THE NERVOUS SYSTEM 557 



is seen in the outgrowth of the neurites of the Rohon-Beard cells in the 

 spinal cord of fishes. These transient nerve fibers extend, devoid of 

 sheath cells, from cells in the roof of the spinal cord to the skin in the 

 extra-embryonic blastoderm. 



The decisive proof of the truth of the process theory, however, was 

 afforded by the beautiful experiments of Harrison, who removed a piece 

 of the spinal cord of a frog or salamander embryo before somatic motor 

 nerves had appeared. Such a fragment, when placed in a sterile nutrient 

 solution, produced elongated protoplasmic processes such as are formed 

 by the Rohon-Beard cells within the normal embryo. No more con- 

 vincing proof of the process theory could be asked and, as a result, this 

 theory is all but universally accepted by students of neurogenesis. 



The origin of the neurolemma and the medullary sheaths, however, 

 presents another problem which also Harrison tried to solve by experi- 

 ment. Observations of sectioned material suggested that, in amphibian 

 embryos, the cells present in embryonic nerves are derived from the 

 neural crest. Harrison therefore removed the skin and neural crest 

 from the back of a salamander larva. The result was striking. Larvae 

 from which the neural crest had been thus removed not only lacked 

 sensory nerves, which develop from the crest, but their somatic motor 

 nerves were wholly devoid of cells. The conclusion seems reasonable that, 

 in amphibia at least, many, if not all, sheath cells are derived from the 

 neural crest and that, in this group, medullary cells do not migrate from 

 the cord into the somatic motor nerves. Later, Harrison was able to 

 observe in the tail of living tadpoles neurolemma cells migrating along 

 naked neurites and thus to confirm the conclusion that neurolemma cells 

 migrate to their definitive position. 



Harrison left unsolved the problem of the origin of the medullary 

 sheath. Speidel, on the basis of observations upon the tails of living 

 tadpoles, has made it seem likely that, in this region, the myelin sheath is 

 formed as a result of a reaction between the neurolemma and the neurite. 

 He has found that the myelin appears soon after the neurolemma encloses 

 the neurite, and the naked neurite seems unable to secrete myelin 

 unaided. 



Histogenesis of Motor Nerves. Somatic motor nerves arise as 

 protoplasmic processes in the manner just described. The neurites 

 become secondarily connected with myotomes by protoplasmic outflow 

 from neuroblast cells, located in the ventral gray column of the neural 

 tube, which are neurofibrillar from their first appearance. The researches 

 of Coghill and others lead to the conclusion that this protoplasmic outflow 

 is a response to stimulus from a bundle of longitudinal nerve fibers which 

 grow caudad along the neural tube. Following the outgrowth of the 

 neurite to the muscle, the dendrites grow in the opposite direction towards 

 the source of stimulus. 



