170 PROCEEDINGS OP THE AMERICAN ACADEMY 



(4) the magnesia was thoroughly stirred in the solution of boric acid, 

 the evaporation carried at once to dryness, and the crucible and resi- 

 due ignited and weighed; in experiments (5) to (8), the magnesia was 

 dissolved, after the addition of the boric acid, in hydrochloric acid 

 sufficient iu amount to prevent the precipitation of magnesium hy- 

 drate on the subsequent addition of ammonia, ammonia introduced in 

 considerable excess in (7) and (8), in distinct excess in (5) and (6), 

 the whole evaporated and ignited, the residue moistened and again 

 ignited, and this last treatment repeated until the residue ceased to 

 yield vapor of hydrochloric acid when heated. 



B2O3 MgO MgO+B^Oa BjOj 



taken. taken. found. found. 



Error. 



grm. grm. grm. grm. gnn. 



I) 0.1734 0.5005 0.6607 0.1602 0.0132— 



/) 0.1804 0.4973 0.6660 0.1687 0.0117— 



3) 0.1793 0.4949 0.6640 0.1691 0.0102— 



I) 0.1794 0.4941 0.6627 0.1686 0.0108— 



f{5) 0.1807 0.4984 0.6542 0.1558 0.0249— 



) (6) 0.1789 0.4974 0.6687 0.1560 0.0229— 



1 (7) 0.1806 0.4944 0.6684 0.1740 0.0066— 



((8) 0.1789 0.4959 0.6672 0.1713 0.0076— 



From these results it appears plain that under the conditions of 

 the experiments neither magnesia alone nor the magnesia mixture is 

 efficient in fixing boric acid; but in experiments (7) and (8), in which ■ 

 ammonia was employed in large excess, the loss of boric acid is least, 

 so that it would seem to be the case that though ammonia is not a 

 perfect preventive of volatilization it does exert a restraining action 

 on the boric acid. That the magnesia mixture should be incapable of 

 retaining entirely the boric acid present is, as has been pointed out, 

 not surprising ; but that the loss should be so great is rather startling, 

 and more than suggests that the errors of Marignac's process are 

 seriously excessive. The failure of magnesium oxide to hold back 

 boric acid under the conditions of the experiment must be due to a 

 cause other than that which determines the loss during the evapora- 

 tion and ignition of the magnesia mixture, and for this it is natural to 

 turn to the insolubility of the oxide, — a quality likely to oppose some 

 difficulty in the way of establishing complete contact between the 

 boric acid and the magnesia during a short exposure. Direct tests of 

 this point showed distinctly that mixtures of boric acid in water and 

 magnesia, when submitted at once to distillation, yielded boric acid to 



