1918.] E. H. WiiiTETiOTTSE : The Caudal Fins of Fishes. 141 



It is far easier to conceive of the very simple process of fusion of radial 

 with the rigidly attached haemal spine, for this is exactly what would 

 be expected. Without labouring the theoretical aspect further, it may 

 be mentioned that there is abundant evidence among Teleostean fishes, 

 as well as Elasmobranchs and Ganoids that such fusion has occurred ; 

 the following examples may be quoted as affording evidence : Acanthias, 

 Galeus, Heterodontus, Acipenser, Poli/odon, Synodontis, Plotosus, An- 

 guiUa, Conger, Gadus, Gadiculus, Molva, Motella, Centriscus, Belone 

 (Stannius' example), Box, Ze^is, and Pleuronecfes. In all these cases 

 the line of fusion between arch and radial can still be seen. 



There is. however, another aspect worth mentioning ; it has been 

 maintained that radials are derived in the first place from neural and 

 haemal arches by segmentation, a view which receives support from the 

 Dipnoi, where the radials rest directly upon the spines. But the most 

 favoured opinion does not lend support to this view since, except for the 

 caudal, in the median fins of Elasmobranchs radials are so far removed 

 from the axis. Were this segmentation theory correct, it might have 

 been argued that the caudal region retains the primitive condition, but 

 in this connection the presence of radials in the caudal fin would present 

 a difficulty. Thus the study of the caudal fin-structure entirely supports 

 the view that radials are elements developed independently of the axial 

 structures. 



Fig. 3. — C-audal extremity of Torpedo ocellata. 

 hy., liypural ; 1. v., last centrum ; n., neural arcli ; n'., neural arches of centra now 

 suppressed ; r., radial. 



In concluding these remarks on homocercy, I would like to refer 

 to the internal structure of the caudal fin of the electric torpedo, Torjwdo 

 ocellata (fig. 3) ; this fin has never to my knowledge been previously 

 described, and I have refrained from discussing it earlier, because material 

 has not been available to allow me to examine more than a single 

 specimen. But even if this particular specimen was abnormal, which 

 I doubt, it certainly is extremely useful in shewing that tendencies to the 

 homocercal type may occur among the Elasmobranchs ; indeed more 

 than mere tendencies, in fact a homocercy which, had it been found in 

 the Teleostei, would not have been considered strange. 



The specimen was dissected by me some years ago and I believe 

 affords the only instance of homocercy yet recorded among the Elasmo- 



