Voi^. 6, 1920 
PHYSICS: F. C. HOYT 
643 
In the present work these results have been confirmed. This leaves as 
still doubtful in platinum 72, 73, 75, and 183. Of these still unclassified 
lines, 72 and 73 are so faint that although well resolved in all pictures, the 
results are still uncertain. The possibility of a fourth critical potential, 
as might be indicated by Sommerfeld's theory, makes these lines par- 
ticularly interesting. 75 is too weak to show except at high voltages, 
with any degree of certainty. For all of these very weak lines it will be 
necessary to cut down the general radiation by using greater dispersion 
(plate farther from crystal) than is at present possible without unduly 
weakening the lines themselves by absorption in the air. For this pur- 
pose a vacuum spectrograph is now under construction. 
/33 is so close to ^2 in platinum as to make work on this line very difficult, 
but in tungsten it lies midway between and /?2. With the platinum 
anticathode used, where the tungsten lines also showed, the coincidence of 
W 7i with Pt 183 was another complication, so that results were obtained 
on this line only in tungsten. 
Pt has given rather anomalous results. Webster and Clark's^ ioniza- 
tion curves showed a critical potential near ¥2, but the uncertainty here 
was greater than with the stronger lines. In Webster's^ photographs, 
however, it showed well below V2, but the presence of a line here could be 
ascribed to the near coincidence of Ir ^2 with Pt ^i,. With the present 
anticathode there were no iridium lines, and yet 184 showed below 1^2- 
We can account for this if we regard it as the combined line and jSy. 
The work with tungsten showed /5 7 to be Li and 184 to be L3, so that the 
part showing below V2 is simply ^-j. That the ionization work indicated 
a critical potential the same as that of can only be ascribed to the uncer- 
tainty of the measurements on a line as weak as /34 and to the disturbing 
effect of /57, which would tend to decrease the apparent critical potential. 
The results on tungsten are well illustrated by the reproductions in 
Fig. 2, c, d, e, and /. In c, d, and e, the product of current through the 
tube, time of exposure, and square of the voltage was made the same in 
all three. In / the time of exposure was shorter. The critical potentials 
calculated for tungsten would be : 
Vi 10.2 kilo volts [Quantum potential of 185 1 
V2 11.6 kilovolts [Quantum potential of 72] 
V3 12.0 kilovolts [Quantum potential of 74] 
These constant voltage pictures show us how /34 and (3 7 are to be classified. 
In / at 13 kv. we have them showing at nearly equal intensities. At 
12.5 kv. /34 is noticeably weaker, and at 12 kv. (F3) 184 is out completely, 
assigning it to L3, while ^7 is still strong. At 11.5 kv., below V2, still 
shows, although faint, assigning it to Li. 
That jSs and /Se belong together in Li is shown by the fact that they 
show with constant intensity ratio down to IIV2 kv. 
183 disappears at 12 kv. placing it in L3. 
