the Faraday Dark Space of Vacuum- Tubes. 565 



paraffin switch-blocks were used throughout. Potentials of 

 the magnitude of the cathode " drop " were made by charging 

 the quadrants with a Clark standard cell and the needle with 

 the potential to be measured. A Hittorf cadmium-iodide 

 resistance * was used to regulate the current, the continuity 

 being tested by the receiver o£ a telephone. 



Preliminary Tests. 



To provide an experimental check on results, two vacuum- 

 tubes were fitted out, alike in all respects, and in open 

 connexion with each other. Observations made in one 

 could thereby be repeated immediately afterwards in the 

 other. 



Platinum exploring-wires (diam. two-tenths ram. ; distance 

 apart 3*5 mm.) were first used, but the gradients obtained 

 from the two tubes seldom agreed. Furthermore, when the 

 current was reversed, the constant gradient of the luminous 

 column appeared often greatly increased (or diminished). A 

 discharge sent from the exploring-wires as cathode tended to 

 reduce these discrepancies, thus proving that they arose 

 from the condition of the surface of the wires. 



Aluminium wires (of same diameter and distance apart as 

 the platinum) were tried, gave very constant results, and were 

 subsequently used, making check-observations in the second 

 tube in general unnecessary. The gradient indicated by 

 these was regularly about 35 per cent, above that calculated 

 from the difference of potential observed between the extre- 

 mities of the luminous column. This discrepancy is probably 

 due to a local disturbance of the discharge caused by the 

 presence of the wires. 



Circular disks of polished iron were used as electrodes. 

 To allow for a considerable variation in the current strength 

 with the cathode " drop " constant, it was found necessary to 

 give these a diameter of 30 mm. 



Faraday Dark Space. 



In Tables I., II., and III., the values of the gradient for 

 various currents and different gas-pressures are given. 

 These all refer to a constant drop at the cathode of about 

 320 volts. The observed gradient is reduced to volts per 

 centimetre, and the positions are as indicated in fig. 1. 

 Though these tables are chosen as the best from several series 

 of observations, at various gas-pressures, yet results obtained 

 at different times under the same conditions varied never 

 more than ten per cent., and usually no more than five. 

 ♦ W. Hittorf, Wied. Ann. vii. p. 559 (1879). 



