1904] BENNETT: ARE ROOTS AEROTROFPIC 251 
there was unquestionably some difference; and, notwithstanding 
this, it may be seen that the roots, although influenced in growth, 
did not make any definite curves with references to these gases, 
even when 100 per cent. CO, was used. In nature one can 
hardly conceive of conditions which would give a greater differ- 
ence in the composition of gases on opposite sides of the roots. 
3. With roots in earth. 
In the lower half of a battery jar was fastened securely with 
sealing wax a glass plate in a vertical position so that gas could 
not pass the septum. Water was placed in the bottom to the 
depth of 2 or 3°, and pieces of broken plant crock were thrown 
in loosely. A glass tube was brought down to the bottom of 
each chamber, with its lower end curved upward under a piece 
of broken crock. The whole jar was now filled with moist earth. 
The seedlings were planted in a line directly above the upper 
edge of the glass partition. A different gas was led into each 
chamber, having passed on its way from the gasometer through 
liter bottles of water, that the rate of flow might be known and 
kept the same for each gas. | 
In this experiment the roots were in practically natural con- 
ditions, except for the gases used. If they were sensitive to the 
gases, and responded by curving toward one or the other, then 
acrotropism would be demonstrated. 
At first, very short seedlings were used and the gases allowed 
to diffuse slowly through the earth; but later, longer roots were 
used and the gases passed in more rapidly. In every case co, 
was passed into one side of the chamber and air into the other, 
the volume of each varying from 18 to 40 liters during the experi- 
ment of 24 hours’ duration. Of the 165 seedlings of Raphanus 
sativus, 31 were positive toward the air, 34 negative, and 100 
were neither positive nor negative, 98 of the last being straight. 
In the last two sets, when the 60 seedlings were planted twelve 
hours before the gases were led into the chambers, the majority 
of the roots were straight at the close of the experiment. 
In a modified form of this experiment the seedlings were 
planted in a vertical layer of earth or sawdust 18™ long by 12™ 
