418 ELECTRO-PHYSIOLOGY CHAP. 



when experimenting on Malapterurus, employed a spark micro- 

 meter, with which two platinum points could be brought as close 

 together as 0*01 mm.; he also made slits in strips of tinfoil, 

 which were not wider than O'O 033-0 '0050 mm. Yet he never 

 succeeded in getting a discharging spark during his investigations 

 with the microscope in a dark room, although an induced current, 

 imperceptible to the tongue, leapt over the same gap with produc- 

 tion of sparks, at 90 mm. distance of secondary coil. 



On the other hand, Santi-Linari and Matteucci on Torpedo, 

 Faraday on Gymnotus, and du Bois-Eeymond on Malapterurus 

 saw separation sparks, when the fish was stimulated by the con- 

 tact of mercury with a platinum point, or by rubbing two files 

 together, or by drawing a spring along a cogged wheel. By 

 means of the frog-interrupter it was possible to open the circuit 

 each time by the twitch of the muscle at the acme of the 

 discharge, when the separation spark always appeared. Dis- 

 charging sparks have been several times observed on Gymnotus 

 only. As early as 1773, Hugh Williamson, in Philadelphia, 

 received a shock through a gap in the circuit, the diameter of 

 which he compared to the thickness of "double post paper"; but he 

 saw no spark. Walsh, on the other hand (as communicated by 

 du Bois-Eeymond, 4 e, p. 158), succeeded in discharging a spark 

 at a slit in tinfoil with a gymnotus brought to London from 

 Guiana in 1775, so infallibly that he was able to demonstrate it 

 1012 times in succession to more than forty members of the 

 Eoyal Society. Sachs again failed conspicuously to produce 

 closing (discharging) sparks in a tinfoil gap of 0-1 mm. Under 

 these Conditions it is not to be wondered at that the shock from 

 Gymnotus fails to pass through rarefied air and to light up a 

 Geissler tube. 



The explanation of all these facts, which are at first sight so 

 remarkable, is simply, as du Bois-Eeymond (I.e. p. 161) showed, 

 that the current of electrical fishes, like all other animal electro- 

 motivity, is due solely to derivation. " In the case of two equal 

 currents A and B, flowing in two conductors of equal resistance, 

 A t however, being in an undivided circuit, while B is completed 

 by derivation, the addition of an equal resistance to both 

 conductors will diminish B more than A, the more so in propor- 

 tion as the resistance of the rest of the circuit is greater." 



" If the fish is connected with a metallic circuit forming a 



