

Dissociation of Compounds in Water. 469 



been concentrated, only 1*5 per cent, remains so nearly con* 

 stant that the error would be entirely inappreciable. 



Further, this correction does not take into account the ice 

 which separated during the time required for the thermometer 

 to become constant. But on account of the form of the 

 apparatus and the large volume of solution employed, the 

 cooling proceeded so slowly that the amount of ice which 

 separated during this time was comparatively small, and no 

 serious error could have been introduced from this source. 



A great advantage in the regulation of the temperature 

 was derived from the use of such a large volume of the solu- 

 tions. The surface of the solution exposed to the cold layer 

 of air was relatively very much diminished, and the solution 

 itself would therefore be affected much less by external 

 changes in temperature. The cooling of the solution pro- 

 ceeded very slowly ; therefore the accuracy of the measurement 

 at any given time would be much increased. When small 

 volumes of solutions are employed, the continued separation 

 of the ice is comparatively rapid, and the solutions are thus 

 ever concentrated by a large and unknown amount. This 

 necessarily affects the accuracy of such determinations very 

 considerably. 



A number of duplicate determinations of the freezing- 

 point of the same solution were made with this enlarged 

 apparatus and improved method, and in no case did the 

 individual determinations differ from each other by more 

 than 0*0004°. For the most dilute solutions duplicate 

 measurements were made, and the figure given is the mean 

 result. 



Whenever an acid was used, a drop of a dilute solution of 

 the same acid was added to the litre of water each morning 

 before the freezing-point of the water was determined. This 

 was for the purpose of neutralizing any trace of ammonia 

 present, and to convert all traces of salts in the distilled water 

 into salts of the acid under investigation, so that no part of the 

 acid to be measured might be neutralized by such impurities. 

 That the water used contained only a mere trace of any 

 alkaline substance was shown by the fact that one drop of dilute 

 acid in a litre of water produced a perceptible acid reaction. 

 The zero of the water containing the trace of salts and free 

 acid was then determined. The very minute quantity of 

 these substances present could not have appreciably affected 

 the dissociation of the acid to be investigated. 



When the alkalies were used, a drop of the same alkali 

 was added to the water before the freezing-point was deter- 

 mined, that any carbon dioxide present might be neutralized. 



