Eesearches on the Discharge of the Electric Organ. 35 



category of a probability curve. Since I cannot ascertain the 

 i^roper method of anal3^sing the curve from tlie standpoint of 

 this hypothesis, I am not able to give the rigorous proof for it. But 

 it is obvious that, qualitatively, the course of the ex})erimental curve 

 agrees with this view.* I am nov/ working on this principle; 

 I only refer to it here, and sliall leave the problem for a later report. 

 Next we shall treat the relation between the Ijreadth of the 

 stimulus and the height of the discharge. With reference to the 

 problem, the fdm first examined was that of No. 37, an anah^sis of 

 whicli was made in the preceding section. Tlie organ-prep- 

 aration was the same as that of No. 35 and No. 3G, in which we 

 knew that the effect of the progressive change was very small. 

 Assuming Hoorweg's decrement factor to be true and not taking 

 into consideration the efïect of the opening-stimulus separately, 

 wliich Hoorweg, in his condenser investigation, also ignored, we 

 examined whether his factor held good in our case. According 

 to his formula the excitation E caused by our stimulus is 

 equal to 



./ 



t 



ai 







where i represents the maximum current in the secondary of the 

 induction coil, A is a constant determined by the self-induction 

 and the resistance of the secondary circuit, and wliere e~^' means 

 Hoorweg's decrement factor. Of course, in our stimulus there is a 

 steep rising portion of the current before its maximum, which is 

 not considered here. But in treating of tlie variation only, the 

 part common to all stimuli has no influence on the result, if we 

 take the origin of the time at a proper instant. Hence measuring 

 tlie time from a suitable origin the formula becomes 



which, for tlie sake of brevity, we write 



* A trial examination was made on the curves of Oscillogram Xo. 5t. The curve represent- 

 ing the relation between the height of the stimulus and the area of the resulting discharge 

 curve was graphically diä'erentiated by driiwing the tangents to the curve, point by point, 

 and by finding the corresponding rates of increase of the ordinates with respect to the 

 abscissa. Then it was found that the differential curve might be expressed by a formula 

 analogous to that of the discharge curve given in the previous section. 



