112 LIGHT, VEGETATION AND CHLOROPHYLL 



obtained from carefully performed experiments, would 

 certainly have stimulated much more research if the war had 

 not diverted scientific investigation into other channels. Let 

 us consider for the moment both the difficulties of principle 

 and the experimental difficulties. 



The experiments are made on living material and the 

 choice of the plants, and their preparation, are very delicate 

 operations. We have seen how numerous are the factors, in 

 the actual conditions of the experiment or in the previous 

 history of the plant, which have an influence on its photo- 

 synthetic capacity. 



To obtain the maximum efficiency, and to ensure that 

 hght is not the hmiting factor, rather low illuminations are 

 used, e.g., one-thousandth of the solar illumination in 

 Warburg's experiments. Respiration is then much in excess 

 of photosynthesis, so that what is measured is not the 

 liberation of oxygen and the absorption of carbon dioxide, 

 which are the indications of photosynthesis, but the reduction 

 of the rate of the reverse phenomenon — respiration — at the 

 moment when the plant passes from darkness into the light. 



It is supposed, without direct proof, that the rhythm of 

 respiration is not altered by illumination and that the only 

 cause of variation in the gaseous exchanges is the operation of 

 photosynthesis. This question would demand a thorough 

 investigation which would not be easy to undertake. 



In fact, even if the quantity of Hght which falls on a 

 surface can be measured, and if the fraction absorbed can 

 also be measured by deducting from the first quantity the 

 part reflected and the part transmitted (although in this case 

 the methods employed have rarely been beyond criticism), 

 it is difficult to know by what constituent of the plant sub- 

 stance this fraction has been absorbed and in what measure 

 chlorophyll has profited from the absorption. This apparently 

 is a problem to which a complete solution has not yet been 

 found. 



When we consider the nature and extent of these difficulties, 

 we can easily understand that the most recent estimates of the 



