glass tubing, with an oxygen reservoir. This combustion tube is filled half way from the bent end with granulated cupric oxide, which may conveniently be held in position either by plugs of asbestos or by platinum wire gauze, or by a combination of both. The connection with the oxygen reservoir being then made, the greater part of the tube is heated to redness, with the ordinary precautions, and a stream of oxygen (which is first conducted through a long tube containing caustic potash) is passed over the glowing oxide of copper until the issuing gas ceases, after long bubbling, to cause any turbidity in the bright baryta water. As soon as this point is reached, the portion of the combustion tube preceding the layer of cupric oxide is allowed to cool somewhat, and the tube is now ready to be connected with the absorption apparatus. The clean absorption tube is carefully rinsed with water, and is clamped in front of the furnace in such a manner that its bulb end is somewhat higher than the end to be connected to the combustion tube. Both ends must be provided with convenient stoppers, consisting of short pieces of caoutchouc tubing closed with a small piece of glass rod. The stoppers being removed, air, which is first caused to pass through a tube containing caustic potash, is pumped through the tube for about two minutes, and it is then filled with baryta water as follows-The baryta water (of strength 15 per cent.) is kept in a sufficiently large stock bottle, provided with a caoutchouc stopper, through which pass two bent glass tubes, the long one for syphoning, the shorter, to which a potash tube is attached, being connected with a small hand-bellows. In filling the absorption apparatus, the longer syphon tube is connected with it by means of flexible tubing, and the baryta water is forced over by gentle pressure of the bellows, the bulb end of the absorption apparatus being provided with a potash tube. As soon as the absorption apparatus is half filled, the flow of baryta water is arrested; the ends of the Pettenkoffer tube are immediately closed by its stoppers, and it is now ready for use. By these means the tube is filled with perfectly clear and bright baryta water. The absorption apparatus is now connected with the combustion tube, and the combustion proceeded with. The silver dish containing the water residue having been inserted just behind the copper oxide, it is burnt in a slow current of oxygen, and the carbon dioxide is absorbed and converted into baric carbonate in the absorption tube. In order to filter off and convert the baric carbonate, a funnel and filter are arranged to stand over a beaker containing a layer of caustic potash solution at the bottom, the whole being covered by a bell jar, which itself stands in a layer of caustic potash solution. The

mouth of the bell jar, which is immediately over the funnel, is closed by a thick caoutchouc cap with two narrow openings, one of which is provided with a caustic potash tube. (Soda lime apparently answers equally well.) The other, which is temporarily stoppered, contains a straight glass tube, placed immediately over the filter so that, after the whole arrangement has been left some time to itself, in order that all enclosed air may be free from CO2, direct connection may be made with the Pettenkoffer tube by means of flexible tubing sufficiently long to admit of some slight freedom of action. Filtration may thus be carried on without danger of CO, being introduced from the atmosphere, the additional precaution being taken of compelling all air which passes through the Pettenkoffer tube during this process of filtration, to pass through a tube containing caustic potash attached to the tube itself. The washing of the precipitate in the tube and on the filter is effected almost entirely with boiling water, which has been previously saturated with carbonate of barium [solubility 1 in 15,000], but finally with a small quantity of boiling distilled water. After complete washing, the tube is disconnected, and the filter ultimately rinsed round, while still under the bell jar, by means of the long tube already mentioned, and which, when not clamped, may be moved freely in all directions. The bell jar is then removed, and the precipitate is rapidly washed together into the bottom of the filter.

The Pettenkoffer tube, which may contain minute particles of baric carbonate not removed by the washing, is rinsed twice with small quantities of dilute pure hydrochloric acid (about 1 in 50), and finally with distilled water: the rinsings are poured on to the filter on which the greater mass of baric carbonate is already collected. The filter is further washed with dilute hydrochloric acid, and finally with distilled water and the whole of the solution of baric chloride so formed is carefully collected in a small beaker. The quantity of such solution need not exceed 50

cc.

This solution of chloride of barium has next to be evaporated, which is best done in a platinum vessel on the water-bath. It is then transferred, when greatly decreased in bulk, to a much smaller platinum dish, weighing about 5 grms., and finally evaporated to dryness after the addition of a few drops of pure sulphuric acid. The dish and its contents have then to be ignited, the residue moistened with a drop of nitric acid and redried, and the whole re-ignited and weighed to conclude the operation. The amount of carbon present is obtained by dividing the weight of the baric sulphate by 19.4.

Nephalometric Method. This ingenious method we also owe

to Dupré and Hake. The carbonic acid resulting from the combustion of an organic residue is passed into perfectly pure clear solution of basic lead acetate, and the turbidity produced is imitated by known weights of CO2; in fact, the operation is a colour method conducted on the same principle as "Nesslerising," with this important difference, that no success will be obtained unless there are special precautions taken to prevent the contamination of the solutions by the breath and air, &c.

The Author's Method of Estimating Minute Quantities of Carbon. The writer in 1881 made some very extended, and as yet unpublished, experiments on the estimation of organic carbon in the air, and the method was afterwards extended to all estimations of minute quantities of carbon dioxide, in which the balance from the small quantity present was likely to give less accurate results than measurement as a gas.

The method, briefly, consists of a suitable arrangement by which the carbon dioxide is absorbed in a solution of caustic potash, and ultimately evolved as gas. The arrangement for evolving the carbon dioxide absorbed as a gas is the same as that described at page 99, and is simplicity itself; in fact, the materials for the estimation merely consist of a flask with a caoutchouc cork, rod, and Bunsen's valve, an ordinary eudiometer and mercury trough, and lastly, a little test tube with sufficient acid to more than neutralise the potash. The solution in the flask is boiled briskly until all air is expelled, then the beak of the tube is put under the eudiometer, and the glass rod lifted up a little to allow the test tube to fall. A brisk effervescence takes place, and the whole of the gas as pure CO, is boiled out into the measuring tube. At first the author always proved its purity by again absorbing it with KHO, but as the result was always perfect absorption, this was abandoned. Of course, the gas is reduced to standard pressure and temperature.

That this method is applicable to the determination of the minute quantity of carbon in a water residue is obvious.*

Mineral Analysis of Water.-Ordinary drinking water holds dissolved but few saline matters, and when an analyst has determined chlorine, nitrates, sulphates, phosphates, and carbonates, and also lime and magnesia and alkalies, he will

*Other methods of determining organic elements have been proposed; one of the most recent is a proposition to estimate nitrogen as NH3, by first treating the water with the zinc copper couple, expelling the ammonia thus produced by the decomposition of nitrates; boiling the solution to dryness with caustic soda in a copper flask, ultimately raising the heat to incineration; and condensing the products formed, and Nesslerising. - See "On a Method of Estimating Organic Nitrogen," by William BettelChemical News, Jan. 27, 1882.

usually find, on adding the several amounts together, that he gets numbers very nearly equal to the solid saline residue. An excellent method of approximately estimating the various saline constituents of a water is to evaporate down to dryness a known quantity, then to treat the residue with a little hot water, which will dissolve all the soluble salts out, but leave insoluble carbonate of lime and silica. In the soluble portion, the soluble lime, the magnesia, and the alkalies are determined; the chlorides, sulphates, and nitrates, are estimated on the unconcentrated water by the processes already detailed. It is also always open to make the analysis in the old-fashioned way, that is, to evaporate down a large quantity of water, to separate the silica by treatment of the ash or residue with hydrochloric acid, and after separation of the silica to divide the solution into three or four quantities, in which sulphuric acid, lime, magnesia, &c., are determined by the ordinary methods.

P

IV. BIOLOGICAL METHODS.

$320. A. Microscopical Appearances. To make a microscopical examination of water, it is necessary to collect the sediment or deposit which falls to the bottom of the vessel in which the water stands. The most convenient way of doing this is to use the author's tube (fig. 49), which holds a little more than a litre. The little glass cell C is adjusted to the pipette-like end, the rod is removed, and after introduction of the water the tube is covered and set aside for twenty-four hours. At the end of that time any deposit will have collected in the glass cap. On now carefully inserting the rod-like stopper, the cap or cell can be removed with great ease, and its contents submitted to microscopical examination. With very pure waters merely a little sand or formless débris collects in the cap, and there is no life. If, however, in the first place eight or ten gallons are allowed to deposit in a capacious vessel, most of the water run off, and then the last litre rinsed into the tube, in nearly every case there may be a few life-forms and sufficient matter collected to give definite results. It need scarcely be said that an opinion must not be formed upon a microscopical examination without taking into account the amount of water from which the sediment has been collected, and a definite quantity should be generally agreed on by analysts. As for the present writer's practice when a gallon of water throws down only mineral

C

Fig. 49.

matters and a little scanty unrecognisable débris without lifeforms-although kept for at least twenty-four hours at a temperature of from 15° to 17°, and exposed to the daylighthe considers it, in a microscopical sense, pure. The contents of the little cap may be conveniently examined as follows:By the aid of a pipette one or more drops are placed under the microscope without any preparation, others are divided upon several slides, and treated with (1.) dilute iodine solution, which will colour starch cells blue; (2.) aniline violet*—this is par excellence the staining fluid for bacteria; (3.) solution of carmine in glycerine and alcohol, which colours the nuclei of cells red. It will be advisable to work at first with a low power, so as to get a general idea of the nature of the larger and more opaque particles, and then afterwards investigate with the highest powers which the analyst possesses. In using low powers it is not well to place any covering glass over the drop, especially if a binocular be employed, for the convexity interferes in no way with the definition. The matters likely to be found in a water residue are

1. Lifeless Forms.

1. Mineral Matters, especially sand, clay, and not unfrequently fine spicula of glass derived from the glass pipette, &c.

2. Vegetable Matters.-In shallow pools, in rivers, reservoirs, and, in fact, all open waters, the microscopist seldom fails to find vegetable débris in the shape of dotted ducts, spiral vessels, parenchymatous cells, bits of cuticle with the hair still adhering, the down of seeds, roots of duckweed, bits of chara, &c. It depends on the amount as to what conclusions are to be drawn ; but this is certain, that a water showing these matters is not likely to be from a deep spring, but one over which the atmosphere more or less freely plays.

3. Dead Animal Matters-(a.) Purely Animal, such as hairs from domestic or wild animals, striped muscular tissue, the scales of moths, butterflies, or other lepidoptera, eggs of entozoa (which, of course, may, for aught we know, be living).

(b.) Human Débris.-Human hair, human epithelium. (c.) Manufactured Matters.-Wool, silk, &c. All animal matters, whether derived from insect, human, or domestic animal life cannot be considered a favourable indication; and even the presence of cotton, silk, hemp, and the like, though

* The common aniline violet ink answers very well.