NA TURE 



241 



THURSDAY, JULY 12, 1883 



HYDRAULIC MANUAL 

 Hydraulic Manual. By L. D'A. Jackson. 4th Edition. 

 Pp. xiv. ■+■ 307 Text -f 184 Tables. (London : Crosby 

 Lockwood and Co., 1883.) 



THIS well-known text-book having reached its fourth 

 edition, it is unnecessary to review it as a new work. 

 The changes from the third edition are very great ; the 

 chief is the omission of the whole of the " Hydraulic and 

 Meteorological Statistics" (about 224 pp. of tables); 

 these relate chiefly to India, so that their omission is an 

 advantage to the " Manual" as a general text-book, as it 

 has enabled the text to be increased from 221 to 307 pp., 

 and the general working tables from 104 to 184 pp., 

 without increasing the bulk of the volume ; the chief 

 increase of the text is the introduction of an account of 

 the great Roorkee hydraulic experiments. 



Much stress is rightly laid on the small value of the old 

 hydraulic knowledge ; thus (p. 3) it is said, " Taken gene- 

 rally the mass of hydraulic science . . . prior to about 

 1856 may be considered superannuated. . . ." The most 

 useful feature of this work is indeed its freedom from what 

 is " superannuated," and its thorough adoption of recent 

 experiment ; the text is in fact in great part a short 

 account of the great modern experiments. In detailing 

 field operations the author has indeed preferred to give 

 a "brief account of the modes adopted by various 

 hydraulicians " as being " a far better guide to the engi- 

 neer about to undertake the execution of gauging opera- 

 tions than any arbitrary advice or set of rules could 

 possibly be." These concise accounts are on the whole 

 well condensed ; but the recapitulation — in some cases 

 verbatim — of the several experimenters' own conclusions 

 has the disadvantage that in several cases contradictory 

 conclusions appear on different pages ; this is inseparable 

 from the progressive state of our knowledge of the mo- 

 tion of water when stated in this way ; a little more 

 discussion of the contradictory views would have been 

 useful. 



Kutter's general formula for mean velocity was early 

 adopted by the author ; its use as the formula to be pre- 

 ferred to all others for the case of canals (whenever 

 velocity-observation has to be dispensed with) is now 

 insisted on, much evidence in favour of it having been 

 brought out by the recent large Roorkee experiments, 

 with the very fair reservation however that Kutter's 

 rugosity-coefficient («) should at present be determined 

 by actual experiment for each new channel, the data for 

 its a priori determination (from the mere nature of the 

 channel) not being as yet good enough. On the other hand 

 it is rightly said that " to determine with accuracy the 

 discharge of any ordinary or large river independent of 

 velocity-observation is at present impossible." 



A few minor details are worth notice. The units of 

 measure, &c, adopted are an extremely simple and useful 

 decimal system ; they include the foot, the " foot-weight " 

 of 1000 fluid ounces, i.e. the weight of a cubic foot of 

 water at its greatest density, and a " league " of two 

 London miles of 5000 feet each ; this league is particularly 

 suited to measurement of hydraulic slopes, a fall of I, 2, 

 Vol. xxviii. — No. 715 



&c, feet per league being at once seen to give a slope of 

 1, 2, &c, in 10,000. Two new very expressive names are 

 introduced for two velocities, which recur very frequently 

 in discussions on flow of water, viz. "verticalic velocity" 

 and " transversalic velocity" for velocities past any 

 vertical line or any (horizontal) transverse line in a channel 

 section ; these short terms will be a great relief from the 

 wearisome periphrases hitherto in use, and merit general 

 adoption. 



A few suggestions towards improvement of the work 

 may now be made. (1) In a purely professional work 

 such expressions as "international recrimination," and 

 " bureaucratic and heated with vanity" (p. 37) are surely 

 out of place. (2) About one page of text and three of 

 tables are devoted to the variation of gravity in different 

 latitudes and at different heights ; now the variation is so 

 small that for the rough calculations of practical hydrau- 

 lics this is an unnecessary refinement. (3) Among the 

 "general notation" (p. 11) occurs the rather awkward 

 phrase " g = velocity acquired by gravity in one second." 

 (4) In finding the (trapezoidal) " section of maximum 

 discharge " from the expression for discharge Q = A V 

 where V = 100 c s/ R 5 and R = A -j- P, the argument 

 used is that " under the condition of maximum discharge, 

 A will be a maximum, so also will R ; and when these 

 are temporarily constant, P will be a minimum ; " this 

 argument might be considerably improved, somewhat as 

 follows: — "Since Q= 100 c A V R . </S, therefore Q is 

 greatest (provided 5 be kept constant) when c, A, and R 

 are all maxima together ; now c is known (from experi- 

 ment) to increase with R, and R = A -j- P ; hence Q will 

 be greatest when A is a max. and P a min. at same time 

 (provided of course that -S" is constant)": this argument 

 is more general than that in the text ; the effect of the 

 5- variation is unknown. (5) About certain formulas for 

 "mean verticalic velocity," quoted from the Roorkee 

 work, it is said (p. 209) — "The defect in these methods 

 is evident; it consists in making the parabolic curvature 

 dependent on one point or on two points, whereas three 

 points are the least necessary." This last statement is a 

 mistake ; three points are necessary (for finding a mean 

 ordinate) only if they be taken at random, but two points 

 are sufficient when suitably chosen, as in the " two-velocity 

 formulae " quoted on pp. 87, 208 from the Roorkee work ; 

 these formulas are in fact accurate for the parabolic form, 

 and the proof of this (from the Roorkee work) is actually 

 given at p. 87. The " one-velocity formula? " are of course 

 only approximate. It may be mentioned here that the 

 writer has lately * discovered another (and far better) 

 " two-velocity formula," also accurate for the parabola, 

 viz. U= J (i/ 2 n h + v-Twu)t showing that the " mean ver- 

 ticalic velocity " is the arithmetic mean of the velocities at 

 •211 and 789 (or say ^ and j^) of the depth: this new 

 formula has several great practical advantages over any 

 other yet published ; among others, the two velocities can 

 be measured at one operation with a single instrument (a 

 compound " double-float " with two equal subfloats at the 

 depths named), which is itself moreover susceptible of 

 being made a more accurate instrument than any other 

 of its class (double-float). 



Allan Cunningham 



1 SeePree. Inst. CM! Engineers, vol. lxxi. pp. 18. 19, where the formula 

 and instrument are both discussed. 



M 



