278 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9, 
whole series is taken, and this is considered to be loo percent; with this 
average as the standard, the relative rate of production of carbon dioxide 
under the influence of alcohol is calculated. 
20 J ^ — I , 
0 50 100 MINS. 
Fig. I. Curves showing the effect of ethyl alcohol on the production of carbon di- 
oxide by radish seedlings. To the left of the point O on the abscissa the curve shows the 
rate of respiration in tap water; to the right of that point, the rate under the influence of 
alcohol, except in curve A which represents the rate in tap water (without exposure to alco- 
hol). Curve B shows the rate in likf alcohol, curve C in 2M alcohol and curve D in /\.M 
alcohol. Each point represents an average of 12 experiments. Probable error of the mean 
is less than 3 percent of the mean. 
Before exposing the seedlings to a solution of alcohol, control readings 
are made with the seedlings in 3 cc. of tap water. The seedlings are then 
taken out of the apparatus and kept in a large dish of water. A bottle 
with a capacity of 250 cc, filled with a solution of alcohol, is fitted to the 
portion of the apparatus to which the tube which is to contain organisms 
is usually fitted. The liquid containing the indicator is now replaced by 
the same amount of liquid of the same concentration of alcohol as that con- 
tained in the larger bottle and having the same concentration of indicator. 
The whole apparatus is thus saturated with the alcohol vapor in equilibrium 
with the solution which is to be used for the experiment. Then 3 cc. of alcohol 
of the same concentration is placed in the tube and the air is again passed 
through the U-tube containing sodium hydroxide until the whole apparatus 
is again free of carbon dioxide. The organisms are now placed in the tube 
and the air is passed through the sodium hydroxide for three minutes, after 
which the reading is made in the same manner as in the control experiment. 
The relative rate of respiration of the seedlings in alcohol is calculated on 
the basis of the average of the control experiment, which was made immedi- 
ately before the experiment with the alcohol. 
