Vol. 6, 1920 
PHYSICS: DUANE AND STEN STROM 
483 
The wave-numbers of the three critical absorptions in the L series of 
tungsten are 
Lai = 0.8240 ±1 La2 = 0.9323 ^ ^ Las = 0.9766 30, 
each multiphed by 10^. Subtracting these from the critical absorption 
wave-number in the K series and comparing the differences with the wave- 
numbers of the three a emission lines we get 
It appears that in each case the difference between the K critical ab- 
sorption wave-number and one of the L critical absorption wave-numbers 
equals the wave-number of one of the a emission lines to within the limits 
of error of the measurements. This agrees with the results obtained 
last year. If the above mentioned effect due to elliptic orbits exists 
at all, it appears to be too small to detect, with our present methods of 
measurements, in atoms of as high an atomic number as that of tungsten. 
The existence of the third faint line, ^3, and the agreement of its wave- 
number with the difference between the Ka and Las wave-numbers is 
a point of some theoretical interest. According to Sommerf eld's theory 
of elliptic orbits the Kai line is due to the transfer of an electron from the 
L orbit to the K orbit, if the L orbit is circular, and the Kq;2 line is due to 
a similar transfer, if the L orbit is elliptic. The difference between the 
wave-numbers of Koii and Kq;2 calculated from the formulas he gives 
{Atombau und Spektrallinien, chapter 5) amounts to 0.1072 X 10^, and 
this agrees very well with our experimental value for the same difference, 
namely, 0.1080 X 10^. There appears to be no explanation on this theory 
for the third critical absorption, Las, in the L series, nor for the third 
line KcKs in the K emission series. It would seem to be necessary to as- 
sume that there are at least two L orbits. 
The peak corresponding to Kas seems to be quite well marked on the 
curves representing spectra of higher order than the first, especially on 
that representing the fourth order. The line is very faint, having only 
about V25 of the intensity of the Kcci line. 
The Kj8 line is supposed to be due to electrons falling from the M orbit 
(or orbits) into the K orbit. We can not test the above frequency differ- 
ence law in this case directly, for the critical absorption frequencies have 
not been measured in the M series of tungsten. Dr. Stenstrom, however, 
has observed three critical absorption wave-lengths in the M series of 
thorium and uranium. He has measured also the wave-lengths of' emis- 
sion lines in the M series of uranium, thorium and tungsten. Since 
the relative positions of the critical absorption and certain of the emission 
lines are about the same in the high frequency spectra of the different 
Wave-Numbers, 1/\ X 10 
Ka — Lai = 4.792 ± 0.003 
Ka — La2 = 4.685 ± 0.003 
Ka — La3 = 4.639 =t 0.OO6 
Kai = 4.7938 ± 0.001 
Ka2 = 4.6858 ± 0.0007 
Kaa = 4.65 ± 0.02 
