140 Thornton and II ay den — Separation of Zirconium. 



sulphate just as was done in the two series above. In Table 

 III are put forth the results of four experiments. 



Table III. 

 The Separation of Zirconium from Iron and Aluminum. 





Zr0 2 



Fe 2 3 



A1 2 3 



Zr0 2 





Tartaric H 2 S0 4 



VollUlK 



No. 



taken 



taken 



taken 



found 



Error 



acid 



(1:1) 



of soln 





grm. 



grm. 



grm. 



grm. 



grm. 



grm. 



cm 3 



cm 3 



7. 



0-1087 



0-1018 



0-05635 



0-1086 



—o-oooi 



2 



40 



400 



8. 



0-1089 



0-1018 



0-05635 



0-1088 



—o-oooi 



2 



40 



400 



9. 



o-iioi 



0-2036 



0*1127 



o-ino 



+ 0-0009 



2 



40 



400 



10. 



o-iioo 



0-2036 



0-1127 



0-1103 



+ 0*0003 



2 



40 



400 



The technique of the analysis is here connectedly given. 

 Tartaric acid is added to the solution in amount equal to live 

 times the aggregate weight of the three bases to be held by it 

 in solution. It is desirable that the initial volume of the test 

 solution should not much exceed 100 cm3 . To facilitate reduc- 

 tion of the iron the solution is made neutral to litmus paper 

 with ammonium hydroxide and then acid again with 2 cm3 of 

 sulphuric acid (1 : 1). Hydrogen sulphide is then passed in 

 until the iron has all been reduced to the ferrous state. The 

 solution is now made slightly ammoniacal and more hydrogen 

 sulphide introduced till the iron has been completely thrown 

 out as ferrous sulphide, leaving the solution, however, alkaline 

 to test-paper. The ferrous sulphide is thrown on a filter and 

 washed ten times with very dilute colorless ammonium sul- 

 phide. The filtrate is acidified with 40 cra3 to 60 cm3 of sulphuric 

 acid (1 : 1) and the hydrogen sulphide thus liberated boiled 

 out. The solution is then cooled (best by iced water), the 

 volume made to 400 cm3 , and the " cupferron " reagent added 

 gradually in the cold with constant stirring. Without 

 delay the filtration is proceeded with, using a paper filter and 

 mild suction. The filtrate should be tested as to completeness 

 of precipitation by adding thereto a few drops of the reagenr. 

 A white turbidity will appear immediately if the solution still 

 contain zirconium — otherwise a small white precipitate of 

 nitrosophenylhydroxylamine which disappears on standing. 

 After a little experience the analyst will have no difficulty in 

 distinguishing these two precipitates, which are very different 

 in appearance. The precipitate is washed at least twenty times 

 with hydrochloric acid (100 cm3 of acid of sp. g. = 1*20 diluted 

 to one liter), agitating the filter contents as much as possible 

 and not allowing the washing fluid to run through too fast to 

 accomplish much solvent work. After having been sucked 

 free from drainage liquid at the pump, the precipitate along 

 with the filter is placed in a tared platinum crucible and dried 



