foaming within the boiler, which was produced by thus relieving the pressure, was distinctly seen through the glass windows. On opening the third cock, steam and water issued from the second, which was two inches above the water level; and on partially raising the safety valve, water flowed freely from the second cock. A further rise of the valve filled the boiler with foam, water flowed freely out of the third cock, more than three inches and threequarters above the water level, and finally through the opening of the safety valve itself. In these experiments, an opening of .03 of a square inch in area, the lowest cock, which, to the area of the water surface, was as one to thirteen thousand seven hundred, caused water and steam to issue through a cock, below which the water was known to be. A further opening of .03 of a square inch, making, with the first, .06 inch, or one six thousand eight hundred and fiftieth the area of the water surface, brought water from the lowest cock; a total opening of .09 inch, (7th of the area of the water surface) brought water and steam from the middle cock, indicating that the level of the water was nearly two inches higher than it really was.

A first apparatus, which was contrived for applying fusible plates to the boiler, suddenly opened an aperture of .95ths of an inch in diameter. Even at low pressures, the scalding contents of the boiler were violently discharged, through this opening, against the roof of the experiment house.

It is time now to speak of the glass gauge-tube, as a means of indicating the level of the water within a boiler, in connexion with which an experiment bearing upon the performance of the gauge-cocks will be stated.

The form given to the water gauge, on its first trials, was that described to the committee by Mr. Hartshorne, of Cincinnati. This was a prismatic box of brass, of suitable dimensions; one face of which was supplied by a glass plate; this box, being put in communication with the boiler, by two pipes, one entering from the steam, the other from the water, the level of the water was seen through the glass plate. This apparatus was attached to the experimental boiler, and its indications compared with those of the gauge-cocks in the experiments already detailed. On relieving the water from pressure, the water within the gauge was agitated; during the further foaming its oscillations did not amount to half an inch, so that the hydrostatic level was truly shown by it; and further, on closing the openings, the fluid in the gauge became tranquil at the mean level of its oscillations, showing that it had fallen with the fall of the water within, caused by the escape of steam. An instructive experiment to the same purport was made on the occasion of a fracture in one of the glass windows, described as placed in the ends of the boiler. The account taken from the minutes of the day's experiments is as follows:

The temperature being at 292° F., and the pressure indicated by the gauge four atmospheres, the north window of the boiler, which had a flaw in it, cracked across the middle, and nearly horizontally; steam, issued slowly through the crack; on looking into the boiler a foaming at the end where the steam was escaping was observed. The crack rapidly enlarging, the steam issued in quantities through it; the water was in general agitation throughout the boiler, running out at the crack, though its hydrostatic level was at the bottom. of the window, about one inch and a quarter below the crack, and being distinctly seen at the opposite window, foaming near the top of the glass; the water-level gauge began to fall, oscillating not half an inch in its fall. The safety valve was now opened by hand, so as to waste the water with great rapidity. The water still issued through the crack, the water-level gauge falling. On closing the valve the water settled down, becoming comparatively tranquil; the water gauge remained at the same level: it had, therefore, indicated constantly the true level unaffected by the foaming, except in slight oscillations. In fact, this gauge shows truly the height of the water within the boiler, o high as to pour over through the upper connecting tube.

The idea was suggested that by placing the gauge-cocks in a prism, connected above with the steam, and below with the water in the boiler, the true level of the water would be indicated. Such a cock was, therefore, applied to the box of the water gauge; its opening produced a local foaming in the gauge, which brought water through the cock, although the true level was much below it. The area of this cock was nearly equal to the area of that which opened into the steam chamber of the boiler.

In relation to the form of the water gauge, as already described, it does not seem to offer as many advantages as the tube which has been adapted to the boilers of some of the English locomotive engines.* The glass plate requires the support of horizontal bars, which are objectionable, or it must be reduced so much in breadth that the level is obscurely seen through it; the strain upon the plate being unequal, it is very liable to fracture; such fractures repeatedly took place near the centre of the plates in the gauge used by the committee. To the use of the glass gauge for the high pressure engine, an objection occurs, from the effect produced by high steam upon the glass, apparently by its action on the alkali; by which the transparency of the glass is gradually destroyed. A similar effect was recorded by Cagniard de Latour, in his experiments on liquids at high temperatures, confined in glass tubes.† As far as the experiments of the committee have gone they show that green glass is not so readily injured; and as it is easily procured in tubes, a further reason appears for preferring the tube, in practice, to the plate.

An attempt which was made to substitute windows of mica in the boiler for those of glass, bears upon the use of that mineral for the plate of the water gauge; as does also another attempt which was made to protect the glass plates by a lamina of that mineral. The mica exfoliated under the action of the steam which insinuated itself between the laminæ through cracks which were invisible, if existing, before the experiment, or which may have been produced by the steam itself; the lamina were separated, and thus the steam quickly found a more or less direct passage through the plate.

2.

The tube gauge which was substituted for the prism is shown in Plates 1 and wx is the tube of green glass passing into the stuffing boxes, w' and x'; the stuffing enables an adjustment to be made for the unequal expansion of the glass and metal by heat, and prevents fracture on the subsequent cooling of the apparatus. y and z, Plate 1, are passages connecting the tube with the boiler; these have conical terminations, by which the pipe is readily attached to, and detached from, the tubes y and z, which are screwed into the boiler, and are provided with stop-cocks: coupling screws might, in practice, be substituted for these conical terminations. To protect the tube, w x, from currents of air, it was surrounded by a second tube, loosely applied. A scale was attached to w x, to indicate the level of the water within the boiler. The tube being transparent, shows the level of the water more readily than it can be seen in the prism before referred to, which was opaque on three of the vertical sides.

The gauge used was nine inches and three-quarters in length. The upper part being so near the top of the boiler as only to be affected by the foaming, in extreme cases; the lower part so near to the bottom that the level of the water was indicated, unless when very low indeed.

The position of the lower communication of the gauge with the boiler soon

*The application of such a gauge to a locomotive engine can give but little idea of its use for a stationary engine. The jarring in the locomotive must constantly expose the gauge to fracture, and perhaps may prevent its use. The glass water gauge has been adopted in at least one of the boats plying between New York and Amboy, New Jersey, belonging to Messrs. Stevens; and the committee understands, are also in use upon their locomotive engines.

See also recent experiments on the action of water at high temperatures upon glass,

m

C 100

showed a defect, to which the instrument must be always more or less liable, namely, to the obstruction of the lower, passage by sediment. To remedy this, a stop-cock was attached to the lower part of the gauge, as at x', Plate 1, and through it, when open, water could be blown, by the pressure of the steam in the boiler, so as to remove any obstruction. This method is to be preferred to that of closing the upper communication with the boiler, while the lower one remains open; in which case the sediment is driven up into the glass tube, soiling it, accumulating there, and affording only a temporary remedy. When the obstruction in the pipe is not removed on opening the cock, a wire inserted will effectually clear the passage.

In connexion with this subject the committee experimented upon the method proposed by Mr. Thomas Ewbank, of New York, for lessening or preventing the foaming here stated to occur. The remarks of Mr. Ewbank are to the following effect: "When steam is raised in a boiler, and the engine not work“ing, the water within (if the flues are sufficiently covered) is probably on a "level and nearly at rest; but as soon as the steam is admitted into the cylinder, "it causes an ebullition of the water, which rises towards the mouth of the "steam pipe, in consequence of the portion of the pressure upon it being suddenly "removed at every stroke of the piston. This might, I think, be prevented by "continuing the steam pipe an inch or two into the boiler, and then branching "it off towards each end of it, with small apertures in its sides and ends, as in

"the diagram. In this manner, the steam "would be equally withdrawn from every "part of a boiler, instead of being violently "agitated in rushing to one place. Such a "tube attached to the aperture of a safety"valve, would be equally advantageous; or the "valve might be placed on one end of the "tube leading to the cylinder.*

"The inaccuracy of the common gauge-cock as a means of detecting the "true height of the water in a boiler, arises chiefly from two causes; firstly, "from the agitation of the water while steam is being withdrawn from the "boiler to supply the engine, or through the safety valve: and secondly, from "the current or rush of steam produced towards the aperture of a gauge-cock "when it is open; in consequence of which the water, though previously at "rest, is agitated and carried out through it.

"The last mentioned, defect may be lessened "by a perforated tube (see the accompanying figures,) attached to the end of the cock which "is within the boiler. Such a tube would pre"vent the current from being concentrated to"wards the aperture of the cock as the steam "would enter it through the small openings, in "The next figure shows a method by which both the "defects to which I have alluded as affecting the gauge-cock, may be reme

"various directions."

"died. The cock passes through the head of "the boiler in the usual way, and is then united "to a perpendicular pipe, open at both ends, and "about two or three inches in diameter. The "lower end of the pipe is four or five inches "below the surface of the water, and its upper "end is carried as far above the level as may "be convenient." "When this cock is opened no current can be formed in

Journal of the Franklin Institute, vol. 9, p. 366, 1832, letter from Thos. Ewbank,

of New YAAA Com

The idea was suggested that by placing the gauge-cocks in a prism, connected above with the steam, and below with the water in the boiler, the true level of the water would be indicated. Such a cock was, therefore, applied to the box of the water gauge; its opening produced a local foaming in the gauge, which brought water through the cock, although the true level was much below it. The area of this cock was nearly equal to the area of that which opened into the steam chamber of the boiler.

In relation to the form of the water gauge, as already described, it does not seem to offer as many advantages as the tube which has been adapted to the boilers of some of the English locomotive engines.* The glass plate requires the support of horizontal bars, which are objectionable, or it must be reduced so much in breadth that the level is obscurely seen through it; the strain upon the plate being unequal, it is very liable to fracture; such fractures repeatedly took place near the centre of the plates in the gauge used by the committee. To the use of the glass gauge for the high pressure engine, an objection occurs, from the effect produced by high steam upon the glass, apparently by its action on the alkali; by which the transparency of the glass is gradually destroyed. A similar effect was recorded by Cagniard de Latour, in his experiments on liquids at high temperatures, confined in glass tubes.† As far as the experiments of the committee have gone they show that green glass is not so readily injured; and as it is easily procured in tubes, a further reason appears for preferring the tube, in practice, to the plate.

An attempt which was made to substitute windows of mica in the boiler for those of glass, bears upon the use of that mineral for the plate of the water gauge; as does also another attempt which was made to protect the glass plates by a lamina of that mineral. The mica exfoliated under the action of the steam which insinuated itself between the laminæ through cracks which were invisible, if existing, before the experiment, or which may have been produced by the steam itself; the lamina were separated, and thus the steam quickly found a more or less direct passage through the plate.

2.

The tube gauge which was substituted for the prism is shown in Plates 1 and wx is the tube of green glass passing into the stuffing boxes, w' and x'; the stuffing enables an adjustment to be made for the unequal expansion of the glass and metal by heat, and prevents fracture on the subsequent cooling of the apparatus. y and z, Plate 1, are passages connecting the tube with the boiler; these have conical terminations, by which the pipe is readily attached to, and detached from, the tubes y and z, which are screwed into the boiler, and are provided with stop-cocks: coupling screws might, in practice, be substituted for these conical terminations. To protect the tube, w x, from currents of air, it was surrounded by a second tube, loosely applied. A scale was attached to w x, to indicate the level of the water within the boiler. The tube being transparent, shows the level of the water more readily than it can be seen in the prism before referred to, which was opaque on three of the vertical sides.

The gauge used was nine inches and three-quarters in length. The upper part being so near the top of the boiler as only to be affected by the foaming, in extreme cases; the lower part so near to the bottom that the level of the water was indicated, unless when very low indeed.

The position of the lower communication of the gauge with the boiler soon

*The application of such a gauge to a locomotive engine can give but little idea of its use for a stationary engine. The jarring in the locomotive must constantly expose the gauge to fracture, and perhaps may prevent its use. The glass water gauge has been adopted in at least one of the boats plying between New York and Amboy, New Jersey, belonging to Messrs. Stevens; and the committee understands, are also in use upon their locomotive engines.

See also recent experiments on the action of water at high temperatures upon glass, Professor Turner of the University of London. Roval Society's Trans. for 1834.

showed a defect, to which the instrument must be always more or less liable, namely, to the obstruction of the lower, passage by sediment. To remedy this, a stop-cock was attached to the lower part of the gauge, as at x', Plate 1, and through it, when open, water could be blown, by the pressure of the steam in the boiler, so as to remove any obstruction. This method is to be preferred to that of closing the upper communication with the boiler, while the lower one remains open; in which case the sediment is driven up into the glass tube, soiling it, accumulating there, and affording only a temporary remedy. When the obstruction in the pipe is not removed on opening the cock, a wire inserted will effectually clear the passage.

In connexion with this subject the committee experimented upon the method proposed by Mr. Thomas Ewbank, of New York, for lessening or preventing the foaming here stated to occur. The remarks of Mr. Ewbank are to the following effect: "When steam is raised in a boiler, and the engine not work"ing, the water within (if the flues are sufficiently covered) is probably on a "level and nearly at rest; but as soon as the steam is admitted into the cylinder, "it causes an ebullition of the water, which rises towards the mouth of the "steam pipe, in consequence of the portion of the pressure upon it being suddenly " removed at every stroke of the piston. This might, I think, be prevented by "continuing the steam pipe an inch or two into the boiler, and then branching "it off towards each end of it, with small apertures in its sides and ends, as in

"the diagram. In this manner, the steam "would be equally withdrawn from every

part of a boiler, instead of being violently "agitated in rushing to one place. Such a "tube attached to the aperture of a safety"valve, would be equally advantageous; or the "valve might be placed on one end of the "tube leading to the cylinder.*

"The inaccuracy of the common gauge-cock as a means of detecting the "true height of the water in a boiler, arises chiefly from two causes; firstly, "from the agitation of the water while steam is being withdrawn from the "boiler to supply the engine, or through the safety valve: and secondly, from "the current or rush of steam produced towards the aperture of a gauge-cock "when it is open; in consequence of which the water, though previously at "rest, is agitated and carried out through it.

"The last mentioned, defect may be lessened "by a perforated tube (see the accompanying figures,) attached to the end of the cock which "is within the boiler. Such a tube would pre"vent the current from being concentrated to"wards the aperture of the cock as the steam "would enter it through the small openings, in "The next figure shows a method by which both the "defects to which I have alluded as affecting the gauge-cock, may be reme

"various directions."

"died. The cock passes through the head of "the boiler in the usual way, and is then united "to a perpendicular pipe, open at both ends, and "about two or three inches in diameter. The "lower end of the pipe is four or five inches "below the surface of the water, and its upper "end is carried as far above the level as may "be convenient." "When this cock is opened no current can be formed in

Journal of the Franklin Institute, vol. 9, p. 366, 1832, letter from Thos. Ewbank,