688 Professor J. A. Fleming [May 24, 



the arc is blown out, or before it is formed, the condenser is charged 

 by the dynamo.* When the arc is re-established the condenser is dis- 

 charged with oscillations. In the above specification nothing is said 

 about the use of a continuous current arc between carbon poles, but 

 Professor Thomson asserts that oscillations with frequency up to 

 00,000 could be obtained. In 1900 Mr. Duddell showed that if a 

 suitable condenser and inductance was shunted across the poles of a 

 continuous current arc formed with solid carbons, high-frequency 

 alternating currents were set up in the condenser circuit and the arc 

 emitted a musical sound (see Fig. 9). 



Much discussion subsequently took place as to the causes of the 

 effect and as to the highest frequency of oscillation it was possible to 

 secure by this method. Duddell and others based their explanation 



OOOQOQQ _ _ 



Fig. 8. — Elihu Thomson's Method Fig. 9. — Duddell Musical 



OF producing Oscillations. Arc. 



of the phenomenon upon the known fact that a small decrease in the 

 current through the carbon arc is accompanied by an increase in the 

 potential difference of the carbons. The continuous arc with solid 

 carbons was said therefore to have a negative resistance. f 



The explanation of the manner in which the continuous current 

 arc maintains undamped oscillations in the condenser circuit is then 

 as follows : If a condenser and inductance are shunted across the arc, 

 the condenser begins to be charged, and this robs the arc of some 

 current. This change, however, raises the potential difference of the 

 carbon poles and the charging of the condenser therefore continues. 

 When the condenser is full the arc current is again steady. The 



* An interesting and not very dissimilar device has recently been described 

 by Mr. S. G. Brown ; he employs a revolving aluminium wheel against which 

 a copper spring presses lightly. The spring and wheel are connected through 

 an inductance and resistance with a source of direct current supply, and also 

 by a circuit consisting of Leyden jar in series with a coil of wire. When the 

 wheel revolves an arc is formed at the loose contact, and high-frequency 

 oscillations are set up in the Leyden jar circuit, see Tiie Electrician, November 

 23, 190G, vol. Iviii., p. 201. 



f The term negative resistance is a very inappropriate term. It is better 

 to call the curve for an electric arc showing the relation of current through 

 the arc to potential difference of the electrodes or poles the clLar act eristic curve 

 of that arc, following a usual nomenclature in connection with dynamos. 

 This characteristic is a curve sloping downwards when the current is taken as 

 abscissa and the P.D. as ordinate. 



