78 
SIR J. B. LAWES AND PROFESSOR J. H. GILBERT ON THE 
other hand, the Papilionacese should give marked increase with mineral manure, and 
but little with nitrogenous manure. Such is certainly the result in ordinary soils 
containing nitrogen; but if we are to assume that in his experiments the sand was 
not the source of the nitrogen, both the amounts of dry produce, and those of 
nitrogen, in the difterent crops, as shown in the tables which have been given, are 
such as seem to exclude any other explanation than that the air had contributed 
nitrogen in some way. 
At the same time, the conditions of experiments conducted in a not absolutely 
nitrogen-free soil, and with free exposure to the weather, and so subject to accidental 
sources of more or less combined nitrogen which such conditions necessarily imply, 
however appropriate for obtaining initiative results and general indications, seem 
scarcely suitable for the settlement of so delicate a question as that of the source of 
the nitrogen of vegetation. On this point it may be remarked that, according to the 
data given, the unwashed sand put into each cement vessel would contain about 
9 grams of combined nitrogen, whilst, as shown in the table at page 74, the largest 
crops of the lupins and red clover, with mineral but without nitrogenous manure, only 
contained about that amount of nitrogen, and the beans only about half as much. It 
is true that this was the result in the third year, 1886, after somewhat similar 
amounts had already been grown for two years. 
Again, so far as the sand did contain nitrogen, the great difference of result with 
the Graminese and the Leguminosse, under the influence of mineral manure without 
nitrogen, is not absolutely conclusive evidence that the Leguminosse had acquired 
nitrogen from some other source than the soil; for there can be little doubt that 
Leguminosse do utilise nitrogen existing in the soil in a condition in which it is not 
available to the Graminese. The results obtained in the washed sand, must, Lowever, 
be admitted to have much greater significance. 
As to the explanation of the results, Wolff is disposed to attribute the gains of 
nitrogen to the absorption of combined nitrogen from the air by the soils, and to the 
fixation of free nitrogen within the soil under the influence of porous and alkaline 
bodies, as supposed by Cloez and Frank, rather than to the fixation of free nitrogen 
either under the influence of micro-organisms, or directly by the plants themselves. 
In fact, neither were the conditions of his experiment wuth the clover in burnt soil, 
nor those of his later experiments in washed sand, such as to favour the supposition 
of the intervention of micro-organisms. On the other hand, he considers they were 
favourable for the absorption of combined nitrogen from the air, and for the supposed 
fixation of free nitrogen within the soil under the influence of porous and alkaline 
bodies. At the same time, he admits that it is not easy to explain why Graminese do 
not, equally with the Leguminosae, benefit by such absorption, and by such fixation. 
For our part we believe that a careful consideration of all that is involved in this 
undoubted fact, points to the exclusion of the supposition that the gain is either by 
the absorption of ammonia from the air, or by the fixation of free nitrogen within the 
