All artificial light is obtained as a result either of com-bustion or of incandescence ; or it might be more accurate to classify illuminating agents as those which emit light as a result of chemical action, and those which glow, from the presence of a large amount of heat, without thereby giving rise to any chemical change. The materials whence artificial light of the nature of flame has been derived are principally bodies rich in carbon and hydrogen. Wax, fats, and oils, on exposure to a certain amount of heat, undergo destructive distillation, evolving inflammable gases; and it is really such gases that are consumed in the burning of lamps and candles, the wicks bringing small proportions of the substances into a sufficient heat.

Wood and coal also, when distilled, give off combustible gases; and ordinary gas-lighting only differs from illumina-tion by candles and lamps in the gas being stored up and consumed at a distance from the point where it is generated.

Inflammable gas is formed in great abundance within the earth in connexion with carbonaceous deposits, such as coal and petroleum; and similar accumulations not unfre-quently occur in connexion with deposits of rock-salt; the gases from any of these sources, escaping by means of fis-sures or seams to the open air, may be collected and burned in suitable arrangements. Thus the "eternal fires" of Baku, on the shores of the Caspian Sea, which have been known as burning from remote ages, are due to gaseous hydrocarbons issuing from and through petroleum deposits. In the province of Szechuen in China, gas is obtained from beds of rock-salt at a depth of 1500 or 1600 feet: being brought to the surface, it is conveyed in bamboo tubes and used for lighting as well as for evaporating brine; and it is asserted that the Chinese used this naturally evolved gas as an illuminant long before gas-lighting was introduced among European nations. At a salt mine in the comitat of Marmaro in Hungary, gas is obtained at a depth of about 120 feet, and is used for illuminating the works of the mine. Again at Fredonia (New York State) a natural emission of gas was discovered in a bituminous limestone, over the orifice of which a gasholder has been erected, and thus about 1000 cubic feet of a gas composed of marsh gas and hydride of ethyl has been made available for illumination. In the city of Erie (Pa.) there are 13 gas-wells, each yielding from 10,000 to 30,000 feet per day, the gas escaping from one of them at a pressure of 200 lb per square inch. At Bloomfield, Ontario co., New York, there is a spring which yields daily no less than 800,000 feet of gas of an illumin-ating power equal to 14 J candles. The city of East Liverpool (Ohio) is entirely illuminated, and to a large extent heated, by gas-wells which exist in and around the town. The light is of extraordinary brilliancy, and is so abundant and free that the street lamps are never extin-guished, and much of the manufacturing steam-power of the town, which embraces 22 potteries, giving employment to 2000 hands, is derived from the gas. The first " well," 450 feet deep, was opened in 1859, and up to the present year (1879) neither it nor any of those tapped at later dates show any sign of failing, In many other parts of America similar gas-wells exist; and several such natural jets of gas have been observed in England.

By general consent the merit of the discovery and appli-cation of artificial gas belongs to Great Britain, and the name most honourably connected with the beginning and early stages of gas-lighting is that of a Scotchman—William Murdoch. But previous to Murdoch's time there occur numerous suggestive observations and experiments as to inflammable air and its sources. In the Philosophical Transactions of the Boyal Society for 1667 the existence of a " burning spring " in the coal district of Wigan is noticed by Thomas Shirley, who traced its origin to the underlying coal. In the same Transactions for 1739 is printed a letter addressed to the Hon. Robert Boyle, who died in 1691, in which the Rev. John Clayton details a series of experiments he made in distilling coal in a retort, showing, not only that he had observed the inflammable gases evolved, but that he collected and stored them for some time in bladders. In Dr Stephen Hales's work on Vegetable Staticlcs, published in 1726, more precise state-ments are made as to the distillation of coal, he having obtained from 158 grains of Newcastle coal 180 inches of inflammable air. In 1787 Lord Dundonald, in working a patented process for obtaining coal-tar, experimented with the gas evolved in the process, and occasionally used it for lighting up the hall of Culross Abbey. None of these observations, however, led to distinct practical results; and it was not till the year 1792 that William Murdoch, then residing at Redruth in Cornwall, began the investigations into the properties of gases given off by various substances which eventuated in the establishment of coal-gas as an illuminating agent. In 1797 he publicly showed the system he had matured, and in 1798, being then employed in the famous Soho (Birmingham) workshop of Boulton & Watt, he fitted up an apparatus for the manufacture of gas in that establishment, with which it was partly lighted. Thereafter the apparatus was extended, and the gas manufactured by it was introduced to other neighbouring workshops and factories. Among others who helped most materially to develop the infant art in England were Dr Henry of Manchester, and Mr Clegg, who, succeeding Mr Murdoch at Boulton & Watt's, introduced many improvements in gas manufacture, and ultimately became the most skilful and famous gas engineer in the United Kingdom.

In 1801 M. Lebon introduced gas distilled from wood into his own house in Paris, and the success of his experi-ment attracted so much notice and comment as to give rise to an impression that he is entitled to the credit of the invention. Lebon's experiment came under the notice of Mr F. A. Winsor, who took up the subject with a zeal and unwearying patience which led to a recognition of the advantages of the system, and the breaking down of the powerful prejudice which existed in England against the innovation. In 1803, through Winsor's efforts, the Lyceum Theatre was lighted with gas; but it was not till 1810 that he succeeded in forming a public company for manufacturing gas, and in obtaining an Act of Par-liament for the Gas-Light and Coke Company. In 1813 Westminster Bridge was first lighted with gas, and in the following year the streets of Westminster were thus illuminated, and in 1816 gas became common in London. So rapid was the progress of this new mode of illumination that in the course of a few years after its introduction it was adopted by all the principal towns in the kingdom, for lighting streets as well as shops and public edifices. In private houses it found its way more slowly, partly from an apprehension of danger attending its use, and partly from the annoyance which was experienced in many cases through the careless and imperfect manner in which the service-pipes were at first fitted up.

SOURCES OF GAS.

Artificial gas is now distilled from a variety of substances, among which are coal, shale, lignite, petroleum, turf, wood, resins, oils, and fats; and it is also prepared by earburetting or impregnating with volatile hydrocarbons other non-luminif erous gases. Of the very numerous systems of gas-making which have been proposed since the early part of the century, none can compete for general purposes with the ordinary coal-gas process, when a supply of the raw material can be obtained at a moderate expense.

Coal-Gas.—Coals, varying greatly as they do in chemical constitution, differ also, as might be expected, as widely in their value and applicability for the manufacture of gas. Taking the leading varieties of coal to be included under anthracite, bituminous coal, and lignite or brown coal, we find that it is the class bituminous coal alone that yields varieties really serviceable for gas-making. Anthracite may be regarded as a natural coke from which the volatile con-stituents have been already driven off, and the more anthraeitic any coal is, the less is it capable of yielding gas. Lignite also is rarely used for distillation, owing to the large proportion of oxygen and the amount of water in its com-position. Of the bituminous coals again, it is only the caking or pitch coals, and the cannel or parrot coals, that are in practice used in gas-works. These also vary within very wide limits in their gas-making value, not only from the great difference among them in yield of gas, but also in the illuminating value of the gas they evolve. As a rule the coals which yield the largest percentage produce also the most highly illuminating qualities of gas. The cannel coals, which are specially recognized as " gas-coal," are most abundantly developed in Scotland and in Lanca-shire, and the fact of the unequalled qualities of Scotch cannel and of the allied substance, bituminous shale, for gas-making, has had the effect of rendering illumination by gas much more general and satisfactory in Scotland than in any other country. It is only a very imperfect valuation of any gas-coal that can be made from chemical analysis, the really satisfactory test being actual experiment. According to H. Fleck, the coal most available for gas-making should contain to every 100 parts of carbon 6 parts of hydrogen, of which 4 parts are available for forming hydrocarbon compounds. It is desirable that coal used for distillation in gas retorts should be as far as possible free from sulphur, that in the case of coking coal the amount of ash should be small, and the proportion of oxygen should also be low, since that ele-ment abstracts hydrogen to form injurious watery vapour. The amount of ash present, however, in the best forms of Scotch cannel is large; and consequently the resulting coke, if the residue can be so called, is of comparatively little value. Unless coal can be stored in sheds which protect it from the weather, it ought to be used as soon as possible after being raised, rain and sunshine being detrimental to its gas-mak-ing qualities. The following table exhibits the chemical analysis and gas-yielding properties of a few of the principal and typical examples of coal for gas-making :—_

Composition of Coals used in Gas-Making.

Variety of Coal. Disposable Hydrogen. Carbon. Hydrogen. Nitrogen. Sulphur. Oxygen. À <
Newcastle Peareth Gas-Coal Dukinfield, Ashton-under- |
Moid-Leeswood Green Cannel
Methil Brown Shale h„ "i Kelty Gin Seam j JanQ- 6V88 6-19 5-65
12-68 9-33 82-42 78-06
83-25
84-07 77-81 63-10 66-44 76-50 4-82 5-80
5-75
5-71 8-47 8-91 7-54 5-03 i;85 1-36 0-86 2-22
0-86
0-71 0-96 0-84 0-94 11-11 3-12
5-06
7-82 6-32 7-25 10-84 11-68 0-79 8-94
3-48
2-40
19-78 12-98 2-25
Products of the Distillation of 1 ton of Coal.

Cub. feet of Gas. Lbs. of Coke. Lbs. of Tar. Lbs. of Ammonia Liquor. Illuminat-ing power of Gas in Candles.
Newcastle Cannel...
Boghead Cannel 9,883 10,850 13,334 1,426 1,332 715 98-3 218-3 733-3 60-0 161-6 nil. 25-2 19*4 46-2
Products of the Distillation I. Illuminating Gases.
Acetylene, C2H2 ^
Ethylene, C2H4 („
Propylene, C3H6 huases.
Butylène, C4Hf )
Benzol, CGH6
Naphthalin, Ci0H8 ..
Hydrogen, H
Light carburetted hydro
gen, CH4
Carbonic oxide, CO
Carbonic acid, C02
Ammonia, NH3
Cyanogen, C2N2
Bisulphide of carbon, CS2 . Sulphuretted hydrogen, H2S
Oxygen, O
Nitrogen, N
Aqueous vapour, H20

When the bowl of an ordinary clay pipe is filled with small fragments of bituminous coal, luted over with clay and placed in a bright fire, immediately smoke is seen to issue from the stalk which projects beyond the fire. The smoke soon ceases, and if a light is then applied to the orifice of the stalk, the issuing gas burns with a bright, steady flame, while a proportion of a black, thin, tarry liquid oozes out from the stalk. After the combustion ceases there is left in the bowl of the pipe a quantity of char or coke. This simple operation is, on a small scale, an exact counter-part of the process by which the destructive distillation of coal is accomplished in the manufacture of gas. The pro-ducts of the distillatory process classed in the gas-works as gas, tar, and ammoniacal liquor, with a solid residue of coke, are in themselves mixtures of various definite chemical compounds ; and as may be evident from the following list, these substances are very numerous and complex:—
of Coal at high-red heat.

II. Components of Tar.
Benzol, C6H6 \
Toluol, C7H8 ( Liquid
Cumol, C9H12 f hydrocarbons.
_ Vapours.
Cymol, C,„H14 .
Naphthalin, C10H8 ...\ Anthracene, C14H]0... ( Solid
V Diluents.
Pyrene, C1GH10 f hydrocarbons.
Crysene, C18H12 ;
Im-f purities.
Carbolic acid, C„H60. Cresylic acid, C7H80. Rosolic acid, C20H16O3. Pyridine, C5H5N. Aniline, CeH7N. Picoline, CSH7N. Lutidine, C,H9N. Collidine, C8H„N. Leucoline, C9H7N.
Ammonium snlphocyanate, NH4.NCS. „ cyanide, NH4.NC. „ chloride, NH4.C1.
IV. Coke and Ash in Retort. The proportions in which coal yields these products may be indicated by the case of a cannel giving off 11,000 feet per ton of gas of a density of 0-600. From 100 parts of such a coal there would be yielded—
Gas 22-25
Tar 8-50
Ammonia water 9-50
Coke 59-76
The proportions,however, and even the nature of these pro-ducts of distillation are greatly modified by the temperature at which the distillation is effected, a low red heat yield-ing a small proportion of non-condensible gas but a large amount of heavy hydrocarbon oils, whence the distillation of shales and coal in the paraffin manufacture is conducted at a low red heat. By excessive heat, on the other hand, the compounds evolved become simpler in their chemical constitution, carbon is deposited, pure hydrogen is given off, and the gain in amount of gas produced is more than counterbalanced by its poverty in illuminating properties.

Of the gases and vapours which pass out of the retorts in a highly heated condition, some portion, consisting of tarry matter and ammoniacal liquor, precipitates almost immedi-ately by simple cooling, and other injurious constituents must be removed by a system of purification to which the gaseous products are submitted. What thereafter passes on as ordinary gas for consumption still contains some percentage of incombustible matters—aqueous vapour, oxygen, nitrogen, and carbonic acid. The combustible portion also is separable into two classes, viz., non-luminous supporters of combustion, and the luminiferous constitu-ents,—the former embracing hydrogen, marsh gas (light carburetted hydrogen), and carbonic oxide, while the latter includes the hydrocarbon gases acetylene, ethylene (olefiant gas or heavy carburetted hydrogen), propylene, butylene, and vapours of the benzol and naphthalin series.

Formerly it was the habit to regard the proportion of heavy carburetted hydrogen (ethylene and its homologues) as the measure of the illuminating power of a gas. It has, however, been pointed out by Berthelot that the proportion of such compounds in some gas of good luminous qualities is exceedingly small; and in particular he cites the case of Paris gas, which, according to his analysis, contains only a mere trace of acetylene, ethylene, and other hydrocarbons, with 3 to 3-5 per cent, of benzol vapours. Subsequent ex-periments of Dittmar have proved that a mixture of pure ethylene and hydrogen burnt in the proportion of 3 volumes of hydrogen to 1 of ethylene yields little more light than ordinary marsh gas, while benzol vapour to the extent of only 3 per cent, in hydrogen, gives a brilliantly luminous flame. Frankland and Thorne have more recently deter-mined the illuminating power of a cubic foot of benzol vapour burnt for 1 hour in various combinations, with the following results :—
With hydrogen it gave the light of 69-71 candles.
,, carbonic oxide ,, 73-38 ,,
,, marsh gas „ 92-45 „
,, ,, (secondseries) ,, 93-94 ,,
Thus it is highly probable that the illuminating value of coal-gas depends much more on the presence of benzol vapour than on the proportion of the heavy gaseous hydro-carbons, and the estimation of benzol in the gas is a point which has hitherto been comparatively neglected. In view of the inference that the presence of benzol vapour is so intimately related to illuminating power, the fact observed by Dittmar that water readily and largely dissolves it out of any gas mixture is of great consequence. When ben-zolated hydrogen containing 6 per cent, of benzol vapour was shaken up with water, the percentage of the vapour was found on analysis to be reduced to less than 2

The average composition of the gas supplied to London is, on the authority of the late Dr Letheby, thus stated :—
46-0 39-5 3-8 7-5 0-6 2-0 0-1 0-5
277 50-0 13'0 6'8 01 2-0 0-0 0-4
Ordinary Gas, 12 Candles.
Hydrogen
Light carburetted hydrogen, Condensible hydrocarbons...
Carbonic oxide
Carbonic acid
Aqueous vapour
Oxygen
Nitrogen
Cannel gas is now, however, supplied only to the Houses of Parliament and to certain of the Government offices.

MANUFACTURE OF COAL-GAS.

The series of operations connected with the preparation and distribution of coal-gas embrace the processes of dis-tillation, condensation, exhaustion, scrubbing or washing, purification, measuring, storing, and distribution by the governor to the mains, whence the consumers' supply is drawn. In connexion with consumption, pressure of the gas, measurement of the amount consumed, and the burners and other arrangements for lighting are the most important considerations.

Site and Arrangement of Works.—The choice of a site for a gas establishment is necessarily conditioned by local circumstances ; but the facts that a considerable area is re-quired, and that, at best, the works do not improve the amenity of any neighbourhood, are important considerations. A central position with respect to the area to be supplied is certainly desirable, but in the circumstances it is seldom to be obtained. Of even greater consequence for a large work is ready access to a railway or other means of trans-port ; and most of the great establishments are now con-nected by sidings with lines of railway, whereby coals, &c, are delivered direct from the waggons to the store or retort-house, and in the same way the coke and residual products are removed. Where the arrangement is practicable, it is also desirable that the works should be erected at the lowest level of the area to be supplied, since coal-gas, being speci-fically lighter than atmospheric air, acquires a certain amount of pressure as it rises in pipes, which pressure facilitates its distribution, and it is much easier to control than to beget pressure. In the planning of works, regard must be given to economy of space and to labour-saving arrangements, so that the cost of manual labour may be minimized, and operations proceed in an orderly, methodical, and easily-controlled manner. The accompanying ground plan of gas-works (fig. 1) has been kindly furnished by Mr James Hislop of Glasgow, a gas engineer of known skill and experience; and while it shows arrangements of the most approved character, it will also enable the reader to recognize the position of the various erections and apparatus as they follow each other, and as they will now be described.

Retorts.—Retorts for destructive distillation of coal are formed of cast iron, clay, brick, or wrought iron. Various shapes have been adopted in the construction of these vessels; nor have their forms been more varied than the modes in which they have been disposed in the furnaces. In many instances they have been constructed of a cylindrical shape varying in length and diameter. Those first employed were of iron, with the axis vertical, but experience soon showed that this position was extremely inconvenient, on account of the difficulty which it occasioned in removing the coke.

The retorts were therefore next placed in a horizontal position, as being not only more favourable to the most economical distribution of the heat, but better adapted to the introduction of the coal and the subsequent removal of the coke. At first the heat was applied directly to the lower part of the retort; but it was soon observed that the high temperature to which it was necessary to expose it, for the perfect decomposition of the coal, proved destructive to the lower side, and rendered it useless long before the upper part had sustained much injury. The next improvement was, accordingly, to interpose an arch of brickwork between it and the furnace, and to compensate for the diminished intensity of the heat by a more equally diffused distribution of it over the surface of the retort. This was effected by causing the flue of the furnace to return towards the mouth of the retort, and again conducting it in an opposite direc-tion, till the heated air finally escaped into the chimney. This arrangement was continued so long as iron retorts were in use, but on the general adoption of clay retorts the furnaces were constructed to allow the fire to play freely around them.

The cylindrical form of retort a (fig. 2) was long in favour
on account of its great durability, but it is not so well fitted
for rap id de-
composition
of the coal
as the ellip-
tical b, or ^S-2-
the flat-bottomed or D-shaped retorts d, which are now principally in use. Eetorts are also made of a rectangular section with the corners rounded and the roof arched. Elliptical retorts are varied into what are called ear-shaped or kidney-shaped c, and it is not unusual to set retorts of different forms in the same bench, for the convenience of filling up the haunches of the arch which encloses them.
The length of single retorts varies from 6 to 9 feet, but they are now in some cases made 19J feet in length and 12| inches in internal diameter, these being charged from both ends.

Every retort is furnished with a separate mouthpiece, usually of cast iron, with a socket 6 (fig. 3) for receiving the stand-pipe or ascen-sion-pipe, and there is a movable lid attached to the mouth, together with an ear-box cast on each side of the retort for re-ceiving the ears which support the lid. Fig. 3 shows a form of mouth-piece attached to the retort a, and also the method of screwing the lid to the mouthpiece. That part of the lid which comes in contact with the edge of the mouthpiece has applied to it a lute of lime mortar and fire clay, and when the lid is screwed up, a portion of this lute oozes out round the edges and forms a gas-tight joint.

Except for small works, where the manufacture is inter-mittent, and where, consequently, the retort heat has to be got up frequently, iron retorts are now little used. Clay retorts, which at present are in most general use, wear out quickly; they very frequently crack so seriously on the first application of heat that they must be removed from the bench before being used at all, and in scarcely any case are they in action perfectly free from cracks. Numerous attempts have been made to introduce retorts built of brick ; but the difficulty of making and keeping the joints air-tight has proved a serious obstacle to their use. In the carbonize 500 tons cannel coal, or 2000 tons per oven oi four, without any repairs whatever. Decayed bricks may be removed from these retorts and new ones inserted, and when thoroughly repaired they are again equal to new. Thus the durability of each retort is so great that they are calculated to cost about |th of a penny per 1000 cubic feet of gas generated, as against Id. in the case of moulded retorts, and 7d. with iron retorts, for the same production of gas. In the Hislop retort the arched bricks are made plain, without groove or rebate joints—being thus stronger, more readily put together, and also cheaper. Carbon does not collect so rapidly on brick retorts as on those of clay, the bricks being harder pressed and better burned. On first lighting brick retorts, a charge of coke, breeze, and tar mixed makes them perfectly gas-tight.

Retort Setting.—A furnace or bed of retorts is composed of a group or setting, heated by a separate fire. The furnace is lined with the most refractory fire-bricks, and while the whole brickwork is made of such strength and solidity as ensures the safety of the retorts, the internal construction is so planned that the heat has the utmost possible amount of direct play on the retorts. The number of retorts to one furnace varies from 1 to 15, from 4 to 7 being the number most commonly adopted; and these are variously arranged to bring them all as close to the furnace heat as practicable. In some retort-houses the furnaces are built in two stages or stories, from the upper of which the retorts are charged and drawn, while at the lower level the glowing coke is removed and quenched. The whole range of furnaces constitutes the retort bench, having a common flue which leads to the chimney shaft by which the products of combustion are carried away. The gas-coal for charging the retorts is broken into fragments about 1 K> in weight or thereby. Figs. 5 (elevation) and 6 (section) illustrate the brick retort made of Glenboig Star fire-clay, according to the plan of Mr James Hislop, it is claimed that the difficulty is surmounted, and that both the retort and its setting present great advantage and economy. These brick retorts (fig. 4) are Q-shaped, 9 feet long and with diameters of 22 and 13| inches, set four in an oven to one unarched furnace, as in fig. 7. Each retort will, it is affirmed.

retort setting and arrangement of furnace and flues adopted by Mr Hislop for his brick retorts, in which, by the use of centre blocks, as seen in the open front illustration (fig. 7), the necessity for internal arching is avoided.

Retort furnaces are commonly fired or heated with a portion of the coke which forms one of the bye-products of the gas manufacture ; but in works where shale and rich cannel coals are distilled, common coal must be used in the furnaces. At the Ivry Gas Works of the Compagnie

Fio. 6.—Section of Retort Bed on line A A of fig. 5.
Parisienne d'Éclairage et de Chauffage par le Gaz, the retorts are heated by gas on a method modified from the Siemens regenerative gas furnace. Sectional illustra-

FIG. 7.—Retort Setting in Hislop's Furnace.
tions of a retort setting on this plan, and a description of the various arrangements connected with the regenerators and the controlling of the air and gas currents, will be found in the article FURNACE, vol. ix. pp. 846, 847.

Ordinarily the work of charging and drawing the retorts is accomplished by manual labour, by means simply of shovels for charging, and long iron rakes for drawing the spent charge. In the larger works it is usual to charge the retorts with a scoop semi-cylindrical in form, made a little shorter than the retort, and of such a diameter that it can with ease be pushed in and overturned within the retort. The scoop deposits the coal neatly over the sole of the retort, and of course the lid is much more quickly replaced than can be done with shovel charging. Numerous attempts have been made to introduce purely mechanical means of feeding retorts, hitherto with indifferent success,— such devices as a travelling endless sole and a rotating sole having been tried without good effect. A charging machine and a drawing machine, worked by hydraulic power, have been introduced by Mr Foulis, the engineer of the Glasgow Corporation Gas Works, but after prolonged trial both in Glasgow and in Manchester, these have not yet proved satis-factory in action. In West's patent the charging is effected by the introduction of a small waggon within the retort, which distributes the charge evenly and uniformly, Neither has it, however, met general acceptance.

The retorts are kept at a bright red heat, and for coal with a high percentage of volatile matter a higher tempera-ture is requisite than is needed for coal less rich in gas. As the retorts in one setting are necessarily subject to some-what different amounts of heat, the charges in those nearest the furnace fire, and consequently most highly heated, must be drawn more frequently than the others, as otherwise the quality of the gas would be deteriorated, and a large pro-portion of sulphur compounds would be given off from the overburnt coke.

In drawing a charge the lid is first slightly opened and the escaping gas lighted, to prevent an explosion or " rap " that would otherwise ensue. The gas is prevented from escaping outward by the ascension pipe dipping into the hydraulic main as afterwards explained; but in some cases special valves are fitted on the ascension pipe to prevent a back rushing of the gas. A carbonaceous deposit forms on the sides of the retorts, which requires to be periodically removed by " scurfing " with chisels, or burning it off with free admission of air or steam.

The Hydraulic Main.—From the retorts the gas, after its production, ascends by means of pipes called ascension-pipes B (figs. 5 and 6) into what is termed the condens-ing or hydraulic main HH, which is a large pipe or long reservoir placed in a horizontal position, and supported by columns in front of the brick-work which contains the re-torts A. This part of a gas apparatus is intended to serve a twofold purpose:—first, to condense the tar and some ammoniacal liquor, and secondly, to allow each of the re-torts to be charged singly without permitting the gas pro-duced from the others, at the time that operation is going on, to make its escape. To accomplish these objects one end of the hydraulic main is closed by a flange ; and the other, where it is connected with the pipes for conducting the gas towards the tar vessel and purifying apparatus, has, crossing it in the inside, a partition occupying the lower half of the area of the section, by which the condensing vessel is always kept half full of liquid matter. The stand-pipes are con-nected by a flange with a dip-pipe C, arising from the upper side of the coudensing main HH, and as the lower end of it dips about 2 inches below the level of the liquid matter, it is evident that no gas can return and escape when the mouthpiece on the retort is removed, until it has forced the liquid matter over the bend, a result which is easily pre-vented by making it of a suitable length. The tar which is deposited in the hydraulic main overflows at the partition, and is carried by a pipe to the tar well.





Condensation.—The gas as it passes on from the hydraulic main is still of a temperature from 130° to 140° Fahr., and consequently carries with it heavy hydrocarbons, which, as its temperature falls, would be deposited. It is therefore a first consideration in ordinary working to have these condensable vapours at once separated, and the object of the condenser is to cool the gas down to a temperature nearly that of the surrounding atmosphere. The first con-trivances employed for the purpose of condensation were all constructed on the supposition that the object would be best attained by causing the gas to travel through a great extent of pipes surrounded by cold water, and winding through it like the worm of a still, or ascending upwards and downwards in a circuitous manner. An improvement on this form of condenser, and one now in general use, is represented in fig. 8. It consists of a series of upright

Fig. 8.
pipes connected in pairs at the top by semicircular pipes «, e, and terminating at the bottom in a trough X Y con-taining water, and divided by means of partitions in such a way that, as the gas enters the trough from one pipe, it passes up the next pipe and down into the next partition, and so on to the end of the condenser. The cooling power of this air condenser, as it is called, is sometimes assisted by allowing cold water to trickle over the outer surface of the pipes. Annulur tubes for condensing are also used, in which the gas is exposed to a much greater cooling surface, and in some large works the condensers are cooled by a current of water. In passing through the pipes the gas is considerably reduced in temperature, and the tar and am-moniacal liquor condense, the tar subsiding to the bottom of the troughs, and theammoniacal liquorfloating on thesurface. In course of time the water in the trough is entirely displaced by these two gaseous products, and as they accumulate they pass off into the tar-tank, from which either liquor can be removed by means of a pump adapted to the purpose. The New York Gas Lighting Company employ a multitubular condenser, consisting of two sets of eight boxes, each con-taining 100 tubes 3 inches diameter by 15 feet long. Through each set of tubes, up one and down another, the gas travels, cooled by an external stream of water, while it traverses the 240 feet of piping in the condenser.

The practice of condensation and separation of tarry matter by rapid cooling is condemned by Mr Bowditch and many eminent authorities, on the ground that thereby a proportion of light hydrocarbons are thrown down with the heavier deposit, which on another method of treatment would form part of the permanent gas and materially enrich its quality. A system of treating gas has accordingly been introduced by Messrs Aitken & Young, in which the gas, kept at a high temperature, is carried from the retorts into an apparatus termed an analyser, which consists of an enclosed series of trays and chambers arranged in vertical series, in principle like a Coffey still, the lower portion of which is artificially heated. In action the analyser separates the heavier carbonaceous part of the tarry matter in the lower part or chambers, and as the gas gradually ascends from one tray or tier to another, it is at once cool-ing and depositing increasingly lighter fluids, while it is meeting and being subjected to the purifying action of the light hydrocarbons already deposited. Thus on entering the analyser it meets, at a high temperature, heavy tar deposits, and it passes out of the apparatus cooled dowD to nearly atmospheric temperature after being in contact with the lightest fluid hydrocarbons.

Exhaustion.—To the subsequent progress of the gas con-siderable obstructions are interposed in connexion with its further purification and storing in the gas-holders, and the result of which would be that, were it not artificially pro-pelled, there would be a pressure in the retort equal to the amount of the resistance the gas meets with in its onward pro-gress. The relief of this back pressure not only improves the quality of the gas, but also increases its amount by about 10 per cent. Among the numerous methods of exhaustion which have been proposed since the operation was first introduced in 1839, there are several rotary exhausters, hav-ing more or less of a fan action, and recently an apparatus on the principle of a Giffard's injector has been intro-duced, chiefly in Continental works. A most efficient form is found in the piston exhauster, a kind of pumping engine with slide valves, which exhausts the gas in both the upward and the down ward strokes of its piston. The action of the exhauster is controlled by a governor, which passes back a proportion of the gas when the apparatus is working too fast for the rate of production in the retorts; and " pass by" valves are arranged to carry the gas onward without passing through the exhauster should it cease to work from accident or any other cause.

Purification.—The operations embraced under this head have for their object the removal from the gas of am-monia, sulphuretted hydrogen, and carbonic acid as the main impurities, with smaller proportions of other sulphuric and of cyanogen compounds.
The agencies adopted are partly mechanical and partly chemical, the separation of the ammonia being first effected in the " scrubber," from which the gas passes on to complete its purification in the " purifiers." In early times the purify-ing was performed in a single operation by the use of milk of lime in the wet purifier, a form of apparatus still in use where wet purifying is permissible.

The Wet Purifier.—This apparatus was supplied with a cream of lime and water, but, although it was a most efficient purifying agent, the ammonia now of so much value was lost by its use, and the " blue billy," as the saturated liquid holding the impurities was termed, created an intol-erable nuisance, and could be in no harmless way got rid of. Except in small works, wet purifying is not now practised.

The Scrubber.—The object sought in an ordinary scrubber is to cause a large amount of gas to come in con-tact with the smallest possible quantity of water, so as at once to dissolve out ammoniacal gases, which are exceedingly and down the other, and from the top a constant small stream of weak ammoniacal liquor trickles down. Such a scrubber, it is stated, is subject to clogging by deposits of tar, and equally efficient work is done without that draw-back by an apparatus in which perforated iron plates occupy the place of the coke, and in the Livesey scrubber layers of thin deal boards are employed, These boards are set in tiers perpendicularly, slightly crossing each other, with about J of an inch between each tier. Anderson's washer is a form of scrubber recently introduced, in which the interior is occupied with a series of rotating whalebone brushes, which dip into troughs of ammoniacal liquor, and in their revolution meet and agitate the gas in its passage upwards through the tower or column. The scrubber shown in section and plan in figs. 9 and 10 is a form introduced by Mr James Hislop. It contains 10 tiers of trays of cast iron, perforated with |-inch holes at a distance of 2 inches from centre to centre. The gas passes upwards through these, meeting in its course a shower of ammoniacal liquor pumped up and distributed by the rose arrange-ment shown in fig. 9. The bottom part of the scrubber, to the height of the first course of plates, is filled with liquor, which is repumped till it reaches the strength de-sired for the manufacturer of ammonia sulphate.

The Purifiers.-—The ordinary lime purifier, by which sulphuretted hydrogen and carbonic acid are abstracted from the gas, consists of a large rectangular vessel seen in section in fig. 11. Internally it is occupied with ranges of wooden trays or sieves A, made in the form of grids of |~inch wood, with about half an inch between the bars. These are covered with slightly moistened slaked lime B to the deoth of about 6 inches, and from three to six tiers of such sieves are ranged in each purifier. The gas enters at the bottom by a tube C, the mouth or inlet being-protected from lime falling into it by a cover D, and it forces its way upward through all the trays till, reaching the lid or cover E, it descends by an internal pocket F to the exit tube G, which leads to the next purifier. The edges of the lid dip into an external water seal or lute H whereby the gas is prevented from escaping. The purifiers are generally arranged in sets of four, three being in use, through which the gas passes in succession while the fourth is being renewed; and to control the course of the gas current among the purifiers, the following ingenious arrange-ment of centre valves and pipes was devised by Mr Malam (fig. 12).

It lias a cover fitting within it in such a way as to communicate with the pipe a and either of the four inlet pipes, and also to com-municate between one of the outlet pipes and the pipe h, which carries off the purified gas. The inlet pipes, b, d, f, admit the gas from the central case to the bottom of the purifiers; and the outlet pipes, c, e, g, return the gas from the purifiers back to the case, after it has passed up through the layers of lime, and descended at the back of a partition plate in each purifier to the outlet pipes at the bottom, a is the main inlet pipe for conveying the gas from the scrubber or the condenser, and 7t is the main outlet pipe for conveying the gas to the gasholder. The central cylinder contains water to the depth of 10 inches, and the ten pipes rise up through the bottom to the height of 12 inches, so that the mouth of each is 2 inches above the surface of the water. The cover which fits into the cylinder is 4 feet 3 inches in diameter, and is divided into five parts, the first of which, 1, fits over the inlet pipe a, and over either of the inlet pipes leading to the purifiers. The partitions 2, 3, and 5 fit each over an inlet and an outlet pipe, while one parti-tion, 4, fits over one outlet pipe from one purifier, and over the pipe h, which leads to the gas-holder. In fig. 12 the arrangement is such as to open a communication between the inlet pipe a and the purifier A. Now supposing the gas to have passed from the scrubber into the centre of the cylinder, its only means of escape is to pass down the pipe b into the purifier A, where it ascends through the layers of lime, and passing over the top of a dividing plate, descends and escapes from the bottom of the purifier by the pipe c back to the cylinder. Here its only means of escape is by the pipe d, which conducts it to the purifier B, in which it ascends and descends as before, returning by the pipe e to the cylinder, whence it proceeds by the pipe / into the purifier C, then along the pipe ¡7, which is shut off from communication with any pipe except h, by which it is conveyed away to the gas-holder. By this arrangement the three purifiers ABO are being worked, while a fourth purifier D is being emptied and recharged with lime. When it is found, on testing the gas, that the lime is unfit for its office, the purifier A is thrown out of work, and D is brought in. The frame is then shifted so as to bring the triangular division 1 over d, by which means BCD will be the working purifiers, and A will be thrown out of use. In this way, by shifting the frame round its centre over each of the four outlet pipes, any three of the purifiers can be brought into action.

The "oxide" method of purifying the gas, originally introduced by M. Laming, and shortly afterwards patented by Mr Hills, is now largely used in ordinary gas-works. It is based upon the property of the hydrated oxide of iron to decompose sulphuretted hydrogen, a portion of the sulphur forming a sulphide with the iron. Quicklime is also used to separate carbonic acid, and the oxide of iron is mixed with sawdust or cinders (breeze) for the purpose of increas-ing the surfaces of contact, and this mixture is placed in the purifiers. When a sufficient quantity of gas has passed through it, the purifiers are opened, and the mixture is exposed to the air, under which new condition it combines with oxygen, and again becomes fitted for use in the puri-fiers. The chemical changes which occur in these opera-tions are thus stated. The mixture of hydrated oxide of iron, &c, absorbs sulphuretted hydrogen, forming ferrous sulphide and water, and liberating sulphur, thus:— Fe203-l-3H2S = 2FeS-rS + 3H20. The ferrous sulphide, by exposure to the air, absorbs oxygen, and its sulphur is

The Gas-holder.—This, which is frequently designated the gasometer, though incorrectly, since it does not in any way measure gas, but simply stores it for consumption, con-sists of two portions—the "tank" T (fig. 13) and the " holder" G. The tank is a cylindrical pit, surrounding a central core, which is usually covered with concrete c at top, and has its sides built of masonry or brick-work, p, b. The tank is water-tight, and is filled to a high level with water, above which project two tubes m m, one being the inlet and the other the supply pipe which leads to the main governor.

Formerly gas-holders were made of heavy plate iron, strengthened by angle-iron and stays, and of so great a

FIG. 13.—Section of Gas-holder.
weight as to require a complex system of equilibrium chains and counterbalancing weights to relieve the gas from the great pressure to which it would otherwise be subjected. They are now made so light that they require to be loaded in order to supply the required pressure, and their rise and fall are regulated by means of guide-rods i i round the tank. For economy of space holders in which different segments " telescope" over each other are now much employed. This form of holder consists of two or even three separate parts,— the upper having the form of the common gas-holder, and the other being open at the top as well as the bottom. They are connected by the recurved upper edge of the lower fitting into a channel which runs round the bottom of the upper, whereby the entire structure is rendered air-tight at the line of junction. Holders of great capacity are now erected in connexion with large works. The Imperial Company in London possesses two, at Bromley and Hackney, telescopic in form,—the outer segment measuring 200 feet in diameter by 35 feet deep, and the inner 197 feet by 35. These holders are each capable of storing 2 million cubic feet of gas, which at sp. gr. 480 would weigh 73 tons. A still larger holder is at the Fulham station of the Gas Light Company, it being 223 feet in diameter and rising 66 feet, with a capacity equal to 3 million cubic feet.

The Governor.—An efficient control of the pressure of the gas, along its whole course from the gas-holder to the point of consumption, is an object of great importance for the avoiding of leakage, for equal distribution, and for supply-ing the burners at that pressure which yields the largest illuminating effect. Uncontrolled pressure may supply certain levels in a proper manner, but will leave low-lying districts insufficiently supplied, while the pressure in high districts will be excessive. The variations from simple difference of level may be very great. Thus, with a pressure of T7 inch at the Leith works, the gas would be delivered in some parts of Edinburgh at a pressure of 4-5 inches. The varying consumption from dusk onwards also greatly affects unregulated pressure. To control and correct these and other irregularities and disturbances governors are now used,—at the works or station for delivering the gas to the mains, in districts to correct variations owing to level, and beyond the consumers' meters for controlling house supply; while in certain forms of burners a regulat-ing apparatus is also inserted. The principle on which all governors are based consists in causing the gas by its own pressure to act on some form of sensitive surface which opens or closes a valve or aperture in proportion to the variations of pressure exerted on it. Fig. 14 is a dia-grammatic section of the common form of station governor.

The course of the gas is indicated by arrows, d being the inlet and e the outlet pipe; c is a valve of conical form fitted to the seat i and raised or depressed by the weight / working by a cord over a pulley; bb is the bell or holder,—a cylindrical vessel of sheet iron which rises and falls in the exterior vessel aa, in which water is contained to the level represented. The gas, entering at d, passes through the valve, fills the upper part of the inverted vessel bb, which it thus partially raises, and escapes by e. If the pressure from the holder be unduly increased or di-minished, the buoyancy of bb will be in-creased or diminished in like proportion, and the valve being by this means more or less closed, the quantity of gas escap-ing at e will be unaltered. And not only will the governor accommodate itself to the varying pressure of the holder, but also to the varying quantities of gas required to escape at c for the supply of the burners. Thus, if it were necessary that less gas should pass through c, in consequence of the extinction of a portion of the lights, the increased pressure thus produced at the holder would raise the governor, and partially shut the valve, leaving just sufficient aperture for the requisite supply of gas.

Numerous improvements have been made on the ordinary station governor. In the form invented and manufactured by D. Bruce Peebles, the bell or holder is enclosed in a gas-tight case or chamber, and a small portion of the inlet gas flows in and out of this chamber above the holder. The pressure of this small quantity of gas is regulated by passing it through a small separate governor; and, acting on the outer surface of the holder, this, in a very delicate and sensitive manner, performs the duty of weights in the older forms of governor. An arrangement similar in principle is applied to the district governor by Bruce Peebles, the minimum day pressure being secured by means of a stopcock or screw-valve on the apparatus, and the maximum night pressure is controlled by a small subsidiary governor. The principle of the small governor, which thus plays an im-portant part in regulating large flows of gas, will be ex-plained under consumers' governors, the apparatus being shown in section in fig. 18 below.

Supply Pipes.—The street main and service pipes are tubes of malleable or of cast iron, the gauge of which must be arranged according to the quantity of gas to be supplied, the length it has to travel, and the pressure under which it is carried forward. Practical gas-engineers possess elaborated tables of data for the regulation of the size of their various supply pipes. Notwithstanding the utmost care and accuracy in the laying and fitting of street mains, leakage at joints is a constant source of annoyance. Under the most favourable conditions there is a discrepancy of from 7 to 8 per cent, between the gas made and the amount accounted for by consumption, and the greater part of that loss is due to leakage in street pipes. To convey the gas from the main pipes and distribute it in houses, pipes of lead or of block tin are generally used.

Consumers' Meters.—Of these there are two forms in actual use, the " wet" and the " dry." The former, the invention of Mr Clegg, is represented in the two sections (figs. 15 and 16), where cc represents the outside case, having the form

Fig. 15. Fig. 16.
of a flat cyclinder; a is the inlet tube and b the outlet pipe ; g, g are two pivots, and h a toothed wheel fixed upon the pivots and connected with a train of wheel-work to register its revolutions. The pivots are fixed to and support a cylindrical drum-shaped vessel ddd, having openings e, e, e, e, internal partitions ef, ef, ef, ef, and a centre piece ffff. The machine is filled with water, which is poured in at A up to the level of i; and, on gas being admitted under a small pressure at a, it enters into the upper part of the centre piece, and forces its way through such of the openings / as are from time to time above the surface of the water. By its action upon the partition which curves over the opening
a, a rotatory motion is communicated to the cylinder,—the gas from the opposite chamber being at the same time expelled by one of the openings e, and afterwards escaping at
b, as already mentioned. Wet meters work easily, and, when well set and properly supplied with water, measure the gas with much accuracy. But excess or deficiency of water impairs their measuring power, which may also be affected by the meter being lifted off the level. The freezing of the water also frequently occasions trouble, and the action of the water on the gas passing through it by dissolving out part of the valuable illuminating hydrocar-bons on the one hand, and diffusing watery vapour through it on the other, doubly affects its illuminating power.

The dry meter is free from the defects just mentioned, but does not pass the gas with such steadiness as the wet meter. The ordinary dry meter consists of an oblong box enclos-ing two measuring cylinders, with leather sides which con-tract and expand as they are being emptied and filled, on the principle of ordinary bellows. The pressure of the gas entering this meter is sufficient to keep it in operation, and by a system of valves the one cylinder is in process of filling as the other is being emptied through the service pipe. The chambers communicate by means of lever arms with a crank which turns a train of wheels in connexion with the indicator dials on the face of the machine.

Consumers' Governor.—In order to consume gas in a perfectly uniform and economical manner, it is essential that the pressure at the burners should be always in-variably the same. That pressure is liable, however, to variation from a number of causes, such as fluctuation in the number of lights in use, either in the house or in the neighbourhood, or the application or withdrawal of pressure at the works' governor. And as all good burners are fitted with regard to a fixed standard quality and pressure of gas to be consumed, if this is not maintained the conditions of maximum illuminating power are lost. A consumers' governor seeures uniformity of pressure at all the burners supplied by the pipe on which it is placed. The prin-ciple of the governor is identical with that of the station governor already described, increased pressure in both cases causing the orifice through which the gas escapes to be contracted. The mechanical arrangements by which this contraction of orifice is effected are various. In some instances they are in direct contact with the separate burners, while other governors are applied to the supply pipes of a whole establishment. They are separable into pressure governors, which, like the station governors, give a constant or uniform pressure under all variations of con-sumption, and volumetric governors which pass a constant volume or amount of gas under all variations of pressure.

Of pressure governors the forms devised by Sugg and Bruce Peebles are in extensive use, the latter especially being much applied to street lamps. In Sugg's consumers' governor (fig. 17)

\
OUTLET

—. 1
the gas enters at the inlet, and, following the course indicated by the arrows, passes through the regulating plate of the governor into the gas-holder, and thence, by the opening provided for it, it reaches the outlet. The gas-holder has suspended from a disc in the crown a half-ball valve, which closes or opens the opening in the regulating plate as the gas-holder rises or falls. A weight placed on the top of the holder fixes the pressure required to raise it. As a consequence, if the pressure of the gas on the inlet is greater than that required to lift the holder, then the latter rises, carrying the half-ball valve with it, till such time as the opening left between the sides of the valve of the regulating plate is sufficient to allow the passage of the necessary quantity of gas to balance the holder. On the other hand, if the pressure at the inlet falls below that required to lift the holder, the full opening of the regulating plate allows all the gas there is to pass through the governor to the burners. "Where a very perfect control is desirable, the parts of the governor are made in duplicate, and a double control is thus established. With certain structural differences the action of the Bruce Peebles governor (fig. 18) is the same. The gas enters at 1,

Fig. 18.—Consumers' Governor (Peebles).

and passes out at 2 into the pipe leading to the burners. To adjust the governor the brass cap 3 is unscrewed, and the weights 4 taken off or put on until the desired pressure, of say 5-tenths, at the burners is obtained, when the brass cap is again screwed to its place. The weights now keep the valve 6 open so long as 5-tenths pressure is not ex-ceeded in the main; but any variations in the main above that pressure act at once on the diaphragm 5, and partly close or open the valve, thus maintaining under all cir-cumstances a steady outlet pressure.

Of volumetric governors the best known is Giroud's glycerin rheometer, which consists of a closed cylindrical casing containing a very light metal dome or ball dipping into a circular channel filled with glycerin. In the upper part of the dome is a small orifice through which the gas passes, and on its top is fixed a conical valve which works in a seat at the top of the casing. As the pressure from the supply side rises or falls, the bell responsively moves up or down, opening or closing by the conical valve the orifice by which the gas passes outward; and so deli-cately is this compensation adjusted that the gas passed is the same in amount however different the pressure. Bruce Peebles has invented a simple and inexpensive form of volumetric governor (fig. 19), in which the use of glycerin is dispensed with. It consists of a conical dome resting on a needle-pointed stud, the cone having an orifice at C, and there is besides a variable consumption channel at the side ABA, which can be 19.—Volumetric controlled by the external screw. As soon Governor (Peebles), as the stopcock is opened the gas fills the interior of the cone, and momentarily closes the valve; hut, finding its way by the vertical passage, or through the hole C, in the cone, it reaches the chamber above the cone. The cone is therefore now surrounded by gas at the same pressure, and, having nothing to support it, falls, and lets gas pass to the burner. But this only takes place to an extent that allows a differential pressure to be established suffi-cient to support the cone, which is then equilibriated between two pressures ; and the difference between these two pressures remains constant, however much the initial pressure of the gas may vary, unless, of course, it gets so low as not to be able to raise the cone.

Burners.—The question of the arrangements by which the maximum illuminating power may be developed in the consumption of gas, being one which principally affects individual consumers, has not received the attention which their importance merits. As a rule, gas-fitters are ignorant of the principles involved in the economical use of gas, and are often prejudiced by the assertions of certain inventors; and thus it happens that, owing to defective fittings, unregulated pressure, and imperfect burners, an enormous loss of illuminating power is suffered. In their report to the Board of Trade in 1869, the referees under the City of London Gas Act state, of a large number of burners examined by them, that

" The diversity as to illuminating power was surprisingly great, and such as will appear incredible to any one who has not ascer-tained the facts by careful experiment. They also found the kinds of burners in common use are extremely defective, thereby entailing upon the public a heavy pecuniary loss, as well as other dis-advantages. In order to examine this important matter more fully, the referees, with the ready permission of the proprietors, inspected several large establishments in the city, where, owing to the preval-ence of night work, an unusually large amount of gas was consumed. The inspection in every case confirmed the apprehensions which the referees had formed from their examination of the burners which they had procured from the leading gas-fitting establishments. In the offices of two of the leading daily newspapers (establishments which consume more gas than any other), they found that the burners principally in use gave only 55 per cent, of light compared with the Sugg-Letheby burner, or with Leoni's Albert Crutch burner, and yet the price of the last-named burner is almost identical with that of the very bad burners employed in these offices. Tested by the Bengel burner, or by Sugg's new burner, the amount of light given by these imperfect burners is only between 47 and 49 per cent, of what is obtainable from the gas."

In a communication to the Philosophical Society of Glasgow in 1874 Dr Wallace, the official gas examiner of that city, dealing with the rich cannel gas of a minimum illuminating power of 25 candles there supplied, estimated that there is in ordinary consumption a loss of 40 per cent, of illuminating power which, under favourable circumstances, might be obtained, and that in practice, while not more than 16-candle power is procured, from 20 to 23-candle illumination ought to be readily obtainable.





This universal wasteful misuse of gas is not merely a question of economy, although the aggregate pecuniary loss must be very great. It affects in no small degree the health and comfort of the consumers of gas; the products of combustion of the purest gas vitiate the atmosphere, and overheat the apartments in which it is burned. Moreover, the light from gas properly burned is much steadier and purer, and less trying to the eyesight, than that wastefully consumed.

The principal circumstances which demand attention in the fitting of burners are the average pressure and illumin-ating power of the gas to be consumed. How pressure may be controlled has already been shown in connexion with governors. The quality or illuminating power of gas has a most important bearing on the nature of burners proper for use, so that a clear distinction must be drawn between common coal-gas and cannel-gas, the burners for the one kind being quite unsuited for the other variety. The maximum amount of light is obtained from any gas just at that point where the flame is on the verge of smoking, and the conditions under which 14-candle gas would be per-fectly consumed would, with 26 or 30-candle gas, produce a large amount of smoke. Indeed, the richer gas is, the greater is the difficulty in developing its full illuminating power, and at all times it must be burned in a much thinner sheet or stream than is proper in the case of poor gas, which requires less access of air for its complete lumini-ferous combustion. The opening or slit in burners used for common gas is therefore much larger than in those devoted to the consumption of cannel-gas.

There are two principal kinds of burners in use—Argand and flat-flame burners. The Argand burner in its usual form is useful only for common or low illuminating power gas, and it has, in the hands of various inventors, especially by Mr William Sugg of London, been so improved that for amount and steadiness of light it leaves little further improvement to be hoped for. The common Argand consists of an annular tube with a circle of small holes pierced in the end of the ring. It thus produces a circular or tubular flame, which requires to be protected with a glass chimney, by which the admission of air is regulated. The burner made by Sugg in 1869, known as the Sugg-Letheby, or Sugg's No. 1, is the standard burner adopted for the United Kingdom in Acts of Parliament, and the same standard has been adopted in the United States, in Canada, and in various European states. At the time it was made, the Sugg No. 1 was esteemed the best known burner, but since that time Mr Sugg has perfected his London Argand, whereby with London gas results equal to about 2 candles better than the standard are obtained. Fig. 20 is a sectional view of Sugg's London Argand with the latest improvements.

At the point at which the gas enters is a brass nose-piece A, screwed to fit the usual three-eighth thread, intended by the manu-facturers of all kinds of gas fittings to receive the burner. This is drilled through its length, and slightly trumpeted at the top so as to fit the cone-shaped piece of metal projecting from the roof of the inlet cham-ber B. The outside of the upper portion of the nose-piece A is screwed to fit the inside of the inlet chamber B, and thus, by an adjustment of this screw by means of paper washers put on the shoulder at AB, it is possible to enlarge or decrease the area of the passage through which the gas has to pass in order to supply three tubes (two of which, C and D, only are shown in the draw-ing), by which it is further con-ducted to the combustion cham-ber E. This chamber is made of steatite, a material which is capable of resisting the corroding action of heat or damp, and is a good non-conductor of heat. It is pierced with a number of holes, so arranged as regards size and number that the quantity of gas the burner is required to consume shall pass out at an inappreciable or the least possible pressure. This is in order that the oxygen of the atmosphere, slowly ascending through the centre opening F, the annulus formed by the edge of the air cone G, and the outside of the com-bustion chamber E, shall combine with the burning gas by natural affinity only, leaving the nitrogen to pass freely out at the top of the flame. H is one of the three springs which are intended to keep the chimney glass steady in its place. JJ are two of three stubs or rests for a screen, globe, or moon; and K is a peg to steady the current of air which passes up the centre opening F.

With the view of competing in illuminating power with the electric light, Mr Sugg has recently devised a modified form of Argand burner calculated to yield a large illuminat-ing power by increased but still economical consumption of gas. These burners are made of two or more concentric Argand rings, the outer being of large diameter, and in operation they give out a large solid, white, steady flame. With London gas, a two-ring burner consuming 19 feet per hour yields 80-candle light; 3-ring burners which consume 23 feet give 100 candles; 4-ring burners fed with 45 feet of gas gave an illumination equal to 200 candles.

As regular pressure is essential for the proper use of these burners, a self-acting governor is frequently fitted to them. The pressure at which the best results are obtained with London gas is about -7 inch. In a series of experiments with Argand burners made by Mr John Pattinson of Newcastle-on-Tyne the following results were obtained:—

Burner. Cubic feet per hour. Illuminating power in Candles. Illuminating power per 5 cubic feet per hour.
Sugg-Letheby Standard Sugg's London Argand Sugg's Improved Lon-
Do. do. 5-0 5-0
| 4-5
5'0 5-0 7-0 14-10 15-90
16-08
17-80 11-20 17-80 14-10 15-90
17-86
17-80 11-20 12-70

Flat-flame burners, or burners which spread their flame in a broad thin sheet, are of two principal kinds known respectively as "fishtail" (fig. 21) and " batwing" (fig. 22) burners. The fishtail or union burner has two orifices drilled in its surface, which are inclined to-wards each other at an angle of 90°, so that the issuing cur-rents impinge and spread the flame in a broad sheet. The gas in the batwing issues from a narrow slit cut right across the surface. In the FlQS- 21> 22.—Flat-flame Burners, best forms of all kinds of burners now in use steatite or adamas (pottery) tops are employed. In Sugg's Christiania burner the slit is circular, and the light issues in two thin sheets which coalesce in their upper luminiferous part, pro-ducing a most beneficial result when common gas is con-sumed. The common metal and steatite-tipped burners in use permit the current of gas to strike against their orifices without any control or regulation, but in the numerous .patented forms of both fishtail and batwing jets certain mechanical obstructions, or small governors, are inserted, which break or retard the current. Screws, wire gauze, calico, cotton wool, iron filings, and constriction of the lower part of the burner are all devices in use. Of all these one of the simplest and most effective is the plan on which the Bronner burner is constructed, which is simply to have the opening at the lower part of the burner smaller than the upper orifice. For different qualities and pressures of gas the Bronner burner presents a great variety of combina-tions by having several distinct sizes of lower constriction which can be adjusted to a large number of tip orifices. Thus, with six distinct openings at each end, 36 combina-tions can be made. As Argand burners are not suited for measuring the illuminating power of rich cannel-gas, flat flame-burners have to be employed ; and in the Act of Parliament under which the Glasgow Corporation supplies gas, it is provided that "all the gas supplied by the corpora-tion shall be at least of such quality as to produce from a union jet burner, capable of consuming 5 cubic feet of gas •per hour under a pressure equal to a column of water -5 of an inch in height, a light equal in intensity to the light produced by 25 sperm candles of 6 in the pound, burning 120 grains per hour."

Dr Wallace, in a communication on the " Economical Combustion of Coal-Gas" (Proc. Phil. Soc. Glasgow, vol. ix.), tabulates an •extensive series of experiments made with flat-flame burners of various sizes with about 28-candle gas at different degrees of pressure. The general result of these experiments shows that, to obtain the highest luminiferous effect with burners of small aperture, a low pressure of gas (not more than -g inch) must be maintained, although, as the size of the jet increases within certain limits, the pressure may be increased with favourable results. With 9 sizes of Bray's regulator fishtail (a burner having an obstruction consisting of a double fold of cotton cloth) Dr Wallace obtained the following results, calculated to 5 cubic feet per hour :—
28-16 30-2
2S-08 32-0
17-4 13-3 9-8
L'i! « L'Ili! 175
20-15
26-2
23-7
Burner
At J-inch pressure
At. 1-inch „
At li-inch „
gas blows

The gas used in the i-inch experiments was 27'72-candle standard, for the 1-inch series it was 29-05, and for the H-inch set it was 28 61-candle. With 30 combinations of Bronner burners Dr Wallace obtained from 2S-2-candle gas at 1 inch pressure an average of 257, and at l|-inch 25-8-candle power, most of the combinations giving fairly equal results.
Of all burners the ordinary fishtails, and they are the most fre-quently used, give the most inferior results when used for burning common coal-gas. The results tabulated below are derived from

Gas Testing.—The universally recognized and practised method of valuing gas is by comparing its light with that yielded by a standard light, which can be obtained as nearly as possible of an unvarying intensity. In making such a photometric comparison it is essential that the conditions under which the lights to be compared are burned shall be uniform, and that the materials be consumed at a definite rate. The standard recognized by legislative authority in Great Britain and America is the burning of a sperm candle 6 to the B>. consuming at the rate of 120 grains of sperm per hour, compared with gas burning at the rate of 5 cubic feet per hour. The burner prescribed for common gas is the Sugg-Letheby-Argand, in Acts of Parliament defined as a 15-holed Argand with a 7-inch glass chimney; and for rich cannel-gas a union or fishtail jet passing 5 feet per hour is employed. The apparatus employed for making the com-parison is generally the Bunsen photometer, or some modi-fication of that instrument; and the ratio of comparative illumination is established by the well-known principle that the intensity of light diminishes in inverse proportion to the square of the distance from its source. The Bunsen photo-meter consists of a bar of wood 98 inches long, with a candle holder at one end and at the other the standard gas burner. A balance for weighing the candle as it burns, an indexed meter for the gas, and a clock are also provided, The bar is graduated from the centre to each end, and on it is set a sliding holder into which a screen of prepared paper is placed. The screen is so prepared that a spot or disc is more opaque than the remainder of the paper, so that when light passes through it from one side, that particular spot is seen distinctly darker than the rest. When, how-ever, equal amounts of light fall on it from both sides the spot disappears, and the whole surface presents a uniform appearance. Therefore, with both candle and gas burning under the stipulated conditions in a darkened chamber, by moving the screen on the graduated bar from the one light and towards the other till the dark spot on the paper dis-appears, the comparative illuminating power of the light is ascertained by the position of the screen on the graduated bar, or by a simple arithmetical calculation. Thus, the lights being 100 inches apart, if at the conclusion of the experiment the screen is 20 inches from the candle and 80 from the gas jet, since 802 is 16 times 202, the gas is 16-candle power.

Comparisons of the quality of gas are also made by the jet photometer, an apparatus which depends on the prin-ciple that gas of uniform quality burned at invariable pressure, through a small orifice, yields a flame of uniform height. If the flame is to be maintained at a uniform height the pressure in the pipes must increase as the quality of the gas decreases. The jet photometer forms a ready and convenient means of ascertaining any variations in the quality of gas supply ; but it is not available for purposes of comparison.

Analysis of gas does not yield so satisfactory evidence of its illuminating value as photometric comparisons, but various methods of ascertaining the proportion of lumini-ferous olefines contained in any gas are occasionally prac-tised. The absorption of the heavy hydrocarbons by chlorine or by bromine, and Dr Fyfe's durability test, are of theoretical rather than practical importance.

Residual Products.—Under this term are embraced coke, ammoniacal liquor, and gas-tar, all of which are sources of income in the gas manufacture. Indeed the value of these products has increased so rapidly of late years, and they now form the basis of manufactures of such consequence, that the residual products can scarcely be regarded as of secondary importance, and they will certainly play no small part in determining the future maintenance of gas-lighting in the face of other competing systems. The change in the valuation of ammonia and tar liquors is well illustrated by the circumstance that, during the year 1878, the corpora-tion of Bradford was offered £10,000 per annum for these products, which about eight years previously had been disposed of for a yearly payment of £800.

Coke is a substance which varies much in value, according to local circumstances, and the nature of the coal distilled. When shale is used, there remains in the retorts an ashy residue which is absolutely worthless; and the coke of cannel coal is also comparatively of little value, owing to the amount of ash it yields. Indeed, in Scotch works where ashy cannel alone is distilled, the retorts have to be partly fired with common coal. The coke obtained from the distillation of caking coal, on the other hand, is of high value, and after a supply is set aside for heating the retorts there generally remains from 65 to 85 per cent, of the whole amount to be disposed of by sale.

Ammoniacal liquor is more abundantly produced by the distillation of cannel than by common coal, from 18 to 22 Bo of ammonia, as sulphate, being obtained from each ton of cannel distilled, as against about 16 lb derived from ordinary coal. Gas liquor is now almost the sole source of ammonia, which, among other purposes, is very largely employed as an agricultural fertilizer.

Tar liquor yields by destructive distillation a wide range of products possessing a great and increasing industrial value. The cannel coals, and other varieties rich in volatile matter, are also the kinds which yield the largest propor-tion of tar. In the distillation of coal-tar, after some ammoniacal and watery vapours have been given off, there is distilled over a proportion of highly volatile fluid hydro-carbons which consist principally of benzol; and afterwards a large amount of a light oil, known as coal naphtha (also a mixture of various hydrocarbons), is obtained. At this point the residue in the retort is called artificial asphalt, and as such is a commercial article; but if the heat is forced, and the distillation continued, a large amount of "heavy" or "dead oils" is obtained, and the mass left in the still is " hard pitch." The heavy oils are a mixture of naphthalin, phenol (carbolic acid), cresol (cresylic acid), and anthracene, oic. The benzol obtained in the first stage of the distilla-tion is the basis of aniline and its various dyes ; naphtha is used as a solvent, and for lighting and other purposes; carbolic acid, in addition to its employment as an anti-septic, is the basis of many valuable dyes; anthracene forms the source of the now most important dye, artificial alizarin; and most of the substances have other applica-tions of minor importance.
The relative position and value of the various products of the gas manufacture is exhibited by the following con-densed statement of the position and operations of the various London gas companies during the year 1875 :—
Total capital of the companies £12,516,009
Capital called up 11,005,589
Total gas rental 2,606,818
Cost of coal 1,455,407
Receipts for coke and breeze 492,927
„ for tar 162,151
,, for ammonia 111,951
Gas produced 14,888,133 thousand feet.
Gas sold 13,622,639 „ ,,
Coal carbonized (4 per cent, cannel) 1,505,000 tons.
Coke produced, 34 bushels per ton 1,417,654 chaldrons.
Coke used as fuel in retorts, 31 per cent., 440,685 ,,
Coke sold, 69 per cent 976,969 ,,
Average yield of gas per ton of coal 9,892 cubic feet.

GAS FROM SOURCES OTHER THAN COAL.

Petroleum-Gas.—Petroleum being a substance obtained in great abundance, notably in America, is used, not only directly as an illuminating agent, but also for the production of gas ; and as an enricher of common coal-gas it is applied at several works in New York and Brooklyn. Its prepara-tion is effected by distilling it first at a low temperature into a rich vapour, which, when passed into highly heated retorts, is converted into permanent gas of an illuminating power about five times greater than common gas, and which is, moreover, absolutely free from ammonia, sulphur com-pounds, and carbonic acid. On account of its great rich-ness, petroleum-gas must be consumed in special burners of very fine aperture, at a rate varying from 5 to 2 feet per hour.

Oil-Gas.—In the early stages of gas manufacture many attempts were made to substitute gas distilled from inferior oils for coal-gas. The oil was distilled by allowing it to percolate into highly heated retorts, in which a quantity of coke or a like porous solid was placed, and the distillate was a richly luminiferous gas free from hurtful impurities. Although oil in this form yields a convenient and powerful illuminant, its direct combustion is much more economical; and as all oils and fats are highly valuable for many purposes besides illumination, they cannot compete with gas coal as a source of gas. Nevertheless the New York Gas Light Company manufactured oil-gas exclusively from 1824 till 1828, and sold their product at $10 per 1000 feet. The distillation of suint from wool washing, and of re-covered spent soap, are examples of the application of oleaginous substances for gas-making.

Resin-Gas.—In its treatment and results resin, as a source of gas, is very similar to oil. It yields a pure gas of great illuminating power, and for twenty years (1828-48) it was supplied in New York at $7 per 1000 feet. Previous to the civil war of 1861-65 it was a good deal used on the European continent.

Wood-Gas.—The original experiments of Lebon, it will be remembered, were made with wood-gas, but he failed to obtain from his product an illuminating power that would compare with that of coal-gas. Lebon's failure was in later years shown to arise from distilling at a temperature which gave off chiefly carbonic acid with non-luminous carbonic oxide and light carburetted hydrogen, leaving in the retort a tar which the application of a higher heat would have resolved into highly luminiferous gases and vapours. Pettenkofer, who pointed out the fact, devised a system of wood-gas making in which the products of the low-heat distillation were volatilized by passing through a range of red-hot pipes; but now it is found that ordinary retorts, properly heated and fed with small charges, answer perfectly well for the operation. Wood-gas, owing to its high specific gravity and the proportion of carbonic oxide it contains, must be burned at considerable pressure, in specially con-structed burners with a large orifice. It is largely used in Germany, Switzerland, and Russia, where wood is more easily obtained than coal. It was used at Philadelphia gas-works in 1856, where it was affirmed to be cheaper and of greater luminosity than coal-gas.

Peat-Gas is evolved under circumstances the same as occur in connexion with the wood-gas manufacture, but the amount of moisture contained in peat is a serious obstacle to its successful use in this as in most other directions. Earnest and persistent efforts have been made to use peat as a source of gas, but these have met but little commercial success. To a limited extent it is used in various German factories which happen to be situated in the immediate neighbourhood of extensive peat deposits.

Carburetted Gas.—Under this head may be embraced all the methods for impregnating gaseous bodies with vapours of fluid or solid hydrocarbons. The objects aimed at in the carburetting processes are—(1) to increase the illuminat-ing power of ordinary coal-gas; (2) to render non-luminous combustible gases, such as water-gas, luminiferous; and (3) so to load non-combustible gases with hydrocarbon vapour as to make the combination at once luminiferous and a supporter of combustion. The plans which have been proposed, and the patents which have been secured for processes of carburetting, coming under one or other of these heads, have been almost endless ; and while the greater part of them have failed to obtain commercial suc-cess, they are sufficient to indicate that there is still a pos-sibility of doing much to increase the effect and cheapen the cost of production of gas. Further, although for ex-tensive use none of the gas-making plans can compete with coal-gas manufacture, some of them are of much value for private establishments, country houses, factories, and similar places, where connexion with coal-gas works cannot be obtained.

The carburetting of common coal-gas with the vapour of benzol obtained by the distillation of gas-tar was originally suggested by Lowe as early as 1832, and subsequently by the late Charles Mansfield, who showed that by passing gas over sponge saturated with benzol a very great addition was made to the illuminating power; and he introduced an apparatus by which common gas could thus be benzolized at a point very near the burner. The facts, however, that benzol is a highly inflammable liquid, that the benzolized gas varied in richness owing to the gas taking up much more benzol when the carburetter was newly charged than it did afterwards, and consequently that it often produced a smoky flame, and that sulphur compounds accumulated in the carburetter, as well as the trouble connected with charging the apparatus, all combined to prevent the extensive intro-duction of the process. In later times the value of benzol for aniline manufacture and other purposes would have been a serious bar to its use. Mr Bowditch introduced the use of a heavier hydrocarbon—a mixture of naphthalin with cymol—which he called carbolin, anrl which possesses the advantage of giving off no inflammable vapour at ordinary temperatures, and is, moreover, a substance for which no commercial demand exists. The carburetting appliance had to be placed in immediate proximity to the burners, and either heated by them direct, or by a small subsidiary jet, as the vapour of naphthalin solidifies on a very small fall of temperature and chokes up pipes. Carburetting by means of a solid block of naphthalin introduced into a gas-tight box, and partly volatilized by a strip of copper passing from the burner flame into the box, has recently been proposed, and is now being carried into effect with every prospect of great increase of illuminating power, and consequent economy, by the Albo-Carbon Light Company.

The efforts to introduce carburetted water-gas have been numerous and persistent; and the sanguine statements of the various inventors have led to the loss of much capital through experiments undertaken on a great scale which have always resulted unfavourably. The whole of the proposed processes depended on the decomposition of water by passing it over highly-heated surfaces in pre-sence of glowing charcoal, whereby free hydrogen, carbonic oxide, and carbonic acid gases are produced, the carbonic acid being eliminated by a subsequent process of purifi-cation. The combustible gas so obtained was in earlier experiments charged with luminiferous hydrocarbons by being passed into a retort in which coal, resin, or oil was being distilled, as in Selligue's and other proeesses; or, as in White's hydrocarbon process, both steam and coal were treated together in a special form of retort. Since the introduction of American petroleum, however, most methods of carburetting water-gas have been by impregnat-ing it with the vapour of gasolin, the highly volatile portion of petroleum which comes over first in its distilla-tion for the preparation of "kerosene" lamp oil. Water-gas has been proposed, not only as an illuminating agent, but at least as much as a source of heat; but the heat expended in the decomposition of water is much greater than can in practice be given out by the resulting gases.

Several of the processes introduced for rendering ordinary atmospheric air at once combustible and luminiferous, by saturating it with the vapour of gasolin, have been so satis-factory that this air-gas is now largely used both in America and Europe for lighting mansions, churches, factories, and small rural districts. The general principle of the air-machines will be understood from the following description of the " sun auto-pneumatic " apparatus (Hearson's patent), which is in extensive use throughout Great Britain. Hearson's machine is cylindrical in form (fig. 23), and is

FIG. 23.—Sun Auto-Pneumatic Apparatus.

surmounted by two turrets. Internally the cylinder is divided into two compartments by a transverse portion, one being occupied by a rotary blower, an apparatus similar in construction to the drum of a water-meter, and the other by an elevator or dipper wheel, the function of which is to raise gasolin into the blower chamber, where the gasoKn must be maintained at a constant level. The blower and the elevator mechanism are set in operation by being mounted on a spindle which passes through and outside the cylinder, and is turned either by a weight attached to a length of steel wire or, where convenient, by hydraulic power. The turrets contain (1) a gas-holder which sup-plies gas while the machine is being wound up, should any light be then burning, and (2) a governor to regulate the pressure of the issuing gas. The apparatus works only when gas is being burned, and moves in proportion to the demand on it up to its limit of production. There is therefore no necessity for storing, as indeed would be im-practicable with this form of carburetted gas. The function of the blower is not only, by its revolution, to press forward the gas into the supply pipes, but also to carburet the air by exposing continually renewed thin films of the liquids to its influence on the moist metallic surfaces. The revolu-tion of the blower, moreover, maintains an unceasing agitation in the gasolin, vaporizes the liquid in an equal and uniform manner, and keeps the entire volume at the same temperature throughout. The quantity of gasolin operated on being comparatively large, the tempera-ture of the liquid decreases only slowly, and is in ordinary conditions sufficiently recouped from the external air to keep it in good working order throughout any length of time.

M. Tessie du Motay, who for many years advocated a modified system of lime-light, latterly abandoned that system in favour of a form of carburetted gas. His system necessitates two sets of pipes and a special form of burner,— one pipe supplying ordinary coal-gas or highly carburetted hydrogen, and the other leading in a supply of oxygen, whereby a powerful, steady, white light is maintained at the burner. Philipps of Cologne has also utilized oxygen in a comparatively pure state for burning in a lamp with a wick a mixture of heavy hydrocarbons, which in common air would burn with a very smoky flame.

Other sources of gas, such as tar, and even faecal matters, have been proposed ; and many modified forms of gaseous illumination have been brought forward which, even to name here, would occupy space out of proportion to their importance.

THE FUTURE OF COAL-GAS.

The processes involved in the preparation, distribution, and consumption of coal-gas still remain essentially the same as when the system was first elaborated; but in all details of the industry numerous improvements have been introduced, resulting in marked economy and efficiency of the system. In the meantime new applications of import-ance have been found for coal-gas in connexion with heating and cooking, and as a motive power in gas-engines. Further, collateral industries have been superadded to the gas manufacture, which in themselves are of such value and importance that, were the distillation of coal as a source of artificial light to cease, it would certainly continue to be practised as a source of the raw materials of the coal-tar colours, and of carbolic acid, &c. Were coal-gas to cease to be made primarily and principally for artificial illumina-tion, and to become more a heating and cooking agent, or were it to fall into the position of being a mere collateral product of the manufacture of tar, it is certain that the manufacturing processes would be very materially modified. Costly cannel-gas, with its high illuminating power, is no better suited for a gas engine than common gas ; and for heating purposes a much greater yield of gas might be obtained, which, in burning, would evolve more heat than is sought in making illuminating gas. But as matters now stand, the fact that illumination, heat, motive power, and dye-stuffs are all obtained by means of the manufacture as at present conducted is a consideration of much weight in dealing with rival systems of artificial lighting.

Throughout the whole experience of gas manufacture the efforts of inventors have been directed, not only to improve the manufacture of coal-gas, but also to supersede its ordi-nary processes, and to supplant it by gas yielded by other raw materials or by new systems of illumination. The persistent efforts which have been made to improve coal-gas, and the success which many of the plans exhibit in their experimental stage, warrant the conclusion that the pro-cesses and results of the manufacture are still susceptible of much improvement. When it is considered how exceed-ingly small is the total proportion of illuminants in coal-gas to the bulk of the materials dealt with, it is not difficult to imagine that modifications of processes may be devised whereby a great increase of lighting effect might be practi-cally available, and at the same time a greater percentage of the total heat-giving power of the coal secured for domestic and manufacturing purposes. Notwithstanding the confessed imperfections of the system of coal gas-mak-ing,—the evil odours which attach to the works, the yet more offensive exhalations given off from streets through which the main-pipes are led, the destructive accidents which occasionally occur from gas explosions, and the heat and sulphurous fumes evolved during its combustion,—not one of the numerous substitutes which have been proposed has been able to stand in competition against it in any large town or city where coal is a marketable commodity. As against the system of electric lighting, which is now being brought into competi-tion with it, the ultimate fate of gas may be different. It may be regarded as already demonstrated that for busy thoroughfares—almost, it may be said, for open-air lighting generally—and for large halls and enclosed spaces, electric lighting will, in the near future, supersede gas. The advantages of the electric light for such positions in bril-liancy, penetration, and purity are so manifest that its use must ultimately prevail, irrespective of the question of com-parative cost, and of the fact that municipalities and wealthy corporations have an enormous pecuniary stake in gas-property. That the electric light will be equally available for domestic illumination is, however, not yet so certain ; and until it is demonstrated that a current may be sub-divided practically without limit, that the supply can adapt itself to the demand with the same ease that the pressure of gas is regulated, and that the lights can be raised and lowered equally with gas-lights—till these and other con-ditions are satisfied, the disuse of gas-lighting is still out of sight. Should these conditions, however, be satisfied, there can be little doubt that gas-lighting will enter on a period of severe competition and struggle for existence ; and in the end the material which at one time was regarded as a most troublesome and annoying waste— the gas-tar—will, in all probability, exercise a decisive influence on the continuance of the gas manufacture.

Bibliography.—Clegg, A Practical Treatise on the Manufacture and Distribution of Coal-Gas, new edition, London, 1869 ; Hughes, A Treatise on Gas- Works and Manufacturing Coal-Gas, 5th edition by Richards, London, 1875 ; Richards, A Practical Treatise on the Manufacture and Distribution of Coal-Gas, London, 1877 ; Accum, Practical Treatise on Gas-Light, 4th ed., 1818 ; Journal for Gas-Lighting, London ; Bowditch, The Analysis, Technical Valuation, and Purification of Coal-Gas, London, 1867 ; Banister, Gas Manipulation, new ed. by Sugg, London, 1867 ; Sender, Traité pratique de la fabrication et de la distribution du gaz d'éclairage, Paris, 1868 ; Payen, Précis de Chernie industrielle, 6th edition, Paris, 1877; Schilling, Handbuch der Steinkohlen-Gas-Beleuchtung, Munich, 1860 ; Diehland Illgen, Gasbeleuchtung und Gasverbrauch, Iserlohn, 1872; Ilgen, Die Gasindustrie der Gegenwart, Leipsic, 1874 ; Bolley, Technologie, vol. i., Brunswick, 1862; "Wagner's Jahresbericht der chemischen Technologie, Leipsie; Journal fur Gasbeleuchtung und verwandte Beleuchtungsart, Munich; Reissig, Handbuch der Holz und Torf Gas-Fabrikation, Munich, 1863. (J. PA.)