20 MISC. PUBLICATION 257, U. S. DEPT. OF AGRICULTURE 



carried on in a closed container, so that the humidity is constantly 

 increasing. 



The cobalt chloride method, which has been in use for many years 

 but which was revised and standardized by Livingston and his co- 

 workers, depends essentially upon the change in color of cobalt chlo- 

 ride with change in moisture content. Strips of paper impregnated 

 with the salt are applied to the leaf surface and the rate of change of 

 color is taken as an index to the transpiring rate of the plant. This 

 method, which is thus seen to be qualitative rather than quantitative, 

 in contrast with the two methods just described, is now used by a great 

 many workers to compare relative rates of water loss. The collodion 

 method is similar and depends upon the fact that collodion loses its 

 transparency and becomes cloudy when a very small amount of water 

 comes in contact with it. In using this method a solution of collodion 

 in alcohol or ether is applied with a brush in a thin layer to the surface 

 of the leaf. Very shortly the solvent evaporates, leaving a dry film 

 of collodion. As transpiration proceeds, the collodion absorbs water, 

 and the rate of clouding indicates the transpiration rate. 



By far the best method is the direct determination of water loss by 

 noting the change in weight of the plant (method IV). This may be 

 done by comparing parts of plants or by using the entire plant, but 

 the latter method is by far the more reliable and should be used 

 wherever possible. With detached organs the experiment must be 

 limited to a few hours or days, and there is no question but that 

 enormous errors have arisen in transferring the data from cut parts 

 to whole plants (as will be shown under the discussion on water require- 

 ments), but the difficulty of using large plants such as trees is obvious. 



Furthermore the cutting of a leaf or twig from a plant, even 

 though the cut is made under water and the plant part is then left in 

 water, changes the transpiration rate. Although Bartholomew (10) 

 found that when citrus leaves were cut they lost less water than before, 

 most observers report the contrary, which seems more natural to 

 expect. When attached to the plant, leaves are seldom transpiring 

 at a maximum because of an insufficient water supply and because of 

 the energy required to absorb water through the tissues between the 

 leaf and the soil. When cut and placed in water these hindrances 

 disappear. Thus Kamp (119) found in studying 30 species of woody 

 plants that, when the leaves were cut, they transpired up to 130 per- 

 cent more for the first 15 minutes than before cutting, after which 

 the transpiration rate gradually decreased to its former level. Pal- 

 ladin, as recorded by Molliard (157, p. 24%), likewise found that one 

 detached oak leaf transpired 3.2 g in 24 hours while a twig of the 

 same plant with 180 leaves lost only 28.2 g in the same period. The 

 transpiration of the isolated leaf was thus 20 times that of the others. 

 The reason for these changes is not clearly understood. If the 

 leaves are really freed from previous stresses, why does the transpira- 

 tion later decrease to normal instead of continuing at the maximum? 

 Changes at the cut surface as well as in the size of the stomatal 

 openings and in the suction forces involved probably play a role in 

 these phenomena, but this entire question needs further study. 



FACTORS INFLUENCING TRANSPIRATION 



The factors influencing transpiration are either external (light, 

 temperature, etc.), depending upon the environment, or internal and 

 connected with the inherent nature of the plant. 



