Speaking of the great importance of applying the pressure test, the "Metal Worker" says: "A pipe may be tight and apparently sound, yet of so thin a substance that the least pressure will destroy it or break it through. Joints may be tight at the moment, though barely filled with a thin coating of putty blown out almost at a single breath. Such pipes, thought tight for the moment, are not safe against the slightest pressure, and at any time may be liable to have their continuity broken by a slight jar. The longer we study this subject the more completely do we become convinced that the true remedy for this state of things is a test of the soil pipes by pressure. Scamping is so easily done, and is so difficult of detection, that it seems impossible to avoid it, even in the best jobs which may be constructed. A large proportion of the work is done in difficult situations, where the workman has every temptation to save himself labor and discomfort, and in such situations poor work is the rule rather than the exception. * * * The real objection to such a test is to be found in the fact that it calls for perfect workmanship throughout. It demands just what every house builder and house owner wishes to have, but just what it is very difficult to obtain from even the best plumbing establishments in the city. In gasfitting, which is much less difficult than plumbing work, no sane man would dare to trust a large job without carefully testing it under pressure."

It is sometimes required in practice that each pipe used be tested for soundness at the foundry before coating them. The straight lengths can be tested under pressure more easily than the branches and bends. Hence the oil test is resorted to, and the strength or thickness of the pipe is not by this method made known.

In order to save as many joints as possible these pipes are cast in rather long pieces. The attempts to cast pipe of small diameter, say of two, three or four inches, in the usual lengths of five feet result in a frequent serious inequality in the thickness of the metal.

I was obliged to make, at one time, a number of experiments on other kinds of cast iron jointing in connection with some researches on furnace and boiler construction. One of these furnaces, which was advertised by its builders to be absolutely gas tight, contained quite a number of joints. The owners claimed that so much care was bestowed on each joint that leakage was a sheer impossibility. I particularly objected to the joints between the cast and wrought iron pipes in the construction, but the makers claimed that their method of jointing was peculiar and could not fail. "A fine new furnace* was exhibited to show the excellence of the workmanship. I still objected until challenged by the makers to give proof of any of the numerous furnaces put up by the company having ever leaked gas. Without taking the time to visit any or all of the 500 or more gentlemen whose letters of recommendation adorned the descriptive circular of the firm, I expressed myself satisfied if the fine new sample furnace then on exhibition would itself stand the test.

"With the assurance that I was at liberty to make any reasonable test I pleased, I ordered the furnace to be turned over and water poured into it. To the complete astonishment of the proprietors and of the careful workmen standing around, the water which had been poured in poured out again through nearly every one of the score of careful joints until the furnace seemed to dissolve and float away in its own tears."

•"The Open Fireplace in All Ages," by J. P. Putnam.

The lead caulked joint, when faithfully made, is very expensive in material and labor. The amount of lead required for each joint, including waste, is estimated at about a pound for every inch in the diameter of the pipe. Thus an ordinary four-inch soil pipe, caulked, consumes four pounds of lead in each joint. The average length of time required by a skillful pipe layer to make a single joint is estimated at twenty minutes, not including, of course, the planning of the pipe system or the cutting and general arrangement of the pipe sections for their proper positions, a part of the work which has no connection with the kind of joints used.

Another point which is defective is the direction in which the energy spent in caulking must be applied. The surfaces to be welded together, as it were, are at right angles with the power applied instead of being in a direct line with it, as it should be. Hence a great loss of energy is sustained, and in order to render it possible to apply power in sufficient quantity to do the work it must be applied successively at small portions of the joint at a time instead of simultaneously over the whole. From this results a loss of time, and the peculiar form of the hub renders it necessary that the caulking be done through the medium of a tool by hand, which increases the loss both of time and energy. The edge of the tool cannot be made to fit exactly the space between the bell and spigot, but must be considerably smaller in order to allow for variations of casting and setting. Hence it acts like a blunt chisel, partially embedding itself in the lead at each blow of the hammer instead of exerting a uniform pressure exactly at the points desired. The proper use of the tool under these circumstances requires considerable skill on the part of the workman, and, as the effectiveness of the caulking depends much upon the manner in which the tools are handled, the quality of the joint is largely dependent upon the skill of the operator, whereas it should evidently be entirely independent of skill, the required degree of skill not being always at hand and being more expensive when obtained than machine power scientifically applied. A still further loss of energy is accordingly sustained, inasmuch as it is contrary to the theory of chances that, even presupposing the most perfect skill of manipulation, the precise position and direction best suited to the varying conditions of the metals to be welded, should be given to the caulking iron at every blow. From these conditions we deduce the following law regarding the form of pipe joints, namely: