168 HOW CROPS FEED, 
3. Oxygen is often nearly or quite wanting, as in char- 
coal, oxide of tron, alumina, river silt, and whiting. 
4, Carbonic acid, though sometimes wanting entirely, 
is usually abundant in the absorbed gases. 
5. In the pores of charcoal and of soils containing de- 
caying organic matters, carbonic acid is often partially re- 
placed by carbonic oxide. The experiments, however, do 
not furnish proof that this substance is not formed under 
the influence of the high temperature employed (284° F.) 
in expelling the gases, rather than by incomplete oxidation 
of organic matters at ordinary temperatures. 
6, A substance, when moist, absorvs less gas than when 
dry. Inaccordance with this observation, De Saussure no- 
ticed that dry charcoal saturated with various gases evoly- 
ed a good share of them when moistened with water. 
Ground (and burnt?) coffee, as Babinet has lately stated, 
evolves so much gas when drenched with water as to burst 
a bottle in which it is confined. 
The extremely variable figures obtained by Blumtritt 
when operating with the same substance (the figures given 
in the table are averages of two or tliree usually discordant 
results), result from the general fact that the proportion 
in which a number of gases are present in a mixture, in- 
fluences the proportion of the individual gases absorbed. 
Thus while charcoal or soil will absorb a large amount of 
ammonia from the pure gas, it wiil take up but traces of 
this substance from the atmosphere of which ammonia is 
but an infinitesimal ingredient. 
So, too, charcoal or soil saturated with ammonia by ex- 
posure to the unmixed gas, loses nearly all of it by stand- 
ing in the air for some time. This is due to the fact that 
gases attract each other, and the composition of the gas 
condense in a porous body varies perpetually with the 
variations of composition in the surrounding atmosphere. 
It is especially the water-gas (vapor of water) which is 
a fluctuating ingredient of the atmosphere, and one which 
