1906.] Mechanism of Carbon Assimilation in Green Plants. 325 



present, following on the primary reaction which gave rise to the formal- 

 dehyde, and having a much higher velocity. In such a case, one could not 

 expect to find formaldehyde in the final mixture. As the question at issue is 

 whether or not formaldehyde is produced at any stage in the series of reactions 

 which take place when sunlight acts on carbon dioxide under the conditions 

 described, and as it was found that formaldehyde could not in any way be 

 isolated from the mixture, it appeared that a careful study of the reactions of 

 formic acid and aldehyde with such substances as were formed under the 

 conditions of the experiments was the method most likely to settle the 

 question. It may be worth while to point out here that a careful distinction 

 must be maintained between those reactions which are endothermic, and 

 derive the necessary energy from the sunlight, and those in which the action 

 of light is purely catalytic. 



Solutions of uranium salts — especially those like the formate — have always 

 an acid reaction, and are probably hydrolysed to a considerable extent, so that 

 any hydrogen peroxide that is formed is not completely removed from the 

 system, but remains in equilibrium with the difficultly soluble uranium 

 peroxide. "We have, therefore, to consider reactions which may occur 

 between 



(i) Formic aldehyde and uranium peroxide. 



(ii) „ acid 



(hi) „ „ „ hydrogen 



(iv) „ aldehyde 



(v) „ „ „ uranic oxide. 



It has been shown by Geisow* that when hydrogen peroxide reacts with 

 solutions of formaldehyde in neutral or acid solution, hydrogen and carbon 

 dioxide are the sole products, no formic acid being produced. On the other 

 hand, the peroxides of lead and manganese give hydrogen and a metallic 

 formate. It has been found that uranium peroxide reacts in the dark with 

 formaldehyde, and that hydrogen and uranium formate are produced, but no 

 carbon dioxide. Hence, in such a system as the one under consideration, 

 where there is a constant but very small concentration of hydrogen peroxide, 

 there is the possibility of explaining the production of both hydrogen and 

 carbon dioxide, the reaction taking place between the formaldehyde and 

 hydrogen peroxide, though, of course, at the expense of the uranium peroxide 

 and this appears to be the probable explanation, when we consider that the 

 velocity of reaction in the case of hydrogen peroxide must be very much 

 greater than in the case of uranium peroxide. The fact that hydrogen and 



* 1 Ber. deut. chem. Ges.,' 1904, vol. 37, p. 515. 



