HYDROHARMOSE 109 
face exposed to the air. No compensation occurs in the case of submerged 
plants; floating hydrophytes possess a single transpiring leaf surface, while 
the leaves of amphibious plants behave as do those of mesophytes. The 
whole question of response to water stimuli thus turns upon the compensa- 
tion for water loss afforded by -water supply where the latter is moderate 
or precarious, and upon the compensation for water supply furnished by 
water loss where the supply is excessive, submerged plants excepted. 
155. The measurement of absorption. As responses to measured 
stimuli of water-content and humidity, it is imperative that the amount of 
absorption and of transpiration he determined quantitatively. It is also 
extremely desirable that this be done in the normal habitat of the plant. 
A careful examination of the problems to be met quickly discloses the great 
difficulty of obtaining a direct and accurate measure of absorption under 
normal conditions, especially in the field. For this purpose, the ordinary 
potometric experiments by means of cut stems.are valueless. The use of 
the entire plant in a potometer yields much more trustworthy results, though 
the fact that the root is under abnormal conditions can not be overlooked, 
especially in the case of mesophytes and xerophytes. While potometric 
conditions are less abnormal for amphibious plants, the error is not wholly 
eliminated, since the roots normally grow in the soil. The potometer can 
be made of value for quantitative work only by checking the results it gives 
by means of an instrument or a method in which the plant functions nor- 
maily. In consequence, the potometer can not at present be used to measure 
absorption directly, though, as is further indicated in the discussion of 
transpiration, it is a valuable supplementary instrument, after the check 
mentioned has been applied to its use with a particular species. 
An estimate of the amount of absorption may be obtained either in the 
field or in the control house by taking samples from the protected soil at 
different times. Since it is impossible to determine the weight of the area 
in which the roots lie, and since the soil water is often unequally distributed, 
this method can not yield exact results. An accurate method of measuring 
absorption under essentially normal conditions has been devised and tested 
in the control house. The essential feature of the process is the placing a 
plant in a soil containing a known quantity of water, and removing it after 
it has absorbed water from the soil for a certain period. In carrying out 
‘the experiment, a soil consisting of two parts of sod and one of sand was used, 
since the aeration is more perfect and the particles are more easily removed 
from the roots. The soil was completely dried out in a water bath and then 
placed in a five-inch battery jar. The latter, together with the rubber cloth 
used later to prevent evaporatiou, was weighed to the decigram. A 
weighed quantity of water was added, and the whole again weighed as a 
