3 82 



NATURE 



[August 1 6, 1883 



column of gas at various pressures, but also forcibly to the 

 impediment presented by the terminals themselves to the 

 passage of a discharge from gas to terminal and terminal to ga c . 

 It is shown that, at moderate exhausts, the resistance to the 

 passage of the discharge is uniform along the length of the 

 column of gas, and that at high exhausts it is not so, nnd that 

 the total resistance increases but slightly with an additional 

 length of the column ; moreover, that, at these low pressures, 

 the main impediment is in the passage of electricity between 

 gas and terminal or terminal and gas ; this is much greater at 

 the negative than at the positive terminal. 



The authors have next studied the electrical condition of a gas 

 in the immediate vicinity of the negative terminal. Tn order to 

 do this they constructed a tube 45 inches (u'4 centims.) long 

 and 1 J inches (4 '8 centims.) diameter. One terminal is in the 

 form of a point, the other in the form of a ring. The positive 

 pole of the battery was connected with the point, and the 

 negative either to the ring alone or to earth as well ; the 

 ring terminal of the tube was, when the battery was insulated, 

 connected with earth either by means of a >tout wire or 3 feet 

 (9P4 centims.) of fine platinum wire, o - oo2 inch (0x05 centim.) 

 diameter, and offering a resistance of 81 ohms, or a moistened 



DP DM* 



Fig. 5. 



cork offering a resistance of 4,300,000 ohms. In the tube were 

 sealed three idle wires, 1, 2, 3, covered with the exception of 

 their extremities with fine glass tubing (Fig. 6). No. I idle 

 wire is o'o02 inch (o'oo5 centim.) ; No. 2 o'2 inch (o'5 centim.) ; 

 and No. 3 o"6 inch (1*5 centims.) from the ring. The ring 

 terminal, when connected to earth, was found to be always at 

 zero potential ; notwithstanding this there was frequently ob- 

 served, more especially as the exhaust was increased, a negative 

 potential when the idle wires were connected successively with 

 the electrometer, amounting in one case with an air charge, 

 pressure o'ot aiillim., at wire No. 2, to 1068 cells, at wires I 



K.g. 6. 



and 3 to 912 cells. At other times a plus potential was observed. 

 Many experiments were made to determine the precise conditions 

 which developed a negative potential or a positive potential, but 

 unsuccessfully, and it was inferred that this depended on the 

 condition of the discharge itself within the tube. It is certainly 

 very remarkable that, while the potential of the negative ring 

 was absolutely zero, a high negative potential should be deve- 

 I iped in its near vicinity. 



The au'hirs remark that everyone familiar with the appear- 

 ance of a stratified discharge will have noticed when the negative 

 terminal is a rin?, that as the exhaust proceeds a spindle of light 



approaches and at last protrudes through the interior of it 

 (Fig. 7, I, 2, 3, 4, 5); this spindle they regard as a visible 

 exponent of strong action among the molecules of gas composing 

 it. In order to probe its electrical condition they prepared a 

 tube with a central idle wire surrounded by a minute glass tube, 

 except its extremity, and projecting to a distance of § inch (o'95 

 centim.) from the plane of the ring, which was made negative. 

 Anofher idle wire was sealed in the tube 0'I5 inch (038 centim.) 

 from the periphery of the ring. As the exhaust proceeded with 

 a charge of carbonic anhydride, the spindle approached the ring 

 and ultimately protruded through it. It was found that the 



potential of the central idle wire increased with the exhaust, 

 until it nearly or quite equalled that of the whole tube ; while 

 that of the external idle wire was only o - o54 that of the 

 tube. 



A great number of experiments n ere made to test the poten- 

 tial across a stratum a, />, and across a dark space c, d, respec- 

 tively, by two idle wires sealed in suitable positions in a tube, 

 one of which was connected w ith earth, the other with the elec- 

 trometer (Fig. 7, 6). The gapes used were carbonic anhydride 

 and hydrogen respectively. As a mean of a great number of 

 experiments it was found that when a dark space was straddled, 



