OF ARTS AND SCIENCES. 441 



by filtration and washing, ignited, again fused with a little sodic car- 

 bonate, the result of fusion dissolved in strong sulphuric acid, diluted 

 with water, and precipitated by ammonia with tlie subsequent addition 

 of acetic acid, and boiling. Experiment (35) was conducted in like 

 manner, excepting that the sodic titanate was dissolved off the filter 

 by means of a mixture of hydrochloric and oxalic acids, the latter 

 destroyed by potassic permanganate (since there is danger that in 

 presence of ammonic oxalate the titanium may not be completely 

 precipitated by ammonia), ammonia added in excess, then acetic acid 

 to strongly acid reaction, and the liquid boiled. Tliis mode of treat- 

 ing the acid titanate is not desirable, since the titanic hydrate, if pre- 

 cipitated but once subsequently, retains a trace of manganese. To 

 fuse the titanate with sodic carbonate, and then treat the melt with 

 strong sulphuric acid, is probably the best way of getting the titanium 

 into solution again. The test solution of (28) and (29) was emj^loyed 

 in experiment (35), and that of (30) and (31) in the rest. 



From experiments (36) and (37), together with (35), which, as 

 has been remarked, would naturally give figures slightly too high, it 

 appears that the method indicates with accuracy the amount of tita- 

 nium present. Experiments (38) and (39) indicate that the separa- 

 tion from alumina is not unreasonably inexact. 



Weight of Solution. TiO, found. Ti02 by Standard. Error. 



(35) 5.2220 grm. 0.0383 grm. 0.0373 grm. 0.0010 grm. + 



(36) 5.2300 " 0.0371 " 00370 " 0.0001 " + 



(37) 5.3840 " 0.0380 " 0.0381 " 0.0001 " — 



(38) 5.2920 " 0.0307 " 0.0374 " 0007 " — 



(39) 5.2540 " 0.0383 " 0.0372 " 0.0011 " + 



The tediousness of filtration, which is a consequence of the nature 

 of the phosphate precipitate, is the great objection to the method, and 

 on account of it the testing was pushed no further, attention being 

 turned instead to the second line of experimentation. 



The incompleteness of the precipitation of alumina by the basic 

 acetate process in presence of a large excess of acetic acid, suggested 

 the attempt to hold up alumina entirely by means of a sufficient excess 

 of acetic acid while precipitating titanium. Experiments to deter- 

 mine the amount of acetic acid necessary to prevent the precipitation 

 of alumina from a boiling solution of the acetate, indicate that amounts 

 of absolute acid in excess of five per cent by volume of the solution 



