BY THE ACTION OF SUNLIGHT 73 



not artificially enriched with nitrogenous manure ; the experiments 

 given below, and others recorded in later chapters, show that 

 there is also an aerial uptake of nitrites by the green leaves. 



A source of much error and confusion in estimating the so-called 

 " active " oxygen of air, rain, or dew by different observers at 

 meteorological stations has been the use of test-papers, impregnated 

 with starch and iodides, which were moistened and exposed to air, 

 and indicated, by the rapidity of development of a blue colour, the 

 degree of " active " oxygen in air. This " active " oxygen was 

 assumed to be present mainly as ozone or hydrogen peroxide without 

 more proof, but this liberation of iodine from iodides is accomplished 

 quite as readily by nitrogen tri- or tetr-oxide as by ozone or hydrogen 

 peroxide. All the more recent researches 1 indicate that ozone and 

 hydrogen peroxide are absent fr,om the air at the earth's surface, and 

 the important purifying and bleaching reactions hitherto ascribed 

 to them must now be transferred to the intermediate oxides of 

 nitrogen. At high dilutions, such as are shown below to occur under 

 atmospheric conditions, the odour of ozone and that of the oxides 

 of nitrogen are indistinguishable. 



The present experiments show that air, rain, and dew invariably 

 contain a mixture of nitrites and nitrates, and that on keeping the 

 nitrites pass over into nitrates, but by insolation this process is 

 reversed, and nitrites are formed from nitrates. 



The test used was Ilosvay's modification of the diazo-reactions 

 discovered by Griess, yielding compounds deeply coloured even at 

 high dilutions ; one of the best of these reactions for the purpose is 

 that in which solutions of sulphanilic acid and a-naphthylamine in 

 acetic acid are added to the water suspected of containing nitrites. 

 This test is given only by nitrites, and not by either ozone or hydrogen 

 peroxide. 



Ilosvay, by the use of this test, showed that the well-known 

 reaction upon a paper impregnated with starch and potassium 

 iodide often used to show the supposed presence of ozone in the 

 atmosphere was really produced by nitrites, and demonstrated 

 that at the earth's surface both ozone and hydrogen peroxide were 

 normally absent. The same observer 2 found a strong reaction for 



1 Ilosvay, Ann. Chem. Pharm., vol. cxxiv., p. 1 ; vol. clxxxvi., p. 2. 



2 Hayhurst and Pring, Chem. Soc. Trans., vol. xcvii., Pt. I., p. 868 (1910); 

 Pring, Roy Soc. Proc., A, vol. xc., p. 204 (1910), and Strutt, Roy. Soc. Proc., 

 A, vol. xciv., p. 260 (1918). 



