432 



MESSRS. R. S. BUTTON AND J. E. PETAVEL ON ELECTRIC 



When the arc is started in a compressed oxidising atmosphere, the current is at 

 first unsteady and the electrodes must be rapidly fed up. Soon, however, a steadier 

 state is reached. If, after such a run, the furnace is opened and the carbons 

 examined, it will be observed that, by the action of the current, the electrodes 



have been so shaped as to nest one into the other as shown 

 by the dotted lines in fig. 8. In this way a considerable 

 cross-sectional area is produced over which the discharge can 

 occur. 



The E.M.F. of the arc rises as the pressure of the surround- 

 ing atmosphere increases* and at the high pressures used in 

 the course of this work it becomes more than double the 

 normal value. 



It is, however, the first few atmospheres which produce the 

 greatest effect upon the voltage. 



A detailed analytical investigation of the arc would in 

 itself, as can be judged from the great amount of work 

 carried out at ordinary pressures, require considerable ex- 

 penditure of time and is outside the scope of the present 

 research. We therefore limit ourselves here to the few 

 observations recorded in Table I. (p. 451), in which the 

 behaviour of the arc under certain definite conditions of 

 pressure and current is recorded. 



Fig. 8. Configuration of car- An interesting effect is throughout noticeable. Although 

 bon electrodes after use , L i ,, /. , -, 



the maximum length or arc is so much reduced, the voltasre is 

 m an inert gas under ' ,11 



in aJl cases abnormally high, and consequently a large amount 



of power is concentrated in a small space. 



Two very distinct types of arc exist. The most usual in 

 to the shape of the crater these enclosed furnaces is found with a non-oxidising atmos- 

 of the positive and pro- phere such as carbon monoxide or nitrogen, 

 jects into it as shown by j n 8UC h arcg at ordinary current densities the electrical 



conditions are complicated by the rapid growth of a deposit 

 A mushroom-shaped deposit . 



of carbon accumulates on f Carbon ' chiefl y around the ne g a t iv e electrode and often 

 the extremity of the nega- completely enclosing the end of the positive electrode. The 

 tive electrode. difference between a resistance and an arc is then less marked. 



The arc flame is not visible, but is replaced by a zone of 



brightly incandescent carbon ; electrically the conditions are ill-defined and difficult 

 to reproduce. It is consequently only during the early part of the experiment that 

 concordant measurements can be obtained. As time goes on the carbon deposit 

 builds itself up, the voltage shows a tendency to rise, and the general appearance 



* DUNCAN, ROWLAND, and TODD, ' Electrical World,' 1893, vol. 22, p. 101, for 6-ampere arcs up to 

 10 atmospheres pressure. 



high pressure. 



The end of the negative 

 carbon becomes moulded 



