1911.] The Mechanism of Carbon Assimilation. 



103 



Two transparent quartz tubes, of about 20 c.c. capacity, were filled witb 

 the purest " conductivity " water obtainable, and the tubes were inverted in a 

 trough of mercury and placed symmetrically near the mercury-vapour lamp. 

 The water in one of the tubes was as nearly as possible gas-free, and a few 

 cubic centimetres of carbon dioxide were introduced into the other. Both 

 tubes were illuminated for about 12 hours, and the contents of each were 

 then examined for the presence of formaldehyde and hydrogen peroxide. 

 The solution of carbon dioxide was found to contain an easily recognisable 

 quantity of formaldehyde, most of which was in a polymerised form, whereas 

 none, either free or polymerised, could be detected in the water from the 

 other tube. Traces of hydrogen peroxide were present in both. All the 

 reagents used were carefully tested, and negative results were obtained with 

 a solution of carbon dioxide which had not been exposed to ultra-violet light. 



It appears from these experiments that ultra-violet light can effect a 

 measurable decomposition of aqueous carbon dioxide without the intervention 

 of an optical or chemical sensitiser, whilst under normal conditions some 

 such agent is required;* moreover, the results furnish very strong support 

 for the belief that both formaldehyde and hydrogen peroxide are formed in a 

 green leaf. 



A considerable number of experiments which have been carried out with 

 chlorophyll films point to the same conclusion, and the results of four which 

 may be regarded as typical are tabulated below : — 



Description of Experiment. 



(i) Sealed tube containing chlorophyll 



painted over a layer of gelatine, 

 made up with catalase solution, on 

 a glass plate. Air and caustic 

 potash present. 



(ii) The same as (i), but without catalase. 



(iii) The same as (i). 



(iv) The same as (i), but with a solution 



of carbon dioxide instead of caustic 

 potash. 



Remarks. 



Both these tubes were set up at the same 

 time, at 5 p.m., 22/4/09. At 6 p.m. (ii) showed 

 signs of bleaching. At 12 noon, 23/4/09, 

 (ii) was much bleached, whilst (i) was still 

 quite green and distorted with bubbles. 



Both set up at 11.30 p.m., 24/4/09. At 

 11 a.m., 25/4/09, (iv) was considerably 

 bleached and developed a very strong 

 coloration after 5 minutes' immersion in 

 Schiff's reagent, (iii) was still quite green, 

 and showed the faintest coloration only 

 after 15 minutes' immersion. 



* Berthelot and Gaudechon ('Comptes Rendus,' 1910, vol. 150, p. 1690) obtained 

 formaldehyde by the action of ultra-violet rays on carbon dioxide in the presence of a 

 reducing agent. Such a reducing agent may have been present in our experiments in 

 the form of hydrogen resulting from the decomposition of water (Kernbaum, ' Comptes 

 Rendus,' vol. 149, p. 273, and Tian, ibid., 1911, vol. 152, p. 1012). See also in this connec- 

 tion Stoklosa and Zdobnicj^, ' Anz. kais. Ak. Wiss. Wien,' 1910, No. 19, p. 319. 



