228 Prof. Dove's Experiments in Electricity mid Magnetism. 



e. The induced current does not act on the galvanometer, 

 does not decompose chemically, and cannot magnetize soft 

 iron. 



The explanation of these phaenoniena would be as follows; — 



An electric current when it commences produces in iron 

 near it an electric current of very short existence; during its 

 whole continuance it causes magnetism, and at the moment of 

 its cessation another electric current. The actions of this 

 evanescent magnetism, and of the second electric current, are 

 exactly opposed to each other. 



If the current continued longer (/. e. if it were galvanic), 

 that is, if the magnetism could be fully developed, it would 

 overpower the opposing action of the electric current, but 

 if momentary (or electrical) this latter action overcomes 

 the former. The electrical currents in iron have therefore 

 different properties to what the magnetism of the iron pos- 

 sesses, and as far as regards their actions they follow quite dif- 

 ferent laws; so that in one the electric current overpowers the 

 magnetism, and in another the contrary is the case. 



A bundle of isolated iron wires is the nearest possible reali- 

 sation of an Ampere's solenoid, but it does not behave like a 

 magnet, for it wants a conducting unbroken covering, which 

 first causes it to become a magnet. 



As, moreover, magnetism is excited in iron during the con- 

 tinuance of an electric current, where, as is well known, elec- 

 tric currents are never produced by induction, they appearing 

 only at the beginning and the end, it is evident that those 

 currents which have been hypothetically assumed by Ampere 

 to exist in iron are altogether wanting in experimental proof. 



2. On the Induction 'which the Connecting Wire of the Ley den 

 Jar exerts on itself. 



If man represent the con- ^ J^ 



necting wire of a Leyden jar, 

 and c h h d another connect- <Z 

 ing wire which ends in handles 

 at // h, by the induction of the 

 spiral part of the connecting 

 wire upon itself a shock is ob- 

 tained from chh d. That this 



does not arise from a division ^""-O o ^ 



of the current, is proved by the ^ * 



fact that the accessory shock is not produced when the fol- 

 lowing arrangement is employed. This action is greatly 





