278 
Journal of Agricultural Research 
Vol. IX, No. 9 
Each type was represented by four atmometers 1 which were weighed 
independently at two-hour intervals throughout the day. The atmome¬ 
ters were mounted on 500-c. c. bottles closed with rubber stoppers con¬ 
taining a small air vent, and were freely exposed in racks about 1 meter 
above the ground. Distilled water was employed in all the measure¬ 
ments. All the atmometers used were new, and each cup was inverted 
and flushed with water for several hours before it was set up. The 
cups were set up for 36 hours before the detailed measurements were 
begun. 
Evaporation tanks. —Two evaporation tanks of widely different 
types were also employed in the measurements, one containing a layer 
of water approximately 1 cm. in depth, while the depth of the water in 
the other was about 50 cm. 
The shallow tank, which was 91 cm. in diameter and 2.5 cm. high, 
was supported on a wooden disk 4 cm. in thickness and mounted about 
1 meter above the ground on an automatic scale. The tank was black¬ 
ened inside. The depth of water was automatically maintained at about 
1 cm. by means of a Mariotte apparatus mounted on the scale platform 3 
(PI. 5). The automatic scale was sufficiently sensitive to record the loss 
of a layer of water 10 m (o.oi mm.) in thickness. 
The deep evaporation tank, which was 192 cm. in diameter and 60 
cm. in depth, was sunk in the ground to within 10 cm. of the top. This 
tank was equipped with a continuously recording gauge of the float 
type reading to 0.1 mm. 
Filter-paper evaporimeter. —An evaporimeter of special design, in 
which the evaporating surface was a saturated filter paper (12.5 cm. in 
diameter) was also employed in the measurements (Pi. 6). The filter 
paper was supported on a light flat brass disk, with a rim 1 mm. in 
height. The paper was kept saturated by means of a Mariotte apparatus 
connected with a tubulure on the lower side of the disk. This arrange¬ 
ment avoids the drying out of the filter paper, which is sometimes 
encountered in the Piche evaporimeter, due to the failure of the instru¬ 
ment to feed properly. The evaporation was determined by weighing 
the whole apparatus at two-hour intervals. 
1 The identification numbers and the coefficients of the atmometers used in the comparisons were as fol¬ 
lows: 
No. 
Coeffi¬ 
cient. 
No. 
Coeffi¬ 
cient. 
White cylinders. 
Do. 
5-36 
5-37 
5-57 
5-142 
4-17 
4-«5 
4-41 
4-59 
0. 77 
•77 
•77 
•77 
• 74 
•74 
. 76 
•74 
White spheres. 
Do. 
16-16 
16-77 
16-105 
16-146 
! D-137 
D-175 
D-205 
D-206 
0.94 
•94 
•94 
•94 
Do. 
Do.. 
Do. 
Do. 
Brown cylinders. 
Do. 
White Bellani plates. 
Do. 
Do. 
Do. 
Do. 
Do.: 
3 This equipment is the same as that used by the writers in connection with earlier transpiration meas¬ 
urements (Briggs and Shantz, 1915, 1916). 
