essed by nearly all
bodies, of absorbing certain parts of the light that passes through
them. Sunlight passing through a prism is split up into the familiar
seven colours of the rainbow. But if a little blood dissolved in water
is placed in a glass tube, and if the light is made to pass through it
on its way to the prism, the blood takes something out of it; for now
among the seven bright colours are seen two dark bands near the middle
of the yellow ray. Nothing but blood gives these two bands in that
particular place, with the exception of two or three substances that are
not likely to be found on criminals' clothes. These are cochineal, mixed
with certain chemicals, hot purpurin sulphuric acid, and the red dye of
the banana-eater.

Blood, however, changes after it is shed. In stains a few weeks old the
colouring matter changes from what is technically called haemoglobin to
methaemoglobin, and, later still, to haematin. All of these give different
spectra. The analyst has standard spectra already mounted, and he
invariably looks at the mounted or standard specimen and the suspected
liquid at the same time, placing them side by side, so that a mistake is
impossible. All the red colours in the world, in fact, have been tried,
and, with the exceptions named above, none of them gives a spectrum like
the colouring matter of blood in any of its forms.

But though the spectroscope is a certain discoverer of blood, it can
draw no distinction between human and animal blood. That duty remains to
the microscope.

[Illustration: LITTLE LINKS IN THE CHAIN OF EVIDENCE.--THE CORPUSCLES IN
THE BLOOD OF DIFFERENT CREATURES.

Man. Mouse. Horse. Camel.
Toad. Pike. Pheasant. Pigeon.]

With the microscope can be seen those red corpuscles which, in some
mysterious manner, seize on the oxygen of the air as it passes into the
lungs, shoulder it, so to speak, and rush away with it, like so many
ants, to the remotest parts of the body. Unfortunately, they can only be
seen in blood that has not been very long shed--that is to say, some
weeks or months. To see these, the analyst scrapes the little clot from
the piece of cloth, or wood, or iron, and places it on a slip of glass;
over this he carefully lays the little film called a cover-glass; and
then he gently places, at the edge of the latter, the tiniest possible
drop of water. This gradually insinuates itself, and soon dissolves the
blood clot; and, when the mixture is p