Section 79. Errors in the Solutions and their Correction. 205 



nium hydroxide (or free acetic acid) in the solution investigated. It was, 

 however, found that, with solutions containing an excess of acid, this 

 excess underwent no considerable change during the heating. That the 

 method is accurate within 0.1 per cent was shown by an analysis of a 

 known ammonium acetate solution containing a known excess of ammo- 

 nium hydroxide. 



This procedure of boiling the solution to remove all air was not used 

 with pure ammonium hydroxide, acetic acid or sodium- acetate, because 

 there was no especial advantage in it, since it was in any case necessary to 

 titrate the solution after the heating. The procedure was used, however, 

 in the experiments with ammonium chloride ; the correction on the concen- 

 tration for the vaporization of water was found to be the same as in the 

 ammonium acetate solutions, viz., about 0.3 per cent. In the experiments 

 with ammonium chloride at 306 the concentration of the excess of ammo- 

 nium hydroxide after the experiment was determined by titration against 

 hydrochloric acid. 



The measured conductance of the solution includes that of the water 

 and of the small amount of impurities left in the water. The initial 18 

 values were corrected by subtracting the conductance, measured in a small 

 Arrhenius cell, of the particular sample of water used in making the solu- 

 tion ; for the most probable effect of the impurities is to increase rather 

 than to decrease the conductance of weak acids and bases. As a basis for 

 the correction at the higher temperatures the specific conductance of the 

 water and its impurities was determined by making several heatings with 

 pure water, following exactly the same procedure as in the regular experi- 

 ments. Both my results (reported in this section) and those obtained 

 in the other investigations of this series show that the conductance at the 

 higher temperatures does not vary much in successive runs, and that it 

 is not proportional to the conductance at 18. 



At 18 the measured conductance of the water is due almost entirely to 

 the impurities, that due to the hydrogen and hydroxide ions being inap- 

 preciable ; but at 218 and 306 the latter forms a considerable part of the 

 whole. Its amount was determined from a preliminary value of the ioni- 

 zation-constant (Kw) of water, by the formula L = 10 _3 Ch(Ah + Aoh) 

 where Ch (equal to V-Kw) is the concentration, in equivalents per liter, 

 of the hydrogen (or hydroxide) ions in pure water, and (Ah + Aoh) is 

 the sum of the equivalent conductances of hydrogen and hydroxide ions, 

 calculated by adding the A n value for sodium hydroxide to the difference 

 between the A values for hydrochloric acid and sodium chloride. 



