448 Mr. J. Henry on the Deflexion by Magnetic 



That this is the case is shown pretty clearly hy Fed- 

 dersen's observations (Pogg. Ann. ciii. p. 69, 1858) on the 

 appearance presented by a succession of sparks in a revolving 

 mirror, when a blast of air was directed across the electrodes 

 (Recent Researches, p. 132, par. 128). 



Feddersen's sparks were obtained from the oscillatory dis- 

 charge of leyden-jars, and the blast was simply that caused 

 by the rotation of the mirror. Each discharge gave a series 

 of sparks following one another very rapidly, the sparks at 

 the beginning of each series beina* less separated in the 

 revolving mirror than those towards the end of the series. 



In these experiments the first spark in each discharge was 

 observed to pass straight between the electrodes, while the 

 succeeding ones were more and more bent by the blast, as 

 shown in fig. 16 : this figure gives the appearance presented 

 by the sparks at different parts in the series of oscillations 

 forming one discharge; a is the beginning of the series, and 

 b the end. 

 a Fk. 16. h 



These observations show that between successive discharges 

 the ionized gas was blown away or scattered, so that each dis- 

 charge found it easier to begin along the straight path ; but 

 the interval between successive sparks in the same discharge 

 was so small that the thread of ionized gas was bent only a 

 little more by the blast each time while it retained its con- 

 ductivity ; the increase in the deflexion between successive 

 sparks, especially at the beginning of a series, was very small, 

 but before the end of a series was reached the bending was 

 very evident, as at b, fig. 16. 



When the electrodes of the secondary of an induction-coil 

 are placed near enough together to form an arc, the gentle 

 current of air across the electrodes produced by the heat of 

 the discharge bends the arc upwards ; if the current of air 

 be much increased, instead of bending the arc still more it 

 will destroy the arc altogether, and the discharge will pass as 

 a spark straight between the electrodes with an aureole blown 

 aside. A blast of air is used to destroy in this way the 

 sparking- at the commutators in some dynamos. 



