256 Prof. J. A. Fleming on the Poulsen Arc as a 



and generally was about 600 units. This field was so directed 

 as to cause the arc to spring between the upper edge of the 

 carbon and the copper beak, and be forced upwards in an 

 arch. 



Using a direct current supply at a pressure of 400-500 

 volts with a suitable rheostat interposed, it is easy to form an 

 arc having a potential difference between the carbon and 

 metal electrode of 300-350 volts and taking 5-10 amperes. 



These electrodes are then connected by an oscillatory 

 circuit consisting of a condenser made with metal plates 

 separated by sheet ebonite, the whole being immersed in an 

 insulating oil. The capacity of the condenser generally used 

 by me was 0*00289 microfarad. In series with the condenser 

 was an inductance consisting of a coil of insulated copper 

 wire No. 7/22 wound in square form over a wooden cross so 

 as to make a circuit 60 cms. in the side, having 8 turns. 

 The inductance of this coil for high-frequency currents was 

 116,200 centimetres. An additional sliding inductance was 

 used consisting of bare copper wire wound in a spiral on a 

 wooden rod 1\ inch or 3'2 cms. in diameter, having 8 turns 

 to the inch. Two such spirals were placed on a support 

 parallel to each other, and cross-connected by a sliding copper 

 saddle, so that by moving the saddle, more or less inductance 

 could be thrown into the circuit without interrupting it. 

 The full inductance of the double spiral was 200,000 centi- 

 metres. If then an arc is created between the electrodes 

 and adjusted to be of such a length as to take about 8 amperes 

 with 300-350 volts between the carbons, then powerful 

 oscillations are set up in the square circuit in series with the 

 condenser. A hot wire ammeter inserted in this circuit 

 indicated a current of about 5 amperes. If an electrostatic 

 voltmeter is joined across the terminals of the condenser, 

 we find a potential difference of 1200 to 1500 volts (R.M.S. 

 value), vastly exceeding the steady potential difference of the 

 arc electrodes. 



The existence of a powerful oscillatory current in this 

 square circuit may be shown by holding near to it a circular 

 coil of about 30 turns, 30 cms. in diameter, formed of well- 

 insulated wire, to the ends of which a 50-volt carbon filament 

 glow-lamp is attached. When this round coil is held near 

 and parallel to the square coil, the lamp lights up brilliantly in 

 consequence of the secondary current induced in the lamp coil. 



It is, however, seen that the lamp often flickers a good 

 deal, snowing that the primary oscillation current is not 

 absolutely constant, and this flickering varies with the ad- 

 justment. It is, however, possible to adjust the arc so that 



