FORMALDEHYDE HYPOTHESIS 63 



hyde is not formed, provided every precaution is taken to 

 ensure the absence of impurities. If it does form a stage in 

 the photosynthetic process, it is immediately condensed to 

 some other substance, possibly sugar, or is used up in some 

 other way. 



The contention of Willstatter and Stoll based on the as 

 similatory ratio is ingenious, but, since the formaldehyde was 

 not actually demonstrated, the evidence is not convincing. 



Feeding Experiments. — That plants in abnormal condi- 

 tions, deprived of their natural raw food materials, may 

 make use of other substances in the elaboration of their food 

 shows that the plant is a transcendental chemist and in 

 virtue of its powers can to a certain degree make use of the 

 substitutes. The facts do not prove that these substitutes 

 are normal intermediate products in the natural process. If, 

 on the other hand, it is held that this kind of evidence is of 

 direct value, then the weight of evidence is on the side of 

 formic acid. 



The Production of Sugar from Formaldehyde. — The evi- 

 dence for the condensation of formaldehyde to hexose sugars 

 in the laboratory under the influence of various catalysts is 

 convincing : but valuable though such evidence be, especially 

 in providing points d'appui, it does not follow that such a 

 sequence obtains in the plant. The formaldehyde hypothesis 

 requires a hexose as the initial sugar : the general opinion 

 of those who have concerned themselves with the final products 

 of carbon assimilation is that sucrose, not a hexose, is the 

 first recognizable sugar, but, as has been pointed out on an 

 earlier page, the known facts may equally well indicate that a 

 hexose is the first formed sugar. 



SUMMARY OF THE OPINION OF WILLSTATTER AND STOLL. 



Willstatter and Stoll agree with Baeyer that formalde- 

 hyde is the intermediate link between the carbon dioxide 

 supplied to the plant and the carbohydrate synthesized with 

 the help of chlorophyll. It has already been pointed out 

 (p. 60) that their chief argument in support of this view is 

 based upon the constant value, 1, obtained for the so-called 



