ENGINEERING CONSTRUCTION AND RAINFALL. XLIII. 



fairly inclined, whilst the surface of the water as determined by 

 the flood-marks was nearly level. It is only when a stream is 

 "in train," as it is termed, that the slope can be used to deter- 

 mine velocity ; the surface of the water and the bed of the 

 stream are then, to all intents and purposes, parallel. The 

 attempt to obtain velocity from the small difference of level in 

 backed-up water by ordinary slope formulae is useless, and 

 co-efficients derived therefrom are incorrect. In the determina- 

 tion of flood-levels, as a rule, there is little or no difficulty if the 

 subject be properly approached, but very erroneous results can 

 be deduced unless great care is taken. 



The general character and inclination of catchment. — As 

 regards these particulars, two areas, of the same extent, may 

 vary widely as to the amount of rainfall discharged in a given 

 time. The shape of a catchment is an important item ; thus, a 

 catchment running back from the outlet, say, 4 miles, and 

 averaging a quarter-mile wide, will discharge rainfall at a much 

 more regular rate than a catchment of the same area and nature 

 of soil, but greatly differing in shape, which, for instance, might 

 consist of a central creek into which several side creeks dis- 

 charged at different points, although the amount of run-off in 

 each case from a similar rainfall might be identical when the 

 flow ceased. The long, narrow catchment might even never 

 experience rain of sufficiently long duration to cause a maximum 

 run-off at the outlet, whilst the differently-shaped catchment of 

 equal area, supplied with what might be termed " feeder " creeks, 

 would very likely discharge a maximum amount under a rainfall 

 of similarly short duration. Thus the shape of a catchment and 

 its characteristics are often of equal importance to its extent. 



The shape of a catchment can be taken into account in the 

 application of some formulae, such as Biirkli-Ziegler, McMath, 

 Adams, and Hawksley, indirectly, it is true, as no direct means 

 has been provided, whilst the formulae of Burge, Jackson, Steane, 

 and Craig deal essentially with the shape. These latter, however, 

 do not deal with the slope of a catchment, which also is of great 

 importance ; if this be uniform the discharge will begin from 



