CHAPTER IX

COMPOUNDS OF CARBON WITH OXYGEN AND NITROGEN

Carbonic anhydride (or carbonic acid or carbon dioxide, CO₂) was the first of all gases distinguished from atmospheric air. Paracelsus and Van Helmont, in the sixteenth century, knew that on heating limestone a particular gas separated, which is also formed during the alcoholic fermentation of saccharine solutions (for instance, in the manufacture of wine); they knew that it was identical with the gas which is produced by the combustion of charcoal, and that in some cases it is found in nature. In course of time it was found that this gas is absorbed by alkali, forming a salt which, under the action of acid, again yields this same gas. Priestley found that this gas exists in air, and Lavoisier determined its formation during respiration, combustion, putrefaction, and during the reduction of the oxides of metals by charcoal; he determined its composition, and showed that it only contains oxygen and carbon. Berzelius, Dumas with Stas, and Roscoe, determined its composition, showing that it contains twelve parts of carbon to thirty-two of oxygen. The composition by volume of this gas is determined from the fact that during the combustion of charcoal in oxygen, the volume remains unchanged; that is to say, carbonic anhydride occupies the same volume as the oxygen which it contains—that is, the atoms of the carbon are, so to speak, squeezed in between the atoms of the oxygen. O₂ occupies two volumes and is a molecule of ordinary oxygen; CO₂ likewise occupies two volumes, and expresses the composition and molecular weight of the gas. Carbonic anhydride exists in nature, both in a free state and in the most varied compounds. In a free state it is always contained (Chapter V.) in the air, and in solution is in all kinds of water. It is evolved from volcanoes, from mountain fissures, and in some caves. The well-known Dog grotto, near Agnano on the bay of Baiæ, near Naples, furnishes the best known example of such an evolution. Similar sources of carbonic anhydride are also found in other places. In France, for instance, there is a well-known poisonous fountain in Auvergne. It is a round hole, surrounded with luxurious vegetation and constantly evolving carbonic anhydride. In the woods surrounding the Lacher See near the Rhine, in the neighbourhood of extinct volcanoes, there is a depression constantly filled with this same gas. The insects which fly to this place perish, animals being unable to breathe this gas. The birds chasing the insects also die, and this is turned to profit by the local peasantry. Many mineral springs carry into the air enormous quantities of this gas. Vichy in France, Sprüdel in Germany, and Narzan in Russia (in Kislovodsk near Piatigorsk) are known for their carbonated gaseous waters. Much of this gas is also evolved in mines, cellars, diggings, and wells. People descending into such places are suffocated. The combustion, putrefaction, and fermentation of organic substances give rise to the formation of carbonic anhydride. It is also introduced into the atmosphere during the respiration of animals at all times and during the respiration of plants in darkness and also during their growth. Very simple experiments prove the formation of carbonic anhydride under these circumstances; thus, for example, if the air expelled from the lungs be passed through a glass tube into a transparent solution of lime (or baryta) in water a white precipitate will soon be formed consisting of an insoluble compound of lime and carbonic anhydride. By allowing the seeds of plants to grow under a bell jar, or in a closed vessel, the formation of carbonic anhydride may be similarly confirmed. By confining an animal, a mouse, for instance, under a bell jar, the quantity of carbonic acid which it evolves may be exactly determined, and it will he found to be many grams per day for a mouse. Such experiments on the respiration of animals have been also made with great exactitude with large animals, such as men, bulls, sheep, &c. By means of enormous hermetically closed bell receivers and the analysis of the gases evolved during respiration it was found that a man expels about 900 grams (more than two pounds) of carbonic anhydride per diem, and absorbs during this time 700 grams of oxygen.[1] It must be remarked that the carbonic anhydride of the air constitutes the fundamental food of plants (Chapters III., V., and VIII.) Carbonic anhydride in a state of combination with a variety of other substances is perhaps even more widely distributed in nature than in a free state. Some of these substances are very stable and form a large portion of the earth's crust. For instance, limestones, calcium carbonate, CaCO₃, were formed as precipitates in the seas existing previously on the earth; this is proved by their stratified structure and the number of remains of sea animals which they frequently contain. Chalk, lithographic stone, limestone, marls (a mixture of limestone and clay), and many other rocks are examples of such sedimentary formations. Carbonates with various other bases—such as, for instance, magnesia, ferrous oxide, zinc oxide, &c.—are often found in nature. The shells of molluscs also have the composition CaCO₃ and many limestones were exclusively formed from the shells of minute organisms. As carbonic anhydride (together with water) is produced during the combustion of all organic compounds in a stream of oxygen or by heating them with substances which readily part with their oxygen—for instance, with copper oxide—this method is employed for estimating the amount of carbon in organic compounds, more especially as the CO₂ can be easily collected and the amount of carbon calculated from its weight. For this purpose a hard glass tube, closed at one end, is filled with a mixture of the organic substance (about 0·2 gram) and copper oxide. The open end of the tube is fitted with a cork and tube containing calcium chloride for absorbing the water formed by the oxidation of the substance. This tube is hermetically connected (by a caoutchouc tube) with potash bulbs or other weighing apparatus (Chapter V.) containing alkali destined to absorb the carbonic anhydride. The increase in weight of this apparatus shows the amounts of carbonic anhydride formed during the combustion of the given substance, and the quantity of carbon may be determined from this, because three parts of carbon give eleven parts of carbonic anhydride.

For the preparation of carbonic anhydride in laboratories and often in manufactories, various kinds of calcium carbonate are used, being treated with some acid; it is, however, most usual to employ the so-called muriatic acid—that is, an aqueous solution of hydrochloric acid, HCl—because, in the first place, the substance formed, calcium chloride. CaCl₂, is soluble in water and does not hinder the further action of the acid on the calcium carbonate, and secondly because, as we shall see further on, muriatic acid is a common product of chemical works and one of the cheapest. For calcium carbonate, either limestone, chalk, or marble is used.[2]

CaCO₃ + 2HCl = CaCl₂ + H₂O + CO₂.

The nature of the reaction in this case is the same as in the decomposition of nitre by sulphuric acid; only in the latter case a hydrate is formed, and in the former an anhydride of the acid, because the hydrate, carbonic acid, H₂CO₃, is unstable and as soon as it separates decomposes into water and its own anhydride. It is evident from the explanation of the cause of the action of sulphuric acid on nitre that not every acid can be employed for obtaining carbonic anhydride; namely, those will not set it free which chemically are but slightly energetic, or those which are insoluble in water, or are themselves as volatile as carbonic anhydride.[3] But as many acids are soluble in water and are less volatile than carbonic anhydride, the latter is evolved by the action of most acids on its salts, and this reaction takes place at ordinary temperatures.[4]

For the preparation of carbonic anhydride in laboratories, marble is generally used. It is placed in a Woulfe's bottle and treated with hydrochloric acid in an apparatus similar to the one used for the production of hydrogen. The gas evolved carries away through the tube part of the volatile hydrochloric acid, and it is therefore necessary to wash the gas by passing it through another Woulfe's bottle containing water. If it be necessary to obtain dry carbonic anhydride, it must be passed through chloride of calcium.[5]

Carbonic anhydride may also be prepared by heating many of the salts of carbonic acid; for instance, by heating magnesium carbonate, MgCO₃ (e.g., in the form of dolomite), the separation is easily effected, particularly in the presence of the vapours of water. The acid salts of carbonic acid (for instance, NaHCO₃, see further on) readily and abundantly give carbonic anhydride when heated.

Carbonic anhydride is colourless, has a slight smell and a faint acid taste; its density in a gaseous state is twenty-two times as great as that of hydrogen, because its molecular weight is forty-four.[6] It is an example of those gaseous substances which have been long ago transformed into all the three states. In order to obtain liquid carbonic anhydride, the gas must be submitted to a pressure of thirty-six atmospheres at 0°.[7] Its absolute boiling point = +32°.[8] Liquid carbonic anhydride is colourless, does not mix with water, but is soluble in alcohol, ether, and oils; at 0° its specific gravity is 0·83.[8 bis] The boiling point of this liquid lies at -80°—that is to say, the pressure of carbonic acid gas at that temperature does not exceed that of the atmosphere. At the ordinary temperature the liquid remains as such for some time under ordinary pressure, on account of its requiring a considerable amount of heat for its evaporation. If the evaporation takes place rapidly, especially if the liquid issues in a stream, such a decrease of temperature occurs that a part of the carbonic anhydride is transformed into a solid snowy mass. Water, mercury, and many other liquids freeze on coming into contact with snow-like carbonic anhydride.[9] In this form carbonic anhydride may be preserved for a long time in the open air, because it requires still more heat to turn it into a gas than when in a liquid state.[9 bis]

The capacity which carbonic anhydride has of being liquefied stands in connection with its considerable solubility in water, alcohol, and other liquids. Its solubility in water has been already spoken of in the first chapter. Carbonic anhydride is still more soluble in alcohol than in water, namely at 0° one volume of alcohol dissolves 4·3 volumes of this gas, and at 20° 2·9 volumes.

Aqueous solutions of carbonic anhydride, under a pressure of several atmospheres, are now prepared artificially, because water saturated with this gas promotes digestion and quenches thirst. For this purpose the carbonic anhydride is pumped by means of a force-pump into a closed vessel containing the liquid, and then bottled off, taking special means to ensure rapid and air-tight corking. Various effervescing drinks and artificially effervescing wines are thus prepared. The presence of carbonic anhydride has an important significance in nature, because by its means water acquires the property of decomposing and dissolving many substances which are not acted on by pure water; for instance, calcium phosphates and carbonates are soluble in water containing carbonic acid. If the water in the interior of the earth is saturated with carbonic acid under pressure, the quantity of calcium carbonate in solution may reach three grams per litre, and on issuing at the surface, as the carbonic anhydride escapes, the calcium carbonate will be deposited.[10] Water charged with carbonic anhydride brings about the destruction of many rocky formations by removing the lime, alkali, &c., from them. This process has been going on and continues on an enormous scale. Rocks contain silica and the oxides of various metals; amongst others, the oxides of aluminium, calcium, and sodium. Water charged with carbonic acid dissolves both the latter, transforming them into carbonates. The waters of the ocean ought, as the evolution of the carbonic anhydride proceeds, to precipitate salts of lime; these are actually found everywhere on the surface of the ground in those places which previously formed the bed of the ocean. The presence of carbonic anhydride in solution in water is essential to the nourishment and growth of water plants.

Although carbonic anhydride is soluble in water, yet no definite hydrate is formed;[11] nevertheless an idea of the composition of this hydrate may be formed from that of the salts of carbonic acid, because a hydrate is nothing but a salt in which the metal is replaced by hydrogen. As carbonic anhydride forms salts of the composition K₂CO₃, Na₂CO₃, HNaCO₃, &c., therefore carbonic acid ought to have the composition H₂CO₃—that is, it ought to contain CO₂ + H₂O. Whenever this substance is formed, it decomposes into its component parts—that is, into water and carbonic anhydride. The acid properties of carbonic anhydride[11 bis] are demonstrated by its being directly absorbed by alkaline solutions and forming salts with them. In distinction from nitric, HNO₃, and similar monobasic acids which with univalent metals (exchanging one atom for one atom of hydrogen) give salts such as those of potassium, sodium, and silver containing only one atom of the metal (NaNO₃, AgNO₃), and with bivalent[12] metals (such as calcium, barium, lead) salts containing two acid groups—for example, Ca(NO₃)₂, Pb(NO₃)₂—carbonic acid, H₂CO₃, is bibasic, that is contains two atoms of hydrogen in the hydrate or two atoms of univalent metals in their salts: for example, Na₂CO₃ is washing soda, a normal salt; NaHCO₃ is the bicarbonate, an acid salt. Therefore, if M′ be a univalent metal, its carbonates in general are the normal carbonate M′₂CO₃ and the acid carbonate, M′HCO₃; or if M″ be a bivalent metal (replacing H₂) its normal carbonate will be M″CO₃; these metals do not usually form acid salts, as we shall see further on. The bibasic character of carbonic acid is akin to that of sulphuric acid, H₂SO₄,[13] but the latter, in distinction from the former, is an example of the energetic or strong acids (such as nitric or hydrochloric), whilst in carbonic acid we observe but feeble development of the acid properties; hence carbonic acid must be considered a weak acid. This conception must, however, be taken as only comparative, as up to this time there is no definitely established rule for measuring the energy[14] of acids. The feeble acid properties of carbonic acid may, however, be judged from the joint evidence of many properties. With such energetic alkalis as soda and potash, carbonic acid forms normal salts, soluble in water, but having an alkaline reaction and in many cases themselves acting as alkalis.[15] The acid salts of these alkalis, NaHCO₃ and KHCO₃, have a neutral reaction on litmus, although they, like acids, contain hydrogen, which may be exchanged for metals. The acid salts of such acids—as, for instance, of sulphuric acid, NaHSO₄—have a clearly defined acid reaction, and therefore carbonic acid is unable to neutralise the powerful basic properties of such alkalis as potash or soda. Carbonic acid does not even combine at all with feeble bases, such as alumina, Al₂O₃, and therefore if a strong solution of sodium carbonate, Na₂CO₃, be added to a strong solution of aluminium sulphate, Al₂(SO₄)₃, although according to double saline decompositions aluminium carbonate, Al₂(CO₃)₃, ought to be formed, the carbonic acid separates, for this salt splits up in the presence of water into aluminium hydroxide and carbonic anhydride: Al₂(CO₃)₃ + 3H₂O = Al₂(OH)₆ + 3CO₂. Thus feeble bases are unable to retain carbonic acid even at ordinary temperatures. For the same reason, in the case of bases of medium energy, although they form carbonates, the latter are comparatively easily decomposed by heating, as is shown by the decomposition of copper carbonate, CuCO₃ (see Introduction), and even of calcium carbonate, CaCO₃. Only the normal (not the acid) salts of such powerful bases as potassium and sodium are capable of standing a red heat without decomposition. The acid salts—for instance, NaHCO₃—decompose even on heating their solutions (2NaHCO₃ = Na₂CO₃ + H₂O + CO₂), evolving carbonic anhydride. The amount of heat given out by the combination of carbonic acid with bases also shows its feeble acid properties, being considerably less than with energetic acids. Thus if a weak solution of forty grams of sodium hydroxide be saturated (up to the formation of a normal salt) with sulphuric or nitric acid or another powerful acid, from thirteen to fifteen thousand calories are given out, but with carbonic acid only about ten thousand calories.[16] The majority of carbonates are insoluble in water, and therefore such solutions as sodium, potassium, or ammonium carbonates form in solutions of most other salts, MX or M″X₂, insoluble precipitates of carbonates, M₂CO₃ or M″CO₃. Thus a solution of barium chloride gives with sodium carbonate a precipitate of barium carbonate, BaCO₃. For this reason rocks, especially those of aqueous origin, very often contain carbonates; for example, calcium, ferrous, or magnesium carbonates, &c.

Carbonic anhydride—which, like water, is formed with the development of a large amount of heat—is very stable. Only very few substances are capable of depriving it of its oxygen. However, certain metals, such as magnesium, potassium and the like, on being heated, burn in it, depositing carbon and forming oxides. If a mixture of carbonic anhydride and hydrogen be passed through a heated tube, the formation of water and carbonic oxide will be observed; CO₂ + H₂ = CO + H₂O. But only a portion of the carbonic acid gas undergoes this change, and therefore the result will be a mixture of carbonic anhydride, carbonic oxide, hydrogen, and water, which does not suffer further change under the action of heat.[17] Although, like water, carbonic anhydride is exceedingly stable, still on being heated it partially decomposes into carbonic oxide and oxygen. Deville showed that such is the case if carbonic anhydride be passed through a long tube containing pieces of porcelain and heated to 1,300°. If the products of decomposition—namely, the carbonic oxide and oxygen—be suddenly cooled, they can be collected separately, although they partly reunite together. A similar decomposition of carbonic anhydride into carbonic oxide and oxygen takes place on passing a series of electric sparks through it (for instance, in the eudiometer). Under these conditions an increase of volume occurs, because two volumes of CO₂ give two volumes of CO and one volume of O. The decomposition reaches a certain limit (less than one-third) and does not proceed further, so that the result is a mixture of carbonic anhydride, carbonic oxide, and oxygen, which is not altered in composition by the continued action of the sparks. This is readily understood, as it is a reversible reaction. If the carbonic anhydride be removed, then the mixture explodes when a spark is passed and forms carbonic anhydride.[17 bis] If from an identical mixture the oxygen (and not the carbonic anhydride) be removed, and a series of sparks be again passed, the decomposition is renewed, and terminates with the complete dissociation of the carbonic anhydride. Phosphorus is used in order to effect the complete absorption of the oxygen. In these examples we see that a definite mixture of changeable substances is capable of arriving at a state of stable equilibrium, destroyed, however, by the removal of one of the substances composing the mixture. This is one of the instances of the influence of mass.

Although carbonic anhydride is decomposed on heating, yielding oxygen, it is nevertheless, like water, an unchangeable substance at ordinary temperatures. Its decomposition, as effected by plants, is on this account all the more remarkable; in this case the whole of the oxygen of the carbonic anhydride is separated in the free state. The mechanism of this change is that the heat and light absorbed by the plants are expended in the decomposition of the carbonic anhydride. This accounts for the enormous influence of temperature and light on the growth of plants. But it is at present not clearly understood how this takes place, or by what separate intermediate reactions the whole process of decomposition of carbonic anhydride in plants into oxygen and the carbohydrates (Note 1) remaining in them, takes place. It is known that sulphurous anhydride (in many ways resembling carbonic anhydride) under the action of light (and also of heat) forms sulphur and sulphuric anhydride, SO₃, and in the presence of water, sulphuric acid. But no similar decomposition has been obtained directly with carbonic anhydride, although it forms an exceedingly easily decomposable higher oxide—percarbonic acid;[18] and perhaps that is the reason the oxygen separates. On the other hand, it is known that plants always form and contain organic acids, and these must be regarded as derivatives of carbonic acid, as is seen by all their reactions, of which we will shortly treat. For this reason it might be thought that the carbonic acid absorbed by the plants first forms (according to Baeyer) formic aldehyde, CH₂O, and from it organic acids, and that these latter in their final transformation form all the other complex organic substances of the plants. Many organic acids are found in plants in considerable quantity; for instance, tartaric acid, C₄H₆O₆, found in grape-juice and in the acid juice of many plants; malic acid, C₄H₆O₅, found not only in unripe apples but in still larger quantities in mountain ash berries; citric acid, C₆H₈O₇, found in the acid juice of lemons, in gooseberries, cranberries, &c.; oxalic acid, C₂H₂O₄, found in wood-sorrel and many other plants. Sometimes these acids exist in a free state in the plants, and sometimes in the form of salts; for instance, tartaric acid is met with in grapes as the salt known as cream of tartar, but in the impure state called argol, or tartar, C₄H₅KO₆. In sorrel we find the so-called salts of sorrel, or acid potassium oxalate, C₂HKO₄. There is a very clear connection between carbonic anhydride and the above-mentioned organic acids—namely, they all, under one condition or another, yield carbonic anhydride, and can all be formed by means of it from substances destitute of acid properties. The following examples afford the best demonstration of this fact: if acetic acid, C₂H₄O₂, the acid of vinegar, be passed in the form of vapour through a heated tube, it splits up into carbonic anhydride and marsh gas = CO₂ + CH₄. But conversely it can also be obtained from those components into which it decomposes. If one equivalent of hydrogen in marsh gas be replaced (by indirect means) by sodium, and the compound CH₃Na is obtained, this directly absorbs carbonic anhydride, forming a salt of acetic acid, CH₃Na + CO₂ = C₂H₃NaO₂; from this acetic acid itself may be easily obtained. Thus acetic acid decomposes into marsh gas and carbonic anhydride, and conversely is obtainable from them. The hydrogen of marsh gas does not, like that in acids, show the property of being directly replaced by metals; i.e. CH₄ does not show any acid character whatever, but on combining with the elements of carbonic anhydride it acquires the properties of an acid. The investigation of all other organic acids shows similarly that their acid character depends on their containing the elements of carbonic anhydride. For this reason there is no organic acid containing less oxygen in its molecule than there is in carbonic anhydride; every organic acid contains in its molecule at least two atoms of oxygen. In order to express the relation between carbonic acid, H₂CO₃, and organic acids, and in order to understand the reason of the acidity of these latter, it is simplest to turn to that law of substitution which shows (Chapter VI.) the relation between the hydrogen and oxygen compounds of nitrogen, and permits us (Chapter VIII.) to regard all hydrocarbons as derived from methane. If we have a given organic compound, A, which has not the properties of an acid, but contains hydrogen connected to carbon, as in hydrocarbons, then ACO₂ will be a monobasic organic acid, A2CO₂ a bibasic, A3CO₂ a tribasic, and so on—that is, each molecule of CO₂ transforms one atom of hydrogen into that state in which it may be replaced by metals, as in acids. This furnishes a direct proof that in organic acids it is necessary to recognise the group HCO₂, or carboxyl. If the addition of CO₂ raises the basicity, the removal of CO₂ lowers it. Thus from the bibasic oxalic acid, C₂H₂O₄, or phthalic acid, C₈H₆O₄, by eliminating CO₂ (easily effected experimentally) we obtain the monobasic formic acid, CH₂O₂, or benzoic acid, C₇H₆O₂, respectively. The nature of carboxyl is directly explained by the law of substitution. Judging from what has been stated in Chapters VI. and VIII. concerning this law, it is evident that CO₂ is CH₄ with the exchange of H₄ for O₂, and that the hydrate of carbonic anhydride, H₂CO₃, is CO(OH)₂, that is, methane, in which two parts of hydrogen are replaced by two parts of the water radical (OH, hydroxyl) and the other two by oxygen. Therefore the group CO(OH), or carboxyl, HCO₂, is a part of carbonic acid, and is equivalent to (OH), and therefore also to H. That is, it is a univalent residue of carbonic acid capable of replacing one atom of hydrogen. Carbonic acid itself is a bibasic acid, both hydrogen atoms in it being replaceable by metals, therefore carboxyl, which contains one of the hydrogen atoms of carbonic acid, represents a group in which the hydrogen is exchangeable for metals. And therefore if 1, 2 ... n atoms of non-metallic hydrogen are exchanged 1, 2 ... n times for carboxyl, we ought to obtain 1, 2 ... n-basic acids. Organic acids are the products of the carboxyl substitution in hydrocarbons.[18 bis] If in the saturated hydrocarbons, C_nH_{2n + 2}, one part of hydrogen is replaced by carboxyl, the monobasic saturated (or fatty) acids, C_nH_{2n + 1}(CO₂H), will be obtained, as, for instance, formic acid, HCO₂H, acetic acid, CH₂CO₂H, ... stearic acid, C₁₇H₃₅CO₂H, &c. The double substitution will give bibasic acids, C_nH_2n(CO₂H)(CO₂H); for instance, oxalic acid n = 0, malonic acid n = 1, succinic acid n = 2, &c. To benzene, C₆H₆ correspond benzoic acid, C₆H₅(CO₂H), phthalic acid (and its isomerides), C₆H₄(CO₂H)₂, up to mellitic acid, C₆(CO₂H)₆, in all of which the basicity is equal to the number of carboxyl groups. As many isomerides exist in hydrocarbons, it is readily understood not only that such can exist also in organic acids, but that their number and structure may be foreseen. This complex and most interesting branch of chemistry is treated separately in organic chemistry.

Carbonic Oxide.—This gas is formed whenever the combustion of organic substances takes place in the presence of a large excess of incandescent charcoal; the air first burns the carbon into carbonic anhydride, but this in penetrating through the red-hot charcoal is transformed into carbonic oxide, CO₂ + C = 2CO. By this reaction carbonic oxide is prepared by passing carbonic anhydride through charcoal at a red heat. It may be separated from the excess of carbonic anhydride by passing it through a solution of alkali, which does not absorb carbonic oxide. This reduction of carbonic anhydride explains why carbonic oxide is formed in ordinary clear fires, where the incoming air passes over a large surface of heated coal. A blue flame is then observed burning above the coal; this is the burning carbonic oxide. When charcoal is burnt in stacks, or when a thick layer of coal is burning in a brazier, and under many similar circumstances, carbonic oxide is also formed. In metallurgical processes, for instance when iron is smelted from the ore, very often the same process of conversion of carbonic anhydride into carbonic oxide occurs, especially if the combustion of the coal be effected in high, so-called blast, furnaces and ovens, where the air enters at the lower part and is compelled to pass through a thick layer of incandescent coal. In this way, also, combustion with flame may be obtained from those kinds of fuel which under ordinary conditions burn without flame: for instance, anthracite, coke, charcoal. Heating by means of a gas-producer—that is, an apparatus producing combustible carbonic oxide from fuel—is carried on in the same manner.[19] In transforming one part of charcoal into carbonic oxide 2,420 heat units are given out, and on burning to carbonic anhydride 8,080 heat units. It is evident that on transforming the charcoal first into carbonic oxide we obtain a gas which in burning is capable of giving out 5,660 heat units for one part of charcoal. This preparatory transformation of fuel into carbonic oxide, or producer gas containing a mixture of carbonic oxide (about ⅓ by volume) and nitrogen (⅔ volume), in many cases presents most important advantages, as it is easy to completely burn gaseous fuel without an excess of air, which would lower the temperature.[20] In stoves where solid fuel is burnt it is impossible to effect the complete combustion of the various kinds of fuel without admitting an excess of air. Gaseous fuel, such as carbonic oxide, is easily completely mixed with air and burnt without excess of it. If, in addition to this, the air and gas required for the combustion be previously heated by means of the heat which would otherwise be uselessly carried off in the products of combustion (smoke)[21] it is easy to reach a high temperature, so high (about 1,800°) that platinum may be melted. Such an arrangement is known as a regenerative furnace.[22] By means of this process not only may the high temperatures indispensable in many industries be obtained (for instance, glass-working, steel-melting, &c.), but great advantage also[23] is gained as regards the quantity of fuel, because the transmission of heat to the object to be heated, other conditions being equal, is determined by the difference of temperatures.

The transformation of carbonic anhydride, by means of charcoal, into carbonic oxide (C + CO₂ = CO + CO) is considered a reversible reaction, because at a high temperature the carbonic oxide splits up into carbon and carbonic anhydride, as Sainte-Claire Deville showed by using the method of the ‘cold and hot tube.’ Inside a tube heated in a furnace another thin metallic (silvered copper) tube is fitted, through which a constant stream of cold water flows. The carbonic oxide coming into contact with the heated walls of the exterior tube forms charcoal, and its minute particles settle in the form of lampblack on the lower side of the cold tube, and, since they are cooled, do not act further on the oxygen or carbonic anhydride formed.[24] A series of electric sparks also decomposes carbonic oxide into carbonic anhydride and carbon, and if the carbonic anhydride be removed by alkali complete decomposition may be obtained (Deville).[24 bis] Aqueous vapour, which is so similar to carbonic anhydride in many respects, acts, at a high temperature, on charcoal in an exactly similar way, C + H₂O = H₂ + CO. From 2 volumes of carbonic anhydride with charcoal 4 volumes of carbonic oxide (2 molecules) are obtained, and precisely the same from 2 volumes of water vapour with charcoal 4 volumes of a gas consisting of hydrogen and carbonic oxide (H₂ + CO) are formed. This mixture of combustible gases is called water gas.[25] But aqueous vapour (and only when strongly superheated, otherwise it cools the charcoal) only acts on charcoal to form a large amount of carbonic oxide at a very high temperature (at which carbonic anhydride dissociates); it begins to react at about 500°, forming carbonic anhydride according to the equation C + 2H₂O = CO₂ + 2H₂. Besides this, carbonic oxide on splitting up forms carbonic anhydride, and therefore water gas always contains a mixture[26] in which hydrogen predominates, the volume of carbonic oxide being comparatively less, whilst the amount of carbonic anhydride increases as the temperature of the reaction decreases (generally it is more than 3 per cent.)

Metals like iron and zinc which at a red heat are capable of decomposing water with the formation of hydrogen, also decompose carbonic anhydride with the formation of carbonic oxide; so both the ordinary products of complete combustion, water and carbonic anhydride, are very similar in their reactions, and we shall therefore presently compare hydrogen and carbonic oxide. The metallic oxides of the above-mentioned metals, when reduced by charcoal, also give carbonic oxide. Priestley obtained it by heating charcoal with zinc oxide. As free carbonic anhydride may be transformed into carbonic oxide, so, in precisely the same way, may that carbonic acid which is in a state of combination; hence, if magnesium or barium carbonates (MgCO₃ or BaCO₃) be heated to redness with charcoal, or iron or zinc, carbonic oxide will be produced—for instance, it is obtained by heating an intimate mixture of 9 parts of chalk and 1 part of charcoal in a clay retort.

Many organic substances[27] on being heated, or under the action of various agents, yield carbonic oxide; amongst these are many organic or carboxylic acids. The simplest are formic and oxalic acids. Formic acid, CH₂O₂, on being heated to 200°, easily decomposes into carbonic oxide and water, CH₂O₂ = CO + H₂O.[27 bis] Usually, however, carbonic oxide is prepared in laboratories, not from formic but from oxalic acid, C₂H₂O₄, the more so as formic acid is itself prepared from oxalic acid. The latter acid is easily obtained by the action of nitric acid on starch, sugar, &c.; it is also found in nature. Oxalic acid is easily decomposed by heat; its crystals first lose water, then partly volatilise, but the greater part is decomposed. The decomposition is of the following nature: it splits up into water, carbonic oxide, and carbonic anhydride,[28] C₂H₂O₄ = H₂O + CO₂ + CO. This decomposition is generally practically effected by mixing oxalic acid with strong sulphuric acid, because the latter assists the decomposition by taking up the water. On heating a mixture of oxalic and sulphuric acids a mixture of carbonic oxide and carbonic anhydride is evolved. This mixture is passed through a solution of an alkali in order to absorb the carbonic anhydride, whilst the carbonic oxide passes on.[28 bis]

In its physical properties carbonic oxide resembles nitrogen; this is explained by the equality of their molecular weights. The absence of colour and smell, the low temperature of the absolute boiling point, -140° (nitrogen, -146°), the property of solidifying at -200° (nitrogen, -202°), the boiling point of -190° (nitrogen, -203°), and the slight solubility (Chapter I., Note 30), of carbonic oxide are almost the same as in those of nitrogen. The chemical properties of both gases are, however, very different, and in these carbonic oxide resembles hydrogen. Carbonic oxide burns with a blue flame, giving 2 volumes of carbonic anhydride from 2 volumes of carbonic oxide, just as 2 volumes of hydrogen give 2 volumes of aqueous vapour. It explodes with oxygen, in the eudiometer, like hydrogen.[29] When breathed it acts as a strong poison, being absorbed by the blood;[30] this explains the action of charcoal fumes, the products of the incomplete combustion of charcoal and other carbonaceous fuels. Owing to its faculty of combining with oxygen, carbonic oxide acts as a powerful reducing agent, taking up the oxygen from many compounds at a red heat, and being itself transformed into carbonic anhydride. The reducing action of carbonic oxide, however, is (like that of hydrogen, Chapter II.) naturally confined to those oxides which easily part with their oxygen—as, for instance, copper oxide—whilst the oxides of magnesium or potassium are not reduced. Metallic iron itself is capable of reducing carbonic anhydride to carbonic oxide, just as it liberates the hydrogen from water. Copper, which does not decompose water, does not decompose carbonic oxide. If a platinum wire heated to 300°, or spongy platinum at the ordinary temperature, be plunged into a mixture of carbonic oxide and oxygen, or of hydrogen and oxygen, the mixture explodes. These reactions are very similar to those peculiar to hydrogen. The following important distinction, however, exists between them—namely: the molecule of hydrogen is composed of H₂, a group of elements divisible into two like parts, whilst, as the molecule of carbonic oxide, CO, contains unlike atoms of carbon and oxygen, in none of its reactions of combination can it give two molecules of matter containing its elements. This is particularly noticeable in the action of chlorine on hydrogen and on carbonic oxide respectively; with the former chlorine forms hydrogen chloride, and with the latter it produces the so-called carbonyl chloride, COCl₂: that is to say, the molecule of hydrogen, H₂, under the action of chlorine divides, forming two molecules of hydrochloric acid, whilst the molecule of carbonic oxide enters in its entirety into the molecule of carbonyl chloride. This characterises the so-called diatomic or bivalent reactions of radicles or residues. H is a monatomic residue or radicle, like K, Cl, and others, whilst carbonic oxide, CO, is an indivisible (undecomposable) bivalent radicle, equivalent to H₂ and not to H, and therefore combining with X₂ and interchangeable with H₂. This distinction is evident from the annexed comparison:

Such monatomic (univalent) residues, X, as H, Cl, Na, NO₂, NH₄, CH₃, CO₂H (carboxyl), OH, and others, in accordance with the law of substitution, combine together, forming compounds, XX'; and with oxygen, or in general with diatomic (bivalent) residues, Y—for instance, O, CO, CH₂, S, Ca, &c. forming compounds XX′Y; but diatomic residues, Y, sometimes capable of existing separately may combine together, forming YY′ and with X₂ or XX′, as we see from the transition of CO into CO₂ and COCl₂. This combining power of carbonic oxide appears in many of its reactions. Thus it is very easily absorbed by cuprous chloride, CuCl, dissolved in fuming hydrochloric acid, forming a crystalline compound, COCu₂Cl₂,2H₂O, decomposable by water; it combines directly with potassium (at 90°), forming (KCO)_n[31] with platinum dichloride, PtCl₂, with chlorine, Cl₂, &c.

But the most remarkable compounds are (1) the compound of CO with metallic nickel, a colourless volatile liquid, Ni(CO)₄, obtained by L. Mond (described in Chapter XXII.) and (2) the compounds of carbonic oxide with the alkalis, for instance with potassium or barium hydroxide, &c.—although it is not directly absorbed by them, as it has no acid properties. Berthelot (1861) showed that potash in the presence of water is capable of absorbing carbonic oxide, but the absorption takes place slowly, little by little, and it is only after being heated for many hours that the whole of the carbonic oxide is absorbed by the potash. The salt CHKO₂ is obtained by this absorption; it corresponds with an acid found in nature—namely, the simplest organic (carboxylic) acid, formic acid, CH₂O₂. It can be extracted from the potassium salt by means of distillation with dilute sulphuric acid, just as nitric acid is prepared from sodium nitrate. The same acid is found in ants and in nettles (when the stings of the nettles puncture the skin they break, and the corrosive formic acid enters into the body); it is also obtained during the action of oxidising agents on many organic substances; it is formed from oxalic acid, and under many conditions splits up into carbonic oxide and water. In the formation of formic acid from carbonic oxide we observe an example of the synthesis of organic compounds, such as are now very numerous, and are treated of in detail in works on organic chemistry.

Formic acid, H(CHO₂), carbonic acid, HO(CHO₂), and oxalic acid, (CHO₂)₂, are the simple organic or carboxylic acids, R(CHO₂) corresponding with HH and HOH. Commencing with carbonic oxide, CO, the formation of carboxylic acids is clearly seen from the fact that CO is capable of combining with X₂, that is of forming COX₂. If, for instance, one X is an aqueous residue, OH (hydroxyl), and the other X is hydrogen, then the simplest organic acid—formic acid, H(COOH)—is obtained. As all hydrocarbons (Chapter VIII.) correspond with the simplest, CH₄, so all organic acids may be considered to proceed from formic acid.

In a similar way it is easy to explain the relation to other compounds of carbon of those compounds which contain nitrogen. By way of an example, we will take one of the carboxyl acids, R(CO₂H), where R is a hydrocarbon radicle (residue). Such an acid, like all others, will give by combination with NH₃ an ammoniacal salt, R(CO₂NH₄). This salt contains the elements for the formation of two molecules of water, and under suitable conditions by the action of bodies capable of taking it up, water may in fact be separated from R(CO₂NH₄), forming by the loss of one molecule of water, amides, RCONH₂, and by the loss of two molecules of water, nitriles, RCN, otherwise known as cyanogen compounds or cyanides.[32] If all the carboxyl acids are united not only by many common reactions but also by a mutual conversion into each other (an instance of which we saw above in the conversion of oxalic acid into formic and carbonic acids) one would expect the same for all the cyanogen compounds also. The common character of their reactions, and the reciprocity of their transformation, were long ago observed by Gay-Lussac, who recognised a common group or radicle (residue) cyanogen, CN, in all of them. The simplest compounds are hydrocyanic or prussic acid, HCN, cyanic acid, OHCN, and free cyanogen, (CN)₂, which correspond to the three simplest carboxyl acids: formic, HCO₂H, carbonic, OHCO₂H, and oxalic, (CO₂H)₂. Cyanogen, like carboxyl, is evidently a monatomic residue and acid, similar to chlorine. As regards the amides RCONH₂, corresponding to the carboxyl acids, they contain the ammoniacal residue NH₂, and form a numerous class of organic compounds met with in nature and obtained in many ways,[33] but not distinguished by such characteristic peculiarities as the cyanogen compounds.

The reactions and properties of the amides and nitriles of the organic acids are described in detail in books on organic chemistry; we will here only touch upon the simplest of them, and to clearly explain the derivative compounds will first consider the ammoniacal salts and amides of carbonic acid.

As carbonic acid is bibasic, its ammonium salts ought to have the following composition: acid carbonate of ammonium, H(NH₄)CO₃, and normal carbonate, (NH₄)₂CO₃; they represent compounds of one or two molecules of ammonia with carbonic acid. The acid salt appears in the form of a non-odoriferous and (when tested with litmus) neutral substance, soluble at the ordinary temperature in six parts of water, insoluble in alcohol, and obtainable in a crystalline form either without water of crystallisation or with various proportions of it. If an aqueous solution of ammonia be saturated with an excess of carbonic anhydride, and then evaporated over sulphuric acid in the bell jar of an air-pump, crystals of this salt are separated. Solutions of all other ammonium carbonates, when evaporated under the air-pump, yield crystals of this salt. A solution of this salt, even at the ordinary temperature, gives off carbonic anhydride, as do all the acid salts of carbonic acid (for instance, NaHCO₃), and at 38° the separation of carbonic anhydride takes place with great rapidity. On losing carbonic anhydride and water, the acid salt is converted into the normal salt, 2(NH₄)HCO₃ = H₂O + CO₂ + (NH₄)2CO₃; the latter, however, decomposes in solution, and can therefore only be obtained in crystals, (NH₄)₂CO₃,H₂O, at low temperatures, and from solutions containing an excess of ammonia as the product of dissociation of this salt: (NH₄)₂CO₃ = NH₃ + (NH₄)HCO₃. But the normal salt,[34] according to the general type, is capable of decomposing with separation of water, and forming ammonium carbamate, NH₄O(CONH₂) = (NH₄)₂CO₃ - H₂O; this still further complicates the chemical transformations of the carbonates of ammonium. It is in fact evident that, by changing the ratios of water, ammonia, and carbonic acid, various intermediate salts will be formed containing mixtures or combinations of those mentioned above. Thus the ordinary commercial carbonate of ammonia is obtained by heating a mixture of chalk and sulphate of ammonia (Chapter VI.), or sal-ammoniac, 2NH₄Cl + CaCO₃ = CaCl₂ + (NH₄)₂CO₃. The normal salt, however, through loss of part of the ammonia, partly forms the acid salt, and, partly through loss of water, forms carbamate, and most frequently presents the composition NH₄O(CONH₂) + 2OH(CO₂NH₄) = 4NH₃ + 3CO₂ + 2H₂O. This salt, in parting under various conditions with ammonia, carbonic anhydride, and water, does not present a constant composition, and ought rather to be regarded as a mixture of acid salt and amide salt. The latter must be recognised as entering into the composition of the ordinary carbonate of ammonia, because it contains less water than is required for the normal or acid salt;[35] but on being dissolved in water this salt gives a mixture of acid and normal salts.

Each of the two ammoniacal salts of carbonic acid has its corresponding amide. That of the acid salt should be acid, if the water given off takes up the hydrogen of the ammonia, as it should according to the common type of formation of the amides, so that OHCONH₂, or carbamic acid, is formed from OHCO₃NH₄. This acid is not known in a free state, but its corresponding ammoniacal salt or ammonium carbamate is known. The latter is easily and immediately formed by mixing 2 volumes of dry ammonia with 1 volume of dry carbonic anhydride, 2NH₃ + CO₂ = NH₄O(CONH₂); it is a solid substance, smells strongly of ammonia, attracts moisture from the air, and decomposes completely at 60°. The fact of this decomposition may be proved[36] by the density of its vapour, which = 13 (H = 1); this exactly corresponds with the density of a mixture of 2 volumes of ammonia and 1 volume of carbonic anhydride. It is easily understood that such a combination will take place with any ammonium carbonate under the action of salts which take up the water—for instance, sodium or potassium carbonate[37]—as in an anhydrous state ammonia and carbonic anhydride only form one compound, CO₂2NH₃.[38] As the normal ammonium carbonate contains two ammonias, and as the amides are formed with the separation of water at the expense of the hydrogen of the ammonias, so this salt has its symmetrical amide, CO(NH₂)₂. This must be termed carbamide. It is identical with urea, CN₂H₄O, which, contained in the urine (about 2 per cent. in human urine), is for the higher animals (especially the carnivorous) the ordinary product of excretion[39] and oxidation of the nitrogenous substances found in the organism. If ammonium carbamate be heated to 140° (in a sealed tube, Bazaroff), or if carbonyl chloride, COCl₂, be treated with ammonia (Natanson), urea will be obtained, which shows its direct connection with carbonic acid—that is, the presence of carbonic acid and ammonia in it. From this it will be understood how urea during the putrefaction of urine is converted into ammonium carbonate, CN₂H₄O + H₂O = CO₂ + 2NH₃.

Thus urea, both by its origin and decomposition, is an amide of carbonic acid. Representing as it does ammonia (two molecules) in which hydrogen (two atoms) is replaced by the bivalent radicle of carbonic acid, urea retains the property of ammonia of entering into combination, with acids (thus nitric acid forms CN₂H₄O,HNO₃), with bases (for instance, with mercury oxide), and with salts (such as sodium chloride, ammonium chloride), but containing an acid residue it has no alkaline properties. It is soluble in water without change, but at a red heat loses ammonia and forms cyanic acid, CNHO,[39 bis] which is a nitrile of carbonic acid—that is to say, is a cyanogen compound, corresponding to the acid ammonium carbonate, OH(CNH₄O₂), which on parting with 2H₂O ought to form cyanic acid, CNOH. Liquid cyanic acid, exceedingly unstable at the ordinary temperatures, gives its stable solid polymer cyanuric acid, O₃H₃C₃N₃. Both have the same composition, and they pass one into another at different temperatures. If crystals of cyanuric acid be heated to a temperature, t°, then the vapour tension, p, in millimetres of mercury (Troost and Hautefeuille) will be:

The vapour contains cyanic acid, and, if it be rapidly cooled, it condenses into a mobile volatile liquid (specific gravity at 0° = 1·14). If the liquid cyanic acid be gradually heated, it passes into a new amorphous polymeride (cyamelide), which, on being heated, like cyanuric acid, forms vapours of cyanic acid. If these fumes are heated above 150° they pass directly into cyanuric acid. Thus at a temperature of 350°, the pressure does not rise above 1,200 mm. on the addition of vapours of cyanic acid, because the whole excess is transformed into cyanuric acid. Hence, the above-mentioned figures give the tension of dissociation of cyanuric acid, or the greatest pressure which the vapours of HOCN are able to attain at a given temperature, whilst at a greater pressure, or by the introduction of a larger mass of the substance into a given volume, the whole of the excess is converted into cyanuric acid. The properties of cyanic acid which we have described were principally observed by Wöhler, and clearly show the faculty of polymerisation of cyanogen compounds. This is observed in many other cyanogen derivatives, and is to be regarded as the consequence of the above-mentioned explanation of their nature. All cyanogen compounds are ammonium salts, R(CNH₄O₂), deprived of water, 2H₂O; therefore the molecules, RCN, ought to possess the faculty of combining with two molecules of water or with other molecules in exchange for it (for instance, with H₂S, or HCl, or 2H₂, &c.), and are therefore capable of combining together. The combination of molecules of the same kind to form more complex ones is what is meant by polymerisation.[40]

Besides being a substance very prone to form polymerides, cyanic acid presents many other features of interest, expounded in greater detail in organic chemistry. However we may mention here the production of the cyanates by the oxidation of the metallic cyanides. Potassium cyanate, KCNO, is most often obtained in this way. Solutions of cyanates by the addition of sulphuric acid yield cyanic acid, which, however, immediately decomposes: CNHO + H₂O = CO₂ + NH₃. A solution of ammonium cyanate, CN(NH₄)O, behaves in the same manner, but only in the cold. On being heated it completely changes because it is transformed into urea. The composition of both substances is identical, CN₂H₄O, but the structure, or disposition of, and connection between, the elements is different: in the ammonium cyanate one atom of nitrogen exists in the form of cyanogen, CN—that is, united with carbon—and the other as ammonium, NH₄, but, as cyanic acid contains the hydroxyl radicle of carbonic acid, OH(CN), the ammonium in this salt is united with oxygen. The composition of this salt is best expressed by supposing one atom of the hydrogen in water to be replaced by ammonium and the other by cyanogen—i.e. that its composition is not symmetrical—whilst in urea both the nitrogen atoms are symmetrically and uniformly disposed as regards the radicle CO of carbonic acid: CO(NH₂)₂. For this reason, urea is much more stable than ammonium cyanate, and therefore the latter, on being slightly heated in solution, is converted into urea. This remarkable isomeric transformation was discovered by Wöhler in 1828.[41] Formamide, HCONH₂, and hydrocyanic acid, HCN, as a nitrile, correspond with formic acid, HCOOH, and therefore ammonium formate, HCOONH₄, and formamide, when acted on by heat and by substances which take up water (phosphoric anhydride) form hydrocyanic acid, HCN, whilst, under many conditions (for instance, on combining with hydrochloric acid in presence of water), this hydrocyanic acid forms formic acid and ammonia. Although containing hydrogen in the presence of two acid-forming elements—namely, carbon and nitrogen[42]—hydrocyanic acid does not give an acid reaction with litmus (cyanic acid has very marked acid properties); but it forms salts, MCN, thus presenting the properties of a feeble acid, and for this reason is called an acid. The small amount of energy which it has is shown by the fact that the cyanides of the alkali metals—for instance, potassium cyanide (KHO + HCN = H₂0 + KCN) in solution have a strongly alkaline reaction.[43] If ammonia be passed over charcoal at a red heat, especially in the presence of an alkali, or if gaseous nitrogen be passed through a mixture of charcoal and an alkali (especially potash, KHO), and also if a mixture of nitrogenous organic substances and alkali be heated to a red heat, in all these cases the alkali metal combines with the carbon and nitrogen, forming a metallic cyanide, MCN—for example, KCN.[43 bis] Potassium cyanide is much used in the arts, and is obtained, as above stated, under many circumstances—as, for instance, in iron smelting, especially with the assistance of wood charcoal, the ash of which contains much potash. The nitrogen of the air, the alkali of the ash, and the charcoal are brought into contact at a high temperature during iron smelting, and therefore, under these conditions, a considerable quantity of potassium cyanide is formed. In practice it is not usual to prepare potassium cyanide directly, but a peculiar compound of it containing potassium, iron, and cyanogen. This compound is potassium ferrocyanide, and is also known as yellow prussiate of potash. This saline substance (see Chapter XXII) has the composition K₄FeC₆N₆ + 2H₂O. The name of cyanogen (κύανος) is derived from the property which this yellow prussiate possesses of forming, with a solution of a ferric salt, FeX₃, the familiar pigment Prussian blue. The yellow prussiate is manufactured on a large scale, and is generally used as the source of the other cyanogen compounds.

If four parts of yellow prussiate be mixed with eight parts of water and three parts of sulphuric acid, and the mixture be heated, it decomposes, volatile hydrocyanic acid separating. This was obtained for the first time by Scheele in 1782, but it was only known to him in solution. In 1809 Ittner prepared anhydrous prussic acid, and in 1815 Gay-Lussac finally settled its properties and showed that it contains only hydrogen, carbon, and nitrogen, CNH. If the distillate (a weak solution of HCN) be redistilled, and the first part collected, the anhydrous acid may be prepared from this stronger solution. In order to do this, pieces of calcium chloride are added to the concentrated solution, when the anhydrous acid floats as a separate layer, because it is not soluble in an aqueous solution of calcium chloride. If this layer be then distilled over a new portion of calcium chloride at the lowest temperature possible, the prussic acid may be obtained completely free from water. It is, however, necessary to use the greatest caution in work of this kind, because prussic acid, besides being extremely poisonous, is exceedingly volatile.[44]

Anhydrous prussic acid is a very mobile and volatile liquid; its specific gravity is 0·697 at 18°; at lower temperatures, especially when mixed with a small quantity of water, it easily congeals; it boils at 26°, and therefore very easily evaporates, and at ordinary temperatures may be regarded as a gas. An insignificant amount, when inhaled or brought into contact with the skin, causes death. It is soluble in all proportions in water, alcohol, and ether: weak aqueous solutions are used in medicine.[45]

The salts MCN—for instance, potassium, sodium, ammonium—as well as the salts M″(CN)₂—for example, barium, calcium, mercury—are soluble in water, but the cyanides of manganese, zinc, lead, and many others are insoluble in water. They form double salts with potassium cyanide and similar metallic cyanides, an example of which we will consider in a further description of the yellow prussiate. Not only are some of the double salts remarkable for their constancy and comparative stability, but so also are the soluble salt HgC₂N₂, the insoluble silver cyanide AgCN, and even potassium cyanide in the absence of water. The last salt,[46] when fused, acts as a reducing agent with its elements K and C, and oxidises when fused with lead oxide, forming potassium cyanate, KOCN, which establishes the connection between HCN and OHCN—that is, between the nitriles of formic and carbonic acids—and this connection is the same as that between the acids themselves, since formic acid, on oxidation, yields carbonic acid. Free cyanogen, (CN)₂ or CNCN, corresponds to hydrocyanic acid in the same manner as free chlorine, Cl₂ or ClCl, corresponds to hydrochloric acid. This composition, judging from what has been already stated, exactly expresses that of the nitrile of oxalic acid, and, as a matter of fact, oxalate of ammonia and the amide corresponding with it (oxamide, Note 33), on being heated with phosphoric anhydride, which takes up the water, yield cyanogen, (CN)₂. This substance is also produced by simply heating some of the metallic cyanides. Mercuric cyanide is particularly adapted for this purpose, because it is easily obtained in a pure state and is then very stable. If mercuric cyanide be heated, it decomposes, in like manner to mercury oxide, into metallic mercury and cyanogen: HgC₂N₂ = Hg + C₂N₂.[47] When cyanogen is formed, part of it always polymerises into a dark brown insoluble substance called paracyanogen, capable of forming cyanogen when heated to redness.[48] Cyanogen is a colourless, poisonous gas, with a peculiar smell and easily condensed by cooling into a colourless liquid, insoluble in water and having a specific gravity of 0·86. It boils at about -21°, and therefore cyanogen may be easily condensed into a liquid by a strong freezing mixture. At -35° liquid cyanogen solidifies. The gas is soluble in water and in alcohol to a considerable extent—namely, 1 volume of water absorbs as much as 4½ volumes, and alcohol 23 volumes. Cyanogen resists the action of a tolerably high temperature without decomposing, but under the action of the electric spark the carbon is separated, leaving a volume of nitrogen equal to the volume of the gas taken. As it contains carbon it burns, and the colour of the flame is reddish-violet, which is due to the presence of nitrogen, all compounds of which impart more or less of this reddish-violet hue to the flame. During the combustion of cyanogen, carbonic anhydride and nitrogen are formed. The same products are obtained in the eudiometer with oxygen or by the action of cyanogen on many oxides at a red heat.