EXPERIMENTAL KNOWLEDGE OF THE PROPERTIES OF MATTER. 135 



density of iodine at the very high temperature — yet much below 2000° — 

 in this experiment. 



Troost had also recognised the influence of reduction of pressure on 

 the vapour- density of iodine, and had in fact obtained for a constant 

 temperature, 440°, a series of vapour-densities of iodine (relative to air at 

 same temperature and pressures), as under : — 



Pressures . 768 mm. 67-2 mm. 48'6 mm. 48"57 mm. 34"52 mm. 

 Densities . 870 „ 8-20 „ 7-75 „ 7-76 „ 7-35 „ 



The conclusion which was drawn by Crafts and Meier from their ex- 

 periments was that the molecule I2 had been gradually decomposed into 

 molecules I. Troost considered that such a dissociation could not be 

 effected by a simple diminution of pressure ; but there are cases of chemi- 

 cal compounds, which can be formed at a low temperature and by suflBi- 

 cient pressure, which can be decomposed, partially or wholly, either by 

 raising the temperature or hy diminishing the pressure, or by both con- 

 bined, and which can be re-formed by lowering the temperature or in- 

 creasing the pressure. 



A remarkable instance of this is phosphonium chloride, formed by 

 Ogier 1 by combining PH3 with HCl. These two gases do not combine 

 at ordinary temperatures, but were by Ogier brought into combination by 

 a pressure of about 20 atmos at 14'^, also by lowering the temperature of 

 the mixed gases to —30°. In the former method he fills over mercury 

 the ordinary tube of ' Cailletet's elegant apparatus ' with a mixture of 

 equal volumes of PH3 and HCl, and when a sufficient pressure has been 

 applied brilliant crystals of PH4CI appear ; on warming the upper part of 

 the tube, at about 20°, a liquid layer forms which is either liquid phospho- 

 nium chloride or a mixture of the liquefied gases. Ogier says that on 

 gradually cooling the mixed gases the deposit of crystals takes place 

 almost suddenly at —30°; but he thinks it possible that the gases may be 

 in combination at a somewhat higher temperature. That is a matter 

 which experiment has not decided. 



Now Van't Hoff,^ by compressing a mixture of equal volumes of the 

 mixture of PH3 and HCl, got the laboratory-tube of a Cailletet's apparatus 

 half-full of the white phosphoniura-chloride crystals. These he heated 

 with a water-bath ; the crystals melted at 25° ; and on heating further at 

 pressure of between 80 and 90 atmos till the temperature was between 

 S0° and 61°, the boundary between liquid and vapour disappeared ; on 

 again lowering the temperature there was noticed the hazy appearance 

 which is characteristic of the critical point. 



The liquid state (?) of the phosphonium chloride obtained by warming 

 the crystals, or the crystals themselves obtained by pressure at 14°, on 

 gradually diminishing the pressure, gradually disappear, being converted 

 without changing the temperature of 14° into PH3 and HCl ; here is an 

 exact analogy with the case of iodine in Troost's experiments with dimi- 

 nishing pressure, for we may suppose the molecules I2 to gradually de- 

 compose, on the pressure being relieved, into molecules I. 



The theory of Crafts and Meier, accepted by V. Meyer, that molecule 

 I2 is split into two molecules I, though not overthrown by the experi- 

 ments of Troost, is barely proved by Crafts' and Meier's experiments ; for 



' Annalcs de Chiniie ct de Physique, 1880 (-5), xs. p. 63. 

 ^ BericJde der Bev.tschen Chemischeii G. xviii. 2088. 



