admitted that the results so far brought forward are abundantly confirmatory of those obtained by Hellriegel; and that the fact of the fixation of free nitrogen in the growth of Leguminosæ under the influence of microbe seeding of the soil and of the resulting nodule formation on the roots may be considered as fully established."
The results obtained by the inoculation of the prepared quartz sand with the microbes of a fertile soil, or of one in which lupines were growing, as shown in the increased growth of the plants in pots 2, 3, 10, 11, 18, and 19, when compared with those in pots 1, 9, and 17, which were not inoculated, are striking; but a comparison of the plants in the inoculated pots with those in pots 4 and 12 in a garden soil, and pot 20 in a "lupine soil," furnish still more significant indications of the futility of purely chemical considerations in discussing the nutritive processes of plants and their relations to the soil. The peas and vetches in a rich garden soil flowered and seeded, but the plants were not as large, and the root-tubercles were not as numerous, as in the sterile quartz sand inoculated with microbes from a fertile soil; and the lupines made a better growth in the inoculated quartz sand than in soil from a lupine field.
The biological factors concerned in the elaboration of plant food seem to be quite as important as the chemical elements provided in the soil itself; and a revision of the accepted theories of plant growth, and the relations of soils to their processes of nutrition, is evidently needed from this standpoint.
It should be remarked, however, that the root-tubercles produced by microbes are not confined to the Leguminosæ, as they have, in fact, been observed in several natural orders of plants. Moreover, there are indications that several varieties or species of symbiont microbes are concerned in the production of tubercles on the roots of leguminous plants, and it is probable that each species has its own favored form.
Hellriegel failed to grow lupines in a nitrogen-free soil inoculated with a fertile soil-extract; but, when the inoculation was made with an extract of a sandy soil in which lupines were growing, a luxuriant growth was obtained.
In the Rothamsted experiments on land where red clover had been grown repeatedly, and its yield of nitrogen was reduced to but 22 pounds per acre, vetches, on an average for three years, obtained 120 pounds of nitrogen per acre; lucern yielded as high as 340 pounds, and made an average for six years of 150 pounds of nitrogen per acre; and Bokhara clover yielded crops of 130 and 145 pounds of nitrogen per acre. On land where beans had been grown almost continuously for thirty-two years, and had "practically failed "to grow, their yield of nitrogen per acre hav-
