the Prussian law, every new locomotive boiler had to be re-tested to double the working pressure after running 8000 Prussian miles, and afterwards for each 5400 miles. These measures, while they did not entirely prevent explosions, greatly injured the boilers, by straining the stay bolts, and by the resulting excessive caulking required to make the joints tight. On the other hand, the absolute security afforded by drawing the tubes can, under the present mode of construction, be only obtained at the expense of perhaps 300 tubes, costing from 258. to 278. each, besides some injury to the tube plates. Whatever may be said against the hydraulic test, the best argument in its favor is its very general adoption. New government boilers in the United States must be tried to a pressure two-thirds greater than the working pressure; the same measure being carried out with the 3000 boilers in the city of New York. Mr. Anderson, C.E., of Woolwich,* * directs his subordinates to use a test of at least double the working pressure for the boilers in the royal gun factories. Mr. Muntz, of Birmingham, has publicly stated that he has for years adopted an annual hydraulic test, "considering it a duty he owes to his workmen." The Eastern Counties, the South Eastern, the Lancashire and Yorkshire, the Caledonian, the North British, the Edinburgh and Glasgow, and the London and South Western Railway companies employ the hydraulic test for both new and old boilers, using generally double the working pressure. The London and North Western are stated to have used it for only new boilers at any rate until recently. The Great Northern and the Great Western Railway companies do not use it, and it is accordingly on these lines that the greater number of explosions take place. Practical experience thus proves that, though there is just a chance of the test failing to detect a weak boiler when it cannot be examined internally, the danger is greater in not using the hydraulic test at all. Mr. Beattie, of the London and South Western, strips the lagging every two years, and applies a pressure of 190 lbs., the working pressure being 125 lbs. Mr. Fletcher, of the Manchester Boiler Association, employs double the intended working pressure for new, and from 1 to 13 the working pressure with old boilers. The most commonly used test is thus double the working pressure for old boilers, with a diminution according to circumstances as they get old. An objectionable plan in measuring the pressure applied, and, for several reasons, one likely to lead into error, is estimating it from the load on the safety valve lever. A metallic gauge should be used, and very neat pocket instruments of the kind are sold in Paris. In frosty weather the rivet heads are liable to be snapped if the metal be not somewhat warmed by using hot water. The hydraulic ram kind of action on the sides is also much less likely to occur if a rather narrow force pipe be used for the pump. There can be no doubt that it would be a valuable thing to be able to employ some means of measuring the permanent and temporary * "Instructions to be Observed in the Management of Steam Boilers in the Royal Gun Factories." extension of volume, if any, produced by the hydraulic test. It is probable that a boiler, as it gets old, and takes a permanent set under the pressure, also increases in volume; so that it doubtless holds a few gallons more than it did when new. An ingenious plan for measuring the increase of volume is recommended in the Bavarian regulation. After the boiler is filled, the amount of water forced in is measured by pumping it from a vessel marked with divisions. When the pressure is removed, the boiler contracts more or less, forcing out at least a portion of the water; the amount remaining is supposed to give the dilatation of volume of the boiler. The difficulty in the use of this plan would probably consist in the presence of air in the water itself, and any which might chance to remain in the boiler. That in the water might be greatly diminished, or at any rate brought down to a constant amount by boiling; but there would be no precise security against any air in the boiler, and as the weight of the air absorbed by water (according to a well known law) is in proportion to the pressure, it would be taken up by the water, thus falsifying the indications when the pressure was removed. On the other hand, a high temperature of the water would form an impediment to this absorption. The experiment is certainly worth trying. It might be very valuable with tubular boilers inaccessible from the inside, as any permanent set or deflection ought to be indicated by little or no water being compressed out by the contraction of the boiler on the removal of the pressure. As long ago as 1844, M. Jobard, of Brussels, in order to obviate the supposed injurious effects of the hydraulic blow of the water on the plates, proposed to fill the boiler with water, first loading the safety valves, and to then dilate the water, and consequently the boiler, by means of heat applied to the outside.* More recently, Dr. Joule, of Manchester, has used the same plan himself, proposing it for general adoption.† In addition to the loaded safety valve, he used a metallic pressure gauge "to be constantly observed, and if the pressure arising from the expansion of the water goes on increasing continuously without sudden decrease or stoppage until the testing pressure is obtained, it may be infered that the boiler has sustained it without having suffered strain." Another plan, also founded upon the same principle of the irregularities of extension of metals when the limit of elasticity is exceeded, has lately been proposed. This consists in bringing an ordinary steam engine indicator in communication with the pump plunger as if it were a steam engine piston rod. The ordinates of the pencil diagram would thus give the pressure in the boiler, while the respective abscissæ would give the quantity of water pumped in at each stroke. As long as the limit of elasticity was not exceeded there would be a horizontal line, while a curved line would appear as soon as the sides began to take a permanent deflection. There seems to be a sort of contradiction in depending for results like these upon such irregular appearances as * "Technologists,” 1844, page 135. "On a Method of Testing the Strength of Boilers." Journal of the Manchester Philosophical Society, 1862, page 97. Polytechnisches Centralblatt, page 1337, 13th October, 1864. the extension beyond the elastic limit. But all these proposals are undoubtedly worth trial in practice. Dr. Joule's plan has the merit of affecting the plates by both heat and pressure, thus bringing them under every day conditions. (To be continued.) Temperatures and Pressures of High Pressure Steam. By R. A. PEACOCK, C.E., Jersey. From the London Artizan, June, 1865. An examination of the following table-which has never been published before, and, with the exception of Dr. Fairbairn's experiments, is quite new-will satisfy the reader that from 25 lbs. per square inch up to 411.6 lbs. per square inch, the temperature increases as the 43 root of the pressure. The greatest variation from M. Regnaults experiments is with his temperature of 300° Fahr.; such variation, however, is less than per cent., viz: 232, or 1 in 431. It will be further observed that even with regard to that small variation, the result ought to be considered as modified by the very close approximation of the calculation to Dr. Fairbairn's experiments with a temperature only a little less, viz: 292-53° Fahr., for which the calculation gives a difference of only 017 per cent., or 1 in 5882. With M. Regnault's temperature of 410° Fahr., the calculation is all but identical, the difference being only 002 per cent., or 1 in 50,000, while with his last temperature, viz: 447° Fahr., the difference between experiment and calculation is still unimportant, being no more than 159 per cent., or 1 in 629. And it may be further observed, that the formula, Temperature increases as 4 root of pressure, is, so to speak, still more moderate than M. Regnault's own formula; because it will be seen that the former requires 447.71° to give the pressure of 411.6 lbs., whereas M. Regnault's formula gets that pressure with 447° Fahr. only. That is to say, if both formula were to be applied (hypothetically) to calculate very high temperatures of, say, 2000° or 3000° Fahr., M. Regnault's formula would give greater pressures than the other. Probably the means do not exist for ascertaining how far either formula differs from fact at those high temperatures. Yet, although exact knowledge does not exist, we are fortunately not without some indications of the force of steam pressure at very high temperatures. The Rev. John Michell wrote a very valuable paper, in which he contended that earthquakes were caused by steam,* which paper does not appear to have gained as much consideration as it probably deserves. He says that in casting two brass cannon "the heat of the metal of the first gun drove so much damp into the mould of the second, which was near it, that as soon as the metal was let into it, it blew up with the greatest violence, tearing up the ground some feet deep, breaking down the furnace, untiling the house, killing many spectators on the spot with the streams of melted metal, and * Phil. Trans. R. S., 1760, vol. xi., p. 447, &c. VOL. L.-THIRD SERIES.-No. 3.-SEPTEMBER, 1865. 14 scalding others in the most miserable manner." These great effects were evidently produced by the steam of a few ounces of water only, for it is called merely "damp," and it must, therefore, have been very powerful steam. Now, according to the late Professor Daniell, F.R.S., brass melts at 1869° F., but the heat of the steam could not have been as much as that, because, amongst other reasons, a portion of the heat must have been taken up by raising the temperature of the "damp," and the "mould," and the neighboring sand. If the temperature of steam could have been 1869° F., the hypothetical pressure would have been by the formula, 114 tons per square inch; but the real force consequent on the reduced temperature most likely was considerably below 100 tons. What is intended to be inferred is, that the pressure of steam continues to increase at all events as high as up to water converted into steam by an initial heat of 1869° F. The law of increase may either be according to either of the formulæ, or to some of the other well known formulæ, or to some other unknown law, but at all events the force had continued to increase. The columns headed "Pressures per square inch," so far as they relate to M. Regnault's experiments, are reduced from Table I, pp. 259-260, of the Rev. R. V. Dixon's "Treatise on Heat,"* where he gives the pressure or force in inches of mercury. He was elaborately accurate in reducing M. Regnault's temperatures and pressures to English denominations, having calculated the values of the constants from Vlacq's tables, in which the logarithms are given to ten places of decimals, (p. 252.) Further particulars of this formula appeared in the Artizan of January and February, 1864. The present writer exhibited before the Mathematical and Physical Section of the British Association at Bath (but did not read) a MS. containing about sixty evidences of the presence of water in some of its forms, in every species of natural disturbance of the earth's crust. Additional evidences have since been collected, and the whole number now amounts to about seventy; and a summary of these evidences gives the following results: Humboldt, who, perhaps, personally inspected more volcanos than any other man who ever lived, and Sir Humphrey Davy, both condemn the alkaloid theory as co-operative only; and 1. We have Humboldt, Davy, Lyell, Von Buch, Dana, Sir William Hamilton, Dr. Scherzer, and three anonymous writers, testifying to the ejection of abundance of steam from volcanos generally, and some of them from Vesuvius in particular. 2. Steam is ejected from earthquake fissures. 3. It is said that steam is exclusively the moving force in geysers, and proved that it issues from them in great force and abundance. 4. Large rocks are ejected by steam. 5. Submarine volcanos necessarily produce steam, and one of them has been active for 2000 years. Earthquakes often accompany. Steam was very active when Graham Island rose from the Mediterranean. *Hodges and Smith, Dublin, 1849. |