566 



SCIENCE 



[N. S. Vol. XXVII. No. 693 



that in the general ease the moment of 

 momentum per unit volume is proportional 

 to the intensity of magnetization. In the 

 case where all the revolving electrons are 

 of the same sign, the coefficient of propor- 

 tionality reduces to 2m/e where m is the 

 mass and e the charge of an electron. If, 

 as is generally assumed, the revolving elec- 

 trons in all real bodies are alike as to mass 

 and charge, this coeiScient should be the 

 same for all substances. Experiments are 

 being made at Princeton University to de- 

 tect the existence of this moment of mo- 

 mentum, which should give rise to effects 

 within the reach of experimental measure- 

 ment. 



Ionization of Gases at High Pressures: 

 Henet a. Eeikson, University of Min- 

 nesota. 



This paper gives the results of an experi- 

 mental investigation of the ionization pro- 

 duced in gases at high pressures by the y 

 rays of radium. 



The results show that the current ob- 

 tained with a constant difference of poten- 

 tial between the electrodes becomes a maxi- 

 mum at a certain pressure and that a 

 further increase in the pressure causes a 

 decrease in the current. 



The position of the maximum point de- 

 pends upon the difference of potential. 

 When the potential difference is 18 volts 

 between concentric cylindrical electrodes 

 separated by about 6 mm., the maximum is 

 at a pressure of 70 atmospheres, and the 

 current at 400 atmospheres is 60 per cent, 

 of the maximum. With a difference of 

 potential of 1,000 volts the maximum is 

 at 150 atmospheres, the current at 400 

 atmospheres being about 78 per cent, of 

 the maximum. 



The results also show that some of the 

 ions remain uncombined for some time 

 after the rays have been discontinued. At 

 the end of an interval of 20 seconds the 



number of ions obtained from air at 200 

 atmospheres, with a difference of potential 

 of 1,000 volts, is about 8 per cent, of the 

 number obtained at the end of an interval 

 of 0.1 second, and at 400 atmospheres the 

 number is about 13 per cent, for the same 

 intervals. 



The current-pressure curves for carbon 

 dioxide are similar to those for air, up to 

 the pressure at which the gas liquefies. 



The Variation of Apparent Capacity of a 

 Condenser with the Time of Discharge 

 and the Variation of Capacity with Fre- 

 quency in Alternating Current Measure- 

 ments: Beuce V. Hill, Chicago. 

 It has been shown that the capacity of 

 a condenser, as measured by the ballistic 

 method, varies greatly with the time dur- 

 ing which the condenser is connected with 

 the galvanometer. The present study was 

 designed to investigate this further and 

 also to learn the behavior of capacities 

 used in ordinary telephone circuits. 



For measuring short time intervals a 

 falling weight, operating make-and-break 

 keys was used. Intervals as short as 

 1/6,000 of a second could thus be read. 

 Discharge times varied from .0001 second 

 to 11 seconds. A D'Arsonval ballistic 

 galvanometer was used. With a mica con- 

 denser the apparent capacity was found 

 independent of the time of discharge, but 

 in four paper condensers the absorption 

 was considerable. 



A small inductor generator was used in; 

 testing the dependence of capacity upon 

 frequency. The mica condenser fell 1 per 

 cent, in capacity as the frequency varied 

 from 60 cycles to 3,000. A paper con- 

 denser fell 2.5 per cent., which is too small 

 to be taken account of in practical tele- 

 phone work. 



An Examination of Certain Alternating 

 Cu,rrent Circuits including those Con- 



I 



