Vol,. 6, 1920 
PHYSICS: D. L. WEBSTER 
33 
mica films, and behave like Pt (3i and Pt ^2, the changes in relative intensi- 
ties with voltage being very striking in both cases. 
^3. This line is obscured by ^2 except in the calcite films. Its behavior 
there is similar to that of the L2 lines and quite different from that of the 
Li. It must be in either L2 or L3. 
Xi. The films show this line clearly when X^^^ = 0.910, but not at all 
at 0.945, confirming the ionization work that assigned it to L2. 
72 and 73. These two lines are not resolved except in three films at 
35, 25 and 18 kv. In these the tube was removed 3 meters from the mica 
instead of 75 cm. In this case the exposure at 35 kv. was 7 hours and at 
18 kv., 30 hours. In these films 72 and 73 are just resolved in the third 
order and well separated in the fifth. These films were taken for three 
purposes, to confirm the existence of (S^, Ir /5i and Ir (32, to get the absorp- 
tion spectrum discussed below, and to see if any difference could be found 
in the intensity ratio of 72 and 73 at different voltages. 73 in each film is 
fainter than 72, and it appeared slightly more so at 18 kv. than at 35, 
but the difference in the ratio was too small to base any definite conclusions 
on it, though if it exists they must belong to different series. The low 
resolution films show the combined line 72,3 well at high potentials, faintly 
but definitely at X^/„ = 0.895 A, and it is barely visible under the best 
conditions at 0.907. Since X^^ = 0.900 A, this line cannot be in the same 
series with 74 but must have a critical wave-length longer than 0.907 
and, therefore, presumably belong to L2 where X2:2 = 0.935 A. Now 
the absorption spectra discussed below indicate that 73 is distinctly more 
absorbed than 72, showing an absorption limit, A2, corresponding to L2, 
lying between them. This apparently means that 73 cannot belong to 
L2 and, therefore, that 72 is probably the line showing at \min = 0.907 A. 
But it does not mean definitely that 73 has the same critical potential as 
74, although the natural assumption is that it has. This is important in 
connection with Sommerf eld's theory to be discussed in the next paper. 
74. This line is so near the bromine absorption limit that any work on 
it is uncertain, and with the mica spectrograph 3 74 is obscured by 2 Ir ai 
and 5 74 by 4 But since \yi<\L2 or it can hardly belong to either 
of these series, and must be ascribed to L3. 
The Faint Lines. — In this class are included the lines found only by 
Dershem and Overn, in tungsten. For these, further work will be neces- 
sary, with high resolving power and very long exposures. 
The Critical Points. — The question arises whether the critical wave- 
lengths determined by potentials are identical with the absorption limits, 
as they seem to be in the K series. The critical wave-lengths found by 
Clark and me for Li and L2 were identical, within limits of error, with the 
wave-lengths of 185 and 72, respectively, provided h = 6.554 X 10 
erg sec, and, therefore the results of various measurements of absorption 
limits are compared with these lines in the following table : 
