THE OBSERVATIONS. 



19 



channel, five textures of channel bed, six dis- 

 charges, and seven grades of sand and gravel, 

 besides mixtures. There were nearly 300 de- 

 terminations of load. 



APPARATUS AND MATERIAL. 



EXPERIMENT TROUGHS. 



The trough in which most of the experiments 

 were made was of wood, 31.5 feet long, with an 

 inside width of 1.96 feet. The height of the 

 sides was 1 .8 feet at the head and 1 foot at the 

 end, the change being made by a series of steps. 

 Its proportions and general relations are illus- 

 trated by figure 1. The surfaces were planed 

 and painted. At the head, where water 

 entered, the trough was connected with a 

 tank by a flexible joint, a groove of the under 

 side of the trough bottom resting on a* semi- 

 cylindric member of the tank, so as to consti- 



tute a hinge, and the walls of the trough being 

 connected with the sides of the tank by a sheet 

 of flexible rubber. Here also was a gate by 

 which the flow from the tank could be stopped. 

 Close to the opposite end of the trough was a 

 cross trough 1 1 feet long, 2.5 feet wide, and 3 

 feet deep, rigidly attached to the experiment 

 trough and extending below it. A rectangular 

 opening in the bottom of the experiment 

 trough, an opening having the width of that 

 trough, permitted sand in transportation by 

 the current to sink into the cross trough, which 

 contained boxes to receive it. The width of 

 the experiment trough was varied by means 

 of a longitudinal partition which was given 

 various positions. Its width at the end was 

 also varied by means of two oblique partitions, 

 the "outfall contractor," which merged with 

 the sides a few feet from the end and could be 

 adjusted as desired. For certain experiments 



FIGURE 1. Diagrammatic view o( shorter experiment trough, showing relations to stilling tank (A), cross tank (jB), and settling tank (C). 



false bottoms were added, with surfaces spe- 

 cially prepared as to roughness. These will be 

 specifically described in connection with the 

 corresponding observations. A second trough, 

 having the same function as the one just de- 

 scribed was 150 feet long but similar in width 

 and style. Its sides were higher and it was 

 not hinged at the head but remained horizon- 

 tal. By temporary arrangements of parti- 

 tions curved and crooked channels 1 foot 

 wide were constructed within this trough. 

 The shorter trough was installed in the base- 

 ment of the Mining Building of the University 

 of California; the longer one on the campus 

 near by. (See PI. I, frontispiece.) The longer 

 trough was remodeled for the experiments on 

 flume traction. 



A third trough 14 feet long and 0.67 foot 

 wide had its wooden sides replaced for a space 

 of 3.5 feet, at midlength, by plate glass, so 

 that observation could be made from the side. 

 It was provided with a sliding diaphragm, 

 to be described in another place. 



A fourth trough, of iron, was used only in the 

 experiments on flume traction and will be 

 described hi connection with those experiments. 



A few experiments were made also in a 

 trough carrying the waste water of the 150-foot 

 trough. This had a width of 0.915 foot. 



WATER SUPPLY. 



The water was taken from the municipal 

 mains of Berkeley. As it was not practicable 

 to draw freely on this source, a moderate 

 supply was made to serve for a long series of 

 experiments, being stored in a sump and 

 pumped up as required. By repeated use it 

 acquired a certain amount of fine detritus in 

 suspension, but the quantity was not sufficient 

 to obstruct the view of the experiments or 

 rather, when it was found obstructive, a fresh 

 supply of clear water was substituted. 



THE WATER CIRCUIT. 



Starting from the storage tank or sump, 

 the water was lifted by a power pump to a 



