268 GROVE KARL GILBERT— DAVIS [MemoikS [v^xxi^ 



depth is so small that the velocity is no longer competent and the capacity is nil. It is equally evident that 

 when the width is gradually and greatly reduced the field of traction must become so narrow that the capacity 

 is very small, and eventually the current must be so retarded by side-wall friction that its bed velocity is no 

 longer competent and capacity is nil. For all widths between these limits capacity exists, and somewhere 

 between them it attains a maximum. 



A perusal of the above passage, in which a brandling concept is pursued along its diverging 

 paths, must make it clear why Gilbert was a good chess player. A later passage touches a 

 practical application of the same special phase of the total problem: "The ratio of depth to 

 width which gives to a stream its greatest capacity for traction is of importance to the engi- 

 neer whenever he has occasion to control the movement of debris," but unfortunately "no way 

 has been found to extend the quantitative results to rivers. It can hardly be questioned that 

 the optimum ratio [of depth to width] for rivers varies inversely with slope, discharge, and fine- 

 ness of debris, but its absolute amount cannot be inferred from the experimental results. River 

 slopes are relatively very small and river discharges are relatively very large, and the two 

 differences affect the ratio in opposite ways. To compute the joint result we should have 

 definite and precise information as to the laws of dependence, but our actual knowledge is 

 qualitative and vague." The belief is later expressed that "all the generalizations from the 

 laboratory results may be applied to natural streams, but only in a qualitative way; the dis- 

 parity of conditions is so great that the numerical results can not be thus applied." A closing 

 chapter on the application of the study to natural streams therefore hardly goes beyond a 

 refined descriptive and explanatory account of various elements of river behavior. This 

 somewhat disappointing conclusion is, however, in a measure offset by an account of various 

 actual examples which serve to verify the principles involved, one of which may here be 

 reproduced. 



In this connection it is of interest to record a single observation on river efficiency. Where Yuba river 

 passes from the Sierra Nevada to the broad Sacramento Valley, its habit is rather abruptly changed. In the 

 Narrows it is narrow and deep; a few miles downstream it has become wide and shallow. Its bed is of gravel, 

 with slopes regulated by the river itself when in flood, and the same material composes the load it carries. In 

 the Narrows the form ratio during high flood is 0.06 and the slope is 0.10 per cent. Two miles downstream the 

 form ratio is 0.00S and the slope is 0.34 per cent. Thus the energy necessary to transport the load where the 

 form ratio is 0.008 is more than three times that which suffices where the form ratio is 0.06; and it is evident 

 that the larger ratio is much more efficient than the smaller. The data do not serve to define the optimum 

 ratio, but merely to show that it is much greater than 0.008. 



PREPARATION OF FIRST REPORT 



The hydraulic laboratory experiments were conducted, as has already been mentioned, 

 at Berkeley, and were almost completed before Gilbert's illness in 1909; but the discussion 

 of the experimental data was carried on chiefly in Washington; and although it was under- 

 taken after a year of slow recovery from illness, it constituted the most arduous task that 

 Gilbert ever entered upon. The Wheeler reports were avowedly incomplete summaries based 

 on hurried observations. The Henry Mountains volume was based on a brief period of field 

 observation, mostly qualitative in nature, the discussion of which was largely thought out in the 

 field and speedily written up in Washington. The Bonneville monograph was equally Gilbert's 

 own work, and differed from the Henry Mountains report chiefly in the longer time allowed 

 for the collection of facts by himself and his aids and in the overlong delay in preparing the 

 manuscript caused by the distraction of other duties. Various studies on Niagara and the 

 Great Lakes were unlike the basin-range, Henry Mountains, and Bonneville reports, in call- 

 ing for a large acquaintance with the results gained by many local observers, although these 

 results were similar and supplementary to those gained from Gilbert's own intermittent field 

 work; the novelty in these studies lies chiefly in their ingenious generalizations and interpreta- 

 tions. Later work on the basin-range structures was interrupted before it was much more 

 than begun. In contrast to all these earlier investigations, the discussion of the debris experi- 

 ments demanded an extended examination of technical studies by experts in a nongeological 

 field, and a large amount of original analysis and computation; and in both these respects 

 it has no parallel in Gilbert's earlier work. Letters written to his elder son between 1911 and 

 1914 confirm this statement. 



