THE STRUCTURE OF THE NUCLEUS. 173 



In more recent expei-iments, preference was given to a horizontal capsule, the 

 lid of which was a light disc attached to the lower end of the needle and closing the 

 capsule similarly to the disposition in the guard ring electrometer. A pellet of 

 phosphorus w\as placed in the capsule, and both this and the upper end of the 

 bifilar were connected with the water battery, thus charging the needle from both 

 ends. The leakage in the presence of the phosphorus was but slightly greater than 

 in its absence, being with the quadrants at 10 volts, .24 X 10~i2 iq 77 ^ 10"' 2 

 amperes in the former case and .29 X 10^' 2 to .40 X lO'^'^ amperes in the latter, 

 according as the sign of the charge was the same or opposite to that of the needle. 

 The deflection with phosphorus is here unsymmeti-ical, but the constancy perfect 

 and the sensitiveness very high. Indeed it was also feasible to charge the quad- 

 rants permanently and symmetrically with the water battery and to connect the 

 unknown voltaaie with the needle either throuj^h induction alone in the absence of 

 phosphorus or through the ionized emanation, according as high or low potentials 

 were to be measured. 



