54 Art. 1.— K. Fuji : 



The numbers enclosed in the brackets are calculated l)y the present 

 author. 



From these data we calculated M and X. The vahies of the 

 normal latent period could not be discovered in his paper and we 

 therefore found them as follows: — 



Taking thive ordinates //,, y-i and i/i on the figure (Plate XII.) 

 Avith equal intervals in succession, we put 



//i = / + Ö, 

 ?/._, = / + mh, 



1/3= ^ + "i' ^. 



where I is the normal latent period and m a constant to be de- 

 termined by the period of recovery. So in Exp. 3 we may find 

 7=0 0019^ and m=OàOr,, the interval between the successive ordinates 

 being 002^5 sec. From these values we obtain M =00098, A = 646, 

 The numbers calculated from the data are plotted on the curve by 

 the mark . (Plate XII., Fig. o and Fig. 4). In experiment 11 we 

 got / = 0016s and m = 0-48, the interval between the successive 

 ordinates being 0024 sec. and using these constants, we obtain 

 M = 0058, X = 300. Taking the value of 0031 in exp. 3 and of 

 0022 in Exp. 11 as the interval between the summit and the com- 

 mencement of the response, we may calculate the values of J/ 

 and of / in the same units as in our case. These constants are 

 given below: — 



Exp. 3. M = 2 25, 



/ = 646, 



Exp. 11. M = 2 66, 



/ = 300. 



In Plate XII. we sei; that the agreement ])etween the experi- 

 mental and the calculated curves is very good. From these results, 

 we may say that our exp(jnential law for the recovery of tho 

 temporary fatigue holds good in the results obtained by Eucas on 

 the gastrocnemius-sciatic preparation of the frog. 



Now denoting the inter\'al l)etween the stimuli by (, and the 

 normal latent period l)y /, the interval between the first stimulus 



