[SHUTT-DORRANCE] NITROGEN COMPOUNDS IN RAIN AND SNOW 67 



off in the nitrogen compounds is noticed in the later precipitations. 

 In other words, there is a tendency towards the washing out of the 

 atmosphere by a more or less continued rainfall. However, in the 

 absence of unusual atmospheric disturbances, longer periods than 

 two or three days have no marked influence on the concentration. 

 This would point to a certain equilibrium as regards the nitrogen 

 compounds of the air being again established within a comparatively 

 short period. A sample from a short shower usually shows a higher 

 concentration than that from a long shower and in the case of a fall 

 more or less continuous for several days the concentration gradually 

 decreases. Necessarily the amount or inches of rainfall, independent 

 of whether it occurs in short or long showers, determines to a large 

 extent the amount of nitrogen furnished annually per acre, which is 

 the product of the precipitation in inches, the nitrogen content ex- 

 pressed in parts per million and 0-2266, the latter factor being cal- 

 culated from the weight of one inch of rain over one acre, viz., 113 tons, 

 600 pounds. 



Discussion of Results. 



Table I. This table gives for each year of the decade, the 

 precipitation in inches of rain, snow and total in inches of rain, the 

 average nitrogen content in the three combined forms expressed in 

 parts per million, and the total pounds of nitrogen furnished per 

 acre. 



Three years, those ending February 29, 1908, February 28, 1909 

 and February 29, 1916, present certain exceptional figures. The 

 first year's results (1908) both as regards nitrogen content and amount 

 of nitrogen per acre, are considerably lower than the average. This 

 may in some measure be accounted for by the fact that the rain of 

 that year, as will be seen from table III, shows the lowest nitrogen 

 concentration of the series. Regarding the very high free ammonia 

 content of the precipitation of the second year of the investigation, 

 mention has already been made of the peculiar atmospheric condi- 

 tions existing during the months of September and October 1908, the 

 prevailing bush fires filling the air with smoke. Ample proof was 

 obtained that this smoke affected the samples collected at that time 

 and the extremely high figure, 0-834 for nitrogen as free ammonia, is 

 undoubtedly to be attributed to that cause. The nitrogen as albu- 

 minoid ammonia and as nitrates and nitrites for the year ending 

 February 29, 1916, are both much higher than the average for the 

 decade. We have not been able to discover any cause which would 



