In fig. 7, air is admitted from below, and rises through the orifice at A, concurrently with the gas at the orifice B. On being ignited, one long flame is produced, of a dark colour, and ending in a smoky top.

In fig. 8, the air is introduced from below, and into the chamber C, C, from which it issues through a perforated plate, like the rose of a watering pot; thus producing immediate mixture with the gas. On being ignited, a short, clear, and brilliant flame was produced, as in the ordinary Argand gas burner.

The heating powers of the flames were then tested, by placing a vessel of cold water over each. When over fig. 7, it required 14 minutes to raise the water to 200°, whereas, over fig. 8, it reached 200° in 9 minutes.

Now, the difference of effect produced in those three experiments corresponds with what takes place in furnaces and their flues, when the air is excluded, and when it is admitted through a single or through numerous orifices.

Of the importance of mechanical agency, in promoting the rapid diffusion or mixture of the air and the gas, the modes adopted on the continent for rendering the coke gas, or carbonic oxide, available, are conclusive and instructive.

M. Peclet has given ample details of the mode of effecting the combustion of this gas (the existence of which has, for a long time, been practically ignored in this country), in the manufacture of iron, and even in the puddling furnaces, where the most intense heat is required.

M. Peclet states that the process at Treveray, in France, (see figs.

FIG. 8.

9 and 10) is preferable to that adopted in Germany, and for the following reasons, which are quite to the point of our present inquiry.

First, the air and the gas are better incorporated; secondly, the relative quantities of the gas being brought into contact with the air are more easily regulated; thirdly, combustion is effected by the introduction of the smallest excess of air.

In the apparatus, as shown in the section, fig. 10, fifty jets of air issue, each in the centre of fifty jets of the gas (carbonic oxide), led from the cupolas of the melting furnaces. On examination of the process here exhibited, the mixing and combustion, it will be seen, takes place on the instant, and before the flame and heat enter the chamber of the furnace at F. By this arrangement, M. Peclet observes, 'that the highest temperature that the arts can require is here obtained.' It is strange that the practical and commercial value of this gas, which is so wastefully expended at our manufactories, at the summit of the cupolas, but so well understood, and economised in France and Germany, is only just now being recognised in this country."

Another experiment on the operation of the jet principle is illustrated in the following extract:

"The main object being the introducing the air in a divided state to the gaseous atmosphere of the furnace chamber, the following experiment was made:--The centre bar of a boiler, 4 feet long was taken out, and over the vacant space an iron plate was introduced, bent in the form as shown in fig. 11.

Here, the upper portion of the bent plate, projecting 3 inches above the fuel, was punched with five rows of half-inch holes, through which the air issued in fifty-six streams. Adequate mixture was thus instantly obtained, as in the Argand gas burner; the appearance as viewed through the sight-holes at the end of the boilers, being even brilliant, and as if streams of flame instead of streams of air, had issued from the numerous orifices. It is needless to add, that nowhere could a cooling effect be produced, notwithstanding the great volume of air so introduced.

This led to the enlarging the door-end of the furnaces sufficiently to admit the required number of apertures and full supply of air; an arrangement which has been for years in successful operation, both in marine and land boilers.

In practice, the great difficulty lay in adapting the plan to marine boilers, the doorways of which are made so contracted, as to render it impossible to introduce the required number of half-inch orifices, as hereafter will be shown."

In the course of some investigations on the principles laid down by Sir Humphrey Davy, and their application to the tubular boiler, Mr. Williams observes:

"On examination of what passes in furnaces using coal, we see the direct connection between its effect, and what Sir H. Davy so clearly points out as the means of extinguishing the flame. On looking into a flue boiler from the back end, a body of flame will be seen flashing along from the bridge, and if air be properly introduced, extending a distance of 20 to 30 feet. This is the appearance which has to be sustained until the process of combustion be completed, if we would have the full measure of heat developed.

On the other hand, looking into a tubular boiler, across the smokebox, the light of the flame through the

may be seen t

tubes; but on entering

as shown in fig. 12.

The distance, then, to which flame will penetrate tubes, before being extin guished, will depend on the rapidity of the current, the size of the orifices, and the quantity and character of the gaseous products, entering in company and contact with it. These products are, from

Carbonic Acid
and Nitrogen.
Carbonic Acid,
Nitrogen, and

Steam.
Here we have the very
incombustible gases re-
ferred to by Sir H. Davy,
-not even in small, but
in very large quantities,
-forced into the most
intimate possible mixture
with the flame. The
result necessarily must be,
the reduction of its tem-
perature, and consequent
extinguishment.

Impressed with the importance of the connec tion between temperature and ignition, Sir H. Davy dwells on the fact, and repeats, that flame, whether produced from the combustion of large or small quantities of explosive mixture (gas and air), may always be extinguished or destroyed by cooling agencies; and in proportion to the heat required to carry on com bustion, so it is the more easily destroyed."

Mr. Williams does not apply these remarks to those locomotive boilers where coke alone is used. He particularly discriminates between the various constructions of engine, and the various kinds of fuel employed. He says:

"In speaking of the evils of the tubular system, these remarks have no reference to its appli cation in locomotive boilers, where coke alone is used, and for this selfevident reason, that no hydrogen-carbon gas or fuliginous flame has there to be encountered. In the tubes of the locomotive there is, in fact, no chemical or practical reason why the heat may not be abstracted from the po products with the greatest rapidity.

With reference to the size and sectional area of lethe tubes, we even deceive ourselves as to what frefer that sarea practically is. now When, for instance, it is Reilly intended to give an area Bluus 07 304 of 2 or 3 inches, it will rismu to nola not be sufficient merely to provide tubes of 2 or 3

inches diameter. Allowance must be made for the effect produced by the rapidity or pressure of the body (whether it be flame, gas, air, or water), forcing an entrance into a restricted orifice. The true and available area, that which directly influences the mixing operation of the several bodies simultaneously forcing themselves through the orifices, will then be, not as at A, in fig. 13, but as at A'. Thus, in effect, a tube of 3 inches of internal area will practically have one of but 24 or 2 inches, and so in proportion to the size of the tube, and the amount of pressure exercised by the current. So little, however, has this been considered, that not unfrequently the ferrule, or iron ring by which the tubes are fastened to the face-plate, is inconsiderately placed within their orifices; thus still more seriously contracting the working area."

A very important subject and one almost distinct, is that referring to the circulation of water in the boiler, and which as it is treated by Mr. Williams at length, cannot fairly be comprised in the brief space we could now give to it. There is likewise a copious discussion of the merits of the tubular system,

Fig. 13.

and Mr. Williams objects to its application for marine purposes He likewise enters on an investigation of the doctrines as to the value of heated air for feeding furnaces of boilers, and here again he is found in the opponent ranks, and is consequently in direct hostility to the patentees of the numerous inventions for applying heated air.

One ground of objection taken by Mr. Williams to the tubular system is grounded on the generation of a large quantity of water in furnaces in which coal gas is produced and consumed, which being in the form of steam becomes the largest product of that combustion. This results from bituminous coal containing from 5 to 6 per cent. of hydrogen.

The smoke question is only incidentally and collaterally discussed, because Mr. Williams considers that instead of attention being directed to contrivances for consuming smoke, the chief object is to obtain perfect combustion, when no smoke will be produced.

There is scarcely a portion of a subject so important and so difficult, and we may add so perplexing, to the practical man, that Mr. Williams has not elaborately and carefully treated. That of DRAUGHT is one essential to the well-working of the steam engine, of whatever class, and we are tempted to lay before our readers the views of Mr. Williams. This we shall do at some length, though were it not for the well-known zeal and liberality of Mr. Williams on a favourite subject, we should feel some delicacy in extracting so copiously from a work on which much labour has been bestowed, and the circulation of which at any rate interests the reputation of the author, even where it is not one of profit. We must, however, trust to his motives and our own to obtain his excuse.

In treating this branch of the discussion, Mr. Williams has frequently availed himself of the labours of M. Peclet, an authority to whom we have had many occasions to refer in the pages of this Journal. Mr. Williams observes:

"The draught, or current of air passing through a furnace, is occasioned by the difference in weight between the column of air within the chimney, and that of an external column of the same proportions,-the 'ascent of the internal heated air,' as Dr. Ure observes, depending on the diminution of its specific gravity,-the amount of unbalanced weight being the effective cause of the draught.' Since, then, this levity of the inside air is the result of increased temperature, the question here for consideration is, how that temperature may be obtained with the least expenditure of fuel.

In marine boilers numerous cases of deficiency of draught will be found to arise from an injudicious arrangement of the flues, and the conflicting currents of the heated products within them.

Notwithstanding the importance of the subject, still but little attention has been given to the causes of these currents and deficient draught. M. Peclet having examined the subject with great care and practical research, his details are so copious, and his remarks so much to the point, that it will be well to give them due attention, and the more so as the subject has not hitherto been examined by any writer in England. Among the inconveniences experienced, a prominent one may here be mentioned, as being of frequent occurrence, namely, a deficient draught in the side or wing furnaces of boilers.

M. Peclet observes, 'Where several tubes or flues open into one common flue, the currents are continued beyond their orifices, and by their mutual action, affect

or modify their respective forces. If, for example, two flues A and B (see fig. 14) enter the common flue C, by orifices opposite each other, the influence of their currents on each other will be nil, if they have equal rapidity; be- Al cause the whole will pass as if they had struck against a plane fixed between them. If, however, the currents be unequal, that which has the greatest rapidity will reduce the speed of the other, and more or less have the effect of closing the orifice through which the latter flowed.'

'So many proofs of this,' he adds, 'may be adduced, as to put the fact beyond doubt.' "These streams of air,' he continues 'in this respect, act on each other as streams of water. It is already known by the experiments of Savart, that where two streams of water of the same sectional area,

C

FIG. 14.

FIG. 15.

act in opposite directions, and that one of them has even but little more speed than the other, the latter is pushed back, and the influence felt up to its source. The result of this collision in the flue may be avoided by the diaphragm D (fig. 15).' Such conflicting currents may be found in almost all marine boilers, yet pass unnoticed, even where the draught is manifestly deficient in consequence.

Again, 'phenomena of the same kind will be produced where the courses of two flues are at right angles to each other, as in fig. 16. These effects may also be avoided by the diaphragm D (fig. 17).' This also is of frequent occurrence, and seriously affects the general draught, as will hereafter be shown.

Again, 'Where the chimney or flue is common to several furnaces,

'So, where a current of hot products issues horizontally into a chimney, it may happen that its draught would be entirely destroyed, if the rapidity of such current was considerable, as it would then have the effect of shutting the chimney like a damper.' He then describes what occurred at a soda manufactory, with a chimney for general draught, which he had to construct, and which was also connected with a flue from another apparatus, as shown in fig. 19. In this case, 'the currrent from the one flue completely neutralised that of the other.' This he remedied by the partition P.

FIG. 19.

Again, where three flues enter a funnel from three different points, it is evident, he observes, that 'the diaphragms should be so placed as to leave each current an adequate section of the chimney,' as in fig. 20. The circumstance here referred to may be found to exist in almost all marine boilers. Rarely, however, is the interposition of these diaphragms thought of, yet numerous instances of the derangement of the draught, particularly of the wing boilers, must be within the knowledge of all engineers.

Let us now apply these judicious practical observations. The first boilers of the Great Britain, screw steamer, are in point. The arrangements of these boilers have already been noticed with reference to their

impeding the due circulation of the water. We have now to consider them in respect to their influence on the draught.

FIG. 20.

In these boilers, attention was given, almost exclusively, to two objects:-providing the largest possible amount of fire grate areas, and the largest aggregate of internal heating surface. As to the former, almost the entire area of these large boilers may be compared to an aggregate of furnaces. Nevertheless, there was no command of steam, and the engineer stated, that the wing boilers were unequal in draught to the centre ones. The deficiency of draught in the furnaces of the side boilers will easily be accounted for on examining the plan of the upper tier of flues, and the numerous collisions where the heated products from twenty-four large furnaces entered the funnels, as shown in fig. 21.

The flues from the four furnaces of each wing boiler, are here made to enter one common cross flue each thwarting the current of the preceding one. No. 1 being checked by No. 2-which crosses it

at right angles-which, in its turn, was checked by No. 3, and so onthe same mal-arrangement taking place in each of the sixteen flues of the four wing boilers. These, it will be seen, are the direct cases adduced by M. Peclet, where the products and current from one flue act as a damper on the draught of its preceding

one.

Again, the joint products of the four flues of each wing boiler are made to enter the funnel by a single opening, which is not only at right angles with the flues from the four centre boilers, but directly opposite to those of the wing boilers on the other side. Thus the flues of no less than eight furnaces, all entering by a single opening are brought into direct collision with those of the other four, and in the most certain way to affect the draught of all. Here we have a combination of the evils referred to by M. Peclet.

The case of the boilers of the Great Liverpool, is a still greater violation of the rules which should regulate the draught. Here, there being but a single tier of flues, the required

aggregate of heating internal surface was ob tained by the labyrinth of windings shown in fig. 22.

In the first place, the flames from the three centre furnaces of each half of the boiler are forced into a single flue of but 13 inches wide, as shown by the arrows. Again, the gaseous products of each set of three centre furnaces, and which are necessarily the more powerful, are made to enter the single back flue at right angles, and across the current of products from the two wing furnaces, as shown by the enlarged view in fig. 23. It is scarcely possible to conceive a more direct case of collision, or a more effectual damper by the hotter and larger current from three furnaces, on the smaller current from the two wing furnaces. In these boilers, it is manifest that nine-tenths of the steam was produced by the plates in connection with the furnaces alone, and by a system of continued forcing; the long run of flues being filled with dense black smoke. A considerable improvement was affected by constructing furnaces in pairs, as in fig. 24. This had the