also made some trials in this direction. In the _Philosophical
Transactions_ for 1814 (p. 264), he says:

"By some experiments on the light of a few of the stars of the first
magnitude, made in 1798, by a prism applied to the eye-glasses of my
reflectors, adjustable to any angle and to any direction, I had the
following analyses:

"The light of _Sirius_ consists of red, orange, yellow, green, blue,
purple, and violet. _[alpha] Orionis_ contains the same colors,
but the red is more intense, and the orange and yellow are less
copious in proportion than they are in _Sirius_. _Procyon_ contains
all the colors, but proportionately more blue and purple than
_Sirius_. _Arcturus_ contains more red and orange, and less yellow
in proportion than _Sirius_. _Aldebaran_ contains much orange and
very little yellow. _[alpha] Lyrae_ contains much yellow, green,
blue, and purple."

Here the essential peculiarities of the spectrum of each of the stars
investigated by HERSCHEL is pointed out, and if we were to use his
observations alone to classify these stars into types, we should put
_Sirius_ and _Procyon_ into one type of stars which have "all the
colors" in their spectra; _Arcturus_ and _Aldebaran_ would represent
another group of stars, with a deficiency of yellow and an excess of
orange and red in the spectrum; and _[alpha] Orionis_ would stand as
a type of those stars with an excess of red and a deficiency of orange.
_[alpha] Lyrae_ would represent a sub-group of the first class.

HERSCHEL'S immediate object was not classification, and his observations
are only recorded in a passing way. But the fact remains that he
clearly distinguished the essential differences of the spectra of these
stars, and that he made these observations in support of his statement
that the fixed stars, "like the planets, also shine with differently
colored light. That of _Arcturus_ and _Aldebaran_, for instance, is as
different from the light of _Sirius_ and _Capella_ as that of _Mars_ and
_Saturn_ is from the light of _Venus_ and _Jupiter_."

Of course, no special discovery can be claimed for him on these few
instances. We can see, however, a good example of the manner in which he
examined a subject from every side, and used the most remote evidence
exactly in its proper place and time.

_Researches on the Variable Emission of Light
and Heat from the Sun._

It is cer