Researclies on the Discharge of the Electric Orgau. Q3 



The above table sliows that the area of the discharge curve be- 

 comes constant when the stimulus is made strong enough. While 

 the experimental result gives support for the theory of "all or 

 none,'' it seems to reveal two new facts : (1) the opening-stimulus 

 can evoke its response, which could not be caused on closing the 

 same stimulating current, or, otherwise, the opening-stimulus may 

 be summed up into something that occurred at the anode on 

 closing tlie current : ('2) the delay of the second discharge is longer 

 than the normal. The cause of the second phenomenon is not 

 clear. It may be seen that the modal latent period measured on 

 the oscillogram seems rather to have normal value i.e. not prolonged. 

 But since the second summit lies upon the inclined branch of the 

 first discharge curve, the displacement of the apparent summit 

 towards the stimulus is not small. On calculating the displacement 



bv the formula <J= t .... . it mav l)e found that the disiilacement 

 is not smaller than 2 nun. on the film. Therefore the prolongation 

 of the modal latent period of the discharge by opening-stimulus 

 seems certainly to exist. 



Before closing this section we ha\-e tw(» more oscillograms of 

 allied problem to be explained. Since I thought that the rhythmic 

 response of a ner\'e or of a muscle to high frequency stimuli, 

 might be the effect of the refractory period, a few experi- 

 ments were made regarding that subject. (!)scillograms No. ()7 and 

 No. 85 (Plate XXIII.) represent the results obtained. In No. 67 

 the make and break of the high frequency of the primary circuit of 

 an induction coil was made by a contrivance like the commutator 



the plates dischargiag per second at an instant in each column be ?ii, ?*_•, «■,, n,,» respectively. 



Then by Kirchhofes law, we have 



cnie = Rl+rii, 



c)u>e = RI+ri.2, 



£nie = IiI+rto_, 

 cn„ie = El + ri„„ 



where l = ii + ù+i:+ + im, and c is a proportional constant. 



By adding each side of the equations, we have 



ce'^n,u = {Ji + r) I, 

 1 



and therefore / iI/(„/7f = — ^ / LU. 



i. e. the total number of the plates dischar-^ed is proportion il to the area of the curve. 



