EXPERIMENTAL KNOWLEDGE OF THE PEOPEETIES OF MATTER. 467 



For Sn2Cl4 the density would be 13-OG, and for SnClg. 6-53.' The 

 molecular formula for stannous chloride is inferred to be SnCl2. 



The molecular formula CU2CI2 remains unaltered at the highest tem- 

 peratures used by Biltz and V. Meyer. 



Vapour-densities of Ferric Chloride. 



Next appears a paper by Griinewald and V. Meyer,^ in which these 

 authors attempt to find the vapour-densities of ferric chloride at high 

 temperatures with the view of deciding whether these indicate a molecule 

 FeoClg at any temperature-range, and whether there is a molecule FeCls- 

 For this purpose they make a series of determinations at the tempera- 

 tures of boiling sulphur, 448°, boiling phosphoric sulphide, 518°, and boil- 

 ing stannous chloride, 606°. At the temperature of boiling sulphur the 

 ferric chloride was volatilised slowly and completely, and its vapour- 

 density, from the mean of four determinations — 10-67o, 10559, 10-227, 

 10-487— was 10-487 at 448° ; now the vapour-density for Fe2ClG would 

 be 11-2, which is greater than this mean result and than either of the 

 individual numbers. The authors say that determinations at even 

 slightly lower temperatures are inadmissible, for the temperature of 

 volatilisation must be very close to 448° (see Friedel and Crafts further 

 on). At 518° they found 9-569 for the vapour-density as the mean of 

 three determinations ; but they found that even at this temperature about 

 one-tenth part of the ferric chloride had been converted into ferrous 

 chloride and chlorine, the ferrous chloride being much less volatile than 

 the fei'ric ; the vapour-density would be that of the ferric chloride remain- 

 ing as vapour together with the chlorine. 



At 606°, that is, by heating in the vapour of stannous chloride, the 

 six experiments at this temperature gave a mean result 8-383 ; about one- 

 eighth part of the ferric chloride was decomposed. 



At about 750° in a Perrot's furnace the mean of three experiments 

 gave 5-436 ; but here about a third of the ferric chloride was decomposed 

 into ferrous chloride and chlorine. 



At 1077° experiment gave 5-307 ; while at 103G° it gave 4-915, the 

 larger number for the higher temperature being due to volatilisation of 

 some of the ferrous chloride. 



At about 1300° the numbers found by two vapour-density determina- 

 tions were 5-155 and 5-115. 



It is seen that no temperature at which the vapour-density was deter- 

 mined in the above experiments, in which ferric chloride was volatilised. 

 in an atmosphere of nitrogen, gave the vapour-density 56, corresponding 

 to the formula FeClg ; and in such experiments it is hopeless to expect it, 

 as the ferric chloride at high temperatures is decomposed into chlorine 

 and ferrous chloride ; but, as the only conclusion which is consistent with 

 their experiments, the authors infer that there is no molecule FeoClg, and 

 that FeClg is the only formula for ferric chloride. 



Experiments in an atmosphere of chlorine do not modify this 

 conclusion. 



It is noticeable, however, that at about 448° the vapour-density is in 

 some determinations over 10-6, while the formula FeoClg requires 11-2, 

 which is but little larger. The possibility of the coexistence of FeaClg 



• Beil. 1888, p. 412 ; and Xachr. d. Itnl. Ges. d. Wiss. GGttingen, 2, 1888, p. 19. 

 == i?«-. 21,1888, p. C87. 



H H 2 



