ment was found to be attended with a bad draft in the furnaces, and therefore to require a greater quantity of heating surface to produce the necessary evaporation. This entailed upon the machinery the occupation of more space in the vessel in proportion to its power; it has therefore been more recently the practice to give a separate system of flues to each pair of furnaces, or, at most, to every three furnaces. When three furnaces communicate with a common flue, two will always be in operation, while the third is being cleared out; but if the same quantity of fire were divided among two furnaces, then the clearing out of one would throw out of operation half the entire quantity of fire, and during the process the evaporation would be injuriously diminished. It is found by experience, that the side plates of furnaces are liable to more rapid destruction than their roofs, owing, probably, to a greater liability to deposit. Furnaces, therefore, should not be made narrower than a certain limit. Great depth from front to back is also attended with practical inconvenience, as it renders firing tools of considerable length, and a corresponding extent of stoking room necessary. It is recommended, by those who have had much practical experience in steam-vessels, that furnaces six feet in depth from front to back should not be less than three feet in width, to afford means of firing with as little injury to the side plates as possible, and of keeping the fires in the condition necessary for the production of the greatest effect. The tops of the furnaces almost never decay, and seldom are subject to an alteration of figure, unless the level of the water be allowed to fall below them.*

(217.) A form of marine engine was some years since proposed and patented by Mr. Thomas Howard, possessing much novelty and ingenuity, and having pretensions to a very extraordinary economy of fuel, in addition to the advantages claimed by Mr. Hall. In Mr. Howard's engines, the steam, as in Mr. Hall's, is constantly reproduced from the same water, so that pure or distilled water may be used; but Mr. Howard dispenses altogether with the use of a boiler.

A quantity of mercury is placed in a shallow wrought-iron vessel over a coke fire, by which it is maintained at a tem* Tredgold on the Steam Engine, Appendix, I. p. 171.

The surface exposed

perature varying from 400° to 500°. to the fire was computed at three fourths of a square foot for each horse-power. The upper surface of the mercury was covered by a very thin plate of iron in contact with it, and so contrived as to present about four times as much surface as that exposed beneath the fire. Adjacent to this a vessel of water was placed, maintained nearly at the boiling point, and communicating by a nozzle and valve with the chamber immediately above the mercury. At intervals corresponding to the motion of the piston a small quantity of water was injected from this vessel, and thrown upon the plate of iron resting upon the hot mercury. From this it received not only the heat necessary to convert it into common steam, but to give it the qualities of highly superheated steam. In fact, the steam thus produced had a temperature considerably above that which corresponded to its pressure, and was, therefore, capable of being deprived of more or less of its heat without being condensed. (94.) The quantity of water injected into the steam-chamber was regulated by the power at which the engine was intended to be worked. The fire was supplied with air by a blower subject to exact regulation. The steam thus produced was conducted to a chamber surrounding the working cylinder, and this chamber itself was enclosed by another space through which the air from the furnace passed before it reached the flue. By this contrivance the air imparted its redundant heat to the steam, as the latter passed to the cylinder, and raised its temperature to about 400°, the pressure, however, not exceeding 25 lbs. per square inch. The valves, governing the admission of steam to the piston, were adapted for expansive action.

The vacuum on the opposite side was maintained by condensation in the following manner: - The condenser was a copper vessel placed in a cistern of cold water, and the steam was admitted to it from the cylinder by an eduction pipe in the usual way. A jet was introduced from an adjacent vessel filled with distilled water, and the condensing water and condensed steam were pumped from the condenser as in common engines. The warm water thus pumped out of the

H H

condenser was drawn through a copper worm, carried with many coils through a cistern of cold water, so that when it arrived at the end of this pipe it was reduced nearly to the temperature of the atmosphere. The pipe was thus brought to the vessel of distilled water already mentioned, and the water supplied by it replaced. The water admitted to the condenser through the condensing jet being purged of air, a small air-pump was sufficient, since it had only to exhaust the condenser and tubes at starting, and to remove the air which might be admitted by leakage. Mr. Howard stated that the condensation took place as rapidly and perfectly as in the best engines of the common kind.

An engine of this construction was in the spring of 1835 placed in the government steamer called the Comet. It was stated, that though the machinery was not advantageously constructed, a part of the engine being old, and not made expressly for a boiler of this kind, the vessel performed a voyage from Falmouth to Lisbon, in which the consumption of fuel did not exceed a third of her former consumption when worked by Boulton and Watt's engines, the former consumption of coals being about eight hundred pounds per hour, and the consumption of Mr. Howard's engine being less than two hundred and fifty pounds of coke per hour.

The advantages claimed for this contrivance were the following: first, the small space and weight occupied by the machinery, arising from the absence of a boiler; second, the diminished consumption of fuel; third, the reduced size of the flues; fourth, the removal of the injurious effects arising from deposit and incrustation; fifth, the absence of smoke.

(218.) The method by which the greatest quantity of practical effect can be obtained from a given quantity of fuel must, however, mainly depend on the extended application of the expansive principle. This has been the means by which an extraordinary amount of duty has been obtained from the Cornish engines. The difficulty of the application of this principle in marine engines has arisen from the objections entertained in Europe to the use of steam of high pressure under the circumstances in which the engine must be worked at sea. To apply the expansive principle, it is necessary that the moving power at the commencement of the stroke shall considerably exceed the

resistance, its force being gradually attenuated till the completion of the stroke, when it will at length become less than the resistance. This condition may, however, be attained with steam of limited pressure, if the engine be constructed with a sufficient quantity of piston-surface. This method of rendering the expansive principle available at sea, and compatible with low-pressure steam, has recently been brought into operation by Messrs. Maudslay and Field. Their improvement consists in adapting two steam-cylinders in one engine, in such a manner that the steam shall act simultaneously on both pistons, causing them to ascend and descend together. The piston-rods are both attached to the same horizontal cross-head, whereby their combined action is applied to one crank by means of a connecting rod placed between the pistons.

A section of such an engine, made by a plane passing through the two piston-rods P P and cylinders, is represented in fig. 127. The piston-rods are attached to a cross-head c,

which ascends and descends with them. This cross-head drives upwards and downwards an axle D, to which the lower end of the connecting rod E is attached. The other end of the connecting rod drives the crank-pin F, and imparts revolution to the paddle-shaft G. A rod H conveys motion by means of a beam I to the rod K of the air-pump E.

(219.) Connected with this, and in the same patent, another improvement is included, consisting of the application of a hollow wrought-iron framing carried across the vessel above the machinery, to support the whole of the bearings of the crankshaft. A plan of this, including the cylinders and paddlewheel, is represented in fig.128. The advantages proposed by these improvements are simplicity of construction, more direct action on the crank, economy of space and weight of material, combined with increased area of the piston, whereby a given evaporating power of the boiler is rendered productive, by extended application of the expansive principle, of a greater moving power than in former arrangements. Consequently, under like circumstances, greater power and economy of fuel is obtained, with the further advantage at sea, that when the engine is reduced in its speed, either by the vessel being deeply laden with coal, as is the case at the commencement of a long sea voyage, or by head winds, more steam may be given to the cylinders, and consequently more speed imparted to the vessel, all the steam produced in the boiler being usefully employed.

(220.) Another improvement, having the same objects, and analogous to the preceding, has been likewise patented by Messrs. Maudslay and Field. This consists in the adoption of a cylinder of greater diameter, having two piston-rods P P′, as represented in fig.129., of considerable length, connected at the top by a cross-head c. From this cross-head is carried downwards the connecting rod D, which drives the crank-pin E, and thereby works the paddle-shaft s. In this case the paddle-shaft is extended immediately above the piston, and the double piston-rod has sufficient length to be above the paddle-shaft when the piston is at the bottom of its stroke This improvement is intended to be applied more particularly for engines for river navigation, the advantages resulting from