%it>rar\> of the "Wmversitp of 'Misconetn TEXT BOOKS OF ORNAMENTAL DESIGN By LEWIS F. DAY. I. THE ANATOMY OF PATTERN. LEWIS F. DAY'S TEXT BOOKS OF ORNAMENTAL DESIGN. Price Three-and-Sixpence each, Crown 8vo, bound in Cloth. SOME PRINCIPLES OF EVERY-DAY ART: Introductory Chapters on the Arts not F1ne. Second Edition. In great part re-written from 'Every-Day Art,' containing nearly all the original Illustrations and some new ones; forming a Prefatory Volume to the Series. THE ANATOMY OF PATTERN. Fourth Edition and Eighth Thousand, revised, with Cuts and Forty-one full-page Illustrations, several of them first drawn for this Edition. THE PLANNING OF ORNAMENT. Third Edition and Fifth Thousand, revised, with Forty-one full-page Illustrations, many of them re-drawn for this Edition. THE APPLICATION OF ORNAMENT. Fourth Edition and Fifth Thousand, revised. With Forty-eight full-page Illustrations. Thick Crown 8vo, Cloth Gilt. Price 12s. atThird Edition and Fourth Thousand. NATURE IN ORNAMENT. With a Hundred and Twenty-three Plates, and a Hundred and Ninety- two Illustrations in the text. DESCRIPTIVE LIST OF PLATES. 1. THE CONSTRUCTION OF GOTHIC TRACERY PATTERNS— Showing the square, diamond, hexagon, circular, or other plan on which elaborate tracery is built 2. the square—Checks and other diapers built on cross- lines. 3. THE LATTICE AND THE DIAMOND—Plaids, zigzags, &c, built on cross-lines. 4. frets, &c.—Showing their construction on a network of cross-lines. 5. all-over pattern—Showing the cross-lines upon which it is planned. 6. the triangle —Diapers of star, hexagon, and lozenge shapes, built up on the lines of the equilateral triangle. 7. the triangle —Diapers composed of the equilateral triangle and its compounds. 8. the hexagon—Honeycomb and other diapers based upon the hexagon and its compound. 9. diaper —Designed on the lines of hexagon or triangle. 10. the octagon—Simple octagon diapers and the lines of their construction. 11. ARAB lattice pattern—Dissected, and their anatomy laid bare. 12. curvilinear patterns—Showing the construction of the wave, the ogee, the net, &c 2 The Anatomy of Pattern. pattern; certainly it is impossible to plait, net, knit, weave, or otherwise mechanically make, without producing pattern. It may be infinitesimally small, as in weaving, where the warp and weft are often invisible to the naked eye; but it is there; and all that re- mains for us to do is, to efface it as far as we can, or to make the best of it. Out of the determination to make the best of it has grown much of the most beautiful pattern-work. To neglect this source of in- spiration, therefore, to say nothing of the attempt to suppress it, would seem to be wasteful of opportunity to the very last degree. So certainly will the repetition of parts result in some sort of pattern, that one may say, wherever there is ordered repetition there / is pattern. Take any form you please, and s repeat it at regular intervals, and you have, whether you want it or no, a pattern, as surely as the recurrence of sounds will pro- 'duce rhythm or cadence. The distribution of the parts need not even be regular. The wave marks on the sand, the veins of marble, the grain of wood, the crystallisation of the breath upon the window- 6 The Anatomy of Pattern. upon the vertical, horizontal, or diagonal lines, which thus asserted themselves. It was much more likely the result of not working upon definite lines at all. A designer who knew the A B C of his business, would make sure of lines not in themselves offensive; he would counteract a tendency to stripes in one direction by features directing the atten- tion otherwards; and he would so clothe any doubtful line that there would be no fear of its unduly asserting itself, as in its naked- ness it might. He foresees the danger (it is a danger even to the most experienced) and he is fore-armed against it. The mighty man of valour who disdains to be trammelled by principles, or any such encumbrance, is with- out defence against contingencies practically certain to arrive. It is only by a miracle, or a fluke, that he can escape failure. The over- whelming odds are, that the petty considera- tions he has despised will be quite enough to wreck any venture he has dared in defiance of them. Since, then, it is practically inevitable that there shall be definite lines in ornamental design—seeing that if you don't arrange for them they arrange themselves—it is the merest Introductory, 7 common sense to lay down those lines to begin with, and, in fact, to make them the skeleton or framework upon which you build up your pattern. You will see, when they are laid bare for you, that these skeletons are after all very few. 8 The Anatomy of Pattern. II. PATTERN DISSECTION. REPEATED pattern may be classified, I said, according to its structure. First in order of obviousness comes the stripe. It comes also very early in order of invention: the loom must from the begin- ning have suggested the stripe-pattern, which practically grows out of it. The stripe, however, carries us only a very short distance in the direction of design. For as soon as you make any break in the re- peated line, the recurrence of that break gives other lines in the cross direction. Suppose a series of horizontal bands broken at equal intervals by a series of rosettes. It is clear, that if the rosettes fall one under the other, they give upright lines; or if they are shifted you get diagonal cross lines. If the line itself is broken, as in the case of a series of waved lines, or, still more plainly, in a Pattern Dissection. 9 series of Vandykes, the turn of the wave, or the point of the zigzag, when it is repeated, gives the cross line just the same. And so we come at once to the vast order of patterns constructed upon cross lines. This is probably quite the first in point of time, arising as it inevitably does out of the very primitive art of plaiting. You have only to interweave strips of two different colours, and you get at once a check, or what is familiar to us in black and white as the chess-board pattern. (Plate 2.) Suppose the interwoven strips to be all of one colour, then the lines of intersection would make a lattice or basket-work pattern. The simplest form of check or lattice is when the crossing is at equal intervals and at right angles. Vary the interval, and you have all manner of plaids and tartans. Alter your point of view (or turn the design 45 degrees round) and you get the diamond. The difference in point of view makes no real difference in plan: a stripe may take any direction, yet it is always a stripe. But if we alter the angle at which the lines cross, we get not only a fresh variety of shapes, but we may obtain also a diamond Pattern Dissection. II to the pattern, but to its translation into a textile fabric. If instead of the chess-board we take the lines of the lattice, and work upon them, we get, without departing from those lines (only intermitting them) a wonderful range of inter- lacements and the like; some of them of ex- ceeding intricacy, as in the case of the "fret" A number of these are shown on Plate 4. There seems no limit to the ever-increasing range of pattern-work thus disclosed, all built upon the same constructional scaffolding. From the intermission of the lines results a dnd of spot pattern, more or less free, which : might be mistaken for a distinct order of de- ! >ign (Plates 2 and 4). But it is only a variety. In a certain sense it matters little whether [a design is constructed on geometric lines, or only arranged so that it falls within them. I The skeleton, when you come to dissect the two, is the same in either case. Our theory of Iconstruction, therefore, applies quite as much to sprigs, spots, and all so-called free patterns, Us to those in which the constructional lines actually occur as lines. You have not done away with construction when you have suc- ceeded in keeping the scaffolding out of 1 I 9 a^lcb ▲ fffftfTfirl 14 The Anatomy of Pattern. another new shape is evolved. Returning once again to the square lattice, if we cross it diagonally both ways, cross it by itself, that is, so that each square is cut up into four, we get out of those lines the octagon (Plate 10); but not an equal-sided octagon; that is, built on a cross lattice of different proportions. The octagon, however, is not a unit which will of itself form a diaper, as the hexagon will. It is only in connection with a square, diamond, or other four-sided figure, that it will "repeat." Place side by side a series of octagons, and there will appear four-sided gaps between (Plate 10). Nevertheless, this new series of lines gives us new varieties of radi- ated pattern: witness once more the elaborate interlacings of the Arabs; all of which, even the most magnificent, are closely related to the pattern so familiar in the, seat of a common cane-bottomed chair. It is possible to carry the principle of radia- tion further still. You may, for example, cross this more elaborate lattice by a lattice like itself; but you get by that means rather intricacy than variety—especially when the intersecting lines are in part interrupted. In Pattern Dissection. 15 certain Arab patterns, where this ultra-elabora- tion of lines is employed, it appears almost as if a new principle had been introduced (Plate 11); but, upon analysis, the designs resolve themselves into the elements with which we have already had to deal—so few are the plans upon which pattern is con- structed. Already we have come to the end of the straight-lined family. Why, it may be asked, can you not make a diaper on other lines, on the lines of the pentagon for ex- ample? Well, you may put together so many penta- gons—and a very respectable diaper they form—espe- cially if you fur- ther enrich the pentagons with five-pointed stars. Not long since I came upon just such a diaper, which, for a moment, promised to upset all my neatly arranged theories on the subject of pattern anatomy. However, Pentagon diaper and its skeleton. fPl&te 12. 20 The Anatomy of Pattern. figures (Plate 12), very nearly approaching the straight-lined hexagon. In this way the straight-lined series might be derived from the curved (compare Plates 1 and 11); and so once more, by a very different road, we reach always, in this maze of pattern- work, the same point, which is, the limited variety of skeleton on which pattern is built. From the combination of straight lines with curved (Plates 13 and 14) result all manner of new diaper forms; which, however, present nothing very new in the way of skeleton. You might start a scroll pattern, such as that given in Plate 17 (a type common in the sixteenth and seven- teenth centuries), on the lines either of the hexagon or of the ogee, or of a mixture of curved and straight lines which I may call the broken ogee; and in the end it would not be very clear which of them you had taken for a groundwork; or even whether you had not founded your design upon the diamond—such close kindred do those various skeletons betray. I have dwelt at some length upon rudi- mentary diaper forms, for reasons quite apart 22 The Anatomy of Pattern. by the conditions of his work, are, in most instances, not just those which beauty would have decreed. They prove, however, to be identical with the lines already shown to be the basis of all recurring pattern-work; and so we begin to see that, had there been no such thing as pattern design before, and no traditional forms of design for us to follow, those very forms must have been evolved as certainly out of the more complex conditions of modern manufacture as they were out of the simple contrivances of primitive handicraft. That is to say, that the lines first given to us by the primary processes of netting, plaiting, and so on, would equally have been prescribed by the printing roller or the power loom. It is one of the most interesting points in the analysis of pattern design to see how regularly we work round, again and again, to identically the same shapes. You cannot safely dogmatise as to the origin of this or that pattern; there are always so many ways in which it might have been suggested. Put side by side a series of waved lines so that their curves are opposed (Plate 12) and the effect is exactly the same as though you had opened Pattern Dissection. 23 out an ogee diaper; you can deduce either pattern from the other. Or, again (same plate), if the ogees interlace, it is impossible to say whether this was the outcome of the ogee, or of waved lines, or simply of the pro- cess of netting. On Plate 16 are shown six different ways in which one and the same simple star pattern may be arrived at. 1. By the juxtaposition of stars and the addition of cross-lines. 2. By the juxtaposition of diamonds and the addition of cross-lines. 3. By the juxtaposition of right-angled diamonds, each occupied by a star. 4. By the interlacing of two series of octagons, and the addition of cross-lines. 5. By the crossing of two series of zigzag lines, and the addition of cross-lines. 6. By the crossing of two series of diamonds or lozenges, and the addition of cross-lines. And this does not by any means exhaust the number of ways in which the same result might have been reached. To take another instance, of a very dif- ferent kind, you know how common it is to see a waved line with leaves alternating on 24 The Anatomy of Pattern. each side of it. This appears on the face of it, a quite mechanical and arbitrary arrangement; but you have only to note how, in nature, the alternate leaves on a slender stem pull it out of the straight, to see the natural and inevitable origin of the idea. By merely ex- aggerating the slight wave of the natural stem, you get one of the most conventional of ornamental border patterns.' So it would seem that, whether you begin with mechanical construction or with nature it works round to the same thing in the end—in the hands of an ornamentist. * See 'Nature in Ornament,' pp. 55, 56. ^Plate 18 "Photc-Tiht" fcjr J.Ak«mau.6,Qi»«n SipimX 26 The Anatomy of Pattern. and though the printer make use of the roller' instead of the block, the conditions of design remain unaltered; for the roller is, for all practical purposes of design, only a block bent round in the shape of a cylinder. The square plan of the printed curtain design on Plate 18 was prescribed by the roller, which was 30 inches wide and the same in circum- ference. Even the bookbinder of earlier days, who was comparatively free to do what he liked in the way of "tooling," was led, whether by instinct or by his tools, to adopt a rectangular repeat, as in Plate 19; in which also is exemplified what may be done in the way of reversing, and again reversing, the unit of design—so as with comparatively little drawing to produce the effect of an extensive pattern. We have, ordinarily, to reconsider the pos- sible lines of pattern construction in their rela- tion to the rectangular figure,—that being the repeat determined for us by the conditions of nearly all modern manufacture. The base of our operations is, then, usually a parallelogram. Furthermore, this parallelogram is in all flate 19. Practical Pattern Planning. 2 7 cases restricted in size, and in most cases of more or less arbitrary proportions. For example—in the case of wall-paper printing, it is practically determined for us that the printer's block shall be rectangular. Custom has further fixed its width at 21 inches. And, since a block of greater length than that would be unwieldy, we are restricted to a square of 21 inches by 21 inches. The block may represent a fraction only of the design, which can theoretically be made up of as many blocks as you please. But in practice the expense of such a pro- ceeding would make the paper-hangings cost more than paper-hangings are ordinarily worth; and, apart from commercial consider- ations, which would be enough to prevent that kind of extravagance, it is contrary to craftsmanship so to misapply labour. The most capable artist is he who can apply his art to most purpose, and get full value out of his materials. As a matter of fact, the wall-paper designer has to content himself, then, except in very few instances, with a repeat of at most 21 inches square. Within those limits he is comparatively 28 The Anatomy of Pattern. free; but, as I have already shown, do what he may, his repeated pattern will fall into geometric lines, if only those of the parallelo- gram on which it is built. A pattern, such as A, on Plate 20, may seem at first sight to conform to no conditions of restraint; but the actual lines of the repeat reveal them- selves, on closer inspection, in any single feature whose recurrence is to be traced. It is based, you will find, upon the parallelogram —faintly indicated by dotted lines upon the black ground. Apart from the conditions of actual manu- facture, it is found commercially expedient to adopt certain fixed dimensions for the tile, block, roller, or whatever it may be—and we are thus constrained to design tiles (if they are to be of any use) on the usual three-, six-, eight-inch or other accepted scale; textiles to a width fixed by the loom, and a length controlled by the consideration of economy; block-printed fabrics under very similar con- ditions; and roller-printed to a length as well as a width prescribed. The proportion of the parallelogram within which our design must be confined varies, that is to say, with the manufacture for which we are designing. An Practical Pattern Planning. 29 experienced designer could often tell, from its proportion and scale alone, for what par- ticular manufacture a design was made. And it is in the impracticability of his ideas that the novice most infallibly betrays his lack of experience. The production of such a pattern as that on Plate 21 in the form of a wall-paper 21 inches wide, would involve the (prohibitive) cost of four sets of blocks. Whereas to the weaver it would not prove comparatively very costly: as a matter of fact patterns of that relative length occur quite frequently in textiles. In the width of the Sicilian silk, on Plate 22, dogs, lions, and eagles follow one another closely in a row; but in the length of the material each is separated from its like by the other two. The pattern repeats on the lines of a long upright parallelogram; but obvi- ously it was designed on the cross lines shown. There is no occasion to enter more fully into all the various technical reasons for the limitations to which the designer is subject. The practical convenience of them, however, is patent. It is as desirable that the architect, for example, should know what sized tiles may be available, as that he should be able 30 The Anatomy of Pattern. to reckon upon the "bond " of his brickwork; and it is equally clear that without some uni- formity in the width of materials (such as silks, velvets, carpets, chintzes, and so on), it would be difficult to estimate, off-hand, the relative cost of each. The upshot of it is, that the designer has habitually to shape his design according to a rectangular plan, and that of limited, if not fixed, dimensions. It becomes, then, a very serious question with him how far he can avail himself of any other basis. The student might with advantage set himself to tabulate the possibilities in the way of adapting the various units of repeat to repetition, within the square. It would then be seen that, though all things are possible, there are schemes the artist would like to adopt, which, in order to be brought into the repeat permitted, would need to be worked out upon so small a scale as to become quite too insignificant for use. One instance of this it may be worth while to give. Suppose a square block of 21 inches, and you wish to adapt a hexagonal design to 32 The Anatomy of Pattern. duced to half that size; that is to say, there would have to be seven hexagons to the width, measuring each only three inches across. Try to arrange a pattern such as that on Plate 9 to accommodate itself to a rectangular block of given dimensions, and you will realise how inevitably it is born of the hexagon or the triangle, and cannot be made to fit into arbitrary square lines. It will be seen how very strictly the artist is bound by considerations which scarcely occur to the uninitiated, considera- tions which have always had a great deal to do with the design of pattern-work. Fashion has had her say in the matter, too, no doubt—it is a wicked way she has; but, though certain lines have been generally adopted at certain periods and in certain countries, I think it will invariably be found that there was some technical or prac- tical reason for their adoption in the first instance. Out of the conditions of weaving came, for example, the adoption of upright patterns, such as that on Plate 23 (from a coarse woollen fabric of the fifteenth century) and cross colouring such as occurs in Byzantine, Plate 24. Practical Pattern Planning. 33 Sicilian, and Early Italian silks (Plate 22).* Out of weaving comes also the turning over of the design on the two sides of an upright stem, or purely imaginary central line; as in Plates 23, 24, and many others. In Plate 24 may further be seen what great influence material may exercise upon pattern. There was a whole class of pat- terns of this kind schemed in the 15th and 16th centuries, with the obvious purpose of disturbing as little as possible of the rich pile of the velvet for which they were designed. When it is realised that the turning over of the pattern is essentially a weaver's device, it will be obvious that in a pattern similarly planned for printing there is no occasion for such rigid symmetry of the two sides; and that, on the contrary, it is desirable rather to introduce a certain amount of variation in the pattern. That is shown in Plate 25, where, though the main lines take the formal ogee shape, the termination of the branches is not alike on both sides. * See also 'The Application of Ornament,' Plate 23. D The '' Drop" Pattern. 3 5 piece of the stuff will fit on to the left of another, and so on. But it is clear that the design may be so contrived that each succeeding breadth has to be dropped in the hanging. If this drop were only very slight—say three inches—it would take seven breadths, in a pattern of 21 inches deep, before a given feature in the design occurred again exactly on the same level. There would be no dan- ger then of any horizontal tendency in the marked lines which recurring features of the design might take, but, on the other hand, great likelihood of a diagonal line developing itself, with even more unfortunate effect. The design would seem, rather, to step down- wards; and the shorter the steps, the more noticeable would be the line such features might take. This difficulty is avoided if you make the "drop" just one-half the depth of the pattern, so that every alternate strip is hung on the same level. Then the diagonal lines correct themselves. If any line at all asserts itself, it is more likely a zigzag (instead of a step), which, in connection with corre- sponding zigzags above and below, may very possibly form a trellis or lozenge pattern. D 2 36 The Anatomy of Pattern. There is good reason, therefore, for saying the diamond is a useful plan to work on; for upon it is formed the safest variety of drop pattern—that, namely, which drops one-half its depth. Instances of drop patterns are given in Plates 17, 20 (b), 37, 41, and others. One has heard persons, more familiar with the forms of ornament than expert in prac- tical design, complain of the difficulty they experience in scheming a "drop." If they would only think of the problem as the filling of a diamond shape, it would come very easily to them. The diamond is, in fact, only a square turned part way round. That part of your design which extends beyond the margin A B must occur again within the margin C D; that which extends beyond the margin C A ma. Akarjnm. London .W.C. 4i V. SKELETON PLANS. The designer finds it ordinarily more con- venient to design at once upon the lines of the diamond, because their simplicity enables him better to keep in view the effect of his pattern in its repeated form than any other lines on which the "drop" can be worked. Even though one may have no intention of taking advantage of the full width of a block, it may still be found convenient to design within the diamond, if only in order to economise design: and, mind you, economy is an absolute necessity of the case. But for economic reasons there would be no weaving, printing, stamping, and so on; we should confine ourselves to embroidery, tapestry, painting, and other work of our own hands. Assume, for the purpose of explanation, that it is a wall-paper you want to design. If you begin by dividing the width of 21 inches into two, and make your pattern a "drop," 21 inches long by rob wide, as at A (Plate 32 *PhotO-TiHt" by Jan* 'l Alc«nn an. London W C.