“Recherches sur l’assimilation de l’azote libre de l’atmosphère par les microbes.” Arch. des Sci. Biolog. St. Petersburg, 1895, iii, 297-352. For further details and fuller bibliography, see E. J. Russell, “Soil Conditions and Plant Growth,” Longmans, Green & Co. CHAPTER II. SOIL BACTERIA. _A._ OCCURRENCE AND METHODS OF STUDY. To understand the development of our knowledge of soil bacteria, it must be remembered that bacteriology is under the disadvantage that it started as an applied science. Although bacteria were first seen by Leeuwenhoeck about the middle of the seventeenth century, and some of their forms described by microscopists of the eighteenth and early nineteenth centuries, it was only with the work of Pasteur on fermentation, and of Duvaine, Pasteur, and their contemporaries on disease bacteria, that bacteriology may be said to have started. From the outset, therefore, attention has been directed mainly to the bacteria in their specialised relationship to disease or to fermentation and similar processes. As a result, little research was done on the pure biology of the bacteria, so that even now many of the most fundamental and elementary problems concerning them are quite unsolved. In their work on fermentations and disease bacteria, the earlier workers were assisted by the fact that under both sets of conditions the causative bacteria exist, as a rule, either in practically pure culture, or else in preponderating numbers. The study and elucidation of such a mixed micro-population as exists in the soil, became possible only when methods had been devised for isolating the different kinds of bacteria, and thus studying them apart from each other. It was the development of the gelatine plate method of isolating pure cultures by Koch[36] in 1881 that made the study of the soil bacteria practicable. The plating method opened up two lines of research. In the first place, it provided a simple means of isolating organisms from the mixed population of the soil, and thus enabled a qualitative study to be made of each organism in pure culture, and, in the second place, from it was developed a counting technique for estimating differences in bacterial numbers between samples of soil, from which has sprung much of our knowledge of the quantitative side.