328 



PROCEEDINGS OF SECTION B. 



TABr.E IIT. 



In No. 2 the TiS04 solution was boiled just as long as No. 4 and 

 No. 6, and a white precipitate appeared. This test is unnecessarily 

 severe. In all cases ferric iron was tested for by NH^CNS. The H,S, 

 after reduction, was boiled off under a current of CO2, and tested for 

 by lead acetate. The results would indicate that — (1) HjS reduction 

 appears to give a slightlv higher value for iron (experiments 5 and 7) ; 

 (2) HjS does not reduce TiS04 to triad (experiment 3) ; (3) Zn Pt. does 

 reduce TiS04 (experiment 2), but does not appear to have any great 

 effect until all the iron is reduced (experiments 4 and 6). 



In the analysis quoted (Table II.), the titanium results obtained 

 by duplicates on each method of reduction of the iron were identical. 

 The titanium solutions from the above tests were titrated colorimetri- 

 cally, and no difference was found after reduction. The conclusion is 

 that if desired the zinc reduction may be safely used as far as the small 

 amount of titanium usually present is concerned, providing the iron is 

 titrated as soon as it is reduced, as determined by NH4CNS. The reduc- 

 tion by HjS is, however, perfectly simple, and, if properly set up, is 

 as quickly done as the zinc method. If the gas is passed into the hot 

 solution, the platinum will readily come down, and should be filtered 

 off. The HgS must, of course, be completely removed by boiling under 

 CO2 before titration of the iron. 



Titanium. — The TiOj is now easily estimated in the solution which 

 has just been titrated for iron. It is necessary that this solution should 

 have had at least 5 per cent. H2SO4 to prevent hydrolytic decomposition 

 of the titanium sulphate into granular metatitanic acid. To the 

 solution, after titration with KMn04, is added three-fifths ccs. hydrogen 

 peroxide (free from F), and it is made up to 250 ccs. ; 100 ccs. is measured 

 into a Nessler tube, and the standard TiSOi (Ice. = -0005 TiO,) is run 

 into another tube containing 90 ccs. water and 2 ccs. H2O2, until colors 

 almost agree. The blank is then made up nearly to 100, and the titra- 

 tion cautiously finished. 



The results are excellent. Varying amounts of hydrogen peroxide, 

 sulphuric acid, temperature or time of standing appear to have no effect. 

 Presence of vanadium or molybdenum would interfere, counting as 

 TiOj, and the presence of fluorine in the H2O2 prevents wholly or in 

 part the formation of the brown color. 



Alumina. — From the weight of the mixed alumina, iron, &c., 

 precipitate, the Fe203, SiOj, TiOz, P2O5, may now be deducted, and the 

 AI2O3 determined by difference. 



