2 THE ROYAL SOCIETY OF CANADA 



is, however, according to this investigator, a different modification 

 of tlie hydroxide from that precipitated, the solubiUty being due to this 

 difference rather than to the formation of ammonium aluminate. 

 The work of W. Blum^ sheds a great deal of hght upon the question 

 of the formation of the aluminates of sodium and potasium in solution 

 and, by inference, adds to our knowledge of the corresponding am- 

 monium aluminates. A later contribution by the same author^ deals 

 more directly with the formation of ammonium aluminate. The 

 hydrogen ion concentration of solutions from which aluminium is 

 being precipitated, by the addition of ammonium hydroxide, is fol- 

 lowed by means of the hydrogen electrode. The results show that an 

 appreciable amount of aluminium remains dissolved when the hydro- 

 gen ion concentration (H+) is less than a 10"^. The reduction of the 

 alkilinity of the solution containing ammonium hydroxide, by the 

 addition of ammonium chloride, is also clearly shown. 



The influence of the incomplete precipitation of the aluminium 

 hydroxide upon the estimation of aluminium has been discussed by 

 Penfield and Harper^ by Hillebrand^ and Daudt^ and many others 

 and numerous modifications of the usual procedure have been sug- 

 gested. 



Experimental Procedure. 



The procedure adopted in making measurements was about 

 as follows : — Pieces of Jena glass tubing of a suitable size were sealed 

 at one end and drawn out appreciably at the other that they might 

 be the more easily sealed off when filled. A portion of freshly pre- 

 cipitated aluminium hydroxide which had been washed with some of 

 the solution that was to be used in the experiment, was transferred 

 to the glass tube; a sufficient quantity of the solvent was added and 

 the end of the tube then sealed in the blowpipe. Several tubes, pre- 

 pared and filled in this way, were placed in a bath the temperature 

 of which could be controlled within narrow limits by means of a ther- 

 mostat, and here they were rotated until equilibrium had been reached. 

 The tubes were then removed and opened at a deep scratch previously 

 made. The contents were filtered and 10 c.c. of the filtrate- were 

 measured off and evaporated in weighed platinum crucibles. In 

 cases where only volatile ammonium salts were present, the contents 

 of the crucibles were blasted and weighed. This was followed by the 



1 Jour. Am. Chem. Soc, 35, 1499 (1913). 



2 Ibid 38, 1282 (1916). 



3 Am. Jour. Sci., 32, 107 (1886). 



* U.S. Geol. Survey, Bull. 422, 99 (1910). 

 6 Jour. Ind. Eng. Chem., 7, 847 (1915). 



