E. J. Russell and E. H. Richards 



325 



amount of chlorine is not accidental: 

 Cirencester data: 



it is equally manifest in the 



There appears, however, to have been some other factor concerned 

 besides the increase in rainfall, as the concentration of chlorine in the 

 rainwater has also increased: formerly the rain contained about 2-2^ 

 parts of chlorine per million, now the amount varies from 2|-3 parts 

 per million. 



The increase in amount of nitric nitrogen is shown in Tables 5 and 9 

 and in Fig. 4 : it is not as steady as that of chlorine and is intensified by 

 a few abnormally high values especially in two dry months February 

 and July, 1913, in the wet October, 1913, in the dry May, 1914, and 

 in the wet August, 1916. It is, of com-se, impossible now to check these 

 figures: we have gone carefully through the laboratory notes and can 

 find nothing to indicate that they are unreliable. Even apart from these 

 exceptional cases, however, there is a clear upward trend in the nitrate 

 figures. Examination of the curves in detail shows that the tendency 

 to increase is spread over the whole year and is not confined to any 

 one season: formerly there was about 0-12 to 0-19 parts of nitric nitrogen 

 per million of rain, 'now the figures are 0-2 to 0-3 parts. 



Although the rise in nitric nitrogen is similar to that of chlorine 

 the phenomena are not necessarily related. We have no grounds for 

 supposing that sea spray contains nitrates in sufficient quantity to 

 produce the observed effects. On the other hand if some of the chlorine 

 comes from coal a relationship might be expected, as part of the nitrate 

 may come from the same source. 



The ammonia, on the other hand, has decreased in concentration: 

 formerly there were usually 0-4 to 0-5 parts per million, now there are 

 more usually 0-2 to 0-3 parts. This is shown in Tables 4 and 9 and in 

 Fig. 4. 



' Data collected from papers in Agricultural Stttdents' Gazelle and Trans. Chem. Soc, 

 1887, 51, 92 and 1900, 77, 1271. 



Journ. of Agrio. Sci. ix 22 



