pressure cylinder marked B, if we supposed the diameter of the latter cylinder to be 36 inches and the stroke feet, or one-fourth of that which it really is.

This result is set out in Fig. 6

The diagram marked A is reversed in position and repeated on the right-hand side by measuring off a series of horizontal lines, such as c d, and making c d equal 4 CD in every case.

30

15

115

FIG. 6.

In this way the upper shaded area represents the work done in the high-pressure cylinder as

it would appear on the scale adopted in the lowpressure cylinder. The bottom shaded area is merely a repetition of the area B.

When the two diagrams are put together, it will be seen that the two portions of the expansion curves fit very fairly or run into one, and that the expansion commenced above d is carried on throughout the stroke.

It will be noticed that there is a little want of similarity between this diagram and Fig. 7. Here the steam line in b is horizontal at first, and then slopes downwards.

That it is horizontal at all is owing to some peculiarity in passing the steam from one cylinder to the other, as there should be the slope of an expansion curve throughout. But any deviation from theoretical proportions does not affect the general inference to be drawn from the two diagrams when viewed together, and we see that the expansion which has occurred in the highpressure cylinder might very well have taken place in the low-pressure cylinder, as something which preceded the actual expansion therein.

ENGINES WITH CRANKS AT RIGHT ANGLES.

For many purposes it is enough to have an engine with a single steam cylinder, or an equivalent engine, with a pair of cylinders acting as one only; but on the other hand, there are numerous instances where two engines should be

placed side by side and work cranks at right angles to each other.

This is particularly the case in applying steampower where it is of consequence to preserve the rotative pressure on the crank as nearly uniform as possible, and to maintain a smooth and even motion; or again in marine engines, for convenience of starting in any position, the same rule would hold; and before proceeding further it may be useful to point out the reason for the greater uniformity of rotative pressure which is a consequence of working with a pair of cranks at right angles.

e

FIG. 7.

That the variations of tangential pressure on the crank of a direct-acting engine are represented by the vertical lines on a diagram similar to that shown by the dotted curve b ƒ in Fig. 7.

Putting a series of such curves end to end, we obtain a graphical indication of the fluctuations of tangential pressure during the working of an engine with one cylinder.

The force is zero at dead point, and rises to fe, its greatest value, after which it sinks again

to zero.

But if there be a pair of cranks at right angles, a second series of diagrams of rotative pressures must be superposed upon the first series, as shown by the second set of dotted curves, whereof one portion is marked be, and the final result is exhibited by the upper line, not dotted, which is obtained by adding together the pairs of ordinates at each point. For example:

N M+M R=M S.

A B A B A C.

E F+0=E F.

The greater uniformity of rotative force is apparent, and it would be improved by cutting off at half stroke in each cylinder, for then the curve be would be hollowed out and reduced, while the part b ƒ would be unaffected, and the upper resultant wavy line would become more nearly horizontal.

By proceeding in this manner it is easy to set out a diagram of the rotative pressure upon the cranks of any pair of engines working under given conditions.

In applying these principles to direct-acting engines, where two cranks at right angles are to be connected with the cylinders, there are different methods for adoption, each of which has its advocates. One plan very commonly met with has been to place the high and low pressure cylinders in pairs, with their axes in the same straight line, so that one piston rod serves for

both. Thus, in marine engines, with the cylinders vertical, there may be:

(1) The high-pressure cylinder above the lowpressure cylinder.

(2) The low-pressure cylinder at the top.

(3) The low-pressure cylinder encasing the high-pressure cylinder.

But in each of these cases, as also in compound horizontal engines, it is usual to confine the expansion to one pair of cylinders.

THE USE OF AN INTERMEDIATE RECEIVER.

In another class of compound engines there are two cranks at right angles, but only one cylinder connected with each crank.

Here each cylinder forms, as it were, an engine complete in itself; the cylinders (called A and B, as before) are placed side by side, and are of equal length; and the point to be noticed is, that the pistons in A and B no longer move together, but that one leads the other by half a stroke.

It is clear that the mode of exhausting at once from A into B is no longer applicable, and that some special method of distributing the steam, different from anything that we have yet seen, must be arranged.

The difficulty arises from the fact that the directions of motion of the pistons cross each other, whereby, for example, when the piston in