II.] 



CARBON DRA WN FROM THE AIR 



preliminary experiments we did not use ordinary soil, 

 because it contains a certain proportion of compounds 

 of carbon, out of which the plant's increase might 

 possibly have been made, though we can now show 

 that such is not the case. Suppose, instead of a 

 laboratory experiment, we deal with a crop in the open 

 and make out a sort of balance-sheet ; there is so much 

 carbon in the soil at starting, and so much in the seed 

 and the manure, against which we can set off the 

 amount of carbon in the crop at the finish together 

 with that in the soil after harvest. Such was the 

 balance-sheet that Boussingault drew up for the first 

 field experiments that were ever carried out, and he 

 showed that, despite the great amount of carbon 

 removed in a crop, the soil contained no less of this 

 substance at harvest than at seed time. 



Hence followed the inevitable conclusion that since 

 the carbon had not come from the soil, and could not 

 have been derived from the rain (water contains 

 hydrogen and oxygen only), it must have been taken 

 from the air. We can take an example of this kind of 

 balance-sheet from the Rothamsted wheat field, choosing 

 a plot manured only with inorganic salts containing no 

 carbon (superphosphate, sulphate and chloride of 

 ammonia, sulphate of potash, eta), the soil of which 

 had been analysed in 1881 and again in 1893. 



Table III.— Carbon in Soil and Crops, Rothamsted Wheat. 

 (Plot 7.) 



