bathyscaphe if one of the compartments is entirely flooded by water. 

 These considerations limit the maximum volume of the big compart- 

 ment to about 460-486 cu. ft. 



6. In the tropics the sea can have a surface temperature of 86° F. : 

 at the bottom it may have cooled down to the vicinity of 32° F., on 

 account of currents coming from the arctic regions. If the bathyscaphe 

 descends rapidly to 2^ miles, the petrol, which at the beginning had a 

 temperature of 86° F., will be heated by adiabatic compression by 

 another 1 8° F. : its temperature will then be in the neighbo*urhood of 

 104° F. The hull, on the contrary, will take on, approximately, the 

 temperature of the water outside, because the heat-transfer coefficient 

 between moving water and iron is much greater than the heat-transfer 

 coefficient between motionless hydrocarbon and iron. The partitions, 

 on the contrary, surrounded in all parts by petrol, will take on its 

 temperature. We must then envisage the possibility of a difference in 

 temperature between hull and partitions of 72°. If hull and partitions 

 are rigid, because of this considerable stress would result. 



If the hull itself cannot be deformed, the specific pressure in the 

 partitions may be calculated by the formula: 



G=KATE 



where K equals i-i x io~^IC.°, AT equals 40°C. and ^equals 31-3 x 

 lo^lb. per sq. in. which gives 



o-= 13,800 lb. per sq. in. 



Although decreased by the elasticity of its hull, this stress would 

 not be negligible, because it is added to other stresses whose absolute 

 value we do not know, particularly constructional stresses. 



If the metal sheets were flat, buckling could occur accompanied by 

 a sound very much resembling a breakage. We can imagine the effect 

 that sound would have on an observer. He is bending down to his 

 porthole and is admiring the submarine fauna. Suddenly he hears a 

 crack. He jumps. The pilot only half reassures him when he says that 

 that sound is heard every time a rapid descent is made. (According to 

 newspaper accounts and remarks of one of the passengers of the 

 FNRS J, this cracking really occurs during the dives of this sub- 

 marine. I attribute it to the fact that, contrary to my recommendations, 

 flat partitions were used in it.) 



If the partitions were rendered completely rigid by means of 



[ 162] 



