1912] LIVINGSTON & BROWN—TRANSPIRATION 319 
ing constant, a diurnal accumulation of non-aqueous materials, as 
of carbohydrates, oils, proteins, etc., might result in a great increase 
in the contained non-aqueous material, and a corresponding fall in 
the percentage of moisture. Such an accumulation, with its 
accompanying decrease in the percentage of moisture, however, 
would have no marked influence upon relative transpiration. This 
question may be studied quantitatively through the daily fluctua- 
tions in relative water content on the basis of unit leaf area, a 
phase of the problem with which we have not concerned ourselves. 
Experimentation 
Our method of experimentally attacking the problem as to 
whether a definite diurnal fall in the percentage of leaf moisture 
actually occurs was to gather, at different hours of the day and 
night, a large number of similar leaves from plants growing in the 
open soil, and to determine the percentage of moisture therein 
contained by the common method of weighing, drying at 100° to 
105° C., and reweighing. As the leaves were gathered they were 
placed immediately in tarred glass bottles and tightly stoppered. 
After being weighed, the open bottles were placed in the drying 
oven, and the final dry weight obtained without removing or 
handling the leaves. A large number of plants were available for 
this work, so that but few leaves were taken from the same plant, 
and the numerical result may be taken as fairly well approximating 
the average condition of the whole plot of plants of the species 
tested. The plants used were all spontaneous in the open ground — 
near the Desert Laboratory. In some cases the sample leaves were 
taken every two hours, in others less often, and a porous cup 
atmometer was operated and read at short intervals in the vicinity 
of the group of plants from which leaves were taken. The work 
was carried out at the Desert Laboratory of the Carnegie Institu- 
tion at Tucson, Arizona, in the summer of 1910. 
Preliminary tests of the leaf moisture in Physalis angulata var. 
Linkiana Gray and in Martynia louisiana Mill., made on July 28, 
. showed that the moisture content of the Physalis leaves was 744.7 
per cent of the dry weight at 6 a.m., and only 561.0 per cent at 
2 P.M., while the corresponding percentages in Martynia leaves were 
4 
