barium. All the thirteen barium salts which are liable to precipitation under these circumstances are white, with the single exception of chromate of barium. The yellow color of the chromate of barium (§ 38) distinguishes this one precipitate from all the rest. If this color is well marked, the presence of a chromate in the original solution (which must also have been yellow) may be inferred with certainty. In all other cases, however, the precipitate is white, as in the present experiment. The next question is, can anything further be learned from this white precipitate?

Add to the contents of the test-tube dilute chlorhydric acid, until the liquid has a decidedly acid reaction to litmus paper. An effervescence indicates the escape of carbonic acid, displaced by the less volatile chlorhydric acid. The bulky precipitate which the chloride of barium produced will in part disappear, but a portion of it remains undissolved. Filter the contents of the tube. The particles of precipitated sulphate of barium are so very fine that they often pass through the pores of the paper, necessitating repeated filtration through the same filter. To the filtrate add ammonia-water until the liquid has an alkaline reaction. A precipitate will reappear: the phosphate of barium which was dissolved by the chlorhydric acid is reprecipitated as soon as the acid solvent is neutralized by ammonia-water.

48. Of all the barium salts which might, in an actual analysis, be precipitated under conditions similar to those of the preceding experiment, only one, the sulphate of barium, is insoluble in chlorhydric and other strong acids. A separation of sulphur from a hyposulphite will not be mistaken for a barium precipitate (§ 19, p. 20). The fact that any of the original precipitate remains undissolved by the chlorhydric acid demonstrates the presence of a sulphate. The portion of the original precipitate which dissolved in the acid, and was reprecipitated by ammonia, consisted in this particular experiment of phosphate of barium; but in an actual analysis the possible salts represented would be so numerous as to make the indication of but little value.

If ammonia-water should cause no precipitation in the acid liquid which was filtered from the original precipitate, it must not be inferred that the acid of course dissolved nothing. The borate, oxalate, arseniate, arsenite, tartrate and fluore of barium are all moderately soluble in solutions of ammoniacal salts, and may not be precipitated on the addition of ammonia. Of course, if the original solution contained ammonium salts, these six barium salts, if present in small quantity only, might fail to be precipitated. In fact, all the thirteen above-mentioned barium salts, except the sulphate, are more or less soluble in solutions of ammonium salts, so that, whenever ammonium salts are known to be present, there is really but one perfectly satisfactory indication with chloride of barium. The presence of a sulphate is revealed by it with certainty, but the results of the other tests must be received with some distrust.

If the original solution be acid, it is necessary to neutralize it with ammonia-water before the chloride of barium is added. Sometimes a precipitate is produced by the ammonia-water so added; in that case it is necessary to filter and proceed with the filtrate, although the ammonia-water may have thrown down salts of several of the classes above named, such as phosphates, oxalates and fluorides. (See also § 28, p. 37.) Even if the ammonia-water produce no precipitate, the testing is then performed under the disadvantage of the presence of ammonium salts.

If the original solution contained silver or lead salts, or mercurous salts, it would be impossible to use chloride of barium and chlorhydric acid as reagents; they would throw uown the chlorides of those metals. Nitrate of barium (App., § 44) and dilute nitric acid must then be used.

The acids added to the precipitate formed by the barium salt must be always dilute acids. Chloride and nitrate of barium are themselves insoluble in concentrated chlorhydric and nitric acids, and if a strong acid were used as a solvent, the reagent salt might itself separate from the liquid.

49. The Calcium Test. Chloride (or nitrate) of calcium precipitates the same classes of salts as chloride (or nitrate) of barium, with the single exception of the chromates. When sulphates and all ammoniacal salts are absent, or present only in minute quantities, something may be learned by testing a neutral or slightly alkaline solution supposed to contain representatives of some of the other classes of salts enumerated in § 46, with chloride or nitrate of calcium. The calcium salts liable to precipitation under these circumstances are all soluble in acetic acid, except the oxalate and the fluoride. The precipitate produced by the calcium reagent is, therefore, treated with acetic acid; if it completely redissolves, oxalates and fluorides are most probably absent. The presence of notable quantities of ammonium salts renders this test of uncertain value; because the fluoride and many other salts of calcium are soluble in solutions of ammonium salts. Since sulphate of calcium is sparingly soluble in water and acetic acid, the presence of a sulphate, causing precipitation of sulphate of calcium, obscures the reaction for oxalates and fluorides. Nitrate of calcium must be used instead of the chloride whenever silver or lead salts, or mercurous salts, are present in the solution under examination.

50. Illustration of the Calcium Test. Prepare in a test-tube an aqueous solution of phosphate, oxalate and tartrate of sodium or potassium. A very small quantity of each salt will be enough. The solution will be neutral or faintly alkaline. Add to this solution a solution of chloride of calcium (App., § 42) until the precipitation is complete. Collect the white precipitate upon a filter, and, when drained, transfer it to a test-tube and treat it with acetic acid. The phosphate and tartrate of calcium will redissolve, but the oxalate remains untouched. To verify this result, and identify each one of the classes of salts present in the original solution, special tests, to be hereafter described, must be resorted to.

51. The Silver Test. - Nitrate of silver produces a precipitate in neutral or acid solutions with all chlorides, bromides, iodides and cyanides, and in neutral solutions with most of the classes of salts enumerated in § 46. In order to obtain the most comprehensive negative conclusion in the case that nitrate of silver produces no precipitate, it is necessary to operate upon a neutral solution. If, on the addition of nitrate of silver to a neutral solution, no precipitate appears after the lapse of several minutes, neither chlorides, bromides, iodides, cyanides nor sulphides can be present, and the absence of sulphites, hyposulphites, carbonates, phosphates, arseniates, arsenites, chromates, silicates, oxalates and tartrates may also be inferred with considerable certainty.

52. Illustration of the Silver Test. — Prepare in a testtube a weak solution of chloride of sodium, iodide of potassium, cyanide of potassium and phosphate of sodium. Add nitrate of silver (App., § 39) to this slightly alkaline solution, until the precipitation is complete. The dense precipitate is yellowish-white. l'our dilute nitric acid into the mixture, until the solution is strongly acid; shake up the contents of the tube thoroughly, and after the lapse of several minutes collect the insoluble precipitate on a filter, and receive the filtered liquid in a test-tube.

Neutralize the filtrate with ammonia-water: a yellow precipitate of phosphate of silver will reappear.

Wash the precipitate on the filter thoroughly to free it from the superfluous nitrate of silver. Rinse the washed precipitate into a clean test-tube, decant the water from above it, pour over it ammonia-water, and gently heat the mixture. The silver precipitate will visibly diminish in bulk, but a yellowish portion remains undissolved. Filter again, and neutralize the filtrate with nitric acid; a white precipitate will fall.

This experiment proves that a portion, but not all, of the mixed silver salts which are insoluble ir nitric acid, are solu

*

ble in ammonia-water. The chloride, cyanide and bromide of silver dissolve in ammonia-water, the latter with difficulty; the iodide remains undissolved. Special tests, hereafter to be described, are applied in order to confirm the presence of iodine, and to detect each and all of the three substances which are liable to be confounded in the white precipitate just mentioned.

53. In the application of the silver test to the examination of a substance of unknown composition, it is, of course, most advantageous to work with a neutral solution, as in such a case the absence of any precipitate would lead to the inference that all the salts mentioned in § 51 are absent: if the solution is neutral, the nitrate of silver may be added directly to a portion of it. In addition to the classes of salts mentioned in § 51, if the original solution contained any considerable quantity of a borate, the borate of silver would be precipitated under these conditions; but a small proportion of some borate might escape precipitation. If the original solution be acid to test paper, add nitrate of silver to a portion of it in a test-tube, and then pour in upon the liquid some dilute ammonia-water, so gently that the two liquids do not mix at once. At some layer near the junction of the two dissimilar liquids, the fluid must be neutral. If at that layer no precipitate or cloud appear, the twelve kinds of salts above enumerated are absent. If the original solution is alkaline, dilute nitric acid is to be added in precisely the same manner as the ammonia-water in the opposite case. The neutral layer between the two liquids is attentively observed, and the absence of any film or cloud therein justifies the same sweeping conclusion as that above given.

Some conclusions may often be drawn from the color of the precipitate produced by nitrate of silver. Chloride, bromide, cyanide, oxalate, tartrate, and borate of silver are white; iodide, phosphate, and arsenite of silver are yellow; silicate of silver is yellow or white; arseniate of silver is brownish-red; chromate of silver is purplish-red; sulphide