ON LIVING MALAPTERURUS. 381 



smaller than our troughs, must have depended on their greater 

 strength at that time (see above, p. 376). 



In their own home, where the supply of Malapterurus is un- 

 limited, an observer would not scruple to take the fish out of the 

 water, and thus get rid of the derivation due to immersion in water. 

 When the fish is in the air, this derivation is confined to its body. 

 But I could not venture to do this. I could only as far as possible 

 diminish the water-derivation. 



The fish to be observed was caught in the landing-net and 

 removed to a shallow cylindrical tub, the experimental tub, filled 

 with water. This was made of coarse porcelain (Gesundheitsge- 

 schirr) having a diameter of 295 mm. and a depth of 125 mm. 

 That the removal was not effected without violent shocks was 

 proved in a way which will be stated later. For reasons which 

 will also be stated later, an unsilvered glass plate having the same 

 diameter as the tub was placed at the bottom of it. The water 

 was drawn off by a siphon, until the surface just touched the back 

 of the fish. The mass of water surrounding the fish was thus 

 diminished as much as seemed compatible with its well-being 

 during the experiment, and the discharge could be led off as 

 advantageously as possible without raising the fish into the air 1 . 



In order to lead off the discharge, the means used by Faraday 2 

 for the Gymnotus first suggests itself; this is, to apply to the fish 

 two metal saddles which are insulated excepting at the surface of 

 contact. A telegraph wire, insulated by guttapercha, is soldered to 

 the middle of the arch of the saddle, and serves both as a handle 

 and for leading-off the current. The saddles are made of different 

 shape and size, and of different metals 3 as occasion requires. Gener- 

 ally they are bands of platinum 5 cm. long, I cm. wide, bent to 

 the contour of the fish at different parts of the body. But even 

 the inequalities of zinc mostly vanish as compared with the shock ; 

 if they do not, they are compensated. 



If it is desired to lead as large a proportion as possible of the 

 shock into the experimental circuit, then the contrivance for lead- 

 ing-off must be somewhat different. Faraday, who needed to be 

 even more careful with his Gymnotus than I with my Malapterurus, 



1 The late A, v. Bezold, who was helpful to me in many ways in these experiments, 

 proposed to replace ordinary tap water by aerated distilled water, in order to in- 

 tensify the discharge. 



2 Experimental Researches, etc., loc. cit. p. 5, 1758. 



3 See the following paper, i, 3, 4. 



