mouth-pieces N, O, may be easy of access, and one above another: the lower branch L, M, may be placed almost horizontally, and the upper should form with it an angle of about ten degrees. The retort being made red hot, the tar will be decomposed, and the gas, and some other products, will flow from the end M, by the pipe P, into the vessel Q, in which is a partition plate R, fig. 11, extending about half way down, and allowing the heavy products to accumulate for a convenient time before they can interfere with the passage of the gas, which passes to the purifiers, as usual, by the pipe S, S, fig. 12, is a moveable lid for cleansing the vessel. It is not thought necessary particularly to describe the construction of the furnace, which may be varied according as circumstances require.

We may now describe the gas-metre erected by Mr. Accum at the works in the royal mint. It consists of a hollow wheel or cylinder, made of thin iron plate; revolving upon an horizontal axis, in the manner of a grind-stone; this wheel is enclosed in a cast iron air-tight cask containing water.

The cylinder, or wheel, is composed of two circular channels, 1 and 2, fig. 1 plate II. concentric to each other. The larger or outer channel, 1, is divided into three equal compartments, by partition plates, marked a, as shown in the design. The compartments are provided with hydraulic ducts or valves, made at the upper part of every partition plate a, a, a, and by means of them a communication is formed between the larger concentric channel, 1, and the outer case in which the wheel revolves. Similar valves are also placed at the foot of each partition plate, they are seen near the letters a, a, a, and by this means a communication is established, between each compartment or chamber of the larger concentric channel, 1, and the smaller interior circle, 2, of the wheel.

On inspecting the design, it will be seen that the valves are situated in opposite directions to each other; hence there can be no communication wither between the inner smaller concentric channel 2, and the larger compartment of the wheel 1, nor between the latter compartment, and the exterior case, in which the wheel revolves, except through the valves a, a, a, which form the communicating ducts. It will be seen also, that these valves are carried from one chamber of the machine into another, but in opposite directions; the entry into one chamber being in the opposite direction to the hydraulic duct, placed in the other chamber. From these particulars the action of the machine will be obvious.

Let us suppose that the outer case, in which the wheel revolves, be filled with water to about an inch above the axis of the wheel, and that gas is conveyed into the interior small channel, by a pipe, passing along the axis, so as to allow the wheel to turn freely round, and that the pipe is turned up at right angles in the inner chamber, and projects a little way above the surface of the water, as shown in the design. The gas then must enter into the interior chamber of the wheel above the surface of the water, and must press against the adjacent partition; it will therefore cause the wheel to turn round, and, in consequence of this motion, the next par

tition plate will press the gas against the surface of the water, and cause it to pass through the hydraulic opening, in an equal quantity to that which is introduced into the exterior chamber. This alternate filling, and discharging, of the contents of each chamber, will take place once during every revolution of the wheel, and hence the number of times each particular chamber has been filled and emptied of gas may be known. In fact this machine performs the office of three revolving gas-holders, fixed on an horizontal axis, and moving in a cistern, which is the outer case of the machine. One gas-holder, or one compartment of the machine, is always in the act of becoming filled with gas, another is emptying its contents into the outer case, from which it passes into the reservoir, where it is to be stored up, or to the lamps, where it is to be burned, and the third compartment is stationary, or in an equilibrium. The wheel in any situation will therefore always have one of its receiving, and one of its discharging valves open, and consequently it will revolve.

Now to ascertain the quantity of gas discharged by one revolution of the wheel, we need only to know the capacity of the chambers, and add them together. Let us for example suppose, that each chamber contains 576 cubic inches, then one revolution of the wheel discharges a cubic foot of gas. To register the total number of revolutions which the wheel makes in a certain time, a train of wheel-work is connected with the axis of the metre; it consists of a pinion impelling a common train of wheelwork, composed of any number of wheels. The pinion on the axis of one wheel, acts into the circumference of the next wheel, and, the circumference of the wheel being as ten to one, it is obvious while the metre makes 1,000,000 revolutions, if the series consist of six wheels, the last wheel of the series will only have made one revolution. Each axis of the wheels is provided with a finger and dial-plate, divided into ten parts; therefore any number of revolutions may be read off at any time by inspection betwixt 10,000,000 and one. The velocity with which the metre acts, is of course in proportion to the quantity of gas passing through it. Thus suppose there is a burner or gas-lamp connected with the machine, of one foot capacity, lighted, which consumes four cubic feet of gas in an hour, the gas metre performs four revolutions per hour, and so on for every number of burners or lamps, not exceeding the number which the machine is calculated to supply.

The gas-holder, of the original construction, consists of two principal parts; first, of a cistern or reservoir of water, usually constructed of masonry, or of cast-iron plates, bolted and screwed together; and secondly, of an air-tight vessel which is closed at top and open at bottom, inverted with its open end downwards into the cistern of water. This vessel is always made of sheet-iron plates riveted together air-tight, and was suspended by a chain or chains, passing over wheels, supported by a frame work. If the common air be allowed to escape from the inner vessel, when its open end is under the edge of the water in the outer cistern, it will freely descend, and water will occupy the place of the

air; but if the avenue of the escape be stopped, and air be made to pass through the water, the suspended inverted vessel will rise to make room for the air. And, again, if the suspended vessel be counterpoised by a weight, so as to allow it to be a little heavier than the quantity of water which it displaces, it will descend, if the entering gas be withdrawn through an outlet made in the vessel to permit the gas to escape. But if the outlet be stopped, and air again be admitted under the vessel, it will rise again. The apparatus, therefore, is not only a reservoir for storing up the gas introduced into it, but serves to expel the gas which it contains, when required, into the pipes and mains connected with this machine. According to this construction of the apparatus, the interior inverted vessel forms strictly what is termed the gas-holder. It is suspended as already stated in the outer cistern, by a chain or chains, passing over pulleys, supported by blocks and frame work, and to the chain there is affixed a counterpoise balance, of such a relative weight as to allow the gas-holder a slow descent into the water, in order to propel the gas into the mains or vessel destined to receive it, with a very small and uniform weight.

It will be obvious that, when a gas-holder of this construction becomes immersed in the water, it loses as much of its weight as is equal to the bulk of water which it displaces; and hence to render its descent uniform, and to preserve the gas within of an invariable density, at any degree of immersion, a greater counterpoise is required as the gas-holder rises out of the water. Among various methods which have been adopted to attain this object, the ends of the chains by which the gas-holder is suspended, have been fastened in separate grooves, in the edge of a large wheel or pulley, of such a diameter, that the gas-bolder rises to its full height before the wheel makes one revolution. In another groove, in the edge of the same wheel, was fixed the end of another chain, to which a balance weight was suspended. This weight was made nearly equal to the weight of the gas-holder. To equalise the density of the gas within the gas-holder, at any degree of immersion of the vessel, the weight chain was made to pass over a wheel, furnished with a spiral groove, so as to make the radii of the wheel change reciprocally with the relative weight of the gas-holder, and consequently to render the pressure of the gas-holder constant and uniform.

Another and more elegant method of obtaining a uniform elasticity of the gas within the gas-holder, and which has been more generally adopted, consists in passing the chain or chains by which the gas-holder is suspended over a pulley or wheels, and making the weight of that portion of the chain which is equal to the depth of the gas-holder, or that part of it which becomes immersed in the water, equal to onehalf of the weight of the specific gravity of the gas-holder. It is obvious that, before the purified gas can be admitted into the gas-holder, the vessel must be allowed to descend to the bottom of the exterior cistern, in order to get rid of the common air which it contains. This may be effected rapidly by opening the man-hole at the

top of the gas-holder, to cause the vessei to descend completely into the outer cistern filled with water. The man-hole is then screwed up again air-tight, and the machine is ready to receive the gas. It is obvious that the operation of opening the man-hole, for letting out the common air, requires only to be done once prior to the commencing of the working of the apparatus.

The collapsing gas-holder was contrived by Mr. Clegg, and certainly, of all the contrivances which have been invented for collecting and storing up large quantities of gas, this machine must be pronounced to be by far the most simple, economical, and efficient. The striking advantage of the revolving gas-holder is, that it enables the dimensions of the tank to be very much diminished, where the nature of the ground will not permit of a cistern of great depth being sunk, except at an extraordinary expense; but the still superior feature of the collapsing gas holder which we now come to describe, is, that it may be constructed of any required capacity, and adapted to a tank or cistern of such diminished depth, as scarcely to deserve that name. It requires a sheet of water no more than eighteen inches in height, so that it may be constructed in or upon ground of all descriptions, not only with every possible facility, but at an immense saving of expense.

Fig. 2, plate II. GAS LIGHT, exhibits a perspective view of this gas-holder. It is composed of two quadrangular side plates joined to two end plates meeting together at top in a ridge like the roof of a house. The side and end plates are united together by air-tight hinges, and the joints are covered with leather, to allow the side plates to fold together, and to open in the manner of a portfolio. The bottom edges of the gas-holder are immersed in a shallow cistern of water, to confine the gas. By the opening out or closing up of the sides and ends of the gas-holder, its internal capacity is enlarged or diminished, and this variation of capacity is effected without a deep tank of water to immerse the whole gas holder in, as required in the ordinary construction of rising and falling gas-holders. The collapsing gas-holder requires therefore only a very shallow trough of water to immerse the bottom edges of the gas-holder to prevent the escape of the gas introduced into it. The lower edges of the thin gas holder, which dip in water, are made to move in an horizontal plane or nearly so, when they are opened, so that they dip very little deeper in the water when shut or folded together, than when opened out.

For this purpose the top or ridge joints, which unite the two sides of the gas-holder, are slightly raised up when the sides close or approach together, or slightly depressed when the sides open out or recede from each other. To guide the whole gas-holder in this movement two perpendicular rods rise from the bottom of the shallow tank which pass through sockets in the ridge joints at the upper part of the gas-holder. These sockets are secured by collars of leather round the shafts or rods, to prevent the escape of the gas, and they are braced by chains proceeding from their upper extremities and fastened at the ground on each side of the tank

simpler, and yet equally effectual, with the best coal-gas apparatus. The retort is a bent cast-iron tube, which is heated red by a small convenient furnace, and into which oil is allowed to drop by a very ingenious apparatus; the oil is immediately volatilised, and the vapor in traversing the tube becomes perfectly decomposed. A mixture of inflammable gases, which contains a great portion of olefiant gas, passes off; it is washed by being passed through a vessel of water (which dissolves a little sebacic acid, and which seldom requires changing), and is then conducted into the gasometer.

The facility and cleanliness with which gas is prepared from oil, in the above manner, may be conceived from the description of the process. A small furnace is lighted, and a sufficient quantity of the commonest oil is put into a small iron vessel, a cock is turned, and the gas after passing through water in the washing vessel goes into the gasometer. The operation may be stopped by shutting off the oil, or, to a certain extent, hastened by letting it on more freely; the small quantity of charcoal deposited in the retort is drawn out by a small rake, and the water of the washer is very rarely changed.

The gas prepared from oil is very superior in quality to that from coal; it cannot possibly contain sulphureted hydrogen, or any extraneous substance; it gives a much brighter and denser flame; and it is also more effectual, viz. a smaller quantity will supply the burner with fuel. These peculiarities are occasioned, in the first place, by the absence of sulphur from oil, and then by the gas containing more carbon in solution. As the proportion of light given out by the flame of a gaseous compound of carbon and hydrogen is, in common circumstances, in proportion to the quantity of carbon present, it is evident that the gas which contains a greater proportion of olefiant gas, or supercarbureted hydrogen than coal gas, will yield a better and brighter light on combustion. It is necessary, in consequence of the abundance of charcoal in solution, to supply the gas when burning with plenty of atmospheric air; for as there is more combustible matter in a certain volume of it than in an equal volume of coal gas, it of necessity, must have more oxygen for its consumption. The consequence is, that less gas must be burnt in a flame of equal size, which will still possess superior brilliancy; that less is necessary for the same purpose of illumination; and that less heat will be occasioned. From five and a half to six cubical feet of coal gas are required to supply an Argand burner for an hour; two cubical feet to two and a half of that from oil, are abundantly sufficient for the same purpose. One important advantage gained by the circumstance, that so small a quantity of this gas is necessary for burners, is, that the gasometer required may be small in proportion. The gasometer is the most bulky part of a gas apparatus, and that least capable of concentration; and wherever it is placed, it occupies room to the exclusion of every thing else. Some very ingenious attempts have been made to diminish its size and weight, as in the double gasometer, and others, but without remarkable success. Here, however, where the

room required to contain the gas is directly diminished, the object is so far obtained; and when that takes place to one-half, or even onethird, it is of very great importance. It in a great number of cases brings the size of the apparatus within what can be allowed in private houses; and, in consequence of the rapidity with which the retort can be worked, the gasometer may again be reduced to a still smaller size. Another advantage gained by the small quantity of gas required for a flame is the proportionate diminution of heat arising from the lights. The quantities of heat and light produced by the combustion of inflammable gases are by no means in the same constant relation to each other: one frequently increases, whilst the other diminishes, and this is eminently the case when coal gas and oil gas are burned against each other. The quantity of heat liberated is, speaking generally, as the quantity of gas consumed, and this is greatest with the coal gas; but the quantity of light is nearly as the quantity of carbon that is well burnt in the flame; and this is greatest in the oil gas.

The very compact state in which the apparatus necessary for the decomposition of oil can be placed, the slight degree of attention required, its certainty of action, its cleanliness, and the numerous applications of which it admits, in the use of its furnace for other convenient or economical purposes, render it not only unobjectionable, but useful in manufactories and establishments; and these favorable circumstances are accompanied, not by an inferiority in the flame, or increased expense, but by an improved state of the first, and saving in the latter.

Messrs. Taylors have shown great ingenuity in the construction of their whole apparatus, but. the washer and gasometer deserve particular notice for their remarkable simplicity also. In the washer, two planes are fixed in a box or cistern, in a direction not quite horizontal, but inclined a little in opposite directions; the planes are traversed nearly across by slips of wood or metal, fixed in an inclined position on the under surface, and which alternately touch one side of the cistern, leaving the other open and free. These planes being immersed in water, the gas is thrown in under the lowest ridge; and, by its ascending power, is made to traverse backward and forward along the ridges fixed on the planes, until it escapes at the highest part of the uppermost ridge." Thus, with a pressure of five or six inches of water only, it is made to pass through a distance of fourteen or sixteen feet under the surface of the fluid, and become well washed. The smaller gasometers are made of thin plate iron, and, being placed in a frame of light iron work, look more like ornamental stoves than the bulky appendages to gas apparatus, which they supply. The larger ones are made very light, and, when in pieces, very portable, by being constructed of a frame of wood work, in the edges of which are deep narrow grooves; plates of iron fit into these grooves, which, being caulked in and painted over, make a light and tight apparatus. These are easily put together in any place; and may therefore be introduced into a small apartment, or other confined space,

where a gasometer already made up would not

enter.

The general advantages of oil-gas, when contrasted with coal-gas, are as follows:-The material from which it is produced, containing no sulphur, or other matter, by which the gas is contaminated, there are no objections to its use, on account of the suffocating smell, in close rooms. It does no sort of injury to furniture, books, plate, pictures, paint, &c. All the costly and offensive operation of purifying the gas by lime, &c., is totally avoided when it is obtained from oil. Nothing is contained in oil-gas which can possibly injure the metal of which the conveyance pipes are made. The oil-gas, containing no unmixed hydrogen, which occasions the great heat of coal-gas, there is no greater heat in proportion from the flame of oil-gas than from burning oil in lamps, wax-candles, &c.

The apparatus for the production of oil-gas is much less expensive than that necessary to make coal-gas; it occupies much less space; it requires much less labor and skill to manage it; it is not so liable to wear and tear, and not so costly to repair as a coal-gas apparatus; there are no offensive products to remove; and, on its present improved construction, it may be introduced into any dwelling-house without nuisance. The economy of light from oil-gas may be judged of from the following data:-One gallon of common whale-oil will produce about ninety cubic feet of gas, and an Argand burner will require a cubic foot and a half per hour to maintain a perfect light; consequently, a gallon of oil, made into gas, will afford such a light for sixty hours, and the expense, at a moderate price of oil, will be, allowing for coals, labor, &c., not more for one burner than three farthings per hour.

Such a burner will be equal in intensity of light to two Argand oil-lamps, or to ten mould candles. The expense of Argand oil-lamps is usually admitted to be about 14d. per hour each. Supposing ten mould candles to be burning, at four to the lb., will be 2 lbs., costing 2s. 11d., one-tenth part will be consumed in each hour, and the cost of the light is then 34d. per hour. If wax-candles be employed, the expense of a quantity of light equal to a gas-burner, for one hour, by the same mode of reckoning allowing a candle to burn ten hours, and taking the price of wax-candles at 4s. 6d. per lb., will cost about

14d.

gas, which renders so small a volume necessary that one cubic foot of oil-gas will be found to go as far as four of coal-gas. This circumstance is of great importance, as it reduces in the same proportion the size of the gasometers which are necessary to contain it; this is not only a great saving of expense in the construction, but is a material convenience where room is limited. The calculations on the cost of light from oilgas are taken on the usual price of good whaleoil; but, it is to be observed, that cheaper oils will answer the purpose nearly as well, and many of these are often to be procured; and the whole expense may be materially lessened by their use.

In the course of their first experiments, Messrs. John and Philip Taylor were surprised to find that the apparatus they employed gradually lost its power of decomposing oil, and generating gas. On investigation, they discovered that the metallic retorts, which had originally decomposed oil and produced gas in abundance, ceased in a very great degree to possess this power, although no visible change had taken place in them. The most perfect cleaning of the interior of the retort did not restore the effect, and some alteration appears to be produced on the iron by the action of the oil, at a high temperature.

Fortunately the experiments on this subject led to a most favorable result, for it was found that, by introducing fragments of brick into the retort, a great increase of the decomposing power was obtained, and the apparatus has been much improved by a circumstance, which, at one time, appeared to threaten its success. A small portion of the oil introduced into the retort, still passed off undecomposed; and, being changed into a volatile oil, it carried with it a great portion of caloric, which rendered the construction of the apparatus more difficult than was at first anticipated; but, by the present arrangement of its parts, this difficulty is fully provided for, and the volatilised oil is made to return into the oilreceiver, whence it again passes into the retort; so that a total conversion of the whole into gas is accomplished without trouble, or the escape of any unpleasant smell.

The only residuum in the retort is a small quantity of carbon, and the only products besides the gas are a minute quantity of sebacic and acetic acids, and a portion of water, all which are easily separated by passing the gas through a vessel containing water.

The superiority of the light from oil-gas over other artificial lights, is fully shown by its rendering the delicate shades of yellow and green nearly as distinct as when viewed by solar light.

Mr. De Ville of the Strand, who has made many important experiments and observations on gas illumination, with a view of applying it to light-houses, is inclined to estimate the average produce in gas of a gallon of oil, at eighty cubical feet. A single jet burner, giving the light of two candles and a half, consumes half a cubical foot of gas per hour. A double jet consumes three quarters of a foot to give twice the above light, and a treble jet requires one foot. The light of an Argand burner of coal-gas, com