Researches on Vegetable Assimilation and Respiration. 305 



avoiding errors in bubble-counting due to variations in size of the gas 

 bubbles, and (B) Method of diminishing errors due to gaseous diffusion in 

 the water round the plant. 



(A) Errors clue to Variation in Size of BvMles can he Avoided hy the Use of 

 Glass " BubUers " and BvMling Cup. 



Aa. Variations in Size of Bubbles due to Alterations of Cut Stem Aperture. — 

 It is obvious that if, over a period of time, the size of the bubbles given off 

 from the cut stem of a submerged water plant varies, either irregularly or by 

 steady drift, then merely counting the number of bubbles gives no adequate 

 measure of the volume of gas set free by the photosynthetic process. The 

 papers of all workers who have used the bubble-counting method contain 

 lamentations about the difficulty that this spontaneous variation produces in 

 attempting long comparative experiments, yet no proposal has been brought 

 forward for really overcoming it. Palliatives, such as cutting the surface 

 overnight, pinching, scratching, covering with varnish and pricking, and so 

 on, have been employed, but it is generally admitted that many experiments 

 suffer from such obvious variations of bubble-size that they must be 

 abandoned. 



In the course of the experiments now described, it was found practicable 

 to manufacture small glass nozzles to fit over the cut end of the shoot, and 

 provide a fine glass capillary opening of absolutely constant size. These 

 " bubblers," as they will be called, were made by first drawing out glass 

 tubing to an external diameter of 1'5 to 2 mm., with an internal bore just 

 sufficient to fit closely over the shoots to be experimented with. A short 

 region of this tube was then sharply drawn out at right angles to a very fine 

 capillary and broken off, to leave a minute aperture. Fitting such nozzles on 

 to the cut stem ensures a long steady succession of bubbles all of the same 

 ^ize. 



Ab. Variations in Size of Bubbles tvith Alteration of Internal Gas-Fressv/re. — 

 Such changes are well known when the raw end of the plant is used. The 

 openings of the plant capillary spaces are so fine that greatly increased rate 

 of bubbling delivers smaller bubbles than does a slow Mow of gas, and thus 

 quite invalidates the counted numbers of bubbles. This source of error is 

 also eliminated by the glass bubblers, because the apertures are larger than 

 in such plants as Elodea, and the bubbles delivered are larger and come off' 

 much more slowly, so that a rate of flow which would affect their individual 

 size is never reached in normal photosynthesis. 



A nozzle that gives a bubble every 5 to 10 seconds is the most satisfactory 

 size. It is found that, in constant conditions, the variation of timing five 



