104 



Mr. W. B. Hardy. On the Mechanism of 



(3) the composition of the phases is dependent to a lesser degree 

 upon the ratio of agar to water in the entire mass, to a greater 

 degree upon the temperature. 



While recognising as fully as possible that only an approximation to 

 the actual composition of the two phases at different temperatures is 

 obtained by these experiments, it is obvious that they afford reliable 

 information on two points. These are, firstly, the marked "lagging" 

 action or passive resistance to change offered by the system agar- water. 

 The difference in composition of the phases according to whether any 

 given temperature lies in an ascending or a descending series shows 

 how slow the system is in reaching final equilibrium.* Secondly, the 

 experimental results seem to me to indicate pretty clearly the general 

 form of a part of the concentration temperature curve. I give the 

 curve as it appears from the figures in Experiment II. AB and CD 

 are the curves for the system — solution of agar in water, solution of 

 water in agar, and vapour. If they correctly represent the general form 

 of the curve, then, by the theorem of Le Chatelier, it follows that the 

 change from the system solution of water in agar and vapour to the system 

 solution of water in agar, solution of agar in water, and vapour will be 

 accompanied by a liberation of heat when the change takes place along 

 the isotherms from 5° to ±20°, and by an absorption of heat when the 

 change is along the isotherms ±20° to 35°, while the change from the 

 system solution of agar in water and vapour to the system of two 

 solutions and vapour will always be accompanied by absorption of 

 heat. 



I have not established this deduction experimentally, but it finds a 

 considerable amount of support in the following facts. When water is 

 allowed to dissolve in pure dry agar at 14°, a considerable amount of 

 heat is given off. 1 c.c. of dry agar in coarse powder added to 10 c.c. 

 of water gives a rise of more than 6°,t while control experiments with 

 carefully dried finest graphite or sand gave a rise of temperature of 

 0-15° and 0*17° respectively. Wiedemann and LudekingJ also found 



* Tlie systems salicylic acid and water, and thorium sulphate and water are 

 perhaps comparable cases. The former readily yields two fluid phases which, 

 however, are throughout in labile equilibrium (Bancroft, ' The Pliase Rule,' 

 p. 105). In the case of the latter system supersaturated solutions can be obtained 

 over a wide range of temperatures, and even in presence of the stable hydrates it 

 is often hours or days before equilibrium is reached (Bancroft, loc. cH., p. 54, or 

 Soozebrom, ' Zeits. f. Phys. Chem vol. 5, 1890, p. 198). The lagging action in 

 colloids which is so markedly shown by van Bemmelen's researches into the 

 effect of time on the hydrogel of silicic acid, ceases to be extraordinary when one 

 remembers that one of the phases is a solid. Gels reach equilibrium much more 

 rapidly than does, for instance, a bar of metal in which the reaction velocity is so 

 slow that final equilibrium may never bo reached. 



t The mercury in the Beckmann thermometer was driven beyond the scale into 

 the upper reservoir. 



X ' Ann. der Phys. u. Chem.,' N.F., vol. 25, 1885, p. 145. 



