We cannot afford to surrender any more space to these specimens of boilers, and must now proceed to despatch what we have still to say on the subject of boilers in as few words as possible. We have already stated that a cubic foot of water raised into steam is reckoned equivalent to a horse power, and that to generate the steam with sufficient rapidity, an allowance of one square foot of fire bars, and one square yard of effective heating surface, are very commonly made in practice, at least in land engines. These proportions, however, greatly vary in different cases; and in some of the best marine engine boilers, where the area of fire-grate is restricted by the breadth of the vessel, and the impossibility of firing long furnaces effectually at sea, half a square foot of fire-grate per horse power is a very common proportion. Ten cubic feet of water in the boiler per horse power, and ten cubic feet of steam room per horse power, have been assigned as the average proportion of these elements; but the fact is, no general rule can be formed upon the subject, for the proportions which would be suitable for a waggon boiler would be inapplicable to a tubular boiler, whether marine or locomotive; and good examples will in such cases be found a safer guide than rules which must often give a false result. A capacity of three cubic feet per horse power is a common enough proportion of furnace-room, and it is a good plan to make the furnaces of a considerable width, as they can then be fired more effectually, and do not produce so much smoke as if they are made narrow. regards the question of draft, there is a great difference of opinion among engineers upon the subject, some preferring a very slow draft and others a rapid one. It is obvious that the question of draft is virtually that of the area of fire-grate, or of the quantity of fuel consumed upon a given area of grate surface, and the weight of fuel burned on a foot of fire-grate per hour varies in different cases in practice from 3 to 80 lbs. Upon the quickness of the draft again hinges the question of the proper thickness of the stratum of incandescent fuel upon the grate; for if the draft be very strong, and the fire at the same time be thin, a great deal of uncombined oxygen will escape up through the fire, and a needless refrigeration of the contents of the flues will be thereby occasioned; whereas, if the fire be thick, and the draft be sluggish, much of the useful effect of the coal will be lost by the formation of carbonic oxide. The length of the circuit made by the smoke

As

varies in almost every boiler, and the same may be said of the area of the flue in its cross section, through which the smoke has to pass. As an average, about one-fifth of the area of fire-grate for the area of the flue behind the bridge, diminished to half that amount for the area of the chimney, has been given as a good proportion, but the examples which we have given, and the average flue area of the boilers which we shall describe, may be taken as a safer guide than any such loose statements. When the flue is too long, or its sectional area is insufficient, the draft becomes insufficient to furnish the requisite quantity of steam; whereas if the flue be too short or too large in its area, a large quantity of the heat escapes up the chimney, and a deposition of soot in the flues also takes place. This last fault is one of material consequence in the case of tubular boilers consuming bituminous coal, though indeed the evil might be remedied by blocking some of the tubes up. The area of water-level we have already stated, p. 48., as being usually about 5 feet per horse power in land boilers. In many cases, however, it is much less; but it is always desirable to make the area of the water-level as large as possible, as when it is contracted not only is the water-level subject to sudden and dangerous fluctuations, but water is almost sure to be carried into the cylinder with the steam, in consequence of the violent agitation of the water, caused by the ascent of a large volume of steam through a small superficies. It would be an improvement in boilers. we think, to place over each furnace an inverted vessel immerged in the water, which might catch the steam in its ascent, and deliver it quietly by a pipe rising above the water-level. The water-level would thus be preserved from any inconvenient agitation, and the weight of water within the boiler would be diminished at the same time that the original depth of water over the furnaces was preserved. It would also be an improvement to make the sides of the furnaces of marine boilers sloping, instead of vertical, as is the common practice, for the steam could then ascend freely at the instant of its formation instead of being entangled among the rivets and landings of the plates, and superinducing an overheating of the plates by preventing a free access of the water to the metal.

In the Transactions of the Institution of Civil Engineers several papers are given by Mr. Parkes and others descriptive of experiments made by them on steam boilers. We have, in Table 1., collected a few of the principal

results exhibited in Mr. Parkes' very voluminous tables, and we have added the two columns on the right hand side of the table to show at the same time the evaporative economy of the boilers in use at the East London Water Works. One of these boilers is on the Cornish plan, and attached to the Cornish engine there. The other is a waggon boiler, with an internal flue, for supplying steam to a Boulton and Watt pumping engine. The Cornish boilers are cylindrical, with an internal flue; and, as they are generally used with steam of from 15 lbs. to 35 lbs. above the atmosphere, they are made of plates half an inch thick. The left hand column in the table gives the mean results of experiments made on the boilers at the Huel Towan and United Mines in Cornwall. The second column from the left is devoted to Mr. Parkes' experiments at Warwick; and, according to him, about one sixth of the evaporation there given is due to his smoke-consuming apparatus. The third column exhibits the mean of eight experiments on waggon boilers, at the different places indicated at the head of the column, which were all (except the Albion Mills experiments) conducted by Mr. Parkes. The fourth column from the left contains the results of Mr. Smeaton's experiments on his atmospheric engine at Long Benton ; and the fifth column gives the mean of eleven experiments on locomotives by Mr. Pambour. Referring to the sixth line, it will be seen that the mean evaporative economy of the Cornish boilers is about the same as that of the Warwick boilers; but, if we exclude the East London Water Works boiler, the other Cornish boilers will show a decided superiority over all the rest. Lines five and eight show two of the principal peculiarities in the proportions of the Cornish boilers. It will be observed that the extent of their surface exposed to the heat, for each cubic foot of water evaporated, is about seven times as great as in any of the others. The combustion in their furnaces also is carried on at a very slow rate, there being only about 34 lbs. of coals burned on each square foot of grate. The only boiler that makes any approach to them in slowness of combustion is the one at Warwick. Mr. Parkes is a great advocate for slow combustion; and he founds his opinions principally on its effect in the Warwick boilers—at least it is from his experiments on them that he derives his opinion that the principle should be carried so far as it was in that case. We conceive that in this instance he has overlooked one very material circumstance. It will be observed in the table, that the heated surface bears very nearly the same proportion to the water evaporated in these boilers as in the other waggon boilers; but, before Mr. Parkes altered his furnaces, much more water was evaporated per boiler, and, consequently, the heating surface must then have been very small in proportion to the evaporation. We are, therefore, rather inclined to attribute the increased duty of the fuel to the increase of the heating surface of the boiler than to the diminution of the rate of combustion. Nevertheless, from other experiments of Mr. Parkes, we are disposed to think that, for economy of fuel, the combustion in the generality of waggon boiler furnaces is rather too rapid. The very large proportion that the heating surface in the Cornish boilers bears to the weight of water evaporated is, no doubt, to a considerable extent, rendered necessary by the thickness of the plates which the heat has to penetrate, and the high temperature of the water within them-both circumstances that retard the transmission of the heat. The Cornish practice, too, is universally, so far as we have been able to ascertain, in favour of this great extent of surface; yet we can hardly think that a small diminution of it would produce any injurious effect on the economical properties of the boiler, while it would save a considerable part of the original cost.

It will be seen, from what we have already advanced, that but a small part of the superior duty of the Cornish engines can be derived from the boilers; we must therefore look to the engines for the principal sources of their superiority, which may be comprised under these three heads:1st. The use of high-pressure steam cut off when a very small part of the stroke has been performed, and working expansively over the remainder.

TABLE II.

Smeaton's Atmospheric Engine, Long Benton, Northumberland, date 1772.

2nd. The careful clothing of every part of the engine where heat can escape. The cylinder is usually encased with a steam-jacket; and the steam-jacket itself, and all the steam-pipes, top of the boiler, &c., are protected from the cold air by being covered with a layer of three or four inches thick of ashes, saw-dust, or other good non-conductor of heat. The amount of saving effected in this way may be conceived to be very considerable. We have not met with any experiments to ascertain the amount of the saving by clothing the whole engine; but Mr. Wicksteed found that clothing the top of the boilers alone produced a saving of 10 per cent. in the fuel consumed.

3rd. The third main source of the great duty of the Cornish engines is to be found in the excellent system of registering and publishing the duty of each engine, which has for many years been prevalent in Cornwall. It has made both the proprietors and engineers much more careful than they otherwise would have been of a host of details that have elsewhere been considered too trifling to require notice; but which, nevertheless, in the aggregate, are of no small importance.

* W. Welsh.-S. Staffordshire.-L. Lancashire.-N. Newcastle.

The grand secret, however, of the economy of the Cornish engines lies in the large application of the principle of expansion, and the results there obtained are very little aided by any peculiar excellence in the boiler. Upon the merits of expansion, however, or the pitch to which it may be beneficially carried in particular cases, this is not the place to enlarge, but we may here give a table from the Artizan, which shows the relative efficacy of different engines with different degrees of expansion. The first column on the left exhibits the performance of Smeaton's atmo

spheric engine at Long Benton. In the second column will be found similar particulars of Boulton and Watt's condensing rotative engine at the Albion Mills. In the third column we have put down the same data for the Holmbush Cornish engine: and, in the fourth, for a high-pressure non-condensing engine, whose duty was determined by Mr. Wicksteed. The last two columns supply the same data for a Cornish engine and Boulton and Watt pumpingengine at the East London Water Works. The economical effects of expansion will be found to be very clearly exhibited in this table. The duties are recorded in the fifth line from the top, and the degree of expansion in the bottom line. It will be observed, that the order in which the different engines stand in respect of superiority of duty is the same as in respect of amount of expansion. The Holmbush engine has a duty of 140,484,848 lbs. raised 1 foot by 1 cwt. of coals, and the steam acts expansively over 830 of the whole stroke; while the water-works' Cornish engine has only a duty of 105,664,118 lbs., and expands the steam over only 687 of the whole stroke. Again, comparing the two Boulton and Watt engines together, the Albion Mills engine has a duty of 25,756,752 lbs., and no expansive action. The water-works' Boulton and Watt engine, again, acts expansively over one-half of its stroke, and has an increased duty of 4,660,333 lbs. Other causes, of course, may influence these comparisons, especially the last, where one engine is a double-acting rotative engine, and the other a single-acting pumping one; but there can be no doubt that the expansive action in the latter is the principal cause of its more economical performance.

Mr. Parkes has extended his investigations in the papers we have mentioned to the subject of locomotive boilers, but here he has fallen into some strange blunders. He has collected all the experiments of Wood, Pambour, and Lardner, on that subject, and thrown their results into the tabulated form, in the same way that he has treated other kinds of engines and the result is, that he thinks that he is entitled to reject the whole, as being utterly unworthy of credit. This sweeping sentence is pronounced principally on two grounds,-first, that the experiments, by the same person, flatly contradict each other; and, second, that they go to show that a cubic foot of water, as steam, does more effective work in a locomotive than in any other form of engine. Unfortunately for Mr. Parkes's reputation, however, while he was thus zealously engaged in bringing to light the errors of others, he fell into the very common mistake of overlooking his own and in this part of his paper he has scarcely analysed a single experiment that he has not misunderstood and perverted, till, towards the end of his paper, he gets involved in such a labyrinth of errors, that extrication is impossible. Mr. Parkes, however, seems not to have been at all aware of the treacherous ground on which he was standing; but, elated with the idea of having effectually demolished the experiments and theories of three authors who had usually been received as authorities, he proceeds with the utmost complacency to build a theory of his own on an entirely new foundation.

The fundamental principle of this new theory is thus enunciated." The momentum communicated to the entire mass set in motion, represents the useful mechanical effect exerted by the steam." This is the pith and marrow of the whole affair, and when it is explained that "momentum here is used in its ordinary sense, to signify the weight of the moving body, multiplied into its velocity, it will be at once perceived that this measure of the power of the engine takes no account of the resistance encountered by the train on the rails. It may, therefore, be demonstrated, on Mr. Parkes's momentum principle, that the power expended in driving a train at ten miles an hour up an inclined plane, is precisely the same as that expended in pushing it down the same plane at the same velocity. Mr. Parkes, however, sees no difficulties or inconsistencies in his way; he proceeds fearlessly to the application of the new theory. He takes all the locomotive experiments he can collect, and multiplies the weight of the train in tons by the velocity in feet per second. The products he puts down in a table, and calls them the momentums, which are the measures of the power of the engine. He then proceeds, with the utmost confidence, to purge the roll of all the erroneous experiments. He assumes that, whenever the weight of a train is increased its velocity must be diminished, (forgetting, of course, that the light train may be going up hill and the heavy one down,) in a greater proportion than the weight was increased. Hence the greater the velocity the less ought the "momentum" to be, the weight being the same, or the greater the weight the greater should the momentum be, and all the experiments that contradict these canons he condemns at once. Take a specimen :-" A reference to the Fury exhibits that engine as having performed more work at 25 than at 21 miles per hour; it is therefore with certainty we may conclude one or both of these experiments to be erroneous." The excellence of the new method everywhere comes out in strong relief. "I adduce these few comparisons to exhibit the facility and certainty with which they are developed by this method of investigation."" Another instance of the delicacy of this test." "This method of investigation discloses experimental defects as well as errors of fact." If the reader is curious to know more of the theoretical "defects, as well as errors of fact," that abound in this Essay, we beg to refer him to the Introduction to the second edition of Pambour's Treatise on Locomotives, where he will find it tried by such "delicate tests," that very little of it stands the proof,

have already expatiated sufficiently: we shall therefore conclude our remarks upon the subject by introducing a table of the comparative evaporative power of different kinds of coal, which we have derived from Mr. Wicksteed's experiments, and which will prove useful, by affording data for the comparison of experiments upon different boilers when different kinds of coal are used. Without this means of reduction, experiments would be useless for comparison, unless the fuel employed was in every case of equal evaporative power; but when the relation between the evaporative power of different kinds of coals is ascertained, the results of experiments can be easily reduced so as to render them comparable with one another. Table of the Comparative Evaporative Power of different kinds of Coal.

Strength of boilers. The extension of the expansive method of employing steam to boilers of every denomination, and the gradual introduction in connection there with of a higher pressure than formerly, makes the question of the strength of boilers one of great and increasing importance. This topic was very successfully elucidated a few years ago by a committee of the Franklin Institute, and we shall here recapitulate a few of the more important of the conclusions at which they arrived. Iron boiler plate was found to increase in tenacity as its temperature was raised until it reached a temperature of 550° above the freezing point, at which point its tenacity began to diminish. The following table exhibits the cohesive strength at different temperatures.

At 320 to 80° the tenacity was 56,000 lbs. or 1-7th below its maximum.

At 5700

At 7200

At 10500

At 12400

At 13170

At 3000 iron becomes fluid.

=66,500 lbs., the maximum. =55,000 lbs., the same nearly as at 320. = 32,000 lbs., nearly of the maximum. = 22,000 lbs., nearly of the maximum. = 9,000 lbs., nearly 1-7th of the maximum.

We must now, however, dismiss these speculations, and with them the whole subject of the efficacy of different kinds of boilers upon which we

The difference in strength between strips of iron cut in the direction of the fibre, and strips cut across the grain, was found to be about 6 per cent. in favour of the former. Repeated piling and welding was found to increase the tenacity and closeness of the iron, but welding together different kinds of iron was found to give an unfavourable result; rivetting plates was found to occasion a diminution in their strength, to the extent of about one-third. The accidental over-heating of a boiler was found to reduce its strength from 65,000 lbs. to 45,000 lbs. per square inch. Taking into account all these contingencies, it appears expedient to limit the tensile force upon boilers in actual use to about 3000 lbs. per square inch of iron; and in cases where the shell of the boiler does not afford this strength, either stays should be introduced, or the pressure within the boiler should be diminished. The application of stays to marine boilers, especially in those parts of the water spaces which lie in the wake of the furnace bars, has given engineers much trouble; the plate, of which ordinary boilers are composed, is hardly thick enough to retain a stay with security by merely tapping the plate, whereas, if the stay be rivetted, the head of the rivet will in all probability be soon burnt away. The best practice appears to be to run the stays used for the water spaces in this situation, in a line somewhat beneath the level of the bars, so that they may be shielded as much as possible from the fire, while those which are required above the level of the bars should be kept as nearly as possible towards the crown of the furnace, so as to be removed from the immediate contact of the fire. Screw bolts with a fine thread tapped into the plate, and with a thin head upon the one side, and a thin nut made of a piece of boiler plate on the other, appear to be the best description of stay that has yet been contrived. The stays between the sides of the boiler shell or the bottom of the boiler and the top, present little difficulty in their application, and the chief thing that is to be attended to is to take care that there be plenty of them; but we may here remark that we think it an indispensable thing when there is any high pressure of steam to be employed, that the furnace crown be stayed to the top of the boiler. This it will be observed is done in the boilers of the Tagus and Infernal, constructed by Messrs. Miller, Ravenhill, and Co., and figured at pages 66. 71. and 72; and we knew of no better specimen of staying than is afforded by those boilers.

Should steam-boat boilers be of iron or copper? This is a question that has often been agitated, though we think it is practically set at rest by the almost universal preference of iron boilers. In the balances that have been struck between the expense of copper boilers on the one hand, and their durability on the other, we think the durability has in most instances been overstated. Copper boilers are very liable to damage in the furnaces from sulphurous coal, and in the flues from the action of the salt deposited by accidental leaks; and a leak in a copper boiler will not staunch itself by corrosion as in the case of iron, but will rather become worse. In some cases that have come under our notice, holes have been burned in the copper plates of the furnaces by the action of sulphurous coal; and as no other kind of coal was conveniently accessible, iron furnaces had to be introduced into the copper boilers. The danger from letting the water get low, moreover, is very great in the case of copper boilers, and will be almost sure to lead to an explosion if there be any considerable pressure. These considerations are in our eyes sufficient to justify the general preference of iron boilers over those of copper: but there is another point which must not be overlooked. Improvements in boilers are taking place every day, and a boiler that lasted very long would become antiquated before it could be worn out. The general introduction of tubular boilers, for example, within the last two years is one of those innovations to which more ancient methods must yield; but in cases in which copper boilers had been adopted such improvements could not be introduced without casting aside a boiler that still was sound.

With respect to the tubes and tube-plates of tubular boilers, we are of opinion that all things considered it is more expedient to make them of iron than of brass. The scale adheres to the iron with greater tenacity than to the brass, which is a disadvantage; but if the boilers be taken right care of no scale worth speaking of can accumulate on the tubes. The manufacture of iron tubes has been improved so much that there is not much difficulty in getting sound tubes of tough iron, the ends of which may be turned over in the ordinary method practised in the boilers of locomotives, and iron tubes have the great advantage that they will not be suddenly melted should the water within the boiler happen to get too low.

Explosions. This subject has been investigated with much care by the committee of the Franklin Institute, whose experiments on the strength of boilers we have already mentioned with commendation. We are unable, however, to follow these experimentalists in their researches, and have only room to remark that we believe most explosions will be found to have arisen either from an undue pressure of the steam, or from the overheating of the plates composing the boiler. The plates of the boiler may become overheated either in consequence of a want of water in the boiler, or from such a configuration of the internal parts of the boiler that the steam when formed cannot escape freely to the surface. The bottoms of large flues upon which the flame beats down are very liable to injury from this cause; and the iron in such a case will probably be softened by the heat, and in all probability will collapse upwards. Lightning, the sudden disengagement of large portions of scale, and other similar accidents have, we believe, caused explosions sometimes. But these causes are of very unfrequent occurrence in comparison with those that we have indicated,

and which would be oftener recognised as the real causes of explosions, were it not that people think they show their cleverness best by clearing up a difficulty with a new hypothesis that has been coined for the purpose in fancy's mint.

The plugs of fusible metal sometimes introduced into boilers to obviate explosions by melting out before the steam can reach any high temperature, are found in practice to be of but little avail. The compound metal is not homogeneous, and the more fusible of the metals is melted first, and is forced by the pressure of the steam out of the interstices of the less fusible metal, leaving its place to be supplied by the debris which all water supplies. The consequence is that the plug ceases to be fusible metal of the kind originally introduced, and cannot be melted by the steam even at a pressure and temperature much above that fixed as the requisite fusing point. Plugs of fusible metal should, therefore, we think, be discarded, as they are only calculated to mislead by pretending to do what they cannot accomplish. In tubular boilers, however, it is, we think, a good plan to introduce lead plugs in the tops of the fire-boxes-not with the idea that they will be melted by the steam where its pressure gets high, but to be melted out, and give notice of danger should the water fall too low.

Every boiler should be furnished with a steam gauge, which may give indication of danger should the pressure become too great, and the passages leading to the safety-valves should have no connection with the pipes leading to the stop-valves. In some cases stop-valves have been lifted from their seats, and forced into the mouth of the pipe, so that no steam could escape thereby; and in consequence of the safety-valve pipe springing from the pipe connecting the boilers, the boiler thus blocked up was in great danger of bursting, and would have burst if the fires had not been immediately drawn. In the case of any derangement of the safety-valve, or of the cone in the waste steam-pipe of a steam-vessel getting loose, and blocking up the mouth of the pipe, the pressure in the boiler may be eased by opening the blow-through valves of the engines, and the steam gauge will in all cases tell whether any undue pressure exists.

Priming. Priming arises from insufficient steam room, an inadequate area of water level, or the use of dirty water in the boiler: the last of these instigations may be remedied by the use of collecting vessels, but the other defects are only to be corrected either by a suitable enlargement of the boiler, or by increasing the pressure and working more expansively. Closing the throttle-valves of an engine partially will generally diminish the amount of priming, and opening the safety-valve suddenly will generally set it astir. A steam vessel coming from salt into fresh water is much more liable to prime than if she had remained in salt water, or never ventured out of fresh. This is to be accounted for by the higher heat at which salt water boils, so that casting fresh water among it is in some measure like casting water among molten metal, and the priming is in this case the effect of the rapid production of steam.

One of the best palliatives of priming appears to be the interposition of a perforated plate between the steam space and the water. The water appears to be broken up by dashing against a plate of this description, and the steam is liberated from its embrace. In cases in which an addition is made to a boiler or steam chest, it will be the best way not to cut out a large hole in the boiler shell for establishing a communication with the new chamber, but to bore a number of small holes for this purpose, so as to form a kind of sieve, through which a rush of water cannot ascend. Incrustation.-The incrustation of boilers by saline deposits was a much more important subject at one time than it is now, as nothing has been more clearly established, of late years, than that boilers may be preserved effectually from any injurious incrustation by abundant blowing off. Brine pumps are now in extensive use for withdrawing a certain quantity of water at every stroke of the engine; and the water so withdrawn has to pass through or among pipes carrying the feed-water to the boiler, so that some interchange of heat is there effected. These refrigerators, however, as they are grotesquely called, are in some respects bad things: the quantity of heat they save is, we believe, inappreciable; and the small pipes of which they are built up are liable to get choked, thereby endangering the boiler by the unconscious concentration of its contents. To guard against this danger, every engine fitted with brine pumps should be provided with a hydrometer for telling the specific gravity of the water in the boiler, so that the engineer may not be cheated by the defective action of the pumps, or suppose that they are operating when they are really inert. In the case of blowing out a boiler in the usual way, the engineer looks at his glass gauge tube, and keeps the blow-off cock open until the water-level has descended through the required distance, so that, under these circumstances, no doubt can arise that the boiler has been emptied of a certain quantity of water; but there is no such assurance in the case of the continuous extraction of the water, either by brine pumps or by a continuous blow off; and all boilers using either of these expedients should be fitted with hydrometer gauges as a precaution against the contents of the boiler being suffered to reach an injurious concentration. Numerous prescriptions have at various times been given as antidotes to incrustation; such as putting potatoes and other vegetable matters in the boiler, or in the case of a steam vessel, taking the feed-water from the bilge. The application of oil to the flues has also been recommended; and some boilers are fitted with a contrivance to inject oil into them just before the steam is let down. We look upon all such expedients, however, as needless, and are confident that boilers require no other preservation from incrustation than effectual