upon the periphery of the wheel; but the question was, "If the head be variable, what should be the diameter of the wheel to secure the best effect?" The paper showed that a wheel, whose diameter was equal to the total descent, when the head was a maximum, did not always give the greatest average labouring force. The question was therefore independent of the sort of shuttle used; it assumed the power of always admitting the water upon the wheel at the highest point of the total descent, and sought to establish the best relation between the diameter of the wheel and the whole descent, when the head alone was variable, according to given conditions. The results of this part of the investigation, therefore, while they admitted the full value of Mr. Rennie's shuttle, went further, and pointed out the limits of its useful application.

He was fully aware of the prejudice which existed against the circular conduit, and he once participated in it; but his attention had been forcibly drawn to it in his professional practice, and having used it very beneficially upon wheels of 40, 50, and 60 horses' power, which he had constructed for mining purposes, he wished to draw the attention of Engineers to the consideration of its practical merits when adapted to good wheels.

No. 558.

February 7, 1843.

The PRESIDENT in the Chair.

"Description of a Drawbridge, at Bowcombe Creek, near Kingsbridge, Devon."

By George Clarisse Dobson, Assoc. Inst. C E.

This Drawbridge spans one of five openings in a stone bridge, built across a navigable branch of Salcombe Harbour; it is in one leaf, 15 feet 9 inches wide, and 32 feet long, from out to out, supported upon a cast-iron shaft or axle, placed 7 feet 6 inches from the inner end, working in the abutment pier, which is built hollow to receive it, and thus the part within the axle-end acts as a counter-weight.

To the centre of the end cross-beam of the counter-part, a chain is attached, and after passing over cast-iron sheaves in the masonry of the face of the abutment, is coiled on a drum fixed on a horizontal shaft, carrying on one end a pinion, worked by a rack, attached to the piston of the hydraulic press; by this means,

motion is given to the shaft and drum, and consequently to the leaf of the bridge. Balance-boxes are hung to the counter-end, by which the shutting is regulated. The struts for supporting the leaf, when raised, are also thrown in and out of their places by a rack and pinion.

The hydraulic press used for opening and closing the bridge, is simple in its construction, and the whole works so easily, that a female can open and close the bridge in about 15 minutes, without difficulty. The fresh water used for the pump is contained in a cistern, beneath, and seldom wants replenishing, as it is returned into the reservoir every time after being used.

The bridge was designed and erected by Mr. J. M. Rendel, about 12 years since, when he was engaged in improving the turnpike road, in the south of Devon.

The expense of repairing, oiling, packing, &c., since its erection, has averaged under £7. per annum, including a small salary to a neighbouring millwright for occasional inspection.

The communication is accompanied by a drawing, showing a plan and sectional elevation of the bridge and the machinery.

No. 589. "An Investigation of the comparative loss by Friction, in beam and direct-action Steam Engines."

By William Pole, Assoc. Inst. C. E.

In consequence of the comparatively recent introduction of direct-action steam-engines on board the steam-vessels of the Royal Navy, the attention of engineers has been drawn to the advantages or disadvantages they possess, when viewed in comparison with those constructed with side levers. The object of this paper is to investigate the value of an apparently formidable objection which has been frequently urged against the direct-action engine, namely, "that from the more oblique action, consequent upon the shortness of the connecting rod, the loss by the increase of friction is so considerable as to constitute a serious objection to this form of engine."

After explaining to what extent mathematical analysis is applicable for determining the amount of friction, the paper proceeds to show that it may be satisfactorily used in the present case, as it is only the friction caused by the strain, or load, which is involved in the objection, and this is more adapted for theoretical than experimental determination.

The three general laws of friction, as established by the best experiments, are,—

1st. That the friction caused by one solid body rubbing upon another, is independent of the velocity with which the rubbing surface moves.

2nd. It is also independent of the area of the rubbing surface. 3rd. It is proportional to the pressure upon this surface.

From these it will follow, that if the pressure upon a moving body be multiplied by a certain co-efficient of friction (whose value is dependent upon the nature of the rubbing surface), the product will be the resistance from friction; and this multiplied again into any space the rubbing surface moves through, will give the amount of "power, work, or labouring force," expended in overcoming the friction through that space.

If the pressure upon the moving body be variable throughout its motion, the differential calculus must be employed, but the principle of calculation is still the same.

The paper proceeds to deduce general mathematical expressions for the amount of friction on each bearing of an engine, by finding, first, by ordinary statical rules, the pressure thrown on each particular bearing, by a given force applied to the piston, and then combining this with the space through which the rubbing surface moves.

This is done for the beam-engine, and for three modifications of the direct-action engine. Equations are also added for the oscillating or vibrating engine, and for an arrangement in which the connecting-rod is supposed to be indefinitely lengthened.

The numerical values of the expressions for friction thus found, are then calculated for an engine upon each of these different constructions, supposing them to be similar in all other respects, having the cylinders 66 inches in diameter, with a length of stroke of 6 feet; and the results are shown in a table, distinguishing the friction of every bearing.

From this it appears that as respects the friction caused by the strain, if the beam engine be taken as the standard of comparisonThe vibrating engine has a gain of 1.1 per cent.

This difference being so trifling, it is contended that the objection to the direct-action engine, on the ground of its alleged increased friction, has, when investigated, no adequate foundation.

Mr. Field believed that the paper was correct in its view of the comparative amount of friction of the two kinds of engines. He was of opinion that an excessive allowance for friction had hitherto been generally made in calculating their effective power. It was found practically, that when the pressure upon the piston was about 12 lbs. per square inch, the friction did not amount to more than 1 lb. or 11⁄2 lb. per square inch. This was easily ascertained by the indicator, when the engine was working without a load, but when loaded, he knew of no accurate experimental mode of showing it.

At the engines of the Blackwall Railway, the experiment had frequently been tried, by casting off all the load, and so regulating the steam, that the engines should make only the regular number of strokes per minute; the result had invariably shown about 1 lb. per square inch for friction.

Mr. Taylor confirmed the preceding remarks; it had been the custom formerly, in large pumping engines, to allow one-fifth for friction, but modern practice had shown that this was not necessary, particularly since greater precision had been introduced into the construction of all kinds of machinery.

Mr. Miller agreed that the friction of engines generally had been over-rated; he believed that as a simple comparison of the friction of the main parts of two kinds of engines, the results arrived at, in the paper, might be received as correct; but there were several other questions which must be considered, if it was intended to establish a general comparison between the beam and the direct-action engines; this, however, he believed was not the intention of the author.

Mr. Murray contended that the second proposition in the paper which assumed that " friction was independent of the area of the rubbing surface," although supported by Coulomb and the early experimenters, had been proved by Vince and others to be incorrect: it was natural to suppose that, in proportion to the hardness and smoothness of the bodies, there would exist a different ratio for the best proportion of surface to weight for every different body; if a surface carrying a given weight was of less than the

due area, the surfaces would cut into each other, become rough, and thus increase the friction: on the other hand, if the surfaces were unduly enlarged, there must be a loss from the additional amount of friction caused by the extended surface. He conceived that the calculations in the paper must be affected by the incorrectness of the data upon which they were based.

The simple mode of comparing the beam engine with the directaction engine appeared to be, to suppose two engines of the same length of stroke and diameter of cylinder; the proportions being good, it would be indifferent whether the power was transmitted through a direct connecting-rod, or through side levers; the cylinders, air-pump, arrangement of parallel motion, &c., being supposed to be alike, the friction of these parts would be alike in all cases, and the comparison would be limited to the parts employed in transmitting the power from the piston-rod cross-head to the crank-pin; both connecting-rods have the same number of bearings, which in both cases travel with friction over nearly the same distances: it is allowed that the bearings of the shorter connecting-rod have a larger amount of friction, and that from the greater angle it assumnes, more friction is thrown upon all the bearings of the parallel motion, on account of the greater force required to retain the piston in a vertical position. To counterbalance the increased friction on these parts of the direct-acting engine, allowance must be made in the beam engine, for the friction of the beam centres, and of the top and bottom necks of the side rods. The friction being directly as the distance moved through, and the distance in the side-rod ends being so very small, it follows that the amount of friction must be very trifling. The distance travelled by the beam centres is greater, but it is not of importance, as it is the angular distance due to the vibration of the beam, measured on the circumference of the gudgeon. Under these considerations Mr. Murray was disposed to give the preference (if any existed) to the side-lever engine.

In a pamphlet published in 1840, by Mr. John Seaward, it is stated that four-fifths of the whole friction of an engine were absorbed by the packings of the piston, and air-pump bucket, by the slide-valves and by the different packings or glands; consequently one-fifth was due to the whole of the necks or bearings throughout the engine. Now on considering the large proportion of this amount of the friction that is due to the bearings of the main shafts, of the crank-pin, and of the bottom end of the con