boiler had been unduly heated. The case of the explosion of a cylindrical boiler, at the foundry of Mr. M'Queen, in New York, may have been an exception to this remark, but the fact of the water having been very low, though rendered probable, is not entirely made out. The copper boiler exploded in the experiments of the Committee, may be considered, however, as illustrating the possibility of such an occurrence.t The explosions in boilers with interior flues, arising from these circumstances, have been found to affect those flues which "collapse," that is, are crushed, and rent, and are frequently separated from the boiler. The case of the accident on board the steam-boat Patriot, which occurred near the mouth of the river Ohio, in 1828, distinctly illustrates this point. It is stated that the engine was kept in motion after it was known that the water was below the flues; an act of temerity which was followed by the collapse of one of the flues. 30. It is highly probable, as will be seen hereafter, that water is thrown upon the hot metal, in cases where the engine has just been set in motion, or when the safety-valve has been opened, just prior to an explosion, and hence we are not warranted in adducing such cases here. 31. The following is a well authenticated instance in which the pressure was not sufficient to produce explosion, or in which the metal was cooled by its change of figure, but when the danger was obviously imminent. It was that of the boiler of the steam-boat Legislator, of which an oval flue was flattened by exposure to the steam-pressure, when known to be unduly heated. An explosion was avoided by the fire being put out, as soon as the deficiency of water was ascer tained. 32. The Committee next proceed to consider the means which have been pointed out, by which, in the event of an undue heating of a boiler, water may obtain access to the heated metal. Various means by which water may be Thomas Ewbank "On the Explosion of SteamBoilers," Jour. Frank. Inst. vol. x. p. 3. That the pressure was unusually great, is satisfactorily shown; and with or without undue heating, was, no doubt, the cause of the explosion. In the case of the Etna, the boiler was unduly heated, but, whether it gave way to the pressure, ascertained from the working of the engine, to be lower than usual, or whether water had access to the hot metal, is not knowu, Report of the Committee on Explosions, Part I. p. 68. Jour. Frank. Inst. vol. xvii. p. 225. 1 Replies to Circular, &c. No. XX1. Thomas J. Haldermann, Esq. The explosion on board the Tricolour might be cited as another example. The boat was at rest, No. XXI. Replies, &c. Replies to Circular, No. XVII. E. A. Lester, Esq., of Boston. brought into contact with heated metal, have been suggested. This may occur by the intentional or accidental removal of an obstruction, or by some other repair, to a forcing-pump, or by the injection of water by a hand-pump. When it is recollected that one cubic inch of water will produce six hundred and twenty-one cubic inches of steam of three atmospheres, or one hundred and eighty-nine inches of steam of eleven atmospheres, and that this steam is produced far more rapidly than in the usual action of the boiler, we are at no loss to understand that an ordinary safety-valve cannot give vent to it. The explosions on board of the steamboats Grampus, Constitution, &c., should serve as warnings against the introduction of water into a boiler under such circumstances, and the course taken by the engineer of the Legislator as an example to be followed. In this latter case there can be no doubt, from the circumstances stated by Mr. Lester,** that an explosion was prevented by cooling the boilers previous to the introduction of water, when the deficiency was discovered. No circumstance of mere convenience should be suffered to interfere with such a course. 33. A second means assigned for bringing water in contact with unduly heated metal in a boiler, is by the foaming produced by the opening of a safety-valve, or in the ordinary working of the engine. This foaming has been abundantly demonstrated, and a detailed statement of the direct experiments made on the subject by this Committee, may be seen in their Report to the Secretary of the Treasury of the United States. Since the foaming is caused by making an opening in the boiler, it may be supposed that it cannot be adequate to do more than to produce steam to supply the place of that which escapes. This view of the subject derives some support from the experiments incidentally made by M. Arago,++ and directly with this object by the Com In the case of the boilers of the Car of Commerce (No. VII. Replies, &c.) of the Grampus (No. XII. Replies, &c.) &c. See explosion of a boiler at Aston Forge (No. XI. of Replies, &c.) Boiler of the steam-boat Constitution (No. VIII. of Replies, &c.) &c. **See Replies to Circular of Com. on Explosions, No. XVII. Letter of E. A. Lester, Esq. to Sec. of Treasury. + Reply to Query I. "To ascertain, by direct experiment, whether on relieving water heated to, or above, the boiling point, from pressure, any commotion is produced in the finid." See also a paper by Mr. F. Peale, whose observations were contemporaneous with experiments of the Committee, Jour. Frank. Inst. vol. viii. p. 145, and Replies, &c. No, XXI. Potts on Explosions, Jour. Frank. Inst, vol. vi. P 327. ++ M. Arago. Sur les explosions, &c. Annuaire du Bureau des Long. 1830, pp. 148 and 180, and Jour. Frank. Ins, vol, v. p. 404, and vel. vi. p. 47. mittee, but is contradicted by those of MM.Tabareau and Rey.* In the experiments of M. Arago, the boilers were not unduly heated; in those of the Committee, there was present a considerable amount of heated metal, and in those last referred to, and in which an increase in the elasticity of the steam was produced by opening a safetyvalve upon a small boiler, the boiler was surrounded by a charcoal fire. There can be little doubt, then, that the result must depend upon the precise circumstances of the case, and that danger may result, though it does necessarily follow, from making an opening in the boiler when the water is low. This effect from foaming would be increased, if in addition to the agitation produced by the first working of the engine, after stopping, the safety-valve should be opened. This was the common practice on the Hudson a few years since; the safety-valve being opened, by hand, on putting the boat in motion after the landing or taking up of passengers. 34. The successive explosions of connected boilers, such as occurred at the Polgooth+ mines, and on board the steam-boat Rhone,+ are easily explained if referred to the effect of foaming, and difficult to understand on any other principle, since just before the explosion of the second boiler a large opening was made for the escape of steam.§ 35. It has been assumed by our countryman Perkins, in his hypothesis on the subject of the explosion of steam-boilers, that the hot steam formed by contact with unduly heated metal is the true source of danger. This opinion has been shown to be inconsist ent with the deductions from sound theory.|| The injection of water into hot and unsaturated steam, should reduce, not increase, its elasticity. With a view to ascertain if any circumstances had been omitted in the application of theory to this problem, the Committee made direct experiments on the subject. The water was introduced both in a full stream, and through small apertures. In no case, an increase, and in all but one a perceptible decrease, of elasticity in the hot and * Ibid. + J. Taylor, Esq. "On the Accidents incident to Steam-Boilers." Lond. Philos. Mag. vol. i. 1827. Annuaire du Bureau des Long. 1830, and Jour. Frank, vol. v. p. 401. The same opinion is expressed by M. Arago, Annuaire, &c. p. 184. Translated in Jour. Frank. Inst. vol. vi. p. 49. The other cases referred to by him, in which an explosion followed the opening of a stop-cock by hand, as at Lyons, or of a safetyvalve by the steam, as at Essone, may be explained by supposing the openings insufficient to give vent to the steam, which was produced by the action of the boiler, in the circumstances then existing. Dulong. Annales de Chim, et de Phys, vol. xlviii. unsaturated steam, was observed. Fourteen ounces of water injected into steam at 533° reduced its pressure 34 of an atmosphere. ¶ The steam had in this experiment a temperature corresponding to the pressure of 60 atmospheres,** and an actual elasticity of only 6.82 atmospheres. There was besides a fire which supplied heat, as it was absorbed by the vaporisation of the injected water. 36. A correspondent has suggested++ that when a steam-boat is first set in motion, the inertia of the water may cause it to rise at one end of the boiler and then to oscillate, by which it would be thrown upon parts of the boiler which might be unduly heated. This is no doubt a true cause, but it would be difficult to say to what extent it would be effective. The subsequent suggestion that water can take a charge of heat in a latent state, which may be rendered free by mechanical means, the Committee do not conceive to be valid. Experiments which have been referred to as showing this, are fully explicable upon well-established principles. 37. There are two other important circumstances to be examined falling under this division of the subject, namely, the effect of the careening of a boat, especially one having connected boilers, and the effects of the sudden cracking of deposits of mud or sediment, beneath which the metal is unduly heated. These will, however, be treated under separate heads. 38. The Committee now proceed to examine the means proposed for preventing the occurrence of the dangerous circumstances now under discussion. These, of course, have reference, mainly, to the original source of the danger, that is, to the deficiency of water within the boiler, though an avoidance of the secondary causes might prove effectual. 39. First. Various self-regulating apparatus for the supply of boilers have been proposed and partially used. Second.-Methods for ascertaining the level of the water, or of giving notice when it falls to a certain level, are in use, or have been suggested. Third. Some methods for ascertaining the temperature of the boiler, or of particular parts of it, have been contrived. 40. 1st. One of the most common methods of regulating the supply of water to a boiler In a certain theoretical case, namely, that in which all the heat to vaporise the injected water is derived from the hot steam, and the quantity of water which that steam can vaporise, without reduction of temperature below that of saturated vapour of the same elasticity, is injected, the precise reduction of elasticity has been calculated by M. Dulong. See Ann. de Chim. et de Phys. vol. xlviii. p. 378. ** Calculated from the formula deduced by MM. Arago and Dulong from their experiments. tt Replies to Circular of Com. on Explosions, No. XX. is by the use of the float. This is understood to have been entirely successful in the low-pressure boiler, the float being applied to raise a valve connecting a reservoir of water with the boiler to be supplied. A selffeeding apparatus in which a float was used was proposed by Mr. Charles Potts, who exhibited to the Franklin Institute a very neat working model, in which a glass boiler was kept at nearly a constant level by this method. Its application is most difficult in the case of a small high-pressure boiler with interior flues. Engineers differ very much in the amount of confidence which is to be placed in the float: those who have seen it in operation in the large boiler of a low-pressure engine give it implicit confidence; others who have tried it in the small high-pressure boiler, consider its action too uncertain to answer a good purpose, even when in its best form. If the objections to the float are not valid, and we apprehend that they are only partially so, the real difficulty will be found to lie in general objections to all self-regulating apparatus. This obviously is one which is liable to get out of order, since it communicates between the exterior and interior of a boiler, and hence must have a packed joint, liable when the stem is not in constant motion to become tight, and therefore beyond the power of the change of buoyancy in the float to move. 41. A most ingenious method of feeding boilers was patented in 1825, by Mr. Eve.‡ It consisted of a revolving-cock, bored in part through, and playing alternately into the boiler, and into a box of water. It was expected that this cock being placed at a proper level of the water within a boiler, would merely draw out and return water while this was at a due height, but when it sunk too low would draw out steam and return water. The difficulty of condensing the high steam§ drawn out, and of making the returned water flow out of the openings, seems to have rendered this, as well as other promising schemes of the same sort, abortive.|| *Journal of the Franklin Institute, vol. vi. p. 42, and also p. 327, &c. where the apparatus is illustrated by a figure. †D. J. Burr on the explosion of steam-boilers. Jour. Frank. Inst. vol. vi. p. 335. Mr. Redfield objects to its use in steam-boat boilers. See Report to Secy. Tres. U. S. in Doc. H. R., No. 478, session 1831-2.. Lond. Jour. of Arts, vol. xii. p. 230, Lond. Mech. Mag. vol. vii. p. 344, Rep. Pat. Invent. vol. iii. p. 70. A revolving-wheel for the same purpose has been patented by Mr. Jesse Fox. Jour. Frank. Inst. vol. x. p. 161. See also J. S. Williams' patent for supplying 2 boilers with water. Jour, Frank. Inst. vol. vii. p. 183, which though different in action is liable to this objection, in even greater force. Walker's feeding-plug. Trans. Soc. Arts, &c., vol. L. Part I. p. 63. Sliding-valve and box. Lond. Mech. Mag. vol. xxi, p. 376. An attempt to obviate these objections, which was seen by some of the members of this Committee, was unsuccessful. Mr. Charles Potts has recently proposed a plan which is similar in principle. It will have to encounter the difficulty of the flow of water from moderately large openings when the pressure on the two ends of the column is the same, and the necessity for the alternate heating and cooling of the revolving-plug or chamber which acts as a feeder, and of at least a part of its contents. They agree entirely, however, with the Committee on Science and the Arts, that this principle merits further trial.** 42. The Committee are decidedly of opinion that no self-feeding apparatus has been, or is likely to be, invented which can be a substitute for the care of an engineer; and, indeed, they consider the carelessness which is liable to result from their use as a very serious, though not an insuperable, objection to them. 43. 2nd. Methods for ascertaining the level of the water in a boiler, or of giving notice when it falls to a certain level. The imperfection of the gauge-cocks in ordinary use has been often pointed out, and indeed is generally admitted. Originating in the very infancy of the art in Savery's engine, they remain at this day a stain upon its more mature age, At best, when the water is tranquil within a boiler, they only show, roughly, the position of the water-line; and when it is above the highest cock, or below the lowest, they fail entirely; and cannot be placed far apart without making their indications, within these limits, too rude even for practice. When a boiler is in action, particularly if it is small and contains high-pressure steam, the foaming is so considerable as to interfere with their use. In the report of experiments by this Committee, abundant evidence is to be found of this imperfection; as an example of which may be taken the case, where by raising the safetyvalve of the small experimental boiler, indications of water appeared at a gauge-cock, below which the hydrostatic level was known to be nearly two inches.‡‡ -44. The method of indicating the level of Journal of the Franklin Institute, May, 1836, vol. xvii. p. 302, **Report on a "Plan of a new pump for feeding steam-boilers." Jour. Frank. Inst. vol. xviii. p. 3. 1836. tt From the remarks which follow, exclusive of the objection on the score of the effect of foaming, we must except the shifting gauge-cock of Mr. Philos. Tyler, described in Jour. Frank, Inst. vol. XV. p. 178. Report of Com. on Explosions, Part I. pp. 11, 12, &c. Jour. Frank. Inst. vol. xvii. pp. 9, 10. Peale on the height of water in boilers of locomotives. Jour. Frank. Inst. vol. viii. and Replies No. XXI. Potts on Explosions. Jour. Frank. Inst. vol. vi. p. 329. the water by a float is liable to all the objections urged against the feeding apparatus, depending for its action upon [that instrument. It has not, however, except in very rare cases, been used in this country. An alarm float was tried by the Committee, which is not subject to the objection in regard to the stuffing-box, since it is entirely within the boiler. This is by no means a new device, though the particular arrangement was made by Mr. D. H. Mason for the Committee, and is figured and described in the first part of their report. This device is intended to allow the escape of a small jet of steam whenever the water rises above, or falls below, a determinate level.§ The alarm by the issue of steam through a trumpet tube, being only applicable to engines working at very low pressures, does not require special notice here. 45. With due care on the part of the engineer, and the Committee are of opinion that no substitute has yet been found for such eare, the glass tube affords the best means known to us, for observing the level of the water within a boiler. It seems strange that this excellent device which has stood the test both of experiment and of practice, has met with so limited a degree of favour. In the great progress made of late years, in the locomotive engine, it has been so clearly shown that engineers and their assistants can be induced to employ any machinery, the use of which is insisted upon, that the excuse of their indisposition to change should not be urged any longer. In this very case, in which the glass tube is probably more exposed to fracture than any other, it is practically used. The objection on the score of its breaking by unequal expansion and contraction of the glass, and of the metal with which it is connected, has been obviated, by passing the ends of the tube into stuffingboxes; that on the score of its breaking by shocks, by giving it great thickness; and that of its breaking by sudden variations of temperature, by using well annealed glass. The difficulty of the glass clouding when high *The hydrostat described in No. XXX. of replies is inadmissible, from the interior of the boiler being occupied by a second cylinder, leaving only an annular space for the production of steam. For the alarm-floats of J. L. Sullivan, Esq., see Silliman's Journal, vol. xx. p. 1. † See the alarm-float of Siebe. Lond. Jour. of Arts, vol. xiii. p. 273. The first of those known to the Committee. Report, pp. 14, 15, Plate 4, fig. A. Jour. Frank, Inst. vol. xvii. pp. 13, 14. § It is exposed to a slight objection from steam. pressure acting to keep the disks upon the openings; these latter are, however, quite small, and the pressing surfaces of the disks may be regulated accordingly. The float described by Mr. Ewbank, Jour. Frank. Inst. vol. x. p. 7, is also referred to by the Committee as deserving a full trial. steam is used, by the action of the steam on the alkali, is got rid of by using green glass. The experiments which the Committee made on this apparatus were highly satifactory, and they confidently recommend its use to practical men.|| 46. 3rd. The danger resulting from a deficient supply of water, being produced by the undue heating of parts of a boiler, many means have been proposed for showing such an increase of temperature, before it attains a dangerous degree. The fusible plates applied to the top of the boiler, may be intended to indicate the general temperature of the steam when saturated with moisture or not, or the local temperature resulting from the rising of hot and unsaturated steam, produced by unduly heated metal. In any case they will be exposed to pressure, though to less in the second case than in the first. The objections already urged, and derived from experiment, will apply to their use, in the ordinary way, in any one of these cases. Indeed, without this objection, it appears that as the source of danger is the heated metal, to this the indicator of temperature should be applied. 47. Various methods of indicating the temperature of a part or parts of a boiler have been contrived. One of the most simple is to apply the common thermometer, inserting the bulb and as much of the stem as is necessary, in a tube closed at one end and fastened into the boiler. The tube should contain mercury, through which the heat is conducted to the thermometer. Such a tube may be placed at or near the water line of the boiler, at the fire end of a flue, or in general at the place of a greatest exposure to heat from a deficiency of water, of which there will usually be one or more well determined, according to the form of the boiler. A mark upon the scale of the thermometer will show the temperature above which the metal must not be allowed to rise, either from an increased elastic force in the steam, or from a deficient supply of water. The fragility of the instrument, its inconvenient length or position in certain cases, and its not acting as an alarm, are the principal objections to its use. T 48. The expanding rods proposed by Mr. Cadwallader Evans, are ingenious; they give, however, not the local temperature of the boiler, but its general temperature, along the lines to which the rods are applied. A much more appropriate device is the application of fusible metal, proposed by the same gentle See the description of the tube water-gauge used by the Committee in the first part of their Report. Jour. Frank, Inst. vol. xvii. p. 10. Mech. Mag. vol. xxv. p. 88, For a detailed description of the mode of applying the thermometer, see the Report of Com, on Expl. Part I. pp. 7 and 8, and Jour. Frank. Inst. vol. xvii. pp. 5, 6. man.* This is intended to take the place of the ordinary fusible plate, and to avoid the difficulty originally existing, but since remedied,t of replacing the plate when it had fused. In the apparatus submitted to the Committee by their chairman, and subsequently made the object of experiment, the fusible metal is applied to the most exposed part of a boiler; it is so small in quantity, that it will serve to indicate a local temperature, while the motion which indicates its fusion is independent of the quantity of fusible metal. These instruments are intended respectively to give notice when the steam, or the metal of the boiler, is exposed to a temperature much below that which would produce danger. Both of them, after giving an alarm, can be immediately restored to action, if the temperature within has been made to decrease.§ By very simple changes in the apparatus, the fusible metal might be applied to the boilers of locomotive engines. Sir,-In the report of the Select Committee of the Lords, to inquire whether any danger by fire is likely to arise from locomotive engines on railroads, passing through narrow streets, the Committee say, "It does not, however, appear that as yet any plan has been devised, which could be described as being at once completely effectual, and, at the same time, not injurious to the operations of the engine." Mr. Stephenson, and other eminent engineers, who were on that occasion examined, stated to the Committee several plans which had been tried, by putting sieves in the chimney, both at top and bottom; but that the effect was materially to injure the draught, and consequently the speed of the engine. The following method (of which I enclose you a sketch) seems to me to obviate the objections which are inseparable from the contrivances hitherto tried, and which (if you *Communication to Com. on Explosions, No. XXII. of Replies, &c. Jour. Frank. Inst. vol. ix. February, 1832. Patented in May, 1834. See specification in Jour. Frank. Inst. vol. xiv. p. 391. The Committee prefer this to the apparatus acting by the expansion of mercury. Hall's method of applying the fusible plate. Bulletin, de la Soc, d'Encouragement, &c. Described in Jour. Frank. Inst. vol. x. p. 217. The apparatus devised by the Chairman of this Committee was made the subject of experiments, and, with them, described in Part I. of the Report of Com. on Expl. See Mech. Mag. vol. xxv. p. 118. consider it has sufficient merit) I shall be glad to see inserted in your Magazinę, A A represents the hot-air chamber at the end of the boiler, into which the tubes from the fire-box lead, and from which springs the chimney D. C is the blastpipe, which conveys the waste steam by the pipes B B, from the cylinders into the chimney, causing a vacuum, and consequent draught. Now, my plan is, to stretch a wiregauze across the chimney from E to E, just below the orifice of the blast-pipe, and, at the same time, to widen the chimney at the same point, so that when the gauze is in its place, the free-air way shall equal, or, if necessary, surpass in area the rest of the chimney. When the gauze was tried at the bottom of the chimney, it was soon choked with cinders and dust, and speedily destroyed |