that the night shift is working under no handicap, having artificial light as good as daylight, and even better in that the attending heat is absent.

The camp buildings present an imposing picture after dark, for then it is that one realizes the extent of this camp, which is really more like a small compact, well-kept town than an ordinary construction camp. Indeed, what town of equal size-say 1,500 peoplecan boast the advantages that this camp offers: Electric lights, railroad, complete water supply from a pure mountain stream, sewerage system with septic tank, ice plant, meat shop, general store, ice cream parlor, soda fountain, picture shows, school, restaurant, and all conducted under strict sanitary regulations. 'Tis said there was a fly seen in camp in 1912, but he was captured after an exciting chase by the sanitary squad, and is now on exhibition along with some of the curios found in the excavation.

Another meal at 10 p. m. ("noon" for the night shift), and after 45 minutes' rest everything starts up again for the "afternoon" run. Earlier in the evening there have been many visitors and sightseers out around the work, mostly the men on the day shift with their families, or some of the young men who have found attractive companions from among the daughters or sisters of their fellow-workmen; or perhaps some of the local people who have friends visiting from out

side. They all like to see the work at night. But when the "afternoon" shift starts, at 10.45 p. m., most of the visitors have left, and from this time until morning is probably the only time during the whole day when the job is entirely free from spectators.

The last bucket of concrete is placed about 2.45 a. m. The locomotives have already come in and are being coaled up; watchmen at various points over the work have been on duty since dark. The crew at the messhouse is still at work, for there are two more meals to serve yet, before the 6.30 breakfast for the day shift.

Repair men are lined up to overhaul some equipment that can not wait until Sunday for its repairs. The rock drills of the graveyard shift are still pounding away at their work, for by morning there must be some of this rock "shot up" ready to be "mucked out" by the day shift.

We have seen it all, we think, and plan to turn in for a little sleep before we board "Uncle Sam's own railroad" in the morning, to be taken back to the outside world. Although we have put in a long day, there has been so much to see and so many different branches of the work to watch, that the time has gone faster than we imagine. But as we come through camp we meet the early train crew, going over for its 5 o'clock breakfast, and realize that even before one day's work is finished another's has begun.

thin impervious diaphragm placed in the interior of a dam section composed of more or less pervious materials. The term "core wall" is applied in engineering literature to walls equal to or greater in height than the dam itself and also to much lower ones which might more properly be called cut-offs.

In common with other features pertaining to dams, core walls are the subject of much technical discussion; the different types adopted for similar conditions by men of equal ability indicate a wide divergence of opinion concerning them and illustrate most pointedly that what one authority considers correct another considers wrong.

It is quite generally conceded, however, that the use of core walls of any type is theoretically inefficient. The most desirable dam sections theoretically are (a) one entirely composed of impervious material, (b) one whose upstream half or third is thus composed, or (c) one having the upstream face impervious. In many cases the available material is such that none of these sections can be economically constructed. In such cases core-wall sections generally offer practical working solutions.

Core walls are divided into two general types-those built of masonry, concrete, or rubble, and those of "puddled" material. Combinations of both are used as well as diaphragms of steel and wood. In dams properly constructed of suitable material by the hydraulic method a comparatively large part of the central portion is automatically made the most impervious and becomes in effect a large puddle core.

Both of the two general types have inherent advantages and disadvantages, and both have been successfully used under similar conditions. Both are criticized because of their tendency to produce supersaturation in the upstream portion of the dam, a condition which may require either flatter upstream slopes or additional weight such as loose rock to prevent slipping. The masonry type is proof against attacks by burrowing animals; the puddle type is not. This is a very important advantage if the dam will have only occasional inspection. A crack in a masonry wall is not attended by so much danger as a breach or puncture in a puddle core, as it is most probable that the crack will tend to become sealed, whereas in the puddle the breach will tend to enlarge. The masonry wall is better adapted than the puddle to making connections with outlet conduits, rock or masonry

also, if placed on line with the upstream edge of the crown, can be readily extended above the top of the aam to form a parapet and additional freeboard thus secured.

On the other hand, as the puddle core is flexible it is less liable to be ruptured by unbalanced pressures in the dam than the more rigid masonry wall. A puddle core section is more nearly homogeneous than one with a masonry wall and experience seems to indicate that they are somewhat more impervious. It is stated frequently that a masonry wall requires firm rock for the foundation. By correctly designing the footings, however, they may be built successfully upon softer materials. As puddle cores are flexible, they may be founded upon almost any material.

Puddle construction in general requires the exercise of good judgment and great care in the selection of materials and skill in their mixing and placing. This is especially the case in the construction of puddle cores by the hydraulic method. The horizontal pressures produced by the semifluid mass of the core must be resisted by the more stable material on each side. These pressures are not completely determinate and hence conservative assumptions in design and special care in construction must obtain to avoid slides. The construction of a masonry core is more in line with ordinary high-grade work.

Core walls constructed of rubble masonry have proven only comparatively impervious. In 1901 a board of engineers investigated several of the earthen dams of the New York Board of Water Supply.

These dams were all built with rubble masonry cores. The investigation showed the presence of water below the cores in all of the dams. The loss of head attributed to the core wall varied from 7 to 40 feet, the average being about 21 feet, which was the amount the board believed could be assumed safely in other designs.

According to Parker's "Control of Water," a wellmade puddle core produces a drop of 20 to 30 feet in the line of saturation. This is confirmed by tests made in the Wachusett (North) Dike, which indicated a loss of about 30 feet.

The loss of head may, of course, be increased by securing greater density in the core itself, but it is generally admitted that no matter which type is used or what precautions are taken, the core itself should be considered as comparatively and not completely im

pervious and that the necessity for drainage provisions in the downstream portion of the dam is not entirely removed by the adoption of cores of either masonry or puddle.

Most authorities agree that puddle cores (not hydraulicked) should be composed of clayey material mixed with sand and gravel, the theory being the same as in proportioning concrete; that is, to fill the voids between the larger components with the particles of the smaller. In "Irrigation Engineering," by A. P. Davis and H. M. Wilson, the following proportions are given as forming a satisfactory core:

"These proportions, when well mixed and compacted with a small quantity of water and rolled, can be reduced to about 1% cubic yards in bulk.”

Mr. J. T. Fanning, in "Treatise on Water-Supply Engineering," makes the sand and the clay content 0.15 and 0.20 cubic yard, respectively, using the same proportions of fine and coarse gravel. Mr. E. Wegman states in "Design and Construction of Dams" that "pure clay is not suitable material for a puddle core, as it swells when wet and shrinks and cracks when dry, making its use very dangerous in any part of a dam where it may be alternately wet and dry."

The dimensions of puddle cores must be determined by judgment and experience. In existing structures the dimensions as well as the shapes vary greatly. The following dimensions for hand-placed cores are proposed by Mr. Wegman: Four to eight feet thick at high-water line, both sides battered uniformly so that at the ground surface the thickness shall be one-third the head. The thickness at the bottom of the trench should be at least one-half that at natural ground surface, but not less than 4 or 5 feet. In the large hydraulic fill dams of the Miami conservancy district the width of the puddle core at any point is equal to the height of the dam above that point.

The excavation below the natural ground surface should be made wedge-shaped in cross section, as a tighter connection will thus result when the puddle is placed than if the sides be vertical or stepped.

The dimensions of masonry core walls are also largely matters of judgment and experience. One method, although not a complete analysis, which has been suggested as an aid to the judgment is as follows: Let A the portion of the area of the section upstream from the core wall and above the angle or repose of the saturated material.

The thickness assumed as required for percolation may or may not agree with the above and is in turn a matter about which authorities differ. Mr. Clemens Herschell gives 4 or 5 feet for the bottom thickness, 8 at the ground line, and 4 at the top. Mr. Wegman gives 211⁄2 to 6 feet thickness at the top that at the ground surface from one-sixth to one-seventh the head, the increase being equal on each side and accomplished either by batters or by offsets 10 feet apart. Other authorities strongly condemn the use of offsets and insist on straight batters.

Diaphragms of wooden sheathing have been used occasionally instead of puddle or masonry. This type can not be considered as permanent. The plane between the planks and earth can not be made as tight as with masonry or puddle; consequently there is more danger of leakage. For temporary construction and also in low dams where the impounded water has a high silt content which may be expected to seal the dam a wooden core might be favorably considered, especially where extreme economy is first cost is essential, but one should never be placed in an important permanent structure.

Steel plate diaphragms coated with asphalt have been used in a few cases. They can be constructed so as to be completely impervious at first and if well coated might be expected to last for a long time. They can not, however, be considered as permanent as masonry or puddle. Their economy is, therefore, questionable and their use is not increasing.

The masonry core walls built by the Reclamation Service are all of concrete and in general are comparatively thin sections. The highest in service at the present time is in the Strawberry Dam, Strawberry Valley project, Utah. The maximum section of this wall is shown in the accompanying illustration.

The Tieton Dam, Yakima project, Washington, now under construction' will be the highest earth dam built by the Reclamation Service. It will be an earth and rock fill section, approximately 220 feet high with both a concrete core wall and a hydraulicked puddle

1 The Tieton Dam was completed in May, 1925.

portion of sufficient thickness, due to the decreasing percentage of fines and the narrowness of the pond. Loam therefore was borrowed from the valley floor and then sluiced into place between wooden forms which were placed to prevent strata of sand being washed in from the other material.

This puddle core is 8 feet thick at its base 14 feet above the base of the dam; the thickness at its top is 5 feet. In section this core is not vertical, but slopes upstream about 30 feet from its top, which is under the upstream edge of the crown. Its vertical height is about 40 feet.

The

The Bumping Lake Dam was built by hauling the material on to the dam, then separating and consolidating it by hydraulicking, the fines being sluiced toward the center and forming a puddle core. puddle was built by this method up to about spillway level, a distance of about 50 feet from the bottom. At this elevation similar difficulties in the hydraulic method developed as on the Conconully Dam and the rest of the puddle material was dumped and worked by shovels. The maximum height of the Bumping Lake core is about 60 feet. The horizontal dimensions are variable; in the hydraulicked portion they range from about 20 feet near the bottom to about 2 feet at the spillway elevation. The portion placed by hand averages about 4 feet thick.

The core walls and puddle cores constructed by the Reclamation Service have proven adequately tight, and so far no trouble due to saturation has developed in the upstream portions of the dams in which they are built. Somewhat extended reading indicates that the above is generally true of other well designed and constructed dams. Both types have been in service for years and doubtless will continue to be used in the future, offering, as stated before, practical and economical solutions when more theoretically correct sections are not feasible.

A most unusual core was built in the Sherburne Lakes Dam by the Reclamation Service as a part of the drainage system of the dam. This drain consists of a core of screened gravel, 5 feet thick at the top, 10 feet thick at the bottom, and about 83 feet high, with side slopes of 3:100. This core is placed approximately under the downstream edge of the crown and extends the full length and height of the dam. The most impervious material in the dam is placed above the core and the coarser below. A cast-iron pipe with open joints is in the bottom of the core and connects with transverse discharge drains, which lead back into the stream.

This construction thus far has proven satisfactory, although it has not yet been thoroughly tested, having been subjected to a head of only 45 feet and being designed for 68 feet.