THE EXCHANGE OF IONS 
Jena glass beakers were chosen as containers, after preliminary 
experiments had shown that any difference between Jena glass and 
opaque quartz or paraffined glass due to the solubility of the former 
was much less than the experimental error from other causes. For 
example the diffusion of carbon dioxide from freshly distilled water is 
often sufffcient to increase the electrical resistance of water left 
standing in any one of these three types of containers, thus offsetting 
any change in the opposite direction due to the solubility of the 
container. In fact, a common method of improving distilled water 
for use in conductivity determinations is to pass through it a stream 
of air, free of carbon dioxide. The loss of carbon dioxide from dis- 
tilled water left standing in contact with the air is sufficient to lower 
its conductivity very appreciably. The water used for distillation 
was moderately ''hard," i. e., it contained bicarbonates which liberated 
carbon dioxide during the process of distillation. The fresh distilled 
water therefore contained more carbon dioxide than it did later 
after standing in contact with the air. It is self evident that water 
which is to be used for culture solutions should not be freed from carbon 
dioxide, in case it is wished to study the excretion or absorption of other 
electrolytes than carbonic acid from the roots. Since there must be- 
a considerable lag in the attainment of CO2 equilibrium between air 
and a solution to which CO2 is supplied continuously, it follows that 
the electrolytic method shows, as the first effect of placing roots in 
distilled water, an increase in conductivity due to the excretion of CO2, 
It follows that if CO2 constantly diffuses into the air from the solution, 
increase in conductivity on account of excretion of CO2 from the roots 
must reach a maximum and thereafter remain reasonably constant as 
long as the roots continue to supply CO2. The maximum conductance 
due to CO2 will be reached when the rate of diffusion of CO2 from the 
solution into the air equals the rate at which CO2 is supplied by the 
roots. The error in the conductance of a solution due to CO2 can 
not be less than the conductance of water in equilibrium with air 
containing 3 parts in 10,000 of COo. This lower limit of error is 
about equivalent to the conductance of half-million th molecular NaCl. 
The upper limit of error can be approximately estimated as equivalent 
to the conductance of 16 millionths molecular NaCl. This estimate 
is based upon the measurement of absorption from solutions which 
show almost perfect absorption of the salt content by the roots. In 
reality the error is not as great as this figure, for probably no culture 
