tlie addition of 25 nil. of concentrntcd liydrocliloric acid; and warmiitg the solution. After cobling, thc solution is dilutcd to 150 i d . Two drops of indicnlor solution arc added nnd tlic solution is titriited witti tlie stnndnrd bromine solution .iititil it is u iliatinct ycllom colour ; 20 mi, of potassium iodidc soIut;oii iirc tiien ntIrIed nnd the solution ib titrntetl with the standard tltiosulphntc solution, using stnrch solution ns indicator. The difference het\vccn the two rendiiigs gives thc correct nmoinit of bromine solution required : tlicrcforc (nil. required) x 0.001 x 100 + 0.5 = yo nliiminium, i.c., (nil. rcquircd) f 5 = yo nliimiiiiunl on ii 0.5-g.'siimple. a It sltould be iiotcd that 50 nil. of 2% osittc soliitioii will completely precipitntc 0.01, g. of nliimiitiuni.
Tlic nbovo mctliotl is npplicnblc to iiormiil mngnesium nlloys, i.c., in tlic nbscncc of mctiils precipitnted by nmmonilr in group 111, If iiiucli iron is present this may be rcinovcd by passing hydrogen sulpliide through the dknliric tnrtrntc solution wrcl filtering iind ~vnsliing, prior to tlic prccipitiitioii of tlic nliimiiiiiiin by osinc.3
According to H. GotiY uluniinium is qrruntitntivcly prccipitnted nt all ~1 1 1 rnlucs Iietwceii 4.2 nnd 9.8, and it is interesting to note tlint for one huiidrcd detcrtiiinations cnrried out nt tlicsc works, the p I l .vnluc of the solutioiis ol)taincd froin tlic trl)ove method rnngetl from 5.20 to 5.tiO. Of tlicsc 80 fdl Iictwceti the rnngc 5-20 to 8.40.
v
The author's tliunks nre due to tlic Directors of Blugnesium Elektron Liiiiitrd for I)crtiiissiou to piiblisli this paper. h