stature; although he was scalded, his face was above the level of the steam, being between the deck-beams where there was an air cushion. That the boiler exploded was due to the remarkable coincidence of two factors. The box safety-valve jammed, owing to the two different metals of which it was constructed expanding in different degrees. And the pressure-gauge tell-tale, which was fitted in a cogged circle, had the needle forced right round the circle twice or more, so that it showed a normal pressure. The actual pressure must have been terrific. And after I left the ship one of her guns burst. This accident contributed another instance in favour of breech-loading as opposed to muzzle-loading guns. The accident occurred during practice at quarters in the Gulf of Ismid, on 2nd January, 1879, in the fore-turret. Captain Alfred John Chatfield had succeeded Captain Wilson in command. Two officers and nine men were killed, and thirty-five persons injured. The muzzle was blown off from about two feet in front of the trunnions. There was much discussion then and subsequently concerning the cause of the accident. The probability is that the bursting of the gun was due to its having been double-loaded, after a previous miss-fire, which, in the simultaneous discharge of the rest of the guns, had not been noticed. The committee {149} which reported on the matter on 1st March, 1879, adopted this hypothesis, in preference to the theory that there had been a flaw in the material. Captain Edward Seymour (afterwards Admiral of the Fleet the Right Hon. Sir E. H. Seymour), who was then in command of the troopship _Orontes_, in his book, _My Naval Career and Travels_, thus refers to the incident:-- "From Malta I brought home the main part of the ship's company of H.M.S. _Thunderer_, on board which ship the terrible explosion of the 38-ton muzzle-loading gun had lately occurred in her foremost turret. Both turret guns were being fired simultaneously, but evidently one did not go off. It may seem hard to believe such a thing could happen and not be noticed, but from my own experience I understand it. The men in the turret often stopped their ears, and perhaps shut their eyes, at the moment of firing, and then instantly worked the run-in levers, and did not notice how much the guns had recoiled. This no doubt occurred. Both guns were then at once reloaded, and the rammer's indicator, working by machinery, set fast and failed to show how far home the new charge had gone. This, too, may seem unlikely, but no doubt it happened; and the gun on being then fired burst, killing two officers and several men, and wrecking the turret. Experiments made with a similar gun double-loaded, burst it in exactly the same way."