EXTENSION-CURRENTS 71 



theory, a reversed current must be set up, which is directed 

 towards the point of contact with the water. The centre of 

 stream lines must lie in the edge of the drop opposite to the 

 point of contact. Such a reversal of the experiments, however, 

 was not very successful -when the drop was placed in soap solu- 

 tion, which seemed to adhere rather firmly to the surface of the 

 drop, so that the surface was surrounded by water before 

 actual contact took place ; but with alcohol, H^SO^ and NH^O, 

 good results were obtained. It was thereby shown, especially 

 when alcohol was used, that the extension-current set up both 

 with paraffin and olive oil, was very markedly stronger than when 

 the drop was placed in water and approached by alcohol. For 

 the most part the current produced is even violent, and also lasts 

 for a fairly long time. If it becomes slower it can be strengthened 

 again several times by repeated afflux of water. During the pro- 

 cess minute globules of oil become split ofi' in numbers from the 

 surface of the drop, and carried by the current to the hinder end 

 of the drop, i.e. to the end in contact with the water. Here they 

 collect, and gradually spread out again from this point in the 

 form of an arch on each side towards the end in contact with 

 the alcohol. They thus form a figure exactly corresponding 

 to that which has already been described above for the distribu- 

 tion of soot particles, or split off oil droplets (see above, Fig. 10, p. 

 64). When a drop of olive oil was used for the experiment, 

 numerous minute droplets of the surrounding fluid penetrated 

 into the oil, and quickly made it turbid. 



An explanation for the much greater intensity of the exten- 

 sion-current under these conditions of the experiment can per- 

 haps be found in the fact that here the drop has a low 

 tension over a large portion of its surface, and a high , 

 tension over a small portion; while in the experiments first 

 described the conditions are reversed. If, as we assume, a 

 bursting of the surface layer of less tension takes place as a 

 consequence of the difference of surface tension, then the whole 

 phenomenon must of course be much more violent when a large 

 surface is burst in this manner, than when only a small surface is 

 dealt with. 



The same reversal of the experiment can also be carried out 

 with drops of paraffin oil which are put up in concentrated sul- 

 phuric acid or ammonia. Afflux of water then produces the 

 reversed extension-current in the same manner. I could not, 

 however, observe an especial strengthening of the current in 

 these experiments. In the year 1889 some experiments made 

 with ammonia succeeded very well, but recently they have come 

 off badly. 



