IRRIGATION EXPERIMENTS IN IDAHO. 3 
SELECTION OF TRACTS. 
As it was imperative that the final results of the investigation 
should show a duty that could be secured by good average farmers 
without more expensive preparation than would be justified, only 
‘such tracts as were considered typical in regard to soil and practical 
in regard to preparation were included, and they were not especially 
prepared for the investigation in any way. 
DIVISION OF TRACTS AND VARIATION OF WATER. 
Each tract selected, wherever its topography and size permitted, 
was divided into three approximately equal parts, care being taken 
to select only such tracts as had uniform soil conditions and previous 
_ preparation and cropping throughout, in order that the results from 
the different plots might be strictly comparable. In most cases the 
tracts consisted of about 15 acres, divided into three plots of about 
5 acres each, in the same crop on the same soil, with all conditions 
uniform, except that a different volume of water was applied to each 
plot. 
Weirs were installed in the ditches leading to and from each tract 
in order to measure the water applied and wasted. ‘These tracts 
were selected from average irrigators’ farms in representative dis- 
tricts, and the owner of each was allowed to select one of the plots 
into which his tract had been divided and to irrigate and handle it 
during the season and apply the water at such times as had been his 
usual custom. The two remaining plots were irrigated by applying 
more water to one and less to the other than the owner applied to the 
plot he himself had selected. The yield produced usually indicated 
which plot had received the supply of water best suited to the soil 
and crop in question. Usually four or five experimental tracts were 
selected from as many farms in the same neighborhood, since one 
assistant could look after and irrigate that number of tracts pro- 
vided the same were located within a radius of 2 or 3 miles. The 
water wasted from each tract was measured, and all amounts tabu- 
lated in this report unless otherwise specified represent only the 
quantities retained upon the land. 
MEASUREMENT OF WATER. 
Cipolletti weirs were used in the measurements, the weir boxes 
usually being built of 1-inch lumber. The weir plates used were 
16-gage galvanized iron. The head on the weirs was measured each 
hour or oftener, as was required, with small steel rules graduated to 
0.01 foot. Measurements of the head were made from the tops of 
spikes driven vertically into substantial 2 by 4 or 4 by 4 stakes placed 
in the weir pools upstream from the weirs, at about twice their length. 
