250 On the Rain and Drainacjc - Waters at Rothamsted. 
exists in the oxidation of ammonia by ozone and peroxide of 
hydrogen. As the latter substance is evolved when turpentine, 
and possibly other bodies, are oxidised in the air, the neighbour- 
hood of a pine-forest should be favourable to the formation of 
nitric acid in the atmosphere. 
The sulphates of the atmosphere are, according to Angus 
Smith, chiefly derived from the oxidation of the sulphur com- 
pounds evolved during tlie decay of animal matter. In towns 
the sulphates are much increased by the oxidation of the 
sulphurous acid contained in coal-smoke. 
Chlorides are principally furnished by the sea, fine spray of 
salt-water being carried long distances by high winds. To a 
small extent chlorides may also be furnished by the combustion 
of fuel. 
The quantity of ammonia and nitric acid contained in rain- 
water is a question of considerable interest to the agriculturist. 
As ammonia and nitric acid form the chief if not the only 
sources of the nitrogen of plants, and manures containing them 
are purchased by the farmer only at considerable expense, it 
becomes of great interest to ascertain the amount naturally 
supplied to our fields by rain. 
De Saussure, Brandes, and Liebig had called attention to the 
existence of ammonia and nitric acid in rain-water before the 
commencement of the Rothamsted experiments. Some initiative 
determinations of the ammonia in rain were made at Rothamsted 
as early as 1846. Barral, in 1851, made determinations of both 
the nitric acid and the ammonia in the rain which fell in 
Paris during several consecutive months; and in 1852 Boussin- 
gault determined the ammonia in the rain collected at Lieb- 
frauenberg, in Alsace. The opportunity afforded by the erection 
of the large gauge at Rothamsted of collecting considerable 
quantities of rain-water at a distance from any large town* 
was at once turned to account to follow up the subject. A fixed 
proportion of the day's rainfall was regularly set aside, and 
these quantities being mixed at the end of the month, a sample 
was obtained accurately representing the month's rainfall. 
Determinations of ammonia were made in these monthly 
samples for 15 months during 1853 and 1854. The method 
employed consisted in fractional distillation of the water, and 
determination of the ammonia in the distillate with a very 
dilute standard acid and alkali. The results of these analyses 
were communicated to the British Association for the Advance- 
* Kothamsted lies about 25 miles north-west of London, and about four miles 
north of St. Albans. The village of Harpeuden, with a very scattered popula- 
tion of nearly 3000, lies mostly to the north-east and east of the rain-gauge, a 
majority of the houses being probably about three-quarters of a mile distant. 
