LIBEKATION OF NITROGEN. 307 



The losses occasioned by the volatilisation of ammonia produced 

 in this manner may be very considerable, but will not be further 

 considered here. We will now turn to the liberation of uncom- 

 bined nitrogen. 



The first researches on this point were undertaken by JULES 

 REISET (I.) in 1854 and 1855. He asserted that free nitrogen is 

 always evolved during the putrefaction of manure, whilst G. 

 HUFNER (I.) arrived at the contrary opinion, being unable to dis- 

 cover any liberation of free nitrogen when atmospheric air or pure 

 oxygen was led through the putrefying substances. The same 

 result was obtained by ALEXANDER EHRENBERG (I.), 0. KELLNER 

 and T. YOSHII (I.), and BR. TACKE (I.); and this view was also 

 held by H. IMMENDORFF (II.) in 1893. 



Although these discoveries may justify the conclusion that no 

 free nitrogen is disengaged during the putrefaction of albuminoids, 

 it must not, however, be assumed that the same also applies to 

 the decomposition of manures in general under natural conditions ; 

 since, under these circumstances, very considerable quantities of 

 this element can be liberated and become lost to the soil. This 

 result is, however, due to the reduction of nitric salts, and not to 

 the putrefaction of albuminoids. 



This de-nitrification in arable soil was first noticed by GOPPELS- 

 RODER (I.) in 1862, and was long regarded as a purely chemical 

 process. The first reference to the agency of bacteria in this de- 

 composition was made by E. MEUSEL (I.) in 1875, and the earliest 

 pure cultures of such organisms were obtained by U. GAYON and 

 G. DUPETIT (II.) in 1882. In succeeding years a large number of 

 species, all capable of reducing nitrates, was made known ; e.g. by 

 W. HERAEUS (I.) in 1886. Two years later P. FRANKLAND (II.) 

 was able to associate with the group in question 17 out of 32 

 species, and R. WARINGTON (I.) 16 out of 25 species examined, 

 among them being Bacillus ramosus, the so-called "Wurzel- 

 bacillus." All these reduce nitrates into nitrites, but these two 

 chemists do not say whether the latter substances in turn may 

 be still further reduced by the bacteria. For this reason we must 

 revert to the labours of GAYON and DUPETIT (III.), who made 

 pure cultures of two bacterial species, named Bacillus denitrificans 

 a and /3, which exhibit a noteworthy difference in their behaviour 

 towards nitrates. Species a is the more energetic, decomposing as 

 much nitrate as is presented to its action, and reducing the same 

 to nitric oxide and free nitrogen. The /3 species, on the other 

 hand, forms nitrites, and ceases to act before the whole of the 

 nitrate is destroyed, free nitrogen being the only gaseous fer- 

 mentation product. Quite distinct from these two species is the 

 Bacillus denitrificans, isolated from arable soil by E. GILTAY and 

 J. H. ABERSON (I.), which reduces the nitrates to free nitrogen in 

 an almost quantitative degree. When grown on nutrient gelatin 

 the rods measure 0.5 ft in breadth and 1.5-3 /* i n length, but in 



