Mass of the Ions in the Electric Wind in Air. 415 



table of results. By the same token, however, these variations 

 will tend to be symmetrical about the true value of V, and 

 will therefore not affect the mean of a large number of experi- 

 mental results. 



There still remains the question of the effect of the current- 

 strength on V. In the light of the above it is useless to 

 make observations on different currents at different times 

 except in great numbers. For fair comparison the readings 

 for two currents must be thoroughly sandwiched, and this 

 was the case for each of the bracketed pairs in the table of 

 results. The precaution is seen to have been justified. The 

 average difference between the values of V for 3'2 and 1*6 

 microamps. respectively is 2 per cent, for + and 0*2 per 

 cent, for — discharge only ; yet the absolute values of V for 

 the first pair of readings are the worst in the table. What- 

 ever the disturbing cause it thus affects the pressure for both 

 currents equally, and we may conclude that within the limits 

 of the experiments V is unaffected by changes in the current. 

 Above 32 microamps. the wimshurst refused to drive the 

 current for sufficiently large values of z. 



The following are the final mean values of V. In calculating 

 them half weight was given to the bracketed pairs, as they 

 required longer time for taking than the rest. 



V+ =413 centim. per sec. in unit E.S. field. 

 V-=540 „ ' „ 



Their sum is thus 953, which is very close to Rutherford's 

 value, 960. Their ratio is 1*31, which is in less satisfactory 

 agreement with Zeleny's ratio, 1"25. There can be little 

 doubt, however, that the ions are the same in all three cases. 



Relation betiveen Velocity of the Ions and that of the Wind. 



As already pointed out, V is the velocity of the ions 

 relatively to the electrodes ; and as the air through which 

 they move is also in motion, Y is not necessarily the same as 

 the relative motion of ions and air. 



In the " hole in plate " experiments the value of the electric 

 field was 4760 volts per centim. for + and 3730 for — dis- 

 charge. The actual ionic velocities were thus 6560 centim. 

 per second for -f and 6710 for — ions, except close to the 

 point. 



From curves II. b the current-area on the plate is about 

 2*3 sq. centim. From curves III., P + is about 40 and P — 

 27 dynes. Hence if we adopt the worst possible condi- 

 tions, by assuming that the drag of the ions on the air is 

 turned completely into momentum before it reaches the plate, 



