iv ELECTROMOTIVE ACTION IN MUSCLE 371 



tion at an average of 2'927 metres per sec. Its duration is ^^ 



k 3TRJ sec - 



With the aid of this repeating method it is possible to 



decide the magnitude of the negative variation, and to determine 

 whether at the moment when the curve of variation reaches 

 its maximum, the current led off will sink to zero, or become 

 reversed in direction. For this purpose the two mercury dishes 

 are s<> arranged that the closure of the galvanometer circuit (T) 

 is made as short as possible. Moreover, in experimenting, the 

 slider must be placed in such a position that the closure of the 

 galvanometer coincides after eacli stimulus with the maximum of 

 the subsequent negative variation. When this is done, compensa- 

 tion may be shut off, and the first deflection on the galvanometer 

 measured, as produced by the current in the non-excited muscle 

 during the revolution of the rheotome wheel. Now, if the magni- 

 tude of the effect is determined during tetanus at the same 

 rate of revolution, it will obviously depend on the difference in 

 strength between " resting " muscle current and negative variation, 

 in the interval under observation. The direction of the effect 

 shows immediately which current is the strongest. If the current 

 is reversed at the moment of the negative variation, i.e. at the time 

 when it is at its maximum, the scale must turn in the negative 

 direction. Bernstein, however, never found a negative deflection ; 

 the effect was always positive, although, as we should expect, it 

 was much weaker than the corresponding deflection produced by 

 the current in the resting muscle. As a rule, therefore, the 

 curve of variation does not sink to zero. 



The graphic record of these results is a great assistance 

 towards understanding them, as was indeed anticipated in ex- 

 plaining the principle of the rheotome. Let (t, t') (Fig. 117) 

 be the time abscissa, also two consecutive moments of stimu- 

 lation, (h) the height of the resting muscle current, (T) the time 

 of the galvanometer closure, supposed to be movable between 

 (t) and (/). If this occurs in (T') there will be no perceptible 

 deflection, which first begins when the galvanometer closure occurs 

 at T" ; from that point the negative deflection decreases rapidly 

 in magnitude with further alteration of the time of closure, and 

 finally dies out (slowly), so that a curve is produced which falls 

 steeply, and rises up again slowly, without, however (in the after- 

 effect), recovering its original height. The deepest point of this 



