SUMMARY 



19 



By reducing the capacity, I have been able to obtain record 

 of the successive evolution of such small quantities of gas 

 as 0-2 cubic mm. Bubblers for different requirements can, 

 however, be constructed in which each bubble represents 

 o • 2 to 5 cubic mm. I may give data of a particular experi- 

 ment in which the successive bubbles under bright sky light 

 were found to take place at intervals of twenty seconds ; 

 there were thus 3 X 60 = 180 bubbles per hour. In a 

 parallel experiment the total quantity of oxygen collected 

 in six hours from 10 a.m. to 4 p.m. was 1100 cubic mm. or 

 183 cub. mm. per hour. This corresponded to about the 

 180 bubbles of the first experiment. The volume of each 

 bubble in the particular Bubbler was therefore 1 cubic mm., 

 which was also found to be the calibrated value. The only 

 precaution that need be taken to ensure accurate working 

 of the Bubbler is to cover the bent tube with a thick piece 

 of silk so as to maintain it at a uniform temperature, the 

 silk having two small slits for the observation of the move- 

 ment of the oil-valve. The precaution just mentioned is, 

 however, unnecessary when the experiment is carried out 

 in a room protected from draughts of air. 



The evolution of oxygen in photosynthesis is continuous, 

 but the formation of the bubblets under water is discon- 

 tinuous ; the intervals are longer the greater the depth of 

 immersion. The gas-evolution could be made continuous 

 by placing the cut end in air at the lower end of the Bubbler ; 

 but in that case the cut end would become dried up, thus 

 obstructing the passage of the gas. The cut end is therefore 

 placed about 2 mm. below the surface of water, so that the 

 process of bubbling becomes almost continuous. 



In the next chapter is described the special device by 

 which the successive evolution of equal volumes of gas is 

 automatically recorded. 



Summary 



The errors introduced into the method of bubble-count- 

 ing by the spontaneous variation in the frequency and 



