from the crank-shaft, at an angle of 45 degrees, to the sole-plate near the supports for the lower ends of the rockers.

The fly-wheel, 18 feet in diameter, running entirely above the floor, has a flat face 12 inches wide and 2 inches thick, with a central rib 5 inches wide, and is properly very light, as it has very little to do.

All parts of the steam-cylinders, covers, and valve-chests, are steam-jacketed; but live steam is not introduced into the pistons. Normal boiler-pressure is 125 pounds per square inch above the atmosphere, and expansion in both cylinders is about twenty-fold.

Water resulting from condensation in the jacket of the highpressure cylinder is returned directly to the boilers; and the drainage from the receiver, including that from the jacket of the low-pressure cylinder, is re-evaporated and superheated by passing through tubular re-heaters in the smoke-flues, and returned as superheated steam to the receiver. The valves and valve-gears are all of Mr. Corliss's well-known construction.

Steam is furnished by three of Mr. Corliss's patent upright tubular boilers, each 4 feet in diameter and 14 feet long, of the same form and arrangement as those which furnished steam for the Corliss engine at the Centennial Exhibition in 1876.

Each boiler has 560 square feet of heating-surface, and 19.63 square feet of fire-grate area. The escaping gases of combustion pass for a considerable distance through tubes surrounded by steam, and doubtless cause a good deal of superheating.

Two distinct tests of this engine were made by Mr. Walter H. Sears, chief engineer of the Pawtucket Water-works, and Mr. Isaac R. Scott, president of the water board of Waltham, Mass.; one of 10 hours per day during two weeks, Tuesday, Aug. 6, to Monday, Aug. 19, 1878, excluding the two intervening Sundays, Aug. 11 and

18.

The total run was therefore 120 hours, in 12 periods of 10 hours each, with 9 intervals of 14 hours each, and 2 of 36 hours each. The wood used for kindling fires, 2,126.5 pounds (89 pounds each day to each boiler), was estimated as coal at 45 per cent of its weight, equal to 957 pounds of coal, which added to 6,521 pounds of coal used in starting fires (271 pounds per day, per boiler) makes 7,478 pounds of coal, equal to 311 pounds of coal per day for each boiler, principally consumed by reason of the intervals when the engine was at rest.

The coal used for pumping was 26,598.5 pounds; and all the coal used for starting fires and pumping, including the estimated coal

value of the wood, was 34,076.5 pounds, of which 22 per cent was used in starting fires. (Only two boilers were used in this test.)

The duty, calculated upon the capacity of the pumps, the mean resistance, and the whole number of pump-strokes for each 100 pounds of coal consumed, without deduction for ashes and residue, was 104,357,654 foot-pounds.

A weir measurement, made Aug. 28, 1878, revealed a loss of action of 4.12 per cent, which was subsequently much reduced by a change of valves.

The other test, of 24 consecutive hours' duration, begun Oct. 3, 1878, at 9 o'clock A.M., gave a duty, calculated on the capacity, mean resistance, and number of strokes of the pump, of 133,522,060 foot-pounds for each 100 pounds of coal burned while pumping, without deduction for ashes or residue. In this test, all three boilers were used.

These results, which I see no reason to question, are the more remarkable when the small size of the engine, and its low speed, are taken into consideration. The space swept through by the piston at each revolution, or double stroke, is, for the high-pressure cylinder, only 6 cubic feet, for both cylinders, only 30 cubic feet; and at the normal speed, 52 revolutions per minute (the actual speed at the tests was 51.8), the velocity of the piston was only 260 feet per minute.

Sets of diagrams from these cylinders show close conformity to normal action of the highly expanded steam.

Confirmation of the accuracy of the results above given is to be found in the fourth annual report of the water commissioners and superintendent of water-works of the town of Pawtucket, dated Feb. 1, 1884. A table is there given in detail of the performance of this engine for each of the twelve months ending Dec. 31, 1883, of which the sums and means for the year are subjoined.

Number of days when pumping was done, per month, 23.1

Total number of hours' pumping

Hours and minutes per month

Hours and minutes per day
Total number of revolutions
Revolutions per month

277.2

.3,852 h., 49 m.

321 h., 4 m. 13 h., 54 m. 10,698,618

891,551

Consumed when not pumping.

Consumed in pumping

Total coal consumed

Total number of gallons pumped in 12 months.

Mean head in feet against the pump, no allowance for frictions in suction

273,041

660,865

933,906

441,061,225

264.19

Mean duty in pounds of water raised 1 foot per 100 pounds of coal used for all purposes, no deduction for ashes and residue, 106,203,760 Mean duty in pounds of water raised 1 foot per 100 pounds of coal used for all purposes except warming, no deduction for ashes and residue.

108,402,320

HORIZONTAL COMPOUND ENGINE.

In the Nourse Mill of the Social Manufacturing Company, Woonsocket, R.I.

This engine, designed by Mr. George H. Corliss, and built by the Corliss Steam Engine Company under a guaranty to run a year with a consumption of no more than 1.75 pounds of coal per horse-power and per hour, all the coal fired during the year for supplying this engine with steam being divided by the mean indicated horse-power and by the number of hours' run of the engine, — is a fine example of its class.

It is a tandem, receiver, double engine, combining the effort of the two pair of pistons upon a single fly-wheel shaft with cranks set "quartering." The cylinder diameters are respectively 20 and 36 inches. The length of stroke common to both is 6 feet, and there is a space of about 4 feet between the end-casings of the two cylin ders, to give access to the low-pressure piston.

The connecting-rod is 18 feet in length, so that the entire engine, from the centre of crank-shaft to the outer end of the high-pressure cylinder, occupies a length of about 42 feet. The fly-wheel is 30 feet in diameter, and has a rim 110 inches wide with four belt-faces. The weight of the fly-wheel, crank-shaft, cranks, and eccentrics, is about 54.5 tons, say 120,000 pounds.

The normal speed is 57 revolutions per minute; so that the velocity of the pistons is 684 feet per minute, and that of the belt 92 feet more than a mile per minute.

The weight of the reciprocating parts is 7,711 pounds, and at normal speed their momentum is equal to 19.38 pounds pressure per square inch on the area of both pistons.

The normal boiler-pressure is 110 pounds per square inch above the atmosphere; and about 8 pounds pressure above atmosphere is carried in the receiver.

Both cylinders are steam-jacketed all around their sides and ends, -the high-pressure cylinder directly from the boiler, the low-pressure from the receiver; and in both, steam is admitted by the induction-valves to the cylinders from the jackets.

The cylinders are supported by hollow legs at each end, which mask the descending exhaust-pipes. Each cylinder has a projecting zone around it in the middle of its length, about 14 inches, lengthwise of the cylinder, projecting about 7.5 inches all around beyond the jacket-casing, which serves the double purpose of providing for unequal expansion, and of taking in steam and drawing out water conveniently.

Steam from the boilers is admitted to the jacket of the highpressure cylinder at the top; and water formed by the condensation of steam in this jacket is drawn out at the bottom, and pumped directly back to the boilers.

It is obvious, from the construction and setting of the boilers (hereafter described), that the steam, even on entering the jacket, may be to some extent superheated: but a considerable quantity must be condensed in this jacket; since the expansion in this cylinder is about 5-fold, and the mean internal temperature necessarily considerably below that of initial steam.

The exhaust-pipes from the high-pressure cylinder, with a pipe about 12 inches in diameter which receives the steam from these pipes, and conducts it to the middle of the length of the low-pressure cylinder, where it enters the jacket of this cylinder at the bottom, and this jacket itself, constitute the receiver, in which the pressure varies but slightly.

The expansion in the low-pressure cylinder is almost 4-fold, so that the whole expansion in both cylinders is 18 to 20 fold. Steam for warming the mill, for water-pots in the weaving-room, and for all other purposes except for the slashers (for which steam is taken directly from the boilers), is drawn from the receiver, so that less steam enters the second cylinder than is discharged from the first: as it is, however, the effective power developed in the two cylinders is very nearly equal.

The drainage from the receiver, which, it will be remembered, includes the water formed by condensation in the jacket of the lowpressure cylinder, is pumped through a set of tubular heaters at the top of the descending flues from the boilers, where the escaping gases are hottest; and returns to the receiver as steam, probably considerably superheated.

Similar tubular heaters, located immediately under those last

mentioned, receive the feed-water on its way to the boilers, and raise its temperature, probably to that corresponding to boiler-pressure. Feed-water is supplied in part by drainage from the mill, and, for the rest, from the hot-well.

At present, one engine only is in place, and that is exerting only about 400 indicated horse-power (four-fifths of its normal load); but the demand for power is gradually increasing, as new machinery is started. Of the nine boilers ultimately to be erected, -eight for regular use, and one to be held in reserve, only four are now set up. These are vertical, 64 inches in diameter, and 14 feet in height. The hot gases of combustion fill a space around the shell of the boiler, between it and the brickwork, for a height of 5 feet, but without direct outlet. They pass off through 84 flues, 3 inches in external diameter and 14 feet long, into a smoke-box on top of the boiler, whence they flow rearward to, and downward through, first the re-heater for the drainage of the receiver, then the feed-water heater, and finally down through a vertical brick flue to a horizontal flue underground, and so off to the chimney.

The brick-work around the cylindrical shell of the boiler is incased in plate-iron for a height of 10 feet 6 inches from the floor, and by a cast-iron cornice for another foot, which cuts off infiltration of air; and the shell of the boiler above the cornice is plastered with some heat-intercepting material.

The fire-grates, 76 inches in diameter, have an area, in the aggregate, for the four boilers, of 126 square feet; and at the rate of 1.75 pounds of coal per indicated horse-power per hour, to supply steam for 500 horse-power will require a rate of combustion equal to 7.0 pounds per square foot of fire-grate area and per hour.

The ratio of water-heating surface to fire-grate area is about 24 to 1; that of steam-heating surface to fire-grate area, about 9.3 to 1; and the total, about 33.3 to 1: but the relative proportions of water and steam-heating surface will depend on the height at which water is carried. The gases evidently go to the re-heaters very hot, and may not be very effectually cooled on passing to the chimney.

A set of diagrams was taken from this engine by the writer, May 22, 1884; and a set taken April 9, by Mr. John T. Henthorne of the Corliss Steam-Engine Company, has since been furnished by him. When the first set was taken, no steam was diverted at the receiver from the low-pressure cylinder. A little was so diverted when the second set was taken, May 22.