separate it by direct precipitation. This can be accomplished with a mixture of uranium and iron by applying, under suitable circumstances, the method already described for determining uranium

in pure solutions. The operations are similar to those used for separating small amounts of aluminium and chromium from steel as phosphate. The separation can, of course, be used equally well when the proportion of uranium is great.

The iron being in the ferrous state, and preferably as chloride, heat the solution nearly to boiling, add microcosmic salt in the proportion already indicated, and then dilute ammonia until the precipitated flocks of ferrous hydrate are no longer quite redissolved. Clear with the smallest possible excess of hydrochloric acid, add 10 grams of sodium thiosulphate and 20 c.c. acetic acid, and boil for ten to fifteen minutes. Allow the precipitate to settle, filter as quickly as possible, and wash with hot water containing a few cubic centimetres per litre of acetic acid and ammonium acetate.

The residue, after redissolving from the filter and the accompanying sulphur, may be reprecipitated to separate the remaining portion of iron, which, when the operation is properly performed,

is rarely greater and often much less than the amount of uranium. Or, instead, the residue may be ignited, fused with sodium carbonate, dissolved in hydrochloric acid, filtered after evaporation to remove any such insoluble body as silica, and reprecipitated as before. The ignited precipitate has the characteristic green colour, and may, if desired, be converted to the yellow pyrophosphate. Not more than a negligible trace of ferric oxide, carried down by the sulphur in the final precipitation, should be present. The following results were obtained in this way :

The addition of ammonium acetate to complete the precipitation and shorten the boiling is not commendable when much iron has to be separated. In order to avoid violent bumping, I have found the following plan to work well: Slide a small widely perforated filter-plate into the solution, and limit its movement along the bottom of the beaker by putting the drawn-out end of a glass

stirring-rod through one of the holes. A similar plate, allowed to move freely, breaks the steam envelopes in a haphazard way, and may cause the liquid to boil over.

SEPARATIONS FROM NICKEL, COBALT, AND

ZINC.

These metals are usually separated from uranium by passing a stream of sulphuretted hydrogen through a boiling solution of the acetates (Gibbs). Zinc only may be separated by means of ammonium carbonate (Rammelsberg). Nickel and cobalt, on the other hand, remain in solution when an emulsion of barium carbonate is added to a mixture of the metals containing a small amount of free acid and some ammonium chloride. Uranium is precipitated, but must, subsequently, be separated from the excess of barium carbonate. The electro-deposition of the uranium as hydrated oxide from an acetate solution is not interfered with by the presence of zinc.

The presence of either nickel, cobalt, or zinc in no way interferes with the estimation of uranium by the method already given for pure solutions. The following are results of test analyses :

The respective filtrates can be used conveniently for determining the separated metal; those containing zinc or cobalt (after boiling with nitric acid to destroy the thiosulphate) by precipitating as the double ammonium phosphate, and those containing nickel by making alkaline and titrating with potassium cyanide and silver iodide.

SEPARATIONS FROM MANGANESE.

The separation of manganese from uranium does not appear to have been much studied. It can be separated, according to Rammelsberg, with ammonium carbonate alone, but the separation generally described is that applicable to other members of its group, viz. precipitation with sulphuretted hydrogen from a solution containing alkaline carbonates.

When uranium is precipitated as phosphate in

the manner described above, it carries down a small portion of any manganese present in the solution. A series of separations of o'10 gram Mn from 0'2045 gram uranium were made in solutions containing amounts of acetic acid, varying from 0 to 50 c.c. The manganese associated with the uranyl pyrophosphate varied from o'0013 to o'0006 gram. When the Mn,P2O, equivalents of these figures were deducted from the total weights, the remainders calculated to amounts of uranium varying between o'2030 and 0'2050 gram.

SEPARATIONS FROM ALKALIS AND ALKALINE

EARTHS.

The separation of alkalis from uranium is of technical importance, as potash is a very considerable constituent of an industrial ore-carnotite. Alkalis, alkaline earths, and magnesia can be separated by precipitating the uranium several times with ammonia in the presence of ammonium chloride; but, considering how badly ammonium uranate filters, this separation is an

1 The amount is easily determined by dissolving in nitric acid oxidising with sodium bismuthate, and titrating the permanganate formed with ferrous sulphate.