xi ELECTRICAL FISHES 375 



cells in the spinal cord of Gymnotus, and showed that there were 

 twice as many as in other fishes, indicating a probable relation 

 with the electrical nerves that pass out from this region. Du 

 Bois-Keyinond also concludes " that there must be a similar 

 structure in the cord of the electrical eel to the lobus electricus of 

 the electric ray." This is the more probable since the electrical 

 organ of Gymnotus is supplied by intercostal nerves, which are so 

 numerous as to have excited attention from the time of Hunter. 

 Fritsch's thorough investigation of the materials brought by C. 

 Sachs from Calabozo showed in point of fact that at a certain 

 level of the cord, which is differently localised according to 

 individual variations (1223 vertebras), large ganglion -cells, highly 

 characteristic in their entire habit, appear at first singly, and 

 later on as a closed column, in the form of a cylinder, surrounding 

 the central canal, and open anteriorly ; these must undoubtedly 

 be denoted " electrical cells." " They exhibit the usual multi- 

 polar characteristics of well-nourished finely-granulated protoplasm, 

 drawn out into several broad processes and bladder-like nuclei, 

 with conspicuous, strongly-refracting nucleolus." The size of the 

 rounded cell-body, which is invariably drawn out into a well- 

 marked axis-cylinder process, is on an average 0*051 mm. The 

 ordinary motor cells exhibit a more polygonal form, with better 

 developed protoplasmic processes. No sharp division can, how- 

 ever, be predicated, since, according to Fritsch, there are transitional 

 forms at the level of the 616 cervical vertebrae. At about 

 the 30th vertebra the quantity of electrical cells has increased so 

 much " that the entire cavity of the anterior horns, and central 

 mass of the gray matter, seems to be filled with them, and they 

 even cause a thickening of the spinal cord in the sagittal direction. 

 It is only in front of the central canal that the cells do not meet, 

 on either side, so that the cell-group in cross-section resembles a 

 wide half-moon " (Fig. 244). " Here, where the seat of origin of 

 the electrical nerves appears most fully developed, the axis- 

 cylinder given off from the cells forms at each cross -section a 

 distinct group of fibres, which run very obliquely. These coincide 

 in arrangement and mode of origin with the ordinary anterior 

 roots of other bony fishes. There are no special motor roots in 

 the cord of Gymnotus alongside of the electrical, but the fibres 

 from which the muscular nerves originate, succeed immediately 

 upon the electrical nerves (Fritsch). Corresponding with the 



