104 BOTANICAL GAZETTE [FEBRUARY 
tive timothy meadow. In estimating the results the weight of the 
roots and tops was separately taken. 
These results show (with the exception of no. 4) a good correlation 
between transpiration and root growth. The order of responses Is 
the same in the respective columns. In the case of no. 2, where no 
increase or decrease was observed in transpiration, a like relation is 
observed in the root growth. There was a slight increase, however, 
in the transpiration per unit of green growth, due perhaps to the 
fact that the green weight of these plants was below normal. 
The general import of the experiments presented in tables A! 
and VI seems to be that the addition of pyrogallol or tannic acid 
causes a marked acceleration in transpiration, and that this accelera- 
tion is correlated more with the growth of roots than of the green 
parts of the plant. As a consequence of this response, the amount of 
water transpired per unit of green weight appears larger than the 
response to any other agent used. ~~ 
In many respects the growth of wheat plants in a soil extract pre- 
‘ viously treated with finely divided solids resembles that in extracts 
to which pyrogallol has been added. Undoubtedly the action of the 
two is similar in alleviating some unfavorable condition which prevl- 
ously existed, but was removed by the treatment in question. ihe 
pyrogallol must enter into chemical combination with some deleterious 
compounds in the soil extract or cause some process of oxidation, 
while the finely divided solids such as carbon black or ferric hydrate 
absorb and remove from solution the same deleterious substances. 
It is not proposed to dwell upon the general aspect of this problem 
in this place, since it has been treated in other papers and by various 
authors. The point to be demonstrated here is the effect of this 
treatment upon the transpiration per unit of green weight. 
A large increase in transpiration is usually caused by the treat- 
ment with absorbing agents. In most cases, the substance employed 
as an absorbing agent was added to the soil extract, stirred or shaken 
frequently during a half-hour, and then filtered out before the extract 
was used as a culture medium for plants; but the same or slightly 
greater increases were produced by leaving the absorbing agent in the 
solution during the time the plants were growing. 
Ordinary distilled water, which contains small amounts of poison- 
