372 ELECTRO-PHYSIOLOGY CHAP. 



curve does not usually reach the abscissa (the muscle current is 

 not abolished). The length (r, m) corresponds obviously to the 

 period between the moment of excitation and the instant at which 

 the process of the negative variation arrives at the first leading- 

 off electrode, while the time represented by the line (m, o) corre- 

 sponds with the period of the negative variation. In order to 

 conceive the true process graphically, the figures must be imagined 

 one behind the other many times over. While the stimuli follow 

 at equal intervals (t, t\ t 2 , etc.), the closure period (T) is 

 always at the same distance from the corresponding moment of 

 excitation ; the effect on the galvanometer will then be zero. 

 But if the closure of the galvanometer coincides with the beginning 

 of the negative variation, the same impact will be repeated at 



FIG. 117. Schema of rheotome experiment. (Bernstein.) 



each revolution, with a common effect on the magnet. This 

 obviously resembles " that of a constant current, equal in height 

 to the superficial content of all parts of the curves above (T), 

 divided by the time of observation." It is possible in this way 

 to construct the whole curve of variation from consecutive 

 observations, and this has till now, at all events for striated 

 muscle, bee'n the sole means of determining its form and process. 

 This would otherwise have been impossible without applying 

 the method of summation (repetition), since even the most 

 suitable instruments, c.y. the capillary electrometer, are incapable 

 of adequately demonstrating the negative variation which corre- 

 sponds with a single excitation. It is easy, on an aperiodic 

 galvanometer with a very free magnet, to obtain a deflection 

 from the negative variation that accompanies each single 



