ABSTRACT. 



Scope. The finer debris transported by a 

 stream is borne in suspension. The coarser is 

 swept along the channel bed. The suspended 

 load is readily sampled and estimated, and 

 much is known as to its quantity. The bed 

 load is inaccessible and we are without definite 

 information as to its amount. The primary 

 purpose of the investigation was to learn the 

 laws which control the movement of bed load, 

 and especially to determine how the quantity 

 of load is related to the stream's slope and dis- 

 charge and to the degree of comminution of the 

 debris. 



Method. To this end a laboratory was 

 equipped at Berkeley, Cal., and experiments 

 were performed in which each of the three con- 

 ditions mentioned was separately varied and 

 the resulting variations of load were observed 

 and measured. Sand and gravel were sorted 

 by sieves into grades of uniform size. Deter- 

 minate discharges were used. In each experi- 

 ment a specific load was fed to a stream of 

 specific width and discharge, and measurement 

 was made of the slope to which the stream 

 automatically adjusted its bed so as to enable 

 the current to transport the load. 



The slope factor. For each combination of 

 discharge, width, and grade of de'bris there is a 

 slope, called competent slope, which limits 

 transportation. With lower slopes there is no 

 load, or the stream has no capacity J for load. 

 With higher slopes capacity exists; and 

 increase of slope gives increase of capacity. 

 The value of capacity is approximately propor- 

 tional to a power of the excess of slope above 

 competent slope. If S equal the stream's slope 

 and a equal competent slope, then the stream's 

 capacity varies as (S a) n . This is not a de- 

 ductive, but an empiric law. The exponent n 

 has not a fixed value, but an indefinite series 

 of values depending on conditions. Its range 

 of values in the experience of the laboratory 



1 Capacity is defined for the purposes of this paper as the maximum 

 load of a given kind of de'bris which a given stream can transport. See 

 page 35. 



10 



is from 0.93 to 2.37, the values being greater 

 as the discharges are smaller or the de'bris is 

 coarser. 



The discharge factor. For each combination 

 of width, slope, and grade of de'bris there is a 

 competent discharge, . Calling the stream's 

 discharge Q, the stream's capacity varies as 

 (Q K). The observed range of values for o 

 is from 0.81 to 1.24, the values being greater 

 as the slopes are smaller or the debris is 

 coarser. Under like conditions o is less than 

 n; or, in other words, capacity is less sensitive 

 to change3 of discharge than to changes of 

 slope. 



The fineness factor. For each combination 

 of width, slope, and discharge there is a limit- 

 ing fineness of de'bris below which no transpor- 

 tation takes place. Calling fineness (or degree 

 of comminution) F and competent fineness <j>, 

 the stream's capacity varies with (F 0)P. 

 The observed range of values for p is from 0.50 

 to 0.62, the values being greater as slopes and 

 discharges are smaller. Capacity is less sensi- 

 tive to changes in fineness of de'bris than to 

 changes in discharge or slope. 



The form factor. Most of the experiments 

 were with straight channels. A few with 

 crooked channels yielded nearly the same esti- 

 mates of capacity. The ratio of depth to width 

 is a more important factor. For any combi- 

 nation of slope, discharge, and fineness it is 

 possible to reduce capacity to zero by making 

 the stream very wide and shallow or very nar- 

 row and deep. Between these extremes is a 

 particular ratio of depth to width, p, corre- 

 sponding to a maximum capacity. The values 

 of p range, under laboratory conditions, from 

 0.5 to 0.04, being greater as slope, discharge, 

 and fineness are less. 



Velocity. The velocity which determines 

 capacity for bed load is that near the stream's 

 bed, but attempts to measure bed velocity 

 were not successful. Mean velocity was meas- 

 ured instead. To make a definite comparison 

 between capacity and mean velocity it is neces- 



