Near the extreme highest point of the apparatus the "filling cap" is situated, and the expansion tube is situated above this. The filling cap is the cap on the elbow branch shown, and it is up to this point that the apparatus is filled with water. The cap is specially made to admit of frequent removal and replacing, and when the engineer has finished an apparatus he leaves a strong wrench with the caretaker of the building with instructions to remove this cap about once monthly and pour in any water that the pipe will accommodate. The caretaker does this, of course, when the fire is not alight, and it is peculiar that it should be necessary in an apparatus that is sealed at all other times.

Assuming that the apparatus is full of water up to the filling cap, it follows that the expansion tube is left full of air. On lighting the fire the water at once commences to expand - to increase in bulk, and, finding no way of relief anywhere else, it starts forcing its way into the expansion tube with a consequent compression of the air there. As the heating continues so does the bulk of water increase, and the air compression becomes greater, and as the compressed air must press back on the water with exactly the same force as that exerted in compressing it, it will be seen how the water is prevented boiling and the boiling-point never reached. The hotter the water becomes, the greater is the pressure exerted on it by the air, and it is difficult to see how a boiling-point can be arrived at. It might be assumed that the same result could be attained without an expansion tube, but the expansion of water is so strong a force that without a cushion-like substance such as air to exert this force upon, the apparatus would be fractured somewhere in a very brief space of time.

The description given, so far, only applies to quite small works, and although the details of filling and expansion remain the same, there have to be special features in the furnace coil when the undertaking is a large one. If 7/8-inch (or 1-inch) pipe were run in this way for low-pressure heating, a circuit of about 200 lineal feet would be considered a maximum length, if the return pipe was expected to be of a useful heat. With high-pressure work the circulation is more rapid and the conditions more favourable ; therefore a length of 500 feet is allowed as having a sufficiently hot return pipe. If a circuit is of greater length than 500 feet there is a risk - a tolerable certainty, in fact - that part of the return will be of a very limited degree of usefulness, for it will be below the temperature expected of high-pressure heating surfaces.

The above being recognised, it becomes necessary in large jobs to run the piping in sections or circuits of about 500 feet each, yet this must not be done by any system of mains and branches. It is a better plan which is adopted, it being arranged that each circuit have its water come back to the furnace, to be reheated, and then go out again, and this is effected by using a heating coil consisting of two or more intercoiled pipes.

In Fig- 57 is shown a three-pipe furnace coil suited for a 1500 feet job, or thereabouts. It will be seen that this coil has three flow and three return connections, there being, in fact, three distinct coils intercoiled. What has particularly to be pointed out is that each coil must not be given a section independently. If this were done then each circuit would require its own expansion tube and other fittings, and unless they were very carefully proportioned it is probable that, while one overheated the others would be partial failures. What has to be done is to let the circuit which starts out from the flow connection of one coil come home and join the return connection of another coil; and if Fig. 58 is studied it will be seen that, by this arrangement of connections, the whole apparatus remains one endless tube, as it should do. It will be seen that the top flow is made to communicate with the bottom return, and that each flow returns its water to a different coil to that it started from. If the pipes were followed up or straightened out they would be found to be one endless circuit of pipe as much as those in Fig. 56 are. There is no rule to follow as to which flow shall communicate with a certain return ; but a good plan, when the circuits differ in length, is to let that which is longest come back and join the coil which has the shortest run from its flow connection. This helps to equalise the general heating.

Fig. 57.

A detail of Fig. 56 which remains to be explained is the method of jointing the pipe. The screw or thread on the tube is finer than that of ordinary hot water or steam tube, and special dies are employed for this. Each piece of pipe, and socket, has a right-hand thread at one end and a left-hand thread at the other ; so that when the two ends of pipe enter a socket, and the socket is secured up, the ends approach one another, and ultimately meet in the middle of the socket. The ends of the pipe, which meet in the socket, are finished, one quite flat and the other with an annular chisel or coned edge, as shown in Fig. 59 ; and the result of screwing the socket up is to make the end of pipe with the chisel edge embed itself in the flat edge, and this, by itself, makes the joint. No jointing material whatever is used. In preparing the ends of the pipe the fitter commonly relies on a file to finish the flat end, also the outside bevel of the coned end ; while for the inside edge of the latter a tool like a countersink serves. Those who do the work largely sometimes have special tools for both ends, the chief object being to get the ends quite true, so that they precisely meet on coming together in the socket When the fitter relies on a file and countersink for finishing the ends he carries a small steel square with which the ends can be tried as he proceeds with them.

Fig. 58.

Fig. 59.

The Limited Pressure is a modification of the high-pressure system, and in which the pressure - and temperature - can never rise beyond a certain desired point. This is effected by the use of a special form of combined outlet and inlet valve, fixed in the position that the expansion tube otherwise occupies. The valve is used in place of the expansion tube, and, as will be seen, the filling cap can be dispensed with. In other respects the apparatus remains the same as the high-pressure system.

Fig. 60 illustrates the details of a valve suited for this purpose ; but, it may be mentioned, engineers professing this work have, in some instances, valves of their own design and make. The upper part of the valve, it will be noticed, is practically a dead-weight safety valve - a valve which opens and provides an outlet when sufficient pressure is exerted from the inside. The lower part of the appliance resembles a spindle valve, which closes and holds tight when pressure is exerted upon its head, but opens easily to pressure from beneath the head, no weight or spring being needed to control the operation of this.

Fig. 60.

Assuming the apparatus is ready for use and the fire just lighted, the water at once commences to expand. The apparatus being wholly full of water, there is no space for the water to expand into, and consequently it lifts the upper spindle of the valve in the cistern and discharges a certain quantity of water there. Before the discharge, however, the pressure within the pipes must be sufficient to lift the load on the valve spindle, and this being set (usually) to open at 70 lbs. to the square inch, it follows that the pressure the water is under before the valve opens admits of a high temperature being obtained. A pressure of 70 lbs. to the square inch admits of the water reaching a temperature of 316° Fahr.; and this, or a figure near it, is considered as high as the apparatus should be worked at, whether it has a valve or an expansion tube upon it. Occasionally an engineer may say that 350°, requiring a pressure of 120 per square inch, is best, but, unless the idea is to get the rooms warm with the least quantity of pipe, the rule is to work at a lower pressure and temperature.

Supposing, therefore, that the water has reached, say, 316°, with its accompanying pressure of 70 lbs., and the outlet part of the valve is ejecting water, it follows that, sooner or later, the temperature of the water will fall a little, and then, with the accompanying contraction of the water, a vacant space - a vacuum - must occur, or be about to occur, in the apparatus. This is prevented, however, by the immediate opening of the inlet valve ; for directly a state of vacuum is experienced inside the apparatus the inlet spindle is lifted, and water carried in, by the pressure of the atmosphere. Thus, at times, the valve is first ejecting, then drawing in water, quite frequently, as the temperature of the apparatus varies. Almost needless to say, the lower part of the valve at least must always be beneath water, and there is no objection to the whole being submerged.

A pressure gauge is a very desirable adjunct to a high-pressure apparatus of any kind, this being fixed in the stoke-hole, in sight of the stoker, as by indicating pressure it becomes, in a sense, a temperature indicator. A very little practice then enables the stoker to regulate the heat of the water according to the weather, and he can by this means soon obtain a correct warmth in the place to be heated by merely consulting an outdoor thermometer and then working to the figures on the gauge.

Quantities relating to the High-Pressure System of Hot-Water Heating Apparatus.