_ auremrernerreres Te ee SL 
RECORD OF THE EXPERIMENTS. 19 
tion, we have no means of knowing how much has been taken from 
the known components and how much from the unknown. Suffice it 
to say that unknown electrolytes are drawn upon to a considerable 
extent, for certain plants (e. g., beets) actually increase the resistance 
of distilled water in which they are grown. 
The chief electrolytic impurity -introduced into culture solutions 
from the air is carbonic acid. It is well known that water which is 
free from dissolved carbon dioxid can not be prepared by distillation 
in the air. Walker and Cormack‘ have shown that water in equilib- 
rium with normal air has a total concentration with regard to 
M 
80,000 
actual and potential carbonic acid (H,CO, and CO,). Of this total, 
14.4 per cent is dissociated into the ions H+ and HCO,-—, rendering 
M 
500,000 
solution of a univalent salt. In a water culture solution the partial 
conductivity due to dissociated carbonic acid would remain constant 
during an experiment except for the fact that roots are continually 
excreting carbon dioxid into the solution. With very dilute solu- 
tions there must be a considerable lag in the readjustment of the 
equilibrium between dissolved and atmospheric carbon dioxid. It 
is therefore probable that no other error in the work recorded in this 
paper is comparable in magnitude with that due to fluctuation in the 
carbon dioxid concentration of the solutions. It does not follow, 
however, that this error seriously vitiates the results; it is discussed 
merely in order to show that work with exceedingly dilute solutions 
can not be done with sufficient accuracy to warrant the correction 
of small experimental errors. 
the conductivity of the solution comparable with that of a 
Since the culture solutions were not analyzed after our experiments, 
the conductivity readings afford only a rough idea of their salt con- 
tent. In the statement of results the phrase ‘Concentration after 
growth” must be understood to mean that degree of concentration 
of the same salt or salts present in the solution before the growth of 
the plants which would give a solution having the same conductivity 
as that found te exist at the end of the experiment. Since, of course, 
the same salts are not present at the end of an experiment as at the 
beginning, the results as thus stated are at best only approximate. 
It is probable, for example, that in all the experiments reported below 
the salts excreted by the roots were in part univalent, whereas 
bivalent salts were absorbed. If a substitution of univalent salts 
for bivalent should take place equivalent for equivalent, the con- 
ductivity of a culture solution during growth might remain about the 
1 Walker, James, and Cormack, William. The Dissociation Constants of Very Weak Acids. Journal 
of the Chemical Society (London), vol. 77, pt. 1 (Transactions), 1900, pp. 5-21. 
e 231 
