30 COMPARATIVE ELECTRO-PHYSIOLOGY 



galvanometric disturbance. Indeed, as the response is 

 electrical, it is obvious that if we are to obtain unimpugnable 

 results, a non-electrical form of stimulus is almost a necessity. 

 But it is only after providing against various sources of error 

 that the electrical form of stimulation can be used with con- 

 fidence. The stimulation caused by mechanical blows can be 

 repeated, it is true, with uniform intensity. But the point 

 struck is subjected to increasing injury, and its excitability 

 thus undergoes an unknown variation. 



I have, however, been able to devise two different modes 

 of stimulation, in which all these difficulties have been 



FIG. 20. The Vibratory Stimulator 



Plant P is securely held by a vice v. The two ends are clamped by holders 

 C c'. By means of handles H H', torsional vibration may be imparted 

 to either the end A or end B of the plant. The end view (b) shows 

 how the amplitude of vibration is predetermined by means of movable 

 stops, s s'. 



successfully overcome, rendering the results as perfect as 

 possible. These are (i) torsional vibration, and (2) the 

 application of thermal shocks. For the obtaining of perfect 

 responses, it must be said here that there is still another 

 condition to be fulfilled. If we wish to obtain the pure 

 effect of stimulus at one contact, say A, special care must be 

 taken that excitation does not reach the second contact, B ; 

 for otherwise, unknown effects of interference will occur. 

 This may, it is true, be obviated by means of the method of 

 relative depression or method of negative variation, so called, 

 to be described in a subsequent chapter. But the experi- 

 mental mode which I am about to describe, in which a block 



