Atomic Weight of Radium. 281 
The ratios of these intervals are :— 
Sar Ren . 2960 
BE oe as 2°980 
Bee Rss 2-952 
See ae 295 |: 
Sar eT 2°961 
fee eee 27956 
Gripe shin. 2°958 
(ae 2°943 
oe 2-Sag 
Sy ek 3 OO1 
(on eee 2942 
The mean of these numbers is 2°948 and the ratio of the 
squares of the atomic weights is 2°960. 
Calculating the atomic weight of zinc from that of cadmium 
from these data we obtain :— 
See bora 
2 a ee 65:12 
1 ae 65°35 
DP 65°43 
os ae 64:32 
Gio 5, 65rd 
Chios. ba 35 
(‘Cape er 657)2 
i 66°71 
Si aee Ree 64°88 
Se eee 65°d4 
The mean of these numbers is 65°45: the accepted value 
of the atomic weight of zinc is 65°4. 
This second law must then be admitted to be true in some 
cases, but it is certainly not true generally. It may be, 
however, that in other groups of related elements the distances 
apart of the lines of a pair (or triplet) are proportional to 
the atomic weight raised to some other power than two 
exactly. Thus, for copper and silver the power is 2°4765 ; 
for magnesium, calcium, strontium, and barium 1°6214, and 
for aluminium, gallium, indium, and thallium 2°1127. These 
numbers are calculated from the differences between the lines 
of the series dealt with by the formule of Kayser and Runge. 
_ The law employed by Runge and Precht in calculating 
the atomic weight of radium, that the differences of the 
oscillation-frequencies of homologous doublets is proportional 
to some power of the atomic weight, or that the logarithm of 
