TRANSPIRATION STUDIES. 11 



used to cool the metal surface, but in 1912 carbon dioxide* was led into 

 the tube from a high-pressure tank. By this means the dew-point was 

 obtained much more quickly than with ether and the danger of physiological 

 disturbances caused by waste ether was avoided. Frequent tests with 

 this dew-point apparatus showed that the temperature of appearance and 

 disappearance of moisture differed less than 0.1 C.when the temperatures 

 were above 4 and not more than 0.2 degree when the temperatures were 

 below -4. 



The calcium chloride was exposed in shallow weighing-bottles, the covers 

 of which were air-tight when the salt was not in use. Merck's C. P. calcium 

 chloride was used, but to guard against the absorption of an appreciable 

 amount of carbon dioxide from the air by traces of potassium hydroxide 

 which might be present, carbon dioxide gas was made from calcium carbo- 

 nate and sulphuric acid, and passed dry through the calcium chloride until 

 saturation resulted. 



Finally the entire jar was shaded, except a small opening which allowed a 

 beam of light to fall on the whole plant. Constant illumination was main- 

 tained by changing the position of the opening every 10 minutes, so that no 

 shadow ever fell on the plant. By this method of shading it was found easy 

 to keep the temperature of the air within the bell-jar within one or two 

 degrees of that on the outside, provided the volume of the jar was at least 

 1,000 times that of the plant and the jar was opened and fresh air let in 

 every two hours. 



The amount of water given off by the plant was found from the gain 

 in weight of the calcium chloride plus or minus a correction ascertained 

 in the following way. The dew-point was taken at the beginning and at the 

 end of the period and from the Smithsonian tables the weight of water in a 

 cubic meter of saturated air at the temperature of the two dew-points was 

 found. The difference between these two weights was then multiplied by 

 the volume of the jar, proper corrections being made for the space occupied 

 by the objects under the jar. Corrections for altitude were made in a few 

 cases where the error from neglecting the correction amounted to more 

 than 0.001 gm. 



The following conditions are provided in the present method: (1) The 

 water lost by the plant can be measured accurately to milligrams by means 

 of well-constructed weighing-bottles and an accurate dew-point apparatus; 



(2) the duration of the experiment may be from 15 minutes to 2 hours; 



(3) transpiration readings in sunlight may be obtained by means of a special 

 method of shading; (4) the air temperature and humidity within the jar 

 may be made to closely approximate the condition of the outside air by a 

 proper manipulation of the shading device and of the amount of calcium 

 chloride exposed; (5) the influence of air currents can not be measured. 



In order to measure the evaporation rate of a less complex surface under 

 similar conditions and thus to secure data regarding the behavior of the 

 plant itself, a second bell-jar was set up, identical with the first one, except 



* The use of carbon dioxide as a cooling agent was suggested by Dr. B. E. Livingston. 



