J 155 



circuit and by varying either the primary current strength or more 

 conveniently tiie coil distance we finally get a minimal contraction. 

 Now leaving the coil distance and the primary current unchanged 

 we take away the added sel (induction L and put in its place a 

 rheostat from which so much resistance is unstopped lill the minimal 

 contraction appears again. Then we have: 



aI,M={R,, f W) ^,3L,,=R,,+^(L,, -f L) 

 or 



1 L 



J=w <'> 



giving the chronaxia .? as the qnotient of' a resistance into a 

 selfinduction. 



This method can still be applied in two different ways. We can 

 either use a fixed selfinduction and a variable resistance or a fixed 

 resistance and a variable selfinduction. The latter might even be 

 provided with a scale gi-aduated to lOOOO^'' parts of a second so 

 as to make it a direct i-eading instrnment. Or if we take 1000 

 olims foi' the fixed resistance, the readings of the gradnated scale of 

 the induction variometer in Henry will give the value of the chron- 

 axia in 1000''' parts of a second. 



The advantage of the induction coil method may be gathered from 

 the fact that the combination of a rheostat and gradnated selfinduc- 

 tion coil may be used with any medical coil and with any ordinary 

 interrupter. The only restriction to be made is against the use of a 

 secondary coil with less than 3000 windings. The formula used 

 ceases to be accurate as soon as an extremely large resistance is 

 inserted in the secondary circuit. With secondary coils of only a 

 few hundred secondary windings this is already the case with a 

 total secondary resistance of perhaps 1000 ohms. But with coils of 

 more than 3000 windings the secondary resistance may be increased 

 up to 15000 or 20000 Ohms without any appreciable error. The 

 error in these cases is caused by the presence of capacity in the 

 secondary coil, the influence of which is to lower the initial strength 

 of the secondary discharge and to render the calculated chronaxia 

 a trifle too small. But with coils of more than 3000 turns this error 

 is considerably smaller than the eri'or of the measurement itself, as 

 w^as proved by a special physiological and oscillographic investigation. 



Finally 1 wish to add that with the condenser or the induction 

 coil it is possible to arrange for the use of an nngraduated but 

 variable resistance and a special ohmmeter, which maj' be gradnated 

 for the chronaxia to be directly read on the scale. 



