ANATOMY AND CLASSIFICATION OF THE ARENICOLTD^*]. 477 



After completely i^emoving every ganglion-cell from which 

 the motor fibres of the second antenna arise (as shown by the 

 methylene blue method), Bethe still obtained movement of the 

 antenna on stimulation for two days after the operation, but 

 the response became weaker and weaker, and after this 

 interval no reflex action could be induced. This, he suggests, 

 is due to the loss of the normal trophic action of the ganglion 

 cells which were removed; while the efferent impulse is due, 

 not, as is generally thought, to the motor ganglion-cells, but 

 to the neuropile which connects the end of the afferent fibre 

 with the commencement of the efferent one. The ganglion 

 cells are, then, nutritive, and the reflex action can go on for 

 some time when they have been removed so long as this 

 operation is effected without injuring the neui-opile to which 

 they are connected. Slight injury to the neuropile, however, 

 causes a most serious loss of the power of response to stimu- 

 lation. According to this important experiment, of which we 

 have given the merest outline, the series of changes by which 

 an afferent impulse sets up an efferent one, and so determines 

 the character of, e. g., a muscular contraction, takes place in 

 the neuropile which connects the recipient fibre from the 

 surface and the motor fibre to the muscle; while the ganglion 

 cells which are intercalated in the path from one fibre to the 

 other exert an important nutritive influence on the reflex 

 mechanism, but cannot be stated to be essential to the reflex 

 act. 



This experiment certainly suggests that the greater develop- 

 ment of neuropile in the brain than in the cord of Arenicola 

 (and other Polychaetes) may confer upon the brain the power 

 not only of receiving afferent impulses from the prostomial 

 and buccal sense-organs, but also of converting these into 

 efferent ones, which may issue as movements of the somatic 

 or splanchnic musculature as an increased glandular secretion, 

 and in other ways. To determine how fur this is the case, 

 and along what tracts these efferent impulses are carried out, 

 will be the work of future investigation. 



