624 



PROCEEDINGS OF SECTION H. 

 Table I, Continued. 



0=Cubic ft. per sec. 

 R= Rainfall in inches per hr. 

 a=Area of waterway in sq. ft. 

 M=Area of catchment in sq. miles. 

 A=Area of catchment in acres. 

 C= Variable co-efficient. 



Those formulae which do not take into consideration the varia- 

 tion of rainfall for different catchments will not be considered, as 

 they are evidently of local value only. Of the above list only five 

 remain, four of which directly take into consideration the slope of 

 surface, and of these, two adopt Si, one S^ and one S"i~2. When 

 S varies between -0001 and "04 the Si makes the discharge vary from 

 1 to nearly 4-5 times ; while S-pi- makes the variation range from 

 1 to 1-65 time. Manifestly both cannot be reliable for general 

 application. Again one assigns to the rainfall a value of R|, one 

 Rf, and the three others simply R. The values of R determined 

 for the most generally occurring maximum rainfalls show the 

 absurdity of these factors : — 



The above shows that by assigning a suitable co-elticient to R 

 the result ( an be made to differ only about 5 per cent, up to 5 inches 

 of rainfall nnd 10 per cent, for 6 inches of rainfall, and on the safe 

 side. There can therefore be no practical advantage in adopting 

 the more cumbersome factor. The Burkli Ziegler formula is pro- 

 bably the best, but the adoption of a definite value for S for the 

 slope makes its application difficult or indeed impossible in the 

 many cases in which the average slope cannot even be approxi- 

 mated. 



It is thought improbable that variations of average slope from 

 •0001 to -04 should give four times the discharge, in many cases 

 exceeding the possible volume due to actual rainfall, and is contrary 



