pipe below the narrowing: 1-57 in. Rate of outflow: almost no lb. 

 per minute for each valve (in Plate XVIII the outflowing can be seen). 

 Power consumed by the coil : 3-6 watts when idle and 14-4 watts at the 

 moment when the outflow is stopped. In the cabin a rotary switch is 

 arranged in such a way that one can, at will, cause one or other of the 

 two valves to open. 



While the bathyscaphe is on the surface we do not want to use up 

 the electrical energy unnecessarily. A diver therefore goes down to 

 close the mouths of the valve mechanically, by means of a plug. 

 Removal of this plug, however, is achieved from the deck before 

 each dive, by releasing the retaining cable. Later an arrangement was 

 worked out which permitted the closing and opening of the tubing 

 with more ease and without the services of a diver. 



These valves function very well. Elementary prudence, however, 

 demands that we should envisage the possibility of an obstruction. 

 There are indeed different possible causes of a stoppage. If the pellets 

 rust, they can be transformed into a compact mass, which will prevent 

 all outflow. Very fortunately iron pellets practically do not rust under 

 water. Even after having remained for months in sea water, they are 

 perfectly fluid. If a little oxide is formed, it does not induce caking. 

 But it is quite otherwise if wet iron pellets are exposed to the air, be- 

 cause then they are capable of solidifying within a few hours. 



The danger of rust does not appear very serious, but as the 

 phenomena we observe are, in spite of everything, a little mysterious, 

 this danger must not be considered as non-existent. Another sort of 

 stoppage could occur, naturally, if by accident a foreign body, a bit of 

 sacking, for example, were introduced into the ballast tub. Finally, if 

 the bathyscaphe drove very deeply into the mud, the outlet mouth 

 could be obstructed. 



Although all these causes of stoppage may be rather unlikely, we 

 must find a way to save the bathyscaphe, even if the two valves should 

 become blocked. That is why we installed a device which allows us to 

 throw overboard both tubs together with their contents. Here is how 

 this operation could be done reliably. 



Each tub is suspended by a chain under the float in a seating of the 

 latter (see Plate VI). This chain passes up through a tube which 

 traverses the entire float and reaches the deck of the bathyscaphe, 

 where it is attached to a hook and lever system. The lever is held by an 

 electro-magnet, in such a way that the hook releases the chain as soon 



[ 155 ] 



