620 Mr. N. Tl. Campbell : E.rperiments on 



D raised from beneath : the water flowed out of the bulb, 

 leaving it full of the gas except for a layer of water which 

 served as one o£ the electrodes for the discharge. By 

 warming the side tube C excellent insulation could be 

 established between the point and the water. 



The two electrodes were connected to a Wimshurst 

 machine through a key arranged so that either of them could 

 be made positive or negative, and either or both of them 

 earthed at will. An adjustable spark-gap .r, inserted in 

 paridlel, regulated the potential difference, which was 

 measured by the Braun electrostatic voltmeter Y. It was 

 soon found that the effects observed were independent of the 

 potential-difference within the range available [i. e., from 

 the minimum potential of 2000 volts up to 5000 volts) : in 

 no case were any nuclei produced unless a visible discharge 

 had passed. ]t was also quite immaterial whether the point 

 or the water were connected to earth, so long as the direction 

 of the electric force remained unaltered. The experiments 

 were conducted wdth the "vvater connected to earth and the 

 point raised to about + 3500 volts. 



Under these conditions, in air, oxygen, hydrogen, and 

 helium, the discharge from the positive point produced a 

 small faintly luminous fan-shaped ''brush :'^ when negative 

 the point was tipped with a much lighter star-like glow\ 

 In nitrogen^ the negative discharge had the same appearance, 

 but the positive " brush '' was extended in the form of a long 

 violet " tail/^ reaching to the surface of the water. In the 

 hydrocarbons employed, as well as in CO, the light from 

 the discharge was greenish, and the angle of the " fan '^ on the 

 positive point considerably greater than in the other gases. 



Mr. Wilson has pointed out in the paper mentioned above, 

 that the nuclei produced appear to grow after the discharge 

 has stopped : it is therefore necessary to observe the results 

 of expansion while the discharge is passing. The fog-dis- 

 persing powder of the discharge renders the detection of very 

 fine showers somewhat difficult, and hence great accuracy 

 in determining the value of the expansion required to give 

 a fog is impossible. 



Each form of apparatus used was standardized by means 

 of Eontgen rays or radium rays as described by Mr. W^ilson*. 

 The form showm in the figure required no correction : only 

 corrected values are given below- . 



It will be necessary here to mention wdth some detail 

 the chief difficulty encountered in the course of this 

 research. 



* C. T. R. Wilson, Pliil. Trans. 193 A. p. 202 {l&d'd). 



