Rowan quotes the cases mentioned by Miller, in one of which a set of boilers worked for 4 years with sea-water, using the injection condenser, exhibited no signs of damage either from incrustation or corrosion, sufficient time being allowed for "scaling" at the end of each voyage; but when the injection condenser was replaced by a surface condenser, and fresh water, supplemented by occasional supplies from the sea, was fed to the boilers, they speedily lost the protecting crust, and became corroded with pits and blotches in the well-known form. The explanation of the action is found by some in the belief that by repeated distillation the water becomes altered - that is, it is split up, the oxygen attacking the metal of the plates, while the hydrogen escapes. Rowan, however, points out that the water, in the case in question, never was genuine distilled water. It was fresh water to start with, supplemented by supplies of sea water, so that it always had a considerable amount of carbonic acid and oxygen in solution; and besides, when put into the boiler, it commenced to dissolve the chloride of sodium in the salt scale formed by the sea water.

In short, Rowan's contention is, that the scale was dissolve4 off, and that the " distilled " water containing air accordingly was free to attack the iron. But when sea water is introduced into boilers as a supplementary feed, the plates, being clean, are liable to be damaged by the action of hydrochloric acid set free by the decomposition of chloride of magnesium. It is also not improbable that carbonic acid is set free by the mutual decomposition of sulphate of magnesia and carbonate of lime; hence the action of these acids, combined with high temperature and pressure, is sufficient to account for most of the corrosion, in Rowan's opinion.

As to the suggested remedies. Zinc has been shown, especially on the Continent, to be an excellent anti-corrosive in those cases where decomposed grease, or fatty acid, is the destroying agent; but its usefulness is confined to land boilers, or to those marine boilers in which fresh or distilled water is alone used; for zinc is attacked by sea water, and chloride of zinc is formed, and would, if the assumptions above made are correct, merely add to the impurities and evils to be avoided. Filtering the feed is an excellent precaution, and should be universally adopted; but to prevent the corrosive action in marine and other boilers of substances which no filtering can arrest, Rowan thinks nothing better can be suggested than forming on the interior surfaces an artificial coating of calcium sulphate and magnesium hydrate, in proportions varying with the pressure carried in the boiler. The mixture can be easily fed in in the form of a thin whitewash with fresh water; but to obtain the best results it should be supplied at the commencement of the boiler's career, before corrosion has unfitted the surfaces of the plates for the adherence of the protective coat.

It is claimed that when once hardened by heat, the artificial scale thus made with fresh water cannot be dissolved by fresh water, and is not likely to be affected by the small quantity of sea water which may leak in; that its thickness is quite under control; and that it is safe and free from the trouble attending the keeping up of a salt scale.

A writer in Le Technologists discusses the processes employed to prevent boiler incrustations under 3 classes: - (1) chemical, (2) chemical and mechanical combined, (3) physical.

Chemical Processes

These consist in the use of certain solvent substances introduced into the boiler to precipitate salts contained in the water. The nonadherent, muddy deposits thus formed from the calcareous matter are from time to time removed, that they may not by their presence be an obstacle to the action of heat. Colouring matters, dyewoods, and in general all woods containing tannin, can be used for the purpose, when the waters contain neither sulphates nor chlorides. Various other products having for base fecula, lime, and baryta, are also employed with success; but the constitution of the substances used should be suited to the nature of the water. The chief inconvenience in using these products is that most of them corrode the boilerplates, and produce a froth in the water with which they are mixed, containing precipitated fragments, which, in consequence of their small size, are readily carried by the steam into the valves and cylinders of the engine, where they may injure the joints through friction, and cause an escape of the steam.

Chemico-Mechanical Processes

In the processes just referred to, it is of course necessary to open the boiler in order to remove the deposits, and this may sometimes require a stoppage of work for an entire day. This disadvantage would be obviated if, instead of putting the anti-calcareous matters into the boilers, they were put in the feed water, and this water filtered (after heating) before being introduced into the boiler. The Chemin de Fer du Nord Co. have used apparatus of this kind for several years, the precipitating matters being mixed, by mechanical agitation, with the feed water in large reservoirs. When the mixture is sufficiently complete, the muddy water is decanted during 10 or 12 hours, and filtered previous to use.

Physical Processes

By heating the water to a high temperature, it is possible to purify it from all the sedimentary matters contained in it. These matters are decomposed and precipitated, and cannot adhere to the walls of the vessel containing them till the temperature of the water is lowered. The temperatures at which the calcareous matters are precipitated are the following: -

Carbonates of lime, between 176° and 248° F. (80° to 120° C).

Sulphates of lime, between 284° and 302° F. (140° to 150° C).

Chlorides of magnesium, between 212° and 257° F. (100° to 125° C).