ELECTRICAL EXPERIMENTS 55 



myself for a long time, and repeatedly, with this object. Un- 

 fortunately, however, I must confess that the results obtained 

 are not very satisfactory. This may depend partly on the 

 natural difficulties of the object, and partly also no doubt 

 on the inexperience of the observer in this subject. Although 

 therefore I must regret not being able to communicate more 

 definite and satisfactory results, I yet think that I ought 

 not to remain silent about them, even at the risk of what 

 is brought forward here being subjected to a severe criti- 

 cism from more competent authorities. 



In my preliminary communication of 3rd May 1889 it 

 was reported that the froth-drops, when between the poles 

 of the constant current, showed streaming , movements 

 directed towards the negative side. Since, however, ordin- 

 ary oil-drops also under these conditions permitted the observa- 

 tion of feeble streamings, soon dying away, the circumstance 

 seemed in need of further explanation. Even at that time 

 I had an idea that the centre of extension-currents directed 

 towards the negative pole might possibly be only a con- 

 sequence of the free alkalis formed at the negative pole by 

 means of the electrolysis set up an effect which would natur- 

 ally produce a centre of extension-currents directed towards 

 the negative pole as soon as they reached the surface of the 

 drop of foam or oil (on this point see below, p. 62). Now 

 since the glycerine, in which the drops of foam were investi- 

 gated, contains potash soap in solution, a possibility of this 

 kind ought certainly to be taken into consideration. More- 

 over, the water employed for diluting the glycerine is itself 

 not free from traces of alkaline salts. The reasons which, 

 in spite of this, caused me at that time not to ascribe the 

 negative stream to this cause I will not go into more fully 

 at present. My original experiments were of course per- 

 formed on slides, which were provided in the usual 

 way with electrodes of platinum foil, the electrodes being 

 about 4 to 5 mm. apart. In order to avoid the above- 

 mentioned source of error, in later experiments performed 

 in the course of the summer of 1889, I made use of small, 

 so-called non-polarisable brush electrodes, as invented by 

 Dubois Keymond, which, wetted with 1 per cent solution of 



