126 Prof. Bumstead, On the absence of excited radio-activity 



to compensate for this large leak. To accomplish this, a second 

 testing cylinder of the same dimensions as the first, but without 

 the window in the bottom, was mounted on a moveable arm in 

 such a way that it was (like the first) exposed to the rays which 

 came through the lead cylinder, but was screened from the disc. 

 Its central rod was connected to the central rod of the first testing 

 cylinder (and thus to the gold leaf) by means of a wire dipping 

 into a very small cup of mercury; this cup was immediately 

 beneath the pivot about which the moveable arm supporting the 

 second testing cylinder rotated, and thus the second cylinder 

 could be moved closer to or further from the lead screen surround- 

 ing the radium. When the two cylinders were charged to equal 

 and opposite potentials (usually 40 volts), one could balance the 

 leak in one by the opposite leak in the other, the ionization in the 

 second cylinder being adjustable by moving it toward or away 

 from the radium. It took a great deal of time to get exact com- 

 pensation on account of the very slow motion of the gold leaf 

 when near to a balance ; but it was easy to make the resultant 

 leak less than 1 per cent, of the uncompensated leak. The 

 connections between the central rods of the testing cylinders and 

 the electroscope were surrounded by metallic screens joined to 

 earth ; and the rotating disc and motor were upon a separate 

 table, so that the apparatus was not disturbed by the rotation of 

 the disc. 



The ordinary course of the observations was to get an approxi- 

 mate balance with disc at rest, allowing the cylinder which was 

 exposed to the disc to preponderate slightly. The potential 

 acquired by the central rods and gold leaf in 10 minutes was 

 measured by observing the deflection of the gold leaf with a 

 micrometer microscope, and immediately thereafter producing the 

 same deflection by means of a potentiometer. In this way varia- 

 tions in the sensitiveness and zero point of the electroscope were 

 eliminated. A series of such observations was taken with the disc 

 alternately at rest and in motion, and the results compared. 



A considerable number of measurements were taken in .this 

 way, with the copper disc bare, and also when it was covered with 

 tin-foil, lead-foil, and with blotting-paper which had been soaked 

 in a solution of uranium nitrate and dried. The disc was rotated 

 ■at two different speeds ; with the slower speed, the disc made 

 approximately 150 revolutions per minute, so that the interval 

 between the exposure to the rays and the passage under the testing 

 cylinder was about | of a second. It was difficult to determine 

 the higher speed, since the application of the counter appreciably 

 diminished the velocity of rotation. It was however greater than 

 2200 revolutions per minute, so that the time interval in this case 

 was less than 0"009 seconds. With none of the materials tried, 



