Experiments on Residual Ionization. 397 



other gases, a fact which certainly cannot be accounted for 

 on the ground of its density. The present investigation was 

 therefore undertaken to see whether this large ionization 

 could be traced to either of the sources (1) or (3) above. As 

 a result, it has been found that in acetylene made from 

 calcium carbide there is present a slight trace of radium 

 emanation, and this it has been shown accounts for the high 

 residual ionization in the gas. 



Experiments. — The gas used at first in the experiments 

 with acetylene was taken from a commercial Prest-o-Lite 

 tank. The ionization-chamber was made of zinc because it 

 has been shown that this metal contains smaller traces of 

 radioactive impurities than any other. The external dimen- 

 sions of the chamber w 7 ere : — diameter 11*6 cm., length 

 22'S cm., and its volume was 2167 c.c. The thickness of the 

 walls w r as about 3 mm. The chamber was carefully scoured 

 with emery, and washed with dilute hydrochloric acid and 

 water, to remove radioactive deposits. During all the course 

 of the work the chamber was absolutely air-tight. It was 

 provided with a zinc electrode which was connected to a 

 sensitive electrometer in the usual manner. The wall of 

 the chamber was kept at a potential of 240 volts, which 

 ensured that all the currents measured were saturation 

 currents. 



The ionization-chamber was filled with acetylene which 

 had been carefully dried and freed from dust, and the number 

 of ions made per c.c. per second was found to be about 20. As 

 this number was considerably less than that previously found 

 by McLennan and Treleaven, the zinc Wolf electrometer 

 (PI. X. fig. 1) used by them was filled with acetylene from 

 the Prest-o-Lite tank, and the number of ions made per c.c. 

 per second was found to be only 12. The difference between 

 this number and that found with the zinc ionization-chamber 

 was probably due to a radioactive impurity in the walls of 

 the latter. It was thought that the low value of the ionization 

 was due to the commercial acetylene not being pure, and to 

 test this some acetylene was made in the laboratory from 

 calcium carbide, and both the Wolf electrometer and the zinc 

 ionization-chamber were filled with it. The number of ions 

 made per c.c. per second (n) was measured at intervals during 

 a period of about 19 days. The following set of readings was 

 obtained with the Wolf electrometer, and a similar set with 

 the zinc ionization-chamber. 



