GLANDS AND ELECTRIC ORGANS. 241 



at first only between the electrodes, but afterward 

 beyond these, which change then corresponds with a 

 change in the distribution of tension on the surface. 



7. We have yet to consider how far the hj^pothesis 

 under discussion explains the electric phenomena in 

 electric fishes and in the glands. The electric shock 

 of the torpedo must evidently be regarded as analo- 

 gous to negative variation in muscle- and nerve- 

 currents. The apparently great difference that in the 

 latter a current present during a state of quiescence 

 becomes weaker during activity, while in electric fishes 

 an organ which is entirely inoperative during the state 

 of quiescence generates a current when it becomes 

 active, appears, when closely examined from the point 

 of view afforded by our hypothesis, to be of no account. 

 For, from the fact that no current in an organ can be 

 externally shown, it by no means follows that no elec- 

 tromotive forces are present within the organ. A piece 

 of soft iron is in itself entirely non-magnetic ; but as 

 this may at any time bo transformed into a magnet by 

 bringing a magnet into its neighbourhood, or by the 

 influence of an electric current, we suppose that mole- 

 cular magnets are present even in the soft iron, though 

 these a,re not regularly arranged as in a regular magnet, 

 . such as that represented in fig. Gl, p. 230. The action 

 of the magnet which is brought near, or of the electric 

 current, therefore consists solely in the fact that it ar- 

 ranges the irregularly placed molecular magnets within 

 the soft iron, and thus allows their action to appear 

 externally. If no magnetic action were known in soft 

 iron, no one would ever have had an idea that magrnetic 

 forces were present within it. But comparison with the 

 permanent magnet, and the possibility that thoroughly 



