thinner; the decrease in weight which would result would permit a 

 decrease in all the dimensions of the float. 



Directly or indirectly all the compartments of the float must be 

 in communication with the sea. Thus the water will partially take 

 the place of the petrol when this diminishes in volume, either as a 

 result of the increased pressure due to the descent, or following upon 

 a decrease of temperature. The water will be driven out when the 

 petrol expands. The pressure will then always be about the same inside 

 and outside the float, with the consequence that the walls will not be 

 acted upon by the enormous pressures of great depths. Fig. 8 shows 

 the arrangement used. 



cp—ic 



ë 





212 

 eu fi 



B B C B C B 



jr~i-l 



iLjy. 



-L" 



r^ 



Uit 



B C B B B 



,*J 



2Sl 



f^ 



Fig. 8. Compartments in the float of the Trieste 



Let us imagine the bathyscaphe floating on the water. The float 

 is still full of air. The caps of the openings C are unscrewed and the 

 drain-cocks B are opened. In order to avoid a loss due to expansion, 

 the float is not completely filled with petrol. First, therefore, a certain 

 quantity of water is poured into reservoirs 6, 7 and 9. This quantity 

 of water has been calculated in advance, in terms of the density of the 

 petrol and its temperature, as well as of the total weight which the 

 bathyscaphe is to support. (At the time of the first filling of the Trieste^ 

 it was a matter of 510 cu. ft. of water.) For the rest, the float has been 

 calculated rather generously, for, as we have seen, at the moment when 

 the plans were drawn up, neither the weight of the metal plates nor 

 the specific gravity of the petrol were known with precision. 



Through the openings C we introduce nearly all the petrol required 

 to fill the float completely. All the filling-holes are closed except the 



[76] 



