406 



G. ADRIAN HORRIDGE 



Fig. 2. Dorsal view of some sensory axon arborizations in an abdominal ganglion 

 of the larva of the dragonfly Aeschna drawn from methylene blue preparations. 

 This method illustrates the form of the neurons but not their relationships with 

 other neurons and therefore gives no information of the possible occurrence of 

 an anatomically addressed system. The widespread terminations and the general 

 but not detailed symmetry of the two sides are shown. Modified from Zawarzin 



(1924). 



unknown. However, some generalizations may be made about the relevant 

 structures. 



(1) Individual neurons can be recognized both physiologically, e.g. in the 

 crayfish (Wiersma, 1958) or anatomically, e.g. in the ragworm (Smith, 1957) 

 progressively more numerously and more clearly as we pass from the coelen- 

 terates through the worms to the arthropods. 



(2) Widespread activation and depression of probability of motor firing is 

 frequently brought about by widely ramifying neurons, for which only a 

 relatively small degree of anatomical specificity of connexions is required. 



(3) Where neurons can be individually recognized by their definite anatomi- 

 cal relations, their dendrites are nevertheless variable in detail from specimen 

 to specimen and branch apparently haphazardly in the neuropile. The pattern 



