496 LIVING TORPEDOS IN BERLIN. 



acted regularly with slow turning, but more and more feebly with 

 more rapid turning, until their effect was reversed. If it was 

 arranged that both shocks passed, the closing shocks always pre- 

 ponderated, all the more as the turning was more rapid. 



The only question now left was, whether the irregularity which 

 ensued on turning the disjunctor with greater rapidity, was peculiar 

 to this apparatus, or whether it would be found also with a more 

 rapid succession of closing and opening, than can be obtained by the 

 use of Wagner's hammer or Foucault's interrupter. I put into the 

 primary circuit of Buhmkorff' s inductorium, the interrupting spring 

 of an ordinary inductorium which opens the circuit 150-300 times 

 per second, according to a previous determination made by Helm- 

 holtz, and further a C tuning-fork interrupter by Konig of Paris, 

 which performs 256 single vibrations per second, or opens the 

 circuit 128 times. But I obtained regular actions with these 

 means also, as well as with Wagner's hammer and with Foucault's 

 interrupter, although the succession of the closing and opening 

 shocks was much more rapid than with the disjunctor, which gave 

 at most 45 opening shocks per second (see above, p. 493). 



Thus it was proved that in consequence of an unexplained condi- 

 tion, the disjunctor is unable to give series of exclusively closing or 

 of exclusively opening shocks, in which the same quantity of elec- 

 tricity equalizes itself, beyond a certain speed which is difficult to 

 determine. The next problem would be to search out that condition, 

 and to discover what becomes of the large amounts of electricity 

 set in motion at high tension in the opening shocks of the induc- 

 torium, and which seem to disappear and leave no trace. It is 

 essential to ascertain this in order to obtain a basis for the theory 

 and construction of so important an apparatus as the disjunctor ; 

 for this reason, I did not feel justified in withholding this series of 

 experiments although their result is not gratifying. 



It may perhaps be asked, why after failing with the disjunctor, 

 I did not attempt to compare the negative polarisation due to 

 a single closing and opening shock, which, as the above and other 

 observations show, can be obtained. This scheme naturally sug- 

 gests itself, and I intend to revert to it. It will, however, not be 

 at all easy to ensure the closure of the galvanometer circuit at the 

 same time after the polarising shock in the two cases. The end 

 may also be attained by a slowly turned disjunctor, but not with- 

 out further provision for regulating the number of the shocks 

 which reach the polarisation object. It will be as necessary, as in 



