The method, however, does not indicate whether the account is fairly complete, or whether there are other organisms to be found. We might, of course, trust to empirical hunting for organisms, or to chance discoveries such as led Goodey to find the mysterious Proteomyxan Rhizopods, which cannot yet be cultured with certainty, so that they are rarely found by soil workers. It is possible that there are many such organisms, and it is even conceivable that these unknown forms far outnumber the known. The defect of the present method is that it always leaves us in doubt as to the completeness of the list, and so we may have to devise another. Reverting to Table I., it obviously serves no purpose to add the numbers of all the organisms together. We can add up the weights of living organisms, of their dry matter or nitrogen, so as to form some idea of the proportion of living to non-living organic matter, and this helps us to visualise the different groups and place them according to their respective masses. But a much better basis for comparing the activities of the different groups would be afforded by the respective amounts of energy they transform, if these could be determined. It is proposed to attempt such measurements at Rothamsted. The results when added would give the sum of the energy changes effected by the soil population as we know it: the figure could be compared with the total energy change in the soil itself as determined in a calorimeter. If the two figures are of the same order of magnitude, we shall know that our list is fairly well complete; if they are widely different, search must be made for the missing energy transformers. There are, of course, serious experimental difficulties to be overcome, but we believe the energy relationships will afford the best basis for further work on the soil population. Finally, it is necessary to refer to the physical conditions obtaining in the soil. These make it a much better habitat for organisms than one might expect. At first sight one thinks of the soil as a purely mineral mass. This view is entirely incorrect. Soil contains a considerable amount of plant residues, rich in energy, and of air and water. The usual method of stating the composition of the soil is by weight, but this is misleading to the biologist because the mineral matter has a density some two and a half times that of water and three times that of the organic matter. For biological purposes composition by volume is much more useful, and when stated in this way the figures are very different from those ordinarily given. Table II. gives the results for two Broadbalk arable plots, one unmanured and the other dunged; it includes also a pasture soil.