1884.] Electric Discharge with Chloride of Silver Battery. 207 



Fig. 1. 



§ 



In the first place the tube was placed in the magnetic meridian, 

 and the needle of iron wire, n ri, in the same direction ; tested bj means 

 of a very small magnet, both ends were equally attracted, showing 

 that the needle had been thoroughly decarbonised; this was done by 

 heating it to redness for many hours in peroxide of iron, prepared by 

 burning its oxalate. 



The discharge was in the first instance passed from the ring to the 

 point, so that the needle was in the dark space ; no magnetism was 

 developed in the needle, which would have been the case if the dis- 

 charge bad had a spiral motion as we have often observed and 

 described to be sometimes the case.* It was with the object of 

 ascertaining this fact that the apparatus had been made. 



Fig. 2. Fig. 3. 



The needle was now placed at right angles to the tube, and the 

 point made positive ; after a few trials at different exhausts a beautiful 

 tongue-shaped stratification was obtained, and it was then possible to 

 make the apex of a stratum impinge on one or the other end of the 

 needle, figs. 2 and 3, on whichever end the stratum touched, the 

 needle was pushed away from it, showing clearly that the balance of 

 forces which hold together the molecules composing a stratum are 

 sufficient to render it viscous, and unyielding to a small resistance- 

 With disk or saucer-shaped strata the whole length of the needle 

 being touched at the same time, it had no tendency to turn it on its 

 axis, but became agitated when touched by strata. 



* "Phil. Trans.," Part I, vol. 169, pp. 250-253, 255-263, 265. 



