202 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1951 



is indicated in figure 2. Alternatively, the pressure vessel may be 

 forced into its sleeve by an independent hydraulic press coupled in 

 the proper ratio to the press which produces the internal pressure. 

 The choice of method depends on the size of the apparatus, and is 

 determined by various factors such as friction, which we cannot dis- 

 cuss here. The limit to the pressures that have been used in the 

 laboratory in apparatus constructed with conical external support 

 has been about 50,000 atmospheres. At this pressure a new limiting 

 effect appears: this is the incipient extrusion of the pressure vessel 

 through the supporting sleeve by the thrust, or some other sort of 

 fracture of the vessel due to the thrust. 



To reach yet higher pressures, the vessel must receive still more 

 effective support. This may be given by immersing the pressure 

 vessel in a liquid which is itself exposed to high hydrostatic pressure. 

 In this way the pressure vessel receives complete support over its en- 

 tire external surface. This proves effective enough to extend the 

 range from 50,000 to 100,000 atmospheres. The external supporting 

 pressure necessary for this extension may be between 25,000 and 

 30,000 atmospheres. 



So large an extension of range, from 50,000 to 100,000 atmospheres, 

 by a supporting pressure of only 25,000 atmospheres, would not be 

 possible if it were not for the change in the properties of steel and 

 carboloy produced by a pressure of 25,000 atmospheres. Under this 

 pressure, steel becomes much more ductile and also stronger. Car- 

 boloy becomes stronger in compression and loses its brittleness, so 

 that carboloy pistons can be subjected to pressures of 100,000 atmos- 

 pheres or even more. Carboloy also increases in tensile strength, so 

 that the pressure vessel itself may be constructed of this material, 

 supported on the outside by a shruiik-on steel jacket. The apparatus 

 is indicated schematically in figure 3 ; it is in fact an arrangement of 

 one pressure vessel inside another. Theoretically, any extension what- 

 ever of the pressure range would be possible by this method, by 

 making a nest of pressure vessels, each one containing a successively 

 higher pressure supporting the vessels within. Although theoreti- 

 cally possible, no feasible method of constructing more than the first 

 apparatus in such a series has been found. 



It may appear paradoxical that at these very high pressures the 

 problem which at first restricted experiments, namely the problem 

 of preventing leak of the pressure-transmitting medium, has entirely 

 disappeared. The reason for this is that at these pressures fluids no 

 longer exist ; all normal liquids or gases are frozen solid by the pres- 

 sure. All the elements ordinarily gaseous, except helium, have been 

 solidified at 50,000 atmopsheres, and there is good reason to think that 

 at 100,000 atmospheres helium, too, freezes solid. In this region, pres- 



