[RRIGATION E5XPEEIMENTS IN IDAHO. 5 



are well represented in Plate II, figures 1 and 2, which show the 

 average soil of the first, second, third, and fourth foot of two of the 

 experimental tracts. 



MOISTURE DETERMINATIONS. 



The primary object of this investigation having been to determine 

 the quantity of irrigation water required by the soils and crops of 

 a farm or project, it was not considered advisable or practical on 

 account of the large territory involved to go to the expense of mak- 

 ing exhaustive soil-moisture determinations on all tracts throughout 

 the season. A careful determination of the moisture in the first, 

 second, third, and fourth foot of soil of each plot experimented upon 

 was made, however, before the first irrigation and about the time 

 plants started to grow hi the spring. This was done by securing a 

 large number of representative samples from each tract and drying 

 them in an oven at a temperature of 108° C. While the available 

 soil moisture* has not been taken into consideration in averaging the 

 quantities of water used and the yields obtained, these determina- 

 tions have permitted of a careful comparison of the results secured 

 from the different experiments. 



REIMBURSEMENT OF LOSSES. 



The owners of the various experimental tracts included in the 

 investigation have been reimbursed for all crop losses that have 

 been experienced because of variation of the water supply. The 

 yield of the plot selected and handled by the owner himself was used 

 in each case as a basis for determining the loss. Reimbursement 

 was found necessary in connection with less than one-third of the 

 tracts included in the investigation. 



METHOD OF INTERPRETING RESULTS. 



The correct and proper analysis of the results has been the most 

 difficult part of the entire investigation. Even when the greatest 

 care is used in agricultural experiments there are many factors 

 other than those under observation which may influence the yields. 

 Normally it has been assumed that the plot which produced the 

 best yield in each experiment received the best application of water. 

 In many cases, however, the largest yield has exceeded the yield 

 of one of the other plots by not more than 2 to 5 per cent, when 

 the quantity of water applied for the maximum yield exceeded that 

 applied to the next largest yield by as milch as 100 per cent. In 

 such cases it is apparent that if economy of time and water is to 

 be considered, it is better practice to apply the smaller quantity 

 of water. The investigation as a whole has emphasized the fact 

 that the results from single experiments can not be depended upon, 



