134 Prof. J. C. PoggendorfF on the Induction Apparatm 



itself. The proof of this lies in tlie fact, that the current of 

 the induction wire always passes to and fro, even when its poles 

 do not touch the coatings, but send sparks to them across a 

 stratum of air ; in the latter case, indeed, it is the current con- 

 sequent upon breaking the inducing current which alone oscil- 

 lates. If behind the Lcyden jar or the coated glass plates, a gal- 

 vanometer be introduced into the circuit, no deviation is obtained; 

 whereas this never fails when the induction current is simply 

 interrupted by a thickness of air across which sparks can pass*. 



This to-and-fro motion of the current is still more perceptibly 

 and beautifully shown by means of the electric egg. So long as 

 it is alone in the circuit, the knob connected with the negative 

 pole is the only one which is surrounded by the well-known 

 beautiful blue light ; but as soon as the Leyden jar or the coated 

 glass plates are also introduced into the circuit, both knobs ap- 

 pear blue, and when the vacuum is tolerably good, the positive 

 flame is also visible above the blue light ; not, of course, because 

 two opposite currents simultaneously traverse the partial vacuum, 

 but because they follow one another at such short intervals of 

 time that the eye is no longer sensible of the changes f- 



Insulating liquids, such as oil of turpentine, when placed be- 

 tween platinum plates and introduced into the circuit, produce 

 the same effects as the Leyden jar. 



The experiments with the egg are very instructive, especially 

 when a second circuit also exists. When it is introduced into 

 the primary circuit Pc or Nc, or into the branch cc at the point 

 dd, only one knob is blue. When it is introduced in ca or cb, both 

 knobs are covered with the blue light. 



It is worthy of mention, that the jar ab has no influence upon 

 the luminous phseuomenon in the branch cc, provided the egg be 

 well exhausted, and the induction current itself be sufficiently 

 intense. If air is allowed to enter, however, or if the intensity 

 of the current is diminished, by substituting a long for a short 

 induction wire or by dispensing with the condenser, then the 

 introduction of the jar at ab augments the luminous phsenomenon 

 at dd, though in a much less degree than with sparks in free air. 



* If, instead of the coated glass plates, an oiled silk condenser is intro- 

 duced into the induction circuit, a continuous crackling is heard, but at the 

 same time a strong deviation is obtained when a galvanometer is also intro- 

 duced. The crackling in this case, therefore, is caused by an actual 

 passage of electricity, by the breaking through of sparks. At first, when 

 I employed a condenser W'hose coated surfaces were separated from one 

 another by a double layer of paper, covered on each side with amber 

 varnish, I obtained the phEcnomena of the glass plates. 



t As Riess has already shown (Pogg. Ann. vol. xci. p. 290), the same 

 optical illusion is obtained w hen, by means of the pin already described 

 (Phil. Mag. vol. x. p. 6), the galvanic current is made to traverse the in- 

 ducing wire alternately in opposite directions. 



I 



