To George Wynne¹

MY DEAR SIR,

In placing before you the reflections suggested by our joint examination of the boiler² which exploded some time ago³ at the Swindon Station of the Great Western Railway, I have to express my regret that the absence of safe experimental data on the question of steam boiler explosions generally, renders every opinion upon the subject in a great degree conjectural and uncertain. Our two hours’ examination,⁴ instead of enabling us to assign the exact source of the actions which led to the destruction of the boiler, merely suggests the mode of attacking the question. While venturing, therefore, upon the following remarks, I cannot help feeling how much more satisfactory it would be to be able to substitute for supposition, however probable, the surer evidence of experimental fact.

The appearances presented by the boiler were as follows:– The iron plates⁵ as far as the water usually extended, were deeply pitted and furrowed by some corrosive agency. At the centre portion of the boiler, the pits, I think, were deepest and best defined; towards the end at which the cold water entered to feed the boiler, they were less pronounced; towards the end nearest to the fire-box⁶ they were more diffuse and general. While pondering over these curious appearances, my attention was attracted by the tubes which passed through the boiler; these were of brass, and were attached at one end to the fire-box, and at the other end to the smoke-box of the engine. In the normal condition of the boiler, these tubes were separated from the bottom and sides by an interval of an inch or two, this interval being usually filled with hot water.

Without pretending to a degree of confidence unwarranted by the circumstances of the case, I think it may be affirmed that such a boiler constitutes a veritable voltaic couple,⁷ consisting of brass, iron, and the exciting liquid water. Feeble currents of electricity will be established, the direction of which, through the liquid, will be from iron to brass, the iron being what is called the positive element of the couple. A decomposition of the water will take place; hydrogen will be liberated against the brass tubes; oxygen (and acid, if a salt be dissolved in the water,) will be liberated against the iron plates. The quantity thus evolved may be very small, but, acting incessantly for a number of years, it would, I imagine, be sufficient to produce the observed corrosion.

In the course of our examination you directed my attention to the singular preservative influence which the fire-box seemed to exercise on some iron bars which passed through it. There were, I believe, four such bars; solid cylinders, which passed, as longitudinal stays, from the front to the rear of the engine. Up to the point where these bars entered the steam dome they were deeply corroded; in the steam dome they were comparatively unaltered. This seems to be in harmony with the foregoing explanation. Up to the steam dome the iron bars ran side by side with the brass tubes which traversed the boiler; at the steam dome the brass tubes ceased, and from this to the end of the boiler, the iron bars, instead of being electro moters,⁸ were merely the conductors of the currents generated where the bars and the tubes were in close proximity. Other explanations may possibly be suggested, the exact merits of which experiment alone can decide.

Parallel to the row of rivets which united the bottom plate of the boiler to the smoke-box, a deep furrow had been eaten into the plate. A similar corrosive action was not observed parallel to the corresponding row of rivets at the fire-box end. Parallel to one longitudinal row of rivets a furrow had been eaten so deep as to reduce the thickness of the plate from three eighths to less than one eighth of an inch. Along this furrow the boiler is supposed to have given way; the plates were torn asunder, and the total destruction of the boiler was the immediate consequence.

Now, whatever be the cause of the corrosion, whether we refer it to the oxygen held in solution by the water, or to the decomposition of the water itself, science probably furnishes the means for its prevention. The experiments of Pepys, Davy, Van Beck,⁹ and others show that by suitably connecting a more positive metal with iron, the corrosion is diverted from the latter, so that it is thus possible to preserve iron uninjured in a liquid which, under ordinary circumstances, would rapidly attack it.

If the explanation which refers the corrosion to electro chemical decomposition be the correct one, then in boilers where iron tubes are used, instead of brass ones, the pitting and oxidation ought not to take place to the same extent. I say ‘to the same extent’, because a difference in the manipulation of two pieces of iron is sufficient to destroy their electro homogeneity and to produce feeble currents when the pieces are immersed in an exciting liquid. An action of this kind may perhaps be established between the iron rivets and the plates they unite, and, possibly, to this cause may be referred the furrows which are sometimes observed to follow the lines of junction. It would, however, be hasty to infer from a single instance that this is a general result; the whole subject demands a thorough examination, and a far wider acquaintanceship with facts than two hours’ inspection of a single boiler could possibly furnish.

The longer we reflect upon this subject, more deeply must we be impressed with the necessity of associating, in the construction and management of steam boilers, the knowledge of the natural philosopher with that of the engineer. The consciousness of our deficiency in this respect was present with me when, on a recent occasion, I ventured to state, that ‘there are agencies at work in a locomotive of which the maker of it never dreamed, and which may nevertheless convert it into an engine of destruction’.¹⁰ At the present moment it might be difficult to say, how many boilers are on the verge of explosion without any single engineer being aware of the danger. Another point deserving of attention is the following:– Common water always contains a quantity of atmospheric air in solution. This air may be expelled from the liquid by continued boiling; and it is an experimental fact, that water thus freed from air possesses mechanical properties widely different from ordinary water. The cohesion of the liquid is enormously increased, in consequence of which it may be heated to a temperature far beyond its ordinary boiling point without boiling; but when it does boil, it is not with the quiet ebullition of common water, the liquid particles snap suddenly asunder like a broken spring, and ebullition is converted into explosion. It is, I believe, a fact that boilers, after standing for a time, have often exploded at the precise moment when the engineer turned on the steam. The question presents itself, whether the action just referred to may not here come into play? If the water has been sufficiently purged of air and then left at rest, its augmented cohesion may permit of its being heated far above the boiling point due to the pressure upon its surface; the mechanical disturbance produced by turning on the steam would destroy this cohesion, and the superheated liquid would develop a force resembling that of gunpowder on the application of a lighted match. I would wish it to be borne in mind, that we are in total ignorance as to the practical value of these suggestions; but the investigation which the subject so pressingly demands must take cognizance of them all.

I am, &c.

To Lieut.-Col. Wynne, R.E. (Signed) JOHN TYNDALL.

House of Commons Parliamentary Papers, 1854–5 [1951], Two Reports of an Inspecting-Officer of the Railway Department to the Lords of the Committee of Privy Council for Trade, on the Explosion of Boilers of Locomotive Engines, One of which Occurred at the Gloucester Station of the Great Western Railway, On the 7th February 1855, and the Other at the Greenock Station of the Caledonian Railway, on the 5th April 1855, Together with a Copy of a Circular Issued by their Lordships to the Railway Companies on the Subject (London: HMSO, 1855), 63–4.