90 Floods on the River Banvon. 



if the contracted part of the channel be short enough to 

 allow the surface slope to vary, the depth would increase to 

 such an extent only as would discharge the same quantity 

 of water with increased surface slope and depth combined. 



17. Applying these principles to the case of a stream 

 confined to one portion of a channel, the necessity for a rise 

 in the surface level becomes apparent. The great difference 

 of level below the railway embankment, already mentioned 

 (see par. 12), is an example of this. The calculations made 

 at each contraction give the same results, which are shown 

 in the diagram. As a general rule the cross-section of any 

 stream shows a horizontal line for the surface of the water 

 on] y when the depth and velocity right across are uniform, 

 or nearly so. When one part of a channel is much deeper 

 than the rest, and the velocity of the water in that part is 

 much greater than in other parts, the level of the water 

 is raised there, so that the surface line is uneven, and not 

 level, as often supposed. When the stream is confined to 

 only a portion of the channel, the rest being still or back- 

 water, the irregularity of the surface line is greatest. This 

 is exemplified by the flood levels obtained immediately 

 below the railway embankment (see par. 12). This effect 

 may be often observed when pouring water from a jug into 

 a basin ; it may be seen also in some of the large drains in 

 Melbourne, at the bottom of Elizabeth-street for instance. 

 Where the velocity of any stream is greatest, the surface is 

 generally higher than elsewhere on the same, cross-section. 



18. Calculations made on the foregoing principles in con- 

 nection with the flood of 1880 having given results as 

 nearly in accordance with the facts as could be expected, 

 considering the conditions of the case, other calculations 

 were made to show what the levels of the flood would 

 have been, had the two embankments above alluded to not 

 been in existence ; that is, if it had occurred in 1852, when 

 the river was comparatively free from obstruction. The 

 results obtained by the author are shown on the diagram by 

 the shaded red line. The starting point is RL. 15 00 at 

 C.S. No. 1 (see par. 13), and the following are the differences 

 of level worked out ■ — Immediately above the railway em- 

 bankment, the water would have been about 4 J* feet lower; 

 opposite the mills (C.S. No. 5), nearly of *f* feet lower; and just 



* E.L. 21-20— R.L. 16-90 = 4-30 feet. 



f R.L. 21-50— R.L. 17.80 = 3-70 feet. Inside tlie mills the water would 

 have been upwards of 4 feet lower. 



