332 HISTOGENESIS 



supported by experiment. It has long been known from the work of Waller that if nerves, 

 are severed, the fibers distal to the point of section, and thus isolated from their nerve cells 

 will degenerate; also, that regeneration will take place from the central stumps of cut nerves, 

 the fibers of which are still connected with their cells. More recently Harrison (1906), 

 experimenting on amphibian larvae, has shown: (i) that no peripheral nerves develop if 

 the neural tube and crest are removed; (2) that isolated ganglion cells growing in clotted 

 lymph will give rise to long axon processes in the course of four or five hours. 



A second theory, supported by Schwann, Balfour, Dohrn, and Bethe, assumes that the 

 nerve fibers are in part differentiated from a chain of cells, so that the neuron would repre- 

 sent a multicellular, not a unicellular structure. Apathy and O. Schulze modified this 

 cell-chain theory by assuming that the nerve fibers differentiate in a syncytium which inter- 

 venes between the neural tube and the peripheral end organs. Held further modified this 

 theory by assuming that the proximal portions of the nerve fibers are derived from the 

 neuroblasts and ganglion cells and that these grow into, a syncytium' which by differentia- 

 tion gives rise to the peripheral portion of the fiber. 



Efferent Fibers of the Spinal Nerves. At the end of the first month, 

 clusters of neuroblasts separate themselves from the syncytium in the 

 mantle layer of the neural tube. The neuroblasts become pear-shaped, 

 and from the small end of the cell a slender primary process grows out 

 (Figs. 307 and 308). This process becomes the axon or axis cylinder. 

 The primary processes may course in the marginal layer of the neural tube, 

 or, converging, may penetrate the marginal layer ventro-laterally and form 

 the ventral roots of the spinal nerves. Similarly, the efferent fibers of 

 the cerebral nerves grow out from neuroblasts of the brain wall. Within 

 the cytoplasm of the nerve cells and their primary processes, strands of 

 fine fibrils are early differentiated (Fig. 307 B). These, the neurofibrillcz, 

 are usually assumed to be the conducting elements of the neurons. The 

 cell bodies of the efferent neurons soon become multipolar by the develop- 

 ment of branched secondary processes, the dendrons or dendrites. 



Development of the Spinal Ganglia and Afferent Neurons. After the 

 formation of the neural plate and groove, a longitudinal ridge of cells 

 appears on each side where the ectoderm and neural plate are continuous 

 (Fig. 309 A). This ridge of ectodermal cells is the neural, or ganglion 

 crest. When the neural tube is formed and the ectoderm separates from 

 it, the cells of the ganglion crest overlie the neural tube dorso-laterally 

 (Fig. 309 C). As development continues they separate into right and 

 left linear crests, distinct from the neural tube, and migrate ventro- 

 laterally to a position between the neural tube and myotomes. In this 

 position the ganglion crest forms a band of ce?& extending the whole 

 length of the spinal cord and as far cephalad as the otic vesicles. At 

 regular intervals in its course along the spinal cord the proliferating cells 

 of the crest give rise to enlargements, the spinal ganglia (Fig. 358). The 

 spinal ganglia are arranged segmentally and are connected at first by 



