THE OBSERVATIONS. 



21 



SETTLING TANK. 



The function of the settling tank was to 

 catch sand which was carried past the cross 

 trough. It was fitted with a system of parti- 

 tions providing two alternative courses through 

 which the stream could be turned, and with a 

 hinged partition the "deflector" by which 

 the diversion was made. 



GAGE FOR DEPTH MEASUREMENT. 



A frame resting on the experiment trough 

 bore in vertical position a slender brass rod. 

 This was raised and lowered by rack and 

 pinion, and its relative height could be read on 

 a scale. Depth of water was measured by 

 reading the scale first with the rod end at the 

 water surface and again with it at the d6bris ' 

 surface. 



LEVEL FOR SLOPE MEASUREMENT. 



A surveyors' level stood a few yards from the 

 trough, about equally distant from the ends, 

 and was used, with a light rod, to measure 

 relative heights for the determination of slopes 

 of water surface and sand surface. 



PITOT-DARCY GAGE. 



A pressure-gaging apparatus of the Pitot- 

 Darcy type, but of special pattern, was used 

 to measure velocities of current. Its two aper- 

 tures were directed severally upstream and 

 downstream. Its external form was designed 

 to give the least possible resistance to the cur- 

 rent. The reading scale, with rubber-tube 

 connection, had a fixed position, while the 

 receiving member was moved from point to 

 point. A fuller description is contained in 

 Appendix A (pp. 251-256). 



SAND ANP GRAVEL. 



The d6bris used in the experiments was 

 obtained from three streams Sacramento 

 River 7 miles below the mouth of the American, 

 American River 8 miles above its mouth, and 

 Strawberry Creek in Berkeley. The de"bris 

 from the creek was relatively coarse and was 

 used only in the experiments on flume trac- 

 tion. The mean density of the river material 

 was 2.69; that of the creek gravel 2.53. The 

 forms of the grains of sand are shown in Plate 

 II. To prepare the debris for use it was sorted 

 into grades by a system of sieves, and in the 

 laboratory records each grade was designated 

 by the limiting sieve numbers. Thus the 



grains of the 40-50 grade passed through a 

 sieve with 40 meshes to the inch and were 

 caught by a sieve with 50 meshes to the inch. 

 For the sake of brevity the grades are com- 

 monly indicated in this report by letters in 

 parentheses (A), (B), etc. and the same 

 notation is extended to mixtures of two or more 

 sizes. Neither of these notations, however, is 

 suited for the mathematical discussion of the 

 laboratory data, and three others were devised. 

 These are, first, the mean diameter of particles, 

 designated by D; second, the reciprocal of the 

 mean diameter, or the number of particles, side 

 by side, in a row 1 foot long, designated by F; 

 third, the number of particles necessary to 

 occupy, without voids, the space of 1 cubic 

 foot, designated by F t . In the sense that the 

 notation of D distinguishes by magnitudes, the 

 notations of F and F 2 distinguish by mini- 

 tudes. F is otherwise called linear fineness, 

 and F 2 bulk fineness. 



To determine the several constants for a 

 grade of debris, a sample was weighed and its 

 particles were counted. Then, N being the 

 number of particles in the sample, W their 

 weight, G their density, and Wo the weight of 

 a cubic foot of water, 



_ NOW* 

 W 



Defining mean diameter as the diameter of a 

 sphere having the volume of the average 

 particle 



/*, 



In the following table the grades of sand and 

 gravel are characterized by the several nota- 

 tions. 



TABLE 1. Grades of debris. 



