280 
Dr. Marshall Watts on the 
of the principal series for the values 5 and 6, and | therefore 
consider them to be homologous. 
Cesium. 
ne an 25714°67 
25794°94 
ie ee 27646°66 
27686°72 
Rubidium. 
27841°66 
27876°66 
29841°57 
29860°91 
Potassium. 
29006°61 
29015°03 
3107752 
31082°25 
Assuming that these lines are homologous, and using them 
to calculate the atomic weight of rubidium from that of 
potassium 39°15 and that of cesium 133, we obtain 
i), oo LS 
pie Oh 
m=0O...... apie 
85°70 
Mean 85°43 instead of 85:2. 
I must, therefore, respectfully maintain that in some cases 
my first lawis very nearly true. In other cases it is certainly 
not true, and must be modified. 
In regard to the second law, that the difference in 
oscillation-frequency of homologous doublets is proportional 
to the square of the atomic weight, this is also true in some 
cases, notably with zinc and cadmium. ‘The following table 
shows the differences between the first and third lines in the 
corresponding triplets of cadmium and zinc, in which * 
denotes lines of the first subordinate series, and + lines of the 
second subordinate series, the figure before the * or T giving 
the value of m in the formule of Kayser and Runge :— 
Cadmium. Zine, 
"ye age ee lw hea) 78°7 
es ee, TOL a1 
5 fea e112 iow 
Ty ee 16993 Jiao 
TT phe t715°0 578°6 
Dts ea ee 1704:0 576°4 
i eee 1a72 Dione 
bei 1704°7 Sy) 
tip ees 1664°9 586°4 
Spanien 16074. 546°3 
8t 1700°5 Done 
