In measuring hills and inclined surfaces, the horizontal distance must be taken. A plummet should be suspended from the end of a chain when it is levelled.

When hills are very steep, the surveyor should assist the chainmen, and when the best is done in levelling and plumbing the chain, judgment must frequently be called into exercise.

Even when elevations and descents are easy, there is danger of making too much measure. In such cases, chainmen often make allowances, but the surveyor will do better to keep them to close measure, and from the shape of the ground, judge what allowances ought to be made. An experienced, judicious surveyor, will form a more correct judgment in such cases, than inexperienced men.

If a surveyor is acquainted with the rise or descent of hills, the traverse table will aid him. For example, from the surface of a hill which rises on an angle of five degrees, one link for every ten rods may be deducted.

Particular care must be taken to have the chain carried on

a straight line. When a tally is ended, and the hinder chainman carries forward the sticks, they must be counted, When on counting the sticks it is discovered that one is lost, the chainmen should not leave the chain and go back to find it, but, from the last mark, should measure back to the point where the tally commenced, to see whether the full tally was measured, or whether one chain was lost from it. Many blunders have been left undetected by not taking this care. A careful, accurate chainman, never loses a stick, or miscounts a tally.

Distances may be taken by chains and links, or by rods and decimal parts, as best suits the surveyor. When a survey is calculated by chains and links, the numbers by which the process is performed, are smaller than when it is calculated by rods and decimal parts. Every method by which the numbers are diminished, is an improvement.

Young surveyors should practice much for their own instruction, and should make correct practice familiar, before they offer their services.

A young surveyor should bear in mind, that if he is detected in one error in the beginning of his practice, it will be more to his disadvantage than to be detected in two, when he shall become well established. If an error is committed in a survey, it is not against the surveyor, provided he detects and corrects it; but if he cannot do this, it is sufficient evidence of

No. IV.

DIRECTIONS FOR RUNNING AND ESTABLISHING LINES.

Many people suppose that a surveyor, at the beginning of a line, by intuition, or by some magic art, can set his compass directly to the terminating point, whatever obstructions may intervene, and that he needs no assistance; but this is a mistaken idea.

In running a line of considerable length, a surveyor should have two assistants to carry the chain, two to carry flags, in whose ability and correctness he can confide, and a fifth to carry an axe. If the surveyor is not furnished with such a number of assistants, his employer need not place too much confidence in his work.

The flag staves should be as much as two and a half or three inches in diameter, or what would be better, two strips of a board of that width, and seven or eight feet in length. If they are not so wide, they cannot be seen through the sights of the compass at any great distance. On one end of each staff, a flag, either white or red, of a yard in length, should be wound tight, and not left to hang loose and flutter in the wind. The latter colour will be more clearly seen through bushes, especially when the ground is covered with snow, and the brighter the colour the better. Being thus manned and equipped, at the beginning of a line, he must set his compass as near the true line as he can, or, he may set up one of the flags at the place of beginning, and go forward as far as he can have a view of the back flag. There set his compass on the random line, and send the other assistant as far forward as he can conveniently see the flag. When each flag is clearly seen through the sights of the compass, the back flag must be brought forward, and placed where the compass stood. In this manner he must proceed on his random line, taking care each time he sets his compass, to turn the sights to the back flag. Great care must be taken to keep the flags perpendicular; also, the surveyor must keep the staff and the sights of the compass perpendicular. A little leaning of the flags, or turning the sights of the compass from a perpendicular, will make a crooked line. In taking his sights to the flags, he must look as near the ground as he can, and when practicable, the flag should be turned downwards, on account of the danger of being leaned when kept up.

All obstructions, such as bushes, brush, &c., must be cleared

venient distances, perhaps at every twenty rods, stakes must be set directly in it. Every stake must be numbered, that no mistake may be made in calculating, when they are afterwards removed and placed on the true line. If, in the course of the random line, the needle does not traverse as at first, or does not traverse alike at different places, no regard should be paid to it, the flags must direct the course; neither should the surveyor be turned aside or terrified by the cry of either of the parties, you are wrong, you are wrong, (for he will most certainly hear it,) but he must continue his random line, until, turning at right angles, either to the right or to the left, as the case may be, he can exactly strike the bound, or the point where a bound is to be erected; there he may stop, and measure the distance from that place to the bound. He may then calculate the course and distance of the true line, as before taught; but for setting the stakes in the true line, he may take the following directions.

Suppose the whole length of the random line is 200 rods, and the distance from the termination of it to the bound is 90 links, the calculation for setting the stakes on the true line may be made thus:-As the whole distance is to 90 links

so is the distance of any stake to the distance that such stake is to be moved.

If the stakes are 20 rods apart, the answer is, the first stake is to be moved nine links, the second 18 links, and so on, adding nine links at each stake, until the whole are moved at right angles from the random to the true line. Many crooked lines and consequent disputes between farmers, have arisen from the want of this care and attention. When a long line is to be run over a number of ridges, and across intervening valleys, it should first be run and established from one ridge to another, and the intermediate spaces in the valleys may be taken afterwards. By taking long sights, there will be less danger of turning from a straight line.

In all cases, the forward flag should be carried as far as it can be distinctly seen, unless it is at the termination of a line.

On account of the irregularity of the magnetic needle, the difference between compasses, and the incorrect manner in which most of the former surveys were made, no rule can be given which may be relied on, with certainty, in renewing lost bounds. In such cases, the best evidence must be taken which can be produced. When an old line is to be renewed, where the bounds have been lost, the surveyor may verify or prove his accuracy in various modes. If he finds, for example, that the lots on each side of the line contain their due

quantity, or possess their full width, he will have reason to believe that the line is accurately run; if not, he may allow for the excess or deficiency in either of these lots, and thus discover the true position of the line.

It would be difficult to mention all the circumstances which may govern, or which may serve as evidence in such cases.

No. V.

ON BALANCING SURVEYS.

In every survey which is taken without an error, either in courses or in distances, the sum of the northings will equal that of the southings; and the sum of the eastings that of the westings; but this is not always an infallible proof that the survey is accurate, for two errors may be committed, one of which will exactly balance the other, which no rule will detect; but such cases do not often occur.

In practical surveying, it is next to an impossibility, in any case, to work so correctly that the survey will exactly close without some correction. The difference between the two columns of latitude, and between those of departure, are the legs of a right angled triangle, the hypothenuse of which will be the distance which the survey will fail of closing. In a survey of one hundred acres, whatever may be the number of the angles, the difference between the two columns of latitude, and the two columns of departure, ought not to exceed a rod for each, but to come within those limits, if possible.

If, in such a survey, either of the differences should exceed a rod, where the land is valuable and easily surveyed, it would be better to take a re-survey, or so far as to detect the

error.

These differences, as before taught in this work, must be balanced before the column of meridian distances can be formed, or a correct plot of the survey be drawn.

The small differences between the columns in surveys which are substantially correct, commonly arise from an excess of measure, on account of the irregularity of surface, and from errors of the compass, caused by local attraction, and diurnal motion.

Some authors have given rules for balancing surveys, which, in theory, appear plausible, but, in practice, they are not with out exceptions.* In correction, these rules are indiscrimi nately applied to every line in the survey, which pre

*It is argued, by theorists, that by these rules all will be agreed. This

supposes that a proportional error must have been committed on each and all, both in courses and in distances; when in almost every survey, a part of the lines are on land so level and so clear from obstructions of any kind, that if the surveyor and chainmen attend to their business, they will not be likely to commit much error on them, while other lines on other parts of the same survey are attended with so many difficulties, that when they have done their best, it will scarcely be possible for them to avoid some error, and the surveyor who takes the survey will best judge on what lines the errors were committed, and whether they were in the courses or in the distances. In all cases, the corrections should be made on the lines containing the errors.

When the errors are in the courses, they should be corrected, and when the errors are in the distances, the correction should be in them, or the correction may be in both courses and distances, as the surveyor may judge proper.

Surveys are commonly balanced by substracting half the sum of the differences, from the numbers in the larger columns, and by adding them to those in the less.

This, in most instances, is a correct method, but in certain cases, there are exceptions to it. For example, the boundary lines of three sides of a farm are on level land where there are no obstructions. The other side is on rough land filled with broken precipices. In such a case, most, if not all the correction, should be on the lines on the rough side. some instances, most of the error may be in the course of a single long line..

In

It is not uncommon on high hills that the traverse of the needle differs a quarter or half of a degree, or more, from its traverse in valleys and on plains below. In such cases, courses should be corrected. In general, the compass is subject to more error than the chain. As rules cannot be given for correcting surveys in all cases, this must be left in a great measure to the judgment and experience of the surveyor.

As a general rule, corrections should be made principally on long lines, and the correction may be in proportion to the difficulties with which the surveying of them is attended.

When a course is north or south, or east or west, or near either of those points, a few minutes may be added to it or substracted from it, when necessary to favour the balancing. If the distance is of considerable length, it is probable that the course may contain some error.

When a course is nearly equidistant between any two of the four cardinal points, if the correction of it increases the