142 COMPARATIVE ELECTRO-PHYSIOLOGY 



each other. Under such balanced conditions, therefore, on 

 simultaneous excitation of A and B, the resultant response 

 will be zero. If now, under the modifying action of any 

 external agency, the excitability of A be enhanced, it is clear 

 that the resultant response will be ' up,' showing the greater 

 excitability of the right-hand point. A similar effect will 

 also be produced if the excitability of B be depressed. 

 Similarly the depression of the excitability of A, or enhance- 

 ment of B, would cause a resultant response which would be 

 4 down.' If, again, the two waves of excitation be not of the 

 same phase, we shall obtain various di-phasic effects resulting 

 from the algebraical summation of the constituent response- 

 curves. The resultant zero-response may thus be converted 

 into di-phasic, by the action of any agency which is capable 

 of changing the time-relations of either of the constituent 

 responses. 



I shall now proceed to describe the experimental arrange- 

 ments by which two points in connection with E and E' may 

 be excited, and the resulting electrical disturbances made to 

 interfere with each other. For this purpose we may use the 

 vibrational stimulation which has already been described, 

 with certain necessary additions (fig. 99). The angle of 

 torsional vibration which regulates the intensity of excitation 

 is determined by two stops, P and Q. An elastic piece of 

 brass, B, projects from the torsion-head. When a single 

 stroke is given to this, a quick to-and-fro vibration is induced, 

 the backward pull being supplied by the attached spring, S. 

 The amplitude of this vibration remains always the same, as 

 determined beforehand by the setting of the stops P and Q. 

 The stroke is given by the striking-rod R, set in motion by 

 the turning of a handle. What has already been said about 

 the excitation of the right-hand side of the specimen applies 

 equally to the left-hand, arrangements for the purpose being 

 a duplicate of those just described. After deciding on a 

 suitable angle of torsional vibration for the right, and taking 

 the response at that point, we proceed to adjust the torsional 

 angle on the left, so that the response there may be exactly 



