to Electricity and Mag7ietism . 555 



Fig. 12. 



a coil No. 2, h an inverted bell glass, c helices No. 2 and 3. 



the direction of the currents of the different orders. These, 

 in the experiments with the glass cylinders, instead of ex- 

 hibiting the alternations of the galvanic currents (92), were all 

 in the same direction as the discharge from the jar, or, in 

 other words, they were all flus. 



114. To discover, if possible, the cause of this difference, a 

 series of experiments was instituted ; but the first fact deve- 

 loped, instead of affording any new light, seemed to render the 

 obscurity more profound. When the directions of the cur- 

 rents were taken in the arrangement of the coils (fig. 9) the 

 discrepancy vanished. Alternations "were found the same as in 

 the case of galvanism. This result was so extraordinary that 

 the experiments were many times repeated, first with the glass 

 cylinders, and then with the coils; the results, however, were 

 always the same. The cylinders gave currents all in one di- 

 rection ; the coils in alternate directions. 



115. After various hypotheses had been formed, and in 

 succession disproved by experiment, the idea occurred to me 

 that the direction of the currents might depend on the di- 

 stance of the conductors, and this appeared to be the only 

 difference existing in the arrangement of the experiments 

 with the coils and the cylinders*. In the former the distance 

 between the ribands was nearly one inch and a half, while in 

 the latter it was only the thickness of the glass, or about g^th 

 of an inch. 



116. In order to test this idea, two narrow slips of tinfoil, 

 about twelve feet long, were stretched parallel to each other, 

 and separated by thin plates of mica to the distance of about 

 ~th of an inch. When a discharge from the half-gallon jar 

 was passed through one of these, an induced current in the 



* This idea was not immediately adopted, because I had previously ex- 

 perimented on the direction of the secondary current from galvanism, and 

 found no change in reference to distance. 



2 P2 



