When the electric light is generated within an atmosphere of hydrogen, then, at the immense temperature of the electric arc, part of the carbon of the charcoal terminals unites with the hydrogen into acetylene gas, C 2 H 2 , Apart from this isolated fact, which was discovered by Berthelot in 1862, it might be said that the two elements are not capable of uniting directly, although an innumerable variety of hydrocarbons exist in nature, and can be produced artificially from organic substances. Individual hydrocarbons may differ very much in their properties. At ordinary temperature and pressure a few are gases; the majority present themselves as liquids; not a few are solids. But the solids are fusible; and all liquid or liquefied hydrocarbons, at a high enough temperature, volatilize, as a rule without decomposition. To the latter circumstance to a great extent we owe our precise knowledge of their chemical constitution.

In all the numerous series of hydrocarbons the percentages of carbon vary from 75 (in marsh gas) to 94 - 7 (in chrysene). Within this narrow range of some 20 per cent, several dozens of elementary compositions have to be accommodated; and many of these, to be represented in formulae C x H 2 , with an adequate degree of precision, require formulae in which the coefficients x and y are so large that, by means of integers less than these, any fancy composition (within our limits) may be expressed with a degree of exactitude which is quite on a par with the analyses. But these hydrocarbons, in general, can be volatilized into gases, and in regard to these Avogadro’s law tells us that quantities proportional to the molecular weights (i.e., the weights represented by the true chemical formulae) occupy the same volume. Hence, to find the true value, M = C x H i ,, of the formula as a whole, we need only determine the vapour density, and from it calculate the weight of the respective hydrocarbon which, as a gas at t a and P millimetres pressure, occupies the same volume as, for instance, H 2 O parts of steam. This is Μ. The elementary analysis enables us to calculate the weight x × C of carbon contained in Μ parts, and the analysis must be very poor to leave us in doubt as to whether it is for instance 6 × 12 parts of carbon or 7 × 12 parts that we have to deal with. The reader will now understand how it has been possible to ascertain the elementary composition of all pure hydrocarbons with a degree of precision which goes beyond that of the analysis, and to prove what analysis could never have done by itself, namely, that there are numerous groups of hydrocarbons which have absolutely identical elementary compositions,—cases of isomerism, as they are called. We speak of " isomerism in the narrower sense” when the atomic formulae are identical (there are, for instance, two hydrides of butyl, C 4 H 10 ), while we speak of “polymeric” bodies when the several formulae are integer multiples of the same primitive group (e.g., ethylene, 2×CH 2 , and butylene, 4×CH 2 , are polymers to one another).

The following table gives an idea of the several classes of hydrocarbons which for us come more particularly into consideration. [table]

The first column, under “n,” gives the number of carbon atoms per molecule in the compounds whose formulae stand in that horizontal line,—these latter being arranged in a descending series according to the number of hydrogen atoms united with n atoms of carbon. Instead of pointing out those regularities, in regard to the atomic proportions in which carbon and hydrogen can unite into compounds, which the table illustrates so forcibly, let us rather state that the “benzols,” in opposition to all that stands to their left in the table, are things of their own kind. In them six atoms of the carbon are most firmly united (into a “ring,” as a certain theory says), and the rest are, so to say, hooked on to the ring in a less intimate fashion. Thus benzol is (C 6 )H 6 ; each one of the six H’s being tied to one of the six C’s; toluol is (C 6 H 5 )—CH 3 ; it is a benzol from which one of the six hydrogen atoms has been removed, and in which the gap left has been filled by a “methyl,” CH 3 :— g 6 h 6 +ch 4 =h 2 +(C 6 H 6 )-(CH 3 ).

Benzol. Marsh gas.

But similarly two dehydrogenated benzols, C 6 H 5 , can unite into one double ring of diphenyl: 2C 6 H 6 - 2H = (C 6 H 5 )(C 6 H 5 ); and two benzol rings may unite more firmly in such a manner that two carbon atoms of the one ring do sendee for the two rings, and a double ring is formed firmly united by these two common carbons, the four hydrogens of the original two benzols being away. This gives naphthalene :—

C 6 H 6 + C 6 H c - 2C - 4H = C 10 H 8 .

Benzol. Naphthalene.

In a similar manner three benzols may unite into one anthracene;—

C 6 H β + C 6 Hθ + C β H β - 4C - 8H - C 1 .H 10 .

Benzol. Anthracene.

Generally speaking, a hydrocarbon is the more volatile the less the number of carbon atoms and the greater the number of hydrogen atoms in the molecule. Thus, in the series of “paraffins,” CH 4 (marsh gas) and C 2 H 6 (ethane) are gases, C 3 H 8 (propane) and C 4 H 10 (butane) are very volatile liquids, and C 5 H 12 , <fcc., are liquids,—with higher and higher boiling points as we ascend the series. From a certain value of n upwards we find ourselves amongst the paraffins proper, which are solids, more or less easily fusible, but not, in general, volatile without decomposition. Benzol, C 6 H 6 , and its neighbouring homologues are volatile liquids. Naphthalene and anthracene are crystalline solids, fusible at 79° ’ 2 and 180° C., and boiling at 217° and above 300° C. respectively without decomposition.

All hydrocarbons agree in this, that they are practically insoluble in water, but more or less readily soluble (in general) in alcohol and in ether. They are all combustible; the more readily volatile ones are inflammable. Any complete combustion, of course, leads to the formation of only carbonic acid and water, with evolution of a large amount of heat; but the mechanism of the process is more or less complex. Naphthalene and anthracene remain undecomposed at a red heat; only at the very high temperature of their flames, and by the co-operation of the oxygen of the air, they are decomposed with large elimination of charcoal; a similar, though less, stability is exhibited by the benzols. The paraffins, on the other hand, are relatively unstable. Marsh gas, it is true, stands a red heat; but, to pass to the other end of the series, the paraffins proper, and also the higher liquid paraffins to some extent, even when being distilled, and especially when distilled “under pressure,” i.e., at higher temperatures than their natural boiling points, break up into olefines and lower paraffins (Thorpe and John Young). Similar changes take place when the vapours of paraffins are passed through red-hot tubes; only the products formed then suffer deeper-going decomposition with formation of hydrogen, marsh-gas, acetylene, ethylene, and charcoal, and, last not least, benzols and naphthalene. To this latter fact the paraffins owe their pre-eminent fitness as illuminating agents.

When organoid minerals, such as cannel coal, shale, &c., al·e subjected to dry distillation, all the several classes of hydrocarbons are in general produced at the same time; but, from what we have said it will be understood that, even with the same material, the quantitative composition of the complex vapour which comes out of the retort depends on the way in which the distillation is being conducted. If we operate at the lowest practicable temperature, comparatively little gas is produced, and in the condensible part of the vapour the paraffins predominate largely; at a bright red heat, such as is used in making coal gas, and especially if the vapours have to pass along red-hot surfaces before they get into the condenser pipes, more gas is produced, and the place of the liquid paraffins is taken by benzols. These latter, however, are always accompanied by naphthalene, often also by anthracene, and invariably by certain ternary benzol-derivatives, namely, by “phenols,” feebly acid bodies containing hydroxyl groups, OH’s, where the corresponding hydrocarbon bore plain hydrogens (ordinary phenol, C 6 H 5 (OH), derived from benzol, C 6 H 5 H, is a representative example), and, secondly, basic compounds of carbon, hydrogen, and nitrogen. Of the latter aniline and picoline— both C 6 H 7 N, but widely different in their properties —may be quoted as examples. The gas produced in this case through the presence in it of the vapour of higher hydrides, but especially of acetylene, C 2 H 2 , and benzol is highly luminous. Supposing now, as a third instance, the distillation to be conducted at a white heat, and so that the primary vapour has to wind its way through a spiral pipe kept at a bright red heat, the proportion of gas increases largely, and there is an increased yield of retort charcoal; but the liquid hydrocarbons of all classes almost vanish; the gas consists mainly of hydrogen, marsh gas, carbonic oxide, and carbonic acid, and gives little light when kindled.

The aim of the paraffin oil manufacturer is to produce the best possible approximation to a mixture of paraffins, wherefore he conducts his distillation at the lowest working temperature. Of course his paraffin mixture contains more or less of the other classes of bodies referred to, whose removal, however, offers no great difficulty. In the laboratory we should commence by shaking the crude oil with caustic alkali ley, which withdraws the phenols and other acid bodies, as part of a lower layer, the upper being purified oil. By shaking the latter with dilute sulphuric acid the bases are removed as a solution of their sulphates, and a still purer oil results. Application of concentrated sulphuric acid to the latter removes part at least of the benzols and olefines as sulpho-acids, and also of the phenols and all the bases, should the two preceding operations have been omitted. But the most thorough mode of getting quit of the benzols and their derivatives is —after having exhausted the milder agents—to shake the oil with first aqueous and then stronger and stronger nitric acid, which reagent converts the benzol-bodies into nitroproducts, soluble in the acid, or removable, after separation of the acid layer, by aqueous alkali. By all these tortures the paraffins—being what the name implies—are not much affected, so that what ultimately survives all belongs to their family. The separation of the individual paraffins from one another is a very difficult problem which has not yet found a satisfactory solution. What we know of individual paraffins is derived chiefly from the investigation of decompositions of pure chemical substances leading to the formation of that one paraffin principally if not solely. To split up a mixture of paraffins approximately the only known method is fractional distillation (see Distillation, vol. vii. p. 260), preferably by means of an apparatus so constructed that the vapour, before reaching the condenser, ascends through an intermediate inverted condenser or still-head, and there suffers partial condensation at some suitable temperature (enforced in the most perfect form of the apparatus by an oil-bath surrounding the stillhead). In this latter case, singularly—not as a matter of course by any means—what goes over boils very nearly at the temperature of the still-head. This particular form of the method therefore lends itself chiefly for the final purification of an unitary substance of known boiling point already purified by preceding distillations. With mixtures of unknown composition the process is very tedious, and may assume something like this form.

We distil the substance (slowly and with ample chance of partial condensation) and collect as separate fractions what came over at, for instance, 100° to 105°, 105° to 110°, 110° to 115°, <fcc., as I., II., III., IV., &c. Each of these when redistilled yields I. and II. and III. and IV., iιc., which parts are poured into the respective receptacles, and on this principle we continue working. If the substance happens to be of comparatively simple composition, it usually turns out, after a while, that (say) the two fractions II. and VI. increase while the rest get less and less; and by working on we may be able to isolate two bodies of the constant boiling points t 2 and t 6 respectively, with formation of “tails” of other boiling points. Unfortunately, even a constant boiling point is no proof of chemical purity; and, if a constant-boiling substance is a mixture, only chemical methods can help us out of the difficulty.

The following table (extracted from Roscoe and Schorlemmer’s Handbook of Chemistry, German edition) gives the names, specific gravities, and boiling points of the more important paraffins. The first column, “rø,” gives the number of carbon-atoms in the molecule, and consequently the molecular weight Μ and the vapour density S. In the case of “pentan,” for instance, we have n = 5; hence M = C 5 H 12 = 72j and, as H 2 = 2, the gas-density, referred to hydrogen = S = 36, while, as air is 14'45 times as heavy as hydrogen, for the gas-density referred to air the value

⅜M÷14∙45 = 36÷14∙45 = 2'491. [table]

Probably all the paraffins enumerated in the table are present in paraffin oil and in petroleum; those marked * have been actually found in the one or the other. The solid paraffins are not known as unitary chemical substances; no chemist as yet has succeeded in splitting up solid paraffin into its proximate components. The manufacturer, in regard to the liquid paraffins even, does not trouble himself with the isolation of chemical species; he contents himself with splitting up his oil into fractions corresponding to certain ranges of boiling point, and consequently adapted to certain practical applications. But even the boiling point is not much heeded industrially; the several kinds of oil are defined by their specific gravity at 60° F., which, as experience shows, increases as the boiling point rises. But it is as well here to point out that the same (initial) boiling point even, and in a much higher degree the same specific gravity, may be exhibited by oils of widely different proximate composition. Hence a relatively (and in a sense sufficiently) high specific gravity is no guarantee against dangerous inflammability; the degree of inflammability in an oil must be—and in practice always is being—determined by direct experiment. For this purpose it is not sufficient to heat a sample oil in an open vessel gradually to higher and higher temperatures, and to note the temperature at which the atmosphere over the oil proves inflammable when a lighted taper is brought in contact with it. By this method (which formerly was the universally recognized test) the most varying results may be obtained with the same oil. Far more trustworthy is the close test first proposed by Keates about 1870, the principle of which is to heat the oil within a close vessel which is opened only from time to time to apply a light to its atmosphere. For the execution of this test many varieties of apparatus have been proposed. That adopted by Abel, and now (1884) legally recognized in Great Britain, is made of sheet copper, the exact thickness of which is prescribed for every part. The oil is placed in a close cup, suspended in an air-bath, which latter is heated by immersion in a warmwater bath, provided with an air-jacket. The top of the oil cup is pierced with three circular orifices, one in the centre for trying the best flame, and two smaller lateral holes for admitting air at the dose of each trial. The holes are covered by a slide so contrived that when the central hole has become almost uncovered the lateral ones are also open. The slide carries a small colza-oil lamp suspended on trunnions, having a flame of a prescribed size. A pendulum two feet in length vibrates in front of the observer, who, in testing, withdraws the slide slowly during three vibrations, tilts the lamp to bring its flame in contact with the atmosphere of the vessel, and quickly shuts the slide during the fourth vibration. To execute a test the oil at about 60° F. is placed in the cup, which is immersed in the water-bath having water of 130° F. A thermometer plunged into the oil and another in the water-bath indicate their temperatures. When the oil has approached its presumable flashing point, trials are made at each rise of 1° F. in the temperature of the oil. The lowest temperature at which the atmosphere of the cup inflames is the flashing point of the oil tested. The legal minimum flashing point of burning oil by the close test is 75° F., corresponding to about 100°F. by the obsolete open test.

The variety of mixed paraffins which the oil-distiller produces may be arranged under the following heads :— (1) oils too volatile to be available for domestic illumination, serving chiefly as solvents; (2) burning oils, as required for house lamps; (3) oils of very high boiling point, available, and used chiefly, for lubricating purposes; (4) solid paraffin.

The products of the second class have long come to practically supersede the colza oil which used to be the illuminating oil par excellence. Over it they offer the advantages of greater cheapness and of giving, weight for weight, more light. But their drawbacks are that, however carefully refined they may be, they have, when allowed to leak out, or in lamps of inferior construction, a somewhat disagreeable pungent odour, and that there is always a lurking danger in the possible presence of highly volatile inflammable hydrocarbons. Colza oil will never burn without a wick; paraffin oil or petroleum may do so.

Products of the second and third classes, separately or combinedly, are of course available as fuels proper, i.e., for the production of heat. At the time when mineral oil was first produced in great quantity in America, the advantages it would offer as a fuel for marine boilers especially were very emphatically insisted on. Of course mineral oil can be more economically stored than coal, and its combustion-heat is susceptible of more exhaustive utilization. The latter fact forms the raison d’être of those beautiful petroleum kitchen-stoves and culinary lamps which are very much used on the Continent where gas is not at hand. But to talk of...