APPLICATION TO NATURAL STREAMS. 



231 



Information as to suspended loads is fairly 

 abundant, but there are no satisfactory data 

 as to the complementary tractional loads. 



The character of the partition of load depends 

 in each particular instance on the relative 

 proportions of the various grades in the debris 

 with which the stream is supplied. It varies 

 from point to point along the course of the 

 same stream. In the case of Yuba River the 

 variation was duo to the fact that exceptional 

 conditions created by hydraulic mining had 

 thrown the bed profile out of adjustment; and 

 many other streams are dealing with new and 

 man-made conditions; but natural streams 

 also have strongly contrasted load conditions 

 in different parts of their courses. These arise, 

 first, from the diversity of detritus furnished 

 by tributaries, and, second, from the gradual 

 comminution of the load as it is borne along. 



The partition of load is greatly modified by 

 variation of discharge and by temporary con- 

 ditions of debris supply. A number of changes 

 due to varying discharge have already been 

 mentioned. The changes in velocity affect the 

 grade of fineness marking the plane of partition. 

 With reduction of discharge from a flood stage, 

 the conditions of partition come to differ in 

 pools and over shoals, and usually suspension 

 ceases altogether, while traction is still con- 

 tinued on the shoals. With flood discharge, 

 also, the partition is probably not quite the 

 same for deep and shoal, a portion of the load 

 traveling by suspension through the deeps but 

 by traction over the shoals. 



When a flood is occasioned by heavy rains 

 the fine particles of soil are washed to the 

 stream, with the result (!) that the suspended 

 load is relatively large and (2) that the limit- 

 ing grade is relatively fine. When an equal 

 flood is caused by snow melting, without rain, 

 the suspended load is smaller and the limiting 

 grade coarser; in extreme cases there may be no 

 suspension whatever. While a river is at low 

 stage and without load, a local shower may wash 

 to it a temporary supply of debris of which only 

 the finer part will be immediately transported, 

 giving suspension without traction. 



The partition of load also varies greatly from 

 point to point in the same cross section, com- 

 petence for ^suspension and traction both 

 responding to the control of velocity. 



Despite these variable factors, it is quite 

 possible that in the entire load of a stream 



largo enough to be called a river there is a 

 fairly definite ratio between suspension and 

 traction.. Knowledge of that ratio would be of 

 practical importance to the engineer and geolo- 

 gist, because the measurement of the tractional 

 load is always difficult, while measurement of 

 the suspended load is merely a matter of rou- 

 tine and patience. The estimate afforded by 

 the Yuba is a rough approximation and is 

 qualified by the fact that the stream had not 

 an established regimen but was engaged in 

 adjusting its slope to new conditions of load. 



Of other estimates of tractional load, the one 

 most often quoted is that of Humphreys and 

 Abbot. They found that the bar across one of 

 the mouths of the Mississippi increased its 

 width at the top at the rate of 338 feet a year 

 and that the material added was similar to that 

 of the tractional load. Assuming that the 

 deposit had an outward extent to the depth of 

 100 feet and a transverse extent equal to the 

 widths of all the mouths, they made a compu- 

 tation from which the tractional load was 

 estimated at 11 per cent of the suspended 

 load. 1 



An elaborate study of the loads of the Rhone 

 was made by Adolphe Gu^rard, 2 who measured 

 the suspended material by sampling the water 

 for two years and computed the total output 

 from the soundings of the sea bed, the survey 

 of which had been repeated after an interval 

 of 31 years. He found the suspended load less 

 than one-fourth of the whole. Various partial 

 estimates have been based on the march of 

 subaqueous dunes. Where the volume of a 

 dune and its rate of progression are known, 

 their product determines definitely a part of 

 the load, and if the data cover a year they 

 yield at once a minimum estimate for the 

 annual load. Some large dunes in various 

 streams have been reported to persist from 

 year to year, with progressive change of posi- 

 tion, but Arthur Hider, who kept two tracts of 

 the Mississippi channel under observation for 

 about a year, found that the dunes were re- 

 peatedly readjusted in respect to size as the 

 river stage changed, so that none could be 

 identified through long periods, while the 

 epochs of readjustment were characterized by 

 general deposition or general scour, which 

 could not be accounted for as a result of the 



' Hydraulics of the Mississippi, p. 149. 



2 Inst. Civil Eng. Proc., vol. 82, pp. 308-310, 1885. 



