360 Transactions of the Boyal Society of South Africa. 
The zonal structure was not developed by the action of the radium, as 
it can be traced in the control tube, but not of course very distinctly, as 
the glass was not perfectly homogeneous, a sort of schistose structure 
being present. There was evidently some obstacle to the free passage of 
the a-particles, encountered at the junction of the layers of material. 
The colouration extends into the glass for over 2 mm., and at all 
points of optical discontinuity a sharp change in the colouring power 
occurred, as though there was a strongly resistant layer at the junction of 
the two zones. From Professor Eutherford's paper {Phil. Mag., Jan., 
1910) one might conclude that the maximum depth of the colouration 
should be proportional to the range of the a-particles, and he gives the 
number 0*041 as the distance to which the a-particles should penetrate 
in glass. 
If the a-particles j5red into the innermost layer of the tube remain 
embedded there, could they impart a radio-activity to the layer, which in 
its turn would send off more particles, thus making the second layer active, 
and so on ? This point might have been elucidated had the tube been 
examined regularly during the two years. An experiment has been 
started in which the tube will be examined at intervals. 
The purple colouration I take to be due to some manganese compound, 
formed by the action of the rays on the glass in the one case, and in the 
other the increase in depth of the green colouration, to the possible reduc- 
tion of a ferric compound to a ferrous one. Manganese in small quanti- 
ties is frequently added to glass used for making tumblers, etc., in order to 
counteract the green colour due to iron compounds ; the intense colouring 
power of the higher oxides of manganese is well known. 
Plugging the tube with cork seemed quite as efficacious as hermetically 
sealing; the colouring did not proceed much beyond the cork. The 
photograph. Fig. 8, is of a corked tube, and the intensity of the colour- 
ation should be relatively greater than that shown, as the coloured portion 
of the tube was only 1*5 cm. in length, whilst the control tube was 
3-2 cm. 
The amount of emanation present in the tube cannot have been very 
large at any time, as, assuming the purity of the radium salt to be 
yq^qo, this would give off a total amount of emanation which, in two years, 
would have approximately amounted to yqVo X '3 x '1 x "730 cb. mm. of 
helium."' As this was continually decaying, and, also, in the case of the 
corked tubes, escaping, the amount present at any one time must have 
been very small, although it is most probable that the action is a 
cumulative one. The nature of the colouration clearly depends on the 
* This would be the amount of emanation from a gramme of the salt. The actual 
amount used was only a few centigrammes. 
