420 Messrs. J. Frith and C. Eodgers on the 



negative carbon. When the inside purple glow is in good 

 contact with the negative carbon, the resistance of the arc is 

 most negative. When, however, there is a dark space (caused 

 by the presence of foreign matter or of carbon in a different 

 physical state), and hence a lack of contact between this glow 

 and the negative carbon, the resistance becomes increased. 



Hissing Arcs. 



An attempt was made, using Method I., to find the resistance 

 of a direct-current hissing arc produced with accumulators, 

 but it was found that even with the alternator at rest there 

 was a large deflexion of the electrometer, showing that the 

 current through a hissing arc is oscillatory. In order to deter- 

 mine the frequency of the oscillation, a condenser and 

 telephone were put across the arc. The sound heard in the 

 telephone was, however, of no definite pitch. 



A method suggested by Mr. Campbell was used to measure 

 the amount of this oscillatory current. The current taken by 

 the hissing arc was passed through the primary of a trans- 

 former, the secondary of which was put in series with a 

 reflecting twisted-strip ammeter, the detiexion being noted. 

 The arc lamp was then disconnected, and a known alternating 

 current was sent through the primary, of" such magnitude 

 that the deflexion was the same as before. This known 

 current has the same R.M.S. as the oscillatory current passing 

 through the arc, for if the impedance of the secondary is 

 large compared with its resistance, the deflexion of the am- 

 meter is the same for the same R.M.S. current whatever 

 the frequency. The arc was run at 14"5 amperes and 35 

 volts, Apostle carbons +9 mm. —9 mm. being used. While 

 the arc was burning the deflexion of the ammeter varied 

 between 200 and 250 scale-divisions. Primary currents of 

 0*43 ampere and 0*53 ampere, as measured by another 

 twisted-strip ammeter, produced these scale -readings respec- 

 tively. Hence in this case the R.M.S. of the oscillatory 

 current amounted to about 3 per cent, of the continuous 

 current. 



Loiv Frequencies. 



dV 

 On comparing our results with the steady — r- taken from 



the slope of the curves published by Mrs. Ayrton, the agree- 

 ment as to sign, which held in the case of solid carbons, was 

 found to break down in the case of cored carbons ; for while 

 the general shape of the curves obtained by Mrs. Ayrton was 

 the same for cored as for solid carbons, that is both curves 



