Ordinary field soil was sterilised by heat at 100° C. for 1 hour on four successive days; it was then divided into equal portions, one of which was inoculated with three known species of bacteria, and the other inoculated with the same number of bacteria plus the cysts of the common soil amœba _Nægleria gruberi_. The numbers of bacteria in each soil were counted daily for the first eight days and then daily from the 15th to the 21st day after the experiment started. The results are given in Table VII. and Fig. 13. TABLE VII. +------------+---------+--------+ | Numbers of | Control | Control| | Days after |(Bacteria|Bacteria| |Inoculation.| alone).|+ Amœbæ.| +------------+---------+--------+ | 0 | 13·0 | 12·2 | | 1 | 48·6 | 35·4 | | 2 | 97·6 | 117·2 | | 3 | 127·0 | 178·4 | | 4 | 154·8 | 154·4 | | 5 | 196·8 | 177·0 | | 6 | 214·4 | 151·8 | | 7 | 193·4 | 75·6 | | 8 | 165·2 | 65·8 | | 15 | 169·2 | 72·8 | | 16 | 174·8 | 30·2 | | 17 | 175·6 | 53·2 | | 18 | 168·4 | 82·8 | | 19 | 160·4 | 43·8 | | 20 | 171·2 | 70·8 | | 21 | 176·2 | 28·2 | +------------+---------+--------+ The numbers of bacteria are given in millions per gram of soil. [Illustration: FIG. 13.--Numbers of bacteria counted daily in soils containing A. Bacteria alone. B. Same Bacteria as in A + Amœbæ. C. Same Bacteria as in A + Flagellates. (From Ann. Appl. Biol., vol. x.)] It will be noted that the numbers of bacteria in each soil rose steadily until a maximum was reached 6-8 days after inoculation. This is in accordance with expectation, since the reproductive rate of bacteria is much greater than that of the amœbæ, which, until their active forms are numerous, will not exert any appreciable influence on the bacterial population. Further, since the protozoa were inoculated as cysts an appreciable time would elapse before excystation took place. The last seven days of the experiment are of particular interest. During this period the amœbæ were known to be active in the soil, and were depressing the bacterial numbers, for in the control (protozoa-free) soil the variation in numbers was within experimental error, while in the other soil the variations were considerable and well outside experimental error. In fact the variations were comparable with those found from day to day in untreated field soils. Finally, the experiment shows that the bacteria in protozoa-free soil are able to maintain high numbers for a longer period than those living in association with protozoa. SEASONAL CHANGES.