xi ELECTRICAL FISHES 395 



still tilled with yolk. The number of prisms in these was 478, 

 4G7, 443, while in the former it was 471. 



Species of Torpcdinidce (T. marmorata, occllata, panHiera) 

 which are otherwise quite characteristic, exhibit only small and 

 inessential differences in the number of columns. On the other 

 hand, Fritsch confirms the strikingly small estimate (146) in 

 Astrape dipt&rygia, already given by Henle, finding equally low 

 numbers of columns in other species of Narcine, cf. N. tas- 

 iiinniensis, New Zealand, 278; .A 7 , linyula, China, 274; N. timid, 

 230 ; N. indica, 145 ; Astrape capcnsis, 147 ; and Tenter a Hard- 

 wickii, 139. On the other hand, there is an unusually large 

 number of columns in a speckled degenerate type of Torpedo 

 nmnnorata (var. annulata). Fritsch was able in Vienna to 

 investigate two examples of the giant (152 cm. long) American 

 T. (Gymnotorpcdo} occidental-is, in which he found over 1000 

 columns (1037), so that it is natural to regard the example 

 quoted above from Hunter as one of the same species brought to 

 the English shores by the Gulf Stream. To this, the largest species 

 extant, must be added, in view of the number of its prisms, T. 

 (Gymnotorpcdo) hebetans (Lowe), the only specimen of which, in the 

 British Museum, contains 1025 prisms, although it is no larger 

 than a medium-sized T. marmorata. The rare T. (Gymnotorpedo) 

 californica, from the west coast of Africa, is equally distinguished 

 by its small size and large number of columns. 



It is much more difficult to determine the number of columns 

 in Gymnotus, more particularly, according to Fritsch, in the pos- 

 terior section of the body, which presents the greatest structural 

 irregularities. The total sum of all the columns in the large organ 

 seems, from Fritsch's investigations, to vary within a wide range, 

 since it falls below 50 in some instances, while in others it is 

 nearly 100. The greatest number is always found in the smaller 

 individuals of Gymnotus. Whether this is due to arrested develop- 

 ment, or to differences of sex, race, or species, cannot be decided. 



The exact determination of the number of plates in the 

 columns of the organ would be theoretically of great value ; un- 

 fortunately the data are not satisfactory. " There are, on an 

 average, 10 plates to the millimetre in the electric eel, and since 

 the organs are about 80 cm. long in a medium-sized animal, 1 

 metre in length, this would give 8000 plates one behind the 

 other, without reckoning the wide compartments of Sachs' bundle 



