T. Mizuno on the Action of the Coherer. 459 



rattier smaller than expected ; and this is perhaps due to the 

 comparatively bad electric conductivities of the metals. Now 

 according to the above-assumed theory of welding, it follows 

 naturally that such a metal, possessing good electric con- 

 ductivity while at the same time its melting-point is not very 

 high, would suffer a marked reduction in its resistance. The 

 experimental results, so far as I have found out, seem to point 

 to such a conclusion. It was thought that it would be inter- 

 esting to ascertain whether the behaviour of coherers can be 

 modified freely by properly mixing one kind of metallic 

 powder with another, or not. To do this it is natural to select 

 metals whose behaviours are very marked, that is to say, those 

 metals which are either very irregular or very smooth in their 

 action under the influence of electric radiations. In this 

 respect iron belongs to the first, while silver, cadmium, 

 Rose's and Wood's metals belong to the second class. Hence, 

 as shown in the foregoing Table IV., the iron powder was 

 respectively coupled with the just-mentioned metals, and also 

 the couple of the two, so to speak, inactive metals, that is 

 silver and Wood's metal, was formed. 



Experimenting on such mixed coherers, the author found 

 that the irregularity in the behaviour of iron could be actually 

 smoothed at will by taking a greater mass of the coupled 

 metal. Again, here I have to draw attention to the fact that 

 with all the metals and alloys investigated the resistance did 

 not necessarily undergo diminution, but often increased during 

 the experiment; and also that in the case of some metals and 

 alloys the increase went up even to infinity. 



How are these phenomena to be accounted for ? Is it not 

 possible to conceive a state of affairs such that, in one case, 

 some of the conducting chains once formed by melting dis- 

 integrate, while in the other case the chains, acting somewhat 

 like fuse, suffer complete break under the impulse of electric 

 radiations ? 



Of course by "welding " I do not always necessarily mean 

 ■electrical welding in its strict sense ; but it might even, 

 perhaps, be sufficient to simply assume a certain process of 

 melting in order to account for the said phenomena. Though 

 the hidden mechanism of the coherer still remains for further 

 complete investigation, yet I venture to say that such a 

 survey as I have made of the successive stages in the history 

 of each of the several coherers appears, at any rate, to be of 

 no little scientific value. 



Lastly, it is to be noticed that the above experiments were 

 performed during the last winter at the Physical Laboratory 

 of Kyoto Imperial University, Kyoto, Japan. 



Berlin, July 1900. 



