and of the mercury (of which the coefficients are known) the volume 

 of the mercury gives us the total contraction of the petrol; the problem 

 being to determine it for an extreme case which might arise. Unfor- 

 tunately the compressibility of petrol is greater than that of water. 

 From this fact the lifting capacity of a given quantity of petrol neces- 

 sarily diminishes proportionately as the bathyscaphe goes down: 

 its vertical equilibrium is thus unstable. Hence the farther down the 

 bathyscaphe goes, the heavier it becomes and the more it has a tendency 

 to descend. Therefore it is necessary for the pilot to be able to lighten 

 the bathyscaphe by throwing out ballast. 



The bathyscaphe can, however, in rare cases, recover its equilibrium, 

 if in the course of its descent it encounters a layer of water that is 

 much colder, and so denser. But this state of rest can only be 

 temporary, because at the end of a few hours the temperature of the 

 petrol will be equal to that of the surrounding water. 



We observed this phenomenon of rest following upon a cooling 

 of the water, on the 14th August 1953, when my son and I were making 

 a trial descent in depths of 22 fathoms. At about 14 fathoms from the 

 surface the Trieste reached equilibrium in a layer of cold water, where 

 it came to a standstill (see pages 109-10). 



THE BALLAST 



The problem of ballast is vital for the bathyscaphe and from this 

 point of view there is a fundamental difference between it and the free 

 balloon. If the aeronaut has no more ballast to throw overboard he 

 cannot rise any farther and he will perhaps have to land more quickly 

 than he would have wanted to, and perhaps also make a rougher 

 landing. But it is not a catastrophe. On the other hand, in the bathy- 

 scaphe, a breakdown in the unballasting apparatus would prevent it 

 from ever returning to the surface. A method must be found thus, 

 which in any circumstances will allow the pilot to unballast, and which 

 will never be in danger of a breakdown. The ballast is outside the 

 cabin : the pilot must then in some manner work through the wall of 

 the cabin. How is it possible.'^ One could imagine a mechanical 

 system : a push rod or a shaft would pass through the wall and would 

 start the unballasting: that would necessitate a stuffing-box. Now a 

 stuffing-box, watertight at a pressure of some thousands of pounds per 

 square inch and allowing free movement of the shaft, would be a 

 very delicate and dangerous organ. If it is held too tightly, in fact, the 



[32] 



