miser.' It consists of a series of upright tubes through which the feed water passes on its way to the boiler, and is heated above the boiling point, and steam in part generated. The formation of soot on the pipes was the source of the ill success of previous attempts in this direction. This difficulty Mr. Green has overcome by an apparatus of scrapers or cleaners, consisting of rings encircling the pipes, and maintained in constant but slow motion by chains and pulleys driven by a belt from the engine. With this apparatus it is found that when the waste gases escape at a temperature of 400° to 500°, the feed water can be heated to an average of 225°, the temperature of the gases after leaving the pipes being reduced to 250°. To produce this effect 10 square feet of heating surface are provided for each horse power.

Water Gauges.-Every boiler should be supplied with a glass tube, fixed in suitable stuffing boxes, and open at the top and bottom to the boiler to show the level of the water. Gauge cocks at various levels are sometimes employed as supplementary to the glass gauge: both are necessary.

Steam Gauges for indicating the pressure of the steam are also indispensable to the safe working of the boiler. For low pressures an open mercury column is employed on the principle of that used by Regnault in his experiments, and in some cases, to bring the indications within a small compass, the fall of the mercury in the cistern, rather than its rise in the smaller tube, is observed. To avoid the inconvenient length of the open mercury column, the air gauge has been used, in which the mercury in its rise condenses the air in a closed glass tube. This gauge, accurate and sensitive, has yet the fault that the indications decrease in length as the pressure increases, and there is also some difficulty in preserving the quantity of air in the gauge constant. Recently Mr. Allan has overcome these difficulties, by the use of a conical air chamber, so arranged that the indications of the gauge shall be uniform at all pressures, and the air can be renewed at any instant. In M. Bourdon's gauge a curved metallic tube, communicating at one end with the steam boiler, and at the other closed, is used. The curvature of this tube decreases with the increase of pressure in its anterior, and the closed end being free to move is connected with an arrow moving over a graduated

In

arc and marking the pressure. In Schaeffer and Budenberg's gauge the pressure acts on a flat corrugated plate of steel which expands and raises a rack acting on a toothed wheel, and carrying a similar arrow to indicate the amount of pressure. Smith's gauge a flat spiral spring is used, against which the pressure acts through the medium of a plate of india-rubber. All these gauges should be fixed on the boiler with a siphon in which water from the boiler may condense. In this way the pressure in the boiler is transmitted through the column of water, and at the same time the gauge is unaffected by the temperature of the steam.

Safety Valves.-I usually place three safety valves on boilers, as shown in fig. 173; two of these have fixed weights on

Fig. 173.

their levers, and the other is pressed down by a spring balance, and serves to regulate the working pressure of the boiler. The two larger valves for a fifty-horse boiler have each an area of 12 square inches. The third is of only 5 square inches area. These valves are fixed to a common valve seating. The bearing surfaces of the valves are made either flat, conical, or spherical. Flat valves have a tendency to blow off at too

low a pressure, from the steam getting between the bearing surfaces. These valves should always be open to the atmosphere that they may be seen.

Man Holes are required to obtain access to the boiler for purposes of examination and cleaning. In double-flued boilers one must be placed beneath the flues as well as above them. Mud Cocks are placed near the bottom of boilers for the discharge of water and sediment.

Fusible Plugs are portions of metal fusing at a temperature not greatly exceeding the maximum working temperature of the steam, and fixed in that portion of the boiler most liable to be

overheated from deficiency of water. These plugs are of pure lead or of an alloy of bismuth, lead, and tin, according to the temperature they are required to melt at, and they are thought to prevent danger by relieving the pressure of the boiler, and putting out the fire before the plates are injured by overheating. These plugs, however, tend to lose their fusibility, and to become coated with a protecting coat of oxide or sediment, which prevents the communication of heat. They are not a very reliable provision.

Plans for the Prevention of Smoke.-Amongst the earliest of these we may class those which depend on mechanical means for the supply of the fuel.

Of this class is the earliest patent for smoke prevention taken out by James Watt in 1785. By this plan the fire is supplied from above downwards by a reservoir of fuel in contact with the burning mass, the combustion of which is supported by a strong lateral current of air passing direct through the fire to a flue on the other side, aided by a slight downward current beside or through the fuel, which last descends by its own gravity as it is consumed. For the purpose of intercepting and completing the combustion a clear fire is maintained at the entrance into the flues, so that the products of the first fire, being subjected to the intense heat of the second and mingled with atmospheric air, may be effectually consumed.

Apart from the external reservoir, we owe to Watt the dead plate very generally adopted in stationary boilers. The fresh fuel is thrown upon the dead plate, where it gradually cokes, the more volatile constituents distilling over and being consumed by the bright fire beyond. Then the coked fuel is pushed back on to the bars and a new supply introduced in front. This plan, where proper provision is made for the supply of the necessary quantity of air, obviates the production of smoke as effectually as many more complicated contrivances.

The succeeding patentees of the principle of mechanical feeding as a substitute for hand labour, have followed two different plans. Some have made the grate itself to carry forward the fuel, either by revolving horizontally or by rolling forward longitudinally, the grate-bars being connected together to form an endless chain, or by the oscillation of the alternate bars causing

the thrusting forward of the fuel by what has been called a peristaltic movement. Others have made the grate stationary, and have used fans revolving horizontally to distribute the fuel over the grate-bars. In all these cases the coal is supplied slowly and uniformly from a hopper. There is no doubt that the uniform distribution of the fuel over the whole surface of the grate-bars, so far as it is secured by these systems, must be to a large extent advantageous in the diminution of smoke and economy of fuel. At one time they were extensively used, but the complication and expense of the apparatus has led to their general abandonment and the return to hand-feeding.

Other plans for the prevention of smoke depend on a double furnace with alternate firing.

Double furnaces patented by Mr. Losh were in use as early as 1815, and in various modifications have been employed ever since. The principle of double furnaces within the same boiler was first introduced by myself; and the plan adopted has already been described as the double-flued boiler. The two flues enable the stoker to fire alternately, and so maintain a more uniform generation of steam than with a single flue, and the flame passing from one flue mingling with the gases from the other, assists in their combustion. I believe that this simple system of alternate firing, when conjoined with the requisites of the economical generation of steam, viz. plenty of capacity in the boiler, sufficient admission of air, and, what is quite as necessary, careful and attentive stoking, will effect the prevention of smoke without any costly apparatus, so far as that is possible with any given description of fuel. There is this further advantage in double furnaces, that the air required for combustion is necessarily variable. Now a double furnace tends to equalise the supply. The two furnaces fed alternately will not require a maximum or a minimum quantity at the same time, and as the two currents of gaseous products mingle, the surplus air of the one furnace will supply the deficiencies of the other. In this way the tendency is to compensate the supply and demand, and prevent waste from too large or too small a quantity in either furnace.

Others, in seeking the prevention of smoke, have introduced an additional supply of air over the fire.

Mr. C. Wye Williams was one of the earliest, as he has been the most pertinacious and consistent, advocate of the introduction of a large additional volume of air into the furnace, and we have to thank him for the labour he has expended in proving the necessity for air as one of the prime conditions of economy of fuel and success in the prevention of smoke. Mr. Wye Williams contends for a uniform admission of cold air to the furnace, relying upon frequent thin feeding to equalise the needs of the furnace. The peculiar principle of his plan is the mechanical division of the air by causing it to enter the furnace through what he terms a diffusion plate, or partition perforated with numerous small apertures. This is usually placed behind the bridge where the gases needing combustion pass into the flues. There is no doubt this is a convenient method for the introduction of air, and has in many instances effectually prevented the formation of smoke.

Mr. Syme Prideaux contends for a variable admission of air, greatest when the fuel is first thrown on, and decreasing to the ordinary supply through the grate-bars as the fire burns clear. For this purpose he constructs his furnace doors with metal Venetians, which open by a self-acting apparatus when the fuel is supplied. They then gradually close at a regulated speed, altogether independent of the care of the fireman. The air entering through the door is, by an arrangement of plates, warmed as it enters the furnace, and carries back the heat radiating from the door.

All these systems are more or less effective, but I am inclined to think that a judicious engineer, with a careful stoker or fireman, will effect all the objects to be attained with the means placed at his disposal, in a well constructed boiler of sufficient capacity, and with a simple furnace such as has been described in the foregoing chapter, as completely as can be done by any one of the numerous nostrums held forth as the only antidotes for smoke, and promising great economy of fuel.