starting, and be a miniature reproduction of it in all other respects; or, in other words, equal piston movements must be represented by equal movements of the paper throughout the whole stroke. To whatever the cord may be attached, whether to a temporary wooden pendulum fastened by a screw to a post, or to the beam of a beam-engine, it (the cord) must be at right angles to a line between its point of attachment and the pivot of the beam or pendulum to which it is attached, when the piston is in the middle of its stroke. For instance, suppose the engine to be horizontal, and that a wooden pendulum is attached to a light post set up by or on the engine, or some other convenient object, or is suspended from a joist, the lower end being connected to the cross-head, and that the point on the pendulum where the motion is sufficiently reduced for the paper drum is higher than the instrument, so that the cord must incline downward. In such a case, unless a carrying pulley is used to deflect it to a horizontal direction, or unless the point of attachment for the cord is moved as many degrees from the centre line of the pendulum as the cord inclines downwards, the movement of the pendulum will be too fast at one end of its travel, and too slow at the other, and the diagram will be distorted. The effects of such distortion will be to cause the ends to appear unequal when they are not so, or else to conceal or exaggerate inequalities where they really exist.

The length of the pendulum may be from one and a half times to twice the length of the stroke or more. If too short, the ends of the diagram will be distorted, unless the connection between it and the cross-head is sufficiently long. The pendulum may be attached to some object at the side of the engine, so that it may vibrate in a horizontal plane.

A good substitute for a pendulum consists of a drum about six inches in diameter, more or less, on the axis of which is another drum, the diameter of which requires to be as much less than the other as the movement of the paper is less than the travel of the piston. If the drum is mounted in a convenient position, and a cord from the large part is attached to the cross-head, and another

from the small part to the indicator, a spring in the large drum keeps its cord taut, just as the spring in the drum of the indicator keeps its cord taut.

When two instruments (right and left), are used, it is best to fix a sliding-bar alongside of them having the proper motion, carrying pins to which the cords are attached, these pins being placed between the two instruments so that the cord of each may pull towards the other, and the movement of the piston from either end of the cylinder may pull the cord of the instrument attached to that end, in which case the upper line of each diagram will be drawn while the cord is being pulled. But no perceptible advantage in the way of accuracy need be expected from this arrangement, though it is a very convenient one where a large number of diagrams are to be taken.

Analysis of diagrams.-All the various particulars which may be learned from the indicator diagram may be classed under three heads.

1. Those relating to the condition of the engine, such as its construction, adjustment, etc.

2. The mean or average pressure exerted on the piston as an element in calculating the indicated horse-power, I. H. P.

3. The theoretical rate of water consumption.

Here it is necessary to explain the terms used hereafter to designate the various parts of the diagram.

In diagram No. 1, A A shows the atmospheric line which is drawn when both sides of the piston of the indicator are exposed to the atmosphere. When tracing such a diagram it is preferable to pull the cord by hand, in order to make the atmospheric line longer than the diagram; B C is called the steam line. It is formed while steam is entering the cylinder. C is the point of cut-off. It cannot always be located exactly by inspection, as the closure of the part is generally sufficiently gradual to cause considerable fall of pressure before the port is entirely closed. In general, it may be located at the point where the outline of the figure ceases to be convex and commences to be concave. CD

is the expansion line or curve.
like that of the cut-off, may be
flexure, or that point where the expansion line begins to change
the direction of its curvature. DE is the exhaust line. It com-

D is the point of exhaust, which,
located at the point of contrary

mences at the point of exhaust, and may be considered as ter-
minating at the end of the stroke, (though, strictly speaking, it
does not terminate till the exhaust port closes at F.) E F is
termed the counter- or back-pressure line, and by some the vac-
uum or exhaust line; but the former terms are more appropri-
ate, as they are applicable to all diagrams, whether from condens-
ing or non-condensing engines. In the diagrams of non-con-
densing engines it is above the atmospheric line, A A; while
in condensing engines it is below; but in both cases it rep-
resents some counter-pressure, since a perfect vacuum is unat-
tainable. F is the point of exhaust-closure. Its exact loca-
tion cannot be so readily determined as the points C and D,
as, although like the former, it is anticipated somewhat by a
change of pressure, it is not marked by any change in the direc-
tion of the curvature of the line. In perfectly working engines
it may be located geometrically, but it is seldom necessary to do
so, since for all practical purposes it is sufficient to know where
the change of pressure due to the closing of the exhaust begins,
and its final result. F G is the compression curve, and G B is
the admission line. These constitute all the lines which belong
to the diagram proper, and all that are produced by the instru

ment.

For certain purposes the vacuum line V V, and the clearance line HH, diagram No. 1, are drawn, the former parallel to the atmospheric line, and at such a distance below it as will represent, according to the scale used, the pressure of the atmosphere as it was, or was supposed to be, at the time and place at which the diagram was taken. For this purpose it is usual, when great accuracy is desired, to consult a barometer at the time, and record its reading on the card; but, in the absence of a barometer, it is usual to assume the pressure at 147 lbs. per square

inch, which is the average at sea level; but, since the pressure diminishes at the rate of lb. for each 189 feet of elevation, allowance should be made for the known or estimated elevation of the locality. It should also be remembered that the pressure will vary nearly 1⁄2 lb., and sometimes more, from changes

in the weather.

The clearance line HH, diagram No. I, is drawn perpendicular to the atmospheric and vacuum lines, and at such a distance from the induction end of the diagram, that the space between them will bear the same proportion to the whole length of the latter as the whole volume of clearance bears to the piston displacement. When the amount of clearance is unknown, and it is not practicable either to calculate it or measure it by filling the space with water, it must be approximated as near as possible from the known clearance of engines of similar construction. The largest clearance will be generally found in the smaller sized engines of the ordinary single slide-valve type. Five such engines tested at the Cincinnati Industrial Exposition of 1875, had the following amounts 9, 9, 10, 114, and 12 per cent. of the cubic contents of the cylinder. Next to these will be the larger sizes of the same type, in which the clearance will range from 6 to 10 per cent. When two slide-valves are used with short, direct ports, but exhausting under the valves, the clearance will average from 3 to 6 per cent., according to the proportionate length of the stroke, the longest strokes having the smallest per cent. Corliss engines, in which the stroke is about three times the bore, have about 3 per cent. The least amount of clearance is obtained from valves designed to exhaust at both ends of the cylinder, instead of in the centre, as in the case of the ordinary single slide-valve. By such an arrangement of the steam- and exhaust-valves, the clearance has in many instances been reduced to 11 per cent. The clearance in poppet-valve engines is more difficult to calculate than in slidevalve engines; but, as a general thing, it does not exceed 5 per cent. It should be measured with water, when it is desirable to ascertain accurately the cubic contents of the clearance. In poppet-valve

S

engines the cut-off and other events are independent and adjustable; consequently, diagrams taken from this class of engines are free from the limitations attending those taken from slide-valve, because advantage is frequently taken of their freedom of adjustment to give an earlier cut-off, or a later depression than is usually adopted with slide-valve engines. In all such cases the diagram will faithfully state the fact.

The Most Accurate Methods of Testing the Adjustments.

The conditions which are mainly instrumental in determining the conformation of the diagram are the valves and valve-gear, the length and capacity of the steam- and exhaust-pipes and ports, the design of the governor-valve, the condition of the valves and piston as to leakage, the amount of clearance, the speed of the piston, etc. The engineer may be called upon to analyze a diagram with reference to all the above conditions, or only to accidental derangements. In the first place, he must compare the diagram with one of the best form which can be produced in practice from the class of engines to which the one in question belongs; in the second case, he must discriminate between such defects as are due to accidental derangements and those that are due to design and construction, which cannot be remedied without the substitution of new parts. Suppose the engine to be of the throttling kind, of the best attainable construction and adjustment, its diagram should possess the following general features:

1. The initial pressure at B, diagram No. 1, should be as high at least as any subsequent pressure; and if the engine is not driving its maximum load, and the steam is in consequence more or less throttled, the pressure should begin to fall at B, and continue to do so at a tolerably uniform rate, until the point of cutoff, C, is reached.

2. The cut-off, when obtained by means of lap in the slidevalve, cannot, as a general rule, take place with advantage earlier