Oct. 2 3l 1916 Daily Transpiration during Normal Growth Period 
205 
determined between the daily transpiration of each crop and the radia¬ 
tion, temperature, wet-bulb depression, wind velocity, and evaporation 
from the shallow and from the deep tanks, making 6 correlation tables 
for a crop, or about 200 tables in all. In order to avoid in the correla¬ 
tion studies the error arising from the change in the transpiration 
coefficient as the crops develop, the writers have, as outlined above, 
compared the transpiration ratio on consecutive days with the ratio of 
the intensity of the- weather factor on corresponding days. The diffi¬ 
culty arising from the asymmetry of these ratios has been avoided by 
taking the logarithm of the ratio in each case. This is, of course, 
equivalent to taking the difference of the logarithms of the two terms 
of the ratio, which was the procedure actually followed. In other words, 
the data in Tables I and III were first converted to a logarithmic basis, 
and the logarithmic differences then determined day by day. In brief, 
then, the actual correlation is between the logarithm of the transpira¬ 
tion ratio on consecutive days and the logarithm of the ratio of the 
intensity of the given weather factor on corresponding days. The 
transpiration of a crop on a given day thus normally enters twice into 
the correlation table for that crop, once as the numerator and once as the 
denominator of the ratios involving the transpiration on consecutive 
days. The results of the correlation studies are summarized in Tables 
XXI to XXIII, inclusive. 
Smaix Grains.— The small grains in Table XXI show such similarity 
in the response of transpiration to various weather factors that the group 
may profitably be considered as a whole. The correlation coefficients are 
higher in 1914 than in 1915. It will be recalled that the latter season was 
cooler, more cloudy, and included many more rainy days. The correla¬ 
tion of transpiration with air temperature is usually, in 9 cases out of 12, 
slightly higher than with radiation, but the difference is usually less than 
its probable error. 
The correlation coefficients of transpiration of the small grains with 
wet-bulb depression and with evaporation (shallow tank) show a strik¬ 
ing agreement. Considered as a group, they are markedly higher than 
the correlations with either radiation or temperature, and constitute the 
highest correlations in the series. Their agreement appears to be due, 
in part at least, to the fact that the depression in the wet-bulb tempera¬ 
ture is dependent upon the rate of evaporation from the muslin covering. 
It is of interest in this connection to recall the difference in exposure of 
the wet-bulb instrument and the shallow tank, the former being shaded 
from solar radiation, while the tank was blackened and fully exposed to 
the sun’s rays. 
The evaporation from the deep tank showed a lower correlation with 
transpiration than the shallow tank. The deep-tank evaporation is corre¬ 
lated with transpiration approximately to the same extent as radiation 
or temperature. In 1915 the correlation of the deep-tank evaporation 
with transpiration was markedly lower for many of the small grains. 
