138 
Transactions of the Boyal Society of South Africa. 
a millimetre per minute when exhaustion had been carried to O'OOO mm. 
(i.e., less than xo^oo nim.). Fig. 1 is a diagrammatic sketch, showing the 
arrangement. The end of the capillary is surrounded by the vapour from 
the boiling water in the flask F ; U serves to break off a short length of 
the capillary when vapour is required to be admitted. The bulbs Bi, B2 
contain P2O5. Connection between these bulbs and the X-ray bulb is 
made through mercury traps. 
Fig. 1. 
This method of introducing the vapour was found to be unsatisfactory. 
Some vapour or gas, unabsorbed by P2O5, invariably remained after the 
introduced water vapour had been admitted to the P2O5 bulbs ; the 
McLeod gauge indicated pressures varying from a few thousandths to 
a few hundredths of a millimeter. 
Fig. 2 illustrates the second method of introducing vapour. The 
bulb W contained water and water vapour only. To introduce the 
vapour into the X-ray bulb, &c., the mercury was made to fall below 
J by lowering the reservoir E. The tube ab was 80 cm. long. Its 
