AFTER ELECTEICAL STIMULATION. 259 



The positive after- variation may be investigated by three distinct 

 methods. 



1. Without a compensator, the nerve current is led off from 

 the lower end of the nerve, and a short period is allowed after the 

 magnet has taken up its new position in order to be assured of the 

 constancy of this current. When this is satisfactory, the nerve is 

 tetanised near the other end. As a result of stimulation the magnet 

 swings back for a certain distance. At the end of excitation, the 

 magnet returns to and even passes beyond the position which it had 

 before stimulation, but as a rule comes back immediately, and slowly 

 regains its original position. The positive after-variation is now 

 over. 



It is here of course understood that the magnet swings aperiodi- 

 cally, otherwise its excursion beyond the original deflection would 

 afford no proof of a transitory increase of the nerve current. I give 

 an example of experiments conducted in this manner : 



Warm Frog. Du Bois-Reymond coil. Galvanometer tract 5 mm. 

 Intermediate tract 30 mm. 



Deflection caused by current of electrodes . . 500 to 510. 



nerve current .... 500,, 621 = +121. 



Duration of stimulation 10 sees. Eeturn of magnet . 609 = 12. 



Deflection at close of stimulation . . . . ,,625. 



Immediate return of magnet ,,621. 



Positive after-variation therefore amounts to . + 4. 



2. Experiments conducted with the nerve current compensated are 

 much preferable to the above. The negative variation of the nerve 

 current is now shown by deflection of the magnet in the opposite 

 direction under the influence of the compensation current. At the 

 close of stimulation, the magnet not only returns to its position of 

 equilibrium, but passes beyond it, and then immediately, or after a 

 very short interval comes back and slowly regains its original 

 position. This concludes the positive after-effect. When the object 

 is not to observe the whole course of the positive after- variation, it 

 is desirable to break the circuit of the galvanometer as soon as the 

 positive effect has arrived at its maximum, and the magnet is on 

 the point of returning ; for in this way it is more easy to make sure 

 that the magnet has not spontaneously changed position during the 

 period of the two deflections. 



The following is an example : 



Cold frog. Du Bois-Reymond coil. Galvanometer tract 10 mm. Intermediate 

 tract about 30 mm. 



3 2 



