FREQUENCY distributions of numbers of eggs, insects, spores, or other biological entities are descriptive of discrete variation, and can never be represented exactly by a normal distribution or by any other distribution of a continuous variate. Nevertheless, when the number of observations and the range of variation is large, such distributions may often be adequately represented by a continuous distribution. Experience has shown that transformation of discrete-counts to a scale of ✓n, log n, or log (n + 1) may considerably assist the representation by a normal distribution.