122 



Prof. J. A. Fleming. Electric Discharge [Jan. 9, 



(16.) If this last form of lamp is employed as in (4) the sub- 

 sidiary carbon loop being nsed as a middle plate, and a galvanometer 

 being connected between it and either the positive or negative main 

 terminal of the lamp, then when the subsidiary carbon loop is cold, 

 we get a current through the galvanometer only when it is in con- 

 nexion with the positive main terminal of the lamp, but when the sub- 

 sidiary carbon is made incandescent by a separate insulated battery, we 

 get a current through the galvanometer when it is connected either 

 to the positive or to the negative terminal of the lamp. In the first 

 case the current through the galvanometer is a negative current, flow- 

 ing from the middle carbon to the positive main terminal, and in the 

 second case it is a negative current, from the negative main terminal 

 to the middle subsidiary hot carbon. 



(17.) If a lamp having a metal middle plate held between the legs 

 of the carbon loop has a galvanometer connected between the negative 

 main terminal of the lamp and this middle plate, we find that when 

 the carbon is incandescent there is no sensible current flowing through 

 the galvanometer. The 'vacuous space between the middle plate 

 and the hot negative leg of the carbon possesses, however, a curious 

 unilateral conductivity. If a single Clark cell is inserted in series 

 with the galvanometer, we find that this cell can send a current 

 deflecting the galvanometer when its negative pole is in connexion 

 with the negative main terminal of the lamp, but if its positive pole 

 is in connexion with the negative terminal of the lamp, then no 

 current flows. The cell is thus able to force a current through the 

 vacuous space when the direction of the cell is such as to cause nega- 

 tive electricity to flow across the vacuous space from the hot carbon 

 to the coo]er metal plate, but not in the reverse direction. 



(18.) If a vacuum tube is constructed, having at each end horse- 

 shoe carbon filaments sealed into it, and which can each be made 

 separately incandescent by an insulated battery, we find that such a 

 vacuum tube, though requiring an electromotive force of many thou- 

 sands of volts to force a current through it when the carbon loops 

 are used as electrodes and are cold, will yet pass the current from a 

 single Clark cell when the carbon loop which forms the negative 

 electrode is rendered incandescent. It is thus found that a high 

 vacuum terminated electrically by unequally heated carbon electrodes 

 possesses an unilateral conductivity, and that electric discharge takes 

 place freely through it under an electromotive force of a few volts 

 when the negative electrode is made highly incandescent. 



(19.) These experimental results above described led the writer to 

 investigate, in the same manner, the electric arc between carbon poles 

 taken in air. If an electric arc is formed, in the usual way, between 

 carbon poles, and a third insulated carbon pole is allowed to dip into 

 or touch the electric arc, or, better still, has the electric arc projected 



