It will be seen at a glance, and can be easily demonstrated with a pair of dividers, that if the point j of the lower end of the actuating lever were attached directly to the point e1 on the connecting-rod, there would be imparted to the centre m, or fulcrum of that lever, an unequal vibration above and below the centre of the reversing shaft, which would give an unequal port and cut off for the two ends of the stroke. The function of the stirrup link e1 f and its attachment to the valve lever at j is to correct this error, for while the point e on the connecting-rod is performing a nearly true ellipse, the point j is moving in a figure like an ellipse bulged out at one side; but the axis j1 j2 equal to the minor axis d1 d2, which is also equal to n1 n2. This irregularity is so set to correct the above error, and it gives an equal vibration to the centre m, above and below the centre of the reversing shaft. Although the position of the point j may he found by calculation, it is much more quickly found by a tentative process, and to test if the assumed point j be the correct one, we mark off on each side of m vertically the correct equal vibration required, nl n2, which will be the same as d1 d2. Then apply the distance e1 j to j1 and d2 j2. Then if the length j n be applied to j1 n1 measuring from j1 and to j2 n2 measuring from j2, and the point m fall below n1 n2 in each case, it will be necessary to take a point on e1f higher than j; or if, on the other hand, m fall above n1 n2, then the point j must be taken lower. This point will be very soon found, the only stipulation being that the length j m of the lever j m g1 must be such that its centre m vibrates equally on each side of the centre of the quadrant, also marked m.

The point g will be the point of attachment for the valve rod, which may be of any convenient length; but from that length as a radius, the centre of the links must he drawn from a centre ml on the parallel line m m1, which corrects the error caused by the movement of the end of the valve rod. The angle at which this curve is set from the vertical - which is mid-gear - will give forward or backward gear, the angle leaning forward s1 being forward gear and the reverse s2 being backward gear, the centre of these curves being shown at m2 and m3. The amount of the angle, marked on the curve of extreme vibration at s s1 or s s2 will equal a quarter more than the full opening of the port at that angle, and the point of cut-off will be about 75 per cent. Thus if 1 inch opening of the port be required, the amount of the angle s s1 must be 1¬ inch. Laid out in this form, the leads and cut-offs for both ends of the cylinder, and for backward and forward gear, will be practically perfect and equal, and the openings of the port nearly so. Deviations may be made without materially altering the correctness of the results.

During the passage of the point j the point m receives an acceleration due to the decreasing angularity of the rod up to the point j1, and also during the remaining portion of the stroke owing to an increasing angularity. For the return stroke the point m encounters retardation for similar reasons. This is shown in Fig. 309, where R equals the amount due to the vertical movement of the point j, and P that due to the decreasing angle of the rod, whereas in the return stroke the whole of the movement of the point m would be repre-sented it by x, excepting for the retardation represented by the minus quantity y. This peculiar movement occurs just at the right moment. The valve opens and admits steam in ample volume at the commencement, and then pauses when full open, due to this peculiarity. It also opens to exhaust promptly to the full extent, then pauses, so that the steam escapes in the same free manner that it was admitted. These advantages can be more forcibly realised by comparing the valve path diagrams of sister engines, one having this gear and the other the link motion, both having the same amount of lap and cut-off. Further advantages are: - Every portion is always in useful action, cylinders can be placed nearer together and increased bearings attained, it is accessible and easy to reverse, besides having less area by one-quarter exposed to friction than the link motion.

In Fig. 309, if 100 represents the total amount of resistance to the progress of the valve, then the different values of the stress upon the other pins are represented accordingly.

Now that the valves have been set, the holes may be opened out in the frames for the reversing shaft brackets and four 1-inch bolts driven home into each. In the meantime the foot plate end has been receiving attention, for the cab has been fitted, put on, and fastened to the platform with bolts. The cab sides and splashers are formed of one plate outside, the cast-iron sand-box being fitted to this plate inside, the reversing bracket being bolted to the top of the left trailing splasher. Now that the two sides are fixed, it only remains for the front and top to be placed, when the cab will he complete. The vacuum brake cylinder is then placed in position by securing it to the underside of the foot plate by carrier plates or brackets. Then all vacuum and injector pipes, etc. are coupled up, also the handrail round the boiler and smoke-box front is secured to its eye bolts or studs. The brake hangers are then hung on their studs, the cross bars fixed, and all roils coupled up. This completes the work as far as shown in the figures; what remains to be done, which is about six or seven hours' work, is shown on the sectional drawing. This consists of securing the buffer-plate, buffers, draw-bars, fall plate, foot steps, packing glands, etc. The engine is then slung in the cranes and the coupling rods put on, the springs adjusted, carried to the traverser, and then drawn out of the shop. The tender is then coupled to the engine, the brake blocks adjusted, also all necessary pipes, and then fired up ready for trial, which consists of about twenty to twenty-five miles run.