3. The salts of protoxide of iron have in the anhydrous state a white, in the hydrated state a greenish color; their solutions appear colored only when concentrated. Exposed to the air, they absorb oxygen and are converted into salts of the protosesquioxide. The soluble neutral salts redden litmus paper, and are decomposed at a red heat.

4. Acid solutions of salts of protoxide of iron are not precipitated by hydrosulphuric acid; neutral solutions of salt of protoxide of iron with weak acids are precipitated by this reager at the most but very completely; the precipitates are of a black color.

5. Sulphide of ammonium precipitates from neutral, and hydrosulphuric acid from alkaline solutions of salts of protoxide of iron, the whole of the metal as black hydrated PROTOSULPHIDE OF IRON (Fe S), which is insoluble in alkalies and sulphides of the alkali metals, but dissolves readily in hydrochloric and nitric acids: this black precipitate turns reddish-brown in the air by oxidation. To highly dilute solutions of protoxide of iron, addition of sulphide of ammonium imparts a green color, and it is only after some time that the protosulphide of iron separates as a black precipitate.

6. Potassa and ammonia produce a precipitate of HYDRATE OF PROTOXIDE OF IRON (Fe O, H O), which in the first moment looks almost white, but acquires after a very short time a dirty green, and ultimately a reddish-brown color, owing to absorption of oxygen from the air. Presence of salts of ammonia prevents the precipitation by potassa partly, and that by ammonia altogether. If alkaline solutions of protoxide of iron thus obtained by the agency of salts of ammonia are exposed to the air, hydrate of sesquioxide of iron precipitates.

7. Ferrocyanide of potassium produces in solutions of protoxide of iron a bluish-white precipitate of FERROCYANIDE OF POTASSIUM AND IRON (K, Fe,, Cfy,), which, by absorption of oxygen from the air, speedily acquires a blue color. Nitric acid or chlorine converts it immediately into Prussian blue, 3 (K, Fe,, Cfy,) + 4 Cl = 3 K Cl + Fe Cl + 2 (Fe, Cfy,).

8. Ferricyanide of potassium produces a magnificently blue precipitate of FERRICYANIDE OF IRON (Fe, Cfdy). This precipitate does not differ in color from Prussian blue. It is insoluble in hydrochloric acid, but is readily decomposed by potassa. In highly dilute solutions of salts of protoxide of iron the reagent produces simply a deep blue-green coloration.

9. Sulphocyanide of potassium does not alter solutions of protoxide of iron free from sesquioxide.

10. Carbonate of baryta does not precipitate solutions of protoxide of iron in the cold.

11. Borax dissolves salts of protoxide of iron in the oxidizing flame, giving beads varying in color from YELLOW to DARK RED; when cold, the beads vary from colorless to dark yellow. In the inner flame the beads change to bottle-green, owing to the reduction of the newly-formed sesquioxide to protosesquioxide. Phosphate of soda and ammonia manifests a similar deportment with the salts of protoxide of iron; the beads produced with this reagent lose their color upon cooling still more completely than is the case with those produced with borax; the signs of the ensuing reduction in the reducing flame are also less marked.

§ 110.

f. SESQUIOXIDE OF IRON (Fe, O.).

1. The native crystallized sesquioxide of iron is steel-gray; the native as well as the artificially prepared, sesquioxide of iron gives upon trituration a brownish-red powder; the color of hydrate of sesquioxide of iron is more inclined to reddish-brown. Both the sesquioxide and its hydrate dissolve in hydrochloric, nitric, and sulphuric acids; the hydrate dissolves readily in these acids, but the anhydrous sesquioxide dissolves with greater difficulty, and completely only after long exposure to heat.

2. The neutral anhydrous salts of sesquioxide of iron are nearly white; the basic salts are yellow or reddish-brown. The color of the solution is brownish-yellow, and becomes reddish-yellow upon the application of heat. The soluble neutral salts redden litmus paper, and are decomposed by heat. 3. Hydrosulphuric acid produces in neutral and acid solutions of salts of sesquioxide of iron a milky white turbidity, proceeding from separated SULPHUR. This reaction is caused by a mutual decomposition of the sesquioxide of iron and the hydrosulphuric acid, in which the former is deprived of one-third of its oxygen, and thus reduced to the state of protoxide; the oxygen forms water with the hydrogen of the hydrosulphuric acid, and the liberated sulphur separates. Solution of hydrosulphuric acid, rapidly added to neutral solutions, imparts to the fluid a transitory blackening. In solution of neutral acetate of sesquioxide of iron it produces a permanent precipitate of sulphide of iron.

4. Sulphide of ammonium precipitates from neutral, and hydrosulphuric acid from alkaline solutions of salts of sesquioxide of iron, the whole of the metal as black hydrated PROTOSULPHIDE OF IRON (Fe S). This precipitation is preceded by the reduction of the sesquioxide to protoxide. In very dilute solutions the reagent produces only a blackish-green coloration. The minutely divided protosulphide of iron subsides in such cases only after long standing. Protosulphide of iron, as already stated (§ 109, 5), is insoluble in alkalies and alkaline sulphides, but dissolves readily in hydrochloric and nitric acids.

5. Potassa and ammonia produce bulky reddish-brown precipitates of HYDRATE OF SESQUIOXIDE OF IRON (F, O,, H O), which are insoluble in an excess of the precipitant as well as in salts of ammonia. In presence of sesquioxide of chromium, an excess of solution of potassa of ammonia will dissolve part of the sesquioxide of iron along with the sesquioxide of chromium. Generally, however, a portion of both oxides remains undissolved; and, moreover, the dissolved oxides speedily separate again from the solution.

6. Ferrocyanide of potassium produces even in highly, dilute solutions a magnificently blue precipitate of FERROCYANIDE OF IRON, on Prussian blue (Fe, Cfy,), which is insoluble in hydrochloric acid, but is decomposed by potassa, with separation of hydrate of sesquioxide of iron.

7. Ferricyanide of potassium deepens the color of solutions of salts of sesquioxide of iron to reddish-brown; but it fails to produce a precipitate. 8. Sulphocyanide of potassium imparts to neutral or slightly acid solu tions of salts of sesquioxide of iron a most intense blood-red color, arising from the formation of a soluble SULPHOCYANIDE OF IRON. Addition of acetate of soda destroys this color, hydrochloric acid restores it

presence

every

again. This test is the most delicate of all; it will indicate the of sesquioxide of iron even in fluids which are so highly dilute that other reagent fails to produce the slightest visible alteration. The red coloration may in such cases be detected most distinctly by resting the test-tube upon a sheet of white paper, and looking through it from the top.

9. Carbonate of baryta precipitates even in the cold all the iron as a basic salt mixed with hydrate of sesquioxide.

10. The reactions before the blowpipe are the same as with the protoxide.

§ 111.

Recapitulation and remarks. On observing the behavior of the oxides of the fourth group with solution of potassa, it would appear that the separation of the oxide of zinc, which is soluble in an excess of this reagent, might be readily effected by its means; however, in the actual experiment, we find that in the presence of sesquioxide of iron, protoxide of cobalt, &c., some of the oxide of zinc precipitates with these oxides; and if only a small quantity of oxide of zinc is present, it frequently occurs that no trace of this metal can be detected in the alkaline filtrate.

Again, from the behavior of the different oxides with chloride of ammonium and an excess of ammonia, one would conclude that the separation of sesquioxide of iron from the protoxides of cobalt, nickel, and manganese, and from oxide of zinc, might be readily effected by these agents. But this method also applied to the mixed oxides, is inaccurate, since greater or smaller portions of the other oxides will always precipitate with the sesquioxide of iron; and it may therefore happen that, in this process, small quantities of cobalt, manganese, &c., altogether escape detection.

It is far safer therefore to separate the other oxides of the fourth group from sesquioxide of iron by carbonate of baryta, as in that case the iron is precipitated free from oxide of zinc and protoxides of manganese and nickel, and mixed only with a very trifling quantity of protoxide of cobalt

Protoxide of manganese may conveniently be separated from the protoxides of cobalt and nickel and from oxide of zinc, by treating the precipitated sulphides with moderately dilute acetic acid, which dissolves the sulphide of manganese, leaving the other sulphides unacted on. If the acetic acid solution is now mixed with solution of potassa, the least trace of a precipitate will be sufficient to recognise the manganese before the blowpipe with carbonate of soda.

If the sulphides left undissolved by acetic acid are now treated with very dilute hydrochloric acid, the sulphide of zinc dissolves, leaving almost the whole of the sulphides of cobalt and nickel behind. If the fluid is then boiled, to expel the hydrosulphuric acid, and afterwards treated with solution of potassa or soda in excess, the zinc is sure to be detected in the filtrate by hydrosulphuric acid.

Cobalt may mostly be readily and safely detected in presence of nickel by the reaction with borax in the inner flame of the blowpipe. The detection of nickel in presence of cobalt is a less easy task, which may, however, be effected with great accuracy by three different methods. The first method is to add to the solution containing the two metals, nitrate of

potassa in not too small proportion, then acetic acid to strongly acid reaction, and to let the mixture stand for at least several hours in a moderately warm place, when the cobalt will separate as nitrate of sesquioxide of cobalt and potassa; the nickel may then be readily precipitated from the filtrate by soda or sulphide of ammonium. The second method is to saturate with chlorine the very dilute solution of the two metals in hydrochloric acid, having the acid slightly in excess; add carbonate of baryta in excess, and let the fluid stand twenty-four hours. The cobalt is entirely precipitated in this process as black sesquioxide, whilst the nickel remains in solution, and may, after the removal of the baryta by sulphuric acid, be thrown down with solution of soda. The third method is based upon the application of cyanide of potassium. Both cyanide of nickel and cyanide of cobalt dissolve in cyanide of potassium; but cyanide of nickel is precipitated from this solution by acids, whilst cyanide of cobalt is not precipitated if the solution contains free hydrocyanic acid, and has been exposed to the action of heat.* This reaction, i. e., the formation of a precipitate upon adding hydrochloric acid or dilute sulphuric acid to the solution of the two cyanides in cyanide of potassium, prepared under the conditions stated, indicates the presence of nickel. Whether the precipitate be cyanide of nickel or cobalticyanide of nickel, is quite immaterial as far as the detection of nickel is concerned: all we have to bear in mind is simply this, that no precipitate will form if cobalt alone is present in the solution, since cobalticyanide of potassium is not decomposed by hydrochloric acid. In order to get at the composition of the various precipitates which are formed by hydrochloric acid in solutions of the mixed cyanides of nickel and cobalt, and to comprehend the general process of their formation, we have to assume and consider three special and distinct cases, differing from one another in the relative proportions of nickel and cobalt which the solutions under examination respectively contain.

The solution will accordingly, in the first case, contain one eq. of cobalticyanide of potassium (K,, Co, Cy) and 3 eq. of double cyanide of nickel and potassium 3 (Ni Cy, K Cy); now upon the addition of hydrochloric acid in excess to this solution, the double cyanide of nickel and potassium suffers decomposition, and the potassium of the cobalticyanide of potassium transposes with the nickel of the cyanide of nickel the products of this process of double decomposition and transposition are chloride of potassium, hydrocyanic acid, and cobalticyanide of nickel (Ni,, Co, Cy), which latter separates in the form of a dirty-green precipitate containing the whole of the nickel and cobalt present in the solution. In the second case we obtain equally a precipitate of cobalticyanide of nickel; but this precipitate, though containing the whole of the nickel, does not contain all the cobalt of the solution, since the excess of cobalticyanide of potassium is not decomposed. In the third case, lastly, we obtain a precipitate of cobalticyanide of nickel, which contains

In this process the double cyanide of cobalt and potassium (K Cy, Co Cy), which forms at first, is converted by the aid of the free hydrocyanic acid and the excess of cyanide of potassium, into cobalticyanide of potassium (K3, Co2 Cу6) : 2 (Co Cy, K Cy) + K Cy+ H Cy (Kg, Co, Cys) + H.

1=

the whole of the cobalt and a portion of the nickel, mixed with insoluble cyanide of nickel, which contains the rest of the nickel. The cobalticyanide of nickel has been formed here in the same manner as in the first case, whilst the cyanide of nickel owes its formation to the decomposition of the excess of the double cyanide of nickel and potassium. Hence it is evident that the presence of nickel is indispensable to the formation of a precipitate, and consequently that the production of a precipitate is the most positive proof of the presence of this metal.

Protoxide and sesquioxide of iron may be detected in presence of each other, by testing for the former with ferricyanide of potassium, for the latter with ferrocyanide of potassium or, better still, with sulphocyanide of potassium.

In conclusion, it is necessary to mention that alkalies fail to precipitate the oxides of the fourth group in presence of non-volatile organic substances (such as sugar, tartaric acid, &c.). We have already seen that the same applies to alumina.

Sesquioxide of uranium is brick-red, the hydrate is of a yellow color. Upon ignition both are converted into the dark blackish-green protosesquioxide. The solutions of sesquioxide of uranium in acids are yellow; hydrosulphuric acid does not alter them; sulphide of ammonium throws down after neutralization of the free acid, a dark brown precipitate of SULPHIDE OF URANIUM, which subsides slowly, and is readily soluble in acids, even in acetic acid, but does not dissolve in an excess of the precipitant. Ammonia, potassa, and soda produce yellow precipitates of sesquioxide of uranium and alkali, insoluble in excess of the precipitants. Carbonate of ammonia produces a yellow precipitate of carbonate of sesquioxide of uranium and ammonia, which dissolves readily in an excess of the precipitant. Potassa and soda throw down from the solution the whole of the sesquioxide of uranium. Carbonate of baryta completely precipitates solutions of sesquioxide of uranium, even in the cold. Ferrocyanide of potassium produces a reddish-brown precipitate: this is a very delicate test for uranium. Borax and phosphate of soda and ammonia give with sesquioxide of uranium in the inner flame of the blowpipe green beads, in the outer flame yellow beads, which upon cooling acquire a yellowish-green tint.

To separate sesquioxide of uranium from the other oxides of the fourth group, and from alumina, precipitate first with carbonate of baryta, which throws down the sesquioxide of iron, the sesquioxide of uranium, and the alumina, leaving the other oxides in solution. Dissolve the precipitate in hydrochloric acid, remove the baryta from the solution by means of sulphuric acid, filter, and add ammonia to the filtrate until a precipitate begins to form; add now a sufficient quantity of carbonate of ammonia, which has previously been once boiled, to destroy any bicarbonate of ammonia it may contain. Dilute with water and filter off the precipitate, which contains the whole of the sesquioxide of iron and alumina, the filtrate containing all the sesquioxide of uranium.