i62 THE LOSS OF WATER FROM PLANTS 



laboratory experiments potted plants are often used, the pot generally being 

 enclosed in a metal shell, and the soil surface sealed of? so that evaporational 

 water loss can occur only through the plant. For field or large-scale experi- 

 ments, metal receptacles have been found convenient in which case it is only 

 necessary to seal off or cover the soil surface in such a way as to prevent 

 evaporation. The method is limited in practical application to plants which 

 can be grown in readily portable containers. The transpiration rates of plants 

 as large as mature maize plants have been measured by this method. 



The loss in weight of the container and plants for a given time interval 

 may be considered to represent transpiration as the effect of other factors on 

 the weight of the set-up is usually negligible. If the experiment is to be 

 continued for more than a relatively short period it is necessary to provide 

 the container with a watering tube through which known volumes of water 

 can be introduced into and distributed throughout the enclosed soil at appro- 

 priate intervals. 



In employing this method the receptacle in which the plants are growing 

 may be weighed at selected intervals by manual manipulation or may be 

 placed on a balance which is so arranged that each loss of a definite increment 

 of weight (for example i g.) is automatically registered on a recording device 

 (Transeau, 191 1, Briggs and Shantz, 1915, Schratz, 1932, and others). 



2. Method of Collecting and Weighing Transpired Water-Vapor. — This 

 method requires a rather elaborate experimental set-up, but is the only way in 

 which the rate of transpiration can be determined quantitatively for plants 

 growing in out-of-door habitats. The plant is enclosed within a glass cham- 

 ber, the soil surface sealed off, and a stream of air taken directly from the 

 atmosphere circulated rapidly through this container. After passing through 

 the chamber the air is conducted through tubes or vessels containing a water 

 absorbent such as calcium chloride. The gain in weight of these absorption 

 tubes during the experiment represents the water-vapor transpired by the plant 

 plus the water-vapor which was introduced into the system from the outside 

 atmosphere. In order to determine the proportion of the water-vapor which 

 comes from the atmosphere it is necessary to set up a check apparatus in exactly 

 the same manner, except that no plant is enclosed in the glass chamber, and to 

 pass air through it at the sam.e rate that it is circulated through the set-up con- 

 taining the plant. The gain in weight of the water-absorption tubes in the 

 second apparatus represents water-vapor from the atmosphere. Such a method 

 has been used by Minckler (1936) for measuring transpiration from attached 

 branches of trees. 



3. Pototneter Methods. — Of a limited usefulness for the measurement of 

 transpirational water loss are instruments known as potometers, a type of 



