Mr Kaye, The Selective Absorption of Rontgen Rays. 243 
In column 4 (under aluminium screen) is given a set of com- 
parative values for the radiations (still keeping Pt = 100) from 
which the softest rays have been sifted out by an aluminium 
screen 1 mm. thick. Such a screen cut down the radiation from 
platinum about 35 times. 
It is very evident that the radiation from an anticathode of 
low atomic weight, e.g. iron or nickel, is largely made up of rays 
that are very soft in character. Such metals have their relative 
radiation values largely reduced when the screen is used. On the 
other hand, metals of high atomic weight such as lead and silver 
have radiation values which are increased when we deal only with 
the harder type of Rontgen rays. These metals obviously give off 
a larger proportion of hard rays than the metals of less atomic 
weight. 
The outcome is, that all the relative radiation values, for rays 
which have passed through an aluminium screen, now follow the 
order of the atomic weights of the anticathode metals ; indeed the 
two exhibit an approximately linear relationship. 
In the last column of all, the ratio of the radiation value to 
the atomic weight is given and, as will be seen, is roughly 
constant. 
The case of silver must be mentioned. It amalgamated with 
the mercury vapour from the pump before any measurements could 
be made. It is hoped subsequently to make an analysis of the 
surface of the silver, but it is very probable that in the presence of 
a continual supply of mercury vapour the amalgam that silver forms 
• A XT mu . • • u. r .U- • 108 + 200 
is AgHg. Ihe mean atomic weight oi this is ~ - = 154, 
and dividing this number into the radiation value for the silver, it 
will be seen that the quotient comes into agreement with the 
other values tabulated in the last column. 
In the remaining columns are given the relative radiations 
which are transmitted through screens of platinum, copper, nickel 
and iron. 
Consider the case of the platinum screen and compare the 
values with those obtained in the case of the aluminium screen. 
It was arranged that the radiation from the platinum anticathode 
should be cut down as much with the platinum screen as it was 
formerly by the aluminium screen, so that rays of equal hardness 
were employed. It will be noticed that again giving platinum the 
value 100, the values for all the other metals are considerably less 
than they were when the aluminium screen was used. The value 
for lead is now actually less than that for platinum, but with this 
exception, the order is undisturbed. Evidently a platinum screen 
shows a special transparency to Rontgen rays from a platinum 
anticathode. This abnormal transparency is shared in less degree 
