erally fitted in Her Majesty's navy, and have nothing novel about their construction.

The steam-gauges are on the ordinary Bourdon principle, graduated to 35 pounds per square inch, and are fixed in conspicuous places on the fronts of the boilers.

It appears the vessel was tried under steam several times at Pembroke, afterward was steamed around to Portsmouth in charge of the Admiralty officials, and has subsequently been under steam on several occasions.

On the 14th July it was arranged to make the official trial. The vessel lay at Spithead, fires were lighted at 10 o'clock, steam was slowly raised in all the boilers, and about 1 o'clock the engines were started for a preparatory run before going on the measured-mile full-power trial.

In about eight or ten minutes afterward the foremost starboard boiler in the after stoke-hole exploded. The vessel was towed into Portsmouth Harbor and the stoke-holes were strictly guarded until the official inspection took place. I was present at the first inspection and have been present at all subsequent examinations.

On examining the exploded boiler, it was found that almost the whole of the front plates above the smoke-box doors had been forced away, and lay, together with the smoke-box doors and other débris, in a distorted mass on the stoke-hole floors. The plates themselves were torn through the first seam of rivets above the smoke-boxes for almost the whole length of the boiler, rending in a diagonal direction through the solid plate at the man-hole, and again tearing through the seams of rivets connecting them with the crown and other parts of the shell. The front of the uptake was bulged, torn, and driven back against the back plates. The plates were drawn over the stays supporting these flat surfaces. The T-iron fastenings to the stays above the uptake-plates were sheared through the pinholes. The top of the boiler immediately over the uptake was bulged upward. The sides of the smoke-boxes below the ruptured parts, together with the flat surfaces in the combustionchambers, and two of the furnace-crowns were bulged between the stays. Altogether the boiler showed decided symptoms of having been subjected to excessive pressure.

There were no signs of shortness of water.

The workmanship appeared to be of the best description, and from the roof-mark on the front of the boiler, it had been tested in November, 1570, by hydraulic pressure to 60 pounds per square inch.

Both stop-valves on this boiler were found to be shut, and the steamgauge was entirely destroyed.

The safety-valve chest lay on the stoke-hole plates in exactly the posi tion it had fallen. It contained two valves, each 53 inches in diameter, one of which could be lifted by means of a screw beneath it, while the other is inclosed and entirely out of the control of the engineer. The cap over the spindle of the hand-lifting valve was carried away, and the top of the spindle broken off.

The stop-valves being closed, it was evident that none of the steam generated in this boiler, from the time the fires were lighted to the time of the explosion, could have gone to the engines. It must therefore have been accumulating pressure in the boiler all that time, or escaping by the safety-valves. The safety-valves were stated to have been loaded to 30 pounds per square inch, and the pressure-gauge was said to have been observed out of order more than an hour before the explosion occurred.

It was considered very important to make a minute examination of the remains of the safety-valves, to see if there was any evidence of their being jammed or otherwise not in working order. The chest was opened in the presence of all the inspectors, and there were no such things as wedges, stops, or other means for jamming the valves found. Both valve-spindles were found broken and bent, the valves were perfectly free in their seats, but the feathers of these valves were battered and distorted in such a manner as to destroy all evidence of their condition at the time of the explosion.

If the safety valves were not jammed or set fast, the boiler must have exploded owing to structural weakness; and in order to ascertain as far as possible whether it could have arisen from this cause, the whole of the other boilers, with the exception of one small boiler in the forward stoke-hole, were tested by hydraulic pressure to from 60 pounds to 65 pounds per square inch. They were quite tight, and after being carefully gauged before, during, and after the test, they showed no signs of weakness.

Strips of plate were cut from the ruptured parts, and tested by the testing-machine in the dock-yard. The iron was found to be above the average strength of iron used in boiler-making, the mean results of these experiments showing that it would stand a tensile stress of 21 tons per square inch of section.

As a guide to the pressure which this boiler was capable of bearing, a chamber was constructed to represent to some extent the part of the boiler which gave way. From its form it did not, in my opinion, give exact information on that point, seeing that the side plates of the uptake, which very materially add to the strength, were omitted, and consequently the whole strength of the structure was reduced. It was burst at various pressures under four or five different conditions, and although it did not exactly represent the exploded part of the boiler, the results of these experiments show almost with certainty that the pressure within the boiler at the moment of rupture could not have been less than 100 pounds per square inch.

The whole question of the cause of this explosion therefore centers itself upon the condition of the safety-valves at the time of the explosion, and as they were so destroyed by the accident as to make it impossible from their appearance to speak of their condition at that time, it was deemed advisable to make an examination of all the safety-valves of the other boilers. This examination was made, and they were all found to be loaded to about 30 pounds per square inch. They were free to move, and to all appearance were in a workable condition.

The feathers of safety-valves made on this principle require to be an easy fit in their seats; if too slack, the rolling of the vessel shifts them on their faces and causes them to leak. In this case they appeared to be a good working fit, and it was suggested to try them under steam. One of the valve-chests containing two valves was accordingly attached

a land-boiler in the dock-yard. It was found that when steam was raised, one of the valves, although loaded to only 30 pounds per square inch and quite free when cold, did not lift until the pressure reached 52 pounds per square inch, but after blowing a short time and uniformly heating the cast-iron chest, it rose and fell at about the working-pressure.

This conclusively proved that the expansion of the feathers of the brass valve was sufficient to make them grip or set fast, when fitted so close as this valve was fitted.

Although this single valve did stick under this trial, the whole of the

others were tried afterward under similar conditions and were per fectly free; the experiment therefore affords no certain evidence of the exact condition of the safety-valves on the exploded boiler at the time of the explosion. But as there can be little doubt of their inoperativeness, it can be inferred that had they been a little tighter than the valve that was tried and found to stick, or if they had been neglected and scale or dirt allowed to get between the working surfaces, the expansion of the valve would have caused them to jam or set fast, and allow the pressure to accumulate until it overcame the strength of the boiler. The facts of the case may be summarized as follows:

1. The stop-valves were shut, so that the steam generated in the boiler could not escape to the engines.

2. The steam-gauge was discarded as being out of order, and therefore it did not make known any abnormal pressure in the boiler.

3. The fires were burning briskly, and therefore the pressure must have been rapidly increasing in the boiler.

4. The boiler presents unmistakable symptoms that it burst from excessive pressure, and not from faulty construction.

Under these circumstances, I am of opinion that the safety-valves of the boiler in question at the time of the explosion were set fast in their seats, and that the cause of the explosion was excessive pressure, owing to the stop-valves being shut and there being no outlet for the steam which the fires were generating.

Having, therefore, arrived at this conclusion, I think it my duty to urge the necessity of the greatest care being taken (1) in the designing of safety-valves, and (2) in the absolute necessity of their being fre quently examined. There ought to be a certain means of preventing any increase of pressure beyond the stipulated amount.

Since this explosion I have carefully considered this subject, and, with a view to reducing the chances of safety-valves sticking or otherwise being inoperative, I would suggest :

1. That the valve with a central guiding-spindle is less liable to set fast by unequal expansion than a feather-guided valve.

2. That the spindles which support the weights of these valves should be detached from the valves themselves, or that a guide should be fitted between the valve and the weights, so that any inclination of the weights caused by the motion of the vessel should not tend to cant the valves and make them grip in their seats.

3. Easing-gear should be fitted to all the valves, and arranged to lift the valves themselves.

4. That all safety-valve spindles should extend through the covers, and be fitted with sockets and cross-handles, so arranged that it would be impossible to add any extra weight or to control their action, but at the same time allowing them to be lifted and turned round on their seats, and their efficiency tested by the engineer at any time, whether the steam is up or not.

I would, in conclusion, add that loading safety-valves for marine purposes by means of dead-weights is now in the merchant service almost entirely superseded by springs.

WILLIAM PARKER,

Chief Engineer-Surveyor to Lloyd's Register of Shipping. (Lloyd's Register of British and Foreign Shipping, 2 White Lion Court, Cornhill, London, August 16, 1876.)

Of the four several official reports on this explosion the above is given as the most practical.

BOILERS OF THE MERCANTILE MARINE.

Through the kind attention of the general inspecting engineer of Lloyd's, I had the privilege of observing the practice and inspection as applied to the boilers and machinery of the commercial marine of Great Britain. Steamers were visited after long voyages; some vessels in which the machinery and boilers were comparatively new, others having boilers in different stages of deterioration, and still others in which the machinery was undergoing construction. All that relates to the corrosion of boilers will be found under that head at page 199 et sequitur. As to the kind of boilers now employed, no material changes have taken place in the last five years, but the old type of box or wagon-shaped boiler, so familiar to the marine engineer of former days, is no longe seen. It is obsolete, reckoned among the inventions of the past, and years hence will probably be shown in the South Kensington Patent Museum by the side of the marine side-lever engine and numerous other obsolete curiosities stored in that interesting institution.

The pressure of steam carried at the present day on marine boilers varies from 60 to 70 pounds per square inch, and the boilers used to generate it have shells either cylindrical or elliptical. More care is exercised than was formerly considered necessary in the details of construc tion, and better workmanship in fitting and riveting the parts together. Composition tubes are being used to a considerable extent where iron tubes were formerly employed. Safety-valves, having levers with loaded weights attached, have, in consequence of the higher pressures used than formerly, and of the difficulty with them, arising from the pitching and rolling of the vessel in rough weather, been abandoned, and spring safety valves substituted in all steamers. The valve adopted generally for boilers, both of commercial and naval vessels, is known as Adams's. The following cut is a representation of it.

In this connection, it may be interesting to quote from Lloyd's Regis ter a section relating to the designing, building, testing, and fittings of boilers; also their formula for the strength of cylindrical shells of boilers and of circular flues or furnaces.

Lloyd's Register of British and foreign shipping.

[Extract from the rules of the society.]

MACHINERY AND BOILERS OF STEAMSHIPS.

SECTION 73. With respect to the boilers and machinery, the owners are required to submit them to the inspection of the society's engineer-surveyors, who will furnish a report to the committee describing their state and condition, in the manner and form (No. 8) annexed. The committee will thereupon grant a certificate, and insert in the registry-book the notification "Lloyd's MC." (in red), indicating that the boilers and machinery have been inspected by the engineer-surveyors, and certified to be in good order and safe working condition. (See section 81.)

The society's engineer-surveyors are to examine the plans of the boilers, and approve of the strengths for the intended working-pressure.

Any great novelty in the construction of the machinery or boilers to be reported to the committee.

The boilers, together with the machinery, to be inspected at different stages of construction.

The boilers to be tested by hydraulic pressure in the presence of the surveyor to twice the intended working-pressure.

Two safety-valves to be fitted to each boiler; if common valves are used, their com bined areas to be half a square inch to each square foot of fire-grate surface; if inproved valves are used, their efficiency to be tested under steam in the presence of the surveyor, and in all cases set to the working-pressure.

A stop-valve to be fitted, so that each boiler can be worked separately.