(fig. 2). In these the gas first passes into the largest bulb, from which it expels the potassic hydrate, and then passes in single bubbles through the other bulbs, and being thus necessarily in long contact with the potash, is very completely absorbed. The modification of this apparatus known as Geissler's bulbs is also much used. Instead of

potash solution, soda lime is frequently used to absorb carbonic anhydride. This is rubbed in a porcelain mortar until a mixture of fine powder with pieces the size of coarse millet remains. U-shaped light glass tubes (fig. 3), which contain in one limb, between two

U

FIG. 3.

plugs of cotton wool a and b, coarsely powdered calcic chloride, are completely filled up with the soda lime; by tapping the tube with both limbs placed horizontally the separation of the fine from the coarse powder is effected. The latter occupies the upper part of the cross-section, and forms a passage through which the gases pass, and from which the finer powder absorbs the carbonic anhydride. In cases where soda lime tubes are employed a small bulb tube g, containing a drop of sulphuric acid, is usually placed between the calcic chloride and soda lime. tubes; it serves to show the rate at which the combustion is progressing.

Combustions are conducted in many different ways. Two methods will be described, of which the oldest and, as regards apparatus, the simplest will always be connected with the name of Liebig. The solid substance to be analysed, having been dried and previously finely powdered, is weighed (generally 0-2 to 0.3 gram is taken) and mixed with finely powdered cupric oxide in a warm porcelain mortar. The cupric oxide must have been ignited just before the operation, transferred whilst still hot to a wide glass tube, which is then carefully closed, and cooled so far as not to alter the organic substance when mixed with it. A tube of difficultly fusible glass must have been previously prepared for the reception of the mixture. This combustion tube (fig. 4) is drawn out at one end (over the blowpipe) to a tail at an obtuse angle, then carefully cleaned and dried. The end is filled

up to c with pure cupric oxide; then follows the mixture of the organic substance and cupric oxide to b; from there to a is the cupric oxide used in rinsing out the mortar. The tube is then filled up to within an inch of its mouth with pure cupric oxide, and finally a loose plug of asbestos is inserted. The length of the layer of cupric oxide must vary with the combustibility or volatility of the organic body, so that for such bodies as give on heating a large volume of combustible vapours and gases it must be made longer than it otherwise

FIG. 4.

would. The tube is then held horizontally and tapped gently, by which means a canal is formed along the top of the contents, along which the products of combustion can pass. The tube so prepared is then laid in a combustion furnace; the chloride of calcium tube is fitted on by means of a good cork or caoutchouc plug, and in its turn is connected to the potash bulbs by means of a piece of caoutchouc tubing. The whole apparatus is shown in fig. 5. The fore-part of the combustion tube is now heated, and as soon as the cupric oxide is glowing the heating gradually carried backward, so that the mixture of substance and cupric oxide is raised to the temperature necessary for combustion.

In order to prevent any condensation of unburnt volatile decomposition products in the tail, that is kept heated from the boundary of the pure cupric oxide. When, at length, the whole tube is brought to a red heat, and no more gas bubbles are seen in the potash bulbs, the combustion is finished. The tail of the combustion tube is broken off at c and air drawn slowly through, so as to sweep all products of combustion from the tube into the absorption apparatus. At the finish the apparatus is disconnected and the absorption tubes weighed. As carbonic anhydride contains of its weight of carbon, and water of its weight of hydrogen, by multiplying the increase in weight of the respective tubes by these figures, the amount of carbon and hydrogen in the substance employed will be obtained.

9

11. When readily volatile liquids have to be analysed in this apparatus they are placed in small previously weighed bulbs, the point fused, and the bulb again weighed, in order to ascertain the weight of the substance taken. The combustion tube is then filled for a quarter of its length with freshly ignited cupric oxide, the bulb, with its point broken, then dropped in, and the tube filled up with cupric oxide. The apparatus is arranged as in fig. 5; the front and back parts of the tube first brought to redness, and then that part where the bulb is gradually heated, so that the vapour of the liquid shall slowly pass over the glowing oxide and be completely burnt.

Non-volatile liquids, fats, &c., are caused to spread over the sides of the combustion tube, which is then filled up with cuprie oxide.

12. In many cases it is necessary to make some alterations in the method above described. Some organic substances cannot be completely burnt with cupric oxide, either from not admitting of sufficiently intimate admixture or from their yielding much carbon on beating, which cannot come into contact with sufficient cupric oxide for its complete combustion. In these cases plumbic chromate is employed instead of cupric oxide. Immediately before use it must be strongly heated in a porcelain basin. It contains more oxygen in an equal volume than cupric oxide, fuses on heating, and gives off oxygen on

FIG. 6.

a

strong ignition. Plumbic chromate is of use also in the analysis of bodies containing sulphur and halogens, as it completely retains them; whilst with cupric oxide sulphurous anhydride or free halogen is frequently obtained, which being soluble in potassic hydrate, would completely spoil the result for carbon. Nitrogenous bodies on combustion frequently give off a portion of their nitrogen as nitric oxide, which, when mixed with air, is absorbed by potash, and would also bring the result too high. In this case a layer of a mixture of potassic chromate and manganic oxide is placed in the front part of the combustion tube, and kept very gently heated during the combustion, whereby all nitrous fuses are effectually removed.

13. Much saving of time, together with simplification of manipulation and greater certainty in the results of carbon and hydrogen estimation, is obtained by combustion in a stream of oxygen in a tube open at both ends (fig. 6). Between the asbestos plugs d and e a layer of granulated cupric oxide is placed; the ends of the tube are closed by bored corks, in one of which is placed a tube for the introduction of pure air or oxygen, whilst the other serves to connect with the absorption apparatus. The combustion tube is heated to redness in the furnace, whilst a stream of dry air is passed through; by this means the whole apparatus and its contents are obtained thoroughly dry and ready for immediate use. The absorption tubes for water and carbonic anhydride having been attached, the weighed substance contained in a porcelain or platinum boat e is pushed in (volatile liquids.

in bulbs as described), and finally a short glass rod with fused edges b, which nearly fills the tube. The object of this latter is to cause an acceleration of the gas stream at that point, and so prevent any chance of the vapours given off by the burning body travelling backwards, as they might otherwise do, the stream of gas being necessarily slow, to ensure complete combustion. The whole apparatus shown in fig. 7 being arranged, and the substance in its boat placed behind the red-hot cupric oxide, heat is next applied at the place where the glass rod is (b, fig. 6) and then slowly advanced to the boat. At the commencement purified air may be employed, the oxidation being mainly effected at the expense of the cupric oxide; later oxygen is turned on from a second gas-holder, and the complete oxidation of any carbon left behind is then effected; any reduced copper is also re-oxidised. Both air and oxygen are purified by passing through potash and calcic chloride or sulphuric acid. The stream of oxygen is continued until its presence can be distinctly observed at the absorption apparatus; it is then displaced by a current of air. The cooled absorption tubes and the boat, if any ash is left, are then weighed. The combustion tube can then be employed for a fresh analysis, as it is in a state of complete readiness, i.e. completely free from water and carbonic anhydride and charged with pure cupric oxide. In the case of nitrogenous bodies potassic chromate and manganic dioxide are placed in the anterior portion of the tube. With substances containing halogens a roll of silver must be placed in front of the cupric oxide; by this every trace of halogen is absorbed. Bodies containing sulphur must be burnt with plumbic chromate, or if cupric oxide and oxygen be used the gas stream must be passed over pure plumbic peroxide, heated to but little above 100°, by which all sulphurous anhydride is removed.

Estimation of Nitrogen.

14. The determination of the amount of nitrogen in organic bodies is effected either by measuring the nitrogen in the free state and calculating the weight from the volume, or by conversion into ammonia.

The first method can be applied to all organic bodies; the second does not give satisfactory results in those cases where the nitrogen is in union with the oxygen of the organic substance (e.g. in nitro compounds).

15. Estimation of Nitrogen by Volume.-By the combustion of nitrogenous organic bodies with cupric oxide there results, as already mentioned, carbonic anhydride, water, and nitrogen, and probably some quantity of nitric oxide, this latter being reduced to nitrogen by passing over red-hot metallic copper. On cooling the products of combustion, the water is separated in the liquid form; the carbonic anhydride (and also any sulphurous anhydride or halogens) is readily absorbed by a solution of an alkaline hydrate, so that pure nitrogen remains, and can be estimated by measurement. The single difficulty of this method is the complete removal from the apparatus of the indifferent gases of the air before the commencement of the combustion.

This can be effected by complete exhaustion by the mercurial