For the reasons just given, it is always desirable to heat the autoclave, whenever possible, in a suitable bath. Such a bath may contain either oil or solder. Even when no crusts are formed, which would interfere with the heat transference, phenomena occur at higher temperatures, which render the use of a solder-bath very desirable. The solder which is poured in to fix the liner in position always liquefies if the autoclave is heated directly, and the liner rises up until finally it touches the cover, and so, after cooling, renders it difficult to screw the cover down tightly again when closing the vessel. In this way strains are developed in course of time which make the autoclave leaky, and, in addition, the strain on the bolts is a serious source of danger. With incorrect heating not only does the autoclave itself suffer, but also the substances which are being heated; such cases have been discussed in detail in connection with a-naphthol and β-naphthylamine.

1 The β-naphthol melt is an example of this type, and any attempt to carry out this melt without the use of a metal bath will with certainty ruin any autoclave.

The autoclave is charged either through the man-hole, this being then carefully closed, or, if possible, the substance is sucked in by evacuating the vessel so that as little opportunity as possible may be given for the development of leaks. Water expands very considerably upon heating (according to Mendeleef by 20 % of its volume on heating from 0-2500), for which reason no autoclave should be filled up to more than 80 % of its total volume. If the vessel is completely filled such an enormous pressure will be developed that it will be burst open. It is therefore necessary to have a notice above every autoclave, showing clearly the total volume, the maximum pressure, the maximum filling, and the method of charging.

The calculations for an autoclave are matters for engineers who have official standards upon which to base their estimates. It is, however, also advisable to have every apparatus recalculated in a first-class engineering works, after which the official sanction for the scheme may be applied for.

By means of a travelling crane the autoclave is placed in its masonry setting, which has already been erected, and which should be held down by means of iron rods placed about 30 cms. apart, the projecting ends being screwed down to the brickwork by means of iron plates 25 by 25 cms. square. The autoclave or its bath is placed in a counter-sunk ring, as shown at the bottom of Fig. 32, and the apparatus, after being charged, is heated up by means of good coal. If the setting has been correctly done, it is unnecessary to carry out the first heating with excessive caution, although it is always advisable to start with a small flame. The fire-bars must be kept clear, and, if necessary, one or more should be removed if the draught is insufficient. It is also advisable to have a separate chimney for each pair of large autoclaves, so as to be independent of neighbouring plant. When working on the large scale the heating-up always occupies several hours, but once the brickwork is hot a very little wood or coal will suffice to keep the temperature up. For temperatures above 2000 the heat losses by radiation are so considerable that the portion of the apparatus projecting from above must be insulated by means of a tin cover lined with asbestos. To cool, the cover, which is made in several pieces, is removed, and at the same time the furnace door and the dampers are opened. By blowing off a portion of the contents into the air, or in the cases of alcohol and ammonia, into the condenser, the cooling of the autoclave may be greatly accelerated without the brickwork losing too much heat, which is of importance for the next operation. Whilst a melt is being carried out the autoclave must be carefully watched. The temperature of the oil-bath should be about 30o higher than the internal temperature, and the two thermometers and manometers should agree within a couple of points. If greater deviations are shown, the thermometers must be checked, and in some cases the process must be interrupted.

By the use of pyrometers it is possible to superintend the running of a process from the laboratory, and self-registering manometers are coming into use which permit of the subsequent examination of the pressures for purposes of control. A book should be kept in which all happenings are entered, so that documentary proof is always available in case of a break-down or an accident. If, in spite of all precautions, any serious mishap should occur,. such as an unexpected rise of the temperature or pressure, the fire must be raked out at once, all the dampers and flues should be opened, and all the personnel should be evacuated from the shed in which the autoclave is situated and from the surrounding buildings. The explosion of an autoclave, like one of those shown in Figs. 31 and 32, may lay an entire factory in ruins. Since, however, all autoclaves are made with an eight-fold factor of safety, there is really no danger with proper attention. Every year each autoclave is examined by a boiler inspector, being cleaned out and well cooled for the purpose. The inside of such a vessel must not be entered until it has been shown that a candle can burn quietly therein. Usually only the man-hole is open, and compressed air is blown in. Never less than two workmen should be engaged upon the job. The result of the examination should be made the subject of an official report. Frequently the lining is removed in order that any alterations in the wall may be accurately measured; the liner is removed for the purposes of the examination by lighting a fire inside, which is kept going by means of compressed air. As soon as the lead is molten the liner rises up somewhat and is then removed by means of the crane. The lead is removed by means of iron ladles and is cast into bars.