18 



The first observation excepted. the différences between 

 the numbers obtained were not larger than could be expec- 

 ted as a resuit of the many errors in the observation. 

 Indeed the rate of assimilation seemed to be constant under 

 unaltered circumstances. 



The fact that at the first observation the numbers were 

 lower than at the foUowing ones, we hâve observed 

 several times, and will be discussed further on. 



In the previous experiment and the experiments about 

 the influence of température, the velocity of the water- 

 current was kept as constant as possible. However this 

 did not succeed entirely, so it was necessary to examine 

 the casual influence of the velocity of the watercurrent 

 on the results. Table IV gives the results of a séries of 

 experiments, carried out on that purpose. The maximum 

 velocity of the watercurrent, which could be obtained 

 with our apparatus, was gradually diminished to the half. 

 The first flask was as usual fiUed with water, that had not 

 passed along the shoots. 



Table IV. 

 The rate of as sim ila t ion when changing the 

 velocity of the watercurrent (24^ C, 2482 candies). 



') Water coUected directly. 



