186 ESSAYS BIOGRAPHICAL AND CHEMICAL 



copper, and formed water. The apparatus was so arranged 

 that mercury could be allowed to enter the tube from 

 below, so as to sweep before it any remaining gas ; and 

 the water was removed from the gas by making it pass 

 through a tube filled with a suitable absorbing agent, 

 followed up by mercury. The gas finally entered a very 

 small spectrum-tube, entirely made of capillary tubing, 

 like the stem of a thermometer. On passing a discharge 

 from a coil through the spectrum- tube after the emanation 

 had been thus introduced, a spectrum was seen, consisting 

 of some bright green lines ; but it was extremely difficult 

 to prevent the presence of traces of carbon compounds, 

 and at this stage their spectrum was always seen. But 

 the D 3 line of helium was absent. After a couple of days, 

 however, a faint yellow hue began to appear, identical in 

 position with D 3 ; and as time went on, that line became 

 more distinct, and was followed by the other lines charac- 

 teristic of helium, until, after a week, the whole helium 

 spectrum was visible. It was thus proved that the radium 

 emanation spontaneously changes into helium. Of course 

 other substances might have been, and undoubtedly were, 

 formed ; but these it was not possible to detect. 



The next problem was to measure the amount of 

 emanation, resulting from a given weight of radium, in a 

 given time. The method of procedure was similar to that 

 already described, except in one respect: the spiral of 

 oxidised copper wire was omitted, and the excess of 

 hydrogen, mixed with the emanation, was cooled in a small 

 bulb by help of liquid air. This condensed the emanation; 

 and the hydrogen, which of course is not liquefied at the 

 temperature of liquid air, was pumped away. On removal 

 of the liquid air, the emanation became gaseous, and it 

 was forced by means of mercury into a minute measuring 

 tube, like the very narrow stem of a thermometer. It 

 was thus possible to measure its volume. It is a well- 



