The number of units of affinity active in the case of any particular element is largely dependent, however, upon the nature of the element or elements with which it is associated. Thus, an atom of iodine only combines with one of hydrogen, but may unite with three of chlorine, which never combines with more than a single atom of hydrogen; an atom of phosphorus unites with only three atoms of hydrogen, but with five of chlorine, or with four of hydrogen and one of iodine; and the chlorides corresponding to the higher oxides of lead, nickel, manganese and arsenic, PbO2, Ni2O3, MnO2 and AS2O5 do not exist as stable compounds, but the lower chlorides, PbCl2, NiCl2, MnCl2 and AsCl3, are very stable.

The valency of an element is usually expressed by dashes or Roman numerals placed on the right of its symbol, thus: H', O'', B''', C^{IV}, P^V, Mo^{VI}; but in constructing graphic formulae the symbols of the elements are written with as many lines attached to each symbol as the element which it represents has units of affinity.

The periodic law (see ELEMENT) permits a grouping of the elements according to their valency as follows:—Group O.: helium, neon, argon, krypton and xenon appear to be devoid of valency. Group I.: the alkali metals Li, Na, K, Rb, Cs, and also Ag, monovalent; Cu, monovalent and divalent; Au, monovalent and trivalent. Group II.: the alkaline earth metals Ca, Sr, Ba, and also Be (Gl), Mg, Zn, Cd, divalent; Hg, monovalent and divalent. Group III.: B, trivalent; Al, trivalent, but possibly also tetra- or penta-valent; Ga, divalent and trivalent; In, mono-, di- and tri-valent; Tl, monovalent and trivalent. Group IV.: C, Si, Ge, Zr, Th, tetravalent; Ti, tetravalent and hexavalent; Sn, Pb, divalent and tetravalent; Ce, trivalent and tetravalent. Group V.: N, trivalent and pentavalent, but divalent in nitric oxide; P, As, Sb, Bi, trivalent and pentavalent, the last being possibly divalent in BiO and BiCl2. Group VI.: O, usually divalent, but tetravalent and possibly hexavalent in oxonium and other salts; S, Se, Te, di-, tetra- and hexa-valent; Cr, di-, tri- and hexa-valent; Mo, W, di-, tri-, tetra-, penta- and hexa-valent. Group VII.: H(?), monovalent; the halogens F, Cl, Br, I, usually monovalent, but possibly also tri- and pentavalent; Mn, divalent and trivalent, and possibly heptavalent in permanganates. Group VIII.: Fe, Co, divalent and trivalent; Ni, divalent; Os, Ru, hexavalent and octavalent; Pd, Pt, divalent and tetravalent; Ir, tri-, tetra- and hexa-valent. (See also VALENCY.)

Constitutional Formulae.—Graphic or constitutional formulae are employed to express the manner in which the constituent atoms of compounds are associated together; for example, the trioxide of sulphur is usually regarded as a compound of an atom of hexad sulphur with three atoms of dyad oxygen, and this hypothesis is illustrated by the graphic formula

O // O=S O .

When this oxide is brought into contact with water it combines with it forming sulphuric acid, H2SO4.

In this compound only two of the oxygen atoms are wholly associated with the sulphur atom, each of the remaining oxygen atoms being united by one of its affinities to the sulphur atoms, and by the remaining affinity to an atom of hydrogen; thus—

H.O O // S / H.O O .

The graphic formula of a sulphate is readily deduced by remembering that the hydrogen atoms are partially or entirely replaced. Thus acid sodium sulphate, normal sodium sulphate, and zinc sulphate have the formulae

Na.O O Na.O O O O // // / // S S Zn S / / / H.O O , Na.O O , O O .

Again, the reactions of acetic acid, C2H4O2, show that the four atoms of hydrogen which it contains have not all the same function, and also that the two atoms of oxygen have different functions; the graphic formula which we are led to assign to acetic acid, viz.

O H // H.C—C . H O.H

serves in a measure to express this, three of the atoms of hydrogen being represented as associated with one of the atoms of carbon, whilst the fourth atom is associated with an atom of oxygen which is united by a single affinity to the second atom of carbon, to which, however, the second atom of oxygen is united by both of its affinities. It is not to be supposed that there are any actual bonds of union between the atoms; graphic formulae such as these merely express the hypothesis that certain of the atoms in a compound come directly within the sphere of attraction of certain other atoms, and only indirectly within the sphere of attraction of others,—an hypothesis to which chemists are led by observing that it is often possible to separate a group of elements from a compound, and to displace it by other elements or groups of elements.

Rational formulae of a much simpler description than these graphic formulae are generally employed. For instance, sulphuric acid is usually represented by the formula SO2(OH)2, which indicates that it may be regarded as a compound of the group SO2 with twice the group OH. Each of these OH groups is equivalent in combining or displacing power to a monad element, since it consists of an atom of dyad oxygen associated with a single atom of monad hydrogen, so that in this case the SO2 group is equivalent to an atom of a dyad element. This formula for sulphuric acid, however, merely represents such facts as that it is possible to displace an atom of hydrogen and an atom of oxygen in sulphuric acid by a single atom of chlorine, thus forming the compound SO3HCl; and that by the action of water on the compound SO2Cl2 twice the group OH, or water minus an atom of hydrogen, is introduced in place of the two monad atoms of chlorine—

SO2Cl2 + 2HOH = SO2(OH)2 + 2HCl. Water. Sulphuric acid.

Constitutional formulae like these, in fact, are nothing more than symbolic expressions of the character of the compounds which they represent, the arrangement of symbols in a certain definite manner being understood to convey certain information with regard to the compounds represented.

Groups of two or more atoms like SO2 and OH, which are capable of playing the part of elementary atoms (that is to say, which can be transferred from compound to compound), are termed compound radicals, the elementary atoms being simple radicals. Thus, the atom of hydrogen is a monad simple radical, the atom of oxygen a dyad simple radical, whilst the group OH is a monad compound radical.

It is often convenient to regard compounds as formed upon certain types; alcohol, for example, may be said to be a compound formed upon the water type, that is to say, a compound formed from water by displacing one of the atoms of hydrogen by the group of elements C2H5, thus—

{ H { C2H5 O { O { { H { H Water Alcohol.

Constitutional formulae become of preponderating importance when we consider the more complicated inorganic and especially organic compounds. Their full significance is treated in the section of this article dealing with organic chemistry, and in the articles ISOMERISM and STEREO-ISOMERISM.

Chemical Action.—Chemical change or chemical action may be said to take place whenever changes occur which involve an alteration in the composition of molecules, and may be the result of the action of agents such as heat, electricity or light, or of two or more elements or compounds upon each other.

Three kinds of changes are to be distinguished, viz. changes which involve combination, changes which involve decomposition or separation, and changes which involve at the same time both decomposition and combination. Changes of the first and second kind, according to our views of the constitution of molecules, are probably of very rare occurrence; in fact, chemical action appears almost always to involve the occurrence of both these kinds of change, for, as already pointed out, we must assume that the molecules of hydrogen, oxygen and several other elements are diatomic, or that they consist of two atoms. Indeed, it appears probable that with few exceptions the elements are all compounds of similar atoms united together by one or more units of affinity, according to their valencies. If this be the case, however, it is evident that there is no real distinction between the reactions which take place when two elements combine together and when an element in a compound is displaced by another. The combination, as it is ordinarily termed, of chlorine with hydrogen, and the displacement of iodine in potassium iodide by the action of chlorine, may be cited as examples; if these reactions are represented, as such reactions very commonly are, by equations which merely express the relative weights of the bodies which, enter, into reaction, and of the products, thus—

H + Cl = HCl Hydrogen. Chlorine. Hydrochloric acid.

KI + Cl = KCl + I Potassium iodide. Chlorine. Potassium chloride. Iodine.

they appear to differ in character; but if they are correctly represented by molecular equations, or equations which express the relative number of molecules which enter into reaction and which result from the reaction, it will be obvious that the character of the reaction is substantially the same in both cases, and that both are instances of the occurrence of what is ordinarily termed double decomposition—

H2 + Cl2 = 2HCl Hydrogen. Chlorine. Hydrochloric acid.

2KI + Cl2 = 2KCl + I2. Potassium iodide. Chlorine. Potassium chloride. Iodine.

In all cases of chemical change energy in the form of heat is either developed or absorbed, and the amount of heat developed or absorbed in a given reaction is as definite as are the weights of the substance engaged in the reaction. Thus, in the production of hydrochloric acid from hydrogen and chlorine 22,000 calories are developed; in the production of hydrobromic acid from hydrogen and bromine, however, only 844O calories are developed; and in the formation of hydriodic acid from hydrogen and iodine 6040 calories are absorbed.

This difference in behaviour of the three elements, chlorine, bromine and iodine, which in many respects exhibit considerable resemblance, may be explained in the following manner. We may suppose that in the formation of gaseous hydrochloric acid from gaseous chlorine and hydrogen, according to the equation

H2 + Cl2 = HCl + HCl,

a certain amount of energy is expended in separating the atoms of hydrogen in the hydrogen molecule, and the atoms of chlorine in the chlorine molecule, from each other; but that heat is developed by the combination of the hydrogen atoms with the chlorine atoms, and that, as more energy is developed by the union of the atoms of hydrogen and chlorine than is expended in separating the hydrogen atoms from each other and the chlorine atoms from one another, the result of the action of the two elements upon each other is the development of heat,—the amount finally developed in the reaction being the difference between that absorbed in decomposing the elementary molecules and that developed by the combination of the atoms of chlorine and hydrogen. In the formation of gaseous hydrobromic acid from liquid bromine and gaseous hydrogen—

H2 + Br2 = HBr + HBr,

in addition to the energy expended in decomposing the hydrogen and bromine molecules, energy is also expended in converting the liquid bromine into the gaseous condition, and probably less heat is developed by the combination of bromine and hydrogen than by the combination of chlorine and hydrogen, so that the amount of heat finally, developed is much less than is developed in the formation of hydrochloric acid. Lastly, in the production of gaseous hydriodic acid from hydrogen and solid iodine—

H2 + I2 = HI + HI,

so much energy is expended in the decomposition of the hydrogen and iodine molecules and in the conversion of the iodine into the gaseous condition, that the heat which it may be supposed is developed by the combination of the hydrogen and iodine atoms is insufficient to balance the expenditure, and the final result is therefore negative; hence it is necessary in forming hydriodic acid from its elements to apply heat continuously.

These compounds also afford examples of the fact that, generally speaking, those compounds are most readily formed, and are most stable, in the formation of which the most heat is developed. Thus, chlorine enters into reaction with hydrogen, and removes hydrogen from hydrogenized bodies, far more readily than bromine; and hydrochloric acid is a far more stable substance than hydrobromic acid, hydriodic acid being greatly inferior even to hydrobromic acid in stability. Compounds formed with the evolution of heat are termed exothermic, while those formed with an absorption are termed endothermic. Explosives are the commonest examples of endothermic compounds.

When two substances which by their action upon each other develop much heat enter into reaction, the reaction is usually complete without the employment of an excess of either; for example, when a mixture of hydrogen and oxygen, in the proportions to form water—

2H2 + O2 = 2OH2,

is exploded, it is entirely converted into water. This is also the case if two substances are brought together in solution, by the action of which upon each other a third body is formed which is insoluble in the solvent employed, and which also does not tend to react upon any of the substances present; for instance, when a solution of a chloride is added to a solution of a silver salt, insoluble silver chloride is precipitated, and almost the whole of the silver is removed from solution, even if the amount of the chloride employed be not in excess of that theoretically required.

But if there be no tendency to form an insoluble compound, Or one which is not liable to react upon any of the other substances present, this is no longer the case. For example, when a solution of a ferric salt is added to a solution of potassium thiocyanate, a deep red coloration is produced, owing to the formation of ferric thiocyanate. Theoretically the reaction takes place in the case of ferric nitrate in the manner represented by the equation

Fe(NO3)3 + 3KCNS = Fe(CNS)3 Ferric nitrate. Potassium thiocyanate. Ferric thiocyanate.

+ 3KNO3; Potassium nitrate.

but it is found that even when more than sixty times the amount of potassium thiocyanate required by this equation is added, a portion of the ferric nitrate still remains unconverted, doubtless owing to the occurrence of the reverse change—

Fe(CNS)3 + 3KNO3 = Fe(NO3)3 + 3KCNS.

In this, as in most other cases in which substances act upon one another under such circumstances that the resulting compounds are free to react, the extent to which the different kinds of action which may occur take place is dependent upon the mass of the substances present in the mixture. As another instance of this kind, the decomposition of bismuth chloride by water may be cited. If a very large quantity of water be added, the chloride is entirely decomposed in the manner represented by the equation—

BiCl3 + OH2 = BiOCl + 2HCl, Bismuth chloride. Bismuth oxychloride.

the oxychloride being precipitated; but if smaller quantities of water be added the decomposition is incomplete, and it is found that the extent to which decomposition takes place is proportional to the quantity of water employed, the decomposition being incomplete, except in presence of large quantities of water, because of the occurrence of the reverse action—

BiOCl + 2HCl = BiCl3 + OH2.

Chemical change which merely involves simple decomposition is thus seen to be influenced by the masses of the reacting substances and the presence of the products of decomposition; in other words the system of reacting substances and resultants form a mixture in which chemical action has apparently ceased, or the system is in equilibrium. Such reactions are termed reversible (see CHEMICAL ACTION).

III. INORGANIC CHEMISTRY

Inorganic chemistry is concerned with the descriptive study of the elements and their compounds, except those of carbon. Reference should be made to the separate articles on the different elements and the more important compounds for their preparation, properties and uses. In this article the development of this branch of the science is treated historically.

The earliest discoveries in inorganic chemistry are to be found in the metallurgy, medicine and chemical arts of the ancients. The Egyptians obtained silver, iron, copper, lead, zinc and tin, either pure or as alloys, by smelting the ores; mercury is mentioned by Theophrastus (c. 300 B.C.). The manufacture of glass, also practised in Egypt, demanded a knowledge of sodium or potassium carbonates; the former occurs as an efflorescence on the shores of certain lakes; the latter was obtained from wood ashes. Many substances were used as pigments: Pliny records white lead, cinnabar, verdigris and red oxide of iron; and the preparation of coloured glasses and enamels testifies to the uses to which these and other substances were put. Salts of ammonium were also known; while alum was used as a mordant in dyeing. Many substances were employed in ancient medicine: galena was the basis of a valuable Egyptian cosmetic and drug; the arsenic sulphides, realgar and orpiment, litharge, alum, saltpetre, iron rust were also used. Among the Arabian and later alchemists we find attempts made to collate compounds by specific properties, and it is to these writers that we are mainly indebted for such terms as "alkali," "sal," &c. The mineral acids, hydrochloric, nitric and sulphuric acids, and also aqua regia (a mixture of hydrochloric and nitric acids) were discovered, and the vitriols, alum, saltpetre, sal-ammoniac, ammonium carbonate, silver nitrate (lunar caustic) became better known. The compounds of mercury attracted considerable attention, mainly on account of their medicinal properties; mercuric oxide and corrosive sublimate were known to pseudo-Geber, and the nitrate and basic sulphate to "Basil Valentine." Antimony and its compounds formed the subject of an elaborate treatise ascribed to this last writer, who also contributed to our knowledge of the compounds of zinc, bismuth and arsenic. All the commonly occurring elements and compounds appear to have received notice by the alchemists; but the writings assigned to the alchemical period are generally so vague and indefinite that it is difficult to determine the true value of the results obtained.

In the succeeding iatrochemical period, the methods of the alchemists were improved and new ones devised. Glauber showed how to prepare hydrochloric acid, spiritus salis, by heating rock-salt with sulphuric acid, the method in common use to-day; and also nitric acid from saltpetre and arsenic trioxide. Libavius obtained sulphuric acid from many substances, e.g. alum, vitriol, sulphur and nitric acid, by distillation. The action of these acids on many metals was also studied; Glauber obtained zinc, stannic, arsenious and cuprous chlorides by dissolving the metals in hydrochloric acid, compounds hitherto obtained by heating the metals with corrosive sublimate, and consequently supposed to contain mercury. The scientific study of salts dates from this period, especial interest being taken in those compounds which possessed a medicinal or technical value. In particular, the salts of potassium, sodium and ammonium were carefully investigated, but sodium and potassium salts were rarely differentiated.[9] The metals of the alkaline-earths were somewhat neglected; we find Georg Agricola considering gypsum (calcium sulphate) as a compound of lime, while calcium nitrate and chloride became known at about the beginning of the 17th century. Antimonial, bismuth and arsenical compounds were assiduously studied, a direct consequence of their high medicinal importance; mercurial and silver compounds were investigated for the same reason. The general tendency of this period appears to have taken the form of improving and developing the methods of the alchemists; few new fields were opened, and apart from a more complete knowledge of the nature of salts, no valuable generalizations were attained.

The discovery of phosphorus by Brand, a Hamburg alchemist, in 1669 excited chemists to an unwonted degree; it was also independently prepared by Robert Boyle and J. Kunckel, Brand having kept his process secret. Towards the middle of the 18th century two new elements were isolated: cobalt by G. Brandt in 1742, and nickel by A.F. Cronstedt in 1750. These discoveries were followed by Daniel Rutherford's isolation of nitrogen in 1772, and by K. Scheele's isolation of chlorine and oxygen in 1774 (J. Priestley discovered oxygen independently at about the same time), and his investigation of molybdic and tungstic acids in the following year; metallic molybdenum was obtained by P.J. Hjelm in 1783, and tungsten by Don Fausto d'Elhuyar; manganese was isolated by J.G. Gahn in 1774. In 1784 Henry Cavendish thoroughly examined hydrogen, establishing its elementary nature; and he made the far-reaching discovery that water was composed of two volumes of hydrogen to one of oxygen.

The phlogistic theory, which pervaded the chemical doctrine of this period, gave rise to continued study of the products of calcination and combustion; it thus happened that the knowledge of oxides and oxidation products was considerably developed. The synthesis of nitric acid by passing electric sparks through moist air by Cavendish is a famous piece of experimental work, for in the first place it determined the composition of this important substance, and in the second place the minute residue of air which would not combine, although ignored for about a century, was subsequently examined by Lord Rayleigh and Sir William Ramsay, who showed that it consists of a mixture of elementary substances—argon, krypton, neon and xenon (see ARGON).

The 18th century witnessed striking developments in pneumatic chemistry, or the chemistry of gases, which had been begun by van Helmont, Mayow, Hales and Boyle. Gases formerly considered to be identical came to be clearly distinguished, and many new ones were discovered. Atmospheric air was carefully investigated by Cavendish, who showed that it consisted of two elementary constituents: nitrogen, which was isolated by Rutherford in 1772, and oxygen, isolated in 1774; and Black established the presence, in minute quantity, of carbon dioxide (van Helmont's gas sylvestre). Of the many workers in this field, Priestley occupies an important position. A masterly device, initiated by him, was to collect gases over mercury instead of water; this enabled him to obtain gases previously only known in solution, such as ammonia, hydrochloric acid, silicon fluoride and sulphur dioxide. Sulphuretted hydrogen and nitric oxide were discovered at about the same time.

Returning to the history of the discovery of the elements and their more important inorganic compounds, we come in 1789 to M.H. Klaproth's detection of a previously unknown constituent of the mineral pitchblende. He extracted a substance to which he assigned the character of an element, naming it uranium (from [Greek: Ouranos], heaven); but it was afterwards shown by E.M. Peligot, who prepared the pure metal, that Klaproth's product was really an oxide. This element was investigated at a later date by Sir Henry Roscoe, and more thoroughly and successfully by C. Zimmermann and Alibegoff. Pitchblende attained considerable notoriety towards the end of the 19th century on account of two important discoveries. The first, made by Sir William Ramsay in 1896, was that the mineral evolved a peculiar gas when treated with sulphuric acid; this gas, helium (q.v.), proved to be identical with a constituent of the sun's atmosphere, detected as early as 1868 by Sir Norman Lockyer during a spectroscopic examination of the sun's chromosphere. The second discovery, associated with the Curies, is that of the peculiar properties exhibited by the impure substance, and due to a constituent named radium. The investigation of this substance and its properties (see RADIOACTIVITY) has proceeded so far as to render it probable that the theory of the unalterability of elements, and also the hitherto accepted explanations of various celestial phenomena—the source of solar energy and the appearances of the tails of comets—may require recasting.

In the same year as Klaproth detected uranium, he also isolated zirconia or zirconium oxide from the mineral variously known as zircon, hyacinth, jacynth and jargoon; but he failed to obtain the metal, this being first accomplished some years later by Berzelius, who decomposed the double potassium zirconium fluoride with potassium. In the following year, 1795, Klaproth announced the discovery of a third new element, titanium; its isolation (in a very impure form), as in the case of zirconium, was reserved for Berzelius.

Passing over the discovery of carbon disulphide by W.A. Lampadius in 1796, of chromium by L.N. Vauquelin in 1797, and Klaproth's investigation of tellurium in 1798, the next important series of observations was concerned with platinum and the allied metals. Platinum had been described by Antonio de Ulloa in 1748, and subsequently discussed by H.T. Scheffer in 1752. In 1803 W.H. Wollaston discovered palladium, especially interesting for its striking property of absorbing ("occluding") as much as 376 volumes of hydrogen at ordinary temperatures, and 643 volumes at 90 deg. In the following year he discovered rhodium; and at about the same time Smithson Tennant added two more to the list—iridium and osmium; the former was so named from the changing tints of its oxides ([Greek: iris], rainbow), and the latter from the odour of its oxide ([Greek: osme], smell). The most recently discovered "platinum metal," ruthenium, was recognized by C.E. Claus in 1845. The great number and striking character of the compounds of this group of metals have formed the subject of many investigations, and already there is a most voluminous literature. Berzelius was an early worker in this field; he was succeeded by Bunsen, and Deville and Debray, who worked out the separation of rhodium; and at a later date by P.T. Cleve, the first to make a really thorough study of these elements and their compounds. Of especial note are the curious compounds formed by the union of carbon monoxide with platinous chloride, discovered by Paul Schutzenberger and subsequently investigated by F.B. Mylius and F. Foerster and by Pullinger; the phosphoplatinic compounds formed primarily from platinum and phosphorus pentachloride; and also the "ammino" compounds, formed by the union of ammonia with the chloride, &c., of these metals, which have been studied by many chemists, especially S.M. Jorgensen. Considerable uncertainty existed as to the atomic weights of these metals, the values obtained by Berzelius being doubtful. K.F.O. Seubert redetermined this constant for platinum, osmium and iridium; E.H. Keiser for palladium, and A.A. Joly for ruthenium.

The beginning of the 19th century witnessed the discovery of certain powerful methods for the analysis of compounds and the isolation of elements. Berzelius's investigation of the action of the electric current on salts clearly demonstrated the invaluable assistance that electrolysis could render to the isolator of elements; and the adoption of this method by Sir Humphry Davy for the analysis of the hydrates of the metals of the alkalis and alkaline earths, and the results which he thus achieved, established its potency. In 1808 Davy isolated sodium and potassium; he then turned his attention to the preparation of metallic calcium, barium, strontium and magnesium. Here he met with greater difficulty, and it is to be questioned whether he obtained any of these metals even in an approximately pure form (see ELECTROMETALLURGY). The discovery of boron by Gay Lussac and Davy in 1809 led Berzelius to investigate silica (silex). In the following year he announced that silica was the oxide of a hitherto unrecognized element, which he named silicium, considering it to be a metal. This has proved to be erroneous; it is non-metallic in character, and its name was altered to silicon, from analogy with carbon and boron. At the same time Berzelius obtained the element, in an impure condition, by fusing silica with charcoal and iron in a blast furnace; its preparation in a pure condition he first accomplished in 1823, when he invented the method of heating double potassium fluorides with metallic potassium. The success which attended his experiments in the case of silicon led him to apply it to the isolation of other elements. In 1824 he obtained zirconium from potassium zirconium fluoride; the preparation of (impure) titanium quickly followed, and in 1828 he obtained thorium. A similar process, and equally efficacious, was introduced by F. Wohler in 1827. It consisted in heating metallic chlorides with potassium, and was first applied to aluminium, which was isolated in 1827; in the following year, beryllium chloride was analysed by the same method, beryllium oxide (berylla or glucina) having been known since 1798, when it was detected by L. N. Vauquelin in the gem-stone beryl.

In 1812 B. Courtois isolated the element iodine from "kelp," the burnt ashes of marine plants. The chemical analogy of this substance to chlorine was quickly perceived, especially after its investigation by Davy and Gay Lussac. Cyanogen, a compound which in combination behaved very similarly to chlorine and iodine, was isolated in 1815 by Gay Lussac. This discovery of the first of the then-styled "compound radicals" exerted great influence on the prevailing views of chemical composition. Hydrochloric acid was carefully investigated at about this time by Davy, Faraday and Gay Lussac, its composition and the elementary nature of chlorine being thereby established.

In 1817 F. Stromeyer detected a new metallic element, cadmium, in certain zinc ores; it was rediscovered at subsequent dates by other observers and its chemical resemblance to zinc noticed. In the same year Berzelius discovered selenium in a deposit from sulphuric acid chambers, his masterly investigation including a study of the hydride, oxides and other compounds. Selenic acid was discovered by E. Mitscherlich, who also observed the similarity of the crystallographic characters of selenates and sulphates, which afforded valuable corroboration of his doctrine of isomorphism. More recent and elaborate investigations in this direction by A.E.H. Tutton have confirmed this view.

In 1818 L.J. Thenard discovered hydrogen dioxide, one of the most interesting inorganic compounds known, which has since been carefully investigated by H.E. Schone, M. Traube, Wolfenstein and others. About the same time, J.A. Arfvedson, a pupil of Berzelius, detected a new element, which he named lithium, in various minerals—notably petalite. Although unable to isolate the metal, he recognized its analogy to sodium and potassium; this was confirmed by R. Bunsen and A. Matthiessen in 1855, who obtained the metal by electrolysis and thoroughly examined it and its compounds. Its crimson flame-coloration was observed by C.G. Gmelin in 1818.

The discovery of bromine in 1826 by A.J. Balard completed for many years Berzelius's group of "halogen" elements; the remaining member, fluorine, notwithstanding many attempts, remained unisolated until 1886, when Henri Moissan obtained it by the electrolysis of potassium fluoride dissolved in hydrofluoric acid. Hydrobromic and hydriodic acids were investigated by Gay Lussac and Balard, while hydrofluoric acid received considerable attention at the hands of Gay Lussac, Thenard and Berzelius. We may, in fact, consider that the descriptive study of the various halogen compounds dates from about this time. Balard discovered chlorine monoxide in 1834, investigating its properties and reactions; and his observations on hypochlorous acid and hypochlorites led him to conclude that "bleaching-powder" or "chloride of lime" was a compound or mixture in equimolecular proportions of calcium chloride and hypochlorite, with a little calcium hydrate. Gay Lussac investigated chloric acid; Stadion discovered perchloric acid, since more fully studied by G.S. Serullas and Roscoe; Davy and Stadion investigated chlorine peroxide, formed by treating potassium chlorate with sulphuric acid. Davy also described and partially investigated the gas, named by him "euchlorine," obtained by heating potassium chlorate with hydrochloric acid; this gas has been more recently examined by Pebal. The oxy-acids of iodine were investigated by Davy and H.G. Magnus; periodic acid, discovered by the latter, is characterized by the striking complexity of its salts as pointed out by Kimmins.

In 1830 N.G. Sefstrom definitely proved the existence of a metallic element vanadium, which had been previously detected (in 1801) in certain lead ores by A.M. del Rio; subsequent elaborate researches by Sir Henry Roscoe showed many inaccuracies in the conclusions of earlier workers (for instance, the substance considered to be the pure element was in reality an oxide) and provided science with an admirable account of this element and its compounds. B.W. Gerland contributed to our knowledge of vanadyl salts and the vanadic acids. Chemically related to vanadium are the two elements tantalum and columbium or niobium. These elements occur in the minerals columbite and tantalite, and their compounds became known in the early part of the 19th century by the labours of C. Hatchett, A.G. Ekeberg, W.H. Wollaston and Berzelius. But the knowledge was very imperfect; neither was it much clarified by H. Rose, who regarded niobium oxide as the element. The subject was revived in 1866 by C.W. Blomstrand and J.C. Marignac, to whom is due the credit of first showing the true chemical relations of these elements. Subsequent researches by Sainte Claire Deville and L.J. Troost, and by A.G. Kruss and L.E. Nilson, and subsequently (1904) by Hall, rendered notable additions to our knowledge of these elements and their compounds. Tantalum has in recent years been turned to economic service, being employed, in the same manner as tungsten, for the production of the filaments employed in incandescent electric lighting.

In 1833 Thomas Graham, following the paths already traced out by E.D. Clarke, Gay Lussac and Stromeyer, published his masterly investigation of the various phosphoric acids and their salts, obtaining results subsequently employed by J. von Liebig in establishing the doctrine of the characterization and basicity of acids. Both phosphoric and phosphorous acids became known, although imperfectly, towards the end of the 18th century; phosphorous acid was first obtained pure by Davy in 1812, while pure phosphorous oxide, the anhydride of phosphorous acid, remained unknown until T.E. Thorpe's investigation of the products of the slow combustion of phosphorus. Of other phosphorus compounds we may here notice Gengembre's discovery of phosphuretted hydrogen (phosphine) in 1783, the analogy of which to ammonia was first pointed out by Davy and supported at a later date by H. Rose; liquid phosphuretted hydrogen was first obtained by Thenard in 1838; and hypophosphorous acid was discovered by Dulong in 1816. Of the halogen compounds of phosphorus, the trichloride was discovered by Gay Lussac and Thenard, while the pentachloride was obtained by Davy. The oxychloride, bromides, and other compounds were subsequently discovered; here we need only notice Moissan's preparation of the trifluoride and Thorpe's discovery of the pentafluoride, a compound of especial note, for it volatilizes unchanged, giving a vapour of normal density and so demonstrating the stability of a pentavalent phosphorus compound (the pentachloride and pentabromide dissociate into a molecule of the halogen element and phosphorus trichloride).

In 1840 C.F. Schonbein investigated ozone, a gas of peculiar odour (named from the Gr. [Greek: ozein], to smell) observed in 1785 by Martin van Marum to be formed by the action of a silent electric discharge on the oxygen of the air; he showed it to be an allotropic modification of oxygen, a view subsequently confirmed by Marignac, Andrews and Soret. In 1845 a further contribution to the study of allotropy was made by Anton Schrotter, who investigated the transformations of yellow and red phosphorus, phenomena previously noticed by Berzelius, the inventor Of the term "allotropy." The preparation of crystalline boron in 1856 by Wohler and Sainte Claire Deville showed that this element also existed in allotropic forms, amorphous boron having been obtained simultaneously and independently in 1809 by Gay Lussac and Davy. Before leaving this phase of inorganic chemistry, we may mention other historical examples of allotropy. Of great importance is the chemical identity of the diamond, graphite and charcoal, a fact demonstrated in part by Lavoisier in 1773, Smithson Tennant in 1706, and by Sir George Steuart-Mackenzie (1780-1848), who showed that equal weights of these three substances yielded the same weight of carbon dioxide on combustion. The allotropy of selenium was first investigated by Berzelius; and more fully in 1851 by J.W. Hittorf, who carefully investigated the effects produced by heat; crystalline selenium possesses a very striking property, viz. when exposed to the action of light its electric conductivity increases. Another element occurring in allotropic forms is sulphur, of which many forms have been described. E. Mitscherlich was an early worker in this field. A modification known as "black sulphur," soluble in water, was announced by F.L. Knapp in 1848, and a colloidal modification was described by H. Debus. The dynamical equilibrium between rhombic, liquid and monosymmetric sulphur has been worked out by H.W. Bakhuis Roozeboom. The phenomenon of allotropy is not confined to the non-metals, for evidence has been advanced to show that allotropy is far commoner than hitherto supposed. Thus the researches of Carey Lea, E.A. Schneider and others, have proved the existence of "colloidal silver"; similar forms of the metals gold, copper, and of the platinum metals have been described. The allotropy of arsenic and antimony is also worthy of notice, but in the case of the first element the variation is essentially non-metallic, closely resembling that of phosphorus. The term allotropy has also been applied to inorganic compounds, identical in composition, but assuming different crystallographic forms. Mercuric oxide, sulphide and iodide; arsenic trioxide; titanium dioxide and silicon dioxide may be cited as examples.

The joint discovery in 1859 of the powerful method of spectrum analysis (see SPECTROSCOPY) by G.R. Kirchhoff and R.W. Bunsen, and its application to the detection and the characterization of elements when in a state of incandescence, rapidly led to the discovery of many hitherto unknown elements. Within two years of the invention the authors announced the discovery of two metals, rubidium and caesium, closely allied to sodium, potassium and lithium in properties, in the mineral lepidolite and in the Durkheim mineral water. In 1861 Sir William Crookes detected thallium (named from the Gr. [Greek: thallos], a green bud, on account of a brilliant green line in its spectrum) in the selenious mud of the sulphuric acid manufacture; the chemical affinities of this element, on the one hand approximating to the metals of the alkalis, and on the other hand to lead, were mainly established by C.A. Lamy. Of other metals first detected by the spectroscope mention is to be made of indium, determined by F. Reich and H.T. Richter in 1863, and of gallium, detected in certain zinc blendes by Lecoq de Boisbaudran in 1875. The spectroscope has played an all-important part in the characterization of the elements, which, in combination with oxygen, constitute the group of substances collectively named the "rare earths." The substances occur, in very minute quantity, in a large number of sparingly-distributed and comparatively rare minerals—euxenite, samarksite, cerite, yttrotantalite, &c. Scandinavian specimens of these minerals were examined by J. Gadolin, M.H. Klaproth, and especially by Berzelius; these chemists are to be regarded as the pioneers in this branch of descriptive chemistry. Since their day many chemists have entered the lists, new and powerful methods of research have been devised, and several new elements definitely characterized. Our knowledge on many points, however, is very chaotic; great uncertainty and conflict of evidence circulate around many of the "new elements" which have been announced, so much so that P.T. Cleve proposed to divide the "rare earth" metals into two groups, (1) "perfectly characterized"; (2) "not yet thoroughly characterized." The literature of this subject is very large. The memorial address on J.C.G. de Marignac, a noted worker in this field, delivered by Cleve, a high authority on this subject, before the London Chemical Society (J.C.S. Trans., 1895, p. 468), and various papers in the same journal by Sir William Crookes, Bohuslav Brauner and others should be consulted for details.

In the separation of the constituents of the complex mixture of oxides obtained from the "rare earth" minerals, the methods generally forced upon chemists are those of fractional precipitation or crystallization; the striking resemblances of the compounds of these elements rarely admitting of a complete separation by simple precipitation and filtration. The extraordinary patience requisite to a successful termination of such an analysis can only be adequately realized by actual research; an idea may be obtained from Crookes's Select Methods in Analysis. Of recent years the introduction of various organic compounds as precipitants or reagents has reduced the labour of the process; and advantage has also been taken of the fairly complex double salts which these metals form with compounds. The purity of the compounds thus obtained is checked by spectroscopic observations. Formerly the spark- and absorption-spectra were the sole methods available; a third method was introduced by Crookes, who submitted the oxides, or preferably the basic sulphates, to the action of a negative electric discharge in vacuo, and investigated the phosphorescence induced spectroscopically. By such a study in the ultra-violet region of a fraction prepared from crude yttria he detected a new element victorium, and subsequently by elaborate fractionation obtained the element itself.

The first earth of this group to be isolated (although in an impure form) was yttria, obtained by Gadolin in 1794 from the mineral gadolinite, which was named after its discoverer and investigator. Klaproth and Vauquelin also investigated this earth, but without detecting that it was a complex mixture—a discovery reserved for C.G. Mosander. The next discovery, made independently and simultaneously in 1803 by Klaproth and by W. Hisinger and Berzelius, was of ceria, the oxide of cerium, in the mineral cerite found at Ridderhytta, Westmannland, Sweden. These crude earths, yttria and ceria, have supplied most if not all of the "rare earth" metals. In 1841 Mosander, having in 1839 discovered a new element lanthanum in the mineral cerite, isolated this element and also a hitherto unrecognized substance, didymia, from crude yttria, and two years later he announced the determination of two fresh constituents of the same earth, naming them erbia and terbia. Lanthanum has retained its elementary character, but recent attempts at separating it from didymia have led to the view that didymium is a mixture of two elements, praseodymium and neodymium (see DIDYMIUM). Mosander's erbia has been shown to contain various other oxides—thulia, holmia, &c.—but this has not yet been perfectly worked out. In 1878 Marignac, having subjected Mosander's erbia, obtained from gadolinite, to a careful examination, announced the presence of a new element, ytterbium; this discovery was confirmed by Nilson, who in the following year discovered another element, scandium, in Marignac's ytterbia. Scandium possesses great historical interest, for Cleve showed that it was one of the elements predicted by Mendeleeff about ten years previously from considerations based on his periodic classification of the elements (see ELEMENT). Other elements predicted and characterized by Mendeleeff which have been since realized are gallium, discovered in 1875, and germanium, discovered in 1885 by Clemens Winkler.

In 1880 Marignac examined certain earths obtained from the mineral samarskite, which had already in 1878 received attention from Delafontaine and later from Lecoq de Boisbaudran. He established the existence of two new elements, samarium and gadolinium, since investigated more especially by Cleve, to whom most of our knowledge on this subject is due. In addition to the rare elements mentioned above, there are a score or so more whose existence is doubtful. Every year is attended by fresh "discoveries" in this prolific source of elementary substances, but the paucity of materials and the predilections of the investigators militate in some measure against a just valuation being accorded to such researches. After having been somewhat neglected for the greater attractions and wider field presented by organic chemistry, the study of the elements and their inorganic compounds is now rapidly coming into favour; new investigators are continually entering the lists; the beaten paths are being retravcrsed and new ramifications pursued.

IV. ORGANIC CHEMISTRY

While inorganic chemistry was primarily developed through the study of minerals—a connexion still shown by the French appellation chimie minerale—organic chemistry owes its origin to the investigation of substances occurring in the vegetable and animal organisms. The quest of the alchemists for the philosopher's stone, and the almost general adherence of the iatrochemists to the study of the medicinal characters and preparation of metallic compounds, stultified in some measure the investigation of vegetable and animal products. It is true that by the distillation of many herbs, resins and similar substances, several organic compounds had been prepared, and in a few cases employed as medicines; but the prevailing classification of substances by physical and superficial properties led to the correlation of organic and inorganic compounds, without any attention being paid to their chemical composition. The clarification and spirit of research so clearly emphasized by Robert Boyle in the middle of the 17th century is reflected in the classification of substances expounded by Nicolas Lemery, in 1675, in his Cours de chymie. Taking as a basis the nature of the source of compounds, he framed three classes: "mineral," comprising the metals, minerals, earths and stones; "vegetable," comprising plants, resins, gums, juices, &c.; and "animal," comprising animals, their different parts and excreta. Notwithstanding the inconsistency of his allocation of substances to the different groups (for instance, acetic acid was placed in the vegetable class, while the acetates and the products of their dry distillation, acetone, &c., were placed in the mineral class), this classification came into favour. The phlogistonists endeavoured to introduce chemical notions to support it: Becher, in his Physica subterranea(1669), stated that mineral, vegetable and animal matter contained the same elements, but that more simple combinations prevailed in the mineral kingdom; while Stahl, in his Specimen Becherianum (1702), held the "earthy" principle to predominate in the mineral class, and the "aqueous" and "combustible" in the vegetable and animal classes. It thus happened that in the earlier treatises on phlogistic chemistry organic substances were grouped with all combustibles.

The development of organic chemistry from this time until almost the end of the 18th century was almost entirely confined to such compounds as had practical applications, especially in pharmacy and dyeing. A new and energetic spirit was introduced by Scheele; among other discoveries this gifted experimenter isolated and characterized many organic acids, and proved the general occurrence of glycerin (Olsuss) in all oils and fats. Bergman worked in the same direction; while Rouelle was attracted to the study of animal chemistry. Theoretical speculations were revived by Lavoisier, who, having explained the nature of combustion and determined methods for analysing compounds, concluded that vegetable substances ordinarily contained carbon, hydrogen and oxygen, while animal substances generally contained, in addition to these elements, nitrogen, and sometimes phosphorus and sulphur. Lavoisier, to whom chemistry was primarily the chemistry of oxygen compounds, having developed the radical theory initiated by Guyton de Morveau, formulated the hypothesis that vegetable and animal substances were oxides of radicals composed of carbon and hydrogen; moreover, since simple radicals (the elements) can form more than one oxide, he attributed the same character to his hydrocarbon radicals: he considered, for instance, sugar to be a neutral oxide and oxalic acid a higher oxide of a certain radical, for, when oxidized by nitric acid, sugar yields oxalic acid. At the same time, however, he adhered to the classification of Lemery; and it was only when identical compounds were obtained from both vegetable and animal sources that this subdivision was discarded, and the classes were assimilated in the division organic chemistry.

At this time there existed a belief, held at a later date by Berzelius, Gmelin and many others, that the formation of organic compounds was conditioned by a so-called vital force; and the difficulty of artificially realizing this action explained the supposed impossibility of synthesizing organic compounds. This dogma was shaken by Wohler's synthesis of urea in 1828. But the belief died hard; the synthesis of urea remained isolated for many years; and many explanations were attempted by the vitalists (as, for instance, that urea was halfway between the inorganic and organic kingdoms, or that the carbon, from which it was obtained, retained the essentials of this hypothetical vital force), but only to succumb at a later date to the indubitable fact that the same laws of chemical combination prevail in both the animate and inanimate kingdoms, and that the artificial or laboratory synthesis of any substance, either inorganic or organic, is but a question of time, once its constitution is determined.[10]

The exact delimitation of inorganic and organic chemistry engrossed many minds for many years; and on this point there existed considerable divergence of opinion for several decades. In addition to the vitalistic doctrine of the origin of organic compounds, views based on purely chemical considerations were advanced. The atomic theory, and its correlatives—the laws of constant and multiple proportions—had been shown to possess absolute validity so far as well-characterized inorganic compounds were concerned; but it was open to question whether organic compounds obeyed the same laws. Berzelius, in 1813 and 1814, by improved methods of analysis, established that the Daltonian laws of combination held in both the inorganic and organic kingdoms; and he adopted the view of Lavoisier that organic compounds were oxides of compound radicals, and therefore necessarily contained at least three elements—carbon, hydrogen and oxygen. This view was accepted in 1817 by Leopold Gmelin, who, in his Handbuch der Chemie, regarded inorganic compounds as being of binary composition (the simplest being oxides both acid and basic, which by combination form salts also of binary form), and organic compounds as ternary, i.e. composed of three elements; furthermore, he concluded that inorganic compounds could be synthesized, whereas organic compounds could not. A consequence of this empirical division was that marsh gas, ethylene and cyanogen were regarded as inorganic, and at a later date many other hydrocarbons of undoubtedly organic nature had to be included in the same division.

The binary conception of compounds held by Berzelius received apparent support from the observations of Gay Lussac, in 1815, on the vapour densities of alcohol and ether, which pointed to the conclusion that these substances consisted of one molecule of water and one and two of ethylene respectively; and from Pierre Jean Robiquet and Jean Jacques Colin, showing, in 1816, that ethyl chloride (hydrochloric ether) could be regarded as a compound of ethylene and hydrochloric acid.[11] Compound radicals came to be regarded as the immediate constituents of organic compounds; and, at first, a determination of their empirical composition was supposed to be sufficient to characterize them. To this problem there was added another in about the third decade of the 19th century—namely, to determine the manner in which the atoms composing the radical were combined; this supplementary requisite was due to the discovery of the isomerism of silver fulminate and silver cyanate by Justus von Liebig in 1823, and to M. Faraday's discovery of butylene, isomeric with ethylene, in 1825.

The classical investigation of Liebig and Friedrich Wohler on the radical of benzoic acid ("Uber das Radikal der Benzoe-saure," Ann. Chem., 1832, 3, p. 249) is to be regarded as a most important contribution to the radical theory, for it was shown that a radical containing the elements carbon, hydrogen and oxygen, which they named benzoyl (the termination yl coming from the Gr. [Greek: yle], matter), formed the basis of benzaldehyde, benzoic acid, benzoyl chloride, benzoyl bromide and benzoyl sulphide, benzamide and benzoic ether. Berzelius immediately appreciated the importance of this discovery, notwithstanding that he was compelled to reject the theory that oxygen could not play any part in a compound radical—a view which he previously considered as axiomatic; and he suggested the names "proin" or "orthrin" (from the Gr. [Greek: proi] and [Greek: orthros], at dawn). However, in 1833, Berzelius reverted to his earlier opinion that oxygenated radicals were incompatible with his electrochemical theory; he regarded benzoyl as an oxide of the radical C14H10, which he named "picramyl" (from [Greek: pikros], bitter, and [Greek: amygdale], almond), the peroxide being anhydrous benzoic acid; and he dismissed the views of Gay Lussac and Dumas that ethylene was the radical of ether, alcohol and ethyl chloride, setting up in their place the idea that ether was a suboxide of ethyl, (C2H5)2O, which was analogous to K2O, while alcohol was an oxide of a radical C2H6; thus annihilating any relation between these two compounds. This view was modified by Liebig, who regarded ether as ethyl oxide, and alcohol as the hydrate of ethyl oxide; here, however, he was in error, for he attributed to alcohol a molecular weight double its true value. Notwithstanding these errors, the value of the "ethyl theory" was perceived; other radicals—formyl, methyl, amyl, acetyl, &c.—were characterized; Dumas, in 1837, admitted the failure of the etherin theory; and, in company with Liebig, he defined organic chemistry as the "chemistry of compound radicals." The knowledge of compound radicals received further increment at the hands of Robert W. Bunsen, the discoverer of the cacodyl compounds.

The radical theory, essentially dualistic in nature in view of its similarity to the electrochemical theory of Berzelius, was destined to succumb to a unitary theory. Instances had already been recorded of cases where a halogen element replaced hydrogen with the production of a closely allied substance: Gay Lussac had prepared cyanogen chloride from hydrocyanic acid; Faraday, hexachlorethane from ethylene dichloride, &c. Here the electro-negative halogens exercised a function similar to electro-positive hydrogen. Dumas gave especial attention to such substitutions, named metalepsy [Greek: metalepsis], exchange); and framed the following empirical laws to explain the reactions:—(1) a body containing hydrogen when substituted by a halogen loses one atom of hydrogen for every atom of halogen introduced; (2) the same holds if oxygen be present, except that when the oxygen is present as water the latter first loses its hydrogen without replacement, and then substitution according to (1) ensues. Dumas went no further that thus epitomizing his observations; and the next development was made in 1836 by Auguste Laurent, who, having amplified and discussed the applicability of Dumas' views, promulgated his Nucleus Theory, which assumed the existence of "original nuclei or radicals" (radicaux or noyaux fondamentaux) composed of carbon and hydrogen, and "derived nuclei" (radicaux or noyaux derives) formed from the original nuclei by the substitution of hydrogen or the addition of other elements, and having properties closely related to the primary nuclei.

Vigorous opposition was made by Liebig and Berzelius, the latter directing his attack against Dumas, whom he erroneously believed to be the author of what was, in his opinion, a pernicious theory. Dumas repudiated the accusation, affirming that he held exactly contrary views to Laurent; but only to admit their correctness in 1839, when, from his own researches and those of Laurent, Malaguti and Regnault, he formulated his type theory. According to this theory a "chemical type" embraced compounds containing the same number of equivalents combined in a like manner and exhibiting similar properties; thus acetic and trichloracetic acids, aldehyde and chloral, marsh gas and chloroform are pairs of compounds referable to the same type. He also postulated, with Regnault, the existence of "molecular or mechanical types" containing substances which, although having the same number of equivalents, are essentially different in characters. His unitary conceptions may be summarized: every chemical compound forms a complete whole, and cannot therefore consist of two parts; and its chemical character depends primarily upon the arrangement and number of the atoms, and, in a lesser degree, upon their chemical nature. More emphatic opposition to the dualistic theory of Berzelius was hardly possible; this illustrious chemist perceived that the validity of his electrochemical theory was called in question, and therefore he waged vigorous war upon Dumas and his followers. But he fought in a futile cause; to explain the facts put forward by Dumas he had to invent intricate and involved hypotheses, which, it must be said, did not meet with general acceptance; Liebig seceded from him, and invited Wohler to endeavour to correct him. Still, till the last Berzelius remained faithful to his original theory; experiment, which he had hitherto held to be the only sure method of research, he discarded, and in its place he substituted pure speculation, which greatly injured the radical theory. At the same time, however, the conception of radicals could not be entirely displaced, for the researches of Liebig and Wohler, and those made subsequently by Bunsen, demonstrated beyond all doubt the advantages which would accrue from their correct recognition.

A step forward—the fusion of Dumas', type theory and the radical theory—was made by Laurent and Charles Gerhardt. As early as 1842, Gerhardt in his Precis de chimie organique exhibited a marked leaning towards Dumas' theory, and it is without doubt that both Dumas and Laurent exercised considerable influence on his views. Unwilling to discard the strictly unitary views of these chemists, or to adopt the copulae theory of Berzelius, he revived the notion of radicals in a new form. According to Gerhardt, the process of substitution consisted of the union of two residues to form a unitary whole; these residues, previously termed "compound radicals," are atomic complexes which remain over from the interaction of two compounds. Thus, he interpreted the interaction of benzene and nitric acid as C6H6 + HNO3 = C6H5NO2 + H2O, the "residues" of benzene being C6H5 and H, and of nitric acid HO and NO2. Similarly he represented the reactions investigated by Liebig and Wohler on benzoyl compounds as double decompositions.

This rejuvenation of the notion of radicals rapidly gained favour; and the complete fusion of the radical theory with the theory of types was not long delayed. In 1849 C.A. Wurtz discovered the amines or substituted ammonias, previously predicted by Liebig; A.W. von Hofmann continued the investigation, and established their recognition as ammonia in which one or more hydrogen atoms had been replaced by hydrocarbon radicals, thus formulating the "ammonia type." In 1850 A.W. Williamson showed how alcohol and ether were to be regarded as derived from water by substituting one or both hydrogen atoms by the ethyl group; he derived acids and the acid anhydrides from the same type; and from a comparison of many inorganic and the simple organic compounds he concluded that this notion of a "water-type" clarified, in no small measure, the conception of the structure of compounds.

These conclusions were co-ordinated in Gerhardt's "new theory of types." Taking as types hydrogen, hydrochloric acid, water and ammonia, he postulated that all organic compounds were referable to these four forms: the hydrogen type included hydrocarbons, aldehydes and ketones; the hydrochloric acid type, the chlorides, bromides and iodides; the water type, the alcohols, ethers, monobasic acids, acid anhydrides, and the analogous sulphur compounds; and the ammonia type, the amines, acid-amides, and the analogous phosphorus and arsenic compounds. The recognition of the polybasicity of acids, which followed from the researches of Thomas Graham and Liebig, had caused Williamson to suggest that dibasic acids could be referred to a double water type, the acid radical replacing an atom of hydrogen in each water molecule; while his discovery of tribasic formic ether, CH(OC2H5)3, in 1854 suggested a triple water type. These views were extended by William Odling, and adopted by Gerhardt, but with modifications of Williamson's aspects. A further generalization was effected by August Kekule, who rejected the hydrochloric acid type as unnecessary, and introduced the methane type and condensed mixed types. Pointing out that condensed types can only be fused with a radical replacing more than one atom of hydrogen, he laid the foundation of the doctrine of valency, a doctrine of incalculable service to the knowledge of the structure of chemical compounds.

At about the same time Hermann Kolbe attempted a rehabilitation, with certain modifications, of the dualistic conception of Berzelius. He rejected the Berzelian tenet as to the unalterability of radicals, and admitted that they exercised a considerable influence upon the compounds with which they were copulated. By his own investigations and those of Sir Edward Frankland it was proved that the radical methyl existed in acetic acid; and by the electrolysis of sodium acetate, Kolbe concluded that he had isolated this radical; in this, however, he was wrong, for he really obtained ethane, C2H6, and not methyl, CH3. From similar investigations of valerianic acid he was led to conclude that fatty acids were oxygen compounds of the radicals hydrogen, methyl, ethyl, &c., combined with the double carbon equivalent C2. Thus the radical of acetic acid, acetyl,[12] was C2H3.C2. (It will be noticed that Kolbe used the atomic weights H=1, C=6, O=8, S=16, &c.; his formulae, however, were molecular formulae, i.e. the molecular weights were the same as in use to-day.) This connecting link, C2, was regarded as essential, while the methyl, ethyl, &c. was but a sort of appendage; but Kolbe could not clearly conceive the manner of copulation.

The brilliant researches of Frankland on the organo-metallic compounds, and his consequent doctrine of saturation capacity or valency of elements and radicals, relieved Kolbe's views of all obscurity. The doctrine of copulae was discarded, and in 1859 emphasis was given to the view that all organic compounds were derivatives of inorganic by simple substitution processes. He was thus enabled to predict compounds then unknown, e.g. the secondary and tertiary alcohols; and with inestimable perspicacity he proved intimate relations between compounds previously held to be quite distinct. Lactic acid and alanine were shown to be oxy- and amino-propionic acids respectively; glycollic acid and glycocoll, oxy- and amino-acetic acids; salicylic and benzamic acids, oxy- and amino-benzoic acids.

Another consequence of the doctrine of valency was that it permitted the graphic representation of the molecule. The "structure theory" (or the mode of linking of the atoms) of carbon compounds, founded by Butlerow, Kekule and Couper and, at a later date, marvellously enhanced by the doctrine of stereo-isomerism, due to J.H. van't Hoff and Le Bel, occupies such a position in organic chemistry that its value can never be transcended. By its aid the molecule is represented as a collection of atoms connected together by valencies in such a manner that the part played by each atom is represented; isomerism, or the existence of two or more chemically different substances having identical molecular weights, is adequately shown; and, most important of all, once the structure is determined, the synthesis of the compound is but a matter of time.

In this summary the leading factors which have contributed to a correct appreciation of organic compounds have so far been considered historically, but instead of continuing this method it has been thought advisable to present an epitome of present-day conclusions, not chronologically, but as exhibiting the principles and subject-matter of our science.

Classification of Organic Compounds.

An apt definition of organic chemistry is that it is "the study of the hydrocarbons and their derivatives." This description, although not absolutely comprehensive, serves as a convenient starting-point for a preliminary classification, since a great number of substances, including the most important, are directly referable to hydrocarbons, being formed by replacing one or more hydrogen atoms by other atoms or groups. Two distinct types of hydrocarbons exist: (1) those consisting of an open chain of carbon atoms—named the "aliphatic series" ([Greek: aleiphar], oil or fat), and (2) those consisting of a closed chain—the "carbocyclic series." The second series can be further divided into two groups: (1) those exhibiting properties closely analogous to the aliphatic series—the polymethylenes (q.v.), and (2) a series exhibiting properties differing in many respects from the aliphatic and polymethylene compounds, and characterized by a peculiar stability which is to be associated with the disposition of certain carbon valencies not saturated by hydrogen—the "aromatic series." There also exists an extensive class of compounds termed the "heterocyclic series"—these compounds are derived from ring systems containing atoms other than carbon; this class is more generally allied to the aromatic series than to the aliphatic.

We now proceed to discuss the types of aliphatic compounds; then, the characteristic groupings having been established, an epitome of their derivatives will be given. Carbocyclic rings will next be treated, benzene and its allies in some detail; and finally the heterocyclic nuclei.

Accepting the doctrine of the tetravalency of carbon (its divalency in such compounds as carbon monoxide, various isocyanides, fulminic acid, &c., and its possible trivalency in M. Gomberg's triphenyl-methyl play no part in what follows), it is readily seen that the simplest hydrocarbon has the formula CH4 named methane, in which the hydrogen atoms are of equal value, and which may be pictured as placed at the vertices of a tetrahedron, the carbon atom occupying the centre. This tetrahedral configuration is based on the existence of only one methylene dichloride, two being necessary if the carbon valencies were directed from the centre of a plane square to its corners, and on the existence of two optical isomers of the formula C.A.B.D.E., C being a carbon atom and A.B.D.E. being different monovalent atoms or radicals (see STEREO-ISOMERISM). The equivalence of the four hydrogen atoms of methane rested on indirect evidence, e.g. the existence of only one acetic acid, methyl chloride, and other monosubstitution derivatives—until the experimental proof by L. Henry (Zeit. f. Phys. Chem., 1888, 2, p. 553), who prepared the four nitromethanes, CH3NO2, each atom in methane being successively replaced by the nitro-group.

Henry started with methyl iodide, the formula of which we write in the form CI{a}H{b}H{c}H{d}. This readily gave with silver nitrite a nitromethane in which we may suppose the nitro-group to replace the a iodine atom, i.e. C(NO2){a}H{b}H{c}H{d}. The same methyl iodide gave with potassium cyanide, acetonitril, which was hydrolysed to acetic acid; this must be C(COOH){a}H{b}H{c}H{d}. Chlorination of this substance gave a monochloracetic acid; we will assume the chlorine atom to replace the b hydrogen atom. This acid with silver nitrite gave nitroacetic acid, which readily gave the second nitromethane, CH{a}(NO2){b}H{c}Hd identical with the first nitromethane. From the nitroacetic acid obtained above, malonic acid was prepared, and from this a monochlormalonic acid was obtained; we assume the chlorine atom to replace the c hydrogen atom. This acid gives with silver nitrite the corresponding nitromalonic acid, which readily yielded the third nitromethane, CH{a}H{b}(NO2){c}H{d}, also identical with the first. The fourth nitromethane was obtained from the nitromalonic acid previously mentioned by a repetition of the method by which the third was prepared; this was identical with the other three.

Let us now consider hydrocarbons containing 2 atoms of carbon. Three such compounds are possible according to the number of valencies acting directly between the carbon atoms. Thus, if they are connected by one valency, and the remaining valencies saturated by hydrogen, we obtain the compound H3C.CH3, ethane. This compound may be considered as derived from methane, CH4, by replacing a hydrogen atom by the monovalent group CH3, known as methyl; hence ethane may be named "methylmethane." If the carbon atoms are connected by two valencies, we obtain a compound H2C:CH2, ethylene; if by three valencies, HC:CH, acetylene. These last two compounds are termed unsaturated, whereas ethane is saturated. It is obvious that we have derived three combinations of carbon with hydrogen, characterized by containing a single, double, and triple linkage; and from each of these, by the substitution of a methyl group for a hydrogen atom, compounds of the same nature result. Thus ethane gives H3C.CH2.CH3, propane; ethylene gives H2C:CH.CH3, propylene; and acetylene gives HC:C.CH3, allylene. By continuing the introduction of methyl groups we obtain three series of homologous hydro-carbons given, by the general formulae C{n}H{2n+2}, C{n}H{2n}, and C{n}H{2n-2}, each member differing from the preceding one of the same series by CH2. It will be noticed that compounds containing two double linkages will have the same general formula as the acetylene series; such compounds are known as the "diolefines." Hydrocarbons containing any number of double or triple linkages, as well as both double and triple linkages, are possible, and a considerable number of such compounds have been prepared.

A more complete idea of the notion of a compound radical follows from a consideration of the compound propane. We derived this substance from ethane by introducing a methyl group; hence it may be termed "methylethane." Equally well we may derive it from methane by replacing a hydrogen atom by the monovalent group CH2.CH3, named ethyl; hence propane may be considered as "ethylmethane." Further, since methane may be regarded as formed by the conjunction of a methyl group with a hydrogen atom, it may be named "methyl hydride"; similarly ethane is "ethyl hydride," propane, "propyl hydride," and so on. The importance of such groups as methyl, ethyl, &c. in attempting a nomenclature of organic compounds cannot be overestimated; these compound radicals, frequently termed alkyl radicals, serve a similar purpose to the organic chemist as the elements to the inorganic chemist.

In methane and ethane the hydrogen atoms are of equal value, and no matter which one may be substituted by another element or group the same compound will result. In propane, on the other hand, the hydrogen atoms attached to the terminal carbon atoms differ from those joined to the medial atom; we may therefore expect to obtain different compounds according to the position of the hydrogen atom substituted. By introducing a methyl group we may obtain CH3.CH2.CH2.CH3, known as "normal" or n-butane, substitution occurring at a terminal atom, or CH3.CH(CH3).CH3, isobutane, substitution occurring at the medial atom. From n-butane we may derive, by a similar substitution of methyl groups, the two hydrocarbons: (1) CH3.CH2.CH2.CH2.CH3, and (2) CH3.CH(CH3).CH2.CH3; from isobutane we may also derive two compounds, one identical with (2.), and a new one (3) CH3(CH3)C(CH3)CH3. These three hydrocarbons are isomeric, i.e. they possess the same formula, but differ in constitution. We notice that they may be differentiated as follows: (1) is built up solely of methyl and .CH2. (methylene) groups and the molecule consists of a single chain; such hydrocarbons are referred to as being normal; (2) has a branch and contains the group :CH (methine) in which the free valencies are attached to carbon atoms; such hydrocarbons are termed secondary or iso-; (3) is characterized by a carbon atom linked directly to four other carbon atoms; such hydrocarbons are known as tertiary.

Deferring the detailed discussion of cyclic or ringed hydrocarbons, a correlation of the various types or classes of compounds which may be derived from hydrocarbon nuclei will now be given. It will be seen that each type depends upon a specific radical or atom, and the copulation of this character with any hydrocarbon radical (open or cyclic) gives origin to a compound of the same class.

It is convenient first to consider the effect of introducing one, two, or three hydroxyl (OH) groups into the -CH3, >CH2, and ->CH groups, which we have seen to characterize the different types of hydrocarbons. It may be noticed here that cyclic nuclei can only contain the groups >CH2 and ->CH, the first characterizing the polymethylene and reduced heterocyclic compounds, the second true aromatic compounds.

Substituting one hydroxyl group into each of these residues, we obtain radicals of the type -CH2.OH, >CH.OH, and ->C.OH; these compounds are known as alcohols (q.v.), and are termed primary, secondary, and tertiary respectively. Polymethylenes can give only secondary and tertiary alcohols, benzene only tertiary; these latter compounds are known as phenols. A second hydroxyl group may be introduced into the residues -CH2.OH and >CH.OH, with the production of radicals of the form -CH(OH)2 and >C(OH)2. Compounds containing these groupings are, however, rarely observed (see CHLORAL), and it is generally found that when compounds of these types are expected, the elements of water are split off, and the typical groupings are reduced to -CH:O and >C:O. Compounds containing the group -CH:O are known as aldehydes (q.v.), while the group >C:O (sometimes termed the carbonyl or keto group) characterizes the ketones (q.v.). A third hydroxyl group may be introduced into the -CH:O residue with the formation of the radical -C(OH):O; this is known as the carboxyl group, and characterizes the organic acids.

Sulphur analogues of these oxygen compounds are known. Thus the thio-alcohols or mercaptans (q.v.) contain the group -CH2.SH; and the elimination of the elements of sulphuretted hydrogen between two molecules of a thio-alcohol results in the formation of a thio-ether or sulphide, R2S. Oxidation of thio-ethers results in the formation of sulphoxides, R2:S:O, and sulphones, R2:SO2; oxidation of mercaptans yields sulphonic acids, R.SO3H, and of sodium mercaptides sulphinic acids, R.SO(OH). We may also notice that thio-ethers combine with alkyl iodides to form sulphine or sulphonium compounds, R3:SI. Thio-aldehydes, thio-ketones and thio-acids also exist.

We proceed to consider various simple derivatives of the alcohols, which we may here regard as hydroxy hydrocarbons, R.OH, where R is an alkyl radical, either aliphatic or cyclic in nature.

Of these, undoubtedly the simplest are the ethers (q.v.), formed by the elimination of the elements of water between two molecules of the same alcohol, "simple ethers," or of different alcohols, "mixed ethers." These compounds may be regarded as oxides in just the same way as the alcohols are regarded as hydroxides. In fact, the analogy between the alkyl groups and metallic elements forms a convenient basis from which to consider many derivatives. Thus from ethyl alcohol there can be prepared compounds, termed esters (q.v.), or ethereal salts, exactly comparable in structure with corresponding salts of, say, potassium; by the action of the phosphorus haloids, the hydroxyl group is replaced by a halogen atom with the formation of derivatives of the type R.Cl(Br,I); nitric acid forms nitrates, R.O.NO2; nitrous acid, nitrites, R.O.NO; sulphuric acid gives normal sulphates R2SO4, or acid sulphates, R.SO4H. Organic acids also condense with alcohols to form similar compounds: the fats, waxes, and essential oils are naturally occurring substances of this class.

An important class of compounds, termed amines (q.v.), results from the condensation of alcohols with ammonia, water being eliminated between the alcoholic hydroxyl group and a hydrogen atom of the ammonia. Three types of amines are possible and have been prepared: primary, R.NH2; secondary, R2:NH; and tertiary, R3:N; the oxamines, R3N:O, are closely related to the tertiary ammonias, which also unite with a molecule of alkyl iodide to form salts of quaternary ammonium bases, e.g. R4N.I. It is worthy of note that phosphorus and arsenic bases analogous to the amines are known (see PHOSPHORUS and ARSENIC). From the primary amines are derived the diazo compounds (q.v.) and azo compounds (q.v.); closely related are the hydrazines (q.v.). Secondary amines yield nitrosamines, R2N.NO, with nitrous acid. By the action of hydroxylamine or phenylhydrazine on aldehydes or ketones, condensation occurs between the carbonyl oxygen of the aldehyde or ketone and the amino group of the hydroxylamine or hydrazine. Thus with hydroxylamine aldehydes yield aldoximes, R.CH:N.OH, and ketones, ketoximes, R2C:N.OH (see OXIMES), while phenyl hydrazine gives phenylhydrazones, R2C:N.NH.C6H5 (see HYDRAZONES). Oxyaldehydes and oxyketones (viz. compounds containing an oxy in addition to an aldehydic or ketonic group) undergo both condensation and oxidation when treated with phenylhydrazine, forming compounds known as osozones; these are of great importance in characterizing the sugars (q.v.).

The carboxyl group constitutes another convenient starting-point for the orientation of many types of organic compounds. This group may be considered as resulting from the fusion of a carbonyl (:CO) and a hydroxyl (HO.) group; and we may expect to meet with compounds bearing structural resemblances to the derivatives of alcohols and aldehydes (or ketones).

Considering derivatives primarily concerned with transformations of the hydroxyl group, we may regard our typical acid as a fusion of a radical R.CO- (named acetyl, propionyl, butyl, &c., generally according to the name of the hydrocarbon containing the same number of carbon atoms) and a hydroxyl group. By replacing the hydroxyl group by a halogen, acid-haloids result; by the elimination of the elements of water between two molecules, acid-anhydrides, which may be oxidized to acid-peroxides; by replacing the hydroxyl group by the group .SH, thio-acids; by replacing it by the amino group, acid-amides (q.v.); by replacing it by the group -NH.NH2, acid-hydrazides. The structural relations of these compounds are here shown:

R.CO.OH; R.CO.Cl; (R.CO)2O; R.CO.SH; acid; acid-chloride; acid-anhydride; thio-acid;

R.CO.NH2; R.CO.NH.NH2. acid-amide; acid-hydrazide.

It is necessary clearly to distinguish such compounds as the amino- (or amido-) acids and acid-amides; in the first case the amino group is substituted in the hydrocarbon residue, in the second it is substituted in the carboxyl group.

By transformations of the carbonyl group, and at the same time of the hydroxyl group, many interesting types of nitrogen compounds may be correlated.

Thus from the acid-amides, which we have seen to be closely related to the acids themselves, we obtain, by replacing the carbonyl oxygen by chlorine, the acidamido-chlorides, R.CCl2.NH2, from which are derived the imido-chlorides, R.CCl:NH, by loss of one molecule of hydrochloric acid. By replacing the chlorine in the imido-chloride by an oxyalkyl group we obtain the imido-ethers, R.C(OR'):NH; and by an amino group, the amidines, R.C(NH2):NH. The carbonyl oxygen may also be replaced by the oxime group, :N.OH; thus the acids yield the hydroxamic acids, R.C(OH):NOH, and the acid-amides the amidoximes, R.C(NH2):NOH. Closely related to the amidoximes are the nitrolic acids, R.C(NO2):NOH.

Cyclic Hydrocarbons and Nuclei.

Having passed in rapid review the various types of compounds derived by substituting for hydrogen various atoms or groups of atoms in hydrocarbons (the separate articles on specific compounds should be consulted for more detailed accounts), we now proceed to consider the closed chain compounds. Here we meet with a great diversity of types: oxygen, nitrogen, sulphur and other elements may, in addition to carbon, combine together in a great number of arrangements to form cyclic nuclei, which exhibit characters closely resembling open-chain compounds in so far as they yield substitution derivatives, and behave as compound radicals. In classifying closed chain compounds, the first step consists in dividing them into: (1) _carbocyclic_, in which the ring is composed solely of carbon atoms—these are also known as _homocyclic_ or _isocyclic_ on account of the identity of the members of the ring—and (2) _heterocyclic_, in which different elements go to make up the ring. Two primary divisions of carbocyclic compounds may be conveniently made: (1) those in which the carbon atoms are completely saturated—these are known by the generic term _polymethylenes_, their general formula being (CH2)_n: it will be noticed that they are isomeric with ethylene and its homologues; they differ, however, from this series in not containing a double linkage, but have a ringed structure; and (2) those containing fewer hydrogen atoms than suffice to saturate the carbon valencies—these are known as the _aromatic compounds_ proper, or as _benzene compounds_, from the predominant part which benzene plays in their constitution.

It was long supposed that the simplest ring obtainable contained six atoms of carbon, and the discovery of trimethylene in 1882 by August Freund by the action of sodium on trimethylene bromide, Br(CH2)3Br, came somewhat as a surprise, especially in view of its behaviour with bromine and hydrogen bromide. In comparison with the isomeric propylene, CH3.HC:CH2, it is remarkably inert, being only very slowly attacked by bromine, which readily combines with propylene. But on the other hand, it is readily converted by hydrobromic acid into normal propyl bromide, CH3.CH2.CH2Br. The separation of carbon atoms united by single affinities in this manner at the time the observation was made was altogether without precedent. A similar behaviour has since been noticed in other trimethylene derivatives, but the fact that bromine, which usually acts so much more readily than hydrobromic acid on unsaturated compounds, should be so inert when hydrobromic acid acts readily is one still needing a satisfactory explanation. A great impetus was given to the study of polymethylene derivatives by the important and unexpected observation made by W.H. Perkin, junr., in 1883, that ethylene and trimethylene bromides are capable of acting in such a way on sodium acetoacetic ester as to form tri- and tetra-methylene rings. Perkin has himself contributed largely to our knowledge of such compounds; penta- and hexa-methylene derivatives have also received considerable attention (see POLYMETHYLENES).

A. von Baeyer has sought to explain the variations in stability manifest in the various polymethylene rings by a purely mechanical hypothesis, the "strain" or Spannungs theory (Ber., 1885, p. 2277). Assuming the four valencies of the carbon atom to be directed from the centre of a regular tetrahedron towards its four corners, the angle at which they meet is 109 deg. 28'. Baeyer supposes that in the formation of carbon "rings" the valencies become deflected from their positions, and that the tension thus introduced may be deduced from a comparison of this angle with the angles at which the strained valencies would meet. He regards the amount of deflection as a measure of the stability of the "ring." The readiness with which ethylene is acted on in comparison with other types of hydrocarbon, for example, is in harmony, he considers, with the circumstance that the greatest distortion must be involved in its formation, as if deflected into parallelism each valency will be drawn out of its position through 1/2.109 deg. 28'. The values in other cases are calculable from the formula 1/2(1O9 deg. 28' - a), where a is the internal angle of the regular polygon contained by sides equal in number to the number of the carbon atoms composing the ring. These values are:—

Trimethylene. 1/2(109 deg. 28' - 60 deg.) = 24 deg. 44'. Tetramethylene. 1/2(109 deg. 28' - 90 deg.) = 9 deg. 44'. Pentamethylene. 1/2(109 deg. 28' - 108 deg.) = 0 deg. 44'. Hexamethylene. 1/2(109 deg. 28' - 120 deg.) = -5 deg. 16'.

The general behaviour of the several types of hydrocarbons is certainly in accordance with this conception, and it is a remarkable fact that when benzene is reduced with hydriodic acid, it is converted into a mixture of hexamethylene and methylpentamethylene (cf. W. Markownikov, Ann., 1898, 302, p. 1); and many other cases of the conversion of six-carbon rings into five-carbon rings have been recorded (see below, Decompositions of the Benzene Ring). Similar considerations will apply to rings containing other elements besides carbon. As an illustration it may be pointed out that in the case of the two known types of lactones—the [gamma]-lactones, which contain four carbon atoms and one oxygen atom in the ring, are more readily formed and more stable (less readily hydrolysed) than the [delta]-lactones, which contain one oxygen and five carbon atoms in the ring. That the number of atoms which can be associated in a ring by single affinities is limited there can be no doubt, but there is not yet sufficient evidence to show where the limit must be placed. Baeyer has suggested that his hypothesis may also be applied to explain the instability of acetylene and its derivatives, and the still greater instability of the polyacetylene compounds.

Benzene.

The ringed structure of benzene, C6H6, was first suggested in 1865 by August Kekule, who represented the molecule by six CH groups placed at the six angles of a regular hexagon, the sides of which denoted the valencies saturated by adjacent carbon atoms, the fourth valencies of each carbon atom being represented as saturated along alternate sides. This formula, notwithstanding many attempts at both disproving and modifying it, has well stood the test of time; the subject has been the basis of constant discussion, many variations have been proposed, but the original conception of Kekule remains quite as convenient as any of the newer forms, especially when considering the syntheses and decompositions of the benzene complex. It will be seen, however, that the absolute disposition of the fourth valency may be ignored in a great many cases, and consequently the complex may be adequately represented as a hexagon. This symbol is in general use; it is assumed that at each corner there is a CH group which, however, is not always written in; if a hydrogen atom be substituted by another group, then this group is attached to the corner previously occupied by the displaced hydrogen. The following diagrams illustrate these statements:—

CH ^ C.OH OH / / / / HC / CH / HC / CH / HC // CH / HC // CH / // / // / CH v CH v

Benzene. Abbreviated. Oxybenzene. Abbreviated.

From the benzene nucleus we can derive other aromatic nuclei, graphically represented by fusing two or more hexagons along common sides. By fusing two nuclei we obtain the formula of naphthalene, C10H10; by fusing three, the hydrocarbons anthracene and phenanthrene, C14H10; by fusing four, chrysene, C18H12, and possibly pyrene, C16H10; by fusing five, picene, C22H14. But it must be here understood that each member of these condensed nuclei need not necessarily be identical in structure; thus the central nuclei in anthracene and phenanthrene differ very considerably from the terminal nuclei (see below, Condensed Nuclei). Other hydrocarbon nuclei generally classed as aromatic in character result from the union of two or more benzene nuclei joined by one or two valencies with polymethylene or oxidized polymethylene rings; instances of such nuclei are indene, hydrindene, fluorene, and fluor-anthene. From these nuclei an immense number of derivatives may be obtained, for the hydrogen atoms may be substituted by any of the radicals discussed in the preceding section on the classification of organic compounds.

Distinctions between aliphatic and aromatic compounds.

We now proceed to consider the properties, syntheses, decompositions and constitution of the benzene complex. It has already been stated that benzene derivatives may be regarded as formed by the replacement of hydrogen atoms by other elements or radicals in exactly the same manner as in the aliphatic series. Important differences, however, are immediately met with when we consider the methods by which derivatives are obtained. For example: nitric acid and sulphuric acid readily react with benzene and its homologues with the production of nitro derivatives and sulphonic acids, while in the aliphatic series these acids exert no substituting action (in the case of the olefines, the latter acid forms an addition product); another distinction is that the benzene complex is more stable towards oxidizing agents. This and other facts connected with the stability of benzenoid compounds are clearly shown when we consider mixed aliphatic-aromatic hydrocarbons, i.e. compounds derived by substituting aliphatic radicals in the benzene nucleus; such a compound is methylbenzene or toluene, C6H5.CH3. This compound is readily oxidized to benzoic acid, C6H5.COOH, the aromatic residue being unattacked; nitric and sulphuric acids produce nitro-toluenes, C6H4.CH3.NO2, and toluene sulphonic acids, C6H4.CH3.SO3H; chlorination may result in the formation of derivatives substituted either in the aromatic nucleus or in the side chain; the former substitution occurs most readily, chlor-toluenes, C6H4.CH3.Cl, being formed, while the latter, which needs an elevation in temperature or other auxiliary, yields benzyl chloride, C6H5.CH2Cl, and benzal chloride, C6H5.CHCl2. In general, the aliphatic residues in such mixed compounds retain the characters of their class, while the aromatic residues retain the properties of benzene.

Further differences become apparent when various typical compounds are compared. The introduction of hydroxyl groups into the benzene nucleus gives rise to compounds generically named phenols, which, although resembling the aliphatic alcohols in their origin, differ from these substances in their increased chemical activity and acid nature. The phenols more closely resemble the tertiary alcohols, since the hydroxyl group is linked to a carbon atom which is united to other carbon atoms by its remaining three valencies; hence on oxidation they cannot yield the corresponding aldehydes, ketones or acids (see below, Decompositions of the Benzene Ring). The amines also exhibit striking differences: in the aliphatic series these compounds may be directly formed from the alkyl haloids and ammonia, but in the benzene series this reaction is quite impossible unless the haloid atom be weakened by the presence of other substituents, e.g. nitro groups. Moreover, while methylamine, dimethylamine, and trimethylamine increase in basicity corresponding to the introduction of successive methyl groups, phenylamine or aniline, diphenylamine, and triphenylamine are in decreasing order of basicity, the salts of diphenylamine being decomposed by water. Mixed aromatic-aliphatic amines, both secondary and tertiary, are also more strongly basic than the pure aromatic amines, and less basic than the true aliphatic compounds; e.g. aniline, C6H5.NH2, monomethyl aniline, C6H5.NH.CH3, and dimethyl aniline, C6H5.N(CH3)2, are in increasing order of basicity. These observations may be summarized by saying that the benzene nucleus is more negative in character than the aliphatic residues.

Isomerism of Benzene Derivatives.—Although Kekule founded his famous benzene formula in 1865 on the assumptions that the six hydrogen atoms in benzene are equivalent and that the molecule is symmetrical, i.e. that two pairs of hydrogen atoms are symmetrically situated with reference to any specified hydrogen atom, the absolute demonstration of the validity of these assumptions was first given by A. Ladenburg in 1874 (see Ber., 1874, 7, p. 1684; 1875, 8, p. 1666; Theorie der aromatischen Verbindungen, 1876). These results may be graphically represented as follows: numbering the hydrogen atoms in cyclical order from 1 to 6, then the first thesis demands that whichever atom is substituted the same compound results, while the second thesis points out that the pairs 2 and 6, and 3 and 5 are symmetrical with respect to 1, or in other words, the di-substitution derivatives 1.2 and 1.6, and also 1.3 and 1.5 are identical. Therefore three di-derivatives are possible, viz. 1.2 or 1.6, named ortho- (o), 1.3 or 1.5, named meta- (m), and 1.4, named para- compounds (p). In the same way it may be shown that three tri-substitution, three tetra-substitution, one penta-substitution, and one hexa-substitution derivative are possible. Of the tri-substitution derivatives, 1.2.3.-compounds are known as "adjacent" or "vicinal" (v), the 1.2.4 as "asymmetrical" (as), the 1.3.5 as "symmetrical" (s); of the tetra-substitution derivatives, 1.2.3.4-compounds are known as "adjacent," 1.2.3.5 as "asymmetrical," and 1.2.4.5 as "symmetrical."

Di-derivatives Tri-derivatives Tetra-derivatives / / / X X X X X X X X X / / / / / / / / / / X / / / X / X / / X / X / X / /X / /X / X /X /X X /X X / / / / / / / / / / X X X X o m p v as s v as s

Here we have assumed the substituent groups to be alike; when they are unlike, a greater number of isomers is possible. Thus in the tri-substitution derivatives six isomers, and no more, are possible when two of the substituents are alike; for instance, six diaminobenzoic acids, C6H3(NH2)2COOH, are known; when all are unlike ten isomers are possible; thus, ten oxytoluic acids, C6H3.CH3.OH.COOH, are known. In the case of tetra-substituted compounds, thirty isomers are possible when all the groups are different.

Equivalence of four hydrogen atoms.

The preceding considerations render it comparatively easy to follow the reasoning on which the experimental verification of the above statements is based. The proof is divided into two parts: (1) that four hydrogen atoms are equal, and (2) that two pairs of hydrogen atoms are symmetrical with reference to a specified hydrogen atom. In the first thesis, phenol or oxybenzene, C6H5.OH, in which we will assume the hydroxyl group to occupy position 1, is converted into brombenzene, which is then converted into benzoic acid, C6H5.COOH. From this substance, an oxybenzoic acid (meta-), C6H4.OH.COOH, may be prepared; and the two other known oxybenzoic acids (ortho- and para-) may be converted into benzoic acid. These three acids yield on heating phenol, identical with the substance started with, and since in the three oxybenzoic acids the hydroxyl groups must occupy positions other than 1, it follows that four hydrogen atoms are equal in value.