450 MR. J. B. LAWES, DR. GILBERT, AND DR. PUGH ON 
somewhat contracted. Finally, access of air from without was excluded by bandages 
of caoutchouc, which were so secured as to render the whole apparatus air-tight. 
In such an apparatus, BoussincauLt made five separate experiments with White 
Lupins. In all he sowed thirteen seeds, which were estimated to contain, together, 
0:2710 gramme of Nitrogen. The experiments extended over periods varying from six 
to eight weeks. In one instance, burnt brick instead of pumice-stone was used as the 
soil; and in this case, as well as in one where pumice was used, bone-phosphate as well 
as ashes was added as manure. The dry matter of the produce was about three times. 
as much as was contained in the thirteen seeds sown. ‘The Nitrogen in the total pro- 
ducts of the five experiments amounted to 0:2669 gramme. There was, therefore, a 
loss, in the five experiments taken together, of about 4 milligrammes of Nitrogen. In 
two of the cases there was a slight gain of Nitrogen, but in neither instance did it 
amount to 1 milligramme. 
In a similar apparatus, two experiments were made with Dwarf Haricots, a single seed 
only being sown in each case. One of the experiments extended over two months, and 
the other over two and a half months. In both instances flowers were formed, and in 
one of them seed. The dry matter of the produce was three to four times as much as 
that of the seed sown. Taking the two experiments together, the Nitrogen contained 
in the seed was estimated at 0-0652 gramme; and that found in the products amounted - 
to 0:0637 gramme. There was a loss, therefore, of 1} milligramme of Nitrogen. 
There was, then, in this third series of experiments with Leguminous plants, again 
rather a loss than a gain of Nitrogen,—the supplies of it in this case being confined to 
the combined Nitrogen contained in the seeds sown, and to the free or uncombined 
Nitrogen in the fixed and limited volume of air within the apparatus. 
Still in the same apparatus, BoussincauLt sowed Garden Cress. Thirteen seeds were 
sown, all of which germinated, but three. plants only survived. The growth of these 
extended over three and a half months; and flowers and seed were produced. The 
Nitrogen in the products amounted to precisely as much as was estimated to be con- 
tained in the thirteen seeds sown. 
The last experiment in this closed globular apparatus was as follows: Two White 
Lupins were sown to grow; and eight others were applied as manure, after treatment with 
boiling water to destroy their powers of germination. The experiment continued for a 
period of between four and five months. The dry matter of the produce was nearly twice 
as much as would be contained in the ten seeds involved in the experiment. The whole 
ten seeds were estimated to contain 0:1827 gramme of Nitrogen; whilst the total pro- 
ducts yielded only 0:1697 gramme. The loss of Nitrogen was here, therefore, 13 milli- 
gramines; or about one-fourteenth of the whole amount involved in the experiment. 
BoussinGavtt considered that the loss was probably due to free Nitrogen being given off 
in the process of decomposition of the organic matter employed as manure. 
In order to ascertain whether the limitation of growth in the foregoing experiments. 
was due to the limitation in the amount of air, or to a deficiency of available Nitrogen 
in the matters used as soil, Boussineavtt sowed Cress in a good soil, placed the vessel 
