basis for much speculation, and many patents embodying those methods have been taken out. For some purposes it is sufficient to add hydrochloric or sulphuric acids to the crude gas-liquor, to well agitate the mixture, allow impurities to subside, concentrate to the crystallising point, and expose the crystals to a slightly elevated temperature. The resulting salt is still of course very impure, but serves for the preparation of carbonate of ammonia by sublimation with chalk; or certain portions of the impure chloride alone may be farther purified by sublimation. For purposes in which purer ammoniacal salts are desired, the ammonia is displaced from its combinations in gas-liquor by a stronger base, usually lime; it is separated from the resulting lime-salts and impurities by distillatory processes, and its vapour condensed in water containing the acids with which it is wished to unite it. From these solutions the salts are obtained by evaporation and crystallisation. For a list of, and details concerning the various patented processes for "the manufacture of ammonia and ammoniacal salts from the ammoniacal waters of coal gas-works," see 'Pharmaceutical Journal,' vol. xiii., first series, pp. 63 and 113. GAS, MANUFACTURE OF. There are few more remarkable illustrations of the occasional stagnation of human invention for long periods, and of its subsequent feverish activity, than may be discovered in the history of artificial lighting. It is true that from a passage in Ammianus Marcellinus, in which he speaks of the towns in the time of Constantius, A.D. 353, and says that it was easy for conspirators to carry on their treason in them, "ubi pernoctantium luminum claritudo dierum solet imitare fulgorem," it may be supposed that the ancients paid some attention to the lighting of their streets, but the most elaborate critical investigations, and the most careful observations upon the remains of antiquity, have failed to discover traces of any other system of lighting amongst the Greeks or Romans, than very rude oil lamps without chimnies, candles of very imperfect manufacture, torches, or open coal fires. The same methods of public and of private lighting were retained, without improvement, throughout the middle ages; nor was it indeed until about the end of the 17th century that anything like a regular system of public lighting was attempted in any of the capitals of Europe. The first decided steps towards the improvement of this important detail of domestic comfort and public security, seems to have been taken under the guidance of M. de Sartines; for under his protection Laugrin introduced the reflector lamps, which were subsequently improved by Quinquet; and finally Argand perfected the ordinary oil lamps, by the introduction of the cylindrical wick and chimney, with a double current of air. Letters patent for this invention were granted on 5th January, 1787, but they were abolished during the French revolution of 1793. Whilst the methods of artificial lighting were thus slowly progressing in the beaten track, Mr. Murdoch was engaged in experiments on the combustion of coal gas; and in 1792 he lighted his house and offices at Redruth in Cornwall, by that means. In 1798, Murdoch also lighted the shops of Messrs. Bolton and Watt at Soho with "inflammable gas; and in 1802 it appears that a M. Lebon proposed to light a portion of Paris in a similar manner. The Jesuits' College at Stonyhurst, was amongst the earliest public institutions to adopt the "new light;" and it was in the course of executing this work that Mr. Clegg introduced the method of purifying the gas by passing it through lime water; this took place about 1807. In the same year Mr. Winsor, who had taken a very conspicuous part in diffusing the knowledge of the new method of lighting, obtained permission to apply his apparatus to a few lamps in Pall Mall. The first act for the establishment of a gas company was passed in 1810, and under its powers the Chartered Gas Company commenced operations, in spite of the sneers of scientific men and the opposition of practical ones. It was not until the 31st of December, 1813, that gas lighting was employed on a large scale in streets, by the lighting of Westminster Bridge; and on April 1st, 1814, the parish of St. Margaret's Westminster substituted gas for oil throughout their district. About the beginning of the year 1820 Paris imitated the example thus set; and subsequently to that period almost every city of importance in England, and on the continent, has adopted the use of gas. The meter, for registering the consumption of gas, was invented in the year 1815 by Mr. Crosley, and since that period gas has been commonly applied in lighting private houses. It may convey some idea of the importance of the gas manufacture of London to state here that there are no less than nineteen companies established for this purpose, with a paid up capital of nearly six millions sterling. The manufacture of gas is conducted upon rather different principles in the large establishments of such a town as London, from those observed in the majority of small country works; and it will therefore be desirable to notice in detail the various processes of the manufacture, in order to explain some of the reasons for these particular variations. In so doing, the system to be observed is that of the ordinary course of manufacture; or to discuss, 1st, the carbonisation and distillation of the coal; 2nd, the condensation and purification; 3rd, the storage of the gas; 4th, the distribution, under the sub-divisions of the mains and of the fittings; and 5th, the accessory buildings, or processes employed. The conditions connected with the discharge of the public service by private companies, or by municipal bodies, will form the subject of a special article [PUBLIC LIGHTING], on account of the warm discussions which have lately taken place, and of the singular differences of system which prevail in the matter. 1. The distillation of the coal, or the process by means of which the illuminating gases are separated from the solid carbon of the coal, is effected in close vessels known by the name of retorts. These retorts are at the present day long, closed, horizontal vessels of cast iron, or of fire clay, or occasionally in large town-works they are built up with fire bricks; and, as might be expected, they are of every variety of form and cubical capacity. Generally speaking, however, retorts of cast iron are made of the shape of the letter D, and they are about 7 feet 6 inches long, by 20 inches wide, and 12 inches deep; their capacity being such as to allow them to carbonise 120 lbs of coal in each charge. The clay retorts are sometimes made of the same form and dimensions as the iron ones, but they are very frequently made perfectly cylindrical; or in large works they are made in lengths of about 16 inches, joined together with fire clay, so as, in fact, to form ovens about 7 feet long by 5 feet wide, and 18 inches high, as in some of Mr. Grafton's works; or, as in the case of the Phoenix Works, London, they are put together in three pieces, making retorts of 20 feet long, by 16 inches diameter. There are numerous descriptions of retort ovens formed of fire tiles and bricks now in use in the London gas factories, such as those for instance of the Westminster station of the Chartered Company, which are about 22 feet clear length, by about 20 inches wide, and 13 high, and are fed from both ends; or again in Paris, and at the London Gas Works, Vauxhall, a large oven is occasionally used, capable of carbonising as much as six tons of coals at a time, for the purpose of obtaining a description of coke adapted for foundry, or other analogous purposes, at the same time that the illuminating gases are secured. HO Now the differences we thus find to prevail in the shape, dimensions, and materials of gas retorts, extend likewise to the manner of their arrangement in the same bed, as the assemblage of retorts heated by one furnace is called. According to the size of the locality to be supplied, 1, 2, 3, 5, 7, or even as many as 12 retorts are fixed in the same setting, though it is very questionable whether any real advantage be gained by exceeding the number of 7 or 8. The advantage of placing several retorts in one bed, being only that they require, under such circumstances, less fuel to bring the coal to the required temperature, there must be a point at which the increased cost of labour, from the difficulty of drawing (or removing the carbonised coal from the upper retorts) must balance the economy of fuel. Assuming then that seven or eight retorts in a bed are the most convenient and economical arrangement in large works, the setting of the retorts in small works would be regulated on the calculation that each ordinary D retort will yield a fair working result of from 2500 to 2700 cubic feet per day: and the number of such retorts in a bed must be regulated according to the size of the town and its consumption of gas, so as to have the smallest possible number of furnaces alight at the same time. The nature of the fuel burnt under the retorts may modify the ordinary course of proceedings, however, and in some character of the foundations upon which the retort beds are placed may render it unadvisable to concentrate a great weight in one spot. cases the Whatever be the shape of the ordinary retorts (that is to say, of those in which no attempt is made to obtain metallurgic coke), the exposed end of the retort bears a mouthpiece, to which is attached the rising or stand-pipe, or the pipe through which the crude gas passes from the interior of the retort into the hydraulic main, to be described hereafter. As this mouthpiece is frequently exposed to severe shocks, it is made of cast iron, even when clay retorts are used; and it bears two ears which receive the ends of the lids, or covering-plates, and the cross-bars used for fastening the lid (by means of a strong square threaded screw) to the mouthpiece. The imperviousness of the junction of the mouthpieces and lids is, moreover, assisted by covering the bearing surface with a luting composed of the spent lime from the purifiers (when that material is used) mixed with fire-clay. The standpipe is usually formed in a separate casting from the mouthpiece, and it is bolted on to a species of saddle formed to receive it; the diameter at the bottom, in ordinary retorts, being generally about 6 inches, usual, it may be here added, to calculate that, when retorts are heated by the waste coke of the gas factory itself, about from 25 to 35 and at the top 4 inches. In such cases the length of the stand-pipe is about 10 feet; but the number of retorts in a bed must naturally affect this dimension. At the top of it there is fitted one end of a short bent pipe, known as the H-pipe, or the bridge-piece, which by its other end is connected with the descending pipe, and by it with the hydraulic main; the lower end of the descending pipe being carried about two or three inches below the surface of the water and tar of the main, and thus converted into a species of hydraulic seal, to prevent the return of the gas to the retort after it has once risen through the water. The H-pipes bear at the top a moveable cup. per cent. of the coke made is burnt under the retorts. Sometimes tar is burnt in the furnaces; but as the local markets for the residual products of gas-making constantly vary, it is not advisable to attempt to discover any general law in respect of the nature of the fuel, any more than it was in the case of the size and details of the furnaces and flues. The mode of charging the retorts is another detail of the service of a retort house in which many differences prevail; for in small country works, where it is important to keep down the outlay for labour, the stoker is made to charge the retorts by means of a shovel; whereas, in large works, the whole of the charge is frequently placed in a scoop, and is inserted in the retort at once: the object to be attained in either case being to distribute the coal for distillation in a uniform coat of about 5 inches in thickness all over the bottom of the retort, and to leave the mouth open for the shortest possible time, in order to avoid any unnecessary lowering of the temperature. There are, in fact, few details in the process of gas-making of equal importance to the regulation of the temperature of the retorts, and it is impossible to call the attention of the workmen too forcibly to the subject. If, for instance, the heat be maintained for any length of time at a very high point, the retorts will be burnt out rapidly, and the proper quantity of gas per ton of coal will not be obtained. If, on the other hand, the heat of the retorts be too low, a large quantity of tar will be formed, the quality of the gas will be inferior, and the rate of distillation will be protracted. When the temperature of the retorts is properly regu The hydraulic main is a species of iron trough, rather more than half filled with water, which is sometimes carried on columns in front of the mouthpieces of the retorts, but is, generally speaking, placed a little within the line of the retort setting, so as to allow of its being supported from the same. Whether the hydraulic main be within or without the front line of the retort beds, it must run from one end of the retort house to the other; and under any circumstances it must be kept sufficiently high to allow good head-room beneath it, and to remove it from the direct action of the flames which escape when the retorts are being charged. The dimensions of the main must be sufficient to contain the quantity of tar and water which should be able to close the immersed ends of the H-pieces when the mouthpieces of the retorts are opened; for at that period the pressure of the gas into the main will cause the tar to rise in the pipe to a point depending upon the pressure itself. Another condition to be observed in the construction of an hydraulic main is, that the exit pipe should be placed above the surface of the tar, in order to prevent any interference with the passage of the gas to the condensers. A small outlet pipe is inserted at the ends of the main, through which the surplus tar flows into the tar cistern; the lower end of this overflow pipe must be care-lated, and the quality of the coals used is of an ordinarily good descripfully sealed, in order to avoid any escape of gas through it. It would perhaps be dangerous in a general description of the process of gas-making, to attempt to lay down the laws for the construction and dimensions of the furnaces and flues required for the purpose of heating the retorts to the requisite temperature for the distillation of the coal; because the varieties in the dimensions of the retorts and of the materials employed in their construction, as well as of the fuel used for heating the retorts, render it necessary to introduce many modifications in the furnaces. The object to be attained is, to heat cast-iron retorts to a uniform cherry-red heat, in the interior; or to heat clay retorts to a white heat, with the smallest possible expenditure of fuel; and, after all, much must be left in these matters to the tact and skill of the operative engineer who fixes the work. It is ARTS AND SCI. DIV. VOL. IV. tion, it may generally be calculated that a ton of coals will yield between 8700 to 9300 cubic feet, measured by the station-meter before being passed into the gas-holder. 2. From the hydraulic main the gas passes at an average temperature of 120° Fahr. into the condensers. The condensers in the best modern works are formed by means of a series of vertical pipes, whose lower ends dip in water for the purpose of arresting the passage of the tar carried over, and which are connected by a second series of inclined pipes passing from the bottom of the first vertical pipe to the top of the second pipe. In its passage through these pipes the gas parts with some portion of its heat by radiation (and therefore the diameter of the vertical pipes should be made as large as possible) until at last, at the final exit pipe, the gas is lowered in temperature to about 60° (but X covered with screened porous coke; and the gas, passing through these materials, and being in its passage exposed to the action of a stream of water filtering constantly through the coke, parts with the remaining portion of its tar and a considerable portion of the ammonia it holds in solution. In large works, what is called a washer is often used instead of a scrubber, and in it the gas is made to pass through a solution of muriate of manganese, whose action upon the ammonia is the same as, or even more effectual than, that of the scrubber. From thence the gas passes into the purifiers, where the sulphuretted hydrogen, and many of the remaining ammoniacal elements it may contain, are eliminated, so that on leaving the purifiers the gas is fit to be passed through the station-meter into the gas-holders in a state adapted for household consumption. As a general rule, the exposed surface of the condensers should be in the proportion of 150 feet superficial to every 1000 feet of gas operated upon; but every description of coal will require a special treatment in this respect; because the more highly bituminous coals, for instance, part with the tarry matters volatilised in the retorts with more difficulty than do the ordinary coals; and it is desirable to retain as much as possible of those matters, provided that they be not exposed to deposition at any time in the course of their passage through the distributing mains. When the gas leaves the condensers it usually contains impurities, consisting of about 1 parts of ammonia, 8 parts of sulphuretted hydrogen, and 25 parts of carbonic acid in every 1000 measures of the gas, according to the author of the very remarkable papers on the 'Chemistry of Coal Gas,' inserted in the 2nd and 3rd vols. of the 'Journal of Gas Lighting.' For the purpose of removing these impurities it would appear that the most theoretically perfect course would be to pass the gas, firstly, through either the sulphate, or the muriate of lime, and then to pass it through the pure hydrated lime in powder. It happens, however, that some of the impurities which are retained by these forms of lime, namely, the hydro-sulphate of ammonia, and the hydro-sulphate of lime, are either extremely volatile, or are susceptible of decomposition on exposure to the air; and under such circumstances their removal from the purifiers, when the lime itself has been saturated, gives rise to so great a nuisance that in crowded districts, or in the centre of towns, it is necessary to employ some other material than lime as the basis of the purification. It is for this purpose that metallic salts have lately been introduced; and in some of the most scientifically managed works (as at Liverpool for instance), the process adopted within a very recent period, has been described as being as follows: "Common green copperas (sulphate of iron) is put into a mill, a little water being added to make it into a pasty mass. Slaked lime is then added, in the proportions of one part of lime to two of copperas, with water sufficient to make the whole, when ground together, of the consistency of a stiff paste. In this state it is removed, and cut into pieces with a spade, and as it dries it forms a powder which consists of sulphate of lime and hydrated sesquinoxide of iron. The powder, after being sifted is put into the purifier, where the oxide of iron becomes reconverted into sulphuret of iron; and that again by exposure to the air becomes reconverted into oxide This process is of iron with a deposition of sulphur in the mass. repeated about 28 times, and then it is found, that owing to the accumulation of sulphur, the same material cannot again be put into the purifier. In this state the material is roasted in an oven, the collected in a separate form; and during the roasting process, the sulphur being in the first instance distilled from the mass, and iron becomes red hot, and is rendered anhydrous. When taken out of the ovens the iron is thrown into open yards to cool; and is then moistened with water, in order to reconvert it into the hydrous oxide, and again put into the purifiers to commence a fresh succession of repeated actions." In the majority of the London works saw-dust is mixed with the oxide of iron to increase its active surface; and in some cases the gas is made to pass through the muriate of manganese before passing through the iron, whilst in others it is passed through dry lime purifiers after leaving the iron, because the action of the latter material is not sufficiently energetic to remove all the ammonia, nor does it materially affect the carbonic acid, in its various forms of the gas. The lime used for the purification of, gas should be of the purest and richest description, such in fact as would be obtained by the calcination of chalk, or of crystalline carbonates of lime, and which expands in volume very considerably when slaked. It should be kept in store in lumps, but ground before being used, and mixed with a sufficient quantity of water to bring it into a stiff plastic state; and in that state it must be spread upon the screens of the purifiers in an even sheet of, at the most, 24 inches in thickness. The successful working of the lime depends, it may be observed, on the evenness of the layer; but the most economical condition for its use is that it should expose a large surface to the gas passing through. Generally speaking, it is calculated that one bushel of ground dry lime will when slaked, spread over a surface of 25 superficial feet, and suffice for the purification of 10,000 cubic feet of Newcastle coal gas. In the oxide of iron purification, a greater surface is required than when lime only is used; but the proportion usually adopted is to make the area of the purifiers such that they should present a surface of one superficial foot to every 150 cubic feet of gas they are intended to pass; and the diameter of the connecting pipes is made in inches, equal to the square root of the area of the purifiers in feet superficial. The use of wet lime purifiers is now so universally abandoned, that it is not worth while to enter into a description of the mode of forming or of working them. One of the most important duties of the manager of a gas factory is to ascertain that his purifiers operate successfully, and for this purpose he should frequently test the gas by means of paper, prepared by being steeped in a solution of acetate of lead in distilled water, one of the best tests for sulphuretted hydrogen. The papers should not be changed in colour by contact with the gas in the last of the series of purifiers. Sometimes papers steeped in a solution of the nitrate of silver are used as tests of the purity of gas from sulphuretted hydrogen. Test papers of litmus are used to detect ammonia, or carbonic oxide. It may be as well here to add that the order in which the gas is made to pass through the purifier and the scrubber, is frequently the reverse of the one described above; and in small works the use of the scrubber is often entirely dispensed with. 3. The first operation which it is desirable to adopt in the storeage of gas, is to pass it through the station meter, in order to be able to control the working of the retort house, and to check roughly the consumption in the mains and the private distribution. For the former purpose it is necessary that the wheel-work of the meter should carry a "tell-tale," by means of which the passage of the gas is recorded in such a manner as to allow the rate of working, hour by hour, to be identified. The value of the station meter, as a check upon the actual consumption, must vary in a marked degree with the circumstances of the localities, and the mode of manufacture; for the rate of leakage, and the amount of condensation, which may take place after the gas should have passed the meter, differ in almost every town. It is, however, by no means rare to find that the loss, known in gas-working accounts, under the name of "unaccounted for gas," amounts to as much as 20 per cent. of the total quantity passing through the station meter; and the efforts of the superintendent must be earnestly directed to reducing it to a minimum. Even if a portion of the loss be attributable to the condensation of the holder and pipes, it argues an unsatisfactory state of working; for the temperature of the gas ought to have been lowered, before it was passed into the purifiers, to such a point as to render any subsequent condensation impossible unless in intensely cold weather. The gas holder is a large wrought-iron vessel, either of one or more ifts, according to the nature of the locality (but wherever it is possible, of only one), which should be made large enough to hold one day's normal consumption (at the period of the shortest days) of gas. The holder is, in fact, an inverted cup, working in water, and in a close brick or iron tank, and rising by the elasticity of the gas entering through the inlet pipes; the discharge taking place through the exit pipes to the governor, by the mere weight of the holder. The pressure required to raise a holder is ascertained by the formula p = in which p = the pressure in inches (of a column of water); w= the weight of the holder in pounds; and a = the area of the water a 5.2 547 w d2' = surface in feet; the constant 5.2 is the weight in pounds of a super ficial foot of water one inch thick. The effluent pressure is found by the formula p = in which p and w represent as before the constant 547 representing the weight of a column of water in inches, pressure and weight, and d the diameter of the holder in feet; the of the area of the holder. Strictly speaking it would be necessary to allow for the levity of the gas, and for loss of weight in the sides of the holder as it may descend; but these considerations are so habitually neglected in practice as to justify the reference of the practical student to Clegg's' Practical Treatise on the Manufacture of Coal Gas,' for further details. The Governor is a machine for the purpose of regulating the pressure upon the outlet pipe, according to the hours of the day, and the draught upon the mains; and one of these machines must be placed at least upon each leading main of a town distribution, large or small. There are numberless patents for the construction of governors; but practically the system described by Clegg is the best, and therefore a sketch of it is appended. Its action and principles are described in detail in the work above mentioned. Attached to the outlet-pipe of the governor there should be placed a self-registering pressure gauge, in order to control the operations of the men charged with the regularisation of the pressure; and one or more ordinary pressure gauges should be placed upon the leading main, for occasional observations. The self-registering gauge invented by the late Alexander Wright is one of the best, if not actually the best, instrument of this description. 4. The distribution of gas is effected by means of cast-iron mains, in all cases where their diameter exceeds 2 inches; and indeed in all cases where the pipes would be exposed to the action of moisture it is preferable to employ cast, rather than wrought, iron for street mains, even at the risk of employing larger ones than would theoretically be required. The service pipes, or those through which the gas is led into the consumer's meters, are, however, almost always of wrought iron when the diameter exceeds half an inch; below that dimension they are either made of tin or of composition, on account of the greater facility with which they can be bent to the abrupt curves frequently required in house fittings. The joints of the castiron pipes are usually of the description known as socket-joints; those of the wrought-iron pipes are of the description known as screwcouplings; whilst those of the smaller pipes are made by soldering in the ordinary way. The wrought-iron services are tapped and screwed into the cast-iron mains, and the composition pipes are joined to the wrought-iron ones by means of a brass screwed end, which is run upon the composition and fits into a female screw on the service pipe. Before describing the laws which are now admitted with respect to the flow of gas in pipes, it may perhaps be advisable to make some remarks upon the quality, and the mode of manufacture, of gas mains; because the economical results of any operation of this description must, after all, greatly depend upon the manner in which those mains discharge their functions. It has been stated above, that the "unaccounted for gas" frequently amounts to as much as 20 per cent. of the total quantity made; and as very probably one-half or one-third of this loss is to be attributed to the permeation of the gas through the mains, it becomes a matter of serious importance to prevent such a loss. The principal difficulty lies in this case, as in so many other practical ones, in the prices of the various goods considered; and gas companies are too often tempted to use cheap porous pipes, obtained from first runnings, rather than incur the expense of sound second runnings. This is a very mistaken economy; and it may be laid down as a rule in these matters, that no gas mains should be allowed to be made in any works where a blast furnace exists; all mains should, in fact, be made from second runnings, and under the immediate inspection of the engineer of the gas-works; and they should all be cast vertically, with a requisite length of feeder to ensure the solidity of the metal. The very conditions of manufacture of wrought-iron pipes render them less likely to be porous than cast-iron ones frequently are; but the cheap composition pipes are so fearfully defective, that the greatest precautions should be taken in their application, and none but the most respectable and most experienced gas-fitters should be employed. No doubt many of the sad accidents recorded, of explosions in houses, are to be attributed to the use of inferior composition service-pipes for the distribution of gas. In arranging the dimensions of the mains of a gas distribution it is to be observed, first, that there is an economy of working expenses, in making the mains rather larger than would theoretically be required to deliver the quantity of gas they are originally intended to supply; because, in such a case, the gas might be made to work under a diminished pressure, and thus the leakage would be diminished. The diameters of the mains are to be ascertained by dividing the district to be supplied into certain sub-districts, according to their relative consumption, and then applying the formula v = 2g; and с 1) sv2; in which the new terms, M = a co Ρ v2; the former for the purpose of determining the velocity of 2g the gas due to a certain pressure; and the latter, the pressure for a given velocity. In these formula, p= the pressure; v= the velocity; g the force of gravity 32:19;s the weight of a cubic foot of the gas in pounds; but no attention is paid to the retarding effects of friction in the mains, or to those produced by bends, or other incidental causes. Taking the friction alone into account, it has been considered that the total pressure at the end of a pipe may be represented by the formula p= +Ml a efficient determined by experience to be ·00011; 7 the length of the main; c = the inner circumference of the pipe; and a = its area. Clegg enters at length into the mathematical reasoning on this subject; and the reader is referred to his work for further details upon it; but it may suffice for present purposes to state that he finally quotes, for ascertaining practically the quantity discharged by a pipe, the formula ; in which q the quantity sought; d the diameter of the pipe in inches; h=the working pressure in inches; the length of the pipe in yards; and s = the specific gravity of the gas compared to that of atmospheric air as unity. Provided the radius of the bends upon a length of main be large, there is little necessity for taking them into account; and indeed the simple precaution of making the diameter of the pipe rather larger than would theoretically be required, would obviate any necessity for so doing. = 1350 de = = Under any circumstances, when a town is characterised by great irregularities of level, it is desirable to insert, upon the leading mains, species of receivers provided with governors at the points where the marked changes of level occur. When, on the contrary, the town is (like those of Holland) nearly upon a dead level, and close to the water-line of the district, it becomes necessary to insert numerous syphons in order to relieve the mains of the tar and other liquids which may find their way into those mains; and in towns of this character it is desirable to give an inclination of about 1 in 600 towards the syphons. It is of course necessary to maintain an efficient seal in the syphon box, and to provide a small discharge pipe to draw off any excess of tar or water. Another detail of pipe laying, which requires to be considered, is the position of the valves to be placed for the purpose of shutting off the gas during the repairs of the mains. These valves must be placed so as to interfere with the service in the smallest possible degree; but no definite rule can be laid down in any of the matters of local detail. In laying the pipes great care should be exercised in forming a regular and incompressible bed, and in rigidly adhering to the rates of inclination prescribed. As a general rule, the pipes for conveying gas may be laid at the depth of 2 feet from the surface; but of course the latitude of the locality considered may render it advis. able to modify this law. If any water-pipes should be laid in the same street as the gas-pipes, the latter must be kept as far from them as possible, and under any circumstances they must be placed above the water-pipes; serious accidents have indeed occurred from the neglect of these precautions, for occasionally the gas has found its way into water-pipes, by the singular action known by the names of endosmose and exosmose, and explosions have taken place at the outlet. The greatest possible care should be taken in laying the mains to ensure the tightness of the joints, as it is at such places that the gas most commonly escapes. The following tables may be considered to represent, first, the dimensions and weights of gas-pipes of various diameters; and secondly, the quantity of gas which pipes of the respective diameters will deliver, under ordinary circumstances, with the pressure of 1 inch. Table I. Dimensions and Weights of Pipes. Column No. 1 represents the diameter in inches; No. 2 the number of belts in the length of pipe; No. 3, the thickness; No. 4, the mean weight of each pipe; No. 5, the tolerated deviation of weight of each pipe; No. 6, the net length of the pipes; No. 7, the ordinary price for laying and jointing, without earthworks, calculated per yard lineal. With respect to the consumers' fittings, it may be as well to observe that the public lamps in streets should be made to consume at the rate of 5 feet per hour on the average of the night; and that gas companies would find it desirable to use regulators upon these lamps, both for the sake of checking the waste of gas, and of improving the light from them. As a rule, the lamp-posts in the streets of populous towns should be placed alternately on either side of the street, and at a distance of from 22 to 30 yards (measured on the axis); though in the less frequented districts, the distance asunder of the posts (measured as before) may be carried, without inconvenience, even to 60 yards. The height of the flame in the lamps should be about 12 feet above the ground line. In private houses, the character of the fittings, and the quantity of gas burnt to light a given space, depend so much upon the taste of the consumer, and the details of the burners are exposed to such constant changes from the whims and fancies of the gas fitters, that it is im possible to lay down any positive rules with respect to them, other than a few general ones derived from the ordinary principles of lighting. Thus, it is preferable to distribute the light of a room by means of a number of burners consuming small quantities of gas, rather than to concentrate it in one central burner consuming a large quantity; because the effect of artificial light diminishes in the ratio nearly of the square of the distance from its source. Of the various kinds of burners used, the ordinary Argand burner seems, quantity of gas for quantity, to be the most convenient and the most economical. The bats-wing burners are the next in the order of relative economy; The so-called solar lamps are and the fish-tail burners the last. admirably adapted for lighting large rooms, wherein it may not be desirable to distribute the sources of light, or wherein it is desired to establish an active upward ventilation; but they are expensive, and on account of the heat they evolve, they render it necessary to adopt certain precautions in the construction of the building. Whatever description of burner be used, attention must be paid to regulate the escape of the gas, in such wise as to prevent its issuing with too great a pressure, because in the latter case a large portion of the gas would only be partially burnt, and would thus cause much smoke. As to the asserted injury to furniture, books, &c., from the use of gas in dwellings, it is curious that in an old French work, published at Lausanne in 1770, precisely the same accusation was brought against the use of coal in fire-places as is now brought against gas; and the inference to be drawn from this tale is, that certainly the latter cannot be at fault, whilst probably there may be exaggeration in both cases. The heat of gas is also at times objected to, but if the same amount of light were obtained by any other method of artificial lighting, even greater heat would be evolved; and in either case attention is required to ensure an efficient ventilation. This is the more necessary from the fact, that by whatever means artificial light is obtained, its brilliance can only be secured by the consumption of oxygen; and therefore it is essential that an ample supply of fresh air should be introduced to maintain the combustion, if for no other purpose. A great deal of attention has lately been directed to the question of gas-meters, and a special act of parliament has been passed (22 & 23 Vict. c. 66) to regulate them. It may suffice here to say that meters are of two sorts, dry or wet; and that the latter work satisfactorily, if kept properly filled, for a longer period than the former. Both of them, however, must now be made with such accuracy that they should only have a range of error of 2 per cent. above, or below, the mathematical quantity indicated: and if the best compensating meters be used, there need be no fear of any subsequent tampering on the part of the gas companies. It is usually calculated that the con |