i74 



NA TURE 



\June 22, 1882 



discharge-measurements of the river nearly on the lines 

 of the Mississippi work, viz. by direct velocity-measure- 

 ments at numerous points of certain selected sites. 

 Besides the practical value of these as necessary data for 

 the embankment projects, the details may be of great use 

 in the study of the flow of water. And indeed this forms 

 the most interesting portion of the work in a scientific 

 sense, being a mass of original experiment on the flow in 

 a mighty river. Much credit is due to the author for the 

 zeal with which he had these experiments carried on for 

 several years, in the face of great difficulties and dis- 

 couragements. The experiments are discussed only so 

 far as necessary to explain the application of the results 

 (chiefly discharge-measurements) to the embankment 

 projects. A further special report upon the experiments 

 themselves is promised, which should be of great value. 



The velocity-measurements appear to have been entirely 

 made with the "double-float," whereof the surface-float 

 was a wood disc 6" X 6" X 1 " joined by a cord Vie" thick, of 

 various lengths, to a cylindric wood sub-float 6"x6"X 12" 

 loaded with clay, and sunk to various depths from 1 to 24 

 metres. At moderate depths this instrument would be 

 pretty efficient ; unfortunately the efficiency of all double- 

 floats decreases with the depth of immersion, and at the 

 greatest depth of 24 metres this one must have been very 

 inefficient; for — supposing even that the sub-float retained 

 its most favourable (the upright) position— the relative 

 areas of connector and sub-float exposed to direct current- 

 action would be as 73 to 100, and to lateral current-action 

 as 52 to 100 ; so that the observed velocity of the instru- 

 ment was certainly not that of the current at 24 metres 

 depth (as it is taken to be). Notwithstanding the inherent 

 objections to the double-float, there seems to be as yet no 

 better instrument available for mighty rivers. 



' Two sorts of velocity-measurements were undertaken, 

 viz. (1) at one metre depth at many points (from 30 to 60) 

 across the channel ; and (2) at every metre of depth upon 

 selected verticals in the channel. The latter were con- 

 sidered the more important. This sort of work must 

 necessarily have been very tedious in flood-seasons on a 

 might)- river ; at such times only ten complete series 

 could be done daily ; altogether about 10,000 such series 

 were done. This is a collection of experimental data 

 quite unique in river hydraulics, of which the author may 

 justly be;proud. From these data, together with the 

 cross-section figure, the discharges were computed ; the 

 mode of computation seems to have been as good as the 

 data admit of. 



The mode of presenting the results is open to some 

 objection— <?.£•. many of the velocities are carried to four 

 decimals of feet per second, a degree of accuracy quite 

 unattainable; again the discharges are given in several 

 different forms, viz. in cubic feet per second, in cubic 

 metres, and also in tons and in "metre- tons" per day, 

 per month, and per year, and in some tables the unit is 

 not stated. The author points out that one cubic metre 

 or water weighs about 55H-s6ths of a ton, so that the 

 two measures (cubic metres and tons) may be used indif- 

 ferently with an error of less than 2 per cent., and he 

 emphasises this coincidence by the use of a new term, 

 "metre-ton"; but by all modern usage this term means 

 either the "moment of one ton of pressure at one 

 foot leverage," or the " work done in raising one ton 



weight through one foot height," so that this new usage 

 is inconvenient. 



As to the theory of running water some novel views are 

 brought forward. It is stated that, speaking broadly, two 

 theories of flow in open channels have existed ; viz. that 

 previous to Du Buat's time the motion had been sup- 

 posed due solely to pressure, and since his time has been 

 supposed due solely to surface-slope, so that the earlier 

 formula? involve pressure, and the later surface slope. 

 The author himself is the advocate of a " new theory," 

 viz. that the motion is due to both pressure and surface- 

 slope ; his arguments appear to be chiefly two, viz. (1) 

 that formulae involving only one of these elements all fail 

 under varying conditions ; (2) that in many cases the 

 ratio of the deep-seated velocities to those near the surface 

 (which is usually < 1) rises with increase of depth of the 

 stream, and may sometimes even exceed unity in very 

 deep streams. These views can hardly be admitted. 

 Firstly, as to causality, surface slope is really only a 

 property of running water, not a cause of motion. All 

 change of motion is due to and is evidence of the action 

 of some unbalanced force (or pressure). In the case of 

 running water the unbalanced active force (effective in 

 forward motion) is the part of the earth's attraction not 

 directly balanced by the normal resistances (ultimately of 

 the margin), i.e. the resolved part thereof parallel to the 

 motion, the measure of which is gp sin /', and actually 

 enters into all modern hydraulic formula; in various equi- 

 valent forms, e.g. as (p - p'), {dp -i-dx).S x, gp (h - h'), 

 gp sin i, gpS, &c. ; it cannot therefore be said that 

 pressure is excluded from modern formulae (although, 

 after substituting numerical values for gp, the evidence 

 of it is apparently lost). In the argument it seems also 

 to be implied that the increase of pressure due to increase 

 of depth should cause increase of velocity, but the fact is 

 that increase of pressure does not of itself affect motion 

 at all, unless the increase be (at least in part) unbalanced. 

 An interesting series of discharge-measurements was 

 made at three sites in concert, viz. at Saiktha, near the 

 head of the Delta, and at Zaloon and Thapangyo, which 

 are situate on the two largest Delta streams. There 

 appear to be only some minor local affluents into and 

 effluents out of the space between the upper and lower 

 sites. It would seem therefore that the discharge- mea- 

 surements at the upper and at the two lower sites together 

 should be nearly equal ; this affords a valuable test of the 

 consistence of the results. The field work seems to have 

 been done at each of the three sites on seventy-three days 

 in 1872-73, thus giving seventy-three pretty complete 

 results ; other eighty-three days' results are also given, 

 but these are partly interpolated, and therefore of less 

 value. The discrepancies are sometimes very large, 

 ranging from a gain of 27 per cent, to a loss of IS per 

 cent, in the daily results ; most of them (97 out of 154) 

 are on the side of gain. After making allowance for the 

 utmost possible supply from the minor affluents between 

 the sites (by adding in the whole rainfall all over their 

 drainage-basins) the residual discrepancy is attributed to 

 the "storage power" of the river area (about 305 square 

 miles) between the sites. While there is no doubt some 

 (temporary) " storage power," it seems more likely that 

 most of the discrepancy is due to real error in the results 

 themselves, the fact being that only very rough approxi- 



