384 ELECTRO-PHYSIOLOGY CHAP. 



cross-section ; and finally the " alveolar " layer, once more richly 

 set with nuclei, which Ewart (10) aptly compares with the bayed 

 inner wall of the frog's lung. From this layer there is, moreover, 

 a long stalk-like process which extends backwards through the 

 mucous tissue of the compartment, and ends in the connective 

 tissue wall of the chamber (Fig. 251). 



In very young embryos of E. latis there is no trace of any 

 electrical organ as such. It is only at a length of about 

 7 cm. that a horizontal longitudinal section through the tail 

 exhibits, between the ordinary normal fibres, others which must 

 be regarded as in the first stage of transformation into electrical 

 plates (Fig. 252, a). This is at first shown only by a club-like 

 swelling of the anterior end (where a nerve enters with copious 

 bifurcation), and a striking proliferation of nuclei, resulting 

 finally in a sort of end-plate or cap, with which the head of the 

 club is invested. The further stages of transformation are intelli- 

 gible without more description from the figures (Fig. 252, 5, c, d), 

 drawn from preparations from an older embryo 10 cm. long. 

 Here it is easy to recognise all the essential parts of the developed 

 electrical plate the nucleated nerve end-plate at the top, the 

 non-nucleated, striated meandering layer, with its undoubted rela- 

 tion to the original cross-stria tion of the muscle-fibre, and lastly 

 the alveolar stratum (e,f) derived from an exuberance of the sarco- 

 plasm and (muscle-) nuclei at the base of the club, with the tail- 

 like appendage of the atrophied stalk of the plate as the remains 

 of the original fibre. This long exhibits indications of cross- 

 striation, and may under certain conditions be detected even in 

 the perfectly developed plate. 



The electrical organ of E. circularis is longer, but much 

 smaller. It consists of curved goblet-shaped plates, correspond- 

 ing essentially with those in E. batis, where there are already 

 indications of a goblet-like pitting of the initially club-shaped 

 and swollen ends of the muscle-fibres. 



The approximately parallel lamellae of the meandering stratum 

 in E. latis are much curved in E. circidaris, and the alveolar 

 layer also exhibits a different structure, and presents a fine 

 radial striation at its base (Fig. 253). The first trace of a 

 transformation of muscles into electrical tissue is, however, 

 ascribed by Ewart to Raja radiata, as shown on the one 

 hand by the much later commencement of the process, on 



