( 
34 McCiretianp anpd Hacxerr—NSecondary Radiation from Compounds. 
change of slope as we pass from one period to the next. At first sight these 
curves indicate a possible peculiarity for the period containing barium, inasmuch 
as the secondary radiation from the last element (iodine) of the preceding period 
is very little less than that from the first element tested (tungsten) in the 
following period. 
If, therefore, the elements of the period containing barium are to take their 
place in the general scheme, little room is left for increase of secondary radiation 
as we pass along this period in the direction of increasing atomic weight. This 
is exactly the result indicated by the comparatively few determinations we have 
been able to make for barium, cerium, and didymium. ‘TTiere is scarcely any 
difference between the secondary radiations from the elements of this period; 
but the period as a whole takes what is its proper place according to the previous 
results. On the same scale as that employed in fig. 2, the values found for 
barium, cerium, and didymium, are 86, 86°5, and 86:5, respectively. This 
period as a whole, therefore, does not present any exception to the general rule 
that the secondary radiation increases as the atomic weight increases; but it is 
different from the other chemical periods, inasmuch as there is not a marked 
increase of secondary radiation as we pass along the period in the direction of 
increasing atomic weight. 
In discussing the results in the previous paper from the point of view that the 
atoms of all elements are composed of electrons, it was pointed out that as we go 
from one element to another of greater atomic weight in the same period, the 
additional electrons necessary to form the heavier atom are added on in a way 
that increases considerably the secondary radiation; but in passing from one 
period to the next, the additional electrons must be added on in some different 
way which is not so effective in increasing the secondary radiation, thus 
producing the change of slope in passing from one period to the next. Also, 
since the chemical properties change from element to element in any period, we 
see that the method of adding on additional electrons which produces a rapid 
increase of secondary radiation also produces a marked change of chemical 
properties. From this point of view the behaviour of the period containing 
barium is very interesting. Here we have an exception to what happens in the 
case of the other periods; in this period there is practically no increase of 
secondary radiation from element to element. But from the chemical point of 
view, this period presents an exactly parallel exception; in this period there is 
so little difference in the chemical properties of the different elements that it is 
difficult to separate them. 
As the only element in the first chemical aardied previously tested was 
carbon, some further determinations were made with compounds in the same 
way as above, in order to fix the slope of the curve corresponding to this period. 
