an important factor. The bird form was probably in familiar use long before it was employed in the decoration of basketry. In Fig. 345 I present an ordinary graphic representation. It is copied from a Zuni shield and is the device of an order or the totem of a clan. The style is quite conventional, as a result of the various constraints surrounding its production. But what a strange metamorphosis takes place when it is presented in the basketmaker's language. Observe the conventional pattern shown upon the surface of a Moki tray (Fig. 346). We have difficulty in recognizing the bird at all, although the conception is identical with the preceding. The positions of the head and legs and the expanded wings and tail correspond as closely as possible, but delineation is hampered by technique. The peculiar construction barely permits the presentation of a recognizable life form, and permits it in a particular way, which will be understood by a comparison with the treatment of the human figure in Fig. 344. In that case the interlaced combination gives relievo results, characterized by wide, radiating ribs and narrow, inconspicuous, concentric lines, which cross the ribs in long steps. The power of expression lies almost wholly with the concentric series, and detail must in a great measure follow the concentric lines. In the present case (Fig. 346) this is reversed and lines employed in expressing forms are radiate.

The precise effect of this difference of construction upon a particular feature may be shown by the introduction of another illustration. In Fig. 347 we have a bird woven in a basket of the interlaced style. We see with what ease the long sharp bill and the slender tongue (shown by a red filament between the two dark mandibles) are expressed. In the other case the construction is such that the bill, if extended in the normal direction, is broad and square at the end, and the tongue, instead of lying between the mandibles, must run across the bill, totally at variance with the truth; in this case the tongue is so represented, the light vertical band seen in the cut being a yellow stripe. It will be seen that the two representations are very unlike each other, not because of differences in the conception and not wholly on account of the style of weaving, but rather because the artist chose to extend one across the whole surface of the utensil and to confine the other to one side of the center.

It is clear, therefore, from the preceding observations that the convention of woven life forms varies with the kind of weaving, with the shape of the object, with the position upon the object, and with the shape of the space occupied, as well as with the inherited style of treatment and with the capacity of the artist concerned. These varied forces and influences unite in the metamorphosis of all the incoming elements of textile embellishment.

It will be of interest to examine somewhat closely the modifications produced in pictorial motives introduced through superstructural and adventitious agencies.

We are accustomed, at this age of the world, to see needlework employed successfully in the delineation of graphic forms and observe that even the Indian, under the tutelage of the European, reproduces in a more or less realistic way the forms of vegetal and animal life. As a result we find it difficult to realize the simplicity and conservatism of primitive art. The intention of the primitive artist was generally not to depict nature, but to express an idea or decorate a space, and there was no strong reason why the figures should not submit to the conventionalizing tendencies of the art.

I have already shown that embroidered designs, although not from necessity confined to geometric outlines, tend to take a purely geometric character from the fabric upon which they are executed, as well as from the mechanical processes of stitching. This is well shown in Fig. 348, a fine specimen given by Wiener in his work Perou et Bolive.

A life form worked upon a net does not differ essentially from the same subject woven in with the web and woof. The reason is found in the fact that in embroidery the workman was accustomed from the first to follow the geometric combination of the foundation fabric step by step, and later in life delination he pursued the same method.

It would seem natural, however, that when the foundation fabric does not exhibit well marked geometric characters, as in compactly woven canvas, the needlework would assume free hand characters and follow the curves and irregularities of the natural object depicted; but such is not the case in purely aboriginal work. An example of embroidery obtained from an ancient grave at Ancon, Peru, is shown in Fig. 349. A piece of brown cotton canvas is embellished with a border of bird figures in bright colored wool thread. The lines of the figures do not obey the web and woof strictly, as the lines are difficult to follow, but the geometric character is as perfectly preserved as if the design were woven in the goods.

So habit and association carry the geometric system into adventitious decoration. When the ancient Peruvian executed a design in color upon a woven surface (Fig. 350), using a pencil or brush, the result was hardly less subject to textile restraint.

As a matter of course, since there are two distinct styles of decorative design--the textile and the free hand--there exist intermediate forms partaking of the character of both; but it is nevertheless clear that the textile system transforms or greatly modifies all nature motives associated with it, whether introduced into the fabric or applied to its surface.

In countries where the textile art is unimportant and the textile system of decoration does not obtrude itself, free hand methods may prevail to such an extent that the geometric influence is but little felt. The Haidah Indians, for example, paint designs with great freedom and skill, and those applied to woven surfaces are identical with those executed upon skins, wood, and stone, but this art is doubtless much modified by the means and methods of Europeans. Our studies should be confined wholly to pure indigenous art.

EXTENSION OF TEXTILE ORNAMENT TO OTHER FORMS OF ART.

I have now dwelt at sufficient length upon the character of the textile system of ornament and have laid especial stress upon the manner in which it is interwoven with the technical constitution of the art. I have illustrated the remarkable power of the art by which decorative elements from without, coming once within the magic influence, are seized upon and remodeled in accordance with the laws of textile combination. Pursuing the investigation still further it is found that the dominion of the textile system is not limited to the art, but extends to other arts. Like a strong race of men it is not to be confined to its own original habitat, but spreads to other realms, stamping its own habits and character upon whatever happens to come within its reach. Its influence is felt throughout the whole range of those arts with which the esthetic sense of man seeks to associate ideas of beauty. It is necessary, before closing this paper, to examine briefly the character and extent of this influence and to describe in some detail the agencies through which the results are accomplished. First and most important are the results of direct transmission.

House building, or architecture as it is called in the higher stages, is in primitive times to a great extent textile; as culture develops, other materials and other systems of construction are employed, and the resultant forms vary accordingly; but textile characters are especially strong and persistent in the matter of ornament, and survive all changes, howsoever complete. In a similar way other branches of art differentiated in material and function from the parent art inherit many characters of form and ornament conceived in the textile stage. It may be difficult to say with reference to any particular example of design that it had a textile origin, for there may be multiple origins to the same or to closely corresponding forms; but we may assert in a general way of the great body of geometric ornament that it owes something--if not its inspiration, its modes of expression--to the teachings of the textile system. This appears reasonable when we consider that the weaver's art, as a medium of esthetic ideas, had precedence in time over nearly all competitors. Being first in the field it stood ready on the birth of new forms of art, whether directly related or not, to impose its characters upon them. What claim can architecture, sculpture, or ceramics have upon the decorative conceptions of the Digger Indians, or even upon those of the Zuni or Moki? The former have no architecture, sculpture, or ceramics; but their system of decoration, as we have seen, is highly developed. The Pueblo tribes at their best have barely reached the stage at which esthetic ideas are associated with building; yet classic art has not produced a set of geometric motives more chaste or varied. These examples of the development of high forms of decoration during the very early stages of the arts are not isolated. Others are observed in other countries, and it is probable that if we could lift the veil and peer into the far prehistoric stages of the world's greatest cultures the same condition and order would be revealed. It is no doubt true that all of the shaping arts in the fullness of their development have given rise to decorative features peculiar to themselves; for construction, whether in stone, clay, wood, or metal, in their rigid conditions, exhibits characters unknown before, many of which tend to give rise to ornament. But this ornament is generally only applicable to the art in which it develops, and is not transferable by natural processes--as of a parent to its offspring--as are the esthetic features of the weaver's art.

Besides the direct transmission of characters and forms as suggested in a preceding paragraph, there are many less direct but still efficacious methods of transfer by means of which various arts acquire textile decorative features, as will be seen by the following illustrations.

Japanese art is celebrated for its exquisite decorative design. Upon superb works of porcelain we have skillful representations of subjects taken from nature and from mythology, which are set with perfect taste upon fields or within borders of elaborate geometric design. If we should ask how such motives came to be employed in ceramic decoration, the answer would be given that they were selected and employed because they were regarded as fitting and beautiful by a race of decorators whose taste is well nigh infallible. But this explanation, however satisfactory as applied to individual examples of modern art, is not at all applicable to primitive art, for the mind of man was not primarily conscious of the beauty or fitness of decorative elements, nor did he think of using them independently of the art to which they were indigenous. Now the ceramic art gives rise to comparatively few elements of decoration, and must therefore acquire the great body of its decorative motives from other arts by some process not primarily dependent upon the exercise of judgment or taste, and yet not by direct inheritance, as the techniques of the two arts are wholly distinct.

Textile and fictile arts are, in their earlier stages, to a large extent, vessel making arts, the one being functionally the offshoot of the other. The textile art is the parent, and, as I have already shown, develops within itself a geometric system of ornament. The fictile art is the offshoot and has within itself no predilection for decoration. It is dependent and plastic. Its forms are to a great extent modeled and molded within the textile shapes and acquire automatically some of the decorative surface characters of the mold. This is the beginning of the transfer, and as time goes on other methods are suggested by which elements indigenous to the one art are transferred to the other. Thus we explain the occurrence, the constant recurrence of certain primary decorative motives in primitive ceramics. The herring bone, the checker, the guilloche, and the like are greatly the heritage of the textile art. Two forms derived from textile surfaces are illustrated in Figs. 351 and 352. In the first example shown, herring bone patterns appear as the result of textile combination, and in the second a triangular checker is produced in the same way. In Fig. 352 we see the result of copying these patterns in incised lines upon soft clay.

Again, the ancient potter, who was in the habit of modeling his wares within baskets, seems to have conceived the idea of building his vessels by coiling just as he built his baskets. The surface exhibits coiled ridges like basketry, as shown in Fig. 353, and the textile character was further imposed upon the clay by marking these coils with the thumb and with implements to give the effect of the transverse series of filaments, and the geometric color patterns of the basketry were reproduced in incised lines. When these peoples came to paint their wares it was natural that the colored patterns native to the basketry should also be reproduced, and many more or less literal transfers by copying are to be found. A fine example of these painted textile designs is shown in Fig. 354. It is executed in a masterly style upon a handsome vase of the white ware of ancient Tusayan. Not only are the details reproduced with all their geometric exactness, but the arrangement of the designs upon the vessel is the same as in the textile original. Nine-tenths of the more archaic, Pueblo, ceramic, ornamental designs are traceable to the textile art, and all show the influence of textile convention.

Another peculiar class of transfers of a somewhat more indirect nature may be noticed. All the more advanced American nations were very fond of modeling the human form in clay, a large percentage of vessels having some trace of the human form or physiognomy. Now, in many cases the costume of the personage represented in the clay is also imitated, and generally in color, the details of the fabrics receiving their full share of attention. Such an example, from a sepulcher at Ancon, is shown in Fig. 355. Here the poncho or mantle thrown across the shoulders falls down upon the body in front and behind and the stripes and conventional fishes are accurately reproduced. In this way both style and matter of the textile decoration are introduced into the ceramic art.

It will be seen by these illustrations that there are many natural methods, automatic or semiautomatic in character, by which the one art receives aid from the other; that in the beginning of the transfer of textile ornament to fictile forms the process is purely mechanical, and that it is continued automatically without any very decided exercise of judgment or taste. As a result, these borrowed decorations are generally quite as consistent and appropriate as if developed within the art itself. Later in the course of progress the potter escapes in a measure from this narrow groove and elaborates his designs with more freedom, being governed still to a certain extent by the laws of instinctive and automatic procedure. When, finally, intellect assumes to carry on the work independently of these laws, decoration tends to become debased.

Turning to other branches of art, what traces do we find of the transfer to them of textile features? Take, for example, sculpture. In the wood carving of the Polynesians we observe a most elaborate system of decoration, more or less geometric in character. We do not need to look a second time to discover a striking likeness to the textile system, and we ask, Is it also derived from a textile source? In the first place let us seek within the art a reason for the peculiar forms. In carving wood and in tracing figures upon it with pointed tools the tendency would certainly be towards straight lines and formal combinations; but in this work there would be a lack of uniformity in execution and of persistency in narrow lines of combination, such as result from the constant necessity of counting and spacing in the textile art. In the presentation of natural forms curved lines are called for, and there is nothing inherent in the carver's art to forbid the turning of such lines with the graver or knife. Graphic art would be realistic to an extent regulated by the skill and habits of the artist. But, in reality, the geometric character of this work is very pronounced, and we turn naturally toward the textile art to ask whether in some way that art has not exercised an influence. The textile arts of these peoples are highly developed and were doubtless so in a degree from very early times, and must have had a close relation with the various arts, and especially so in the matter of ornament. Specific examples may be cited showing the intimacy of wood carving to textilia. Bows, spears, arrows, &c. are bound with textile materials to increase their strength. Knives and other weapons are covered with textile sheaths and handles of certain utensils are lashed on with twisted cords. In ceremonial objects these textile features are elaborated for ornament and the characteristic features of this ornament are transferred to associated surfaces of wood and stone by the graver. A most instructive illustration is seen in the ceremonial adzes so numerous in museums (Fig. 356). The cords used primarily in attaching the haft are, after loss of function, elaborately plaited and interwoven until they become an important feature and assume the character of decoration. The heavy wooden handles are elaborately carved, and the suggestions of figures given by the interlaced cords are carried out in such detail that at a little distance it is impossible to say where the real textile surface ceases and the sculptured portion begins.

All things considered, I regard it as highly probable that much of the geometric character exhibited in Polynesian decoration is due to textile dominance. That these peoples are in the habit of employing textile designs in non-textile arts is shown in articles of costume, such as the tapa cloths, made from the bark of the mulberry tree, which are painted or stamped in elaborate geometric patterns. This transfer is also a perfectly natural one, as the ornament is applied to articles having functions identical with the woven stuffs in which the patterns originate, and, besides, the transfer is accomplished by means of stamps themselves textile. Fig. 357 illustrates the construction of these stamps and indicates just how the textile character is acquired.

Textile materials are very generally associated with the human figure in art, and thus sculpture, which deals chiefly with the human form, becomes familiar with geometric motives and acquires them. Through sculpture these motives enter architecture. But textile decoration pervades architecture before the sculptor's chisel begins to carve ornament in stone and before architecture has developed of itself the rudiments of a system of surface embellishment. Textile art in mats, covers, shelters, and draperies is intimately associated with floors and walls of houses, and the textile devices are in time transferred to the stone and plaster. The wall of an ancient Pueblo estufa, or ceremonial chamber, built in the pre-esthetic period of architecture, antedating, in stage of culture, the first known step in Egyptian art, is encircled by a band of painted figures, borrowed, like those of the pottery, from a textile source. The doorway or rather entrance to the rude hovel of a Navajo Indian is closed by a blanket of native make, unsurpassed in execution and exhibiting conventional designs of a high order.

The ancient "hall of the arabesques" at Chimu, Peru, is decorated in elaborate designs that could only have arisen in the textile art (Fig. 358), and other equally striking examples are to be found in other American countries. The classic surface decorations known and used in Oriental countries from time immemorial prevailed in indigenous American architecture at a stage of culture lower than any known stage of classic art.

It may appear that I have advocated too strongly the claims of the textile art to the parentage of geometric ornament and that the conclusions reached are not entirely satisfactory, but I have endeavored so to present the varied phenomena of the art that the student may readily reach deductions of his own. A correspondingly careful study of other branches of art will probably enable us finally to form a just estimate of the relative importance of the forces and tendencies concerned in the evolution of decoration.

* * * * *

INDEX

Alaskan Indians, illustration of ornamentation by 199 Ancon, Peru, examples of ornamentation from graves at 212, 230, 231, 236, 243, 248 Apache, illustrations of ornamentation by 198, 213, 223 British Guiana Indians, illustrations of ornamentation by 217 Chimu, Peru, ornamentation of "hall of arabesques" at 251, 252 Clallam Indians, illustrations of ornamentation by 207 Color in textile art 201, 202 Color phenomena in textile ornament 215-232 Form in textile art and its relation to ornament, with illustrations from Indian work 196-201 Geometric design, relations of, to textile ornament 202-244 Holmes, W. H. paper by, on textile art in its relation to the development of form and ornament 189-252 Klamath Indians, illustrations of ornamentation by 208, 209, 227 McCloud River Indians, illustrations of ornamentation by 221 Moki, illustrations of ornamentation by 197, 205, 224, 225, 226, 238, 240 Northwest Coast Indians, illustrations of ornamentation by 213, 218, 227, 230 Penn wampum belt 233 Peruvians, ancient, illustrations of ornamentation by 211, 212, 214, 228, 230, 231, 235, 236, 237, 242, 243, 248 Pima Indians, illustrations of ornamentation by 220 Piute Indians, illustrations of ornamentation by 198, 205 Polynesian ornamentation, illustrations of 249, 250 Seminole Indians, illustrations of ornamentation by 207 Textile art in its relation to the development of form and ornament, paper by W. H. Holmes on 189-252 Tule River Indians, illustrations of ornamentation by 219 Tusayan ornament, illustrations of 247, 248 Wiener, cited 242 Yokut Indians, illustrations of ornamentation by 233, 234 Zuni, illustrations of ornamentation by 239

* * * * *