A. The requisite buoyancy is given to the metallic pyramid, a, which is solid, by the weight b, acting as a counterpoise over the fulcrum c. The whole apparatus is attached to the top of the boiler by the screw d, and the nut e, and the working parts are thus entirely within the boiler. When the water is at the proper level fg, the shoulders h and i are in the same horizontal line, and the disks k l, which are pressed against the shoulders by two springs shown in the figure, close the apertures k m, and I n, which, when open, permit steam to escape from the boiler. Should the water sink below its proper level, the equilibrium of the pyramid a, being destroyed, the shoulder i, would press against the disk 1, remove it from the aperture, and permit steam to escape through In; should the water, on the contrary, rise above the proper level, steam would escape through k m. The force of the springs which close these openings, should, of course, be duly proportioned, as they will determine the sensibility of the apparatus. The details of construction are clearly shown in the figure, which is drawn to a scale.*

The quantity of steam which would escape by the small opening In, while it would serve as an alarm, would not materially diminish the supply of water within the boiler. The float used by the committee, was found to be sensible to less than three-tenths of an inch in the change of level: it could have been made more sensitive by increasing the breadth between the shoulders, so as to bring them in contact with the disks, as shown in the figure.

Effect of Foaming on the Elasticity of the Steam within the Boiler.

This point was the next proposed for examination. When an opening is made in a boiler, of which the sides are heated, will the effect be to diminish the elasticity of the steam within, by permitting its escape, or will the water thrown upon the heated sides by the foaming which results, be converted so rapidly into steam as actually to increase the elasticity of the vapour within? It is obviously difficult to obtain an answer to a query involving so many conditions. It might be expected, however, that a small boiler would afford satisfactory means of making a fair trial of the question, since the size of the openings could be varied very easily, so as to make them comparatively small, or very great. The position of the boiler used by the committee in its furnace was such, that the sides could be very readily heated; thus placing it in favourable circumstances to increase the elasticity of the steam by producing a foaming within. The apparatus was therefore adapted to make the desired trial.

M. Arago, in his Essay on the Explosions of Steam-Boilers, states, that MM. Tabareau and Rey, at Lyons, found on opening a large stop-cock, connected with a small high pressure boiler, that the safety valve rose, showing an increase of pressure within. The boiler was placed naked upon a fire of charcoal, and the part not containing water was surrounded by flame. The experiments of MM. Arago and Dulong, at Paris, were attended with a contrary result, the opening of a safety valve being always accompanied by a diminution in the elasticity of the steam within. The circumstances, however, were not the same as those in the experiment of MM. Tabareau and Rey.

To repeat this experiment, a hot fire was made beneath the boiler, and when the water had fallen to about three inches above the lowest line of the cylinder, the experiment was commenced, the pressure being about three atmospheres and a half. A stop-cock of .03 sq. inches in area, 10th part of the area of the water surface at the beginning of the experiment, delivering per second, at three

* In the figure, the shoulders h and i, do not rise high enough; they should overlap the disks more, that no depression or elevation of the water may carry them clear of

the disks.

and a half atmospheres, about four hundred and nine cubic inches of steam was first opened; next the safety valve was raised, either in part or entirely, the area when entirely raised, being .208 sq. inches, or oths of the water surface, and capable of delivering, in one second, at three and a half atmospheres, a bulk of steam nearly nine times that of the steam chamber. The water level falling by the waste caused in the experiments, the steam soon became surcharged with heat; and the iron of the boiler, from near the water line to more than one-third of the distance from the lowest line to the middle of the convex surface, became, on each side of the water line, heated until it attained redness, passing, of course, through the temperature of maximum vaporization of the water thrown by the foaming, upon the iron. The experiments were made at intervals, until all the water was exhausted. Water was then injected in small quantities, and with the bottom of the boiler for the most part red hot, the trials were repeated.

It will be seen from the following table, that the result was uniformly a diminished elasticity of the steam within, as shown by the fall of the mercury in the steam gauge. The pressures varied, in the former part of the experiments, from three and a half to eight atmospheres.

The first column of the table contains remarks referring to the level of the water within the boiler. The second to the opening made. The third is the temperature indicated by the thermometer, M, Plate 1, before referred to, as passing nearly to the bottom of the boiler. The fourth, the height of the mercury gauge, before making the opening. The fifth, the height immediately after making the opening, unless the contrary is stated in the sixth column, which contains remarks relating to the effect on the gauge. The seventh column contains general remarks.

The thermometer at first indicated the temperature of the water, then that of the surcharged steam, and finally was affected by the heat radiated from the bottom of the boiler.