422 ELECTRO-PHYSIOLOGY CHAP. 



of the discharge from the anterior and posterior half of Gym- 

 notus, in the same sense as that previously noted by du Bois- 

 Eeymond in Malaptcrurus (4 d, p. 630), where the anterior 

 half gave much stronger deflections of the galvanometer than 

 the posterior (in the ratio of about 11:6). Since this differ- 

 ence, as shown by du Bois-Eeymond, disappears with increasing 

 resistance of the experimental circuit, there is no reason, in 

 Malapterurus at any rate, to assume a different E.M.F. in the 

 two halves. The diminishing diameter of the fish (or of the 

 organ) in the antero- posterior direction, with the consequent 

 diminution of resistance in the same direction, sufficiently explains 

 the reaction. In Gi/mnotus there is the further possibility that 

 the posterior prisms of Sachs' bundles (with wide compartments) 

 may give a different electromotive reaction from those with 

 small chambers. 



Under all circumstances the shock increases here with the 

 length of the fish, and the question then presents itself whether this 

 is due to diminution of resistance, or to increase of E.M.F., or both. 

 As appears from comparison of the length and weight in different 

 animals, Gi/mnotus grows more in length than in diameter, so 

 that its cross-sections are relatively smaller in proportion as their 

 length increases ; and since we may assume that the reaction of 

 the electrical organ will be the same, its resistance also will 

 diminish more slowly than if the organs remained parallel in 

 their growth, or it may even be augmented. In any case the 

 greater intensity of shock in longer fishes must be referred to 

 increase of E.M.F. , and not to diminution of resistance (du Bois- 

 Eeymond). 



The anatomical relations of innervation in the electrical organs 

 of the several electrical fishes show considerable differences as 

 regards the initiation of spontaneous (voluntary) and reflex dis- 

 charges. In Torpedo it might be predicated that the discharge 

 after destruction of the electrical lobe or sensory nerves leading to 

 it could only occur from excitation of the electrical nerves, or 

 of the electric lobe itself. In Malapterurus, too, the property of 

 spontaneous and reflex discharges must be associated with the 

 integrity of the two giant ganglion-cells. In Gymnotus, on the 

 other hand, the innervation of the organ is evidently more analo- 

 gous to the muscular innervation of the fish. Humboldt found 

 no shock from the decapitated gymnotus, so that when an animal 



