two short, open glass tubes. Attach one of these tubes to a gas bottle in which carbonic acid is being generated, so that a slow, steady current of carbonic acid may flow continually through the flask. Heat the contents of the flask in such manner that the iron may dissolve without violent frothing. When the iron has all dissolved, remove the lamp, and place a dish of cold water beneath the flask in order to cool the solution as quickly as possible. But do not interrupt the current of carbonic acid until the contents of the flask are cold. Pour the cold iron solution from the flask into a beaker of about 600 c.c. capacity, wash the flask thoroughly with water, and add water enough to the solution that the beaker may be half-filled.

Fill a capacious Gay-Lussac's burette to the 0° mark with the solution of permanganate of potassium, and pour the latter, little by little, into the iron solution in the beaker. Set the beaker, meanwhile, on a sheet of white paper and stir its contents continually with a glass rod. At first, the red drops of permanganate disappear the moment they touch the iron solution, but after a short time their color is destroyed more slowly. The solution of iron, which was well-nigh colorless in the beginning, gradually acquires a yellowish tint. The permanganate solution must be added the more slowly in proportion as the color of the drops disappears less rapidly. Towards the close of the operation the permanganate should be added drop by drop, until the last drop imparts to the whole of the liquid in the beaker a faint but unmistakable reddish color. As soon as this reddish coloration appears, place the burette in an upright position, and leave it at rest for a moment, in order that the liquid adhering to its walls may drain down; then note how many cubic centimetres of the liquid have been expended.

Weigh out again 1 gramme of the iron and dissolve it as before; again fill the burette with the solution of permanganate, and repeat the experiment in all its details. The number of c.c. of permanganate consumed in this second experiment should not differ materially from the result obtained in the first trial. If there be any marked discrepancy between the results of the two experiments,

the operation must be repeated a third time, or until concordant results are obtained.

The results of our experiments are as follows:

In the first trial 1 gramme of iron wire required 182 c.c. of the solution of permanganate, and in the second trial 183 c.c. Taking the mean of the two trials, 182.5 c.c., the value of the permanganate solution, in terms of iron, may be found by the proportion:

With the chameleon thus standardized we tested a sample of the crystallized double sulphate of iron and ammonium [(NH4)2SO4; FeSO,+6H,0], with the following results: Three portions of the salt, each weighing 7 grammes, were taken, and there was consumed of the chameleon 190,190.5 and 191 c.c. The mean of the three results, 190.5,- corresponding to 1.0439 grammes of iron, is equal to 14.91 per cent. Theory requires 14.29 per cent of iron in the double sulphate.

The large difference between the amount of iron found in this case, and that required by theory, indicates that the salt tested was not perfectly pure. If the permanganate had been standardized against the double salt, as is recommended by some writers, instead of by means of metallic iron, the result actually obtained would have seemed to be exceedingly exact. The final statement would then have read as follows: Per cent of iron found in the salt = 14.2847; per cent required by theory 14.2857. In case the process be applied to the assay of iron ores,—in which a part of the iron, at least, would usually be in the condition of ferric oxide, dissolve the ore in hot dilute sulphuric acid in a flask, and after solution has been effected, drop a small piece of metallic zinc into the solution, and immediately connect the flask with an apparatus for generating carbonic acid. When the zinc has all been dissolved, remove the stopper from the flask, drop in a fresh bit of zinc, replace the stopper, and, if need be, heat the liquor, to hasten the solution of the zinc. The zinc reduces

any ferric salt, which may be present, to the condition of ferrous salt, while hydrogen is set free. When the whole of the ferric salt has been deoxidized, the hot solution will appear colorless, or in any event will exhibit no yellow tint.

Cool the liquid by immersing the flask in water, as before, without interrupting the current of carbonic acid; transfer to a beaker, dilute with water, and pour in the standard solution of permanganate as long as the latter is decolorized. For every c.c. of permanganate expended there must be grammes of iron in the amount of ore weighed out for the assay.

EXPERIMENT VII. - Estimation of Tin and Lead in ordinary

Solder.

Weigh out four or five grammes of the alloy, put it in a beaker of 300 or 400 c.c. capacity, pour upon it 40 or 50 c.c., of ordinary nitric acid, free from sulphuric acid, and heat the mixture until chemical action ceases. Transfer the mixture of solid oxide of tin and of liquid to a porcelain dish and evaporate to dryness upon a water bath, in order to render the whole of the tin insoluble. Stir up the dry residue with water, collect the oxide of tin upon a filter, and wash it with water until the washings no longer redden litmus paper, or scarcely redden it. Dry the precipitate, ignite it intensely in the crucible, and weigh it.

To determine the lead, add a quantity of dilute sulphuric acid to the filtrate from the oxide of tin, and evaporate the mixture until so much of the nitric acid has been expelled that the odor of nitric acid is no longer perceptible. Dilute the acid residue with water, collect the precipitated sulphate of lead upon a filter, and wash it first with diluted sulphuric acid (in which sulphate of lead is inoluble), and afterwards with water, or better with alcohol.

In one experiment 4.895 grammes of solder yielded 3.2215 grammes of stannic oxide, and 3.4025 grammes of sulphate of lead ; equal, respectively, to 2.534 grammes of tin and 2.324 grammes of lead, or in terms of per cent. 51.76 and 47.47. Total 99.23. In this experiment the sulphate of lead was washed with much dilute sulphuric acid, and finally with a little water. A second experiment, in which the sulphate of lead was washed with a com

paratively small quantity of dilute sulphuric acid, and afterwards with alcohol, gave the following results: Solder taken 5.164 grammes; oxide of tin found 3.4475 grammes. Sulphate of lead found 3.5235 grammes. Equal to 2.712 grammes tin, and 2.407 grammes lead. Or in per cents 52.51 and 46.61. Total, 99.12 A small quantity of iron was observed, but not estimated in the filtrate from the sulphate of lead.

EXPERIMENT VIII.— Assay of Missouri Galena.

Weigh out 2 or 3 grammes or more of the finely powdered ore and place the powder in a tall beaker of 200 or 300 c.c. capacity, together with a smooth lump of metallic zinc. Pour upon the mixed mineral and metal 100 or 150 c.c. of dilute chlorhydric acid, which has been previously warmed to 40° or 50° (C.). Cover the beaker with a piece of window-glass and put it in a moderately warm place.

The chlorhydric acid required for this experiment may be prepared by mixing 1 volume of the ordinary commercial acid with 4 volumes of water. The lump of zinc should be about an inch in diameter by a quarter inch thick; it may be readily obtained by dropping a small quantity of melted zinc from a crucible upon a smooth surface of wood or iron.

Let the zinc and acid act upon the ore during 15 or 20 minutes in order that the decomposition may be complete. Meanwhile, stir the mixture from time to time with a glass rod; and with the wash bottle wash back into the liquid any particles of galena which may be thrown up against the cover or sides of the beaker.

as may

pour out the

When the whole of the galena has been decomposed, be determined by the facts that the liquid has become clear, and that sulphuretted hydrogen is no longer given off, – liquid (but not the precipitate) from the beaker into a six-inch filter of smooth paper, in which a small lump of zinc has been placed, and wash the lead and zinc in the beaker rapidly, by decantation with repeated portions of hot water. Pour all the wash-water into the filter in order to catch any small particles of lead which may be suspended in it, and continue to wash until the

water running from the filter ceases to give an acid reaction with

litmus paper.

As soon as the washing is finished, scrape out the lead from the beaker into a porcelain crucible by means of a glass rod, followed by the wash bottle; taking care to rub off, either with the glass rod or with the finger, any particles of lead which may adhere to the zinc. Wash out the filter into an evaporating dish; remove the fragment of zinc, and add the particles of lead, thus collected, to the contents of the crucible.

Suck up as much water as possible from the crucible, by means of little strips or rolls of filter paper, and finally dry the lead in the crucible at a moderate heat in an atmosphere of ordinary illuminating gas, and weigh it. During the process of drying, the crucible may be placed in a beaker provided with a closely-fitting pasteboard cover, or in an ordinary tin mustard-can. In either event, the cover of this "drying chamber" should be provided with an inlet tube and an outlet tube of glass, both tubes reaching almost to the bottom of the can or beaker. The pasteboard cover of the beaker may readily be made tight enough for the purpose described, by means of a rim or washer of soft paper or cloth.

There is no need of heating the lead in the crucible intensely; it is enough to dry it out of contact with the air.

Since the zinc of commerce is usually contaminated with lead, it will be best to weigh the lump of zinc used for decomposing the galena, both before and after the experiment, having previously ascertained the proportion of lead in the zinc, either by assaying a tolerably large quantity of it by the method now under consideration, or by precipitating the lead as sulphate from a nitric acid solution, as described in Experiment VII.

If the galena to be analyzed is contaminated with a silicious or other insoluble gangue, the metallic lead may be dissolved in dilute nitric acid after weighing, and the insoluble contamination may then be collected upon a filter and weighed by itself. In the case of galenas which contain silver, antimony, copper or other metals precipitable by zinc, the proportion of each may be deter