304 
THE QUESTION OF CARBON 
LINE STARS. 
“| HE spectrum of carbon is one which is subject to very great 
changes when examined under different experimental con- 
ditions, and an acquaintance with these variations is essential to 
an adequate discussion of the spectra of the heavenly bodies, 
IN BRIGHT 
NATURE 
| JANUARY 28, 1897 
vapour at a relatively high pressure; and among the most 
notable differences from the lower pressure spectrum at the bot- 
tom of the diagram, is the enhancement of the more refrangible 
part of the group of flutings commencing at 4737. This 
is still more marked in the spectrum at intermediate 
pressures, as shown in the middle spectrum in Fig. 2. In 
short, the brightest part of this group now agrees, very 
nearly, with the blue band 
bisa 
Headed 
| ene | 
i | 
a eat 
ames 
of some of the bright line 
L stars. It is very difficult to 
Fr] estimate the middle of such 
CnH,+co 4 broad, diffuse band as 
} that in question; but the 
wave-length of the brightest 
| Comet part is approximately 4685. 
“ : f| —-| _As Dr. Vogel had pre- 
: } | viously done for comets, I 
3 if | Cy, Hy was particularly careful to 
a 2 eae] point out that the carbon 
band in the bright line 
Fic. 1.—(1) Spectrum of mixture of hydrocarbons and oxy-carbon obtained from meteorites, with small coil and stars was not seen under the 
Jar, (2) Spectrum of Comet III. 1881. (3) Spectrum of gases from meteorites, with large coil and jar (Vogel). same conditions as that in 
Every one knows the considerable development of the spec- 
trum of carbon in most cometary phenomena; and there is a 
band which Dr. Vogel some years ) attributed to carbon, 
although it does not coincide with the most familiar carbon 
spectrum, that of the Bunsen burner. Dr. Vogel gave his 
reasons for this allocation, and illustrated them by.a diagram ! 
in which it is shown that in the spectrum of the comet, the blue 
band has its maximum about 470, and fades away nearly equally 
in both directions. If one’s knowledge of the carbon spectrum 
were limited to that of the Bunsen burner, indicated at the 
bottom of the diagram, the comet band could not be ascribed 
to carbon. 
But another spectrum of carbon, obtained by Dr. Vogel 
(given at the top of the diagram), shows the blue band of 
exactly the same form, and in the same position as that in 
the comet. Hence, Dr. Vogel argued that the blue band in 
the comets, though not coinciding with the carbon group at 
wave-length 4737, seen in a Bunsen flame, was still due to 
carbon, 
When I was discussing the 
the Bunsen flame.? 
I next append some ob- 
servations of the band in both comets and bright line 
Stars : ; 
Maxi srichte: 
Comet. / mete Observer. Star. I rightest Observer. 
of band. |. , part. 
Winnecke. 186: 469 | Huggins ) : Ricks 
Comet IV. 1873 469 Vogel { No. ‘ $99 Huggins 
Comet III. 188r 468"5 Vogel in “ re 
Comet III. 1881 4 Copeland |) j N4oor 468 | Huggins 
Comet IV. 1881 470 Copeland 
brightest part of a diffuse band, it will be seen that the bands 
in two of the stars are exactly coincident with bands which have 
been measured by trustworthy observers in three comets. In 
the fifth comet, named above, the variation in the wave-length 
of the band is not greater than that between two individual 
measures in stars. 
spectrum of the bright line stars 
in my general survey, a band in 
very nearly, if not absolutely 
the position of the comet- 
ary band, was found recorded. 
Most unfortunately I had com- 
pletely forgotten Dr. Vogel's 
paper of 1881, and I set to work 
to study its origin for myself. 
In the course of the previous 
thirteen years I had taken some 
hundreds of photographs of the 
spectrum of carbon compounds 
under a great variety of con- 
ditions; and I was driven to 
carbon because one of the most 
conspicuous features of the spec- 
trum of many of the bright line 
stars is a broad blue band, a part 
of which falls within the limits 
of the group of carbon flutings 
at 4737, which is seen in the 
Bunsen, although its brightest 
part in the stellar spectra is 
about wave-length 468 that is, 
some distance further towards 
the blue than the brightest 
part of the Bunsen group. Forgetting Vogel’s prior labours, I 
looked through my photographs, and found what he had found, 
that, under certain conditions, the maximum of the band is 
shifted, under some conditions of pressure and temperature, 
from 4737 to about 4685. One of my photographs, taken in 
December 1886, is reproduced in Fig. 2. 
The spectrum at the top of Fig. 2 is the spectrum of alcohol 
w 
1 Potsdam, Observations, 
NO. 1422, VOL. 55] 
r, Ih. p. 173. 
2,—Spectrum of alcoho! vapour. 
) 
| 
| 
(x) Highest pressure. (2) Intermediate pressure. 
pressure, 
(3) Lowest 
The fact that, so far as I know, this explanation, the whole 
credit of which is due to Dr. Vogel, has never been ‘called in 
1 In 1888 I wrote: ‘* This band is evidently the bright band of carbon, 
commencing at 474, with a maximum about 468, as observed and photg- 
graphed at Kensington” (Rey. Soc. Proc., vol. xliv. p. 35). Later in the 
same year I added: ‘*‘Itis necessary to-state that the maximum luminosity 
of the blue band, under some conditions, is about 468, . . . The conditians 
under which this band has its maximum luminosity at 468 in Geissler tubes 
seem to be those of maximum conductivity” (Rey. Sec. Prod., vol. xlv. 
p- 167). 
