separated. Furtlier, since silver ehloride is
known to be easily soluble in aqueous ammonia,
while nierturous chloride is converted into a
black, insoluble mass containing free mercury,
one might assume that treatment of the two
chlorides with ammonia solution would aflfect
an easy separation of silver chloride from mer-
curons chloride. This case, however, well illus-
trates one of the difficulties of analytical work.
If the amount of mercurous chloride is large in
])roportion to the amount of silver chloride, the
metallic mercury set free b_y the action of am-
monia causes the formation of metallic silver,
which is practically insoluble in ammonia. If,
therefore, ammonia has failed to extract any-
thing from the precipitate in question, we can-
not conclude that silver is absent. We must,
tlien, treat the black mass with a mixture of
nitric and hydrochloric acids, which dissolves the
black substance containing mercury; while the
silver, if at all present in the original substance,
remains behind, again in the form of silver chlo-
ride, but this time unmixed with anything else.
Such cases frequently occur. So often is the be-
havior of a substance toward a reagent modified
by the presence of other substances, that no
scheme of analysis worked out at the writing
table possesses any value until thoroughly tested
in the laboratory.
The filtrate obtained on precipitating out the three metals just spoken of is treated with sul- phuretted hydrogen. This precipitates a second group of metals, which are separated from one another by methods analogous to those employed for tlie first group. The filtrate obtained on pre- cipitating out the metals of the second group is usually treated with ammonium sulphide, and the filtrate from the anuuonium sulphide group with ammonium carbonate. Thus the metals that may be present in the original substance are separated into several groups, and then special .methods are employed to separate and test for the several metals composing each group.
The acid radicals are tested for in a somewhat similar manner, but usually less systematically; because by the time all the metals present have been identified, the analyst usually is able to ex- clude the possibility of the presence of a large number of acids.
The spectroscope (q.v. ) is usually applied to identify the metals potassium and lithium, and is quite indispensable when substances are to be examined to ascertain whether they are in the purest condition possible, since the instrument is capable of revealing the presence of the merest traces of substances. See Spectrum Analysis.
The system of analysis usually followed may be carried out mechanically and almost without intelligence, if the substance examined contains only the more familiar metals and acids, and those in considerable quantities. In fact, quali- tative analysis is criticised by teachers on this account, wlien used as a discipline, or as a means of acquiring a scientific knowledge of chemistry. The ordinary scheme, however, overlooks even some elements of common occurrence, as tita- nium ; and when the chemist has to take into consideration small amounts and the less famil- iar elements, all his chemical knowledge and acuteness find full field for exercise.
It may be seen from the above that the chemist relies on two sets of properties for the identification of a substance. First, those that belong to the substance itself under ordinary conditions; for example, the yellow color and the lightness of sulphur. Such properties may be called properties of condition. On the other hand, if sulphur is heated sufficiently without access of air, it assumes the form of a red vapor ; if heated with access of air, it forms with the oxygen of the air a colorless gas possessing a characteristic odor. The first of these changes is physical; the second, chemical. Physical or chemical changes uuiy thus serve to Ijring out certain properties that are just as characteristic of the substance as the properties of condition. Such properties may be called jiroperties of re- action. They are far more niunerous than prop- erties of condition, and far more useful to the analyst. In the case of sulphur, for instance, the properties of condition are only apparent when the stil]ihur is in a nearly pure form; but the two properties of reaction just mentioned as an example enable us to identify sulphur even when mixed with so much foreign matter that the characteristic color and lightness are quite masked.
Quantitative Analysis. Before beginning a quantitative analysis the chemist must knowj in part at least, the qualitative composition of the substance to be analyzed. This knowledge may be obtained by a special qualitative analysis, or, more frequently, from the results of numerous analyses of similar substances.
Methods of quantitative analysis which involve weighing (see B.vlance) are termed gravimetric. Methods that involve measuring the volumes of solutions are termed volumetric. Finally, meth- ods involving the decomposition of substances by means of an electric current are termed clec- trolyiic.
As an illustration of the methods of gravi- metric analysis, we may take the analysis of an alloy of silver and copper, such as is" used for silver coins in the United States. If high-class w-eights and a balance are at the disposal of the analyst, not more than half a gram (less than one-fourth of a dime) is the most suitable weight to be taken of the alloy. If the weights or the balance is inferior, a larger amount must be taken, so that the errors of weighing may remain proportionately small. The alloy is dissolved in nitric acid, the insoluble residue (carbon and tin oxide) filtered oft' and weighed, and the fil- trate is treated with hydrochloric acid to pre- cipitate silver chloride, just as in qualitative work. In quantitative work, however, certain precautions nuist be taken in carrying out this simple operation. Thus, only a slight excess of hydrochloric acid must be added, since silver chloride is somewhat soluble in a large excess of that acid ; the liquid must be vigorously stirred and warmed to cause the precipitate to assume a form in which it can be easily filtered and washed, etc. The silver chloride is then filtered ofl', dried, and weighed, proper corrections being made for the weight of the ash of the filter. The amount of silver in the alloy is then readily calculated from the weight of silver chloride yielded. The filtrate from the silver chloride contains copper and usually a small amount of lead. The exact amount of copper contained in this filtrate may be best determined by electrolysis. For this purpose the filtrate is first evaporated to dryness, in order to get rid of the hydrochloric acid; the residue is taken up with dilute nitric acid, and the solution thus obtained is subjected to the action of an electric current passing between two
