ANALYSIS OF A VOLATILE ORGANIC LIQUID.

19

and blown into a small bulb about as big as a good-sized pea ; this is to be cut off, leaving a capillary neck of about two inches (5 centimetres) long. Having made a sufficient number of these

FIG. 380.

a

small bulbs, two of them, which it has been ascertained will freely enter the combustion-tube, are to be accurately weighed; a little of the liquid to be analysed is to be put into a small tube, and the capillary neck of the bulbs inverted into the liquid; the bulbs are then to be warmed by the flame of a spirit-lamp, so that on cooling they shall be about three-fourths filled with the liquid. The necks are now to be sealed by the blowpipe-flame, and the bulbs again weighed; the increase of weight gives the quantity of the liquid which has entered, and which is to be analysed. The oxide of copper having been heated, and allowed to cool with the usual precautions, about an inch and a half (36 millimetres) of the retort is to be filled with pure oxide; next a file is to be drawn across the capillary neck of one of the bulbs, and the bulb having been placed in the retort, the neck is to be broken by pressure against the glass, and the broken portion of the neck to be dropped in with the bulb; dried oxide of copper is then to be poured in to the depth of two inches more; the second bulb is to be introduced in the same manner as the first, and the tube is to be filled up with oxide; after which it is to be corked and struck smartly upon its side on the table, to secure free air-way. The combustion tube is now to be adjusted in the furnace, and the first half of the tube is to be gradually heated; when this is redhot, the liquid must be volatilized by cautiously approximating a piece of ignited charcoal to the part of the tube where the bulb nearest the open end of the retort lies, taking especial care not to raise the temperature too rapidly: when by degrees all the liquid in the first bulb has been expelled, the other is to be proceeded with in like manner; the whole tube is finally to be heated carefully, and the gases are to be swept out of the apparatus by air or oxygen in the manner already described.

If the liquid be not volatile it may be weighed in a small tube the weight of which, when empty, is known, and after the tube with its contents has been introduced into the retort tube, the liquid is distributed over the sides of this tube; dry oxide of

20

ANALYSIS OF A BODY CONTAINING NITROGEN.

copper is then filled in, and the combustion is proceeded with in the usual way.

Substances which contain a great excess of carbon sometimes escape complete combustion by the foregoing process; when this is feared, the tube is connected with a receiver filled with oxygen, upon the plan of Dumas, already mentioned; or some finely pulverized, carefully dried potassic chlorate may be mixed with about four times its weight of oxide of copper, and the portion of the retort near a (fig. 379) may be filled with it for about an inch; the tail-like prolongation may in this case be dispensed with : at the close of the operation, instead of sucking air through the apparatus, heat is to be very cautiously applied to the chlorate e; by this means oxygen is evolved; this burns the last traces of carbon and displaces the gas and aqueous vapour contained in the tubes. The heat must not be too suddenly applied, for if it be, a portion of the chlorate is apt to be carried forward mechanically, a circumstance which constitutes the chief objection to its use.

In some cases chromate of lead may be advantageously substituted for oxide of copper, where chlorine, bromine, or iodine is present, as the chloride, bromide, or iodide of copper are somewhat volatile, and would be carried by the current of gas into the condensing tubes. The chromate is also useful with substances difficult of combustion, since by a bright red heat alone it gives off a portion of its oxygen. It is easily prepared by

a

precipitating the potassic chromate or dichromate with solution of acetate of lead: it should be well washed and heated to incipient fusion before it is used for analysis. Chromate of lead has the advantage of being much less hygroscopic than the oxide of copper. In some cases the body to be burned may be mixed with potassic dichromate (K,Er,,) which parts with oxygen by heat more freely than the lead salt.

3. Analysis of a Body containing Nitrogen.

(1046) The presence of nitrogen in an organic compound is ascertained by mixing a small quantity of it, with a portion of solid caustic potash, and heating it in a small test tube: vapours of ammonia will be evolved if nitrogen be present. Two separate analyses are in this case required: the first, to discover the proportion of carbon and hydrogen; and the second, expressly for the nitrogen. When bodies containing nitrogen are burned with oxide of copper, a variable proportion of the lower oxides of nitrogen is formed, which being retained by the chloride of calcium, or by

METHOD OF VARRENTRAPP AND WILL.

the potash, would render the analysis incorrect.

21

A precaution is

therefore employed which renders it necessary to make use of a retort-tube rather longer than common; this tube is to be charged as usual to within six inches (15 centimetres) of the opening, and then filled up with clean copper turnings; the apparatus is to be arranged as before directed, the copper turnings brought to full redness, and the analysis is to be cautiously conducted in the ordinary manner. As the oxides of nitrogen pass slowly over the ignited copper they are decomposed, the oxygen combining with the copper, while pure nitrogen escapes: the quantity of carbon and hydrogen is determined exactly in the manner previously described.

(1047) Method of Varrentrapp and Will.-The most accurate method for ascertaining the proportion of nitrogen is that devised by Varrentrapp and Will; and the fundamental fact upon which this method is based consists in the observation of GayLussac, that when azotised matters are heated with a large excess of potassic or sodic hydrate, the whole of the nitrogen is expelled in the form of ammonia. In order to render it available for the purposes of analysis, the subjoined precautions are requisite.

A mixture of two parts of quicklime and one of sodic hydrate is prepared by slaking some well-burned lime with the necessary quantity of a solution of soda; the whole is evaporated to dryness, and ignited; the dry mass is pulverized as quickly as possible, and then transferred to well-closed bottles, in order to exclude carbonic acid and moisture. When an analysis is required, the substance under examination is very intimately mixed in a warm mortar with a portion of this alkalized lime, or sodalime as it is frequently termed, instead of with cupric oxide: the accidental presence of a little moisture after the weight of the material for analysis has been accurately determined, is of no consequence in this case.

Having introduced the mixture into the retort tube, it is better to plug the aperture loosely with a few fibres of asbestos, (which has been ignited just before) to prevent any mechanical transport of the mixture into the apparatus through which the gases are made to pass; on applying heat to the combustion-tube in the ordinary way, and with the usual precautions, the substance is decomposed, and the whole of the nitrogen escapes as ammonia. The drying-tube and potash apparatus are dispensed with, and the ammonia is collected in a bulb-tube of the form represented at f, fig. 381, which is connected, air-tight, by means of a good cork, with the retort-tube a; the bulbs ƒ having been previously charged with hydrochloric acid of sp. gr. 1'I.

22

ANALYSIS OF A COMPOUND CONTAINING NITROGEN.

Pure hydrochloric acid is easily procured for this purpose by diluting the ordinary acid of the shops till it has a sp. gr. of 1'1, and distilling it in glass vessels; the first eighth may be rejected, and the distillation proceeded with until three-fourths of the acid

FIG. 381.

d

d

employed have passed over. It is better for the operator always to rectify his own acid, in order to be quite sure of the absence any trace of ammonia.

of

The tube connecting the bulbs f should be somewhat larger in diameter than that of the ordinary potash apparatus, in order to allow the liquid to be poured out readily. When the operation is complete, absorption will take place, and the liquid will rise in the bulb nearest the fire; at this moment the tail of the combustiontube should be nipped off, and air carefully drawn through When the combustion is

the apparatus in the usual way. completed, the contents of the bulb-tube are to be emptied into a small evaporating dish, and the bulbs washed out first with a mixture of alcohol and ether, and afterwards several times with water; some solution of platinic chloride is to be added, and the whole evaporated to dryness in a water-bath or a chloride of calcium bath; the residue when dry is to be digested with a mixture of two parts of alcohol, sp. gr. 0.833, and one of ether, which dissolves the excess of platinic chloride, and leaves the double chloride of platinum and ammonium in a crystalline form. This must now be brought upon a weighed filter, and washed repeatedly with the mixture of two parts of alcohol and one of ether until nothing further is taken up; the precipitate and filter must be dried by a heat of 212°, and the weight accurately observed; 220 52 parts of the ammonia-chloride of platinum are equivalent to 14 of nitrogen, or ICO parts to 6.348 of nitrogen. Instead of collecting the ammonia in hydrochloric acid, it may be condensed in a certain volume of dilute sulphuric acid of known strength (about 150 grains, or 10 grammes, of an acid of which 100 parts will neutralize 25 parts of weight of ammonia, will answer the purpose). The acid liquor is then to be washed out of the

bulbs f, diluted, and carefully neutralized by a solution of caustic soda, or by a solution of lime in dilute syrup, the strength of which has been accurately graduated.

(1048) Method of Dumas.-The foregoing method of determining nitrogen answers for nearly all cases excepting those in which this element occurs in the form of cyanogen or nitric acid, when it must be determined by volume, and its weight thence deduced. For this purpose the process of Dumas is the most trustworthy :-A retort-tube of about 30 inches (75 centimetres) long is employed, not drawn out into a tail, but sealed with a rounded extremity; about 3 inches (75 millimetres) of the tube are filled with hydro-sodic carbonate (NaHCO3) and then the mixture with oxide of copper is added, and covered as usual with a layer of pure oxide; beyond this, the last 3 or 4 inches of the tube are filled with clean copper turnings, as already directed, with a view of decomposing any of the oxides of nitrogen which may be formed. The retort-tube is then connected with a threelegged apparatus of brass or copper (380, b), one limb of which is furnished with a stopcock, c. The connexion with the retorttube a is best made by passing a piece of glass tubing through a cork fitting accurately into the mouth of the tube a, and connecting the brass apparatus to this small glass tube by a caoutchouc connector, d; the third limb is fastened by a similar joint to a glass tube e, bent at right angles near one end, with a straight portion upwards of 30 inches (760 mm.) long, the other extremity of which is turned up at an acute angle for the convenience of safely delivering the gas this tube is placed with its long portion in a vertical direction and with its lower upturned extremity dipping into a small mercurial trough; the stopcock tube, c, is connected with an exhausting syringe, by which a vacuum is produced; the apparatus is left for half an hour to ascertain that all the joints are tight if the mercury after this lapse of time still stands at the same level, the experiment may be proceeded with. A moderate heat is then applied by a spirit lamp at the end of the retort containing the carbonate; by this means carbonic anhydride is set free, and displaces the last portions of air; the exhaustion and the disengagement of gas are repeated alternately three or four times, care being taken to leave an amount of the carbonate undecomposed sufficient to renew this expulsive process at the termination of the experiment. The stopcock, c, is now closed, the air-pump is removed, and a graduated jar containing some solution of potash is inverted in the mercury over the recurved extremity of the long glass tube. The copper turnings are then

: