86 



NA TURE 



[November i8,, 1909 



tend to be -felted ; together.. The temperature condition 

 giving birth to the most favourable number of spontaneous 

 centres is not necessarily the. temperature at which crystals 

 grow to the largest size, so there is really no escape from 

 finding by direct trial the most effective way to go to 

 work. 



Another possibility is brought to light by an examination 

 of a case of pseudo-equilibrium, such as that of phenol. 

 Here we have three regions — in one No. i alone is stable, 

 in another No. 2, and in the third both Nos. 1 and 2 

 are stable. The case of iodide of silver is similar but 

 more complicated. If in the area C we change the 

 pressure, the temperature remaining constant and the 



30' 



20 



Phenol 



lO" 



-10 



-20 



800 



1200 



1600 



2000 



Kg. per Sq. Ci 



material consisting of a mi.xture of the two stable phases, 

 we can alter the proportions in which these phases exist, 

 but we cannot cause either of them to disappear.' 



A notable case of this kind is that of graphite and 

 diamond, both perfectly stable in presence of each other 

 at atmospheric pressure up to a temperature nearly that 

 of the electric arc, say about 3000° C. If there be any 

 similarity between the carbon and phenol diagrams, 

 diamond would correspond to variety No. 2 of phenol and 

 graphite to variety No. i, heat being evolved in both 

 cases when the less dense modification changes into the 

 denser. If we desire to obtain phenol 2 from phenol i, 

 NO. 2090, VOL. 82] 



we note that, down to a temperature of —20° C, we 

 should require to keep the pressure always above about 

 boo kg. /cm.-, otherwise the operations would be similar 

 to those described in the case of rhombic sulphur. '. 



Similarly, to convert graphite to diamond on this 

 analogy we should have to raise the temperature and 

 pressure together to some unknown values, and then let 

 the product cool, keeping up the pressure meanwhile. 



The apparatus which I have used in making the experi- 

 ment is based on the transmission of pressure by crystal- 

 line graphite or the softer metals. In order to ascertain 

 how much pressure is lost during transmission, I have 

 arranged an apparatus in which the material to be tested 

 is exposed to a known pressure, tending to force it 

 through a cylindrical space, identical in figure with the 

 space in which the heating is intended to be carried out. 

 The pressure transmitted is transferred by a simple device 

 to a piston with a hard steel point, and this is forced 

 by the pressure to penetrate a soft steel plate. In a sub- 

 sequent experiment the same piston is forced by a known 

 pressure into the same steel plate so as to penetrate to the 

 same depth as in the main experiment. It is then possible 

 to compare the pressure transmitted with the pressure 

 applied. 



Experiments of this kind have been made with lead and 

 with graphite as pressure-transmitting substances. 



So far as I know, there is no substance other than 



graphite combining the property of a certain amount of 

 Huidity with the capacity to resist high temperatures, and 

 our hope of studying chemistry at really high pressures 

 and temperatures appears at present to depend largely 

 upon it. It is true that some attempts have been made 

 to use compressed gases, but the apparatus is vastly more . 

 complicated, and the experiments themselves become reallv 

 dangerous in view of the immense potential energy 

 possessed by ga^s at pressures of 100 tons per square 

 inch. As illustrating this, I may mention that 100 tons 

 per square inch is about the highest instantaneous pressure 

 noted by Sir .\ndrew Noble in his well-known experiments 

 on the exploding of cordite in closed vessels. The density 

 of nitrogen at 100 tons per square inch is, taking Boyle's 

 law as a very rough appro.ximation, 15,240 times its 

 density under standard conditions. This works out to 

 rather more than iq, i.e. about the same as gold, and the 

 energy stored is of the same order as that contained in 

 an equal volume of cordite, though its availability is 

 lower. 



The construction of the apparatus I have used can be 

 easily followed from the drawings. It consists essentially 

 of a steel cylinder divided perpendicular to the longitudinal 

 axis by a thin plate of mica, the two halves being clamped 

 tightly together by an insulated ring and clamps at top 

 and bottom. Pressure can be applied by an ordinary 



