We arrive now at that feature which, of all others in dam construction, is an art rather than a science. Practically all dam foundations are more or less wet, and to handle the water properly while starting the masonry is a process to tax one's experience, ingenuity and patience. Every difficult foundation is a masterpiece. The desired result may indeed be simply and positively stated, namely, keep the water out of and away from the masonry until the masonry has set; but a correspondence course might as successfully attempt to qualify one for painting a Mona Lisa as for building a difficult foundation.

However, some of the obvious steps and processes will be briefly described in the hope that it will assist in an appreciation of the art. Most of the water entering the pit enters along the up- and downstream sides at or near the foot of the slope of the loose material, i.e., above the rock surface. It may be necessary to, in effect, pave portions of the slope with rock from the excavation in order to hold it from flattening out indefinitely and coming into the pit. If the loose material itself contains enough stone of suitable size, a practically equivalent paving will naturally result in a short time from the action of the water coming through the slope. The finer material will be washed in, while the bank ravels and recedes until it becomes stable.

The main body of water should be intercepted at the edge of the rock excavation and led to a sump situated preferably outside of the masonry lines. This can probably be accomplished best by a small masonry wall. Having selected the area, preferably in the lowest part of the pit, upon which to start the masonry, examine the adjoining or surrounding rock bottom in order to determine the source and quantity of water entering the area which it is possible to divert or keep out. This diversion may, like the first, be accomplished by small masonry walls leading to the same or other sumps, or a small masonry dam may divert the water through suitably sized pipes. If the water cannot readily be made to flow to a lower sump, it may be necessary to pump or bail from a small reservoir thus formed.

After all practicable diversion of water from entering the area has been accomplished, some that must be overcome in some other way will remain. Water may still be entering through seams all over the area, the previous operations being in the line of reducing as much as possible the quantity to be finally overcome. In this last process appears the art of the artist or the helplessness of the tyro. The essence of the ultimate element of the process is the actual smothering of the water by the masonry, and the consequent forcing of the water to take a course via the rock seams to remote and higher outlets. Determining considerations are the quantity of masonry to be laid, the height to which it must be brought, and the time required to do it; also the fact that for a short time a given depth of concrete will balance the head of a greater depth of water. No stream of water is so small that its possible effect may be ignored.

The aim should be to handle it so that its effect, if any, will be not under the masonry, but on top or at the side where it can be readily observed and remedied later. Not only that but the builder must know after the operation has been completed that it has been successful.

The effect of moving water in contact with fresh masonry is of course to remove the cement, and if given time enough it will simply leave a mass of sand or gravel or stone without any cement. If forced to find an outlet through a mass of concrete, the water will merely render it nothing but a mass of sand and gravel, while if the water runs over or at the side of the concrete it will simply remove only enough cement so that its channel will be paved with the sand and gravel. The first condition may be more or less complete, and may or may not be recognized, while the second condition will be readily recognized and as readily repaired at some future time. It may often be found desirable to introduce pipes in the masonry to serve as vents or channels for the water. These should be carefully placed so that there will be a free connection between their end and any water-bearing seam; also that they may serve their purpose with the least possible head of water against the masonry. They should be extended carefully by adding short lengths at a time till the masonry around their inlet end has acquired some strength. The stream of water itself may be trusted to keep free its own passage into the pipe, and (if the pipe is properly set) with an insignificant surrounding area of bottom affected by the moving water. The various water-bearing seams and their obvious or obscure connections should be carefully observed in order to determine how the water may be successfully backed through them. Depending upon the size and configuration of the area first to be built upon, it may be desirable that all of the higher part be built upon first, thus chasing the water to the lowest part, which may be easily done. In the area thus first covered introduce the necessary pipe vents, then, when the masonry has hardened sufficiently, concentrate upon the relatively small lower area, bringing it up in a short time, to such a height that the course of the water will be reversed so as to outlet through the vents previously prepared. In an operation like the above, the element of time is an important consideration in connection with the rate of flow. Obviously, some appreciable period of time is required for the water to produce a harmful effect; therefore reverse the stream before it gets that time.

Forty cu. yd. per hour of masonry construction may accomplish successfully in one hour what 5 cu. yd. per hour for a month might not do. Under such circumstances it may be well to mix the mortar or concrete considerably drier and richer in cement than would be otherwise advisable or necessary in order that it may absorb some of the entering water and even lose some of the cement without harm. It has been attempted above to outline only some of the more common and obvious methods of procedure. In actual practice other expedients, as well as many combinations and variations of the above, may be resorted to.

With the gradual spreading and raising of the masonry until the water-bearing bottom is covered, the various vent pipes should also be brought up. No attempt should be made to discontinue them till there is abundant weight of masonry to hold whatever pressure might result from closing them. While this elevation may be a matter of conjecture, it should be at least well above the original surface of the rock. While these pipes are being brought up a most excellent opportunity is afforded for studying the water-bearing courses under the dam, in order to ascertain the connections between the various seams and how best to discontinue the pipes by grouting or otherwise. The streams from some of the pipes may very readily respond to any raising or lowering or shutting off of the outlets of adjacent pipes, and a free connection of seams shown most clearly. Others may respond more slowly or even not at all.