spadix, generally conical in shape, sometimes, however, altogether replaced
by a perfect thistle, at other times again by a pomegranate. Auberville, in
his magnificent work "L'Ornement des Tissus," is astonished to find the
term pomegranate-pattern almost confined to these forms, since their
central part is generally formed of a thistle-form. As far as I can
discover in the literature that is at my disposal, this question has not
had any particular attention devoted to it except in the large work upon
Ottoman architecture published in Constantinople under the patronage of
Edhem Pasha. The pomegranate that has served as the original of the pattern
in question is in this work surrounded with leaves till it gives some sort
of an approach to the pattern. (There are important suggestions in the book
as to the employment of melon-forms.) Whoever has picked the fruit from the
tender twigs of the pomegranate tree, which are close set with small
altered leaves, will never dream of attributing the derivation of the
thorny leaves that appear in the pattern to pomegranate leaves at any stage
of their development.

[Illustration: FIG. 14]

It does not require much penetration to see that the outline of the whole
form corresponds to the spathe of the Araceæ, even although in later times
the jagged contour is all that has remained of it, and it appears to have
been provided with ornamental forms quite independently of the rest of the
pattern. The inner thistle-form cannot be derived from the common thistle,
because the surrounding leaves negative any such idea. The artichoke theory
also has not enough in its favor, although the artichoke, as well as the
thistle, was probably at a later time directly pressed into service. Prof.
Ascherson first called my attention to the extremely anciently cultivated
plant, the safflor (_Carthamus tinctoris_, Fig. 15), a thistle plant whose
flowers were employed by the ancients as a dye. Some drawings and dried