I may remark that this test has been a very trying one, for I estimate
that it is equal to more than a year's exposure in this country.
During the whole period there was cloudless sunshine, without any
rain, and each evening heavy dew. I have pleasure in acknowledging the
services of Mr. W. Reid, a former student, who superintended the
exposure of the patterns, and from time to time took notes of the rate
at which individual patterns faded.

These diagrams contain, perhaps, the most complete series of both old
and new dyes, on the three fibers, which have been simultaneously
exposed to sunlight, and they form an instructive object lesson.

Let me first direct your attention to the diagram containing the
_natural coloring matters_--those dyestuffs which were in use previous
to 1856. Broadly speaking, they are of two kinds; those which dye
textile materials "direct," and those which give no useful color
without the aid of certain metallic salts, called "mordants."

Now, among the natural coloring matters, these "mordant dyes," as they
may be conveniently termed, are much more numerous than the "direct
dyes;" but be it observed, we have fast and fugitive colors in both
classes.

Referring first to the wool patterns and to the "direct dyes," we find
that the only really fast colors are Prussian blue and Vat indigo
blue. Turmeric, orchil, catechu, and indigo carmine are all extremely
fugitive.

As to the "mordant dyes," some yield fast colors with all the usual