At the up-stream toe of the embankment, two wells or sumps (shown at “S” and “K,” Fig. 4) were cut 10 or 12 ft. deeper than the main trench, which received the water entering the inner toe puddle trench during construction. This water was disposed of partly by pumping and partly by means of the 2-in. branch pipes leading into and from these wells. At their junction (J) a 2-in. stand-pipe was erected, which was carried vertically up through the embankment, and finally filled with cement. The branch pipes from the wells were finally capped and the wells filled with broken stone, as previously mentioned. EMBANKMENT.–As has been said, the upper surface of the slate bed rock was found to be badly fissured, especially near the up-stream toe of the dam, and as the average depth below the surface of the ground was not very great, it was thought best to lay bare the bed rock over the entire upper half of the dam site. Had the depth been much greater, it would have been more economical and possibly sufficient to have put reliance in a puddle trench, alone, for securing a water-tight connection between the foundation and the body of the dam. At the axis of the dam and near the inner toe, where the puddle walls abutted against the hillsides, the excavation always extended to bed rock. Vertical steps and offsets were avoided and the cuts were made large enough for horses to turn in while tramping, these animals being used, singly and in groups, to mix and compact the puddle and thus lessen the labor of tamping by hand. In plan, the hillside contact of natural and artificial surfaces presents a series of corrugated lines, (as is clearly shown in Fig. 4.) After all loose and porous materials had been removed, the stripped surface and the slopes of all excavations were thoroughly wetted from time to time by means of hose and nozzle, the water being delivered under pressure. Fig. 10 is a view of the dam taken when it was about half finished and shows the work in progress. The face puddle shown in Fig. 11 was used merely to “make assurance doubly sure” and was not carried entirely up to the top of the dam. The earth of which the dam was constructed may be described as a red gravelly clay, and in the judgment of the author is almost ideal material for the purpose. Physical tests and experiments made with the materials at different times during construction gave the following average results: