Jan. 3,1916 
Hourly Transpiration Rate on Clear Days 
625 
Since transpiration and evaporation are similarly affected by en¬ 
vironmental factors, the loss of water from a free-water surface affords a 
good summation of the intensity of such factors. The total evaporation 
from a tank 8 feet in diameter with the water surface at ground level at 
Akron during the months from April to September, inclusive, is 44 inches, 
based on the records for seven seasons, compared with 33 inches at Dick¬ 
inson in western North Dakota, 53 inches at Amarillo in the Panhandle 
of Texas, and 57 inches at Yuma, Ariz. In general, the evaporation 
increases as one proceeds from north to south through the Great Plains, 
and the same condition, though less marked, prevails from east to west. 
The transpiration conditions at Akron are probably as severe as may be 
found in cultivated areas east of the Rockies in this latitude (40° N.) or 
to the north of this parallel. 
Hourly evaporation measurements with the shallow, blackened tank 
were not made in 1912. The evaporation rate in 1914 was highest during 
the amaranthus period, as would be expected from a consideration of the 
intensity of the environmental factors. The mean maximum evapora¬ 
tion rate for the different periods during the hours near midday ranged 
from 700 to 900 gm. per hour from a tank of 6,540 sq. cm. in area. 1 
The highest temperatures and the greatest saturation deficits were 
encountered during the sorghum and amaranthus transpiration periods; 
yet these conditions produced no flattening of the peak of the transpira¬ 
tion curve of either plant, which is so marked in the case of wheat and 
rye. The lowest mean temperature and the smallest saturation deficit 
1 A loss of 1,000 gm. from the small tank corresponds to a loss of 0.0386 inch from the 8-foot tank referred 
to above, based on continuous records for the period, June 16 to September 19,1914. The large tank loses 
more slowly during the forenoon, but more rapidly during the night. This is due to the heat capacity of 
the large tank. The records based on 24-hour periods show good agreement between the two tanks. To 
those who are more familiar with evaporation as measured by Livingston's atmometer, the following 
comparison with the shallow blackened evaporation tank used in our experiments will be of interest. 
The hourly evaporation graph of the porous-cup atmometers does not agree in form with the evapo¬ 
ration graph from the tank. The atmometers show a marked lag during the middle of the day as 
compared with the evaporation taking place from the tank. This might be anticipated, since the tank 
receives only the vertical component of the radiation, while the candle type of atmometer receives a smaller 
percentage of the total radiation at midday in midsummer than earlier or later in the day, due to the verti¬ 
cal walls. The difference is, however, very pronounced even with the new spherical form of porous cup. 
It is consequently impossible to establish a definite ratio between the evaporation from the Livingston 
atmometers and the shallow tank used in our experiments. The average ratio may, however, be given. 
From 6 a. m. to 6 p. m., on August 13 and 14, 1915, an evaporation of 1,000 c. c. from the tank corresponded 
to an evaporation of 6.5 c. c. from the white candle-type atmometers (1913); of 7.5 c. c. from the same type 
(1915); of 8.3 c. c. from the white, spherical type (1915); and of 10.9 c. c. from the black candle type (1915). 
The loss from the atmometers corresponding to 1,000 gm. loss from the shallow tank for different parts of 
the day is as follows: 
Type of atmometer. 
White candle type ( 1913 ). 
White candle type (1915). 
White spherical type (1915). 
Blade candle type ( 191 s). 
6 to 10 
zo a. m. 
2 to 6 
a. m. 
to 2 p. m. 
p.m. 
7.2 
5 -i 
8.6 
8.2 
5-8 
ia.0 
9.1 
6.7 
10.3 
14.0 
8.4 
12.9 
During the night the atmometers each lost about 3 gm. of water, while the tank showed a slight gain due 
to deposition of dew. None of these atmometers had ever been used in other measurements, and dis¬ 
tilled water was used in all cases. The values given are based on the means of determinations with four 
atmometers of each type, after the observed evaporation from each atmometer had been multiplied by 
the standardization coefficient supplied with the apparatus. 
