236 METABOLISM 



(PLINY, Historia Naturalis, vol. 8 ; cited by JACOBITZ, 1901), but were for the 

 first time clearly comprehended by ScHULTZ-Lupixz (1881), who obtained fifteen 

 crops of lupins one after the other from sandy soils on his estate merely by 

 mineral manuring without any nitrogen, and without noticing any diminution 

 in their yield. He also noticed that, after a crop of lupins, the cereal harvest 

 was doubled or tripled in amount. SCHULTZ did not remain content with 

 these results ; he made an analytical estimate of the nitrogen in the soil, with 

 the following result : 



1. Land which had neither been manured or cultivated for fifteen years and 

 which had been used as sheep pasture contained 0-027 per cent, of nitrogen in the 

 arable layers to 6" deep, and 0-21 per cent, in the subsoil from 6" to 24" deep. 



2. The same soil after being cultivated for fifteen years with lupins and 

 manured with minerals only, gave to 8" deep 0-087 per cent., and from 8" to 

 24" deep, 0-025 P er cent. 



It will be noticed that a very apparent gain in nitrogen had been effected 

 in the upper soil layers, and this was confirmed later by FRANK (1888) when 

 he examined the same fields after twenty years of lupin culture. 



Both husbandman and agricultural chemist may therefore conclude 

 from the statements of SCHULTZ that the Leguminosae, and especially lupins, 

 must have the power of combining atmospheric nitrogen. Botanists, on the 

 other hand, may refer to an experiment of BOUSSINGAULT'S, of the accuracy of 

 which there can be no doubt, which shows that peas and lupins, which germinate 

 in soils containing no nitrogen, and which have access to no other source of 

 nitrogen than that present in the air, exhibit neither gain or loss in their nitro- 

 genous contents after a long continued experiment. Exact though this ex- 

 periment may be, it in no way demonstrates the inability of Leguminosae to 

 combine nitrogen in general, but only under the conditions of this experiment. 

 BOUSSINGAULT'S experiments are in reality not antagonistic to the classical 

 researches of HELLRIEGEL and WILFARTH, although these latter authors have 

 advanced definite proof of the power of the Leguminosae to combine free nitrogen. 



HELLRIEGEL and WILFARTH (1888) used as a culture medium very pure 

 quartz sand, which was completely sterile, save for the addition of minerals. 

 The Leguminosae experimented with were compared most carefully with 

 cereals (oats and barley), which are known to be incapable of growth in the 

 absence of nitrates and other sources of nitrogen present in soils, and which also 

 have been specially shown to be unable to make any use of the nitrogen of the air. 



Certain Leguminosae (Serradella, peas and lupins) were planted in this sand, 

 after it had been freed from all micro-organisms by heat, i. e. sterilized, and kept 

 free from them, and an experiment was carried out in precisely the same way 

 as with cereals, when growth took place only on the addition of nitrates. This 

 agrees with the behaviour of the Leguminosae in BOUSSINGAULT'S experi- 

 ment referred to above. An important difference was noticed, however, as soon 

 as a small quantity of an extract of arable soil had been added to the sterile 

 soil free from nitrogen ; then the crop showed a remarkable gain in nitrogen, 

 which could only have arisen from the employment of atmospheric air. A 

 numerical illustration will make this clearer (HELLRIEGEL, 1888, p. 145): 



Without addition of soil extract. With addition of soil extract. 



Dry weight. Gain in N. Dry weight. Gain in N. 



g g g g 



Serradella 0-092 0-022 16-864 +0-326 



Lupins 0-919 0-049 44-7*8 +1-077 



Peas 0-779 0-025 17-616 +0-449 



If the experiments are carried out with oats, instead of with Leguminosae, 

 the addition of the soil extract produces no result. 



The effect of the soil extract cannot be due to the amount of nutrient 



