Some of the most important of a Clerk of Works' duties are those which have to be undertaken at the earliest stage of the contract, for it is then that the building is set out on the ground. He has not to do the setting-out himself, but it is essential that he should check it in every possible way, as once it has been started, and the foundations dug, it is exceedingly difficult to correct any error that has been made. It is possible, for instance, for a Contractor to misplace his buildings altogether on the site, especially on a tolerably open field or amongst underwood which has to be cleared away for it. One of the main points of the building should be fixed by the Contractor or his Foreman, and this should be checked by the Clerk of Works by measuring up to it from other well-marked points shown upon the block plan. At least two such measurements should be taken; but it is much better to check these two by a third, and in case of difficulty this should always be done. Tape measurements are usually trusted; but a tape is not a reliable instrument unless it be new, stretching considerably in course of time. A new tape should consequently be used (or preferably a steel tape, though it breaks easily), and at this early stage it is well to mark against a wall - or some other object which is not likely to be destroyed by the works - the exact tape length of 66 feet, measuring it first with the tape and then checking it with great care by means of the 5-feet rod, making marks along the wall at every 5 feet or 10 feet. If at any subsequent time there be any doubt as to the correctness of the tape, it can then always be checked against this standard marking; and, in fact, this ought to be done occasionally, for the reason above stated - that a tape rarely retains its true length for very long.

When one point of the building has been fixed in this way a second point must be similarly fixed, preferably at the other end of a long wall, of which the first point was one of the corners. It is not sufficient to trust to fixing the first point alone correctly, for if this is done it is possible to lay out the house with a wrong aspect, turning upon this point as on a swivel.

It occasionally happens that one of the walls is described to face one of the cardinal points, or that, as in the case of a church, the centre line shall run due east and west. It does not do in such a case to trust to the compass, for it must be borne in mind that compass bearings are very far from accurate. At the present time the N. point on the compass is directed about 16 degrees west of true north, and this varies slightly day by day, while there is a gradual tendency, though it is a very small one, towards greater correctness. Even were the compass an accurate guide, the small instrument carried on a watch-chain would be utterly unreliable, and even that upon an ordinary level, though it will come to rest (which the little hand instrument will not do), is still liable to be affected by all sorts of accidental causes, such as the presence of an iron spade in the hand of a labourer who is standing near. It is in the writer's experience that an iron railing about 5 feet from his instrument once drew it out of truth to the extent of over 12 degrees.

For most purposes the shadow cast by the sun at noontide may be considered to give a north and south line with sufficient accuracy. In order to obtain this all that is necessary is to erect a vertical pole on a sunny day. Note, and drive a small stake or pin into the ground at the end of the shadow which it casts about half an hour before noon; then, taking a piece of string and tying it to the pole, stretch it, and holding a sharp stick or some other pointer where the pin has been driven in, describing a circle with the pole as centre, marking it on the ground. It will be noticed that, as noon approaches, the shadow will recede within this circle, and then will gradually come out to it again. When the circle is reached a second time a second pin is driven in, and midway between these two pins will be a point which lies due north from the pole. It must be noticed that the first pin should be driven in a good half-hour before noon, as, although the shortest shadow is always cast when the sun is due south, this only happens at noon twice in each year, being sometimes before and sometimes after that time. Of course, this can be done much more accurately by means of a theodolite; but such extreme exactitude is very rarely needed in building works. The North Star can also be taken as an index; but this varies at different times of the year, though its extreme error is only something like 2 degrees from exactitude.

There are cases in ecclesiastical work when it is desired to set out the centre line to the point of the sunrise on some particular Saint's day; but in such a case the Clerk of Works does best to consult with the Architect, and leave the responsibility of checking in his hands. He may then very likely demand that a true meridian line be given him - say by means of the sun's shadow, as has been just described; - and from this he will give instructions as to the angle of declination, which will have to be laid out by the help of a theodolite or some other angular instrument.

If one has not already been supplied, in all probability the Builder or his Foreman will make a foundation plan, showing the outline of the trenches which are to contain the concrete; and it is from this plan that the setting-out would be proceeded with. If this be done a good deal of trouble might be saved by the Clerk of Works going carefully over it, checking each point to see that the proper widths have been provided everywhere. This can easily be done if either it or the ground-floor plan be upon tracing-paper. It is now usually the practice to supply both Contractor and Clerk of Works with a set of sun-prints of the contract drawings; and it is these with which the foundation plan has to be compared. If they are inaccurate, so in all probability will be the foundation plan; and as sun-prints are no more reliable than tracing-cloth in the matter of scale, it is well that they themselves even should be checked against the figured dimensions, and in case of any discrepancy the figured dimensions must always be taken in preference. It may consequently become necessary to draw out a fresh brickwork plan from the ground-floor plan, and to make, or to instruct the Contractor to make, a new foundation plan from this. There should be no necessity to consult the Architect so long as the drawings are sufficiently dimensioned, but in case of real doubt he should be communicated with at once, for, as already said, a mistake at this early stage of the work is a very serious matter indeed.