362 Transactioiis of the Boyal Society of South Africa. 
The foregoing was written in August, 1910, and since then observa- 
tions have been taken upon other tubes in which impure radium salt had 
been sealed. The glasses used were : (1) a greenish coloured barometer 
tubing, (2) white thermometer tubing, (3) Jena combustion tubing, (4) 
tubing of the same kind used in the experiments detailed, (5) ordinary soft 
glass tubing. After introduction of the radium preparation the tubes were 
plugged with cork, and small cover-glasses fastened on the ends with 
Canada balsam. Control samples of tubing were cut off at the same time, 
which were kept in the dark. These were made of various lengths, so 
that any increase in the depth of colouring produced by the action of the 
radium could be contrasted with the original depth of colour by comparing 
the experimental tube with others of different length in the manner usual 
in colourimetric methods. 
1. The length of the experimental tube was 3'3 cm., small plugs of 
cork 0*3 cm. closed each end, and the radium salt occupied about 0*3 cm. 
The control tubes were 4, 5, 6, and 8 cm., and the time taken for the depth 
of colouration to reach that shown by a greater length of control tube, 
which had not been acted upon by radium, found. This time cannot be 
stated with any degree of certainty, but a progressive colouration was 
shown, and the contrast presented the same difficulty as that found in 
comparing intensities of illumination of lights of different colour. The 
green colour was changed to a browner shade, especially close to the 
inner wall of the tube. 
The following figures must be taken as approximate only : — 
Time Length of control tube 
in months. to match colour. 
0 3-3 
3 4-0 
6 5-0 
8 6-2 
12 8-0 
A brown layer was formed close to the inner wall of the tube, and 
this gradually merged into the general green colour of the glass and 
reached the outer wall. The thickness of the wall was 0-23 cm. 
2. Jena hard glass tubing. This glass had a bluish green-colour, and 
no change occurred in twelve months in a tube 2-1 cm. in length. 
3. White thermometer tubing 2-2 cm. in length. This specimen was 
nearly white. A faint violet colouration was seen in the layer immediately 
surrounding the bore, and a darkening occurred throughout the mass. 
The length at the end of twelve months of the control tubing which 
matched that of the experimental tube was 5 cm. 
