504 



Frederick Guthrie on some Thermal and 



Table LX. 

 Separation of Alcohol and Bisulphide of Carbon by Cold. 



Per cent, of 



Per cent, of 



Temperature at which 



bisulphide. 



alcohol. 



separation begins. 



94-94 



5-06 



Remains clear down to —18° "4. 



89-54 



10-46 



Becomes turbid at — 14° *4. 



84-89 



15-11 



-15°-9. 



79-96 



20-04 



-16°-1. 



6511 



34-89 



-17°-7. 



59-58 



40-42 1 





4991 

 39-96 



50-09 1 

 60-04 f 



Remains clear down to —20°. 



29-92 



70-08 J 





In these experiments, as the thermometer sinks it is noticed 

 that at several degrees above that at which separation begins 

 the liquid becomes opalescent, or rather fluorescent ; for in 

 this state it is still perfectly transparent, and colours the 

 transmitted light of a light orange -yellow. It is pale blue by 

 the fluoresced light. On reaching the critical temperature of 

 true turbidity, the change is abrupt. At only the fraction of 

 a degree lower the liquid divides into two with surprising sud- 

 denness and precision. In the case of the 79*96-per-cent. 

 mixture the volumes of the two separated liquids are about 

 equal at —17°, although at —15° only the above-mentioned 

 fluorescence betrays the impending change. The suddenness 

 of the change argues that the quantity of heat lost during the 

 change is inconsiderable. Again, on removing from the 

 freezing-mixture and shaking up, I have not been able to 

 detect anything like a stationary thermometer during remix- 

 ture, although all analogy shows that there must be a tendency 

 that way. Indeed the phenomenon closely resembles that 

 described in §§ 238-241a attending the decomposition of an 

 aqueous solution of triethylamine. But it is here presented 

 with a falling thermometer, there with a rising one. Here it 

 resembles the separation of a solid salt containing water of 

 crystallization from an aqueous solution which still retains 

 some : there it was due to the decomposition by heat of a 

 subcryohydrate. Here the original mixing absorbed heat, as 

 when a salt dissolves in water : there it liberated heat. Here 

 such mixing gained volume: there it lost volume. The one 

 case is, as it were, a real image of the other, and inverted, 

 because real. 



Only such a little bit of this behaviour is revealed by an 

 ice-salt cryogen that it would be almost useless to draw its 

 curve from these four points. Laid out in percentages of 



