xi ELECTRICAL FISHES 417 



had already proved by potassium-iodide electrolysis that " every 

 point of the fish in water, or of its immediate vicinity, is negative 

 to every point anterior to it on the fish, and positive to every 

 posterior point, the effect being stronger in proportion as the 

 points with which contact is made are farther apart, while it 

 disappears on leading off symmetrically to the sagittal plane." 

 This is intelligible if, at the moment of discharge, the anterior 

 surfaces of all the electrical plates are positive, the posterior 

 negative, as du Bois showed upon a submerged model of prisms 

 made of zinc and platinum elements soldered together (4 d, ii. 

 p. 683). The current is accordingly ascending ("positive" in 

 direction) in the columns of the organ, i.e. directed from tail to 

 head. 



Bilharz, having convinced himself that the nerve entered by 

 the posterior surface of each plate in Malapterurus, also concluded 

 forthwith that the direction of discharge would correspond with 

 that of Grymnotus, without actually being able to perform the 

 experiment. Du Bois-Eeymond showed, on the contrary, that 

 the .discharge in the Malapterurus organ is invariably directed 

 from head to tail, i.e. the opposite of Pacini's rule. This is also 

 true of Raja. 



It was stated at the beginning of the chapter that Faraday 

 had succeeded in demonstrating all the signs of a true electrical 

 discharge (as laid down by him) with one exception, in the shock 

 of electrical fishes (Grymnotus). He obtained physiological action, 

 deflection of magnetic needle, magnetisation, production of heat, 

 spark, electrolysis, attraction and repulsion ; conduction through 

 hot air (flame) alone seemed impossible, a fact already observed by 

 Cavendish, and of which he had failed to find any explanation. 

 Du Bois-Eeymond subsequently pointed out that this is only a 

 special instance of the general fact that, notwithstanding the 

 frequently enormous power of the discharge from electrical fishes, 

 it is unable to overcome even slight hindrances to its passage. 

 This is expressed inter alia in the fact that it is seldom 

 possible in Torpedo and Malapterurus to elicit so-called discharg- 

 ing and closing sparks from the shock ; while, on the other hand, 

 it is easy to get separation sparks. In the first case there is a 

 gap between the stationary or approximating metal points, which 

 the current bridges over at closure ; in the second case a circuit 

 in which current is flowing is interrupted. Du Bois-Eeyniond 



VOL. II 2 E 



