3i6 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9. 
system of twelve Livingston's porous-cup atmometers which were fed from 
an elevated water tank. The bottom of the chamber was also covered by 
two large flat trays of water. Under these conditions it was possible to 
keep the relative humidity between 70 and 100 percent. The temperature 
varied from 15° to 31° C. during the daytime, usually averaging about 25° 
C. The atmospheric humidity in the dry chamber was kept low by the 
use of anhydrous calcium chloride. Under these conditions it was possible 
to keep the relative humidity between 30 and 60 percent. The tempera- 
ture variations in the dry chamber were about the same as in the humid 
chamber, but often the temperature was slightly higher in the former. The 
evaporating power of the air was determined by standardized Livingston's 
dark porous-cup atmometers. The average quantity of water lost per 
day for seven days was 13 cubic centimeters in the dry chamber as com- 
pared with 6 cubic centimeters in the humid chamber. 
The cultures growing in the sunlight were placed about 20 centimeters 
apart on a greenhouse bench. The relative humidity under these condi- 
tions varied from 25 to 60 percent. The temperature was usually between 
22° and 26° C, but the extremes were 14° and 30° C. The cultures grown 
in the shade were on the same greenhouse bench, but were covered with a 
tent of the same size and shape as the dry and humid chambers, made of 
two layers of cheesecloth, so as to reduce the sunlight. Within this tent 
the relative humidity was usually about 5 percent higher than in the open 
sunlight. The temperature was usually from i to 5 degrees lower than in 
the open sunlight. The average daily evaporation from standardized 
atmometers was 29 cubic centimeters in the sunlight and 17 cubic centi- 
meters in the shade. 
Barley (Hordeum vulgare L.) was used in this experiment. In order to 
avoid as far as possible error due to individual variation, seed from a pure 
line of barley was obtained from the department of plant breeding, Cornell 
University. The seeds, selected for uniformity in size and shape, were 
sterilized by formalin treatment, and germinated. When the roots were 
about four centimeters long the seedlings were planted in culture jars. The 
culture jars used throughout these experiments were quart fruit jars of the 
"Improved Mason" brand which were covered with black paper. Four 
seedlings were planted in each culture jar. After standing on a greenhouse 
bench for one week, those cultures which contained four healthy plants were 
divided into four similar lots of 28 cultures and placed under the following 
conditions : 
a. Dry chamber 
h. Humid chamber 
c. SunHght 
d. Shade tent 
These cultures were grown for five weeks, August 4 to September 8, 
1920. During this time the water lost by transpiration was replaced with 
