552 M. A. dc la Rive on the Aurora Borealcs, 



from the N. to the S. It is essential that care be taken, before 

 exhibiting; this second derived current, to assure oneself that the 

 metallic plates have lost the secondary polarities which they had 

 derived from the effect of the first. For these polarities give 

 rise of themselves to a current moving likewise from the S. to 

 the N., since it is the contrary of that which has produced them. 

 This current is added to that which is derived from the altera- 

 tion of the position of the discharge, and greatly augments its 

 intensity. It shows itself alone if the discharge ceases momen- 

 tarily at the north pole without taking place immediately at the 

 south pole ; and, as I have already remarked above, it may of itself 

 alone be almost as strong as that which has given rise to it. The 

 same phenomena present themselves, precisely in the same way, 

 except that the direction of the current is reversed, when the two 

 metal plates are fixed in the southern hemisphere of the sphere, 

 instead of being in the north. 



To resume, if we fix, on the moistened band which represents 

 a meridian, two similar and equidistant pairs of metal plates, one 

 on one side of the equator, the other on the other side, and ob- 

 serve the derived currents which they exhibit, we ascertain that 

 the current exhibited by the plates of the northern hemisphere 

 is directed from the N. to the S. when the discharge takes place 

 at the north pole, and from the S. to the N. when the discharge 

 is transferred to the south pole ; we ascertain likewise that that 

 which is exhibited by the plates situate in the south hemisphere 

 is directed from the S. to the N. when the discharge takes place 

 at the south pole, and from the N. to the S. when it is removed 

 to the north pole. It is readily conceived that a difference in 

 the intensity of the discharges which take place at the two poles 

 simultaneously, is sufficient to produce the same effects, only in 

 a slightly less degree than when there is a complete cessation of 

 the discharge at one of the poles, accompanied by its appearance 

 at the other. This is precisely what takes place in nature, and 

 it explains all the variations in the movements of the galvano- 

 meters placed in the circuit of telegraphic wires, which accom- 

 pany so faithfully the different phases through which pass the 

 electric discharges constituting the auroraB boreales and australes. 



I have already indicated how these variations explain the dis- 

 turbances of the needle, which I have also succeeded in reprodu- 

 cing artificially, either independently of the other phenomena, 

 or simultaneously, by causing the same discharge which is com- 

 municated to the apparatus just described, to pass through a 

 surface of mercury above which a magnetic needle is delicately 

 suspended. These disturbances, being the result of the direct 

 action of the terrestrial currents upon the magnetic needle, are 

 independent of the secondary polarities, which play an important 



