yo 



ments. Transpiration was determined by weighing the 

 entire culture at the beginning and end of the observa- 

 tion period, 1 The method was to run air through the 

 culture solution for from 20 to GO minutes, then to 

 run through a stream of the gas tinder test for a like 

 period. The difference in the transpiration indicated the 

 effect of the gas. Experiments by this method with pure 

 carbon dioxide gave decreases of transpiration of approxi- 

 mately 50 percent, ilixtures of yO and 80 percent carbon 

 dioxide with air gave somewhat smaller decreases of 

 transpiration. .Hydrogen also decreases the transpiration 

 btit to a less extent than did carbon dioxide. The 

 average decrease was about 25 percent, i'/hen carbon 

 dioxide was followed by air the transpiration rate went 

 lip again almost to its initial value. When carbon di- 

 oxide was followed by hydrogen the transpiration rate 

 rose to about the same value as when hydrogen was used 

 following air. Experiments on plants the roots of which 

 had been killed by dippirig in boiling water showed a de- 

 crease of transpiration rate with carbon dioxide but 

 none with hydrogen. Experiments with cut twigs of 

 Tilia pubescens . apa mania africana and Abut i Ion striatum 



1. It is obvious that this does not measure 

 root absorption if there was loss or gain of water by 

 the plant tissue, as, for instance, by wilting. However, 

 the error is probably without effect on the general 

 meaning of the results. 



