Vol. 6, 1920 
PHYSICS: A. S. KING 
63 
From these facts and from closely similar observations on aviators^ 
during the war, it appears highly probable that in normal persons the blood 
alkali is controlled by the dissolved CO2: more or less alkali being called 
into use in the blood to satisfy the equation above quoted, and to keep 
the Ch of the blood constant. The amount of dissolved CO2 in the blood 
is controlled by the pulmonary ventilation; and fundamentally the venti- 
lation is adjusted to the oxygen partial pressure of the air at the altitude 
at which the person lives. 
Thus, as data from the report of the Pike's Peak expedition^ tend to 
confirm, the partial pressure of oxygen in the lungs, the alveolar CO2, 
the C02(H2C03) dissolved in the blood, and the amount of the blood alkali 
(the so-called "alkaline reserve"), each multiplied by a constant of its 
own, tend to vary in direct proportion to the mean barometer, minus the 
water vapor tension of the pulmonary air (about 45 mm.) at all altitudes. 
1 Haggard, H. W., and Henderson, Yandell, /. Biol. Chem., 39, No. 1, August, 1919 
(163-261). 
2 Henderson, Yandell, Science, New York, N. S., 49, No. 1271, May 9, 1919 (431-441). 
2 Douglas, Haldane, Henderson, and Schneider, London Phil. Trans. Roy. Soc, 
B. 203, 1913 (310). 
A STUDY OF ABSORPTION SPECTRA WITH THE ELECTRIC 
FURNACE 
By Arthur S. King 
Mount Wilson Observatory, Carnegie Institution of Washington 
Communicated by G. K. Hale, December 9, 1919 
The tube resistance furnace has found a place as a useful light-source 
in spectroscopic work, but thus far it has been used mainly for the produc- 
tion of emission spectra. The long column of vapor, whose temperature 
can be controlled, offers interesting possibilities, however, in the field 
of absorption phenomena. In the experiments to be described, the con- 
tinuous background was supplied by a plug of graphite placed in the center 
of the furnace tube, the heated portion of which was 20 cm. long and 12.5 
mm. internal diameter. A close approach to black-body conditions is 
thus obtained, and the emission of the plugged tube is stronger than that 
of the metallic vapor filling the tube. An absorption spectrum results, 
which, by simply removing the plug, may be compared with the emission 
spectrum given by the vapor at the same temperature. 
It is known from observations of reversed emission lines that in a given 
range of spectrum, some lines are more subject to reversal than others. 
The present experiments have shown this to result from a close connection 
between the tendency of a line to reverse and the two factors which enter 
into the classification of lines in furnace spectra, viz., the temperature at 
which a line is first radiated and its response to increase of temperature. 
