AFTER ELECTRICAL STIMULATION. 267 



In order to get more accurate data of the relationship in amount 

 between negative and positive variation, it would be necessary to make 

 observations with two galvanometers at the same time, as already 

 mentioned. With a stimulation of short duration, on account of the 

 inertia of the magnet, the deflection of the needle does not cor- 

 respond to the amount of negative variation : before it has reached 

 a point in its deflection which would correspond in measurement to 

 the decrease of the nerve current, the stimulation is already at an end. 

 On this account all values recorded and quoted above of negative 

 variation were far too small when the duration of stimulation was 

 short (less than about 10 seconds). This drawback does not hold good 

 for positive variation, because in this case one may always await the 

 maximum deflection of the magnet. Yet even the value so found does 

 not give a correct representation of the amount of positive variation, 

 because this, as has been already stated, rapidly declines, and indeed 

 in all probability too quickly for the sluggish magnet to make an 

 excursion corresponding to the original amount of the variation. 

 Hence it is desirable to repeat these experiments under more favour- 

 able circumstances. 



The relationship in amount of positive and negative variation 

 is best determined by experimenting with stimulations of equal 

 strength and duration alternately with and without a switch, as has 

 already been discussed. In this manner it was shown that the 

 amount of positive variation not unfrequently reached half the 

 amount of negative variation, seldom exceeded that proportion, but 

 generally fell short of it. 



The duration of positive variation under different circumstances I 

 have not accurately determined. In cases where positive variation 

 was considerable, minutes might pass before it fully died away. In 

 most cases its duration is measured by seconds. 



The variation of the nerve current may assume a totally different 

 aspect if stronger induction currents be employed to tetanise the 

 nerve, for then unipolar actions mix with them, especially if the 

 break-shock is used. With what distance of coils and with what 

 strength of primary current this occurs, must be specially deter- 

 mined for each form of experiment by special control experiments. 



Du Bois-Reymond has already pointed out that, even with the 

 most careful insulation of the circuit of the secondary coil on the one 



