July I, 1880] 



NATURE 



197 



a discharge from 2,400 cells, the deviation to the left on 

 passing down e and/" was 20 divisions. 



Tube 199, with a hydrogen charge, was now substituted ; 

 pressure 2 millims., 2,632 M, 5,100 cells, current 0-01639 W. 

 This tube has already been described (Nature, vol. -\xii. 

 p. 176) ; it is 37 inches (94 centims.) long, and 5}J inches 

 <I4'8 centims.) in diameter. The distance between the 

 terminals, a ring and a straight wire, is 33'S inches (85 

 centims.). In the first place the battery was connected 

 direct to the galvanometer, the positive to e and the 

 negative to / (that is, in the same direction as if the 

 positive were attached to c and the negative to d). A 

 short piece of wire was inserted between e and / as a 

 shunt, and the 5,?,^ shunt was also used with the galvano- 

 meter, the direction of deviation was found to be to left, 

 which it was desired to know. 



On pressing down e andy, the deviation was to the left, 

 and only amounted to 2 divisions. On another occasion, 

 at the same pressure, 3,900 cells, current 0-02925 W., the 

 deviation to the left on pressing down e and f was from 

 3 to 5 divisions. 



The same tube, with a charge of coal-gas, pressure 

 3 millims., 3,947 M, current o'oi705 W., deflection, on 

 pressing down e andy; 7 divisions to left. 



The same tube, with air, pressure i'5 millims., 1,974 M, 

 current o'0272S W., on pressing down e and /, deviation 

 20 divisions. The deviation was, therefore, greatest with 

 air ; but if due to a chemical polarisation, it would have 

 been a prion' expected to be greatest with coal-gas, which 

 is a mixture of decomposable molecules. The result of the 

 experiments with tube 199 gave the following deviations : — 



Hydrogen. Coal-gas. Air. 



3 7 20 



Again, tube 199, 2,400 cells, pressure i millim., 1,316 M, 

 current 0-02456 W., deflection on pressing down the keys, 

 16 to left ; with 1,500 cells no discharge, yet a deflection 

 of 10 divisions to the left was produced on pressing down 

 the keys. The deflection in the latter case being clearly- 

 due to a static charge. 



Experiments were now made in air with an apparatus 

 sii-nilar to that already described in Nature, vol. xxii. 

 p. 174, but with disks instead of points, in order 



to obtain a vacuum vessel with large terminals. The 

 terminals are two disks a little over 4j inches ( 1 1 centims.) 

 in diameter, and in the first instance w-ere placed 3 inches 

 (7-62 centims.) apart. Pressure 4-5 millims., 5,921 M, 

 2,400 cells, current 0-04366 \V., deviation on connecting 

 the disks with the galvanometer, 10 divisions to left. 



With 1,200 cells, no discharge took place, but on con- 

 necting the galvanometer a deviation of 27 divisions to 

 the left was, nevertheless, produced. 



At a pressure of 20 millims., 26,316 M, with 3,900 cells, 

 no current passed, but the deviation on connecting the 

 galvanometer, was 66-6 divisions to the left. 



The disks were now placed at a distance of 2 inches 

 {5-03 centims.), the pressure being still 20 millims., and 

 the battery 3,900 cells, the discharge now took place, 

 current 0-03896 \V., deflection on connecting galvanometer 

 8-5 to left. 



On reducing the battery to 2,400 cells, the discharge 

 did not take place, but a deviation of 43 divisions was 

 produced on pressing down e and/ 



'^|0^" 



^y 



The disks were placed at i inch (2-54 centims.) apart, 

 same pressure, with 3,900 cells, the current was 0-04201 

 \\'., deviation to left on connecting the galvanometer 13 

 divisions. 



2,400 cells would not pass, but a deviation to the left 

 of 47 divisions was obtained on connecting the galvano- 

 meter. 



The following are the results of these experiments with 

 the disks at various distances 



