526 Cannon : A method of measuring transpiration 



Time Relative humidity Temperature 



10:41 43 74"" F- 



10:47 ^^ 74 



11:10 63 74 



It will be observed that the humidity increased 18 per cent in 

 6 minutes (10:41-10:47). From 10:47 to 11:10 the instrument 

 was allowed to settle, during which time it show^ed an increase of 

 2 per cent. This was due to the tardy response of the polymeter, 

 since the moisture was equally distributed throughout the bell- 

 glass by reason of the manner of introducing it 



At 11:10 the stream of moist air was again sent through the 

 bell-glass ; it was stopped at 11:17. 



11:17 68 74 



11:37 67 75 



The rise of 5 per cent of the humidity in 7 minutes (i 1:1 

 was closely followed by the polymeter, since it showed no change 

 except only a fall of one per cent, due to the higher temperature, 

 in the succeeding period of rest 



At 1 1:37 a portion of the moisture was removed from the bell- 

 glass and collected in calcium chloride drying-tubes. This proc- 

 ess required 12 minutes, during which the humidity fell 25 per 



cent 



11:49 42 75-5 



According to calculations based upon the polymeter, the 

 amount of moisture withdrawn was 42 mg.; the drying-tubes 

 gained 55 mg. That is to say, the polymeter, on account of the 

 too rapid decrease in the humidity, lagged considerably in its 



recording. 



Best methods of conducting the experiments 



When an experiment is performed it is necessary to compro- 

 mise between the desirability of making it as short as possible, in 

 order to maintain conditions for the plant as nearly normal as 

 practicable, and the fact that the instrument, for reasons given 

 above, does not record instantly the humidity of the air. Conse- 

 quently the rate of increase in the relative humidity should be as 

 nearly constant for all experiments as possible, particularly when 



