64 
PHYSICS: A. S. KING 
PROC. N. A. S. 
The phenomena may be briefly described by saying that when a tempera- 
ture is reached sufficiently high to produce an absorption spectrum, the 
Hues appearing are those of low-temperature class. As the temperature 
is raised, lines of successively higher classes appear. The lines of low- 
temperature class thus show the largest degree of absorbing power. At 
a given temperature, the absorption spectrum shows fewer lines than are 
given by the emission spectrum at the same temperature. The absent 
lines are of the higher temperature classes, which the vapor is able to emit, 
but which have relatively low absorbing power. For the iron and titanium 
spectra, the absorption spectrum is very similar to that given in emission 
at a temperature about 400° lower. As lines which are strong in the fur- 
nace and faint in the arc are usually lines of large absorptive power, the 
difference between the arc and the dark-line furnace spectrum is very strik- 
ing. 
The dependence of absorptive power on temperature class was well 
brought out by placing the graphite plug beyond the center of the tube 
away from the spectrograph, so that some of the metallic vapor was hotter 
than the plug. It was then possible to obtain emission and absorption 
lines at the same time. The high-temperature lines appeared in emission, 
while the plug was still hot enough to cause those of low-temperature class 
to show as absorption lines. 
Lines which are being emitted by a vapor may fail to appear in absorp- 
tion owing either to a low absorptive power and consequent quenching by 
the continuous spectrum, or to a balancing of emission and absorption 
which may occur when the continuous background is reduced in tempera- 
ture so that its emission is of about the same strength as that from the 
metallic vapor. This latter condition was pro.duced in the furnace by 
withdrawing the plug nearly to the end of the tube. Lines produced by 
the vapor in the central portion were then frequently neutralized by the 
balanced emission and absorption and were not to be seen. In this way 
a line may seem to be totally absent when in reality it is being strongly 
radiated by the vapor. 
A method of producing absorption spectra without the use of a plug may 
be mentioned in this connection. When the furnace is operated in the 
usual way, with a metallic powder or salt in the open tube, the emission 
spectrum increases in intensity and number of lines as the temperature is 
raised, until it compares in richness with that of the arc. A still higher 
temperature causes a rapid increase in a continuous ground, which is 
probably due in part to the long column of dense metallic and carbon vapor 
and partly to light from the walls of the tube reflected by the vapor particles. 
This obliterates the emission lines and we have an absorption spectrum 
consisting of the lines which in the high-temperature emission spectrum 
were self-reversed. In general these are low-temperature lines, although 
they result from a high-temperature condition. This phenomenon may 
