52 BULLETIN 339, U. S. DEPARTMENT OP AGRICULTURE. 



A study of the results shows, among other things: 



(1) Small laterals carrying 1 second-foot and less almost inva- I 

 riably lose a large part of the water carried, and the percentage of 

 loss decreases rapidly as the volume carried is increased, thus empha- I 

 sizing the desirability of rotation systems where the necessity of ! 

 carrying small amounts is eliminated. 



(2) Since certain types of soil have a fairly uniform loss per square I 

 foot of canal bed, canals should be designed, other things being 

 equal, with as small a wetted perimeter as possible in comparison to I 

 their cross sections. 



(3) Porous irrigated land above a canal may cause it to gain 

 instead of lose. 



(4) Canals in average southern Idaho soil, which is a medium clay ; 

 loam, should be designed to withstand a loss of 0.5 to 1.5 cubic feet I 

 per square foot of canal bed in 24 hours; 0.5 cubic foot per square foot I 

 per day is a safe basis for impervious clay loam soil, about 1 cubic I 

 foot per day for medium soil, and 1.5 to 2 cubic feet per square foot I 

 per day is a safe basis for somewhat pervious soils. 



(5) One per cent per mile is a safe basis for the loss in medium i 

 southern Idaho soil with capacities in excess of 200 second-feet. 



(6) Canals hi gravelly soil should be designed to withstand a loss 

 of 2.5 to 5 cubic feet per square foot of canal bed in 24 hours, depend- | 

 ing upon the porosity of the gravel, although it is probable that lining 

 would be profitable if the higher loss were experienced. The pro- | 

 cedure must be determined by local economic conditions. 



(7) A project having a comparatively long main canal, constructed I 

 through earth and unlined, may lose from 20 to as high as 50 per I 

 cent of the water diverted before it reaches the farms even in the I 

 impervious soils. (See fig. 13.) 



SURVEY OF WASTE LAND IN IDAHO. 



It has long been known that not every acre of the gross area con- j 

 tained in a project is irrigated. Engineers, for lack of more accurate [ 

 data, have commonly assumed that 20 per cent of a typical project | 

 is unirrigated because of high spots, corrals, county and private roads, 

 railroad rights of way, etc. Water has been increasing in value to • 

 such an extent that there is a great incentive to base projects upon 

 narrower margins each year, and it was evident that the above factor I 

 should be carefully determined in order to allow of a design that 

 would not only be economical but would safeguard the future of the 

 newer projects. For the purpose of determining the extent of unirri- 

 gated land which exists in a high-class, intensively cultivated project 

 in any one year from all causes, 16,067.8 acres of typical irrigated 

 land, located in two contiguous bodies in the best part of two typical 



