When the spectra of thousands of stars were recorded for study, such a
variety was found that some system of classification was necessary. The
commencement of such a system was made by Secchi in 1863. It was based
on the observed relation between the color of a star and the general
character of its spectrum.
Arranging the stars in a regular series, from blue in tint through
white to red, it was found that the number and character of the
spectral lines varied in a corresponding way. The blue stars, like
Sirius, Vega and α Aquilæ, though they had the F lines strong, as well
as the two violet lines H, had otherwise only extremely fine lines.
On the other hand, the red stars, like α Orionis and α Scorpii, show
spectra with several broad bands. Secchi was thus led to recognize
three types of spectra, as follows:
The first type is that of the white or slightly blue stars, like
Sirius, Vega, Altair, Rigel, etc. The typical spectrum of these stars
shows all seven spectral colors, interrupted by four strong, dark
lines, one in the red, one in the bluish green, and the two others in
the violet. All four of these lines belong to hydrogen. Their marked
peculiarity is their breadth, which tends to show that the absorbing
layer is of considerable thickness or is subjected to a great pressure.
Besides these broad rays, fine metallic rays are found in the brighter
stars of this type. Secchi considers that this is the most numerous
type of all, half the stars which he studied belonging to it.
[Illustration: FIG. 1. SPECTRUM OF SIRIUS.]
[Illustration: FIG. 2. SPECTRA OF α AURIGÆ AND SUN.]
[Illustration: FIG. 3. SPECTRA OF α BOOTIS AND β GEMINORUM.]
The second type is that of the somewhat yellow stars, like Capella,
Pollux, Arcturus, Procyon, etc. The most striking feature of the
spectrum of these stars is its resemblance to that of our sun. Like
the latter, it is crossed by very fine and close black rays. It would
seem that the more the star inclines toward red, the broader these
rays become and the easier it is to distinguish them. We give a figure
showing the remarkable agreement between the spectrum of Capella, which
may be taken as an example of the type, and that of the sun.
The spectra of the third type, belonging mostly to the red stars, are
composed of a double system of nebulous bands and dark lines. The
latter are fundamentally the same as in the second type, the broad
nebulous bands being an addition to the spectrum. α Herculis
may be taken as an example of this type.
[Illustration: FIG. 7. SPECTRUM WITH BOTH BRIGHT AND DARK LINES.]
It is to be remarked that, in these progressive types, the brilliancy
of the more refrangible end of the spectrum continually diminishes
relatively to that of the red end. To this is due the gradations of
color in the stars.
To these three types Secchi subsequently added a fourth, given by
comparatively few stars of a deep red color. The spectra of this class
consist principally of three bright bands, which are separated by dark
intervals. The brightest is in the green; a very faint one is in the
blue; the third is in the yellow and red, and is divided up into a
number of others.
To these types a fifth was subsequently added by Wolf and Rayet, of
the Paris Observatory. The spectra of this class show a singular
mixture of bright lines and dark bands, as if three different spectra
were combined, one continuous, one an absorption spectrum, and one an
emission spectrum from glowing gas. Less than a hundred stars of this
type have been discovered. A very remarkable peculiarity, which we
shall discuss hereafter, is that they are nearly all situated very near
the central line of the Milky Way.
[Illustration: FIG. 4. SPECTRA OF α CYGNI AND α TAURI.]
[Illustration: FIG. 5. SPECTRUM OF α ORIONIS.]
[Illustration: FIG. 6. SPECTRUM OF γ CASSIOPEIÆ]
Vogel proposed a modification of Secchi’s classification, by
subdividing each of his three types into two or three others, and
including the Wolf-Rayet stars under the second type. His definitions
are as follows:
Type I is distinguished by the intensity of the light in the more
refrangible end of the spectrum, the blue and violet. The type may be
divided into three subdivisions, designated _a_, _b_ and _c_:
In I_a_ the metallic lines are very faint, while the hydrogen lines are
distinguished by their breadth and strength.
In I_b_ the hydrogen lines are wanting.
In I_c_ the lines of hydrogen and helium both show as bright lines.
Stars showing this spectrum are now known as helium stars.
According to Vogel, the spectra of type II are distinguished by having
the metallic lines well-marked and the more refrangible end of the
spectrum much fainter than in the case of type I. He recognizes two
subdivisions:
In II_a_ the metallic lines are very numerous, especially in the yellow
and green. The hydrogen lines are strong, but not so striking as in
I_a_.
In II_b_ are found dark lines, bright lines and faint bands. In this
subdivision he includes the Wolf-Rayet stars, more generally classified
as of the fifth type.
The distinguishing mark of the third type is that, besides dark lines,
there are numerous dark bands in all parts of the spectrum, and the
more refrangible end of the latter is almost wanting. There are two
subdivisions of this type:
In III_a_ the broad bands nearest the violet end are sharp, dark and
well-defined, while those near the red end are ill-defined and faint.
In III_b_ the bands near the red end are sharp and well-defined; those
toward the violet faint and ill-defined. The character of the bands is
therefore the reverse of that in subdivision _a_.
This classification of Vogel is still generally followed in Germany
and elsewhere. It is found, however, that there are star spectra
of types intermediate to all these defined. Moreover, in each type
the individual differences are so considerable that there is no
well-defined limit to the number of classes that may be recognized. At
the Harvard Observatory a classification quite different from that of
Vogel has been used, but it is too detailed for presentation here. The
stars of type II are frequently termed Capellan stars, or Solar stars.
Certain stars of type I are termed Orion stars, owing to the number
of stars of the type found in that constellation. The stars which
show the lines of helium are known as helium stars. We mention these
designations because they frequently occur in literature. It would,
however, be outside the object of the present work to describe all
these classifications in detail. We therefore confine ourselves to a
few illustrations of spectra of the familiar types described by Secchi
and Vogel.
There are many star spectra which cannot be included in any of the
classes we have described. Up to the present time these are generally
described as stars of peculiar spectra.
As the present chapter is confined to the more general side of the
subject, we shall not attempt any description of special spectra.
These, especially the peculiar spectra of the nebulæ, of new stars, of
variable stars, etc., will be referred to, so far as necessary, in the
chapters relating to those objects.
The most interesting conclusion drawn from observations with the
spectroscope is that the stars are composed, in the main, of elements
similar to those found in our sun. As the latter contains most of the
elements found on the earth and few or none not found there, we may
say that earth and stars seem to be all made out of like matter. It
is, however, not yet easy to say that no elements unknown on the earth
exist in the heavens. It would scarcely be safe to assume that, because
the line of some terrestrial substance is found in the spectrum of a
star, it is produced by that substance. It is quite possible that an
unknown substance might show a line in appreciably the same position
as that of some substance known to us. The evidence becomes conclusive
only in the case of those elements of which the spectral lines are so
numerous that when they all coincide with lines given by a star, there
can be no doubt of the identity.