34 
PHYSICS: D. L. WEBSTER 
it 
Proc. N. a. S. 
TABLE 1 
ELEMENT 
^72 
^72-^A2 
^A3 
^74 
^74 ~ ^A3 
Tungsten 
1.230i« 
1.20973 
—0.020 
1.08110 
1.05963 
—0.021 
1.02510 
1.02633 
+0.001 
Tungsten 
I.2151' 
1.20973 
—0.005 
Platinum 
1.07011 
1.0726 
+0.002 
Platinum 
1.07212 
1.0726 
0.000 
0.93412 
0.0336 
—0.001 
LxOla 
l.Uoo" 
— U.UOd 
0.901^^ 
A OAOfi 
0.898" 
A A AO 
— 0.003 
0.86111 
0.8d9'' 
+0.008 
Gold 
1.04212 
1.0356 
—0.007 
0.91412 
0.8986 
—0.016 
Mercury 
1.00911 
1.01313 
+0.004 
Thallium 
0.97711 
0.9776 
0.000 
0.84311 
0.8446 
+0.001 
Lead 
0.94811 
0.95313 
+0.005 
O.8I411 
0.8206 
+0.006 
Lead 
0.94914 
0.95313 
+0.004 
O.8I314 
0.8206 
+0.007 
0.78114 
0.7926 
+0.011 
Bismuth 
0.92411 
0.9236 
—0.001 
0.78911 
0.7946 
+0.005 
0.75611 
0.7626 
+0.006 
Thorium 
0.76011 
0.76713 
+0.007 
0.62711 
0.63713 
+0.010 
O.6O711 
0.61013 
+0.003 
Uranium 
0.72111 
0.72613 
+0.005 
0.59111 
0.59813 
+0.007 
0.56711 
0.57113 
+0.004 
Mean 
—0.001 
0.000 
+ 0.006 
From this table it would appear that Ai, A2 and are probably at nearly 
the same wave-lengths as jSs, 72 and 74, respectively, in every element. 
But most of these results were obtained with different apparatus for the 
two kinds of measurement, and there is considerable disagreement. As 
even a slight difference in wave-length one way or the other is of great 
theoretical interest, it seemed desirable to check this point by using a 
screen of H2PtCl6 on blank films. This was attempted in two of the high 
resolving power spectrograms. The results were not very satisfactory, 
owing to the strong absorption by the other substances than platinum in 
the screen, and the presence of the less absorbed radiation of higher fre- 
quencies, in the fifth order, at the most interesting part of the third. But 
they seem to indicate that Ai is within a very few thousandths of an 
Angstrom of (3^, and that A2 is between 72 and 73, 73 being apparently 
the more strongl}^ absorbed of these lines. This is consistent with the 
conclusion reached above, that 7« belongs to L2 and 73 to L3, but the evi- 
dence is unsatisfactory and none was obtained on A 3. This apparatus 
was not designed for high resolving power but it is hoped that this point 
will be settled with the other apparatus now being constructed by Mr. 
F. C. Hoyt for work on the fainter lines. 
Summary of Results. — We have assigned the lines to the various series 
as follows: 
To Li, /, a2, ai, ^2, jSe and (3^; 
To L2, rj, (3a, /3i, 7i, 72 and perhaps ^3; 
To L3, 74, probably 73 and perhaps /Ss. 
Unassigned, all of Dershem's and Overn's faint lines except /Se, which is 
probably one of them. 
