46 
SIR J. B. LAWES AND PROFESSOR J. H. GILBERT ON THE 
correlative with the amounts of either. Further, calculations showed that the 
amounts far exceeded those which could be due to ammonia in the air, or to ammonia 
and nitric acid in the rain ; whilst the gains in the closed flasks showed that they 
could not be due to combined nitrogen from the air or rain. 
The author considers the results establish the fact that there is gain of nitrogen 
quite independently of any absorption of combined nitrogen. 
From the evidence so far it might be concluded that the gains in the meadow and 
on the tower were due to electrical action; but the fifth series of experiments, in 
which the soils were sterilised by heat, and then left in the balloons from July 10 to 
October 6, 1885, indicate another influence. In the case of each of the four soils so 
sterilised, and afterwards exposed, there was, instead of any gain, a slight loss of 
nitrogen, which was attributed to the heating at the commencement. The cause of 
the fixation of nitrogen had at the same time been destroyed; nor did the soils 
recover the power of fixing nitrogen, either by exposure to the am of the chamber, or 
when a small quantity of the unsterilised soil was added. 
It was concluded that there was a fixation of free nitrogen due to living organisms. 
It was shown that the action was not manifested during the winter, that it was the 
most effective during the periods of active vegetation, and that it was exercised in 
closed vessels as well as in the fi’ee air. 
M. Berthelot estimated that the gains corresponded to gains of nitrogen per 
hectare of 20 kilog. by sand No. 1, of 16 and 25 kilog. by sand No. 2, and of 
32 kilog. by the kaolin No. 3. These estimates are, however, said to be much too 
low, as they are on the assumption of only 0'08 or OTO m. depth of soil, whilst the 
action extends much deeper. He compares these amounts with 17 kilog. the amount 
of combined nitrogen in the rain, &c., at Montsouris in 3 883 ; and with 8’ kilog., 
the amount formerly estimated as annually so coming down at Rothamsted ; which, 
however, more recently we have estimated at less than this. On the other hand, 
taking the amount of nitrogen removed in a crop of hay at from 50 to 60 kilog. per 
hectare (= 45 to 54 lbs. per acre), he estimates that the loss to the soil will be from 
40 to 50 kilog, per hectare (=36 to 45 lbs. per acre). Hence, if it were not for com¬ 
pensation by fixation of free nitrogen, the soil would gradually become exhausted. 
He considers that the results bring to view not only one of the methods by wdrich 
fertility is maintained, but that they also show how argillaceous soils, which are 
almost sterile when first brought into contact with the air, come to yield more and 
more flourishing crops, and in time become vegetable moulds. 
Quite recently, March, 1887 (‘ Compt. Bend.,’ vol. 104, pp. 625 et seq.), M. Berthelot 
has published the results of experiments on the fixation of free nitrogen by vegetable 
mould supporting vegetation. The experiments were commenced in May, and con¬ 
cluded in November, 1886. He determined the nitrogen by the soda-lime method, 
and also as nitric acid, in the soil before and after the growth; also in the initial 
plants [Amaranthus pyramidalis), and in the final products. He also determined 
