The soil is not, however, all made up of disintegrated rocks.  It contains, besides, various ingredients which combine to make it fertile.  Among these are various sulphates which form important parts of plant foods.  These sulphates appear to be formed, in part, at least, by bacterial agency.  The decomposition of proteids gives rise, among other things, to hydrogen sulphide (H2S).  This gas, which is of common occurrence in the atmosphere, is oxidized by bacterial growth into sulphuric acid, and this is the basis of part of the soil sulphates.  The deposition of iron phosphates and iron silicates is probably also in a measure aided by bacterial action.  All of these processes are factors in the formation of soil.  Beyond much question the rock disintegration which occurs everywhere in Nature is chiefly the result of physical and chemical changes, but there is reason for believing that the physical and chemical processes are, to a slight extent at least, assisted by bacterial life.

A more important factor of soil fertility is its nitrogen content, without which it is completely barren.  The origin of these nitrogen ingredients has been more or less of a puzzle.  Fertile soil everywhere contains nitrates and other nitrogen compounds, and in certain parts of the world there are large accumulations of these compounds, like the nitrate beds of Chili.  That they have come ultimately from the free atmospheric nitrogen seems certain, and various attempts have been made to explain a method of this nitrogen fixation.  It has been suggested that electrical discharges in the air may form nitric acid, which would readily then unite with soil ingredients to form nitrates.  There is little reason, however, for believing this to be a very important factor But in the soil bacteria we find undoubtedly an efficient agency m this nitrogen fixation.  As already seen, the bacteria are able to seize the free atmospheric nitrogen, converting it into nitrite and nitrates.  We have also learned that they can act in connection with legumes and some other plants, enabling them to fix atmospheric nitrogen and store it m their roots.  By these two means the nitrogen ingredient in the soil is prevented from becoming exhausted by the processes of dissipation constantly going on.  Further, by some such agency must we imagine the original nitrogen soil ingredient to have been derived.  Such an organic agency is the only one yet discerned which appears to have been efficient in furnishing virgin soil with its nitrates, and we must therefore look upon bacteria as essential to the original fertility of the soil.  But in another direction still does the farmer depend directly upon bacteria The most important factor in the fertility of the soil is the part of it called humus.  This humus is very complex, and never alike in different soils It contains nitrogen compounds in abundance, together with sulphates, phosphates, sugar, and many other substances.  It is this which makes the