Curves of the Oscillating Spark. Ill 



This low temperature enables us to explain the occurrence 

 of the glow p.d. between the oscillations. If the temperature 

 of the anode were the same as that of the cathode, there is 

 no reason to suppose that on the reversal of the current the 

 arc character of the discharge should not remain unchanged, 

 for the air in the gap must still be strongly ionized. But on 

 the reversal of the current at the first instant what is now 

 the cathode is comparatively cool, and it cannot supply the 

 necessary corpuscles at a sufficient rate. Since the air in 

 the gap is already ionized, we have now the exact conditions 

 necessary for the glow discharge. By the removal of the 

 negative ions from the neighbourhood of the electrode, the 

 potential drop becomes concentrated in that region until it 

 reaches such a value that the positive ions ionize the air 

 there, giving the necessary supply of negative ones at the 

 electrode. In the rest of the field at this potential only 

 the negative ions will be able to ionize. The approximate 

 agreement of the glow p.d. with the cathode fall of potential 

 in vacuum-tubes, and its constancy under varied conditions, 

 confirms this explanation of it. The glow p.d. must be of 

 very short duration, for the new cathode is now being- 

 bombarded by the heavy positive ions, its surface rapidly rises 

 in temperature, and when sufficient corpuscles are emitted to 

 convey the current from the electrode to the gas, the arc 

 period supervenes. 



We shall now consider briefly a number of features of the 

 spark which are elucidated by this view. 



1. The Core and Sheatli. — The condensed spark in air 

 exhibits a very bright white central part or core, surrounded 

 by a much less bright sheath or aura. On the view taken 

 above, the core marks the region of the spark in which the 

 initial ionization by positive ions occurs. The maximum 

 field naturally occurs only at the centre of the spark, 

 surrounding this there is a region in which the field is only 

 sufficient to produce ionization by negative ions, and the 

 sheath may be associated with this region*. If we consider 

 a series of sparks in which the self-induction of the circuit is 

 gradually diminished from large to A T ery small values, at the 

 larger inductances there is no differentiation of the spark 

 into core and sheath; then the core forms, and gradually 

 grows brighter and brighter up to a certain point, after that 

 its width begins to increase, its intrinsic brightness remaining 

 as far as can be judged constant. In fig. 9, PL XX., 

 photographs of a series of sparks, 7 mm. in length, and 



* The glow discharge in the oscillations no doubt also contributes to 

 the sheath. 



