306 ELECTRO-PHYSIOLOGY CHAP. 



of columns" (du Bois-Keymond). Valentin gives only 5150, 

 Pacini only 4000. Hunter reckons 150 plates in a column 25*4 

 mm. long of a medium-sized ray; Leukart, 180; Pacini, reckon- 

 ing the height of column at 40 mm., counted 2000 plates, while 

 Valentin only finds about 300 plates in medium columns (of 11*3 

 mm.). The figures thus vary considerably, as is not surprising 

 when one reflects on the difficulties of enumeration, even with 

 the most favourable conditions of preservation. Fritsch (12 g, ii. 

 p. 1105) estimates the number of plates in a column of Torpedo 

 (Finibriotorpedo) marmorata (length of body, 265 mm.) 13 '5 mm. 

 high, at about 375; since the organ contains 479 columns, the 

 total number of plates would be 179,625 ; in T. ocellata, with an 

 average column number of 433 (height of column usually 6 '2 5 

 mm.) and a content of 380 plates, the sum total would be 

 164,540 plates. These measurements also bring out the further 

 and striking point that " the plates are closer together in the lower 

 than in the higher columns of the same organ," so that the 

 growth of the latter is in this respect also " a process of swelling, 

 leading to the divergence of the plates," which on growing increase 

 in diameter, as found by Boll. 



There is another method of determining the number of plates 

 in the organ of Torpedo. If, as cannot be doubted, each fibre 

 of the electrical nerve is to be regarded as the axis-cylinder 

 process of a ganglion- cell of the electric lobe, it is evident that 

 definite and regular relations must exist between the number of 

 cells and the number of plates in the entire organ. If the total 

 number of cells = N", these will, by means of the correlative N 

 axis-cylinders, which each divide into 18 branches, and supply 



1 8 



the 6 corners of each plate, innervate NX = 3 N plates. From 



6 



this point of view the enumeration of ganglion-cells in the lobe 

 is of great interest. After Boll had undertaken a research in 

 this direction, estimating a number of 53,760 cells, which is far 

 too low, Fritsch adopted the much safer method of counting the 

 axis-cylinders in the electrical nerves by photographing sections 

 of the four nerve-trunks. He obtained a total of 58,318 nerve- 

 fibres, which on multiplying by 3 gives the number of plates as 

 174,964. This agrees with the number as given above at 

 179,625 sufficiently to justify the method. 



