Vol. 6, 1920 PHYSICS: DUANE AND PATTERSON 
515 
Judging from this data alone, there is some question as to whether the 
emission line L^s belongs to the absorption difference Laz-Mai or to 
La^-Mas. 
No emission lines have been observed in the spectrum of either thorium 
or uranium that correspond to the other absorption differences contained 
in the table. 
Four emission lines have been found by Siegbahn and Friman in the 
spectra of both thorium and uranium in addition to those contained in 
our Table 3, namely, l,jS2,Ti, and 72,3. The way in which the frequencies 
of the Hues for different chemical elements vary with their atomic num- 
bers^ and the researches of Webster and Clark^'^ on the voltages required 
to produce the lines in the spectrum of platinum indicate that 1 and ^2 
belong to the same X-ray group as ai and 0:2, and therefore also to the 
absorption Lai. Similarly, 71 belongs to the absorption La2. If these 
lines obey the frequency difference law, there must be critical absorption 
wave-lengths at X = 2.37 X IQ-^ cm. and at X = 2.23 X IQ-^ cm. for 
thorium and uranium, respectively, to correspond with the emission 
line 1. These wave-lengths lie within the range of possible measure- 
ments, but the spectrum in their neighborhood has not been examined 
for critical absorption. Similarly, there should be critical absorption 
wave-lengths at X = 16.2 X 10 ~^ cm. for thorium and at X = 15.8 X 
10"^ cm. for uranium to correspond with the emission lines (32 and 71. 
These wave-lengths, however, lie beyond the range of wave-lengths that 
can be measured at present by means of crystal spectrometers. 
It is well known that Bohr's theory of radiation explains very simply 
the above relation between emission and absorption wave-numbers. 
The relation, however, should not be absolutely exact, if, as Sommerfeld 
supposes, some of the atoms contain M orbits that are elliptic, and other 
atoms of the same chemical element contain M orbits that are circular. 
In this case the L absorption drops should have a complex structure. We 
have not been able to observe such a structure in the drops on our curves, 
and the effect must be exceedingly small, if it exits at all. 
Theoreies have been developed by Rubinowicz^^ and Bohr,^^ according 
to which transfers of electrons between certain pairs of orbits cannot 
take place. If we apply these theories to the data contained in Table 3, 
we find 'that according to both theories the transfers represented by Mai-Lai 
and Ma2-La2 should occur. As a matter of fact, the corresponding lines, 
ai and jSi, are the strongest lines in the L series. According to Rubi- 
nowicz's theory the transfer Ma2-Lai should occur, but according to 
Bohr's theory it should not, unless the atom is in a field of force. The 
corresponding line, 0:2, has been observed, but is relatively very faint. 
According to both theories the transfer Mai-Lai should not occur, and, 
as noted above, the experiments do not indicate conclusively whether the 
line belongs to this combination or not. Both theories allow the transfer 
