CHAPTER XIV. PROBLEMS ASSOCIATED WITH RHYTHM. 



RHYTHM IN STREAM TRANSPORTATION. 



This chapter is concerned with certain prob- 

 lems upon which the Berkeley investigation 

 touched, but which were not seriously attacked. 



The low precision of the observations on 

 stream traction, a precision characterized by an 

 average error of about 11 per cent and an. aver- 

 age probable error after adjustment between 2 

 and 3 per cent, had for its chief cause the failure 

 to eliminate from the experiments the influ- 

 ence of rhythm. The slope of the water sur- 

 face, the slope of the channel bed, and the load 

 of debris transported were all subject without 

 intermission to rhythmic fluctuations. If ma- 

 terially better observations of the same sort are 

 to be made, this difficulty must be successfully 

 dealt with, and the first step toward mastering 

 it is to understand it. The removal of a diffi- 

 culty, however, is neither the sole nor the most 

 important result to be expected from the study 

 of fractional rhythms. Underlying them are 

 physical principles which are of importance in 

 the dynamics of rivers, and their study con- 

 stitutes one of the available lines of approach 

 to the broader subject. 



For their empiric study the general plan of 

 the Berkeley apparatus is well adapted, but our 

 experience indicates that certain details should 

 be modified. The use of a long trough is ad- 

 visable, with contraction at the outfall, and 

 with delivery of the load to a settling tank be- 

 yond the outfall. The appliances and methods 

 should be such as to secure uniformity in dis- 

 charge, in character of d6bris, and in rate of 

 feed. 



The apparatus for regulating discharge, de- 

 scribed on pages 20 and 257, was one of the most 

 satisfactory parts of the Berkeley equipment. 

 Its most important feature, as affecting pre- 

 cision, was the delivery of the water through an 

 aperture under a considerable head. 



Uniformity of debris can hardly be secured 

 without the employment of an artificial, nar- 

 rowly limited grade, and the available means 

 of sorting is the sieve. It is to be observed, 

 however, that after a grade has been separated 



by sieves it is still subject to sorting by current, 

 the current recognizing differences of form and 

 density which the sieves ignore. When d6bris 

 that has once been handled by the stream is to 

 be used a second time, remixing may be ad- 

 visable. 



None of the devices we employed to feed 

 debris to the current achieved uniformity. 

 Those which depended on the flow of wet de- 

 bris through an aperture failed because the 

 proportion of water could not be kept constant. 

 The others depended on handwork and ex- 

 perienced the irregularity usual to handwork. 

 An apparatus planned near the end of the ex- 

 perimental work, but never tried, is of such 

 promise that its essential features are here 

 described. 



20921 No. 8614- 



-16 



FIGPBE 72. Suggested apparatus for automatic feed of debris. 



A drum, A in figure 72, is turned slowly by 

 power, its rate being regulated by clockwork. 

 Its position is above the trough containing the 

 experiment stream, B. The surface of the 

 drum is uniformly roughened. Above it is a 

 vertical rectangular shaft, filled with moist 

 debris. The shaft does not touch the drum. 

 The width of the separating space at D is con- 

 trolled by some suitable device. As the drum 

 turns, a debris layer of uniform thickness is 

 carried with it, and this falls into the stream. 

 To prevent irregularities due to adhesion, a 



241 



