Attention is drawn to the need for caution in interpreting experimental results.
Frequency curves are discussed, chiefly from the point of view of their bearing on the legitimacy of averaging results.
The method of calculating probable error is described and its meaning explained.
The application of probable error methods to questions of sampling for analysis, to field experiments and to feeding experiments, are illustrated by instances.
The probable error of one animal on a fattening ration is found to be about 14 per cent, of the live-weight increase produced, from which it is calculated that to obtain a precision of 10 per cent, in an ordinary feeding experiment 29 animals must be fed on each ration.
The probable error of field experiments is investigated by two independent methods, and found to be about 5 per cent, of the crop. This figure is shown to be independent of the size of the plot employed, provided this is acre or larger. A table is given showing the number of duplicate plots which must be employed to give any desired precision in the result.
It is also suggested that accurate results may be obtained by employing large numbers of very small plots, even as small as one square yard. This method is useful for nursery work in testing the cropping power of new varieties of cereals where very little seed is available.