Rainfcdl and Flood Discharge. 153 



The objection to this formula is, that when applied to 

 very large areas and long rivers, the high power of the 

 length reduces the quantity too rapidly ; I would therefore 

 alter a and I to miles, and adopt different co-efficients. 



The following formulae are given in Mr. Jackson's 

 " Hydraulic Manual : — " 



1st. Q = k, 27 (K/- 

 2nd. Q = k 100 (K)' 

 Srd. Q = L 1300 K (L)-^ 



Q = discharge cubic feet per second ; K area square miles ; 

 L length miles ; and kj^ h \ local co-efficients. 



The first is most simple, but k varies so much as to 

 make it inconvenient, and no attention is paid to the shape. 



The second is a modification of Col. Dickens' formula, 

 which was suited to Bengal, but ^• co 1 to 24. 



The third was deduced by Mr. Burge, of Madras. 



4th. Mr. Jackson proposes Q = k^y 100 (K^ 



I don't know the object of a numerical constant and a 

 variable constant. (?) 



5th. Mr. Hawkesley, an eminent authority, supplies a 



formula for the diameter of outlet pipes log. dia. inches = 



3 log. area acres -f- log. length, m which sewer falls 1 ft. + 6 "8. 



_ 



by using Mr. Hawkesley 's formula for discharge D = G \/ g'^ !>, 



H 



it is easily proved that the formula 5 is constructed on the 

 assumption that the discharge varies as K^ without regard 

 to form. 



6th. Mr. B. Zeigler, of Zurich, supplies the following 

 formula R = r x c ^V s 



a 



T being average rain, C= coef , varying from "75 for cities and 

 •31 for suburbs, s = fall in area or gTade per 1000, and a = 

 area drained, giving R resultant rain discharge, and this I 



find to vary as ^S'- K^, but by this if - under the sign becomes 



a 



greater than 1, the result is incorrect. Three inches of rain 

 with grade j^ and \ acre will give a rain discharge ot 

 5 '3 inches, evidently wrong for small areas. 



