There is always the danger of the current being cut ofF, by either a 

 short-circuit or a poor contact, but in our case this would not have 

 catastrophic consequences. The ballast would fall out and the bathy- 

 scaphe would reach the surface sooner and more rapidly than its 

 occupants would like, but that is all. 



To economize the ampere-turns necessary to produce the magnetiza- 

 tion, thus to economize the energy in our accumulators, we arranged 

 along the lines of magnetic force of the mild steel a, a, a, in the form of 

 a central tube, two annular plates and an external cylinder. At the point 

 where the central tube narrows, the iron is replaced by a non-magnetic 

 substance c, plexiglas or stainless steel. Thus the lines of force, not 

 finding any more solid iron on their route, are obliged to go through 

 the iron pellets, this having a much greater magnetic permeability 

 than the non-magnetic substance, even though less than that of solid 

 iron. 



The construction of the coils d sets a delicate problem, by the fact 

 that they will be in sea water at high pressure and that all danger of a 

 short-circuit must be avoided. We gave up the idea of protecting the 

 coils by strong steel containers, thus preserving them from external 

 pressure. To give an adequate protection, this construction would have 

 been too expensive, too heavy and too complicated. The whole coil 

 could be placed in an insulating liquid (as we did for the electric motors 

 and for the tachometer, pages 37 and 99). However, we preferred a 

 still simpler construction. The winding is composed of copper wire 

 insulated in polythene. It was naturally necessary that this insulation 

 should not present the slightest defect through the whole extent of the 

 wire. Thanks to the assistance of the Marelli Company of Milan, this 

 condition was completely fulfilled. Thus the sea water could penetrate 

 between the wires of the coil without damage. It is even necessary 

 that it should be able to enter everywhere, because if air remained 

 caught between the turns without the water being able to get to it, the 

 pressure of the water would produce deformations in the winding 

 which, in their turn, would jeopardize the continuity of the insulat- 

 ion. To avoid all danger, while the wire was being wound, we inter- 

 posed little sheets of celluloid between the layers and upon the side 

 of the coil. Along these pieces of celluloid there remains a little free 

 space through which the water has access to the whole coil. 



The whole assembly at no time gave us trouble. It is true that the 

 insulation, to be perfectly safe, must be rather thick, which decreases 



[153] 



