366 Mr. R. K. McClung on the Relative Amount of 
an alternate spark-gap. When a balance was being obtained, 
the length of this spark-gap was adjusted at the same time 
so that the discharge would pass across this air-space just 
about as easily as through the bulb. This sparking-distance 
was the distance between two brass spheres. The length of 
this alternate spark-gap depends upon the pressure in the 
bulb, and the hardness of the rays also depends upon this 
pressure. Although the relation between this alternate 
sparking-distance and the hardness of the rays is not a very 
definite one, yet the greater this sparking-distance the harder 
are the rays, and the length of this spark-gap therefore gives 
some indication of the hardness of the rays. It is not an 
absolute measure, but serves as an indication of the relative 
hardness of the rays. The experiments made by this method 
showed that the relative ionization in air and hydrogen 
depended in a very definite manner upon the hardness of the 
rays. A number of experiments were made on this question, 
and a series of readings is shown in Table II. just as an 
example. 
TaBLeE II. 
| Length of Alternate Pressure in vessel 
| Spark-gap in mms. containing Air. 
— | ee ae as ie = 
3°99 41°5 mm. 
8:02 50°35 _s—,, 
10°77 554 5 
14:06 63°4,, | 
17°24 664 ,, | 
The first column shows the alternate sparking-distance, while 
the second one shows the pressure of the air in the one 
cylinder at which the ionization in it just balanced the 
ionization in the hydrogen in the other one at atmospheric 
pressure. As the rays became harder more air had to be 
let into the cylinder containing air in order to maintain a 
balance. These results given in this table are only relative 
ones, for in doing these experiments a balance was not obtained 
in each case with both cylinders filled with air in the first 
place before introducing the hydrogen into one of them. 
The numbers representing the pressure of the air given in 
the table do not therefore indicate the absolute values of the 
pressure, but only relative values to show how the pressure 
varied as the rays varied. These results show conclusively 
that the variation of the pressure in the bulb producing the 
