The second or dilution method is the one, therefore, that has been most extensively developed. Cunningham obtained concordant results in this way, and his method, modified by L. M. Crump, was as follows: 10 grams of soil were added to 125 c.c. of sterile tap-water and shaken for three minutes. This gives a 1 in 12·5 dilution. From it further dilutions were made until a sufficiently high one was obtained. Petri dishes, containing nutrient agar, were inoculated with 1 c.c. of each of the dilutions and incubated. At intervals covering 28 days the plates were examined and the presence or absence of protozoa on each recorded. In this way the approximate number of organisms per gram of soil could be found. By methods essentially similar to this numerous counts have been made of the bacteria and protozoa in field soil and in partially sterilized soils. They were, however, inconclusive; thus, on the one hand, Goodey and several American observers, found no correlation between the numbers of protozoa and bacteria, while Miss Crump and Cunningham obtained evidence pointing to the reverse conclusion. Such divergence of opinion was probably mainly due to two causes: firstly, that the time elapsing between the successive counts was too long, for it has been shown recently that the number of bacteria and protozoa fluctuate very rapidly; and secondly, the method was not completely satisfactory since only the total numbers of protozoa were considered, no means having been found of differentiating between the cystic and active forms. This was a particularly serious source of error for it is possible for soil to contain large numbers of bacteria and protozoa, of which a high percentage of the latter are in the form of cysts. A count made on such a soil would give results apparently opposed to the theory that protozoa act as depressors of bacteria. This difficulty has, however, been overcome by a further modification of the dilution method, and it is now possible in any soil sample to count both the numbers of cysts and active forms. Also a further advance in technique has made it possible to recognise and enumerate the common species of protozoa, instead of simply grouping them as Ciliates, Flagellates, and Amœbæ, as was done in the past.[7] Briefly the method consists in dividing the soil sample into equal portions (usually 10 grams each) one of which is counted, thus giving the total numbers of protozoa (active + cystic) present. The second portion is treated over-night with 2 per cent. hydrochloric acid, the HCl used being B.P. pure 31·8 per cent. Previous experiments have shown that such acid kills all the active protozoa, leaving viable the cysts. The number of cysts is therefore found by counting this treated sample, and the number obtained subtracted from the total gives the active number.[F] [F] The proof of the accuracy of this method will be found in the following papers:-- (1) Cutler, D. W. (1920), Journ. Agric. Sci., vol. x., 136-143.