THE CENTRAL NERVOUS SYSTEM 



663 



cells of the cerebral, and the Purkinje's cells of the cerebellar cortex, 

 have small swellings, the so-called lateral buds or gemmules, on their 

 course. Their significance is unknown. The dendrites terminate at a 

 ttle distance from the cell, where they come into relation with the end 

 brushes of the axons of other neurons. In this way two or more 

 neurons are linked together to form a nervous path. The relation is 

 not one of actual anatomical continuity, but the processes come so 

 close together that nerve impulses are able to pass across from the 

 terminal brush of the axon of one nervous element to the dendrites 

 or cell-body of another. Foster and Sherrington call the connection 

 a synapsis. 



It has been suggested that the contact may be rendered more or 



FIG. 229. 



a e shows the development of the pyramidal nerve-cells of the cerebral cortex in a 

 typical mammal ; a, neuroblast with commencing axon ; b, dendrites appearing ; d, com- 

 mencing collaterals. A-D shows the different dt-gree of complexity in the fully- 

 developed pyramidal cells in different vertebrates : A, frog ; B, lizard ; C, rat ; D, man. 

 (Donaldson, after Ram6n y Cajal.) 



less close through amoeboid movements of the dendrites, and that in 

 this way the nervous impulse may be switched like a railway train 

 from one path to another. But there is at present scarcely any 

 experimental basis for this fascinating hypothesis. Whatever the 

 nature of the relation between two superposed neurones may be, it 

 does not permit the conduction of nerve-impulses indiscriminately 

 in both directions. For instance, stimulation of the central end of 

 the posterior root of a spinal nerve causes a negative variation (p. 630) 

 in the anterior root of the same segment, while no electrical change 

 is produced in the posteiior root by stimulation of the anterior. 

 We shall see later on (p. 682) that some of the fibres of the 



