FLUME TRACTION. 



211 



already drawn (p. 207), that the transportation 

 of fine debris over a fixed bed of coarser debris 

 particles is essentially of the nature of stream 

 traction. It may fairly be inferred that if we 

 were able to extend this curve into the region 

 of debris coarser than the gravel of the bed, a 

 minimum would be developed. 



If the curves were to be traced toward the 

 right, by means of additional experiments with 



1,400 



200 

 Linear fineness 



600 



FIGURE 69. Curves illustrating the relation of capacity for flume trac- 

 tion to fineness of de'bris. Data from bed of planed wood are recorded 

 by crosses; from wood-block pavement by circles; other data by dots. 



finer de'bris, there can be little doubt that they 

 would be found to continue their ascent ; but 

 eventually, as curves of traction, they would 

 come to an end with the passage of the process 

 of transportation from traction into suspension. 

 In the opposite direction they may be conceived 

 to attain a maximum and then drop suddenly 

 to the base line; for despite the law of increase 

 of capacity with coarseness, there must be a 

 degree of coarseness for winch the force of the 

 current is not competent, and when that is 

 reached the ordinate of capacity becomes zero. 

 The position of this limit, which I have in earlier 

 pages called competent fineness, evidently 

 depends on slope and discharge, as determining 

 the force of the current, and on the degree of 

 rounding of the de'bris. 



The double ascent of the curve of flume 

 traction is susceptible of plausible explanation, 

 by means of considerations connected with the 



process of rolling. The process of rolling in- 

 volves a question of space. Each rolling pebble 

 occupies an area of the channel bed somewhat 

 larger than its sectional area, even if the 

 pebbles are arranged in the closest possible 

 order. If we conceive the channel bed to be 

 occupied by rolling pebbles of a particular size, 

 separated by spaces which bear a definite ratio 

 to the diameters of the pebbles; and again 

 conceive it to be occupied by rolling pebbles 

 of a larger size, with the same ratio between 

 interspace and diameter; it is evident that the 

 total volumes or masses of pebbles in the two 

 cases will be proportional to the diameters. 

 If the larger pebbles have twice the diameter 

 of the smaller, then a given area of bed will 

 contain twice as much rolling load of the larger 

 pebbles as of the smaller. It is also true, as 

 stated on an earlier page, that the rolling speed 

 is somewhat greater for larger pebbles than for 

 smaller. The tendency of these two factors is 

 the same, to make the load greater for large 

 particles than for small, when the process of 

 transportation is rolling. The analysis is doubt- 

 less too simple the degree of crowding on the 

 bed, for example, may not be the same for 

 different sizes, and the degree of crowding may 

 affect the speed of rolling but qualifying 

 factors can hardly impair the qualitative 

 inference that the rolling load increases with 

 coarseness. 



It is a matter of observation that, under 

 similar conditions determining force of current, 

 the dominant process in flume traction is for 

 coarse debris rolling and for fine debris salta- 

 tion. When the process is rolling, as just 

 shown, capacity increases with coarseness of 

 de'bris. When it is saltation, as illustrated by 

 the body of experiments on stream traction, 

 capacity increases with fineness. With the 

 passage from saltation to suspension the effect 

 is even heightened, and it is probable that in 

 a number of the recorded experiments the 

 process was largely that of suspension. Thus 

 the double ascent of the capacity-fineness curve 

 is determined by the distinctive properties of 

 two (or three) modes of propulsion. 



If the preceding explanation is well founded, 

 the nature of the law connecting capacity with 

 the degree of comminution of the debris in any 

 particular case depends on those conditions 

 which determine the dominant process of con- 

 veyance. If the channel bed is smooth, and if 



