148 



The first experiments of Berthelot date from 1885. Tlieir object 

 was the fixation of nitrogen by denuded soils, leaving out, conse- 

 quently, all idea of vegetation. The soils used for the purpose were 

 chosen from among the poorest in nitrogen. They were sandy clays 

 taken from INIeudon or from Sevres, below the level of the quarries, 

 or, again, porcelain earths, crude kaolins not yet crushed in the mills. 



These soils, four in number, were submitted to five series of ex- 

 periments. They were left to themselves in glazed pots, either 

 within a Avell-closed room or in the open air in a meadow, either 

 without shelter or under a little glass roof, merely to protect them 

 from vertical rains, or on the top of a tower 29 meters above the 

 ground and without anv shelter, or finally, in corked flasks, so as to 

 exclude all possibility of absorption of ammoniacal or nitric vapors. 

 In the fifth series of experiments the same soils had first been ex- 

 posed to a temperature of 100°, so as to destroy from the first all the 

 organic germs that they might contain. The quantity of nitrogen, de- 

 termined with great precision in each of the samples at the very 

 beginning of the experiment, was again analyzed after tAvo months, 

 and again after renuiining five months under the conditions indi- 

 cated above, allowance being made for exterior additions attribut- 

 able to air and to the rains when the pots were not sheltered. 



The results obtained did not leave the slightest doubt. In every 

 case in which the earth had been left in its normal state it had be- 

 come enriched, and sometimes to a very great extent more than 

 doubling the quantity of the initial nitrogen ; when, on the contrary, 

 the soil had been sterilized by heat, it became constantly more 

 impoverished. In a word, then, poor clayey soils are able to absorb 

 atmospheric nitrogen directly. This absorption is not accompanied 

 by any increase in the previous proportions of ammonia or of nitric 

 acid ; it is, then, due to the formation of com]ilex organic substances. 

 Finally, it is the work of a micro-organism, since it ceases to be pro- 

 duced as soon as the soil has been sterilized. 



To what sum per hectare does such a fertilization correspond? 

 Berthelot estimates at 20 or 30 kilograms for a thickness of one 

 decimeter of soil. Hence for a thickness of 0.35 meter it would 

 suffice to compensate for the losses inherent to drainage and cultiva- 

 tion ; but before going further it is well to remark that the experi- 

 ments which we have just described relate to particidarly poor soils, 

 which are therefore of a nature to enrich themselves. In truly 

 arable soils, averaging from 1 to 2 grams of nitrogen per kilogram, 

 Berthelot has also observed a perceptible fixing of niti'ogen, which, 

 hoAvever, is relatively less than in sandy clays, and it is probable 

 that this phenomencm Avould cease to be apparent after a certain 

 limit, Avhich, doubtless, is not very high. 



The conditions which, according to Berthelot, apear the most 

 favorable to the fixing of nitrogen by the naked soil are : 



1. The presence of a quantity of water comprised betAveen 3 and 15 

 per cent of total saturation ; 



2. A sufficient porosity to assure the free jienetration of air 

 throughout the wliole mass of earth: 



3. A temperature of betAveen 10° and 40°^ C. 



Th?se conditions define the microbe AA'hich secretes or fixes the 

 nitrogen as an aerobic organism (i. e., one that feeds on the atmos- 

 phere or is aerobiotic) . 



