CORRESP ONDENCE. 



2 37 



ing remarks and discussions to render them 

 intelligible. 



Again : " It was determined by actual 

 experiment that the greatest velocity of 

 current is at the surface, and the least at 

 the bottom ; and that the increase of veloci- 

 ty ' is in the simple ratio of the distance 

 from the bottom.' " 



Taking this ratio as true, draw a verti- 

 cal right line, to any convenient scale of 

 equal parts, to represent the depth of the 

 river at any given place. Then the veloci- 

 ties at different depths would be repre- 

 sented graphically by an inclined right 

 line, the lower extremity coinciding with 

 that of the vertical and the upper at 

 a distance from the upper extremity of 

 the vertical equal to the surface-velocity. 

 Humphreys and Abbot did not find the 

 relation so simple ; they demonstrate that 

 " the velocities at different depths below 

 the surface, in a vertical plane, vary as the 

 abscissas of a parabola, whose axis is par- 

 allel to the water-surface," and that " the 

 position of the axis in calm weather is 

 about -f\,- of the depth below the surface, 

 whatever be the mean velocity of the river." 

 In order to deduce the relation between the 

 velocity and depth, the river-depth was di- 

 vided into ten equal parts, and the velocity 

 of each was determined by 222 observa- 

 tions; the mean velocity, which was 0.297 

 D (depth) below the surface, was 3.26 feet ; 

 the mean depth 82 feet. 



Very many other points of interest are 

 brought out in the Report ; such as proving 

 that the curve of surface-velocities is also a 

 parabola, deducing the values of the para- 

 meters of these curves, showing how the 

 curves are affected by winds, irregularities 

 of the bottom, etc., etc. The conclusions 

 were deduced from an immense number of 

 observations taken at different times of the 

 year, different stages of high and low river, 

 and at numerous points, by a number of 

 careful and trained observers. In the math- 

 ematical discussion of results, the most re- 

 fined methods were employed. 



The current-metre used by M. Revy, 

 although it is stated that he improved it 

 greatly, is open to grave objections, double 

 floats being preferable in point of accuracy 

 in determining the laws regulating the flow 

 of water in river-channels. I have never 



used any of the patent current -metres, 

 though I have seen and examined them ; 

 but I have used the double-floats often 

 enough to be convinced of their utility. I 

 have, however, used an instrument very 

 similar in principle to the current-metre, 

 one of the numerous patent deep-sea leads. 

 This instrument registered depths in fath- 

 oms and quarters on an index-wheel turned 

 by the rotation of a propeller, the latter 

 being so arranged as to turn the index- 

 wheel on going down, but, on being drawn 

 up through the water, was thrown out of 

 gear, so as not to unwind the register. 



This lead I tested by over one hundred 

 careful soundings, taken at depths rang- 

 ing from six to fifty fathoms, simultaneous 

 soundings being taken with the patent lead 

 and an ordinary lead-line. The soundings 

 given by the patent lead were so utterly un- 

 reliable that I discarded it, and used the 

 ordinary lead-line for the entifte work of 

 some pretty extensive hydrographic sur- 

 veys. To show that this is not merely my 

 own individual opinion about this style of 

 instrument, I will quote, from page 621, Re- 

 port of the Chief of Engineers for 1870, the 

 criticism of General Abbot upon the same 

 class of instruments : 



"In my opinion, founded on a some- 

 what close study of the subject, instru- 

 ments of this class are pretty toys, which 

 have contributed more to retard the prog- 

 ress of discovery in the science of river 

 hydraulics than any other one cause. This 

 is due principally to the fact that they regis- 

 ter their results in a kind of cipher, to which 

 we can by no means be sure that we pos- 

 sess the key. To translate a given num- 

 ber of revolutions of a submerged wheel 

 into velocity per second, and by this means 

 to detect laws whose existence is denoted 

 only by differences of a few tenths of feet 

 in this velocity, is so delicate an operation 

 that errors in the coefficient have usually 

 masked the laws." 



The subject of river hydraulics is quite 

 an intricate one, and not likely to be inves- 

 tigated by people generally. Therefore, an 

 article on this subject in a magazine is apt 

 to be read and its statements accepted with 

 less of questioning than one upon almost 

 any other scientific topic. 



I have been a constant reader of The 



