636 
Journal of Agricultural Research 
Yol. V f No. 14 
the maxima referred to in the investigations cited. One maximum occurs 
in the morning about 7 or 8 o’clock, and a second and greater maximum 
is found in the afternoon between 4 and 6 p. m. 1 In other words, the 
transpiration graph shows a tendency to rise earlier in the morning and 
fall later in the afternoon than the evaporation graph. This is evident 
in each of the three graphs presented in figure 20. 
This result is capable of two quite dissimilar interpretations. If the 
assumption is made that evaporation constitutes a correct summation of 
the influence of environment on transpiration, it follows logically that the 
departure of the transpiration-evaporation ratio from a constant value is 
due to a decrease or increase in the transpiration coefficient. It must, 
Fig. 20.—Graphs showing hourly ratio of transpiration to evaporation as plotted in figure 19. 
however, be recognized that all evaporimeters do not respond to their 
environment in the same way. A large deep tank does not have the same 
daily graph as a shallow tank. A filter-paper evaporimeter does not fol¬ 
low the graph of the porous-cup atmometer. If none of these agree, can it 
be said without further proof that the evaporation rate of any one of them 
is proportional to the transpiration rate of a plant which responds freely 
to its environment? The fact that the transpiration graph is so uni¬ 
formly asymmetrical with respect to noon in our determinations and that 
the evaporation graph is so uniformly symmetrical would indicate that 
the two processes were not controlled in the same way by the physical 
factors of the environment. The writers are inclined to the belief that 
1 The hourly values of transpiration and evaporation at night are so small that the observational errors 
make the ratio uncertain, and the night ratios will consequently not be considered at this time. 
