any appreciable effect. It must be remembered, however, that the variable twisting strains on the crank-shaft are reflected to a greater or less extent throughout all the working parts of the engine; and it seems certain that the beneficial influence of three cranks is far beyond what is generally accepted. There seems to be an impression that the length of an engine-room is increased by the introduction of three cranks. But by placing the valve casings at the side, thus allowing the intermediate and low-pressure cylinder covers to be brought close together, the total length of a three-crank engine is made no greater than in nine-tenths of the existing compounds. In the Orient liner Lusitania, since she has been converted to three cranks by the writer, the distance longitudinally over the cylinders is one foot shorter than before, and the indicated horse power is 800 greater.

Valve Gear. The arrangements for working valves are becoming very extensive, and a review of the various devices which are employed to distribute the steam in modern engines would by itself occupy a very lengthy paper. The requirements of a good valve gear are that it shall give at both ends of the cylinder an equal distribution of steam at all grades of expansion, with a minimum of working parts and with no undue strains. The four principal methods adopted to work valves are: first, by the single

eccentric; secondly, by the double eccentric; thirdly, by taking the motion from the connecting rod; and fourthly, by a compound motion derived from both the piston rod and the connecting rod. All four of these have their advantages and defects, and vary considerably in complexity and in multiplicity of parts. The single eccentric valve gear, giving almost perfect steam distribution, having few working parts, and being independent of the connecting rod or piston rod, seems eminently suitable to fulfill all the desired conditions. A sliding block gear of this kind has given every satisfaction in actual working; and the arrangement is such as to allow free access to all the working parts, while occupying a minimum of space. The end of the eccentric rod reciprocates on guide bars, which are inclined at various angles to suit the desired action of the valve. Motion is imparted to the valve spindle from an intermediate joint in the eccentric rod, which moves in an approximately elliptic path. A slot in the tail of the guide bars admits of varying their inclination to different extents for a given movement of the reversing engine, thus altering independently the cut-off in each cylinder. The long leverage and easy motion reduce the wear and tear to a minimum, as results in actual practice have proved. An objectionable feature, however, in an engine fitted with this sliding block gear is that the

valves are at the front, over the starting platform; and the exhaust has to be led by a belt around the low pressure cylinder to the condenser. To overcome this objection of the sliding block gear, a swinging link gear was designed. The eccentric rod, as in the last case, is placed diagonally over the condenser, but is here guided in an arc of a circle, by suspending it by a swinging link, centered on a pin, which pin is adjustable by the reversing engine into various positions for varying the grade of expansion either ahead or astern.

The movements for working the valve are transmitted from a joint at the end of the eccentric rod by a compensating link connecting the joint with one arm of an oblique lever, of which the other arm is jointed to the valve spindle. The compensating link is an essential and distinguishing feature of this gear; it is so placed and proportioned relatively to the other parts as to produce practically equal port opening and cut-off at each end of the stroke. There is a quick and a slow movement of the valve at each end of its travel; the slow movement being at the maximum port opening, and the quick movement at the cut-off. The lead is also constant at all grades of expansion.

Practical Results.-In the engines of the screw-steamer Para, belonging to Messrs. Steel, Young & Co., which made her maiden voyage

to the River Plata about 31⁄2 years ago, the cyl

and 53 inches in These being the engines, special ascertaining the

inders are 19 inches, 35 inches, diameter, with 33-inch stroke. writer's first triple-expansion arrangements were made for actual coal consumption per indicated horsepower. Although, in order to get a correct estimate of an engine's performance, the quantity and temperature of the feed and circulating water ought of course to be considered, yet there are so many practical difficulties in the way of getting these particulars on board ship, that it has been found impossible to obtain any reliable data on these points. This steamer still continues on the same run, averaging 9 knots an hour on 10 tons of coal; and has not yet cost anything beyond the usual overhaul for repairs. The great saving in coal consumption with the triple engines is apparent when comparison is made with two sister ships fitted with compound engines, the Ingram and the Wandle, belonging to the same company, and built by the same builders; the comparison is shown in Table I., which gives the average working over a period of three years.

TABLE I.

Comparative Results from Three Similar Steamers with Compound and with Triple Expansion Engines.

The diagrams are an expanded set taken during the three days' trial of the Para; but owing to the intermediate cylinder being rather too large, the equalities of temperatures and of initial stresses are disturbed; and the drop of the steam pressure in the intermediate cylinder is excessive, in consequence of the steam velocity being too great.

In Table II. is given a comparative statement of results from an approximately similar trio of boats in the same trade and under the same management, on a round voyage to Java under average conditions. The triple engines in the Jacatra are of the same general design as those in the Para, but of greater power; the compounds in the two other boats are of the ordinary type.