Always begin grade-lines at the summit, and work down. For such service, carry habitually a slip of profile paper, say six inches wide and two feet long. Rule the proposed gradeline on it, assume a summit cut, mark the stations, and start down. When at fault, the elevation can be spotted on the profile, which will show at a glance, without any calculation, how you stand in relation to grade.

The work of each day should be compiled and recorded in the evening, that no delay may result from the loss or defacement of a field-book.

On location, check the transitman's calculation of the length of each curve and the fractional deflections.

The senior assistant must be qualified to locate a line accurately on the ground from the project furnished him. Lateral deviations exceeding five feet on ten-degree slopes, three feet on fifteen-degree slopes, and two feet on twenty-degree slopes, will be considered errors requiring correction. Measurements to the experimental line should be made and noted frequently, in order not only to check the field-work, but that the line may by means of them be laid down on the map.

The senior assistant will supply himself with drawing instruments, colors, brushes, and the like personal furniture of an engineer. He will take care also that the stationery, fieldbooks, instruments, and other articles of outfit supplied by the company, are not misused. His field equipment should always

added a straight, round staff, five or six feet long, steel pointed; it will be found exceedingly useful.

If without a topographer, he should make sketches of irregular ground, of streams, buildings, roads, and the like, to help in compiling the map.

In hilly or wooded districts, the front chainman carries the flag on survey, and is at the head of the line. In open, plain country, work is greatly forwarded by detaching an axeman with flag, to accompany the senior in advance, and set turningpoints for the transit. The transitman follows as rapidly as possible, and the chainmen come after, lining in their stakes by the eye from point to point. The whole force is thus kept pretty steadily in motion.

On wide plains, a set of chain-pins may be used, and surveystakes placed five hundred or a thousand feet asunder. Very often stakes at intervals of two hundred feet are sufficient, the levels being taken every hundred. Location stakes are put in every hundred feet.

4. THE TRANSITMAN will be expected to keep his instrument in adjustment, and to be quick and accurate in its manipulation. It is not needful to plant it as if for eternity. On the contrary, it should be set gently, the legs thrust but slightly into the ground, and the screws worked without straining.

On long tangents it is a good plan to reverse the instrument at each new point, putting the north and south ends forward alternately. Small errors in adjustment are thus balanced in some measure. Select also, in such a case, some distant object in range, when practicable, to run by. The telescope, in wind or sun, will sometimes warp a little out of line.

Never omit to note both the calculated and magnetic bearings of the lines on survey, and of the tangents on location. Guard against the error of reading deflections or bearings from the wrong ten mark; as, for instance, 34 instead of 26.

At the beginning of a curve, let the rear chainman know the plus of the P. T. Tell the front chainman the degree of curve, and instruct him how, by multiplying 1.75 by the degree, he can find the distance of each full station from the range of the last two. A quick fellow will soon pick this up, and become wonderfully skilful in practice. Thus accomplished, he is a check on wrong deflections.

In running curves, a tangential angle of fifteen degrees from one point should seldom be exceeded: twenty degrees is to be

Carry a pocket-compass, and observe with it the magnetic bearings of streams and roads crossed.

Record daily each day's run; fill out the distance column, transcribe the chain-book, and, on location, record the apex distances also in the column of remarks.

On survey, do not erase from the field-book the notes of abandoned lines. Simply cancel, and mark them "abandoned," in such manner that they may still be legible.

When required by the senior assistant, the transitman will aid in the making of maps.

5. THE LEVELLER must be familiar with the adjustments of his instrument, keep it in order, and handle it rapidly.

On survey, establish and mark benches at half-mile intervals; on location, four to the mile when practicable.

Note the surface elevations, the depths, and the flood heights, of all considerable streams crossed. Take a rod in the beds of small streams.

Six hundred feet each way should be regarded as the maximum sweep of the level.

Carry a hand-level, and thus save the time required to peg across narrow hollows, or over heights which can be turned with the instrument.

The leveller should record his work, and make up the profile daily.

6. THE RODMAN will give his intermediates close by the stations, observing the number of each one as a check on the chainmen, and calling it out to the leveller. He should have an eye to abrupt irregularities in the ground, and give plus elevations when necessary.

He will keep note of bench-marks and turning-pegs, describing the latter occasionally with reference to the nearest stake, that the levels may be taken up speedily in case of a revision of the line.

When unaccompanied by an axeman, the rodman is equipped with belt and hatchet. Sometimes he is furnished also with a steel pin for turning on. The pin has a ring through the head, by which it may be hung to a spring hook in the belt.

The rodman will assist the leveller at record and profiles, and transcribe the slope-book daily.

If stakes of survey are set at intervals of two hundred feet, give rods every hundred feet, as nearly as the midway points

7. THE SLOPEMAN will give backsights, and take the cross slopes for one hundred feet on each side of the line at every station.

8. THE REAR CHAINMAN will carry a book in which to note the turning-points, the crossings of roads, streams, swamps, woodland, and, when convenient, property lines also. He will

hand it in daily to the transitman for record. As each successive chain is stretched, the rear chainman calls out the number of the stake it is stretched from: this assures the selection of the right number for the stake ahead.

9. ONE AXEMAN will be detailed to make stakes, another to mark and drive them. Additional axemen may be employed at the discretion of the senior assistant, as the work requires them. Wanton destruction of timber, fences, growing crops, or other property, should not be allowed. Axemen must be careful, in passing through the country, to do as little damage as possible.

XXV.

THE CURVE-PROTRACTOR, AND THE PROJECTING OF LOCATIONS.

1. The curve-protractor is simply an eight-inch, semi-circular horn protractor, upon which a series of twenty-three curves, from half a degree up to eight degrees, is finely engraved, with radii of 400 feet to an inch. After some years' use in his own practice, the contrivance was transmitted by the writer to the well-known firm of James W. Queen & Co., mathematicalinstrument makers, New York and Philadelphia, by whom it is now manufactured. It greatly facilitates the projecting of lines and solution of field-problems on location. It enables the engineer, for example, by a short, graphical process and rapid inspection, to find the curve which shall close an angle between tangents, or terminate a compound curve, and pass at the same time through some fixed intermediate point, without liability to the errors, and free from the loss of time, involved in a tedious calculation. Other applications, such as the nice adjustment of line among buildings, on precipitous steeps, and

2. For office use, the writer prefers a home-made curveprotractor of isinglass, prepared as follows: Take a thin, clear sheet, say six by ten inches, free from bubbles and cracks. Block it securely on the drawing-table with thumb-tacks, setting the shanks close against the edge of the sheet, but not piercing it, and the heads lapping its edge. From a centre, midway of one of the long sides and near its margin, strike the curves from 12° or less, varying outwards by half-degrees to 6°; thence by quarter-degrees to 4°; and thence by ten-minute differences to 21°. This covers one side of the sheet, the scale being 400 feet to an inch. Now release the sheet, turn it over, and on its other face strike the remaining curves, down to ten minutes, from centres on the table, in the reverse direction, so that they shall cross the first series at a large angle. Space them about three-eighths of an inch asunder at the middle. Use a needle-point centre for the first series, to avoid boring a large hole in the sheet. Add also, on that face, two radial lines drawn towards the corners. Score the fractional curves very lightly, the full figure curves a little deeper, but all of them with steadiness and delicate stress. Practise beforehand on a separate slip, for the right intensity of stroke. Engrave the numbers with a stiff steel point on the opposite side of the sheet to that upon which the corresponding curve is traced. Bring the work out by rubbing it with India ink. If preferred, the flat curves on the reverse side may be colored with carmine. Duplicate protractors will be found useful in projecting compound curves. Clip off the four corners, reenforce the edges with a narrow ribbon of tracing-linen, folded over them and glued fast, and the article is complete. It is perfect for its use; durable, flexible, spotlessly transparent, not liable to warp or change dimensions with changes in the temperature or moisture of the air, and, withal, takes and preserves a visible line, thin as the gossamer.

3. To experienced locating engineers, the curve-protractor needs no wordy commendation. Contrasted with the inconvenient appliances of the old method, — cardboard, veneer, glass, or dividers, its advantages will be manifest. A few hints as to the manner of using it may be in place.

4. First of all, let the experimental line approximate to the probable line of location; and, upon that base, construct a contour map, with reference to which special observations