62 HOW CROPS FEED. 
Boussingault found in the first tenth of a slow-falling 
rain (24th Sept., 1853) 66 parts of ammonia, in the last 
three-tenths but 13 parts, to 10 million of water. In dew 
he found 40 to 62; in fog, 25 to 72; and in one extraordi- 
nary instance 497 parts in ten million, 
Boussingault found that the average proportion of am- 
monia in the atmospheric waters (dew and fogs included) 
which he was able to collect at Liebfrauenberg (near Stras- 
burg, France) from the 26th of May to the 8th of Nov. 
18538, was 6 parts in 10 million (Agronomie, etc., T. II, 
238). Knop found in the rains, snow, and hail, that fell at 
Moeckern, near Leipzig, from April 18th to Jan. 15th, 
1860, an average of 14 parts in 10 million. ( Versuchs- 
Radeon: Vol. 3, p. 120.) 
Pincus ne Roéllig obtained from the atiitodplient wa- 
ters collected at Inster burg, North Prussia, during the 12 
months ending with March, 1865, in 26 analyses, an average 
of 7 parts of ammonia in 10 million of water. The average 
for the next following 12 months was 9 parts in 10 million. 
Bretschneider found in the atmospheric waters collected 
by him at Ida-Marienhiitte, in Silesia, from April, 1865, to 
April, 1866, as the average of 9 estimations, 30 parts of 
ammonia in 10 million of water. In the next year the 
quantity was 23 parts in 10 million. 
In 10 million parts of rain-water, etc., collected at the 
following places in Prussia, were contained of ammonia— 
at Regenwalde, in 1865, 24; in 1867, 28; at Dahme, in 
1865, 17; at Kuschen, in 1865, 54; and in 1866, 73 parts. 
paix oe d. Deanilectest ious 1867.) The monthly 
averages fluctuated without regularity, but mostly within 
narrow limits. Occasionally they fell to 2 or 3 parts, once 
to nothing, and rose to 35 or 40, and once to 144 parts in 
10 million. 
Quantity of Ammonia per Acre Brought Down by Rain, 
etce.—In 1855 and ’56, Messrs. Lawes & Gilbert, at Roth- 
amstead, England, collected on a large rain-gauge having 
