LAND DRAINAGE BY MEANS OF PUMPS. 33 



It is considered that it will be of value to introduce at this point 

 the results of very careful and extensive investigations of rainfall 

 and run-off on pumping drainage districts in southern Louisiana. 

 While various local conditions are different, it is believed that for 

 the six summer months mentioned above the relations between the 

 rainfall and the run-off are quite similar; that is, in both sections 

 the same percentage of the rainfall would appear as run-off. A 

 total of 45 heavy storm periods occurring in southern Louisiana were 

 therefore selected and the results averaged. The average precipita- 

 tion during the storms selected was 4.90 inches over an average period 

 of six days. The average run-off from these storms was 2.45 inches, 

 distributed over a period of 7.5 days. An increase of another day 

 in the period over which the run-off was taken would have increased 

 the total run-off very little, while a decrease of a day in the average 

 period of run-off would have decreased the total run-off very mate- 

 rially. Very few of the storms included in this list occurred after a 

 prolonged dry period, while perhaps 20 per cent of them followed 

 periods of rainfall that had saturated the soil. Therefore the result 

 undoubtedly is somewhat too large, and if provision be made for 

 such a percentage any error will be on the safe side. It will be noted 

 that the period covered by the run-off figures is one and one-half days 

 longer than the length of the average storm. The districts from 

 which these records were taken average about 2,500 acres in area. 

 As most of the districts in the upper part of the Mississippi Valley 

 are much larger than this, it is assumed that the time required for 

 the run-off to reach the pumping plant will be somewhat increased; 

 therefore in the discussion following the difference between the length 

 of the storm period and the length of the period during which 50 per 

 cent of the rainfall would run off is assumed to be two days instead 

 of one and one-half days, as found for the districts in southern Lou- 

 isiana. It will be noted in the above table for storm intensities that 

 a 4-inch rainfall during a three-day period occurs a little less fre- 

 quently than once in two years. Decreasing this frequency by one- 

 third, as explained above, a 4-inch rainfall in a three-day period 

 would occur in the growing season once in three years. Fifty per 

 cent of it, or 2 inches, would appear as run-off in five days. The aver- 

 age storage capacity of the canals and ditches on such districts is 

 about 0.25 inch; thus 1.75 inches of water would have to be pumped 

 in five days to prevent flooding, requiring a pumping capacity of at 

 least 0.35 inch per 24 hours. 



If it were considered that flooding of the- lower portions of the dis- 

 trict to a small depth for 24 hours once in three, years would not be 

 greatly injurious, the time for removing the water could be made 

 six days, requiring a capacity of 0.2!) inch. It is believed that a 



