OVERFLOW OF EED RIVER OF TlIK NORTH. 17 
were tabulated or plotted as curves showing various relations. Run- 
off records were studied and the problem of the relation between pre- 
cipitation and run-off was approached from various angles in an 
, ilort to reach reliable conclusions as to maximum rates of flow 
which may occur in the different parts of the Red River, the coinci 
dence of heavy How in the various tributaries, and the probable time 
intervals between damaging flood stages. 
DRAINAGE. 
The drainage situation over the watershed was studied in consid- 
erable detail. The flat land along both sides of the main watercourse 
was considered separately from the higher, backlying areas, as its 
drainage is more intimately related to the conditions of flow in the 
main channel. Areas affected by present drainage of these flats were 
determined and estimates made as to probable future drainage and 
its effects. 
Considerable areas of land along the Bois de Sioux River lie only 
slightly higher than the normal water surface in that stream. Under 
high -water conditions these lands are submerged and form a part of 
the bed of Lake Traverse. At all times these areas are marshy and 
unlit for agricultural purposes. Estimates were made as t<> the 
amount of such land reclaimable and the extent of improvement pos- 
sible. Details for the complete drainage of the areas were not worked 
out. but plans for the main channel improvement of the Bois de 
Sioux include the layout of main outlet ditches necessary for the 
sal of overflow water and of the flow from future local drains: 
The general conditions of drainage systems in all parts of the 
watershed were investigated. The question of tile drainage, its prob- 
able development and generaleffect on the outlet channels, was also 
considered. 
FLOODS. 
In studying the flood situation the main object was to. determine as 
accurately as possible the extent of damaging flow with correspond- 
ing river stages, the probable frequency and magnitude of future 
floods and the maximum flow that may be expected, Idle studio of 
precipitation, run-off', and effects of drainage were made mainly for 
this purpose. Information concerning floods from whatever source 
onsidered and given such weight a- it- importance and relia- 
bility seemed to warrant. 
MEASURES FOR RELIEF. 
Relief measures were considered under two bead-, drainage and 
flood relief. The question of draining the wet lands along the Bois 
9 oux River was taken up along with that of the detention 
70352°— 22 2 
