EXPERIMENTS WITH MIXED GRADES. 



175 



The composite ratios have been arranged in 

 the order of the ratio of fineness, but the com- 

 parison shows no correspondence. They prove 

 equally inharmonious when compared with the 

 fineness of the .finer component, the fineness of 

 the coarser component, or the fineness of the 

 mixture. Their irregularities must be as- 

 cribed to observational errors and to causes not 

 at present to be discriminated from observa- 

 tional errors. It is of interest, however, to note 

 that the large measure of capacity change as- 

 sociated with the (CG) combination might be 

 inferred also from a comparison which involves 

 a different viewpoint and also some practically 

 independent data. If we think of grades (A), 

 (C), and (E) as modifiers of capacity for grade 

 (G), we may compare their efficiencies by means 

 of the following quantities, taken from Table 61 : 



The superiority of grade (C) as a modifier for 

 (G) is thus brought out without making use of 

 the capacities for the uncombined grades (A), 

 (C), (E), and (G); and the result from mix- 

 tures of 1 : 1 is supported by that from mixtures 

 of 1 : 2. 



CONTROL BY SLOPE AND DISCHARGE. 



The preceding comparisons are conditioned 

 by a discharge of 0.363 ft. 3 / sec., a slope of 1.4 

 per cent, and a channel width of 1 foot. With 

 a different set of conditions a different set of 

 quantitative relations would be found, and the 

 qualitative also would doubtless be modified. 

 The observations on mixtures included no other 

 width, and there was but a single set of ex- 

 periments using a different discharge, but the 

 range in slope was coordinate with that for the 

 separate grades. 



Figure 57 shows the capacity-slope curves 

 for the (AG) set of experiments, figure 58 for 

 the (BF) set, and figure 59 for the (CE) set. 

 In figure 57 the curves for mixtures form a 

 graded series between those for the component 

 grades, and there is almost perfect harmony of 

 form and attitude. As the points representing 

 capacities associated with a slope of 1.4 per 

 cent all lie in the same vertical line, and as 

 similar points for another slope lie in some 



other vertical line, it is evident by inspection 

 that inferences from data of any other available 

 slope would be practically identical with those 

 from the slope of 1 .4 per cent. It is also evident 



200 



o 



ID 



Q- 



1:1 



I 2 



Slope 



FIGURE 57. Curves of capacity in relation to slope for grade (A), grade 

 (G), and mixtures of those grades. The ratios of components in the 

 mixtures are indicated. 



that indexes of relative variation, \ or / for 

 the mixtures constitute, with those for the 

 components, an orderly system. The same 

 remarks apply also to the (CE) groups of 



200 





100 



ifs> 



1:1 



1:2 



I 2 



Slope 



FIGURE 58. Curves of capacity in relation to slope for grade (B), grade 

 (F), and mixtures of those grades. The ratios of the components in 

 the mixtures are indicated. 



curves, but they do not apply to the (BF) 

 group in figure 58. The attitudes of the curves 

 for mixtures are there out of harmony with the 

 attitudes of the (B) and (F) curves. The 

 curves for the mixtures seem to belong to a 



