DR. A. SCHUSTER OH THE SPECTRA OF METALLOIDS. 
47 
5. The Continuous Spectrum of Oxygen. 
Wullner (Pogg. Ann., 137, p. 350) has shown that when the pressure in his vacuum 
tube is gradually increased, the background begins to be illuminated at a pressure of 
280 m.m. This illumination, according to him, first extends from the solar line E to a 
point about half way between F and G. As the pressure was increased, the continuous 
spectrum extended more towards both sides, but chiefly towards the blue. I do not 
know whether this continuous illumination of the background is due to the same 
cause as the continuous spectrum, which is seen under the circumstances presently 
to be described. There is nothing antagonistic to known facts in the assumption 
that the vibration of the gas, though chiefly taking place in certain definite periods, 
can be disturbed by mutual impacts in such a way that a slight admixture of vibra¬ 
tions of all periods is produced which increases as the pressure is increased. At any 
rate, if this is the cause of the continuous spectrum which is seen under great pressure, 
it has nothing to do with the continuous spectrum which appears by a reduction of 
temperature, and which is not the background to bright lines, but forms the whole 
visible spectrum. The following facts seem to me to prove that oxygen at the 
lowest temperature at which it is luminous gives a continuous spectrum :— 
1. The wide part of a Geissler tube filled with pure oxygen generally shines 
with a faint yellow light; when analysed by the prism, this yellow light gives a 
continuous spectrum. 
2. Experiment 6.—A spark was passed through oxygen gas at the atmospheric 
pressure. The maximum striking distance of the coil was about one inch. No jar 
was included in the circuit. The poles were about 4 m.m. apart. A lens concentrated 
the image of the spark on the slit on the spectroscope. Under these circumstances 
the elementary line spectrum of oxygen, which is the spectrum commonly observed 
at atmospheric pressure, is seen. But when the break of the induction coil was put 
out of adjustment, so that the spark decreased considerably in strength, it took a 
yellow colour, and the spectrum was perfectly continuous. The phenomena were 
very striking when I kept the wheel of the break in my hand, while looking through 
the spectroscope. A slight turn would change the line spectrum into a continuous 
spectrum, and vice versd. When the line spectrum was seen, the spark was con¬ 
siderably stronger and the poles became red hot. When the continuous spectrum 
was seen, the poles were not sufficiently heated to become luminuous. The maximum 
intensity of the light was in the yellowish green, and decreased on either side. 
I attach some importance to this experiment, as the sudden conversion of the line 
spectrum into a continuous spectrum cannot I think, be explained on any other 
hypothesis than of molecular combination. 
3. An interesting observation of Becquerel, which he describes in a foot-note 
(‘La Lumiere,’ vol. i., page 196), bears on this point. He says that an excess of 
oxygen in the oxyhydrogen flame makes itself visible by a yellow streak of light 
