NEAL: NERVOUS SYSTEM IN SQUALUS ACANTHIAS. 209 



and the motor components of the dorsal nerves. Furthermore, we 

 must determine the primitive relations of medullary nerves, not by the 

 place of exit of their fibres (i. e. by their roots), for we know these to be 

 variable, but by the position of their "Kerne" in the walls of the neural 

 tube.^ 



There is considerable difference of opinion as to whether nerves 

 ("roots") arise primarily from the expanded portion of the encephalo- 

 mere (or myelomere), or from the constrictions between these segments. 

 As early as 1878 Marshall said. " My investigations tend very strongly 

 to prove that all the nerves arise primitively from the widest parts of 

 the dilated vesicles, whether of brain or cord, and never from the inter- 

 vening constrictions." Later, McClure ('89), who is in agreement with 

 ^larshall as well as with Orr, Beraneck, and Waters, said, " The dorsal 

 roots of spinal nerves take their origin from the apex of their respective 

 myelomeres in exactly the same manner as the nerves of the medulla 

 do from their respective encephalomeres." Miuot (92) criticises McClnre 

 for overlooking the fact that the " ueuromeres can have no genetic rela- 

 tion to the ganglionic nerves." The ground of Minot's statement does not 

 seem to me to be so self-evident as not to be in need of explanation.^ 



In disagreement with McClure, Miss Piatt ('89) claimed that "the 

 concavity in both medulla and spinal cord is the source from which the 

 nerve originates," and her conclusion, which Minot accepts, is that 

 the origin from the expanded portion of the neuromere is secondary. 

 In view of this difference of opinion it is of interest that Balfour ('85) 

 stated that in Selachian embryos the dorsal and ventral roots of spinal 

 nerves alternate with each other, the dorsal roots being intersegmental 

 (intersoraitic) and the ventral roots segmental (somitic) in position. 

 Miss Piatt did not, however, in her statement of nerve relations make 

 a distinction between dorsal and ventral nerves. 



1 Tlie most serious obstacle to the use of this criterion is the difficulty of apply- 

 ing it in those early stages of development when raetameric relationships appear 

 least modified. Martin ('90 and '91, p. 230) has noted an ontogenetic ventral 

 shifting of motor " Kerne " in the eat. 



2 It is to be regretted that McClure gave no figures of tlie nerve relations of 

 myelomeres. Minot apparently assumes that the neuromeres are constituted 

 solely in adaptation to a motor segmentation, and therefore that the neuromeres 

 are segmental localizations of ganglionic cells (i.e. motor "Kerne") in the wall of 

 the neural tube, just as are the segmental ganglia of Annelida. It seems to me 

 therefore that McClure might have met Minot's criticism by reminding him that 

 neurologists have recognized in the medulla groupings of ganglion cells which are 

 in relation with sensory fibres, i. e. sensory '" Kerne " or " Endkerne " (see Edin- 

 ger, '96, p. .366), and may well contribute to the metameric enlargements. 



VOL XXXI. NO. 7. 5 



