﻿washing of Aluminum. • 115 



Weight of A1 2 3 . 



Weight of 



•1041 



•1093 



•1057 



•1080 





•1096 



•1045 



•1099 





•1107 



•1032 



•1093 



•1037 



•1093 



•1031 



•1083 



Average -1040 



•1093 



Standard -1002 



•1107 



The precipitates were in all cases ignited for a long time to 

 full redness over a ring burner. It will be seen that the BeO 

 is quite constant and the average only a little below the stand- 

 ard. The A1 2 3 is in all cases too high. If the quantity of 

 mixed beryllium and aluminum oxides be determined in any 

 compound, and then the BeO separated according to the above 

 method and the A1 2 3 determined by difference, the results will 

 be very nearly correct. The determinations of A1 2 3 and BeO 

 in beryl given earlier in this article were made in this way. 



To account for the excess of A1 2 3 we thought that perhaps 

 appreciable quantities of silica were dissolved from the glass 

 by boiling the alkaline solutions of the mixed oxides for one 

 hour. We therefore in every case dissolved the oxides in hy- 

 drochloric acid and deducted from their weight a slight quan- 

 tity of insoluble matter amounting to about '0010 grams in 

 the A1 2 3 and *0004 grams in the BeO. The excess of A1 2 3 

 must, therefore, have come from some other cause than silica 

 dissolved from the glass. The alkaline solutions are so dilute 

 that the glass is very little acted on, and determinations made 

 in a platinum dish convinced us that the separation can be 

 made just as well in beakers. We have concluded that the 

 beryllium is not completely precipitated from a solution in soda 

 by boiling and that it is not possible to wash out all of the soda 

 from the precipitate. The following experiments on the deter- 

 mination of BeO in beryl will show this. The first column 

 gives the weight of BeO obtained by boiling the soda solution, 

 the second the same oxide dissolved in hydrochloric acid, repre- 

 cipitated with ammonia and ignited. 



•1476 -1440 



. -1464 -1422 



•1330 -1300 



•4270 -4162 



The total BeO in these experiments weighed 4270 on first 

 precipitation, '4162 on second, a loss of '0108. Calculating by 

 Am. Jour. Scl— Third Series, Vol. XXXII, No. 188— August, 1886. 



