[shutt] NITROGEN COMPOUNDS IN RAIN AND SNOW 87 



fires have occurred more or less every year during the investigation, 

 usuallv in the autumn. Naturally they are chiefly in times of great 

 heat and dryness. After a week or so without rain, when such fires 

 prevail, the subsequent rains will be very rich in ammonia and may 

 continue so for a period of several days or even more, unless the pre- 

 cipitation is very heavy. 



Another factor is frequency of precipitation, and this is more 

 particularly true during the summer months. A scanty rainfall 

 after a period of hot, dry weather is invariably rich, and on the other 

 hand the later collections after several days of showery weather show 

 a rapidly decreasing nitrogen content. 



While the direction of the prevailing winds, during the falling 

 of rain and snow, as for instance toward or from the city, does not 

 apparently markedly influence the nitrogen content, its velocity and 

 violence may, and frequently does, very appreciably affect the char- 

 acter of the rain in this respect. The rain during thunder storms is 

 invariably rich in nitrogen, and this we have attributed chiefly to the 

 presence of an increased amount of dust in the air, though to some 

 extent the nitrates may be increased by the electric discharges (light- 

 ning flashes) of the storm. It has been repeatedly noted that the 

 rain falling during or immediately following cyclonic storms of great 

 severity and which "filled" the atmosphere with dust particles, had an 

 exceedingly high nitrogen content, more particularly present as free 

 and albuminoid ammonia. 



In this work every precipitation of rain and snow that would 

 yield a sufficiency for analysis from the catchment area has been 

 chemically examined. Early in the investigation it was found that 

 the results from composite samples, representing the rain of a week 

 or a month, were not reliable, and hence that plan, though economical 

 of labour, was not adopted. 



The collection of the samples of rain was made on a leaden tray 

 or basin, placed about twenty feet from the ground which for some 

 distance around is lawn and shrubbery. It is approximately 60 x 30 

 inches. The water as it falls is conducted from the bottom of the basin 

 by means of a glass tube into a glass jar, from which the sample for 

 analysis is taken at the end of each rainfall. 



