ELECTRICAL BREAK-EXCITATION. 123 



is easy to observe, for both ascending and descending currents, that 

 within the current-strengths used by us (^ Daniell to about 

 7 Daniells) imperfect break-contractions appear the earlier the 

 stronger the currents are. Thus with a strong current excitation 

 is caused by a fall of a smaller fraction of its height than is 

 necessary with a weaker one. (I shall not give detailed numbers 

 here, since these facts are only mentioned for the sake of complete- 

 ness. They can be as well explained by Hermann's theory as by 

 ours). For if a current of strength 8 is allowed to fall a half, it 

 produces the polarisation-current of which the initial strength is 

 half /S; whereas a current of half the strength similarly allowed to 

 fall a half gives only a current of a quarter S. The first current 

 has not therefore to fall a half in order to produce the same 

 effect as is produced by the second when suddenly diminished in 

 the same proportion. This fact is however evidently also in 

 accordance with Hermann's theory ; for in the one case there is 



TT TT 



a fall of current-strength of , in the second of . For produc- 



2 4 



ing weak electromotive forces we used a Noe's Thermopyle, the 

 flame of which was placed at different distances from its middle. 



As the reader will now see, the results of all these experiments 

 on the imperfect break-contraction are not only intelligible on our 

 theory, but present almost insuperable difficulties to the ordinary 

 explanation. For the present I can see no single objection which 

 can be brought against my own and Tigerstedt's conception of the 

 nature of the break-excitation. Of course I shall not persist in my 

 opinion if any one can prove to me that the mere disappearance of 

 anelectrotonus produces excitation. In any case no one can deny 

 that I have referred about 80 per cent, of all so-called break- 

 excitations to make-excitations, and that the polarisation after- 

 current has great significance as regards the causation of the 

 remaining 30 per cent., as even Hermann acknowledges. But it 

 is more than probable, and to me it seems certain, that the polari- 

 sation-current is the sole cause. I must, however, add that it 

 would have been desirable if I could have directly proved this by 

 experiment for the incomplete break-excitation. The experiment 

 would require to be arranged so that when the fall of current- 

 strength that first produced an incomplete break-excitation was 

 reached, the current passing through the derivation-circuit 

 DAD"e'e should be immediately led to a galvanometer. This 

 current must then be in the opposite direction to that of the 



