Mad line for producing High Tension Electricity. 189 



London building ; but with my multiple machine I have succeeded in 

 reproducing the experiment in a modified form. Taking two glasses 

 of the form shown in fig. 6, placed near to each other as in the 

 original experiment, I inserted in one of them a block tin cup in the 

 manner shown in the figure. Into this cup I coiled a string composed 



Pig-. 7. 



of ten strands of fine lamp cotton, which, when laid together, were 

 ( f a sufficient section to make an easy fit for a hole of | inch in 

 diameter. The negative wire was put in contact with the tin cup 

 and the positive wire was inserted in the other glass. The upper 

 end of the cotton string was then laid over the edges of the two 

 glasses, and the tin cup and the positive glass were then each filled 

 with chemically pure water up to the same level. When the current 

 was turned on the cotton string commenced to crawl over the edges of 

 the two glasses, and never ceased to travel until it was bodily trans- 

 ferred from the negative cup into the positive glass. It was natural to 

 expect that the water with which the cotton was loaded would travel 

 with the cotton and raise the level of the water in the positive glass 

 above the level of that in the tin cup, but the contrary was the case. 

 The water lowered in the positive glass and overflowed in the tin cup, 

 the surplus descending into the empty space beneath the cup, where 

 its quantity could be seen arid estimated. In short, the cotton 

 travelled one way and the water the other, notwithstanding that the 

 flow of the water was in opposition to the motion of the cotton. No 

 spark passed between the glasses until the tail of the cotton ceased 

 to touch the water in the cup, and then sparks passed in profusion. 

 When the cotton was restrained from travelling, the water came over 

 from the positive side in greater abundance, but when the cotton was 

 dispensed with, and the gap between the vessels was bridged by a 



