Boltwood — Radio-activity of Natural Waters. 381 



The air in the electroscope is exhausted until the pressure is 

 about one-half atmosphere. To one of the stopcocks of the 

 electroscope a small T-tube is attached by a short length of 

 rubber tubing. The branch of the T-tube connected with the 

 rubber tube contains a filling of granulated, anhydrous calcium 

 chloride held between plugs of cotton wool. The rubber tube 

 j of the gas burette is slipped over another branch of the T- 

 tube, and the third branch is closed by a short rubber tube 

 plugged with a glass rod. The pinchcocky is first opened and 

 then the stopcock of the electroscope is opened, so that the gas 

 is drawn very slowly into the electroscope. When the water 

 in the burette has risen to the junction of the side tube, the 

 pinchcock^' is closed and the plug removed from the side tube, 

 permitting the external air to pass into the electroscope and 

 sweep with it any of the gases from JE remaining in the tube 

 and connections. When the atmospheric pressure is estab- 

 lished in the electroscope the stopcock is closed and the appa- 

 ratus disconnected. 



In the tests which will be described later in this paper the 

 average volume of gas obtained was about 150 cc . The capacity 

 of the electroscope used was about 530 cc , so that under the 

 conditions which have been described about 100 cc of atmos- 

 pheric air swept through the connecting tubes before the normal 

 pressure was established in the interior of the electroscope. 



The activity of the gas from the water was measured in the 

 electroscope, which has been described in a previous paper.* 

 The gold leaf was charged by a battery consisting of 216 small, 

 lead storage cells, at a potential, therefore, of about 432 volts. 

 The positive terminal of the battery was earthed and con- 

 nected with the case of the instrument ; the negative terminal 

 was connected with the top of the rod supporting the gold- 

 leaf. The normal air-leak of the instrument was low and 

 quite constant, and was equal to approximately 0*012 division 

 per minute. The readings obtained with each measurement of 

 gas were corrected by this quantity. 



Because of the initial rise in the activity when the emana- 

 tion is first introduced into the electroscope, due to the forma- 

 tion of " emanation X " on the walls of the vessel, the rate of 

 leak at the end of three hours was taken as the measure of the 

 activity. 



Standard. 



One of the most important points in the measurement of 

 the radio-activity of the gas contained in a water is that the 

 results shall be given in the terms of some standard which will 

 permit the direct comparison of the results obtained by dif- 

 ferent experimenters. In the investigations conducted by 

 * This Journal, xviii, 97 (1904). ' 



