The solution containing the zinc is contained in an evaporating dish, and freed from sulphuretted hydrogen by boiling, and, if necessary, from an excess of acid by evaporation. The evaporating dish must be a large one. Cautiously add sodium carbonate to the hot, moderately dilute solution, until the liquid is distinctly alkaline, and boil. Allow the precipitate to settle, decant on to a filter, and wash with hot water. Dry, transfer to a porcelain crucible (cleaning the paper as much as possible), add the ash, ignite, and weigh. The substance weighed is oxide of zinc, which contains 80.26 per cent. of the metal. It is a white powder, becoming yellow when heated. It must not show an alkaline reaction when moistened. If it contains manganese this metal will be present as sesquioxide (Mn_{2}O_{3}). Its amount can be determined by dissolving in dilute acid and boiling with an excess of sodic hydrate. The oxide of manganese will be precipitated, and can be ignited and weighed. Its weight multiplied by 1.035 must be deducted from the weight of oxide of zinc previously obtained. The results yielded by the gravimetric determination are likely to be high, since the basic carbonate of zinc frequently carries down with it more or less soda which is difficult to wash off. VOLUMETRIC DETERMINATION This method is based on the facts that zinc salts in an acid solution decompose potassium ferrocyanide, forming a white insoluble zinc compound; and that an excess of the ferrocyanide can be detected by the brown coloration it strikes with uranium acetate. The method resembles in its working the bichromate iron assay. The standard solution of potassium ferrocyanide is run into a hot hydrochloric acid solution of the zinc until a drop of the latter brought in contact with a drop of the indicator (uranium acetate) on a white plate strikes a brown colour. The quantity of zinc in the solution must be approximately known; run in a little less of the ferrocyanide than is expected will be necessary; test a drop or two of the assay, and then run in, one or two c.c. at a time, until the brown colour is obtained. Add 5 c.c. of a standard zinc solution, equivalent in strength to the standard "ferrocyanide," re-titrate, and finish off cautiously. Of course 5 c.c. must be deducted from the reading on the burette. The precipitate of zinc ferrocyanide formed in the assay solution is white; but if traces of iron are present, it becomes bluish. If the quantity of ferrocyanide required is known within a few c.c., the finishing point is exactly determined in the first titration without any addition of the standard zinc solution. Unfortunately this titration serves simply to replace the gravimetric determination, and does not, as many volumetric processes do, lessen the necessity for a complete separation of any other metals which are present. Most metals give precipitates with ferrocyanide of potassium in acid solutions. If the conditions are held to, the titration is a fairly