It is. generally claimed that the question of labor is the most difficult and uncertain the carpenter is called upon to solve. Material can often be figured very closely, but just how long it will take to work up a lot of material and place it in position in a building can not be so easily determined. The cost of labor depends upon the time required to perform a certain amount of it. All men do not work alike; some will do easily one-third more than others - hence the time required to perform a certain amount of labor depends largely upon the ability of the men employed, the advantages they take in doing work and the skill of the foreman in the management as it progresses day by day. It is an easy matter to find four men who will do as much in a day as five others, and to illustrate the surprising result of the difference in the ability of men to perform labor, I will give a practical example.

Suppose two contractors, A and B, each have a job of work exactly the same. A takes his job for \$900 and B his for \$800. Each pays wages at the rate of \$2.50 per day, and each employs five men ; but four of B's men are equal to five of A's and it takes 60 days to complete his job. Which will make the most money, and how much? The solution of this problem is as follows: If A employs five men at \$2.50 per day for 60 days, the labor will cost him \$750 ; as he took his job for \$900, his profit is \$ 150. Now if four of B's men are equal to five of A's, B will complete his job in one-fifth less time than A, which will be 48 days. Now, if B employs five men at \$2.50 per day for 48 days, the labor will cost him \$600, and, as he took his job for \$800, his profit is \$200. Thus we can see how one man can underbid his competitor \$ 100 on \$900 worth of work and still make the most money. Again, suppose it required B 52 days to complete his job; even then he could bid \$100 lower than A and still make as much money. The above example shows at least one chance for the surprising difference in builders ' estimates on the same work. It also shows how the difference in the ability of the workmen employed and the management of the work can make a vast difference in the cost of a building. Under such circumstances how can a contractor make estimates upon which he can rely?

In all kinds of work there must be an average, and this average is what is wanted as a standard in estimating. If labor cannot be estimated from what is known to be an average day's work, then we naturally conclude it must be estimated by comparison or guessed at. The best way for a contractor to obtain facts and figures that he can rely upon in estimating is to keep a record of all the work he does. It will not do to trust to memory, for in a few months or a year he will not know whether such and such work cost \$42 or \$54, or what it cost. If he would profit by experience he will keep a record of the cost of his work, so that he can refer to it at a moment's notice. To keep a record that will give the best and most reliable facts and figures prepare a list of all kinds of work, having two sets of money columns, one for estimated cost and one for actual cost.

When estimating a job put down the estimated cost, and when the actual cost is found from experience in doing the work put it down, and keep each particular kind of work or portions of a job separate from the entire job. By so doing one will soon be able to see where he has estimated too high or too low, and will have facts and figures which will enable him to make a proper average. Some parts of a building are easily estimated by the "square," which contains 100 square feet. Some parts are easily estimated by the lineal foot, while other portions are best estimated by the piece. Keep a record of the time required by different men in doing work by the square, lineal foot or piece. In this way one will find the average day's work from actual experience, which is the only plan that can be followed with success.

When it is known what it is worth to do work by the square, lineal foot or piece, any person of ordinary skill in figuring ought to be capable of making an estimate reasonably accurate. As I have said before, the average day's work of all kinds is what is wanted as a standard in estimating. Accordingly I have prepared a table with the average day's work of each kind and the average rates to figure on. The table is made on a basis of ten hours for a day's work and as near as practical to average \$3.50 per day. If an estimate is wanted for nine hours add one-tenth to the price; and if for eight hours add one-fifth. The prices can easily be made for any rate per hour or any number of hours per day. To those who want to test the advantage of a table of this kind I would say, do not take it for granted that my rates and averages are the best in the world, or that they are just the thing for a guide, but prepare a similar list and begin entering rates and averages as they are found from actual experience. Then one will have something that will suit the locality in which he lives, and there can be no doubt that in a short time he will have something that will be much to his advantage in estimating. Let me say however, that the average day's work as found in the table is a reasonable average, as I have found from experience, and considerable dependence can be placed on estimates made from it.