THE ASSIMILATION OF CARBON BY AUTOTROPHIC PLANTS. I 113 



that formaldehyde arises from carbon-dioxide directly, without going through 

 the intermediate product, carbon-monoxide. The direct origin is all the more 

 plausible if we start, not from carbon-dioxide (CO2), as we have been doing 

 hitherto, but from carbonic acid (Hg CO3). We are certainly entitled to do this, 

 since chemists tell us that carbon-dioxide exists in aqueous solution only in the 

 form of Hj CO3 ; we know, further, that carbon-dioxide never occurs in the 

 plant cells as a free gas, but always in the dissolved condition. The con- 

 struction of formaldehyde would thus be represented by the equation : — 



H2CO3 = CHjO + O2 ; 

 Carbonic acid = Aldehyde + Oxygen ; 



so that the hypothesis of Erlenmeyer (1877), which assumes formic acid as 

 an intermediate product would be in accord with this view (compare Czapek, 

 1902) : — 



H2CO3 = CH2O2 + O; CH2O2 = CH.jO + O. 



Carbonic acid = Formic acid + Oxygen ; Formic acid = Aldehyde + Oxygen. 



Many arguments have been advanced from the physiological standpoint 

 against the origin of formic acid and formaldehyde as intermediate products 

 in carbon-dioxide assimilation ; for example, the regular occurrence of these 

 substances in the assimilating leaf has been doubted, and further it has been 

 pointed out that formaldehyde is very poisonous to plants. More recently, 

 PoLLACCi (1900, 1902) has demonstrated with certainty the constant occurrence 

 of formaldehyde in green leaves, while Curtius and Reinke (1897) observed 

 the appearance of aldehyde in the green leaf in light and its disappearance in the 

 dark, and especially noted (Reinke, 1899) that it occurred not as formaldehyde, 

 but as a substance with the formula CgHjOj. [Pollacci also (Atti Istit. d. 

 Pavia, 1904) has very strongly upheld the aldehyde hypothesis in a paper not 

 long published, but a critical re examination of his work is still wanted (compare 

 Czapek, 1, 502-6).] In whatever way the question may be finally settled it would 

 appear to us that the deficiency in a certain substance is a much better proof 

 of its being an intermediate product in assimilation than its plentiful occurrence, 

 since such intermediate stages would in all probability be passed through rapidly. 

 It is important to remember in the discussion of this question that although 

 formaldehyde, even in very small quantities (i : 20,000), is intensely poisonous 

 to plants, its poisonous effects could be avoided by its rapid construction into 

 higher compounds. Further, it is extremely significant that, as yet, all attempts 

 to make the chloroplasts use formaldehyde in the manufacture of starch have 

 failed, and yet it must certainly take place if the hypothesis be correct. Accord- 

 ing to BoKORNY (1897), Spirogyra certainly can (although only in sunlight) 

 manufacture starch out of sodium formaldehyde sulphite, and also from methylal. 

 Since both these substances yield formaldehyde readily, Bokorny concluded 

 that the experiment furnishes evidence of the manufacture of starch from 

 formaldehyde. Even if his opinion were well grounded, still Bayer's hypo- 

 thesis would not be directly proved thereby, for one could hold with equal 

 justice that glycerine was the first product of assimilation (Pfeffer, Phys. I, 

 p. 341), since starch is manufactured from glycerine, and that, too, in the dark. 

 Experiments carried out by Treboux (1903) on Elodea show that this plant 

 can live perfectly well in a 0-0005 P^r cent, solution of formaldehyde, but that 

 it was quite unable to manufacture starch from it either in the dark or in the 

 light. 



The settlement of the question as to what is the first product of carbon- 

 assimilation is not of great importance in physiology, and we will there- 

 fore confine our attention in the following pages to the consideration of 

 the carbohydrates, which have been shown to be undoubtedly products of 

 assimilation ; and our knowledge of the phenomena taking place will be con- 



JOST I 



