d and
symmetric motive of an anthemion character.

[Wall-Paper Design]

Take, however, a wall-paper. The problem is to construct a design
pleasant to the eye in line, form, colour, and suggestion; which will be
interesting in detail, and yet repeat upon a wall-surface without flaw,
and without becoming wearisome. Moreover, one which will lend itself to
being cut upon wood, if for block-printing, and which may be reproduced
with a due regard to economy of means. The designer may have a square of
twenty-one inches in which to make his design.

[Illustration (f026): Diagram Showing the Use of a Geometric Basis in
Designing Repeating Pattern.]

A useful way to begin with is to rule out a sheet of paper into squares,
say on the scale of 1-1/2 inch to the foot, and upon this jot down your
first ideas of linear arrangement and colour motive, and get the
general effect, and test the plan of repeats. When you are satisfied
with one, enlarge it to full size, correct and amplify it, and improve
it in form and detail. Changes will probably be found necessary in
drawing it upon the larger scale, sometimes additions, sometimes
omissions. Now in sketching out the general plan, one builds, as before
said, upon some basis or plan, however simple, since one cannot put a
simple spot, sprig, or spray upon paper intending to repeat, without
some system of connection to put them into relation.

[Controlling Forms]

In designing one's sprig, too, the best plan to secure good decorative
effect is to see that its general form is inclosed or bounded by an
agreeable linear shape, although itself not actually visible. Simple
leaf and flower forms are generally the best to use for these
controlling boundaries. Sprays designed on this principle may be relied
upon for repeating pleasantly and safely when they are placed upon, and
connected by, the controlling geometric plan. A good practical test of
the truth and completeness of your square repeat is, when the design is
done, or even in progress, to cut it into four equal parts (supposing it
to be a twenty-one inch square). This will enable you to get the joints
true, and also, by altering the position of the squares, to give you a
very good idea of the effect of the repeat full size. (See the diagrams
on p. 41.[f028])

These things must be considered, of course, merely as practical ai