s the light (that is the best position for it);
we shall then first perceive that the whole surface is covered, or, so to
{22} speak, tiled over with distinct, sharply cut _scales_, arranged as in
fig. 16, Plate II., with the outer or free edges of one row overlapping the
roots of the next. These roots being all planted towards the base of the
wing, if we place that end next the light (as above directed), the free
edges of the scales throw a strong shadow on the next row, which brings out
the imbricated effect most strikingly.

Beginning our observations at the outer edge of the wing, we first notice a
delicate fringe of scales or plumes, more elongated and pointed than the
surface scales, and of a quiet brown colour. This tint is continued inwards
for a short space, gradually lightening, when (as we shift the field of
view towards the centre of the wing) the colour of the scales suddenly
changes to an intense black; then a little further, and the black ground is
all spangled with glittering sapphires, then strewed deep with amethyst
round a heap of whitest pearls. Golden topaz--(jewels only will furnish apt
terms of comparison for these insect gems)--golden topaz ends the bright
many-coloured crescent, and in the centre is enclosed a spot of profoundest
black, gradating into a rich unnameable red, whose velvet depth and
softness contrast deliciously with the adjacent flashing lustre; then comes
another field of velvet black, then more gold, and so on till the gorgeous
picture is complete.

Subject a piece of finest human painting to the scrutiny of a strong
magnifying glass, and where is the beauty thereof? Far from being
magnified, it will have wholly vanished: its cleverest touches turned to
coarse, repulsive daubs and stains.

[Illustration: IV.]

{23}

Now, bring the microscope's most searching powers to bear upon the painting
of an insect's wing, and we find only pictures within pictures as the
powers increase; the very pigments used turn out to be jewels, not rough
uncut stones, but cut and graven gems, bedded in softest velvet.

If by gentle rubbing with the finger-tip the scales be removed from both
sides of the wing (for each side is scale-covered, though generally with a
very different pattern), there remains a transparent membrane like that of
a bee's or fly's wing, tight stretched between stiff branching veins, but
bearing no vestige of its late gay painting, thus showing that the whole of
the colouri