IRRIGATION EXPERIMENTS IN IDAHO. 25 



cated that the maximum yields were made with approximately 1 

 foot of water per acre, and that applications in addition to this quan- 

 tity made a considerable decrease in the yield. There is no doubt 

 that had the fertilized plots of the Gooding experiment been more 

 fertile the largest yield would have been made with the 1 foot per 

 acre. 



This Gooding experiment, together with the other experiments 

 which have been made, seems to prove conclusively that less water 

 is required for the production of a given yield of crop on fertile soils 

 than is required for the same yield of crop on infertile soils and empha- 

 sizes the necessity of maintaining a high state of fertility in cases 

 where a high duty of water must be secured. It does not mean, 

 however, that less water is required by the fertile soils, for, as shown 

 by figure 3, fully as much is required for maximum crop production, 

 a greatly increased yield being produced on the fertile soils. 



RESULTS SECURED WITH INDIVIDUAL CROPS. 



In order to show the results secured from single crops on the same 

 type of soil, curves are shown in figures 5, 7, and 8. Each of these 

 curves is based upon five years' results at the Gooding experiment sta- 

 tion. The average amount of precipitation during the growing sea- 

 son for four of these five years was 2.92 inches. Considering that 

 each of these curves is based upon a number of experiments, that this 

 station lies at -the average altitude for southern Idaho, and that the 

 soil, which is a medium clay loam, rather impervious, and averaging 

 6 to 8 feet in depth, is representative, it is believed that the results 

 plotted are typical of what may be expected throughout the State, 

 arid that they show within a reasonable limit the yield of crop which 

 may be produced with a given quantity of water. These curves do 

 not necessarily indicate a proper duty of water under good practice 

 in Idaho, since, as is later stated, local economic features enter largely 

 into all such problems. 



The curves in figure 5 show the average results secured from 96 

 plots of staple varieties of spring wheat grown during the five seasons. 

 One curve shows yields per acre, and the other yields per acre-foot of 

 water applied. Considering the large number of plots upon which 

 these curves are based, they should be very dependable. The curve 

 showing yield per acre agrees with the data secured elsewhere through- 

 out the State, in that it shows that the yield of grain will normally 

 increase as the water applied is increased until an amount varying be- 

 tween 1.4 and 1.8 acre-feet per acre has been applied, after which the 

 application of more water will decrease the yield of gram and in many 

 cases the yield of straw as well (fig. 6). The curve showing yields per 

 acre-foot of water (fig. 5) shows that the highest efficiency from water 

 applied to spring grams is secured where the smaller quantities are 



