Electrification at Liquid-Gas Surfaces. 



301 



prevent it slipping. The drifting of: the babble due to 

 gravity -was controlled by levelling-screws in the base o£ the 

 apparatus. It was found that when all adjustments 'were 

 made, the bubble, observed through a low-power microscope, 

 would remain on the hair-line for some time. 



With this cell an examination was made ol (1) the effect 

 of the size of the bubble on its velocity, (2) the relation 

 between the velocity and the potential gradient, (3) the value 

 of the velocity in distilled water of conductivity about 5.10 -6 . 

 The velocity was measured by noting the time—with a stop- 

 watch — taken by the bubble to move over a fixed number of 

 scale-divisions in the microscope. By a simple device the 

 watch was started (or stopped) and the electric field applied 

 (or removed) simultaneously. To eliminate a possible error 

 due to the tube not being exactly level, the bubble was made 

 to move a definite distance in one direction and to retrace it 

 in the opposite direction. The diameter of the bubble was 

 measured on the microscope-scale as the tube was rotating. 

 The following readings on bubbles of air of various sizes 

 show the accuracy obtained. In the first column is the 

 diameter in mm. ; in the second, five successive readings of 

 the time required to move over 20 scale-divisions (45 divi- 

 sions=l mm.). Since the field between the electrodes was 

 not uniform, the readings were taken at a fixed distance from 

 one electrode. 



Diaui. 





T] 



me in sees. 





Average. 



■06 



9h 



10i, 



10i, 





101 



'08 



10 , 



10± 



10i, 10|, 



n 



102 



•JO 



10 , 



10|, 



10i, 10±, 



m 



10-3 



•1l> 



10|, 



10i, 



ioi m, 



n 



10-5 



•14 



10 . 



10 , 



10 , 10|, 



10 



100 



•16 



10 , 



10 , 



10 , 10 , 



10 



100 



•16 



10|, 



11 , 



11 , 11 , 



10| 



109 



If the difficulty of maintaining a uniform speed of rotation 

 were fully overcome, the average time for bubbles of different 

 sizes would probably be the same. Bubbles larger than 

 •2 mm. gave inconsistent results, due apparently to eddy 

 currents set up by their failure to remain exactly on the axis. 

 Very small bubbles, formed sometimes by the gradual absorp- 

 tion of the gas by the water, were observed to leave the axis 

 in an irregular way as if their buoyancy were no longer 

 sufficient to keep them on the axis. It may be concluded 

 then, that, within the limits given in the table above, the 

 velocity is independent of the size. 



