[WRIGHT] VARIATIONS IN THii: CONDUCTIVITY OF AIR 91 



both with and without an auxiliary condenser of known capacity attached 

 to the electrode. 



To show how the determination of '"' q " with this instrument was 

 arrived at and to give an estimation of the magnitudes involved, a 

 measurement was undertaken with the same cylinder and under the 

 same conditions as obtained in the case of the determination with the 

 Wilson instrument, the actual values being given in Table III. These 

 values for " q," as obtained with the two instruments, are seen to be 

 practically the same and so afford a conclusive test of thp accuracy of 

 the measurements. 



TABLE III. 



(1)— Date Mar. 19, 4 p.m. 

 (2)— Lead Cylinder No. 1. 



(3) — Locality — Basement of Physics Laboratory. 

 (4) — Potential applied to needle + 250 volts. 



(5) — Rate of motion of needle without condenser = 13.0 mm. per min. 

 (6) — Rate of motion of needle with condenser = - 7.4 mm. 



(7) — Rate of motion of needle due to defective insulation cylinder being removed 

 and electrode capped = — 1.0 mm. 

 . • . Rate due to conduction in air in cylinder alone = - 12.0 mm. 

 (8) — Sensitiveness = 832.2 mm. per volt. 

 (9) — Capacity of auxiliary condenser = 128.04 cms. 

 (10) — Temp. 10°C, voltage on cylinder + 85. 

 (11)— Volume of cylinder = 26700 ces. 

 From this we obtain: — 



Capacity of electrometer 169.1 cms. 



The calculation to determine " q " is then 



169. 1 X 12 = q X 3 . 4 X 10 — ^^x 60 x 26700 



832 . 2 X 300 



Whence : q = 14.93 



Before making the final measurements in these comparisons, pre- 

 liminary obseiwations were made to ascertain what voltage it was neces- 

 sary to apply to a receiver in order to obtain saturation currents. A 

 pet of the different voltages applied and the corresponding values of 

 the currents obtained with a lead cylinder of comparatively high activity, 

 but having dimensions the same as cylinder Ko. 1 mentioned above, are 

 given in Table IV and a curve to represent them in Fig. 6. From 

 the latter it may be seen that a potential of 60 volts gave a reading 

 well over tlie knee of the curve, and as the voltage applied in all the 

 determinations lay between 80 and 250 volts, it is clear that saturation 

 currents were always obtained. 



