Vegetable Assimilation and Respiration. 3 1 1 



intensity of the other factors, so that CO2 is a Hmiting factor in Blackman's 

 sense, then the photosynthetie rate will necessarily be high at first and steadily 

 decline as the diffusion gradient slackens, until an approximately steady 

 gradient is established, bringing along CO2 at the rate which is that appro- 

 priate to the general concentration. Here we have an initial aberration whicli 

 is exactly the inverse of the oxygen aberration. This is illustrated in fig. 3, 

 drawn from the present work. This phenomenon repeats itself after every 

 movement or stirring, which brings up more highly charged layers of CO2 

 solution, and also after every period of darkening during which CO2 has 

 again accumulated round the plant. This type of initial effect has also been 

 often observed, but hardly ever precisely interpreted. 



It is interesting to note that it only occurs when the conditions of experi- 

 mentation are such that CO2 is limiting. When weak light is limiting and 

 the CO2 concentration is in excess, these variations of C02-supply cannot, by 

 definition, manifest themselves as changes of bubble rate, and it is found 

 throughout the present work that the falling initial curve occurs only in the 

 one class of ca.ses and not in the other. 



This initial aberration cannot be removed by any preliminary treatment. 

 If investigation is to be made on the relation of assimilation to CO2 concen- 

 tration or to KHCOa solutions, this effect is bound up with the problem stated. 

 The only theoretical way would be to use solutions which were always in 

 violent movement, so that there were practically no layers less charged with 

 CO2 than their neighbours, but in such cases it would be very difficult to 

 keep the lighting of the shoot absolutely constant. 



Be. Combination of Initial Oxygen Diffusion Effect with Initial CO2 Diffusion 

 Effect. — We now see the experimenter is faced with two initial effects of 

 entirely opposite nature, one giving a rising series of readings and the other a 

 falling series. The combination of them suggests interesting complications as 

 each is independent of the other. It would be possible for the two to neutralise 

 one another if the solution started with exactly the right deficit of oxygen 

 and excess of CO2. It seems highly probable that in the cases here and there 

 in the literature where the bubble rate starts at once at the rate that it continues 

 to maintain we have such a relation. What appears to be an ideal experiment 

 will then be only a chance cancelling out of two unmeasured sources of error. 



The experiments in this paper show big CO2 effects as compared with other 

 workers because special trouble was taken to eliminate the oxygen effect. No 

 one is directly interested in the relation of bubble rate to the oxygen-content 

 of the water, and everyone in its relation to the C02-coutent, so the best 

 procedure seems to be to eliminate one source of errpr and so get a clear 

 measure of the magnitude of the other source. 



