Nipher — On the Nature of the Electric Discharge. 5 



two plates where in each gap of Fig. 1, the plates were 

 exposed at the opposite terminal. Here we have over 

 each film a negative outflow from the pin-head. In the 

 negative line the outflow is from the pin-head, which is 

 in communication with the machine. In the positive line 

 it is an outflow from the grounded pin-head. This out- 

 flow is induced by the copper plate below, which is in an 

 exhaust condition. Franklin's fluid has been drawn out 

 of it. These discharge lines are also alike in character. 

 These results might have been expected, if the ground 

 connections in Fig. 1 were broken, and the lines were 

 connected to form a circuit between the machine termi- 

 nals. This result is similar to that described in the 

 former paper 2 for the Crookes tube. 



It can hardly be claimed that these four photographic 

 plates contain in themselves evidence that they are pro- 

 duced by au outflow at one terminal of each gap, and by 

 an inflow at the other terminals. Such plates have long- 

 ago been produced at the gap terminals of an electric cir- 

 cuit, and they have not suggested such an explanation. 

 But when it is known that the so-called positive discharge 

 is an inflow of negative electricity, the plates themselves 

 seem to be suggestive of such a condition. 



If both terminals in either of the gaps in the discharge 

 lines of Fig. 1 be placed on the film of a photographic 

 plate, the terminals of the two forms of discharge lines 

 will unite with each other as should be expected if one is 

 an outflow and the other an inflow. In such an exposure, 

 a sheet of glass may be placed between the metal plate 

 and the photographic plate. In the exposure of the four 

 photographic plates thus far described, the spark gaps 

 at the machine were between large knobs. No apprecia- 

 ble brush discharges which could affect the film preceded 

 the disruptive discharge. 



The exposure table was screened from external sources 

 of light due to sparks at the machine. It was also wholly 

 surrounded by a metal screen of wire netting, which was 



2 These Trans. XIX, 1, p. 7. 



