. tgt1] LIVINGSTON—LIGHT INTENSITY AND TRANSPIRATION 435 
about the same part of each day. It is quite obvious that in other 
regions the results might have been different, but I am convinced 
that the present data would agree fairly well with those for the 
hottest summer days in most parts of the United States. It is to 
be remembered, however, that the work was done in the arid region, 
albeit in the moist season, and that humidity has not yet been 
investigated with reference to its quantitative effect on plant 
transpiration. In the absence of a better method for describing 
weather conditions, it may be stated that the temperature varied 
within the limits 30-35° C., and the sky was not without haze, 
though clouds were rare. We have considered nothing weaker 
than strong diffuse light, that obtained under a screen of tent 
canvas. The plant stomata were probably always in the day con- 
dition throughout these experiments, and incipient wilting, if it 
occurred, was probably not generally a controlling factor in the 
transpiration rates. 
Several different shade intensities were included in the tests, 
but an inspection of the tables will convince one that the fluctua- 
tions in the correction coefficients do not appear to be related to 
any particular shade. In the following derivation of conclusions 
all tests will be considered as tentatively equivalent, and no attempt 
to weight the averages will be made. The coefficients of correction 
will be treated as the main criterion for judging of the relative 
degrees of sensitiveness of the plants and instruments toward 
variations in light intensity. 
1. Considering all coefficients (tables III, V, and VI, not the 
averages) with values between 0o.go and 1.10, inclusive, as equal 
to unity, we see that all of those greater than unity have reference 
to the porous cup atmometers. The other instruments always 
recorded greater differences between the two light intensities than 
did any of the plants. To study the distribution of the different 
forms of coefficient more in detail we may proceed to classify them 
in each of these two groups of instruments. The frequencies of 
occurrence of coefficients less than, equal to, and greater than unity, 
for the three atmometers and the three plants, are presented in table 
VII. For example, there are six coefficients greater than one 
occurring for the brown atmometer and Physalis, only one for 
