465 



rubber tubing from the shaking vessel into the ionization chamber and back 

 again to the shaking vessel until the emanation is mixed through the air of 

 both chambers in the same proportion. Knowing the constants of the elec- 

 troscope and the observed change of deflection of the leaf, the amount of 

 emanation in the ionization chamber is known. Knowing this and the various 

 volumes of air and water the amount of emanation per liter of water can be 

 calculated. The shaking vessel is made of a can with two brass stop cocks 

 soldered into it. One cock is placed near the top the other is placed on the 

 side about half way up. For convenience the position of the lower stop cock 

 can be calculated so that the vessel will hold a certain quantity of water 

 when the vessel is filled full and then placed on a level stand with both stop 

 cocks open. In this manner the volume of the water is determined easily 

 and can be made the same in each experiment. The volume of the air above 

 the water can be had by determining the total volume of the can. To pump 

 the air around a rubber bulb pump such as is used in pyrography outfits 

 answers well. The volume of the air in the tubes and pump must be esti- 

 mated and used in the calculations. 



The formula for calculating the amount of emanation per liter, which 

 can be derived easily in connection with Fig. 6, is as follows: 



I V 2 + a Vi V 2 + V 3 + V 4 



E = - - ( -) (- -) e, 



Vi V: V 4 



Where V, = Volume of water in shaking can, expressed in liters. 

 V> = Volume of air in shaking can, expressed in liters. 

 V 3 = Volume of bulb, pump, and connection tubes. 

 V4 = Volume of ionization chamber. 



a = Absorption coefficient of water for radium emanation, 

 e = Amount of emanation in chamber, V4. 

 E = Amount of emanation per liter of water. 



The quantity alpha, a, has been determined experimentally and has 

 been found to depend upon the temperature. The value at any temperature 

 can be had by referring to the curve (Fig. 7). The data for this curve is 

 taken from M. Koflcr (Akad. Wiss. Wien, Ber. 121, 2a pp. 2193; Sci. Abs. 

 Vol. 16, 1742, 1913), and Boyle (Phil. Mag., 22, p. 840, 1911.) 



As a test of the above equation the following will serve (Table 4). 

 Three tests were made at the spring under the ordinary conditions. The 



:J0— 4966 



