Section j8. Preparation of the Solutions. 199 



A barium hydroxide solution was prepared for the purpose of titrating 

 acetic acid solutions. This was found by titration against the standard 

 hydrochloric acid, using phenolphthalein and excluding carbon dioxide, to 

 have 80.73 milli-equivalents per kilogram of solution.* 



The ammonia solutions were from two independent sources : first, a 

 special preparation of specific gravity 0.90, obtained from Baker and 

 Adamson, marked " free from amines, carbonate, and silicate" ; second, 

 redistilled liquid ammonia. The solutions were made by filling a 6-liter 

 bottle with conductivity-water, displacing this completely with pure air, 

 then forcing in water of conductance less than 0.9 X 10~ 6 . When the 

 strong ammonia solution was used, it was introduced by means of a pipette, 

 under the surface of the water. When liquid ammonia was employed, it 

 was first drawn off into an iron cylinder and allowed to stand in contact 

 with metallic sodium for several weeks. From this cylinder it was dis- 

 tilled, passing through plugs of asbestos into a small glass bulb from 

 which the air had been previously evacuated ; this bulb stood in a tube of 

 liquid ammonia, which was kept rapidly evaporating by a current of air 

 over the surface of the liquid. The ammonia within the bulb was thus 

 condensed until the proper quantity was obtained (about 15 c.cm.) ; the 

 air current was then stopped, the ammonia surrounding the bulb part'y 

 removed, and the pure ammonia within was allowed to distill through a 

 plug of cotton directly into the water, through the exit tube of the bottle. 

 The last cubic centimeter was rejected. A procedure adopted later con- 

 sisted in distilling the ammonia from the iron cylinder into a flask contain- 

 ing solid ammonium nitrate, and kept in a freezing mixture. The nitrate 

 readily absorbs its own weight of ammonia, and the mixture has a rela- 

 tively low vapor pressure, so that the ammonia could be preserved in the 

 flask, which was closed by a glass stopcock, and could be redistilled there- 

 from at room temperature when needed. 



The concentration was determined by titrating the standard hydro- 

 chloric acid with the ammonia, using as indicator at first phenacetolin, 

 and later Congo red, both of which gave a better end-point than methyl 

 orange. The solutions could not be kept long, as they began to increase 

 in conductance after about three weeks, probably because of action on 

 the glass, and they were not considered trustworthy after the conductance 

 had risen 0.2 per cent. The following list gives the date of making the 

 stock solutions, and the source from which the ammonia was obtained ; the 

 number corresponds to that in section 82. 



*Derived from the following data : 



Grams HCl solution 70.25 82.38 74.36 



Grams Ba(OH), solution 78.76 92.39 83.36 



Milli-equivalents per kilogram 80.74 80.76 80.68 



