2^2 COMPARATIVE ELECtRO-PHYSlOLOGY 



different tissues. The constancy of the speed of rotation of 

 the alternating-current dynamo is secured by means of the 

 resonating index, I. This consists of a short steel spring 

 with a long index. When the frequency of alternation is 

 the same as the natural period of vibration of the spring, the 

 resonator is thrown into strong sympathetic vibration. At 

 first the rheostatic resistance, which determines the speed of 

 the motor, is made slightly too large. The movable plate is 

 now gradually brought nearer, till the proper speed has been 

 arrived at, and this point is at once indicated by the induced 

 vibration of the resonator. 



A further difficulty has to be overcome in the main- 

 tenance of the uniformity of speed. When the open circuit 

 of the alternating dynamo is closed, by the interposition of 

 the primary of the RuhmkorfFs coil, the speed undergoes 

 a sudden diminution, owing to the work which the dynamo 

 has now to perform. In order to avoid this fluctuation, 

 then, the dynamo circuit is kept closed by means of an idle 

 primary coil, p f , which is a duplicate of the primary, P, of the 

 Ruhmkorff's coil. When the key, K, is pressed, the alter- 

 nating current is transferred from P' to P. There is thus no 

 fluctuation in the speed of the dynamo, and the duration of the 

 closure determines that of the stimulus. I may mention 

 here, that instead of employing a separate motor to drive 

 the alternating-current dynamo, I have sometimes used, with 

 equal success, a motor transformer, giving rise to alter- 

 nating currents. It is easy to construct a very compact and 

 portable form of this latter apparatus. 



In this manner we may apply uniform stimuli of equi- 

 alternating shocks at regular intervals of time, say of one 

 minute. The usual preliminary test of the successful 

 elimination of all sources of disturbance may here be made 

 in the following way. The kaolin ends of the non-polaris- 

 able electrodes are connected with each other, without the 

 interposition of a specimen, and alternating shocks from the 

 secondary are passed through the circuit. These should 

 give rise to no deflection in the galvanometer. It may be 



