The working periods, and the causes for the various delays when work was suspended, were as follows: Contract let September 23, 1905, winter of 1905-1906 built flume, flood destroyed flume June 13, 1906, after which other contractors took over the work and rebuilt flume. Flood July 5, 1907, took out 100 ft. of dam, repairs made so that excavation was started December 2, 1907, and completed April 1, 1908. First concrete laid March 30 to April 12,1908. Flood till April 26,1908. Concrete April 29 to May 2,1908. Flood conditions till August 28, 1908. Concrete September 25 to November 30,1908. Severe winter weather till March 16,1909. Concrete March 16 to 29, 1909. Flood March 29 to September 1, 1909. Concrete September 1 to January 16, 1910, completed. Thus during something over four years elapsed time, one winter season of four months saw the excavation all taken out once, and a total of 202 working days divided into five or six periods was required to build all of the masonry. It is evident that between the winter seasons on one hand and the flood seasons on the other, complicated further by the necessity for repairs after the floods, the actual working periods were short and precarious, and might easily have been rendered ineffective by a slight adverse departure from actual conditions. Depending on the value of time in completing the dam, it might have been economical to spend much more time and money on diverting works that would have handled the floods, and then built the dam in one year.

At the Pathfinder dam the river diversion was accomplished through the permanent outlet tunnel. (See Fig. 52.) This tunnel was 480 ft. long, 13 ft. wide, 9 ft. side walls, 10 ft. to arch, lined with concrete and cost about $33,000. The width of the canyon at river level was about 90 ft., and the bottom of the foundation was only 20 ft. below. The water was diverted into the tunnel by means of a rock fill cofferdam. As this was constructed during the winter when much ice was running in the river, it was found to leak considerably with the advent of warm weather. The leakage was intercepted by sand-bag dams and conducted across the site in a small flume.

1 Originally planned 200 ft. long X 12 ft. high, contract price with flume was $37,000. See page 249.

At the Hales Bar dam recently completed for the Chattanooga & Tennessee River Power Company, rather peculiar and very adverse conditions were encountered. The dam is some 1200 ft. long and 37.5 ft. in height above low water. There was about 5 ft. depth of water, and below that about 10 ft. of loose material overlaying a bottom of limestone which contained many seams and fissures. For portions of the length of the dam, extending from each end toward the center, the foundation was satisfactorily unwatered by surrounding them with ordinary crib cofferdams as described under Type 1 of diversion works. (See Plate V).

The bottom of the river was entirely destitute of any fine, banket-ing material which would have in large measure sealed the fissures. Further, for somewhat over half of the total length of the dam the fissures carried so much water that it was impossible to prosecute the work by ordinary methods. The method finally adopted was unique in its application to such a situation, was entirely successful and reflects credit upon those who conceived and executed it. Briefly the method was to sink reinforced concrete caissons which were later incorporated with the permanent structure of the dam. The caissons were usually 70 ft. up- and downstream by 54 parallel to the dam; though for a portion of the work two rows of smaller caissons were used. The cutting edge is beveled inside at an angle of 45 deg., so that the roof of the working chamber which is 4 1/2 ft. above the edge has dimensions 9 ft. less each way. The caissons were sunk through the loose material and into the rock to a maximum depth of 36 ft. below low water to the cutting edge; though after preparing the bottom, and in treating the crevices, concrete was put in for some 10 ft. or 12 ft. below that elevation. Concrete was added on top as the caisson was sunk, and on reaching and preparing an acceptable bottom the entire working chamber was filled with concrete. The caissons were started upon loose material above the water level, a crib structure surrounding the site first having been built and filled with sand and gravel. The crib served to hold the loose material from being washed away by the current.