[cline] radiation at the surface of the earth 117 



selected potential. Attached also to the leaf system was the inner 

 tube of the sUding condenser C, for details of which the reader is re- 

 ferred to the paper by C. S. Wright mentioned above. In the measure- 

 ments to be described the receiver was charged negatively to about 180 

 volts, which was found sufficient to insure a saturation current. The 

 small quartz Ley den jar was kept at a negative potential of approxi- 

 mately 50 volts, and the slide tube condenser C to varying negative 

 voltages depending on the sensibility desired. 



By moving this condenser any charge acquired by the gold-leaf 

 system through the conductivity of the air enclosed in R could be 

 annulled. In the experiments the condenser was always moved over 

 a standard distance and the time was taken for the conduction current 

 to annul the deflection of the gold-leaf produced by the displacement of 

 the sliding condenser. 



The charge annulled per unit voltage applied to the compensator 

 tube was .00501 e.s.u., a number which was determined by using the 

 auxiliary parallel plate condenser supplied with the instrument. 



Assuming the charge carried by an ion to be 3.4 x 10~^" e.s.u., it 

 follows when the volume of the receiver is known and also the time 

 required for the conduction current to annul a given c^uantity of 

 electricity, that the number of ions per cc. per second " q " in the 

 receiver R can be readily found. 



III.— EXPERIMENTS. 



Series I. 



In this series of experiments the receiver R was made of sheet zinc. 

 The potentials used were obtained from a set of small dry cells which 

 remained constant over the whole range of measurements. 



The receiver R, whose capacity was 31180 cc, was kept at a con- 

 stant potential of 184.5 volts throughout, the Leyden Jar Q at 51.5 

 volts and the sliding condenser C at 13.2 volts. In this series of 

 measurements the receiver was not hermetically sealed, so that baro- 

 metric changes were necessarily followed by changes in the air content 

 of the receiver. 



The readings were taken visually and were commenced by a few 

 preliminary observations over periods of from 4 to 6 hours' duration 

 selected from different parts of the day. These were afterwards fol- 

 lowed by observations taken continuously over a twenty-four hour 

 period. 



In the measurements for the shorter periods the apparatus was 

 set up in a room in the Physical Laboratory and rested on a soHd stone 

 table which constituted the sill of one of the windows in the room. 



