CAUDAL DIPLOSPOKDTl.T OF SHARKS. 47 



the dorsal nerve-roots. In some forms, such as SoylUum, the 

 nerve-roota pass out, not through the plates, but between them ; 

 5'et the alternation exists the same (see fig. 1, p. 50). 



In Hasse's monograph on the vertebras ot Elasmobranchs, to 

 which one naturally turns for information on such a point, t!ie 

 figures (7. pi. 34. fig. 14, Scylliiim catulus, and fig. 22, Scyllium 

 canicuJa) give an incorrect idea of the actual appearance of the 

 vertebrae, for the differences between the calcified and uncalcified 

 parts are exaggerated, while the margins of the cartilage plates, 

 which are of far more niorphological importance, are not shown 

 at all. And, further, although Hasse was fully aware of the 

 occurrence of diplospondyly in the tails of these animals (7. p. 21), 

 he has indicated in these figures nerve-apertures on consecutive 

 vertebrae. The error cannot be excused on the ground that the 

 part of the vertebral column figured is anterior to that where 

 dipljspondyly obtains, for the presence of large haemal arches 

 proves the contrary. Tiie figures given by Mayor (13. pi. 18. 

 fig. 12, and pi. 19. fig. 1, Scyllium stellare) are considerablv^ cleirer 

 than those of Hasse, but even they leave much to be desired in 

 the delineation of the boundary lines between the neural plates 

 and spines. 



The portion of the vertebral column of Scyllium catulus 

 ( = stellare) depicted in fig. 1 is taken immediately below the 

 second dorsal fin, and shows the diplospondyious condition in its 

 most typical form. The haemal arches (h) and the crural plates 

 (cr) are fused on the centra (c), but the intercrural plates (ic), 

 alternating with the crural plates, remain distinct. The haemal 

 spines (hs) are not separate from the haemal arches, but the 

 neural spines (ns) are small cartilages which fit with great regu- 

 larity over the intervals between the crural and intercrural plates. 

 The dorsal roots of the spinal nerves issue through the foramina 

 {d) on the posterior margin of every alternate intercrural plate, 

 while the foramina for the ventral roots {v) lie at the lower end 

 of the posterior edge of the crural plates. The dorsal foramina 

 of the right and left sides lie in the same transverse plane ; 

 and similarly wdth the ventral foramina. 



This condition of diplospondyly obta'us not only in the tail of 

 Scyllium catulus, but, as I can testify from personal observation, 

 also occurs in Mustelus vulgaris, Q-aleus canis, Carcharias lati- 

 cauda, Scyllium canicula, Cestracion PJiilippi, Acanthias vulgaris, 

 Spinax niger, and Scymnus lichia. Considerable confusion has 



