GASEOUS SPECTRA IN VACUUM TUBES. 99 
any oxygen either, if that had to be derived, simultaneously with the hydrogen, 
from the decomposition of water; and yet the bands of Nitrogen (said by the 
opposite school to be bands of an Oxide merely, because they were bands and 
not lines) were magnificently developed, broad, spreading, and true bands. 
So also with the Carbon bands of the same tube, derived from its Cyanogen 
combination of Nitrogen and Carbon being dissociated. The Carbon line 
spectrum consists of only eleven lines, and never shows except in a very 
powerful and condensed spark. But its band spectrum can be called up by 
any, even the smallest, spark ; and that band spectrum (said by ANGsTRom and 
THALEN to be necessarily belonging to an oxide of carbon because it is in 
bands) was brilliantly visible in this tube, where there could have been no 
oxygen for the carbon to oxidise with, 7.¢., if, as before, the oxygen had to be 
derived from the decomposition of water; and the absence of hydrogen lines, 
inherently far brighter than those of oxygen, proved that such decomposition 
had not taken place. Hence, after virtually clearing my observations from 
Hydrogen and Oxygen, I proceeded in the same manner to get them free from 
traces of Nitrogen, Carbon, and the peculiar compound Carbo-hydrogen,— 
wherever these gases had no right of intended standing place. But still there 
were many lines left, and some of them very pronounced, common to several 
tubes with most diverse labels. What lines could they be ? 
By far the greater number turned out to be low temperature lines of 
Hydrogen. Almost a new class of lines in the spectroscopic world ; even 
denied by some persons, yet clearly visible simultaneously with the four great 
and almost classical Hydrogen lines; which are properly high temperature 
emanations, but of such an intensity of vital force, as to be capable of living 
on, down through low temperatures also. And whenever they, the high- 
temperature lines, appeared in my low-temperature, but brilliantly lighted 
end-on tubes, there and then, in nearly the same proportions of relative 
intensity, appeared the crowds of the new low-temperature lines; not three 
or four only, but rather three or four hundred. 
This discovery is involuntarily, but exemplarily given in a few of its principal 
features in the Tables of 
APPENDIX II. (see the end of this paper) ; 
which Tables show likewise the degree and manner in which impurities are 
distributed among the several tubes ; an instructive thing in itself. 
I have also prepared, but refrain here, for the cost’s sake, from printing the 
practical deductions from Appendix I., in the shape of a set of resulting 
standard Tables of the places of leading features of gaseous spectra. The 
foundation for these places is always taken from the admirable observations of 
