494 LIVING TORPEDOS IN BERLIN. 



vided with clay shields, to the object to be polarised. Blotting-- 

 paper pads, white beech wood, pumice-stone, roofing- tiles were made 

 into bars 50 mm. long and one square centimeter in transverse 

 section, and served for such objects. The blotting-paper was soaked 

 with distilled water, and the wood, pumice-stone and tiles were 

 boiled in it until they sank below the surface 1 . In the same 

 circuit, secondary in regard to the induction, primary to the polari- 

 sation, was placed the galvanometer P with 53 turns of wire at a 

 distance of 20 mm. from the mirror. 



A second pair of conducting vessels with wedge-shaped pads, the 

 edges of which, protected with physiological clay, were laid on the 

 object to be polarised, formed as usual the terminals of the gal- 

 vanometer S. 



At first it seemed that the negatively polarising effect of the 

 closing shocks exceeded that of the opening shocks. It was 

 however soon proved that this could not be relied on. If 

 the experiments went right, the deflection of the galvanometer 

 P, due to the closing shocks, ought to have been sensibly equal 

 to that due to the opening shocks. But it was observed that 

 the closing shocks always produced by far the greater deflec- 

 tions. As at first I employed ordinary inductoria, in which 

 no special measures are adopted for the insulation of the turns of 

 the wire from each other in the secondary coils, I supposed that a 

 channel for sparks had formed itself in the interior of the coils, in 

 which the high tension electricity due to the opening induction, 

 sprang over, instead of making its way through the object to be 

 polarised and the galvanometer P. In order to work under 

 simpler conditions, I abandoned the polarisation object and the 

 conducting vessels in connexion with it, and replaced them by one 

 or two resistance coils out of a plug rheostat, which gave a resist- 

 ance of 60 and 1 20 Siemens units, respectively. Instead of the 

 ordinary induction apparatus, I took a Ruhmkorff inductorium by 

 Siemens and Halske, and as its shocks, even with only two Bunsen 

 chromic acid cells in the primary circuit, seemed likely to damage 

 the galvanometer, I drew the primary coil somewhat out of the 

 secondary, so that I gradually lessened the effect as in the ordinary 

 inductorium. In this apparatus, it is certain that no sparks sprang 

 over in the interior of the secondary coil ; but, on the other hand, it 

 was very difficult to put a stop to all sparks and lateral discharges 

 owing to the length of the conducting wire, which was necessary 



1 Cornp. Untersuchungen, vol. ii. part ii. p. 430 f. 



