NERVE ELEMENTS AND THEIR FUNCTIONS. 79 



of the spinal ganglion cell there are but two processes from 

 the cell and both these are slender, like neurites. As shown 

 in the last chapter (p. 50) the process which receives the impulse, 

 the peripheral fiber, corresponds in form and time of development 

 to the dendrite of a cell of the usual form (Fig. 39, A and B). 

 In the retina and in the olfactory epithelium the sensory cells 

 seem to have no dendrites but instead only rod-like or hair-like 

 peripheral processes. 



In the cells which are usually regarded as typical nerve cells 

 in the central nervous system of vertebrates the impulse starting 

 in the dendrites passes through the cell-body and along the 

 neurite. Probably in most cells in the central nervous system, 

 however, and in the unipolar spinal ganglion cells the neurite 

 arises from a dendrite at a longer or shorter distance from the 

 cell-body. In these cases the most direct route for the nerve 

 impulse is to pass from the dendrites to the neurite without entering 

 the cell-body. Although it has long been supposed that such is 

 probably the course of the impulse, it has not been possible actually 

 to demonstrate it in vertebrates. In the crab Carcinas, Bethe 

 has found a ganglion in which it was possible to secure exper- 

 imental proof of this hypothesis. In the ganglion which supplies 

 a sensory nerve to the second antenna the sensory cells have a 

 form analogous to that of the unipolar spinal ganglion cells of 

 higher vertebrates. Each cell body gives off a single process 

 which is very much elongated and divides into a dendrite which 

 serves as a sensory fiber to the antenna and a neurite which makes 

 connection with motor cells in another part of the ganglion (Fig. 

 40). Without explaining the arrangement in 'detail, it may be 

 said that the unipolar sensory cells are so placed that by very 

 careful manipulation with fine instruments it was possible to 

 remove the cell-bodies without disturbing the point of division 

 of the single process into dendrite and neurite or the course and 

 terminations of these processes. By artificial stimulation of the 

 antenna, now, it was found that the same muscular reflexes 

 followed as in the normal animal. Evidently the nerve elements 

 perform their functions in the usual way in spite of the cell-bodies 

 being cut away. The phenomenon of summation was also wit- 



