330 Rain Falling at Rotliamsted 



July and August, and lowest during January, February, March and 

 April (Tables 1, 2, 3, and Figs. 1 and 2). 



The nitric nitrogen is on an average one-half the ammoniacal, viz., 

 1-33 lb. per acre per annum. The amounts fluctuated year by year 

 and month by month in the same way as the ammoniacal nitrogen and 

 the rainfall until 1910, since when there has been no simple relationship 

 (Tables 1, 2, 3, and Figs. 1 and 2). 



Reasons are adduced for supposing that the ammonia arises from 

 several sources. The sea, the soil, and city pollution may all contribute. 

 Neither the sea nor city pollution seems able to account for all the 

 phenomena : the soil is indicated as an important source by the fact that 

 the ammonia content is high during periods of high biochemical activity 

 in the soil, and low during periods of low biochemical activity. 



The close relationship between the amounts of ammoniacal and nitric 

 nitrogen suggests either a common origin or the production of nitric 

 compounds from ammonia. 



The average amount, of chlorine is 2-43 parts per million bringing 

 down 16 lb. per acre per annum (Table 6). The fluctuations closely 

 follow the rainfall both month by month and year by year, but the 

 general level is much higher during the months September to April than 

 during the summer months (Fig. 3). It seems probable that the chlorine 

 comes from the sea, but some may come from fuel. 



Since 1888, when the experiments began, to 1916, when they ter- 

 minated, there has been a rise in the amounts of nitric nitrogen and of 

 chlorine in the rain. (Tables 5, 7, 8 and 9; Figs. 3 and 4.) In the case 

 of chlorine a parallel series of determinations made at Cirencester over 

 the same period shows a similar rise (p. 325). There is no rise of ammonia 

 but on the contrary a tendency to drop (Tables 4 and 9; Fig. 4): the 

 sum of ammoniacal and nitric nitrogen shows little change over the 

 period. This seems to suggest that a former source of ammonia is now 

 turning out nitric acid : it is possible that modern gas burners and grates 

 tend to the formation of nitric oxides rather than of ammonia. 



Rain contains on an average 10 parts of dissolved oxygen per million, 

 the amount being higher in winter than in summer: 66-4 lb. per acre 

 per annum was brought down during the two years over which the 

 determinations extended (p. 328). 



The marked difference in composition between summer and winter 

 rainfall suggests that these may differ in their origin. The winter rain 

 resembles Atlantic rain in its high chlorine and low ammonia and 

 nitrate content: the summer rain is characterised by low chlorine but 



