240 



TRANSPORTATION OF DEEMS BY BUNKING WAITER. 



suspension gives to the work of traction (p. 234) . 

 For estimation of optimum form ratio some- 

 thing should be deducted from the laboratory 

 indication to allow for the greater resistance 

 of the channel walls. 



SUMMARY. 



Natural streams of alluvial type differ from 

 the streams used in the laboratory in ways con- 

 nected with the bondings of their courses and 

 with variations of discharge. The differences 

 affect forms of cross section, the distribution of 

 velocities within the section, and the partition 

 of the load between suspension and traction. 

 The two portions of load are carried at the 

 expense of the stream's energy, each reduces 

 the velocity, and the reduction of velocity de- 

 termines the limit of carrying power. The 

 whole burden of the stream includes not only 

 two divisions distinguished by mode of trans- 

 portation but as many minor divisions as there 

 are grades of debris, and the load carried of 

 each grade reduces the capacity for all the 

 grades. 



These and other complexities make it diffi- 

 cult to apply the laboratory results to natural 

 streams. It is probable that the forms of the 

 laboratory formulas are applicable, with limi- 

 tations, to the comparison of one stream with 

 another, but the availability of the exponents 

 is problematic. There are special difficulties in 

 attempting to use the formulas for the compari- 

 son of capacities of the same stream at different 

 stages, and in such comparisons the tractional 

 load can not be considered by itself, because 

 much material "which is swept along the bed 

 at lower stages is lifted by flood velocities into 

 the body of the current. 



It is thought that the laboratory formulas 

 may be applied to natural streams which 

 are geometrically similar to the laboratory 

 streams -that is, to streams having the same 

 slopes and form ratios and carrying d6bris of 

 proportionate size. The class of streams to 

 which the formulas apply by reason of simi- 

 larity is necessarily restricted, being character- 

 ized in the main by high slopes and coarse 

 debris. It can include few large streams. 



CONCLUSION. 



It was a primary purpose of the Berkeley in- 

 vestigation to determine for rivers the relation 



which the load swept along the bed bears to the 

 more important factors of control. As a means 

 to that end it was proposed to study the mode of 

 propulsion and learn empirically the laws con- 

 .nectirig its output with each factor of control 

 taken separately. The review of results in the 

 present chapter shows that the primary purpose 

 was not accomplished. In the direction of the 

 secondary purpose much more was achieved, 

 and a body of definite information is contrib- 

 uted to the general subject of stream work. A 

 valuable outcome is the knowledge that the 

 output in tractional load is related to the con- 

 trolling conditions in a highly complex manner, 

 the law of control for each condition being qual- 

 ified by all other conditions. 



With the aid of the Berkeley experience it 

 would be possible to avoid certain errors of 

 method and arrange experiments which should 

 yield more accurate measurements of the same 

 general class and it is natural that the experi- 

 menter should feel the desire to do his work over 

 in a better way but I am by no means sure that 

 adequate advantage would reward a continu- 

 ance of work on the same or closely related lines. 

 The complex interactions could be given better 

 numerical definition, but it may well be 

 doubted whether their empiric definition would 

 lead to their explanation. It is possible that 

 the chasm between the laboratory and the river 

 may be bridged only by an adequate theory, 

 the work of the hydromechanist. It is possible 

 also that it may be practically bridged by ex- 

 periments which are more synthetic than ours, 

 such experiments as may be made in the model 

 rivers of certain German laboratories. (See 

 p. 16.) 



The practical applications for results from 

 experiments in stream traction belong almost 

 wholly to the field of river engineering. For 

 the transportation of detritus and related mate- 

 rials by artificial currents, stream traction will 

 rarely be used, because flume traction is more 

 efficient. Our results in flume traction have 

 therefore an immediate practical application, 

 and as they were limited in range the advantage 

 of extending them can hardly be questioned. 



The report on the Berkeley investigation 

 properly closes with this chapter, but there are 

 several by-products which seem worthy of rec- 

 ord. Some of them are presented in the fol- 

 lowing chapter, and others are contained in 

 appendixes. 



