Section 62. Preparation of the Solutions. 757 



solution whose content had been determined by precipitation with silver 

 nitrate to be 137.37 milli-equivalents per kilogram of solution.* 



Solution No. 1 was thus found to contain 151.00 milli-equivalents 

 NaOH per kilogram ; solution No. 2, 105.32 milli-equivalents. The more 

 dilute solutions (0.05 to 0.002 normal), whose conductance was actually 

 measured, were prepared from these stock solutions by adding weighed 

 portions of them to weighed quantities of water. The water used for 50 

 and 20 milli-normal solutions had a specific conductance at 18 of 

 0.9 X 10" 6 ; that used for the diluter solutions one of 0.6 X 10" 6 - Two or 

 three independently diluted solutions were always prepared at the concen- 

 trations of 4- and 2-milli-normal. The reduction of content from milli- 

 equivalents per kilogram to milli-equivalents per liter at 4 was made with 

 the help of the specific gravity (1.002) in the case of the 50 milli-normal 

 solution; but this correction was insignificant and was not applied in the 

 case of the more dilute solutions. f 



Three separate stock solutions of ammonium hydroxide, each approxi- 

 mately 0.1 normal, were prepared in the course of the work from water of 

 conductivity 0.5-0.8 X 10" 6 and from a "strictly chemically pure" sample 

 of ammonia water (sp. gr. 0.90) furnished by Baker and Adamson of 

 Easton, Pennsylvania, and stated to be free from amines and hydrocar- 

 bons. Ten cubic centimeters of the strong ammonia yielded only 1 mgm. 



*Derived from the following data : 



Grams solution taken 112.34 106.58 91.31 



Grams AgCl obtained 2.2121 2.0988 1.7994 



Milli-equiv. per kilogram. .137.34 137.34 137.44 



The weights of silver chloride were reduced to vacuo. The atomic weights used 

 were Ag = 107.93, CI = 35.45. The equivalent conductance of the solution was also 

 determined at 18 in a U-shaped cell and found to be 348.0 while Goodwin and 

 Haskell (Phys. Rev., 19, 382, 1904) found 347.8 for the same solution. 



fAfter the conductance measurements were completed 481.3 grams of the second 

 solution were (December 12, 1904) acidified with hydrochloric acid, evaporated to 

 dryness, and the residue was heated to 200 and weighed; it amounted to 2.997 

 grams, which if regarded as NaCl would correspond to a content of 106.5 millimols 

 per kilogram solution while that found earlier by titration was 105.3. Exactly one- 

 half of this excess of 1.1 per cent was found to be due to silica (or some other sub- 

 stance insoluble in strong acid), which had apparently been taken up from the bottle 

 since the solution was first prepared (October 22, 1904). About one-half of the 

 remainder consisted of barium sulphate or sodium sulphate ; for this amount was 

 precipitated when barium chloride was added to the filtrate from the silica after 

 evaporation to remove the acid present. These impurities certainly could not have 

 nearly so great a percentage effect on the conductance as upon the total solid-con- 

 tent, both because of their having a lesser equivalent conductance than the base, 

 and because they tend to compensate each other, acids like silica_ reducing and salts 

 increasing the conductance. Still their presence might possibly give rise to an error 

 of 0.5 per cent in the equivalent conductance of the base. That the stock solution 

 was not much contaminated with carbonic acid was shown by the fact that even at 

 the end of the measurements it gave no precipitate with a large quantity of barium 

 hydroxide. The conductance of a 2-millimolal solution prepared from the second 

 stock solution on December 1 to 4 was greater than that of one so prepared on 

 November 10 by only 0.2 per cent at 18, which is within the experimental error at 

 that concentration, showing that little if any change with the time took place. 



