The most important investigation of this kind carried out at Rothamsted was organised by Mr. Cutler.[3] A team of six workers was assembled, and for 365 days without a break they counted every day the ciliates, the amœbæ, the flagellates, and the bacteria in a plot of arable ground, distinguishing no less than seventeen different kinds of protozoa. The conclusions arrived at were carefully tested by the Statistical Department. Of the protozoa the flagellates were found to be the most numerous, the amœbæ came next, and the ciliates were by far the fewest. The numbers of each organism varied from day to day in a way that showed conclusively the essentially trophic nature of the protozoan population. The numbers of amœbæ--especially _Dimastigamœba_ and of a species called α--were sharply related to the numbers of bacteria: when the amœbae were numerous the bacteria were few, and vice versa. Detailed examination showed that the amœbæ were probably the cause of the fluctuations in the bacterial numbers, but Mr. Cutler has not yet been able to find why the amœbæ fluctuated; it does not appear that temperature, moisture content, air supply or food supply were determining causes. The flagellates and ciliates also showed large fluctuations, amounting in one case--_Oicomonas_--to a definite periodicity, apparently, however, not related to bacterial numbers, or, so far as can be seen, to external conditions of moisture, temperature and food supply, and showing no agreement with the fluctuations of the amœbæ. However, one cannot be certain that lack of agreement between curves expressing protozoan numbers and physical factors implies absence of causal relationships: the observations (though the best that can yet be made) are admittedly not complete. If we saw only the end of the bough of a tree, and could see no connection with a trunk, we might have much difficulty in finding relationships between its motion and the wind; whatever the direction of the wind it would move backwards and forwards in much the same way, and even when the wind was blowing along the plane of its motion it would just as often move against the wind as with it. Meanwhile evidence was obtained that the twenty-four hour interval adopted by the protozoological staff was too long for bacteria, and accordingly the Bacteriological Department, under Mr. Thornton, refined the method still further. Bacterial counts were made every two hours, day and night, for several periods of sixty or eighty hours without a break. The shape of the curve suggests that two hours is probably close enough, and for the present counts at shorter intervals are not contemplated. But there is at least one maximum and one minimum in the day, although the bacterial day does not apparently correspond with ours, nor can any relationship be traced with the diurnal temperature curve.