murphy: IRRKiATION IN WESTERN KANSAS. 12[ 



the greatest maxitnum flow of the river as given in table III. The 

 question naturally arises, where does the water come from to supply 

 these canals? It is true that at times there is not water enough in the 

 river and its tributaries to supply this capacity, but there has been 

 plenty this year. It has not been necessary this year to restrict the 

 amount which any canal could take, and near the end of July when I 

 passed through, there was enough in the river east of the Colo- 

 rado line to supply all the canals in Kansas along the river. May it 

 not be that some of the water is used two or more times for irrigating? 

 That such is the case on the Cache la Poudre, a tributary of the 

 Platte in northern Colorado, the following measurements will show. 

 Measurements of this stream in 1889 showed that the discharge at 

 one point was 127.6 cu. ft. per sec. while at a point lower down the 

 stream it was found to be 214.5 cu. ft. per sec. after supplying 15 

 canals and without receiving additional natural drainage. This is 

 an increase in the flow of this stream of 60 per cent, after supplying 

 the 15 canals. In the upper valley of the Arkansas much of the 

 water which some canals take from the river is returned directly to it. 

 A man who has a large farm in the upper valley and who owns two 

 canals taking water from the Arkansas, told the writer that he kept 

 his grass lantl under slowly moving water much of the time. The 

 water from his ditches flows gently over his land and then directly 

 back into the river. 



It is evident from the above tables and discussion that the amount 

 of water which can be taken from the Arkansas in western Kansas 

 is small and is available only during June and a part of May and 

 July, therefore water to irrigate this region must be gotten mainly 

 from beneath the surface. The amount of underground or under- 

 flow water and its distance from the surface is therefore very impor- 

 tant. 



The subject of the underflow of the Plains has been investigated 

 by the U. S. Dept. of Agriculture. This investigation for the Ar- 

 kansas valley consisted in running three stadia lines north and south 

 across the valley, thus getting the elevations of wells and their dis- 

 tances apart on or in the vicinity of the lines. The distance from 

 the surface to water in each well, and the strata passed through were 

 noted. Thus the position and inclination of the upper surface of 

 this underground water was ascertained. These lines are known as 

 the Garden City, Dodge City, and Great Bend lines. The latter 

 being east of the eastern limit of the Arid Region will not be con- 

 sidered. 



The Garden (Mty line extends from a point on Ladder Creek, a 

 branch of the Republican, about 42 miles north of Garden City, to 



