58 Art. l.-K. Fuji : 



As a mere experimental result, it seems that l)efui'e the so- 

 called refractory period there exists a period in wliieli the response 

 may occur. The modal latent period or the latent period of the 

 second discharge seems to decrease even to a smaller value than 

 that of the first for some value of the interval between the two 

 stimuli. Indeed the shortening of the latent period from the 

 normal is equal to (3 nnn. in the maximum length on the oscillo- 

 gram. But in other experiments we know that, while the modal 

 latent period gives very consistent values, the latent periods differ 

 widely according to the height of the curve. It will not therefore 

 be useless to calculate roughly the value of the displacement of the 

 summit of the second discharge curve on account of its lying on 

 the descending branch of the first. Suppose a discharge curve 



y = Ae~^''^^^''Tj superposed on an inclined straight line 



ij = (p — x)ta,n(f. 



Then the resultant curve is expressed by 



y = (p-x)tan<f + Ae-^''^''^''X' 



Differentiating the expression with respect to x, and equating the 

 result to zero, we have 



t ■'■0 J^ , 



A value of .i- satisfying the equation must be that for tlie maximum 

 value of //. 



Let tho displacement of the summit l)e represented l)y o, 

 which, in the first approximation, may be considered very small 

 when compared witli .^o. l^utting x=^Xa—d in the above expression 



and neglecting the higher orders of — than the first, we have 



*- Xq 



^n^tan^ 



2.46'-' ^ ^ 



Taking the third experiment in the oscillogram No. öO, we may 

 put very roughly 



