quite sufficient for us altogether to abandon the proceeding in determining the acids, and without regret, as we possess volumetrical and other methods of the highest accuracy, by means of which the amount of any of the acids in combination with the fixed bases in the urine may be determined.

There

A physiological law makes this substance the common accompaniment of our food. Salt is a necessary to man and animal, and nature finds means and ways to supply it to both. Like oxygen, which composes two thirds of our globe, salt is found everywhere. Its solubility in water equals the diffusibility of oxygen; and thus it penetrates the masses. is probably no water on the globe which does not contain it, and consequently it pervades the vegetable and the animal kingdoms. Of all waters, sea and mineral water contain the greatest abundance of it. It occurs, moreover, in a solid crystalline state, mostly imbedded in layers of gypsum, and then is termed rock-salt. Of this we have to make an important use in the quantitative analysis of the chloride of sodium of the urine. For, no salt being so pure as rock-salt, a solution of it in water serves as the basis for the preparation of all the tests which are required for the volumetrical method.

Crystallography.

Chloride of sodium crystallizes in the cubic system. The cleavage of rock-salt leads to a cube, and crystallizations of salt from pure solutions always assume this form. If, however, certain organic substances are mixed with the solution, the crystallization will ensue in the form of the regular octahedron. Chloride of sodium, therefore, when crystallizing out of the urine or other animal fluids, always appears in octahedra. The latter are best obtained by evaporating a large amount of (one day's) urine to a thin syrupy consistence, and letting it stand over night. The crystals will be found in the sediment on decanting the supernatant fluid, and are good objects for microscopical observation.

Chemical properties.

Chloride of sodium dissolves in water of from 54° to 75° F. (12° to 24° C.), invariably at the rate of 31.84 parts in 100 parts of the saturated solution. For procuring the latter clear rocksalt is best used, as already described under Urea (p. 57).

In solutions of chloride of sodium, nitrate of silver produces a white precipitate, which is insoluble in nitric and hydrochloric acids. We make use of this test for removing chlorides from the urine when we wish to ascertain accurately the amount of urea by the volumetrical method. In the same solutions the nitrate of the suboxyde of mercury produces a precipitate of calomel (subchloride of mercury).

If a concentrated solution of chloride of sodium is mixed with a similar solution of nitrate of protoxyde of mercury, nitrate of soda and bichloride of mercury (corrosive sublimate) are formed, the latter transforming the fluid into a white magma of crystals. The same juxtaposition takes place in dilute solutions, with the difference, however, that the fluid remains clear, because the sublimate is kept in solution. In solutions which contain chloride of sodium and urea at the same time, no precipitation of urea by nitrate of protoxyde of mercury will take place, as long as any chloride of sodium is yet present untransformed into nitre and corrosive sublimate. Upon this test is based the volumetrical analysis of chloride of sodium by Professor Liebig.

Diagnosis in Urine.

As it is, in some cases, of interest to know whether chloride of sodium be present in the urine or not, it is useful to have a short qualitative test at hand for ascertaining the fact. This test is nitrate of silver. One caution only is required, namely, to strongly acidulate the urine to be tested with nitric acid, in order to prevent the precipitation of phosphate of silver. Or the nitric acid may be added after the nitrate of silver, when any phosphate of silver will be immediately dissolved.

Liebig's Method of determining the quantity of Chlorine in neutral fluids by means of Nitrate of Protoxyde of Mercury.

From the analysis of urea we are already acquainted with the fact, that nitrate of protoxyde of mercury produces a

copious white precipitate in a solution of urea. This precipitate is not produced with corrosive sublimate.

On mixing a chloride of any of the alkali metals with nitrate of protoxyde of mercury, a transmutation of the two salts into corrosive sublimate and a nitrate of the alkaline base takes place. We have already seen the result of this process as regards chloride of sodium.

If a solution of urea is mixed with some chloride of sodium, and a dilute solution of nitrate of protoxyde of mercury is then added in small portions, a white turbidity occurs on the spot where the two fluids mix with each other; but this turbidity immediately disappears if the fluid is shaken a little, and the latter remains as clear and transparent as before the addition of the nitrate; without the chloride of sodium it would have remained permanently thick. On the addition of the nitrate being continued, the precipitate will disappear until the whole of the chloride has been transformed into corrosive sublimate. Beyond this limit a single drop of the mercurial solution produces a permanent turbidity of the fluid.

From this it is evident that, if we know the amount of mercury contained in the solution of the nitrate of the protoxyde of mercury, which has been added to the solution of urea containing an unknown amount of chloride of sodium, until the permanent turbidity was produced, the amount of chlorine, or chloride of sodium, contained in that solution may be also known. One equivalent of mercury of the mercurial solution consumed, exactly corresponds to one equivalent of chlorine or chloride of sodium.

On the contrary, if the amount of chloride of sodium contained in the solution of urea be known, and the amount of mercury contained in the mercurial solution be unknown, it is easy to calculate the amount of mercury contained in the mercurial solution used.

This proceeding for ascertaining the amount of chloride of sodium is particularly applicable to the urine, because the addition of urea is here not required. It may, of course, be used with advantage for ascertaining the amount of chlorine contained in brine or sca-water, and, generally speaking, in all cases where a large number of analyses have to be made in the shortest possible time. If, however, the amount of chlorine in fluids not being urine is to be determined, the proceeding has to undergo some modification.

I have already described (pp. 54, 55) the simplest modes of obtaining solutions of the nitrate of the protoxyde of merCare must be taken not to use the common mercury cury.

of commerce, because it always contains lead and bismuth, which render the analysis of chloride of sodium uncertain. If either lead or bismuth be present in a solution of mercury, it will, on the latter being mixed with a solution of urea containing chloride of sodium, immediately cause a white turbidity or opalescence, which makes it impossible to see distinctly the point at which the combination of urea and oxyde of mercury begins to be precipitated.

If, therefore, it is the intention of the operator to use the common mercury of commerce, it will be best for him to transform it into crystallized protonitrate or nitrate of suboxyde of mercury, by boiling an excess of the metal in dilute nitric acid, concentrating and cooling the solution. The crystals of this salt are then separated from the motherliquor, which contains the foreign metals; they are washed with dilute nitric acid, afterwards with a little water, by which process a part is transformed into basic salt. If the commercial nitrate of the suboxyde be used, this process of washing must always be gone through, because the manufacturers simply remove the crystals from the mother-liquor without washing them. Small pieces of the crystallized salt should not be used at all, because the yellowish mother-liquor adheres to them with such pertinacity, that it is difficult to remove it by washing without dissolving the greater part of the salt also.

The crystals of nitrate of suboxyde of mercury are now dissolved in nitric acid, and heated until the evolution of vapours of nitrous acid has entirely ceased, and a drop of the solution is no longer precipitated by chloride of sodium. The solution, after evaporation on the water-bath to a syrupy consistence, is diluted with ten times its own volume of water. If, after the lapse of twenty-four hours, any basic salt of the protoxyde has been precipitated, it may be removed by filtration.

In order to make this solution serviceable for the quantitative analysis of chloride of sodium it must be graduated, so as to contain a definite amount of nitrate of protoxyde of mercury in a given volume. This may be effected in two ways. It is either graduated directly by means of a solution of chloride of sodium of known strength, or, after the amount of protoxyde it contains has been determined, it may be diluted with as much water as is necessary, in order to make one cubic centimètre of this dilute mercurial solution indicate exactly 10 milligrammes of chloride of sodium. For both proceedings a solution of chloride of sodium is required, containing a known amount of this salt. The prepa

ration of the standard saturated solution has already been described at p. 56. Of this saturated solution we take with a pipette, observing the usual caution, 20 c.c., and add 298 4 c.c. of water, whereby we obtain 3184 c.c. of dilute solution of chloride of sodium, containing in all 2×3184 milligrammes of chloride of sodium; 10 c.c. of this solution contain, therefore, 200 milligrammes of chloride of sodium.

Preparation of mercurial solution graduated for chloride of sodium.-Ten cubic centimètres are measured by means of a small pipette delivering exactly that amount of fluid after having been filled up to the mark on the narrow tube. These 10 c.c. are poured into a small beaker; to this are added 3 c.c. of a solution of urea, containing in 100 c.c. 4 grammes of urea, in 1 c.c. therefore 40 milligrammes of urea. For measuring this latter solution a narrow test-tube is very serviceable, when marked with a file at the point to which it will be filled by any 3 c.c. of fluid. It does not matter whether a few drops more or less are taken.

The dilute solution of mercury to be graduated is now filled into a dropping glass or burette, and from this, and after noting down the level, it is added in drops to the solution of chloride of sodium containing urea, which is kept in a rotatory motion. The formation of a distinct and permanent precipitate indicates the completion of the test. An opalescence of the fluid must not be mistaken for the precipitate of urea and protoxyde of mercury. It is caused by a trace of foreign metal; it may easily be recognised as not proceeding from the completion of the test by the circumstance, that after its appearance the turbidity is not increased by the addition of a few more drops of the mercurial solution. If the precipitate has been caused by the compound of urea, every additional drop of the mercurial fluid produces an increase of the precipitate, and therefore makes the fluid thicker than it was before.

In graduating these fluids, I generally take the following caution. I measure 10 c.c. of water into a beaker, add 3 c.c. of the solution of urea, and then one or two drops of the mercurial solution to be graduated. The amount of precipitate thus produced shows the limit to which the addition of mercurial solution to the fluid containing a known amount of chloride of sodium must be carried, in order to be safe against the error from the opalescence of the mixture.

Suppose that there have been used for the production of the precipitate, in 10 c.c. of the solution of chloride of sodium, 7.8 c.c. of the mercurial fluid, the latter is too concentrated to