26 



TRANSPORTATION OF DEBRIS BY RUNNING WATER. 



estimate of depth was obtained by subtracting 

 one profile from the other. This mode of de- 

 termination avoided the error incident to the 

 gage work and was on the whole satisfactory, 

 but unfortunately the number of experiments 

 to which it could be applied was not large. 

 During the greater part of the experimentation 

 the importance of the water profile was not 

 recognized, and this particular use of it was 

 essentially an afterthought. 



The values from profiles being assumed to 

 have relatively small errors, both systematic 

 and accidental, it is possible to measure by 

 their aid the precision of the values from gage 

 readings. Of 118 depths which were meas- 

 ured by both methods, the gage gave the 

 greater value for 36, the lesser for 78 ; and the 

 average for gage values was 0.0045 0.0007 

 foot less than the average for profile values. 

 Independently of this apparent systematic er- 

 ror, the probable error of a single measure- 

 ment with the gage was 0.007 foot. 



MEASUREMENT OF VELOCITY. 



The mean velocity of the current is compu- 

 ted by dividing the discharge by the area of 

 the cross section, or the product of width and 

 depth. Its precision depends on those of the 

 determinations of discharge, width, and depth; 

 and as the precision for discharge and width 

 is relatively very high, the precision of mean 

 velocity may be regarded as identical with that 

 of depth. 



The attempts to measure velocity close to 

 the channel bed were not successful. This is 

 much regretted, because it is believed that bed 

 velocity is a prime factor in traction and that 

 slope and discharge exert their influence chiefly 

 through bed velocity. The mode of measure- 

 ment to which most attention was given was 

 that by the Pitot-Darcy gage, and special 

 forms of that instrument were constructed for 

 the purpose. The difficulty which seemed in- 

 superable was essentially the same as that en- 

 countered in the measurement of depth. As 

 the instrument approached the current-molded 

 bed of ddbris, the bed retreated, with the for- 

 mation of a hollow. In the presence of the 

 instrument the normal velocity at the bed did 

 not exist. Inseparable from this difficulty is 

 a property of the instrument. When it is held 

 close to the bottom or side of a channel its con- 

 stant is not the same as in the free current. 



The system of flow lines and velocities with 

 which the stream passes the obstructing object 

 determines the instrument's constant, and when 

 that system is modified by a neighboring ob- 

 ject the constant changes. The nature of these 

 difficulties is such that it was not thought 

 worth while to experiment with other gages 

 and meters which limit the freedom of the 

 current. 



Other devices tried were of one type. Small 

 objects, such as currants or beans, only slightly 

 denser than water, were placed in the current 

 and watched. The lighter ones would not 

 remain near the bottom. The heavier ones 

 were visibly retarded when they touched the 

 bed and were also retarded when close to the 

 bottom by the cloud of saltatory sand, which 

 has a slower average velocity than the water it 

 suffuses. 



MODES OF TRANSPORTATION. 



MOVEMENT OF INDIVIDUAL PARTICLES. 

 ROLLING. 



In stream traction sliding is a negligible fac- 

 tor. The roughness of the bed causes particles 

 that retain contact to roll. When, as in most 

 of the experiments, the grains are of nearly 

 uniform size, each moving grain has to sur- 

 mount obstacles with diameter like its own, 

 and when it reaches the summit of an obstacle 

 it usually possesses a velocity which causes it 

 to leap. So rolling is chiefly the mere prelude 

 to saltation. With mixed d4bris the same is 

 true for the finer grains, but the coarser may 

 roll continuously over a surface composed of 

 the finer, and the coarsest of all, those close to 

 the limit of competence, move solely by rolling. 



The large particle, as it rolls over the bed of 

 smaller particles, indents the bed, and its con- 

 tact involves friction. The energy thus ex- 

 pended comes from the motion of the water, 

 and its communication depends on differential 

 motion between water and particle. Except 

 under special conditions, to be mentioned later, 

 the load travels less rapidly than the carrier, 

 and it is also true that in a load of mixed debris 

 the finer parts outstrip the coarser. 



SALTATION. 



In stream traction the dominant mode of 

 particle movement is saltation. Because salta- 

 tion grades into suspension it has often been 



