24:0 HOW CPvOPS FEED. ' ' 



tested it for ammonia. In but two instances did tliey find 

 sufficient to weigh. In all cases, however, they were able 

 to detect it, though it was present in very minute quanti- 

 ty. The two experiments in which they were able to 

 weigh the ammonia were made in a light, sandy soil from 

 which potatoes had been lately harvested. On the 2d of' 

 September the field was manured with stable dung ; oni 

 the 4th the first experiment was made, the air being taken, 

 it must be inferred from the account given, at a depth of 

 14 inches. In a million parts of air by weight were found 

 32 parts of ammonia. Five days subsequently, after rainy 

 weather, the air collected at the same place contained but 

 13 parts in a million. 



b. Ammonia phi/slcall^ condensed in the Soil. — Many 

 porous bodies condense a large quantity of ammonia gas. 

 Charcoal, which has an extreme porosity, serves to illus- 

 trate this fact. De Saussure found that box-wood char- 

 coal, freshly ignited, absorbed 98 times its volume of 

 ammonia gas. Similar results have been obtained by St en- 

 house, Angus Smith, and others (p. 1G6). The soil cannot, 

 however, ordinarily contain more than a minute quantity 

 of physically absorbed ammonia. The reasons are, first, a 

 porous body saturated with ammonia loses the greater share 

 of this substance when other gases come in contact with 

 it. It is only possible to condense in charcoal 98 times its 

 volume of ammonia, by c(Joling the hot charcoal in mer- 

 cury which does not penetrate it, or in a vacuum, and then 

 bringing it directly into tlie ^:)r<>'e ammonia gas. The 

 charcoal thus saturated with ammonia loses the latter rap- 

 idly on exposure to the air, and Stenhouse has foimd by 

 actual trial that charcoal exposed to ammonia and after- 

 wards to air retains but minute traces of the former. 

 Secondly, the soil when adapted for vegetable groAvth is 

 moist or wet. The Avater of t!; > soil which covers the 

 particles of carlh, ra'lur tliaii (lie particles themselves, 

 must contai.i any absorbed auuuonia. Thirdly, there are 



