Cannon: A method of measuring transpiration 525 



Testing the polymeter method 



Laboratory experiments for the purpose of testing the accuracy 

 of the method indicate that the polymeter used for the present 

 object generally registers too low. This is due to the slow diffu- 

 sion of moisture in a confined space as well as to a tardy reaction 

 of the instrument itself. But these experiments show, on the 

 other hand, the very important fact that when the instrument is 

 handled properly, in a uniform manner, it will give results which 

 are consistent. 



The succeeding experiments, which were conducted in the 

 laboratory in a manner to be detailed, give some idea of the 

 behavior of the instrument in regard to the rate of increase or 

 decrease of moisture in the bell-glass. 



Experiment No. i. — Reaction of the polymeter to a change in 

 the relative humidity zvliich is approximately one per cent, a minute. 

 . The apparatus was arranged in the following manner: A bell- 

 glass was placed over a polymeter and sealed to a ground-glass 

 plate. Inlet and outlet tubes were arranged so that the amount 

 of moisture in the bell could be increased or diminished at pleas- 

 ure. At the beginning of the test the humidity of the atmosphere 

 of the bell-glass was increased to a certain per cent, which was 

 indicated by the polymeter, after which a portion of it was with- 

 drawn and collected in sulphuric acid drying-tubes. The poly- 

 meter was again read. The amount of vapor which was taken 

 away from the bell-glass was then computed by the hygrometric 

 niethod as given in the preceding pages, and it was compared 

 with the amount as determined by the increase in the weight of 

 the drying tubes. According to the polymeter the vapor removed 

 was %Z mg. ; the tubes gained 90 mg. in weight. Thirty-seven 

 minutes were taken to remove the moisture from the bell-glass, 

 during which the polymeter recorded a decrease of 33.5 per cent. 

 in the relative humidity. 



Experiment No. 2. — Reaction of the polymeter to a variable 

 change in the relative httmidity. The apparatus was set up as 

 before. Calcium chloride drying-tubes w^ere used. A stream of 

 moist air was passed through the bell-glass from 10:41 until 

 10:47, when all openings were again closed. 



The following data were recorded : 



